Scribd Logo
Upload

Welcome to Scribd!

  • 77 views

    Oxford Textbook of Cancer in Children 7th Ed

    Uploaded by

    Zuleyma Cristina Cimbron Garcia
    This document is the preface to the seventh edition of the Oxford Textbook of Cancer in Children. It was edited by Hubert N. Caron, Andrea Biondi, Tom Boterberg, and François Doz. The preface notes that outcomes for children with cancer have greatly improved since the first edition in 1975 due to international collaboration between clinicians and researchers. However, cures often come with long-term health effects, and new immunotherapies and genomic data need to be translated into effective new treatments. The editors thank all those who contributed to producing this edition.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    Oxford Textbook of Cancer in Children 7th Ed

    Uploaded by

    Zuleyma Cristina Cimbron Garcia
    77 views313 pages
    This document is the preface to the seventh edition of the Oxford Textbook of Cancer in Children. It was edited by Hubert N. Caron, Andrea Biondi, Tom Boterberg, and François Doz. The preface notes that outcomes for children with cancer have greatly improved since the first edition in 1975 due to international collaboration between clinicians and researchers. However, cures often come with long-term health effects, and new immunotherapies and genomic data need to be translated into effective new treatments. The editors thank all those who contributed to producing this edition.

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document is the preface to the seventh edition of the Oxford Textbook of Cancer in Children. It was edited by Hubert N. Caron, Andrea Biondi, Tom Boterberg, and François Doz. The preface notes that outcomes for children with cancer have greatly improved since the first edition in 1975 due to international collaboration between clinicians and researchers. However, cures often come with long-term health effects, and new immunotherapies and genomic data need to be translated into effective new treatments. The editors thank all those who contributed to producing this edition.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    77 views313 pages

    Oxford Textbook of Cancer in Children 7th Ed

    Uploaded by

    Zuleyma Cristina Cimbron Garcia
    This document is the preface to the seventh edition of the Oxford Textbook of Cancer in Children. It was edited by Hubert N. Caron, Andrea Biondi, Tom Boterberg, and François Doz. The preface notes that outcomes for children with cancer have greatly improved since the first edition in 1975 due to international collaboration between clinicians and researchers. However, cures often come with long-term health effects, and new immunotherapies and genomic data need to be translated into effective new treatments. The editors thank all those who contributed to producing this edition.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 313

    Oxford Textbook of

    Cancer in Children
    Oxford Textbook of

    Cancer in
    Children
    SEVENTH EDITION

    EDITED BY
    Hubert N. Caron
    Professor of Paediatric Oncology, Emma Children’s Hospital AMC,
    University of Amsterdam, Amsterdam, The Netherlands
    Group Medical Director for Pediatric Oncology, Hoffmann-La Roche, Basel, Switzerland

    Andrea Biondi
    Professor of Pediatrics, University of Milano-​Bicocca, and Director, Department of Pediatrics,
    Fondazione MBBM/​Hospital San Gerardo, Monza, Italy

    Tom Boterberg
    Associate Professor, Department of Radiation Oncology,
    Ghent University Hospital, Ghent, Belgium

    François Doz
    Professor of Pediatrics, School of Medicine, University Paris Descartes,
    and Pediatric Oncologist, SIREDO Center, Institut Curie, Paris, France

    1
    3
    Great Clarendon Street, Oxford, OX2 6DP,
    United Kingdom
    Oxford University Press is a department of the University of Oxford.
    It furthers the University’s objective of excellence in research, scholarship,
    and education by publishing worldwide. Oxford is a registered trade mark of
    Oxford University Press in the UK and in certain other countries
    © Oxford University Press 2020
    The moral rights of the authors have been asserted
    First Edition published 1975
    Second Edition published 1986
    Third Edition published 1992
    Fourth Edition published 1998
    Fifth Edition published 2005
    Sixth Edition published 2012
    Seventh Edition published 2020
    Impression: 1
    All rights reserved. No part of this publication may be reproduced, stored in
    a retrieval system, or transmitted, in any form or by any means, without the
    prior permission in writing of Oxford University Press, or as expressly permitted
    by law, by licence or under terms agreed with the appropriate reprographics
    rights organization. Enquiries concerning reproduction outside the scope of the
    above should be sent to the Rights Department, Oxford University Press, at the
    address above
    You must not circulate this work in any other form
    and you must impose this same condition on any acquirer
    Published in the United States of America by Oxford University Press
    198 Madison Avenue, New York, NY 10016, United States of America
    British Library Cataloguing in Publication Data
    Data available
    Library of Congress Control Number: 2020935774
    ISBN 978–​0–​19–​879721–​0
    Printed in Great Britain by
    Bell & Bain Ltd., Glasgow
    Oxford University Press makes no representation, express or implied, that the
    drug dosages in this book are correct. Readers must therefore always check
    the product information and clinical procedures with the most up-​to-​date
    published product information and data sheets provided by the manufacturers
    and the most recent codes of conduct and safety regulations. The authors and
    the publishers do not accept responsibility or legal liability for any errors in the
    text or for the misuse or misapplication of material in this work. Except where
    otherwise stated, drug dosages and recommendations are for the non-​pregnant
    adult who is not breast-​feeding
    Links to third party websites are provided by Oxford in good faith and
    for information only. Oxford disclaims any responsibility for the materials
    contained in any third party website referenced in this work.
    Preface

    This is the seventh edition of Cancer in Children, first published in We think that this book will offer readers a clear understanding of
    1975. The outlook for children with cancer has greatly improved what is now possible and what is still needed to take things forward
    since then, mainly due to strong international collaboration be- into the future.
    tween clinicians of various relevant specialties as well as preclinical We would like to thank the contributors to this edition, our col-
    researchers. Still, we are not able to cure all children with cancer, leagues and collaborators, and all others who have helped with the
    and cure comes at a cost, with many survivors of childhood cancer preparation of this book. We would also like to thank the many
    facing significant health problems later in life. However, the last dec- members of the International Society of Paediatric Oncology (SIOP)
    ades have brought an explosion of biological knowledge and gen- who contributed to this work and appreciate SIOP’s continued ef-
    omic data. Translating these into clinically useful tests and therapies forts on behalf of children with cancer in all countries.
    remains a challenge. Also, incorporating the emerging new cancer
    Hubert N. Caron
    immunotherapies in paediatric oncology requires innovation and
    Andrea Biondi
    collaboration.
    Tom Boterberg
    François Doz
    Acknowledgement

    We thank the many SIOP members who contributed to this work and appreciate SIOP’s continued efforts on behalf of children with cancer.
    Contents

    Abbreviations ix 13. Cancer in Adolescents and Young Adults 97


    Contributors xv Lucy J.W. Jones, Emmanouil Saloustros, Andrea Ferrari,
    and Dan Stark

    1. The Epidemiology of Cancer in Children and 14. Late Effects of Therapy and Survivorship Issues 106
    Adolescents 1 Lars Hjorth, Riccardo Haupt, Gisela Michel, Maria Luisa Garrè,
    Charles A. Stiller and Gemma Gatta Leontien C.M. Kremer, and Rod Skinner

    2. Imaging in Paediatric Oncology 12 15. Acute Lymphoblastic Leukaemia 117


    Kieran McHugh and Thierry A.G.M. Huisman Denis Schewe, Mignon Loh, Rob Pieters, and Martin Schrappe

    3. Chemotherapy and Other Anti-​Cancer 16. Acute Myeloid Leukaemia 124


    Drugs: Current Knowledge and New Gertjan Kaspers and Dirk Reinhardt
    Perspectives 21 17. Myelodysplastic Syndrome, Myeloid
    C. Michel Zwaan, Gareth J. Veal, and Lucas Moreno
    Leukaemia of Down Syndrome, and Juvenile
    4. Radiotherapy in Paediatric Oncology 31 Myelomonocytic Leukaemia 134
    Mark Gaze and Tom Boterberg Henrik Hasle and Charlotte M. Niemeyer

    5. Clinical Trials in Childhood Cancer 38 18. Non-​Hodgkin Lymphomas in Children and


    Maria Grazia Valsecchi, Stefania Galimberti, and Pamela Kearns Adolescents 142
    Véronique Minard-​Colin and Catherine Patte
    6. Evidence-​Based Paediatric Oncology 46
    Leontien C.M. Kremer, Erik A.H. Loeffen, and Robert S. Phillips 19. Hodgkin Lymphoma: Modern Management of
    Children and Young Adults 157
    7. Allogeneic Stem-​Cell Transplantation in Children
    Judith Landman-​Parker and Françoise Montravers
    and Adolescents with Malignancies 52
    Adriana Balduzzi, Giovanna Lucchini, Jean-​Hugues Dalle, 20. Childhood Histiocytoses 162
    and Rupert Handgretinger Maurizio Aricò, Cor van den Bos, and Sheila Weitzman

    8. Cancer Immunotherapy in Children 64 21. Symptoms and Emergencies in Children with


    Koichi Hirabayashi, Gianpietro Dotti, and Barbara Savoldo Brain and Spinal Tumours 169
    Michel Zerah, David Walker, and Shaarna Whitton
    9. Supportive Care During Treatment 71
    Marianne van de Wetering and Robert S. Phillips 22. Glial Central Nervous System Tumours of
    Childhood 179
    10. Psychosocial Care 78
    Dannis van Vuurden, Darren Hargrave, Dominik Sturm,
    Martha Grootenhuis, Momcilo Jankovic, Esther van den Bergh,
    and David T.W. Jones
    and Femke Aarsen
    23. Embryonal Tumours 188
    11. Palliative Care for Children with Advanced
    Maura Massimino, Eric Bouffet, and Vijay Ramaswamy
    Cancer 83
    Michelle Koh, Finella Craig, and Joanne Wolfe 24. Other Central Nervous System Tumours of
    Childhood 198
    12. Cancer Treatment in Low-​and Middle-​Income
    Cecile Faure Conter, Didier Frappaz, Kristian W. Pajtler,
    Countries 90
    and Stefan M. Pfister
    Elizabeth Molyneux, Trijn Israels, and Scott C. Howard
    viii Contents

    25. Soft-​Tissue Sarcomas 206 29. Germ-​Cell Tumours 253


    Gianni Bisogno and Hans Merks Gabriele Calaminus and James C. Nicholson

    26. Bone Tumours 219 30. Childhood Liver Tumours 262


    Stefan Bielack, Michael Paulussen, and Lee Helman Giorgio Perilongo and József Zsiros

    27. Wilms and Other Renal Tumours 231 31. Retinoblastoma 270
    Norbert Graf and Christophe Bergeron Guillermo Chantada and Carlos Rodríguez-​Galindo

    28. Neuroblastoma 241


    Angelika Eggert, Garrett M. Brodeur, Index 277
    and Gudrun Schleiermacher
    Abbreviations

    6-​MP 6-​mercaptopurine BM bone marrow


    6-​TG 6-​thioguanine BMD bone mineral density
    A actinomycin D BOIN Bayesian optimal interval
    AAPCOB American Academy of Pediatrics Committee on BSA body surface area
    Bioethics Bu bulsulfan
    ABVD doxorubicin, bleomycin, vinblastine, dacarbazine BV brentuximab vedotin
    aCPP atypical choroid plexus papilloma BW body weight
    ACTH adrenocorticotropic hormone BWS Beckwith–​Wiedemann syndrome
    ADC apparent diffusion coefficient C cyclophosphamide
    ADCC antibody-​dependent cellular cytotoxicity CAR chimeric antigen receptor
    ADV adenovirus CB cord blood
    AER absolute excess risk CBTRUS Central Brain Tumor Registry of the
    AF aggressive fibromatosis United States
    AFP alpha fetoprotein CBV cyclophosphamide, BCNU, and etoposide
    AJCC American Joint Committee on Cancer CCC Cochrane Childhood Cancer
    ALCL anaplastic large-​cell lymphoma CCG Children’s Cancer Group
    ALK anaplastic lymphoma kinase CCNU procarbazine, 6-​thioguanine, dibromodulcitol, and
    ALL acute lymphoblastic leukaemia lomustine
    ALT alternate lengthening of telomeres CCS childhood cancer survivor
    AMKL acute megakaryoblastic leukaemia CCSK clear-​cell sarcoma
    AML acute myeloid leukaemia CCSS Childhood Cancer Survivor Study
    AMPK activation of AMP-​activated protein kinase CDS complement-​dependent cytotoxicity
    APL acute promyelocytic leukaemia CEUS contrast-​enhanced ultrasound
    APO adriamycin (doxorubicin), prednisone, and CEVAIE carboplatin, epirubicin, vincristine, actinomycin D,
    vincristine ifosfamide, and etoposide
    A-​RMS alveolar rhabdomyosarcoma CHC choriocarcinoma
    ASPS alveolar soft-​part sarcoma CHOP cyclophosphamide, doxorubicin, vincristine, and
    ASR age-​standardized rate prednisone
    ATO arsenic trioxide CHL classical variant of Hodgkin lymphoma
    ATRA trans retinoic acid
    all-​ CHS Chédiak–​Higashi syndrome
    ATRT atypical teratoid rhabdoid tumour CI confidence interval
    AYA adolescent and young adult CIR cumulative incidence of relapse
    BCD bleomycin, cyclophosphamide, and actinomycin D CKD chronic kidney disease
    BCH benign cephalic histiocytosis CML chronic myelogenous/​myeloid leukaemia
    BCNU bis-​chlorethylnitrosourea CMML chronic myelomonocytic leukaemia
    BCR B-​cell receptor CMN congenital mesoblastic nephroma
    BCRP breast cancer resistance protein CMV cytomegalovirus
    BEACOPP bleomycin, etoposide, doxorubicin, CNS central nervous system
    cyclophosphamide, vincristine, prednisolone and COG Children’s Oncology Group
    procarbazine COJEC cisplatin, vincristine, carboplatin, etoposide, and
    BEAM BCNU, etoposide, cytarabine, and melphalan cyclophosphamide
    BEP cisplatin, etoposide, and bleomycin COMP cyclophosphamide, vincristine, methotrexate, and
    BFM Berlin–​Frankfurt–​Munster prednisone
    BITE bispecific T-​cell engaging COPAD cyclophosphamide, vincristine, prednisone, and
    BL Burkitt lymphoma doxorubicin
    x Abbreviations

    COPDAC vincristine, cyclophosphamide, dacarbazine, and EICNHL European Intergroup for Childhood Non-​Hodgkin
    prednisone Lymphoma
    COPP vincristine, cyclophosphamide, procarbazine, and EMA European Medical Agency
    prednisone EMT epithelial-​mesenchymal transition
    COSS Cooperative Osteosarcoma Study ENTYAC European Network for Teenagers and Young Adults
    CPC choroid plexus carcinoma with Cancer
    CPG clinical practice guideline EpSSG European paediatric Soft-​tissue sarcoma
    CPP choroid plexus papilloma Study Group
    CPT choroid plexus tumour ES Ewing sarcoma
    CR complete remission ESR erythrocyte sedimentation rate
    CRM continual reassessment method EtD evidence to decision
    CRS cytokine release syndrome ETV endoscopic third ventriculostomy
    CRT cranial/​conformal radiotherapy EU CTD EU Clinical Trials Directive
    CsA cyclosporine A FA fibrillary astrocytoma
    CSC cancer stem cell FAB French–​American–​British
    CSF cerebrospinal fluid FAP familial adenomatous polyposis
    CSI craniospinal irradiation FDA Federal Drug Agency
    CT computed tomography FDG 18-​fluorodeoxyglucose
    ctDNA circulating tumour DNA FHL familial haemophagocytic lymphohistiocytosis
    CTL cytotoxic lymphocyte FISH fluorescent in situ hybridization
    CTR clinical trial regulation FL follicular lymphoma
    CTV clinical target volume FN febrile neutropenia
    CWS German Soft Tissue Sarcoma Cooperative Group FSH follicle-​stimulating hormone
    CXR chest X-​ray G-​CSF granulocyte colony-​stimulating growth factor
    D doxorubicin GA general anaesthetic
    DC dyskeratosis congenital GBM glioblastoma multiforme
    DC dendritic cell GCB germinal centre B-​cell
    DDS Denys–​Drash syndrome GCT germ-​cell tumour
    DEXA dual X-​ray absorptiometry GFR glomerular filtration rate
    DFP dermatofibrosarcoma protuberans GG ganglioglioma
    DFS disease-​free survival GH growth hormone
    DHAP dexamethasone, high-​dose cytarabine, and cisplatin GIST gastrointestinal stromal tumour
    DI diabetes insipidus GM-​CSF granulocyte macrophage colony-​stimulating
    DIA/​DIG desmoplastic infantile astrocytoma/​ganglioglioma growth factor
    DIPG diffuse intrinsic pontine glioma GNB ganglioneuroblastoma
    DLBCL diffuse large B-​cell lymphoma GO gemtuzumab ozogamicin
    DLI donor lymphocyte infusion GP general practitioner
    DLT dose-​limiting toxicity GPOH German Society for Paediatric Oncology and
    DNT dysembryoplastic neuroepithelial tumour Haematology
    DS Down syndrome GRADE grading of recommendations, assessment,
    DSRCT desmoplastic small round-​cell tumour development, and evaluation
    DVH dose volume histogram GSII Griscelli syndrome type II
    DWI diffusion-​weighted imaging GTV gross tumour volume
    DXA dual X-​ray absorptiometry GvH graft-​versus-​host
    E etoposide GvL graft-​versus-​leukaemia
    eIPV enhanced inactivated polio vaccine GvT graft-​versus-​tumour
    E-​RMS embryonal rhabdomyosarcoma GWS genome-​wide association
    EBM evidence-​based medicine HART hyperfractionated accelerated radiotherapy
    EBMT European Group for Blood and Marrow HbF fetal haemoglobin
    Transplantation HBV hepatitis B virus
    EBRT external-​beam radiotherapy HC haemorrhagic cystitis
    EBV Epstein–​Barr virus HCG β-​human chorionic gonadotropin
    EC embryonal carcinoma HCV hepatitis C virus
    ECG electrocardiogram HD high-​dose
    EFS event-​free survival HDAC histone deacetylase
    EGF epidermal growth factor HDCT high-​dose chemotherapy
    EGFR epidermal growth factor receptor HER2 human epidermal growth factor receptor 2
    hFLMPC human fetal liver multipotent progenitor cell
    Abbreviations xi

    HGG high-​grade glioma JCML juvenile chronic myeloid leukaemia


    HHV8 human herpesvirus 8 JEB carboplatin, etoposide, and bleomycin
    HIC high-​income countries JMML juvenile myelomonocytic leukaemia
    HIPEC hyperthermic intraperitoneal chemotherapy JNBSG Japanese Neuroblastoma Study Group
    HIV human immunodeficiency virus JPA juvenile pilocytic astrocytoma
    HL Hodgkin lymphoma JXG juvenile xanthogranuloma
    HLA human leukocyte antigen KIR killer-​cell immunoglobulin-​like receptor
    HLH haemophagocytic lymphohistiocytosis LAE late adverse event
    HPA hypothalamic pituitary axis LAIP leukaemia-​associated immunophenotype
    HPF high-​power field LC large-​cell
    HPV human papilloma virus LCH Langerhans cell histiocytosis
    HR hypophosphataemic rickets LD-​PCR long-​distance polymerase chain reaction
    HR hazard ratio LDH lactate dehydrogenase
    HRS Hodgkin/​Reed Sternberg LFS leukaemia-​free survival
    HSC haematopoietic stem cell LFS Li–​Fraumeni syndrome
    HSCR haematopoietic stem-​cell rescue LGA low-​grade astrocytoma
    HSCT haematopoietic stem-​cell transplantation LGG low-​grade glioma
    HSV herpes simplex virus LGGNT low-​grade glial-​neuronal tumour
    HVA homovanillic acid LH luteinizing hormone
    HvG host-​versus-​graft LL lymphoblastic lymphoma
    I ifosfamide LMIC low-​and middle-​income countries
    IBMFD inherited bone-​marrow failure disorder LN lymph node
    IBMTR International Bone Marrow Transplant Registry LOH loss of heterozygosity
    ICCC-​3 International Classification of Childhood Cancer, LOI loss of imprinting
    Third Edition LRP lung resistance-​related protein
    ICD-​O International Classification of Diseases for Oncology LSC leukaemic stem cell
    ICE ifosfamide, carboplatin, and etoposide LTFU long-​term follow-​up
    ICG Italian Cooperative Group LyP lymphoid papulosis
    ICP intracranial pressure MALT mucosa-​associated tissue lymphoma
    ICU intensive-​care unit MA myeloablative
    IDRF image-​defined risk factor mAb monoclonal antibody
    IEP ifosfamide, etoposide, and prednisolone MAPK mitogen-​activated protein kinase
    IESS Intergroup Ewing Sarcoma MAS macrophage activation syndrome
    IFS infantile fibrosarcoma MBEN medulloblastoma with extensive nodularity
    IGEV gemcitabine, prednisone, ifosfamide, and vinorelbine M-​CSF macrophage-​colony stimulating factor
    IGF insulin-​like growth factor MCV mean corpuscular volume
    IGHG International Guideline Harmonization Group MD matched donor
    IGRT image-​guided radiotherapy MDD minimal disseminated disease
    IMAT intensity-​modulated arc therapy MDP methylene-​diphosphonate
    IMC immunohistochemistry MDR multidrug resistance
    IMRT intensity-​modulated radiotherapy MDS myelodysplastic syndrome
    IMT inflammatory myofibroblastic tumour MDS-​EB myelodysplastic syndrome with excess blasts
    INPC International Neuroblastoma Pathology MDT multidisciplinary team
    Classification MFD matched family donor
    INRG International Neuroblastoma Risk Group MGCT malignant germ-​cell tumour
    INRGSS International Neuroblastoma Risk Group MHC major histocompatibility complex
    Staging System MIBG metaiodobenzylguanidine
    INSS International Neuroblastoma Staging System miRNA microRNA
    IPSS International Prognostic Scoring System ML-​DS myeloid leukaemia of Down syndrome
    IR intervention radiology MMAE monomethyl auristatin E
    IRSS International Retinoblastoma Staging System MMD mismatched donor
    IT intrathecal MMR measles, mumps, rubella
    ITCC Innovative Therapies for Children with Cancer MoAb monoclonal antibody
    ITD internal tandem duplication MPNST malignant peripheral nerve sheath tumour
    ITT insulin tolerance test MPO myeloperoxidase
    IV intravenous MRC Medical Research Council
    IVAD ifosfamide, vincristine, actinomycin D, and MRD minimal residual disease
    doxorubicin MRI magnetic resonance imaging
    xii Abbreviations

    MRP multidrug resistance protein PFS progression-​free survival


    MRT malignant rhabdoid tumour PFT pulmonary function test
    MS metabolic syndrome PGC primordial germ cell
    MS multisystem PHPV persistent hyperplastic primary vitreous
    MSD matched sibling donor PIP paediatric investigational plan
    MSKCC Memorial Sloan-​Kettering Cancer Center PIS patient information sheet
    MSTS Musculoskeletal Tumor Society PLAP placental alkaline phosphatase
    MTD maximum tolerated dose PMBL primary mediastinal large B-​cell lymphoma
    MTPI modified toxicity probability interval PNET primitive neuro-​ectodermal tumour
    mTOR mammalian target of rapamycin pNT peripheral neuroblastic tumour
    MTP-​PE muramyl tripeptide POG Paediatric Oncology Group
    MTX methotrexate PPB peuropulmonary blastoma
    MUD matched unrelated donor (p)PNET (peripheral) primitive neuroectodermal tumour
    MZL marginal zone lymphoma PPR prednisone poor response
    NBL neuroblastoma PR partial remission
    NCI National Cancer Institute PRETEXT pre-​treatment extent
    ND neurodegeneration PTC principal treatment centre
    NDI nephrogenic diabetes insipidus PTCL peripheral T-​cell lymphoma
    NF1 neurofibromatosis PT-​Cy post-​transplant cyclophosphamide
    NGGCT non-​germinomatous germ-​cell tumour PTSD post-​traumatic stress disorder
    NGS next-​generation sequencing PTV planning target volume
    NHL non-​Hodgkin lymphoma PVB cisplatin, vinblastine, and bleomycin
    NICE National Institute of Clinical Excellence PXA pleomorphic xanthoastrocytoma
    NK natural killer RA refractory anaemia
    NLPHL nodular lymphocyte predominant Hodgkin RAEB refractory anaemia with excessive blasts
    lymphoma RAEB-​t refractory anaemia with excessive blasts in
    NMDA N-​methyl-​D-​aspartate transformation
    NMSC non-​melanomatous skin cancer RARS refractory anaemia with ringed sideroblasts
    NMZL nodal marginal-​zone lymphoma RC refractory cytopenia
    NOPHO Nordic Society of Paediatric Haematology and RCC refractory cytopenia of childhood
    Oncology RCC renal-​cell carcinoma
    NOS not otherwise specified RCT randomized controlled clinical trial
    NPM nucleophosmin RDD Rosai–​Dorfman disease
    NRM non-​relapse mortality REAL Revised European–​American Lymphoma
    NRSTS non-​rhabdomyomatous soft-​tissue sarcoma Classification of HL
    NS Noonan syndrome RECIST response evaluation criteria in solid tumours
    NSAID non-​steroidal anti-​inflammatory drug RFS relapse-​free survival
    NSCLC non-​small-​cell lung cancer RIC reduced-​intensity conditioning
    NSE neuron-​specific enolase RMS rhabdomyosarcoma
    NTV neuroendoscopic third ventriculostomy RR relative risk
    OEPA vincristine, etoposide, prednisone, and adriamycin RR relapsed and refractory
    OMS opsoclonus-​myoclonus syndrome RT radiotherapy
    ORR objective response rate RTA renal tubular acidosis
    OS overall survival RTK rhabdoid tumour of the kidney
    OS osteosarcoma RTK receptor tyrosine kinase
    PA pilocytic astrocytoma RT-​PCR reverse transcriptase polymerase chain reaction
    PAS periodic-​acid-​Schiff SC sclerosing cholarigitis
    PB peripheral blood SC stem cell
    PBSC peripheral blood stem cells SCF stemcell factor
    PCP pneumocystis pneumonia SCT stem-​cell transplantation
    PCR polymerase chain reaction SCN severe congenital neutropenia
    PCU paediatric cancer unit SDD selective decontamination of the digestive tract
    PDGF platelet-​derived growth factor SDS Shwachman–​Diamond syndrome
    PDGFR platelet-​derived growth factor receptor SE subependymoma
    PD-​L1 programmed death ligand 1 SEER surveillance, epidemiology, and end results
    PEI ifosfamide, cisplatin, and etoposide SEGA subependymal giant-​cell astrocytoma
    PET positron emission tomography SEN subependymal nodule
    PF posterior fossa SHH sonic hedgehog
    Abbreviations xiii

    SHML sinus histiocytosis with massive lymphadenopathy TRH thyrotropin-​releasing hormone


    sIg suface immunoglobin TRM transplant-​related mortality
    SIOP International Society of Paediatric Oncology TSC tuberous sclerosis complex
    SIOPEL International Childhood Liver Tumor TSH thyroid-​stimulating hormone
    Strategy Group TTL target toxicity level
    SMN secondary malignant neoplasm UCH uncommon histiocytosis
    SMR standardized mortality ratio UNCRC UN Convention of the Rights of the Child
    SNP single nucleotide polymorphism URD unrelated donor
    SOS sinusoidal obstructive syndrome US ultrasound
    SP spine UV ultraviolet
    SS synovial sarcoma V vincristine
    ST supratentorial VAC vincristine, actinomycin, and cyclophosphamide
    STS soft-​tissue sarcoma VDC-​IE vincristine, doxorubicin, cyclophosphamide,
    STS sodium thiosulfate ifosfamide, and etoposide
    STSC Soft Tissue Sarcoma Committee VEGF vascular endothelial growth factor
    SVC superior vena cava VIP vinblastine, ifosfamide, and cisplatin
    TAM transient abnormal myelopoiesis VHR very high risk
    TBI total body irradiation VMA vanillylmandelic acid
    TCR T-​cell receptor VOD veno-​occlusive disease
    TDI total diagnostic interval VP ventriculo-​peritoneal
    TDM therapeutic drug monitoring VTC topotecan, cyclophosphamide, and vincristine
    TdT terminal deoxynucleotidyl transferase VZV varicella zoster virus
    TKI tyrosine kinase inhibitor WAGR Wilms tumour, aniridia, genital deformity,
    TLCT transitional liver cell tumour retardation
    TLP traumatic lumbar puncture WBC white blood cell
    TLS tumour lysis syndrome WCC World Child Cancer
    TMP/​SMZ trimethoprim/​sulfamethoxazole WGS whole-​genome sequencing
    TMZ temozolomide WHO World Health Organization
    TN tenascin WT Wilms tumour
    TNFR tumour necrosis factor receptor WVI whole ventricular irradiation
    TNM tumour-​node-​metastasis XLP X-​linked lymphoproliferative syndrome
    TPMT thiopurine methyltransferase YST yolk sac tumour
    Contributors

    Femke Aarsen Pediatric Clinical Finella Craig Consultant in Paediatric Palliative Henrik Hasle Professor, Pediatrics and Adolescent
    Neuropsychologist, Princess Máxima Center for Medicine, Great Ormond Street Hospital for Medicine, Aarhus University Hospital, Aarhus,
    Pediatric Oncology, Utrecht, The Netherlands Children, London, UK Denmark
    Maurizio Aricò Coordinator of Rare Pediatric Jean-​Hugues Dalle Head of HSCT Pediatric Unit, Riccardo Haupt Epidemiology and Biostatistics
    Diseases Center, Meyer University Hospital, Hematology and Immunology Department, Unit, DOPO Clinic, IRCCS Istituto Giannina
    Florence, Italy Robert-​Debre Hospital, Assistance Publique, Gaslini, Genova, Italy
    Adriana Balduzzi Pediatric Clinic University of Hopitaux de Paris and University of Paris, Lee Helman Attending Physician, Children's
    Milan Bicocca, MBBM Foundation, San Gerardo Paris, France Center for Cancer and Blood Diseases, Children's
    Hospital, Monza, Italy Gianpietro Dotti Professor, Department of Hospital Los Angeles, Los Angeles, USA
    Christophe Bergeron Paediatric Oncologist, Microbiology and Immunology, University of Koichi Hirabayashi Lineberger Comprehensive
    Institute of Hematology and Pediatric Oncology North Carolina, Chapel Hill, USA Cancer Center, University of North Carolina,
    (IHOPE), Centre Léon Bérard, Lyon, France François Doz Professor of Pediatrics, School Chapel Hill, USA
    Stefan Bielack Medical Director, Department of Medicine, University Paris Descartes, and Lars Hjorth Paediatric Oncology and Haematology,
    of Pediatric Oncology, Hematology and Pediatric Oncologist, SIREDO Center, Institut Department of Paediatrics, Skåne University Hospital,
    Immunology, Stuttgart Cancer Center, Stuttgart Curie, Paris, France Clinical Sciences, Lund University, Sweden
    Hospital—​Olgahospital, Stuttgart, Germany Angelika Eggert Director, Department of Pediatrics, Scott C. Howard Professor and Associate Dean for
    Andrea Biondi Professor of Pediatrics, University Division of Oncology and Hematology, Charité Research in the College of Nursing, University
    of Milano-​Bicocca, and Director, Department University Medical Center Berlin, Berlin, of Tennessee College of Health Sciences,
    of Pediatrics, Fondazione MBBM/​Hospital San Germany Memphis, USA
    Gerardo, Monza, Italy Andrea Ferrari Pediatric Oncology Unit, Thierry A.G.M. Huisman Radiologist-​in-​Chief,
    Gianni Bisogno Associate Professor, Department Fondazione IRCCS Istituto Nazionale dei Texas Children's Hospital, Houston, USA
    of Women’s and Children’s Health, University of Tumori, Milan, Italy
    Trijn Israels Pediatric Oncologist, Coordinator
    Padua, Padua, Italy Didier Frappaz Pediatric Oncologist, Institute of Clinical Research Sub-​Saharan Africa,
    Tom Boterberg Associate Professor, Department of Pediatric Hematology and Oncology (IHOPE), Outreach Programme, Princess Máxima
    Radiation Oncology, Ghent University Hospital, Lyon, France Center for Pediatric Oncology, Utrecht, The
    Ghent, Belgium Stefania Galimberti Professor of Medical Netherlands
    Eric Bouffet Garron Family Chair in Childhood Statistics, School of Medicine and Surgery, Momcilo Jankovic Pediatric Hemato-​Oncologist,
    Cancer Research, Director, Paediatric Neuro-​ University of Milano-​Bicocca, Milan, Italy Pediatric Clinic University of Milano-​Bicocca,
    Oncology Program, and Professor of Paediatrics, Maria Luisa Garrè Department of Neuro-​Oncology, Foundation MBBM, Monza, Italy
    Hospital for Sick Children, Toronto, Canada IRCCS Istituto Giannina Gaslini, Genova, Italy Lucy J.W. Jones Medical Oncology Registrar, St
    Garrett M. Brodeur Director of the Cancer Gemma Gatta Evaluative Epidemiology Unit, James’s University Hospital, Leeds, UK
    Predisposition Program, Children's Hospital of Fondazione IRCCS Istituto Nazionale dei David T.W. Jones Group Leader, Pediatric Glioma
    Philadelphia, and Perelman School of Medicine, Tumori, Milan, Italy Research Group, Hopp Children’s Cancer
    University of Pennsylvania, Pennsylvania, USA Mark Gaze Consultant Clinical Oncologist, Center Heidelberg (KiTZ), and German
    Gabriele Calaminus Department of Pediatric University College London Hospitals and Great Cancer Research Center (DKFZ), Heidelberg,
    Hematology and Oncology, Bonn University Ormond Street Hospital for Children, London, UK Germany
    Children’s Hospital, Bonn, Germany Norbert Graf Department of Pediatric Oncology Gertjan Kaspers Director of the Academy, Princess
    Hubert N. Caron Professor of Paediatric Oncology, and Hematology, Saarland University, Homburg, Máxima Center for Pediatric Oncology, Utrecht,
    Emma Children’s Hospital AMC, University of Germany and Professor of Pediatric Oncology, Amsterdam
    Amsterdam, Amsterdam, The Netherlands Martha Grootenhuis Princess Máxima Center for UMC/​Vumc, The Netherlands
    Guillermo Chantada Head of Precision Pediatric Oncology, Utrecht, The Netherlands Pamela Kearns Professor of Clinical Paediatric
    Medicine, JP Garrahan Hospital, and Principal Rupert Handgretinger Chair, Department of Oncology, Institute of Cancer and Genomic
    Researcher, CONICET, Buenos Aires, Argentina; Hematology/​Oncology, Children's University Sciences, University of Birmingham, and
    Associated Member, Hospital Sant Joan de Deu, Hospital, University of Tuebingen, Tuebingen, Honorary Consultant, Birmingham Children’s
    Barcelona, Spain Germany Hospital, Birmingham, UK
    Cecile Faure Conter Pediatric Oncologist, Institute Darren Hargrave Haematology and Oncology Michelle Koh Paediatric Palliative Care Consultant,
    of Pediatric Hematology and Oncology (IHOPe), Department, Great Ormond Street Hospital for University Hospital Southampton NHS
    Lyon, France Children, London, UK Foundation Trust, Southampton, UK
    xvi Contributors

    Leontien C.M. Kremer Department of Pediatrics, Catherine Patte Department of Pediatrics, Institut Dan Stark Associate Professor of Cancer
    Amsterdam UMC, Amsterdam, The Netherlands; Gustave Roussy, Paris, France Medicine and Honorary Consultant in Medical
    Princess Máxima Center for Pediatric Oncology, Michael Paulussen Medical Director, Datteln Oncology, Leeds Institute for Medical Research
    Utrecht, The Netherlands Children’s Hospital, Datteln, and Chair of Pediatrics, and Leeds Teaching Hospitals NHS Trust,
    Judith Landman-​Parker Pediatric Hematology Witten/​Herdecke University, Witten, Germany Leeds, UK
    and Oncology Department, Armand Trousseau Giorgio Perilongo Department of Women's Charles A. Stiller Lead on Childhood Cancer,
    Hospital, Paris, France and Children's Health, University of Padua, National Cancer Registration and Analysis
    Erik A.H. Loeffen Pediatrician-​in-​training, Beatrix Padua, Italy Service, Public Health England, UK
    Children's Hospital, University Medical Center Stefan M. Pfister Hopp-Children’s Cancer Center Dominik Sturm Physician Scientist, Pediatric
    Groningen, Groningen, The Netherlands Heidelberg (KiTZ); Division of Pediatric Glioma Research Group, Hopp Children’s
    Mignon Loh Professor of Clinical Pediatrics, Neurooncology, German Cancer Research Center Cancer Center Heidelberg (KiTZ), Clinic
    Division of Hematology Oncology, (DKFZ); Department of Pediatric Hematology for Pediatric Oncology, Hematology, and
    Helen Diller Family Comprehensive Cancer and Oncology, Heidelberg University Hospital, Immunology, University Hospital Heidelberg,
    Center, University of California San Francisco, Heidelberg, Germany and German Cancer Research Center (DKFZ),
    San Francisco, USA Heidelberg, Germany
    Robert S. Phillips Centre for Reviews and
    Giovanna Lucchini BMT Department, Great Dissemination, University of York, York, UK, Maria Grazia Valsecchi Professor of Medical
    Ormond Street Hospital for Children, London, UK and Leeds Children's Hospital, Leeds Teaching Statistics, School of Medicine and Surgery,
    Hospitals NHS Trust, Leeds, UK University of Milano-​Bicocca, Milan, Italy
    Maura Massimino Pediatric Unit Director,
    Fondazione IRCCS Istituto Nazionale dei Rob Pieters Pediatric Oncologist, Board of Marianne van de Wetering Pediatric Oncologist,
    Tumori, Milan, Italy Directors, and Chief Medical Officer, Princess Princess Máxima Center for Pediatric Oncology,
    Máxima Center for Pediatric Oncology, Utrecht, Utrecht, The Netherlands
    Kieran McHugh Consultant Paediatric Radiologist,
    Great Ormond Street Hospital for Children, The Netherlands Esther van den Bergh Clinical Psychologist,
    London, UK Vijay Ramaswamy Assistant Professor, Division of Princess Máxima Center for Pediatric Oncology,
    Haematology and Oncology, Hospital for Sick Utrecht, The Netherlands
    Hans Merks Pediatric Oncologist, Princess Máxima
    Center for Pediatric Oncology, Utrecht, The Children, Toronto, Canada Cor van den Bos Pediatric Oncologist, Princess
    Netherlands Dirk Reinhardt Director, Pediatrics III and Máxima Center for Pediatric Oncology, Utrecht,
    Department of Pediatric Oncology, Hematology, The Netherlands
    Gisela Michel Department of Health Sciences
    and Medicine, University of Lucerne, Lucerne, Pneumology, Stem Cell Therapy, Cardiology, and Dannis van Vuurden Pediatric Neuro-​Oncologist,
    Switzerland Rheumatology, University Hospital Essen, Essen, VU University Medical Center, Amsterdam, The
    Germany Netherlands
    Véronique Minard-​Colin Department of Pediatric
    and Adolescent Oncology, Gustave Roussy, Carlos Rodríguez-​Galindo Executive Vice-​ Gareth J. Veal Professor of Cancer Pharmacology,
    Université Paris-Saclay, Villejuif, France President, Chair of the Department of Global Newcastle University Centre for Cancer,
    Pediatric Medicine, and Director of St Jude Newcastle upon Tyne, UK
    Elizabeth Molyneux Honorary Professor of
    Global, St Jude Children’s Research Hospital, David Walker Professor of Paediatric Oncology,
    Paediatrics and Child Health, University of Memphis, USA Children’s Brain Tumour Research Centre,
    Malawi College of Medicine, Malawi, and Strategic
    Advisor for East and West Africa, Royal College of Emmanouil Saloustros Department of Oncology, University of Nottingham, Nottingham, UK
    Paediatrics and Child Health, London, UK University Hospital of Larissa, Larissa, Greece Sheila Weitzman Pediatric Oncologist, The
    Françoise Montravers Head of the Nuclear Barbara Savoldo Lineberger Comprehensive Hospital for Sick Children, Toronto, Canada
    Medicine Department, Assistance Publique Cancer Center, University of North Carolina, Shaarna Whitton Children’s Brain Tumour
    Hôpitaux de Paris, Paris, France Chapel Hill, USA Research Centre, University of Nottingham,
    Lucas Moreno Clinical Director, Paediatric Denis Schewe Christian-​Albrechts University Nottingham, UK
    Oncology and Haematology Division, Vall Kiel and University Medical Center Schleswig-​ Joanne Wolfe Director, Pediatric Palliative Care,
    d’Hebron Hospital, Barcelona, Spain Holstein, Kiel, Germany Boston Children's Hospital, and Division
    James C. Nicholson Department of Paediatric Gudrun Schleiermacher SIREDO Care, Innovation Chief, Pediatric Palliative Care Service,
    Oncology, Cambridge University Hospitals NHS and Research for Children, Adolescents and Department of Psychosocial Oncology and
    Foundation Trust, Cambridge, UK Young Adults with Cancer, Pediatric Oncology Palliative Care, Dana-​Farber, Cancer Institute,
    Center; Deputy Director for Translational Boston, USA
    Charlotte M. Niemeyer Professor, Division of
    Research; Team Leader, SiRIC RTOP Michel Zerah Head of the Department of Pediatric
    Pediatric Hematology and Oncology, Department Translational Research in Pediatric Oncology, Neurosurgery, Hȏpital des Enfants Malades,
    of Pediatrics and Adolescent Medicine, Medical INSERM U830 PARIS, France Paris, France
    Center, Faculty of Medicine, University of
    Freiburg, Freiburg, Germany Martin Schrappe Chair of Pediatrics, Christian-​ József Zsiros Pediatric Oncologist, Princess
    Albrechts University Kiel and University Medical Máxima Center for Pediatric Oncology, Utrecht,
    Kristian W. Pajtler Hopp-Children’s Cancer
    Center Schleswig-​Holstein, Kiel, Germany The Netherlands
    Center Heidelberg (KiTZ); Division of
    Pediatric Neurooncology, German Cancer Rod Skinner Department of Paediatric and C. Michel Zwaan Professor of Paediatric Oncology,
    Research Center (DKFZ); Department Adolescent Haematology and Oncology, Great Princess Máxima Center for Pediatric Oncology,
    of Pediatric Hematology and Oncology, North Children’s Hospital, and Newcastle Utrecht, and Erasmus MC, Rotterdam, The
    Heidelberg University Hospital, Heidelberg, University Centre for Cancer, Newcastle Netherlands
    Germany University, Newcastle, UK
    1
    The Epidemiology of Cancer in Children
    and Adolescents
    Charles A. Stiller and Gemma Gatta

    Introduction million children, giving a cumulative risk of 1 in 438 of developing


    cancer during the first 15 years of life.
    Childhood and adolescent cancers account for less than 2% of all Overall, incidence is almost 15% higher in boys than in girls.
    cancers in industrialized countries, but the young age at which they The male excess is greatest for lymphomas. Extracranial germ-​cell
    occur means that they account for a much larger proportion of total tumours, melanomas, and carcinomas of several sites are mark-
    population life years potentially lost to cancer. We consider here the edly more frequent in girls than in boys. Within childhood, total
    classification of childhood and adolescent cancer, incidence and incidence of cancer is highest in the first five years of life. There
    survival rates, late effects, and aetiology. are also marked early age peaks for acute lymphoblastic leukaemia
    (ALL) and for all the distinctive embryonal tumours. In contrast,
    Hodgkin lymphoma and bone sarcomas are virtually never seen
    Classification before the age of two years, and their incidence increases steeply
    throughout childhood. Gonadal germ-​cell tumours are most fre-
    The current standard classification for those under the age of 15 years quent among boys in early childhood, whereas they are rare among
    is the International Classification of Childhood Cancer, Third Edition girls until the postpubertal increase, which begins earlier than
    (ICCC-​3), with groups defined by the codes for morphology and among boys.
    topography in the third edition of the International Classification of Table 1.2 gives numbers of cases and incidence rates for the main
    Diseases for Oncology (ICD-​O). Most of these groups are limited to groups and subgroups of the Birch classification among adolescents
    malignant neoplasms, but benign and unspecified intracranial and aged 15–​19 years in England during 2003–​2012. The total annual
    intraspinal tumours are also included, since it is hard to distinguish incidence was 200 per million. About 23% of all adolescent cancers
    between malignant and non-​malignant tumours and they are re- are lymphomas, predominantly Hodgkin lymphoma. Leukaemias,
    corded by many cancer registries. central nervous system tumours, and carcinomas each account for
    In the most recent version of the standard classification for cancers 14–​16% of all cancers, and germ-​cell tumours for 11%.
    of adolescents and young adults (15–​24 years), diagnostic groups are
    International variations
    defined by morphology and topography codes in the second edition
    of ICD-​O. The grouping of diagnoses is more logical than ICCC-​3 Total incidence of cancer in children and adolescents tends to be
    for cancers affecting adolescents and young adults especially at age higher in industrialized countries and in parts of tropical Africa, and
    20–​24 years, but some comparability with data on childhood cancer lower in developing countries elsewhere. The relative frequencies of
    is thereby sacrificed. different types of cancer also vary considerably between world re-
    gions and between ethnic groups in the same country.
    Leukaemias
    Incidence In the industrialized countries of all continents, leukaemias form the
    largest diagnostic group of childhood cancer, with an ASR of 40–​
    Incidence in England 60 per million, and often account for one third of all malignancies.
    Table 1.1 gives numbers of cases and incidence rates for the main Incidence is lower in less industrialized countries of Asia and Africa,
    groups and principal subgroups of ICCC-​3 among children aged 0–​ and among black children in the United States. ALL is the most com-
    14 years in England during 2003–​2012. The pattern of incidence is monly occurring leukaemia, and variation in the magnitude of the
    typical of that found among mainly white populations of industrial- early childhood peak accounts for much of the international vari-
    ized countries. The total age-​standardized rate (ASR) was 156.5 per ation in total incidence.
    2
    Table 1.1 Childhood cancer in England, 2003–​2012: numbers of registrations, age-​specific and age-​standardized (World Standard
    Population) incidence rates, and sex ratio of age-​standardized rates (not calculated for groups with fewer than 20 registrations)

    Diagnostic group Total registrations Annual rate per million by age group Age-​standardized rate Sex ratio
    per million (M/​F)
    0–​4 years 5–​9 years 10–​14 years
    All cancers 14,113 207.4 116.1 133.4 156.5 1.1
    Leukaemias, myeloproliferative and 4,219 74.1 36.0 26.7 48.1 1.2
    myelodysplastic diseases
    Lymphoid leukaemias 3,277 59.5 29.0 17.8 37.6 1.2
    Acute myeloid leukaemias 698 10.6 5.3 6.7 7.8 1.1
    Chronic myeloproliferative diseases 63 0.2 0.4 1.4 0.6 0.9
    Myelodysplastic syndrome and other 115 2.4 0.8 0.5 1.3 1.3
    myeloproliferative diseases
    Lymphomas and reticuloendothelial neoplasms 1,513 8.8 14.3 25.7 15.5 2.1
    Hodgkin lymphomas 620 1.3 4.4 14.2 6.0 1.7
    Non-​Hodgkin lymphomas including Burkitt 776 5.3 9.0 10.9 8.1 2.6
    lymphoma
    Central nervous system and miscellaneous 3,510 42.7 37.4 33.8 38.4 1.1
    intracranial and intraspinal neoplasms
    Ependymomas and choroid plexus tumours 347 6.4 2.9 2.0 4.0 1.1
    Astrocytomas 1,466 17.8 16.2 13.6 16.1 1.0
    Embryonal tumours 638 9.2 7.0 4.6 7.1 1.4
    Other gliomas 378 3.6 4.6 4.1 4.1 1.1
    Other specified 533 3.7 5.5 8.1 5.5 1.0
    Unspecified 148 2.1 1.2 1.4 1.6 1.1
    Neuroblastoma and other peripheral nervous 836 23.0 3.1 0.9 10.2 1.0
    cell tumours
    Neuroblastoma and ganglioneuroblastoma 830 23.0 3.1 0.8 10.1 1.0
    Retinoblastoma 371 11.4 0.4 0.1 4.6 0.9
    Renal tumours 793 19.4 4.9 1.5 9.5 0.8
    Wilms tumour and other non-​epithelial renal 771 19.2 4.8 1.0 9.3 0.9
    tumours
    Hepatic tumours 176 4.4 0.5 0.8 2.1 1.1
    Hepatoblastoma 144 4.1 0.4 0.2 1.8 1.3
    Malignant bone tumours 605 1.3 5.5 12.7 6.0 1.1
    Osteosarcomas 334 0.3 2.8 7.6 3.2 1.1
    Ewing tumour and related bone sarcomas 225 0.7 2.3 4.3 2.3 1.1
    Soft-​tissue and other extraosseous sarcomas 928 11.8 7.9 10.3 10.1 1.3
    Rhabdomyosarcomas 461 7.2 4.6 3.1 5.2 1.5
    Fibrosarcomas, peripheral nerve sheath 93 1.1 0.5 1.4 1.0 1.3
    tumours, and other fibrous neoplasms
    Other specified soft-​tissue sarcomas 298 2.9 2.1 4.6 3.1 1.1
    Germ-​cell tumours, trophoblastic tumours, 506 7.1 2.0 7.1 5.5 0.7
    and neoplasms of gonads
    Intracranial and intraspinal germ-​cell tumours 164 1.4 1.0 2.9 1.7 1.4
    Malignant extracranial and extragonadal 130 3.7 0.2 0.3 1.6 0.6
    germ-​cell tumours
    Malignant gonadal germ-​cell tumours 191 2.0 0.7 3.4 2.0 0.6
    Other malignant epithelial neoplasms and 544 1.5 3.2 12.7 5.3 0.7
    malignant melanomas
    Thyroid carcinomas 106 0.3 0.6 2.6 1.0 0.4
    Malignant melanomas 97 0.4 0.7 2.0 1.0 0.6
    Skin carcinomas 111 0.3 0.8 2.5 1.1 0.8
    Other and unspecified carcinomas 187 0.3 1.0 4.7 1.8 0.8
    Other and unspecified malignant neoplasms 112 1.7 0.8 1.1 1.2 1.0

    Source: data from National Cancer Registration and Analysis Service (NCRAS), Public Health England.
    Incidence 3

    Table 1.2 Adolescent cancer in England, 2003–​2012: numbers of registrations, age-​specific incidence rates, and sex ratio of rates
    (not calculated for groups with fewer than 20 registrations)

    Diagnostic group Total registrations Annual rate per million, 15–​19 years Sex ratio (M/​F)
    All cancers 6,596 200.4 1.2
    Leukaemias 894 27.2 1.5
    Acute lymphoblastic leukaemia 420 12.8 2.1
    Acute myeloid leukaemia 258 7.8 1.1
    Chronic myeloid leukaemia 76 2.3 1.1
    Myelodysplastic syndrome and other 93 2.8 0.9
    myeloproliferative diseases
    Lymphomas 1,533 46.6 1.2
    Non-​Hodgkin lymphoma 451 13.7 1.9
    Hodgkin lymphoma 1,082 32.9 1.0
    Central nervous system and other 1,065 32.4 1.3
    intracranial and intraspinal neoplasms
    Astrocytoma 343 10.4 1.6
    Other glioma 163 5.0 1.4
    Ependymoma 54 1.6 2.1
    Embryonal tumours 71 2.2 1.6
    Other specified 362 11.0 0.9
    Unspecified 72 2.2 1.2
    Bone tumours 484 14.7 1.6
    Osteosarcoma 245 7.4 2.0
    Ewing sarcoma 178 5.4 1.3
    Soft-​tissue sarcomas 307 9.0 1.1
    Rhabdomyosarcoma 97 2.9 1.9
    Other specified soft-​tissue sarcomas* 136 4.0 0.8
    Germ-​cell tumours 720 21.2 4.7
    Gonadal 638 18.8 4.6
    Non-​gonadal 82 2.4 5.6
    Melanoma 467 13.8 0.6
    Carcinomas 971 28.6 0.5
    Thyroid 292 8.6 0.3
    Other head and neck carcinomas 138 4.1 0.8
    Genito-​urinary tract 183 5.4 0.1
    Gastro-​intestinal tract 268 7.9 1.0
    Miscellaneous, specified 77 2.3 0.6
    Unspecified 78 2.3 1.0

    * Except fibrosarcoma
    Source: data from National Cancer Registration and Analysis Service (NCRAS), Public Health England.

    Incidence of leukaemia among adolescents is generally between for 15–​25% of all childhood cancers in Africa and the Middle East,
    20 and 35 per million, and international variations are again mostly 10–​13% in Europe, the Americas, and Oceania, and 8–​10% in Eastern
    attributable to ALL. Incidence of acute myeloid leukaemia (AML), and South-​Eastern Asia. The most common subtype in countries with
    the second most frequent subtype in children and adolescents, varies the highest incidence is Burkitt lymphoma. Hodgkin lymphoma is rare
    relatively little. among children in tropical Africa and Eastern and South-​Eastern Asia.
    In developing countries, the incidence of Hodgkin lymphoma some-
    Lymphomas times peaks at 5–​9 years, whereas in industrialized countries it increases
    The ASR of lymphomas in childhood is as high as 30–​60 per million more steeply at the onset of adolescence.
    in parts of tropical Africa, 20–​25 per million in Mediterranean re- In adolescence, incidence of lymphomas is highest in Southern
    gions, and 12–​16 per million in most other populations. They account Europe and lowest in India, South-​Eastern Asia, and among Native
    4 CHAPTER 1 The Epidemiology of Cancer

    Americans. Hodgkin lymphoma is more common than non-​ million in children and adolescents. Elsewhere, incidence of soft-​
    Hodgkin lymphoma (NHL) in most world regions, but NHL pre- tissue sarcomas in childhood rarely exceeds 11 per million and
    dominates in sub-​Saharan Africa and in Eastern and South-​Eastern rhabdomyosarcoma is the most frequent type. Children of Asian
    Asia. Hodgkin lymphoma is around twice as frequent among boys origin have a lower incidence of soft-​tissue sarcoma.
    under 15 years of age as it is among girls, but at age 15–​19 years in-
    cidence among females is usually at least as high as among males. Germ-​cell tumours
    NHL is nearly everywhere more frequent among males than females Childhood germ-​cell tumours are most frequent in East Asia. In
    throughout childhood and adolescence. adolescence, testicular cancer has an incidence of 25–​30 per mil-
    lion in Europe, North America, and Oceania, and less than 10 per
    Brain and spinal tumours million in most other populations. Ovarian germ-​cell tumours have
    Brain and spinal tumours are the second most frequent child- an incidence of 5–​10 per million at age 15–​19 years and are one of
    hood cancer in industrialized countries, accounting for 20–​25% of the very few types of cancer whose peak incidence occurs in this
    all cases. They also occur in substantial numbers in adolescence. age group. Intracranial and intraspinal germ-​cell tumours have an
    Incidence in the United States, in both children and adolescents, incidence of five per million among adolescents in Eastern Asia and
    is considerably higher in non-​Hispanic whites than in other ethnic below three per million elsewhere.
    groups. Recorded incidence is consistently lower in developing
    countries, where under-​diagnosis is likely to be a factor.
    Trends in incidence
    Neuroblastoma
    Neuroblastoma occurs most frequently in infancy. Its incidence The best-​documented trend concerns ALL, the most common childhood
    is higher in industrialized countries than in developing countries, cancer in all developed countries. The early childhood peak emerged in
    probably in large part because of a higher rate of diagnosis of asymp- mortality data in England and Wales in the 1930s, and was well established
    tomatic tumours. Higher incidence among infants has been found in among white children in the USA by the early 1940s. Incidence continued
    several industrialized countries compared with the UK, possibly be- to rise, particularly in early childhood, in some western countries until at
    cause of increased detection of otherwise silent tumours at routine least the end of the twentieth century. Small increases have also been ob-
    health checks. Mass screening proved to be ineffective, resulting in served in western populations for many other childhood cancers, but part
    especially high incidence in infants, but without lowering the inci- of these increases may be attributed to changes in diagnostic and registra-
    dence of advanced stage neuroblastoma at older ages. tion practice rather than in underlying risk.

    Retinoblastoma
    Retinoblastoma has relatively constant incidence worldwide, al- Survival rates in Europe
    though it tends to account for a higher proportion of all childhood
    cancers in developing countries. Survival of European children and adolescents diagnosed with
    cancer between 1978 and 2007 has been analysed in the EUROCARE
    Wilms tumour project. The population-​based data reported here are from the
    Incidence of Wilms tumour tends to depend on ethnicity rather EUROCARE-​5 database and include 29 European countries covered
    than geographical area. Incidence is highest in black populations, by 74 cancer registries. About 78,000 European children and adoles-
    with an ASR of 9–​12 per million, and lowest in those of Asian origin, cents diagnosed with cancer between 2000 and 2007, and followed
    with an ASR often under five per million. up to the end of 2008, were considered in this analysis.

    Liver tumours Survival of children and adolescents by cancer type


    Hepatoblastoma has an ASR of under three per million children In children, five-​year survival was high for lymphoid leukaemia
    in all world regions. The incidence of hepatocellular carcinoma in (86%), Burkitt lymphoma and other NHL (83% for both), and
    young people is highest in regions where there is a high proportion Hodgkin lymphoma (95%), but only 61% for AML. Five-​ year
    of chronic carriers of hepatitis B in the population, namely East Asia, survival for CNS cancers was modest (57%), with limited variation
    Melanesia, and sub-​Saharan Africa. in relation to morphology. Five-​year survival for retinoblastoma was
    very high at 98%. Survival was also good (89%) for nephroblastoma
    Bone tumours and chondrosarcoma, but lower for neuroblastoma (71%), osteosar-
    Incidence of malignant bone tumours increases with age, reaching coma, Ewing sarcoma, and rhabdomyosarcoma (all 67%) (Table 1.3).
    a peak in late adolescence. The two most common subtypes in In adolescents, five-​year survival was above 90% for Hodgkin
    children and adolescents are osteosarcoma and Ewing sarcoma. lymphoma, gonadal germ-​cell cancer, and skin melanoma, inter-
    Osteosarcoma has fairly constant incidence worldwide, whereas mediate for NHL (78%), and relatively low for lymphoid leukaemia
    Ewing sarcoma is rare in East Asian and black populations. (62%) and AML (52%).
    Poorer survival in adolescents than in children was statistically
    Soft-​tissue sarcomas significant for ALL, AML, NHL, astrocytomas, osteosarcoma, Ewing
    In parts of East and Central Africa with high levels of human her- sarcoma, rhabdomyosarcoma, fibrosarcomas, and gonadal germ-​
    pesvirus 8 (HHV8) infection, Kaposi sarcoma is the most frequent cell tumours. For Hodgkin lymphoma and skin melanoma, how-
    cancer in young people, with incidence sometimes above 40 per ever, survival figures were similar between children and adolescents,
    Survival rates in Europe 5

    Table 1.3 Comparison of five-​year relative survival (RS) with standard error (SE) by major diagnostic category for the
    European pool of children and adolescents with cancer diagnosed in the period 2000–​2007

    Diagnostic group (ICCC-​3) 0–​14 years 15–​19 years p value


    15–​19 vs 0–​14
    N RS (SE) N RS (SE)
    Lymphoid leukaemias 15,089 85.8 (0.4) 1378 62.2 (1.6) <0.0001
    Acute myeloid leukaemias 2,944 60.5 (1.0) 704 52.2 (2.2) 0.0007
    Hodgkin lymphomas 2,995 95.1 (0.5) 3,541 94.3 (0.5) 0.21
    Non-​Hodgkin lymphomas (except Burkitt lymphoma) 2,407 83.0 (0.9) 1,217 78.0 (1.4) 0.0023
    CNS and miscellaneous intracranial and intraspinal 8,856 57.2 (0.6) 1,464 61.8 (1.5) 0.0090
    neoplasms
    Astrocytoma 2,584 61.9 (1.1) 604 50.8 (2.5) 0.0003
    Intracranial and intraspinal embryonal neoplasms 2,951 56.3 (1.1) 233 67.0 (3.8) 0.0074
    Medulloblastomas 2,156 63.2 (1.3) 158 72.8 (4.5) 0.041
    Retinoblastoma 1,627 96% (0.5) -​ -​ -​
    Nephroblastoma and other non-​epithelial renal tumours 3,554 89% (0.6) -​ -​ -​
    Neuroblastoma and ganglioneuroblastoma 4,588 71% (0.8) -​ -​ -​
    Osteosarcomas 1,430 66·8 (1·5) 765 60.3 (2.2) 0.012
    Chondrosarcomas 66 89.4 (3.4) 140 80.7 (3.8) 0.092
    Ewing tumour and related sarcomas of bone 1,322 66.6 (1.5) 448 51.1 (2.7) <0.0001
    Soft-​tissue and other extraosseous sarcomas 3,871 69.3 (0.9) 1,185 63.0 (1.6) 0.0007
    (excluding Kaposi)
    Rhabdomyosarcomas 2,124 66.6 (1.3) 280 39.6 (3.4) <0.0001
    Fibrosarcomas 209 83.8 (3.6) 47 72.8 (9.8) 0.29
    Germ-​cell tumours, trophoblastic tumours, and 1,805 91.5 (0.8) 2,238 92.2 (0.7) 0.55
    neoplasms of gonads
    Intracranial and intraspinal germ-​cell tumours 466 85.9 (2.2) 158 79.5 (4.2) 0.18
    Malignant gonadal germ-​cell tumours 821 96.8 (0.8) 2,011 93.6 (0.7) 0.0015
    Malignant skin melanoma 394 92.2 (1.6) 1,248 91.2 (1.2) 0.64

    Adapted with permission from Trama, A. et al. ‘Survival of European adolescents and young adults diagnosed with cancer in 2000–​07: population-​
    based data from EUROCARE-​5’, The Lancet Oncology, Volume 17, Issue 7, pp. 896–​906. Copyright © 2016 Elsevier Ltd. DOI: https://​doi.org/​10.1016/​
    S1470-​2045(16)00162-​5

    and adolescents had significantly higher survival for intraspinal radiation therapy, and increased risk of death from toxic effects
    and intracranial embryonal tumours including medulloblastoma. mostly due to infection.
    Poorer survival in adolescents has been attributed to various factors
    including no or few clinical trials conducted in adolescents, dearth European geographical variation
    of specific treatment guidelines, differences in cancer biology, vari- Because countries varied widely in how they attributed malignancy
    ations in the pharmacokinetics of chemotherapeutic agents, and of CNS tumours and how data for non-​malignant CNS tumours
    delays in diagnosis and treatment. Adolescents with cancer are, in were collected by population-​based registries, five-​year survival by
    many ways, neglected by both paediatric and adult oncologists, yet country is presented for all childhood cancers, except CNS tumours.
    effective disease management necessitates a multi-​professional ap- For most countries, five-​year survival ranged between 80% and
    proach incorporating expertise from both specialties. 85%, but all eastern European countries (except Poland) had lower
    survival (60–​77%) (Figure 1.1). Many factors could explain the poor
    Survival in children by age survival in Eastern Europe. These countries have a lower gross do-
    Infants younger than one year old had the lowest five-​year survival mestic product (GDP) than countries in other European regions,
    for several cancers, particularly ALL, AML, NHL, ependymoma, and this is likely to have several effects. Drugs might be unavailable,
    embryonal CNS cancers, and all CNS cancers. By contrast, children or supplies might run out so that treatment has to be halted before
    younger than age one year with neuroblastoma had good survival completion. Specialized centres for childhood cancers might not
    (91%), whereas less than 60% of older children survived for five have enough beds or resources to treat all patients in their catchment
    years. Five-​year survival was worst for children aged 10–​14 years areas, and poorer countries might be unable to afford the multidis-
    with astrocytomas, nephroblastoma, and Ewing sarcoma (Table 1.4). ciplinary teams typical of paediatric oncology units in richer coun-
    Several factors may explain poor outcome in infants, including tries. Treatment could be delayed because diagnosis is delayed, or
    increased risk of metastatic disease at presentation, characteristic patients might die of drug toxicity because modern equipment is not
    genetic mutation, increased rate of subtotal resection, not receiving available to accurately measure serum drug concentrations.
    6 CHAPTER 1 The Epidemiology of Cancer

    Table 1.4 Country-​weighted five-​year survival (%) by ICCC-​3 diagnostic category and age, with 95% confidence intervals (95% CI),
    for children in Europe with cancer diagnosed in the period 2000–​2007

    Diagnostic group N Age <1 year Age 1–​4 years Age 5–​9 years Age 10–​14 years
    5-​yr 95% CI 5-​yr 95% CI 5-​yr 95% CI 5-​yr 95% CI
    survival survival survival survival
    All cancers combined 57,956 77.9 76.4–​79.4 79.3 78.4–​80.0 77.6 76.6–​78.5 76.6 75.7–​77.5
    I(a) Lymphoid leukaemias 15,860 61.8 56.0–​67.1 90.6 89.5–​91.7 88.1 86.8–​89.3 77.7 75.5–​79.7
    I(b) Acute myeloid leukaemias 3,094 53.5 47.0–​59.6 65.1 61.4–​68.5 67.9 63.5–​71.9 59.5 55.1–​63.5
    II(a) Hodgkin lymphomas 3,142 -​ -​ 95.5 91.1–​97.8 94.1 89.9–​96.6 95.8 94.5–​96.8
    II(b) Non-​Hodgkin lymphomas (except 2,544 63.3 49.8–​74.0 78.1 72.7–​82.5 87.0 83.8–​89.6 85.4 82.7–​87.8
    Burkitt lymphoma)
    II(c) Burkitt lymphoma 1,443 -​ -​ 89.3 85.3–​92.3 91.1 88.8–​93.0 87.2 84.0–​89.8
    III CNS and miscellaneous intracranial 9,277 48.3 43.8–​52.7 57.4 55.0–​59.8 57.0 54.6–​59.3 60.3 57.8–​62.7
    and intraspinal neoplasms
    III(a) Ependymomas and choroid plexus 1,233 42.4 30.0–​54.3 55.3 50.6–​59.8 74.7 66.5–​81.1 76.2 68.6–​82.2
    tumours
    III(b) Astrocytomas 2,714 64.1 56.3–​70·9 79.4 75.6–​82.7 55.6 51.1–​60.0 49.3 45.0–​53.5
    III(c) Intracranial and intraspinal 3,119 33.3 26.6–​40.2 46.5 42.3–​50.5 67.3 63.3–​71.0 67.3 62.2–​71.9
    embryonal tumours
    IV(a) Neuroblastoma and 4,588 91.1 89.6–​92.5 58.7 54.8–​62.5 52.1 45.8–​58.0 55.7 45.5–​64.6
    ganglioneuroblastoma
    V Retinoblastoma 1,627 98.3 96.6–​99.1 94.6 89.9–​97.2 96.4 79.9–​99.4 -​ -​
    VI(a) Nephroblastoma and other 3,554 84.3 80.2–​87.6 91.4 89.7–​92.9 88.2 85.4–​90.4 76.7 66.0–​84.5
    nonepithelial renal tumours
    VIII(a) Osteosarcomas 1,500 -​ -​ 59.8 47.6–​70.1 72.1 66.7–​76.8 68.5 64.9–​71.9
    VIII(c) Ewing tumour and related 1,397 70.6 58.8–​79.6 73.7 64.6–​80.7 76.3 71.5–​80.4 62.1 57.1–​66.6
    sarcomas of bone
    IX(a) Rhabdomyosarcomas 2,197 61.0 49.7–​70.5 71.2 66.2–​75.5 70.6 65.5–​75.2 62.3 56.6–​67.5
    -​No cases

    Adapted with permission from Gatta, G. et al. ‘Childhood cancer survival in Europe 1999–​2007: results of EUROCARE-​5—​a population-​based study’, The Lancet Oncology, Volume 15,
    Issue 1, pp. 35–​47. Copyright © 2013 Elsevier Ltd. DOI: https://​doi.org/​10.1016/​S1470-​2045(13)70548-​5

    Survival trends 1999–​2007 oncologists, developing national policies for managing adolescents
    Figure 1.2 shows trends in five-​year survival among children and ado- with cancer, and setting up specific treatment units—​have been im-
    lescents for all malignancies together, the major haematological tu- plemented in several European countries and worldwide. For can-
    mours, CNS tumours, and skin melanoma. Five-​year survival during cers in adolescents that also affect children, it has been suggested
    the period 1999–​2007 improved by 2% and 3% per year in children that adolescents should be treated in an integrated paediatric–​adult
    and adolescents, respectively. Despite faster improvement for adoles- multidisciplinary setting. This change should increase the likelihood
    cents during the study period, survival remained significantly worse of adolescents being included in clinical trials, and improve family
    than in children for several cancers (Table 1.3). Significant progress and social support. The present situation could be improved by
    was reported in ALL and NHL for both children and adolescents, intensifying international collaboration with twinning programmes
    and for AML and soft-​tissue sarcomas (not shown in Figure 1.2) and by motivating politicians to create and develop sustainable in-
    in children only. No significant progress was observed for CNS tu- frastructure, recognizing the specific needs of children with cancer.
    mours or melanoma. Among children, survival also did not improve The rarity of most childhood cancers creates particular difficulties
    for Hodgkin lymphoma, neuroblastoma, nephroblastoma, Ewing for small countries, since good results for these cancers are usually
    sarcoma, and osteosarcoma. The absence of substantial survival in- obtained only when treatment is centralized in high-​volume centres
    creases for several cancers during 1999–​2007 may confirm the beliefs of excellence. Collaborative programmes in which all patients are
    of some paediatric oncologists and researchers that optimization of sent to a single centre could be a solution, through the implemen-
    present treatments has reached its limits. New research approaches tation and extension of the European directive on cross-​border
    are therefore needed to improve survival further, especially for high-​ healthcare, particularly to small countries that lack the resources
    risk groups. New approaches to trial design and greater international and infrastructure to treat these diseases and are unlikely to develop
    collaboration are needed to recruit enough participants for studies of such infrastructure because of the very small numbers of children
    small and homogeneous subgroups of patients. with cancer.
    To improve outcomes for adolescents, various initiatives—​ A strength of these studies was that outcomes were evaluated
    including promoting collaboration between paediatric and adult in a large population-​based database of children and adolescents,
    Aetiology of cancer in children and adolescents 7

    archived by European cancer registries. In the future, data provided second primary cancer is about six times that in the general popu-
    by these registries will be vital to assess whether changes in manage- lation, but it varies according to type of first cancer and treatment
    ment policies have had the desired effect of improving survival in given. The risk is highest in those who received radiotherapy and
    European children and adolescents who develop cancer. among survivors of heritable retinoblastoma.
    Survivors of childhood and adolescent cancer are at increased risk
    of a range of chronic diseases, which are discussed in other chapters.
    Follow-​up

    Continuing improvement in survival from childhood cancer has Aetiology of cancer in children and adolescents
    led to a great increase in long-​term survivors, including substantial
    numbers of adults. For childhood cancer overall, excess mortality The short latency of childhood cancers suggests that events affecting
    continues beyond 25 years from diagnosis, principally due to second in utero life and/​or exposures affecting parents may be aetiologically
    primary cancers and circulatory diseases. The risk of developing a important. Intensive investigation since the mid twentieth century

    (a) (b)
    100% 100%

    90% 90%

    80% 80%

    70% 70%

    60% 60%

    50% 50%

    40% 40%
    1999–2001 2002–2004 2005–2007 1999–2001 2002–2004 2005–2007

    (c) (d)
    100% 100%

    90% 90%

    80% 80%

    70% 70%

    60% 60%

    50% 50%

    40% 40%
    1999–2001 2002–2004 2005–2007 1999–2001 2002–2004 2005–2007
    0–14 years 15–19 years

    Figure 1.1 Five-​year relative survival for children and adolescents with cancer in Europe in 1999–​2001, 2002–​2004, and 2005–​2007: (a) all tumours;
    (b) acute lymphoid leukaemia; (c) acute myeloid leukaemia; (d) non-​Hodgkin lymphoma; (e) central nervous system; (f) malignant melanoma;
    (g) Hodgkin lymphoma; (h) malignant gonadal germ-​cell tumours; (i) Ewing tumour and related sarcomas of bone.
    * indicates a statistically significant survival trend
    Adapted with permission from Trama, A. et al. ‘Survival of European adolescents and young adults diagnosed with cancer in 2000–​07: population-​based data from
    EUROCARE-​5’, The Lancet Oncology, Volume 17, Issue 7, pp. 896–​906. Copyright © 2016 Elsevier Ltd. DOI: https://​doi.org/​10.1016/​S1470-​2045(16)00162-​5
    8 CHAPTER 1 The Epidemiology of Cancer

    (e) (f)
    100% 100%

    90%
    95%
    80%

    70% 90%

    60%
    85%
    50%

    40%
    80%
    1999–2001 2002–2004 2005–2007
    1999–2001 2002–2004 2005–2007

    0–14 years 15–19 years

    (g) (h)
    100% 100%

    95% 95%

    90% 90%

    85% 85%

    80% 80%
    1999–2001 2002–2004 2005–2007 1999–2001 2002–2004 2005–2007

    (i)
    100%

    90%

    80%
    0–14 years 20–39 years
    70%
    15–19 years 15–39 years
    60%

    50%

    40%
    1999–2001 2002–2004 2005–2007

    Figure 1.1 Continued


    Environmental factors 9

    Denmark
    Finland
    Iceland
    Norway
    Sweden
    Ireland
    UK - England & Wales
    UK - Northern Ireland
    UK -Scotland
    Austria
    Belgium
    France
    Germany
    Switzerland
    The Netherlands
    Croatia
    Italy
    Malta
    Portugal
    Slovenia
    Spain
    Bulgaria
    Estonia
    Hungary
    Latvia
    Lithuania
    Poland
    Slovakia
    Europe

    0 10 20 30 40 50 60 70 80 90 100
    Figure 1.2 Five-​year survival (%) for all cancers combined (CNS tumours excluded) diagnosed in 2000–​2007, in European children (age 0–​14 years)
    by country. Data adjusted for age, sex, case mix, and period of diagnosis.
    Adapted with permission from Gatta, G. et al. ‘Childhood cancer survival in Europe 1999–​2007: results of EUROCARE-​5—​a population-​based study’. The Lancet Oncology,
    Volume 15, Issue 1, pp. 35-​47. Copyright © 2013 Elsevier Ltd. DOI:https://​doi.org/​10.1016/​S1470-​2045(13)70548-​5

    has indicated the complexity of mechanisms underlying the occur- Naturally occurring background radiation is largely unavoidable
    rence of cancer in children, but the causes of the great majority of and 15% of childhood leukaemia in Britain may be attributable to
    cases remain unknown. Nevertheless, the knowledge accumulated is this source of exposure.
    sufficient to inform public health plans aiming to reduce cancer in-
    cidence. Furthermore, risk factors for cancer in children and adoles- Non-​ionizing radiation
    cents are also involved in other diseases affecting these age groups. Ultraviolet radiation from sunlight can cause melanoma and other
    skin cancers. There is no evidence of raised risk of childhood leu-
    kaemia with exposure to power-​frequency electromagnetic fields
    Environmental factors at the levels experienced by more than 95% of children in western
    countries. A doubling of risk has been consistently found at the very
    Ionizing radiation highest exposure levels but the explanation is unknown, and only a
    Exposure to ionizing radiation is one of the few factors generally tiny fraction of childhood leukaemia cases could be attributable to
    accepted to increase the risk that a child will develop cancer. Recent electromagnetic fields.
    studies do not themselves show an association with antenatal diag-
    nostic X-​ray exposure but are consistent with earlier, more powerful Infections
    studies that do. In the past, antenatal radiography probably caused at Worldwide, the most frequent virus-​associated cancers among chil-
    most about 5% of childhood cancers, even when it was more widely dren and adolescents are Burkitt lymphoma, Hodgkin lymphoma
    used and doses were higher than they are now. The much higher and nasopharyngeal carcinoma (Epstein–​ Barr virus), liver car-
    doses of radiation given during radiotherapy are also carcinogenic cinoma (hepatitis B), Kaposi sarcoma (HHV8 and human immuno-
    but can only account for a tiny proportion of all childhood cancers. deficiency virus), and cervical carcinoma (human papilloma virus).
    10 CHAPTER 1 The Epidemiology of Cancer

    For Hodgkin lymphoma, 60% of cases in children of 0–​14 years range of syndromes and congenital anomalies, but the pattern of in-
    in developed countries and 80% of cases in children aged 0–​14 in heritance is more complicated and the proportion of clearly heredi-
    developing countries can be attributed to Epstein-​Barr virus, com- tary cases much smaller.
    pared to only 30% of cases in people aged 15–​44. Up to 10% of children and adolescents with cancer may have
    It is likely that infection is sometimes involved in the aetiology germline mutations in cancer predisposition genes. Other herit-
    of childhood leukaemia, especially ALL. Two main hypotheses are able syndromes carrying an increased risk of childhood cancer in-
    supported by substantial epidemiological evidence. About 1% of clude neurofibromatosis types 1 and 2, tuberous sclerosis complex,
    children are born with a premalignant clone with the potential to Fanconi anaemia, ataxia telangiectasia, Li-​Fraumeni syndrome, con-
    develop into precursor B-​cell ALL. Under the ‘delayed infection’ hy- stitutional mismatch repair deficiency, xeroderma pigmentosum,
    pothesis, if they are protected from exposure to infection in infancy and Bloom syndrome. There are also well-​documented associations
    then heightened immune response to infection later in childhood with chromosomal abnormalities; the strongest association is with
    acts as a promoter of leukaemia. Evidence supporting this hypoth- Down syndrome, which occurs in a small percentage of cases of
    esis includes consistent associations of breastfeeding and attendance childhood leukaemia.
    at daycare in early childhood with a reduced risk of leukaemia, par- If one child in a family has malignant disease, then, in the absence
    ticularly common ALL. of any further information about the existence of genetic disease in
    Under the ‘population mixing’ hypothesis, childhood leukaemia that family, and excluding twins and retinoblastoma, that child’s sib-
    is an unusual response to one or more specific but as yet unidentified lings have approximately double the risk of the general population
    infectious agents, with higher risk occurring when susceptible and for developing childhood cancer. However, this excess could well be
    infected people are brought together. This is supported by numerous entirely accounted for by familial syndromes. In up to 25% of cases
    studies in which childhood leukaemia rates have increased when where a child who is a monozygotic twin has leukaemia, their co-​
    people from different areas come together. twin will also develop leukaemia. These cases are usually due to in
    utero transfer of leukaemic or pre-​leukaemic cells rather than being
    Chemical exposures genetically determined.
    Many epidemiological studies have found associations of childhood There is no evidence either of a raised risk of cancer or genetic dis-
    cancer with various chemical exposures, notably pesticides and at- ease among the offspring of survivors who do not themselves have
    mospheric pollution from traffic. While more evidence is needed to a hereditary cancer syndrome or that the parents of children with
    establish causality, it would be justified, from the public health point cancer have an increased risk of cancer in the absence of known her-
    of view, to reduce or eliminate exposure to suspected chemicals in editary syndromes.
    children and in young fertile adults, especially since many of these A rapidly growing body of research has found significantly raised
    chemicals are also involved in other diseases affecting children and or reduced risk of ALL, neuroblastoma, and other childhood cancers
    adolescents. associated with polymorphic variants of certain genes, sometimes
    interacting with exogenous factors, though many of these findings
    Other factors associated with pregnancy and birth remain to be replicated.
    There is considerable evidence that high birthweight is associ-
    ated with a raised risk of leukaemia and several other cancers in FURTHER READING
    childhood, possibly as a result of increased antenatal growth rate.
    Barr RD, et al. (2006) Classification schemes for tumours diagnosed in
    Children who are twins have a risk of cancer that is about 80% that
    adolescents and young adults. Cancer 106, 1425–​30.
    in singletons. Bearer CF, et al. (eds.) (2016) Opportunities for cancer prevention
    Several studies have found increased risk of specific cancers during early life. Pediatrics 138 Suppl 1, S1–​S100.
    among children born following assisted reproductive technology, Colt JS, Blair A (1998) Parental occupational exposures and risk of
    but with little consistency with respect to cancer types, numbers childhood cancer. Environ Health Perspect 106, 909–​25.
    of cases of specific types of cancer were usually low, and the as- Draper GJ, et al. (1992) Patterns of risk of hereditary retinoblastoma
    sociations could well have been due to confounding with other and applications to genetic counselling. Br J Cancer 66, 211–​19.
    factors. Fidler MM et al. (2016) Long-​ term cause-​ specific mortality
    among 34,489 five-​year survivors of childhood cancer in Great
    Britain: population-​based cohort study. BMJ 354, i4351.
    Gatta G et al. (2014) Childhood cancer survival in Europe 1999–​
    Genetic epidemiology 2007: results of EUROCARE-​5—​a population-​based study. Lancet
    Oncol 15, 35–​47.
    The most obvious example of a genetically determined cancer is ret- Infante-​Rivard C (2008) Chemical risk factors and childhood leu-
    inoblastoma; about 40% of cases have the heritable form of this dis- kaemia: a review of recent studies. Radiat Prot Dosimetry 132,
    ease. The pattern of inheritance is autosomal dominant, but in fact 220–​27.
    the gene RB1 is the first example of a tumour suppressor gene; about Johnson KJ et al. (2014) Childhood brain tumor epidemiology: a
    90% of individuals who inherit the mutated form of this gene from Brain Tumor Epidemiology Consortium review. Cancer Epidemiol
    a parent subsequently suffer a mutation of the wild-​type (normal) Biomarkers Prev 23, 2716–​36.
    allele, leading to loss of heterozygosity and the development of ret- Kaatsch P, et al. (2006) Time trends of cancer incidence in European
    inoblastoma. These individuals are also at increased risk of a variety children (1978–​ 1997): report from the Automated Childhood
    of other cancers. Wilms tumour is associated with an especially wide Cancer Information System project. Eur J Cancer 42, 1961–​71.
    Genetic epidemiology 11

    Linabery AM, Ross JA (2008) Childhood and adolescent cancer Steliarova-​Foucher E et al. (2017) International incidence of childhood
    survival in the US by race and ethnicity for the diagnostic period cancer, 2001–​2010: a population-​based study. Lancet Oncol 18(6),
    1975–​1999. Cancer 113, 2575–​96. 719–​31.
    Lindor NM, et al. (2008) Concise handbook of familial cancer suscepti- Stiller C (ed.) (2007) Childhood Cancer in Britain. Incidence, Survival,
    bility syndromes, second edition. J Natl Cancer Inst Monogr 38, 1–​93. Mortality. Oxford: Oxford University Press.
    Little J (1999) Epidemiology of Childhood Cancer. Lyon: IARC. Stiller CA (2005) Constitutional chromosomal abnormalities and
    Olsen JH, et al. (1995) Cancer in the parents of children with cancer. N childhood cancer. Ital J Pediatr 31, 347–​53.
    Engl J Med 333, 1594–​9. Trama A et al. (2016) Survival of European adolescents and young
    Reulen RC et al. (2011) Long-​term risks of subsequent primary neo- adults diagnosed with cancer in 2000–​07: population-​based data
    plasms among survivors of childhood cancer. JAMA 305, 2311–​19. from EUROCARE-​5. Lancet Oncol 17, 896–​906.
    Sankila R et al. (1998) Risk of cancer among offspring of childhood-​ Wiemels J (2012) Perspectives on the causes of childhood leukemia.
    cancer survivors. N Engl J Med 338, 1339–​44. Chem Biol Interact 196, 59–​67.
    Siegel DA, et al. (2014) Cancer incidence rates and trends among chil- Winther JF et al. (2001) Cancer in siblings of children with cancer in
    dren and adolescents in the United States, 2001–​2009. Pediatrics the Nordic countries: a population-​based cohort study. Lancet 358,
    134, e945–​55. 711–​17.
    Steliarova-​ Foucher E, et al. (2005) International Classification of Zhang J et al. (2015) Germline mutations in predisposition genes in
    Childhood Cancer, third edition. Cancer 103, 1457–​67. childhood cancer. N Engl J Med 373, 2336–​46.
    2
    Imaging in Paediatric Oncology
    Kieran McHugh and Thierry A.G.M. Huisman

    Introduction Another important role for the plain CXR, notably the lateral film, is
    assessing for airway compression from an anterior mediastinal mass.
    Some general principles apply with regard to the radiological approach to Follow-​up during treatment and later surveillance for metastatic relapse
    all mass lesions, despite the diverse range of pathologies seen in different is also performed with serial CXRs. The CXR is also useful in detecting
    body parts. Suspected abdominal and pelvic masses must be examined secondary infection or other complications of treatment, such as rib
    by ultrasound (US) first, and this may be all that is necessary to evaluate osteochondromas secondary to irradiation, during follow-​up.
    benign entities. Superficial lesions or palpable lumps should also be
    Ultrasound
    assessed by US initially. Lesion vascularity can be easily evaluated by
    Doppler interrogation, increased vascularity being an important finding Ultrasound (US) entails no radiation burden and so can be done re-
    with proliferating haemangiomas in infancy, for example, lesions that peatedly without harm. Soft-​tissue masses can be easily evaluated to
    could otherwise mimic more sinister pathology. Further cross-​sectional assess their cystic or solid nature, and presence of calcification. US is a
    imaging, largely to determine lesion extent, will be guided by the US dynamic real-​time examination. With large abdominal masses, inva-
    findings, tumour location, and local availability of magnetic resonance sion into other organs can be difficult to interpret on CT or MRI, but
    imaging (MRI). The reliable interpretation of imaging requires a radiolo- may be easily evaluated with US. Movement of the liver separate from
    gist with sufficient experience and expertise in paediatric oncology, par- a renal mass is a reliable sign of non-​invasion of the liver by the mass.
    ticularly with the more advanced technologies such as X-​ray computed Similarly, movement of the mass separately from the psoas muscle
    tomography (CT), MRI, and positron emission tomography (PET). means the mass is not adherent to the psoas. With additional Doppler
    assessments, the degree of vascularity of a mass lesion can be iden-
    tified. US is the best technique for detecting and excluding tumour
    thrombus extension into the inferior vena cava from a renal or adrenal
    Imaging modalities
    tumour. US is the most sensitive technique in the detection of early
    fungal infiltration of the liver or spleen in the immunocompromized,
    Plain radiography
    best evaluated with linear-​array high-​resolution transducers.
    Plain films play a very limited role in the diagnostic work-​up of new Most percutaneous imaging-​guided biopsies are done simply with
    paediatric mass lesions. The major exception to this is with bone tu- US guidance. Contrast-​enhanced ultrasound (CEUS), with sonovue
    mours, both benign processes and malignant lesions. In this setting, (Bracco), is not licenced for use in children at present in Europe, but
    the plain X-​ray probably has a bigger role in suggesting the proper is being increasingly used within paediatrics. Its major use in the on-
    diagnosis than other techniques. Bone lysis with ill-​defined margins cology setting is in differentiating benign from malignant liver lesions.
    is highly suggestive of an aggressive process such as osteomyelitis or
    tumour, which may be indistinguishable on imaging, but often have Computed tomography
    very different clinical presentations. Sclerotic well-​defined margins The size and extent of large masses are easily assessed with CT, but
    in an osseous mass suggest a benign process. The dilemma of benign with that comes the added radiation burden of CT. The newer multi-​
    versus malignant for bone tumours is usually relatively straightfor- detector CT scanners are extremely fast at examining any body
    ward at plain radiography, and the role of MRI in this setting is largely region, avoiding the need for sedation or anaesthesia in many circum-
    to define the extent of the lesion, and to evaluate for a soft-​tissue mass stances. CT is the optimal method for assessing the lung parenchyma.
    which is likely to be underestimated on plain radiographs. Calcification in a mass, often not apparent on MRI, is easily detected
    Although the chest X-​ray (CXR) is performed routinely and plays an with CT. For chest and abdominal scanning in children, non-​contrast
    important role in staging for metastatic disease at diagnosis, particular enhanced CT studies are usually of little help due to children’s lack
    reliance is placed on chest CT to exclude pulmonary metastases in virtu- of mediastinal or retroperitoneal fat. Thus, for paediatric chest and
    ally all extra-​cranial malignancies, with the exception of neuroblastoma. abdominal CT, single-​phase post contrast-​enhanced images are all
    Posterior rib erosion and rib separation with an obvious chest mass are that is generally required to evaluate a mass lesion, including liver
    virtually pathognomic of a neurogenic tumour such as neuroblastoma. tumours (although the latter are better evaluated with MRI). Bone
    Tumours outside the central nervous system 13

    detail is very well demonstrated on CT such that, for example with signal on T2W MRI, again with variable contrast enhancement after
    skull base tumours, an MRI to assess soft-​tissue extent and a CT to gadolinium administration. Calcification typically has no inherent
    assess erosion of the skull base or facial bones may be necessary. signal at MRI scanning such that calcification in mass lesions is missed
    or underestimated at MRI. Some densely fibrotic lesions may be of low
    Magnetic resonance imaging signal (hypointense) on both T1W and T2W MR images. Some of the
    MRI is the ideal imaging modality because of its better characteriza- most important issues to consider when imaging tumours beyond the
    tion of tissues, multi-​planar capabilities, and lack of ionizing radi- central nervous system (CNS) will be addressed in this section.
    ation. Tumour extent is best assessed with MRI, although for large
    masses CT can be as accurate. As MRI is sensitive to oedema (seen Renal tumours
    as hyperintense or bright signal on T2-​weighted sequences), it is gen- Ultrasound (US) should be utilized first to assess the origin of a po-
    erally impossible to differentiate tumour margins from oedema. In tential renal tumour, tumour thrombus in the renal vein and/​or in-
    addition, at the end of treatment or after surgery, MRI often cannot ferior vena cava (IVC), the contralateral kidney, and liver. Dynamic
    distinguish residual fibrosis or benign residual tissue from residual real-​time ultrasound is a very useful method for assessing tumour
    tumour. Diffusion-​weighted imaging with apparent diffusion coeffi- invasion of the liver or other organs, but tends to be under-​utilized
    cient (ADC) maps has an established role now in the assessment of in this context. Abdominal CT should be performed only after intra-
    tumours all over the body. Highly cellular tumours with restricted venous contrast medium enhancement. Ipsilateral and contralateral
    diffusion are generally malignant lesions and appear dark on ADC small lesions or foci of nephroblastomatosis may be seen with Wilms
    images. Tumour response after chemotherapy, with or without a tumour (nephroblastoma). A chest CT is required to detect or exclude
    change in tumour size, can be evident with a change in the diffusion pulmonary metastases. The diameter of pulmonary nodules should be
    pattern on follow-​up. MRI studies generally take a minimum of 30 measured on lung window settings, and these same settings should be
    minutes, which means that most children less than six years of age used at follow-​up. Ideally, MRI is preferred over CT in evaluating a pri-
    need general anaesthesia or sedation for the MRI examination. mary renal tumour due to its superior contrast resolution and lack of
    radiation. Tumour extent is equally assessed by CT or MRI, but foci of
    Nuclear medicine
    nephroblastomatosis are much better visualized with MRI (Figure 2.1).
    Wilms tumours and hepatoblastoma seldom metastasize to bone All renal tumours in childhood have similar radiological appearances
    and so do not merit routine bone scanning with technetium99-​MDP and age at presentation is the most important factor in the differen-
    (methylene-​diphosphonate). Suspected neuroblastoma metastases are tial diagnosis. Clear cell sarcoma, because of its propensity for skel-
    best assessed with iodine123-​MIBG (metaiodobenzylguanidine) scans. etal metastatic spread, merits routine radionuclide bone scanning for
    If the primary tumour is not avid for MIBG, then a PET-​CT scan with staging at diagnosis. Renal cell carcinoma, which usually manifests in
    18fluorodeoxyglucose (FDG) is recommended to evaluate for skeletal adolescence, demonstrates calcification in approximately 25% of cases.
    and other metastases in this setting. Most sarcomas and other solid tu-
    mours historically required routine bone scans for staging at diagnosis; Neuroblastoma
    however, following paediatric sarcoma trials in both Europe and North Encasement of the aorta and upper abdominal arteries is the hall-
    America, PET-​CT is now recommended for staging at diagnosis and mark of the majority of abdominal neuroblastomas (Figure 2.2),
    also increasingly at follow-​up to assess for tumour response.

    Interventional radiology
    Paediatric interventional radiology (IR) has developed remark-
    ably in the past 20 years. All children’s oncology centres should
    have access to IR services that provide common procedures such
    as tumour biopsy, line placement, angiography, gastrostomy inser-
    tion, and abscess drainage. More complicated procedures such as
    preoperative tumour embolization to reduce haemorrhage at sur-
    gery, radiofrequency ablation, and chemo-​embolization techniques
    should also ideally be available in the bigger centres.

    Tumours outside the central nervous system

    The differential diagnosis of masses in childhood is largely influenced


    by the age of a child and the organ of origin or location of the lesion.
    Calcification in a neuroblastoma or fat plus calcium in a teratoma are
    useful in suggesting a diagnosis, but the imaging characteristics of the
    majority of tumours are in general relatively non-​specific. Most tu-
    mours are, for example, hypovacular on contrast-​enhanced CT with
    areas of low attenuation and heterogeneity. Homogeneous imaging ap-
    pearances of a paediatric mass suggest a benign cause, such as haem- Figure 2.1 Nephroblastoma (Wilms tumour). Wilms tumours in the
    upper poles of both kidneys are seen as hyperintense heterogeneous
    angioma or fibromatosis in infancy. Tumours are typically of low or round masses on coronal T2W MRI. A third possible tumour is seen in
    intermediate signal on T1-​weighted (T1W) MRI sequences and high the right lower pole as a smaller cystic lesion (arrow).
    14 CHAPTER 2 Imaging in Paediatric Oncology

    (a)

    (b)

    (c)

    Figure 2.2 Neuroblastoma. (a) Axial T2W image showing a mass lesion in the right upper quadrant encasing the aorta and right renal artery (arrow).
    Encasement of more than 50% of the circumference of the aorta and tumour abutting the renal vessels are so-​called imaging-​defined risk factors.
    (b) Axial ADC map shows the mass lesion to be dark on this sequence, indicating restricted diffusion, in keeping with a highly cellular malignant lesion.
    (c) Planar images from a MIBG scan in another patient show uptake in the primary abdominal tumour and widespread skeletal metastases, stage M
    disease.
    Tumours outside the central nervous system 15

    in contrast to renal tumours which tend to displace the aorta Liver tumours
    and IVC. Confinement to the suprarenal area and adrenal gland Between 60% and 70% of all primary liver tumours in children
    is more often seen in infants. MS (formerly 4S) disease causes are malignant, with hepatoblastoma and hepatocellular car-
    hepatomegaly and diffuse hepatic infiltration by tumour. The in- cinoma constituting the vast majority of these malignancies.
    filtration may be subtle however, and only appreciated with a Haemangioendothelioma/​haemangioma lesions and mesenchymal
    higher-​resolution linear-​array US transducer. MRI and CT are hamartoma account for nearly all the non-​malignant tumours in
    probably equivalent when assessing the primary tumour, but MRI children. The pre-​ treatment extent (PRETEXT) staging system
    is superior at evaluating for intraspinal disease and allows easier used by the International Childhood Liver Tumor Strategy Group
    depiction of marrow metastases. Some surgeons prefer contrast-​ (SIOPEL) for hepatoblastoma has gained widespread acceptance.
    enhanced CT prior to surgery as CT tends to give a more reliable PRETEXT is used to describe tumour extent before any therapy,
    surgical roadmap of the vasculature, and more clearly demon- thus allowing comparison between studies conducted by different
    strates the extent of tumour calcification after chemotherapy. collaborative groups. PRETEXT staging is based on Couinaud’s
    MIBG scanning is crucial for staging at diagnosis and for assessing system of liver segmentation. The liver segments are grouped into
    response in patients with metastatic disease. four sections: segments 2 and 3 (left lateral section), segments 4a and
    Rhabdomyosarcoma 4b (left medial section), segments 5 and 8 (right anterior section),
    and segments 6 and 7 (right posterior section). Involvement of only
    Skull base tumours often need CT to assess the bony erosion and one section by a hepatoblastoma mass is denoted as a PRETEXT
    MR to evaluate the soft-​tissue extent. MRI is also superior in I tumour; a tumour involving two, three, or all four sections is classi-
    demonstrating tumour involvement of the neural foramina and fied as PRETEXT II, III, or IV respectively. An increasing PRETEXT
    intracranial spread. Orbital tumours rarely have lymphatic loco-​ number gives, very roughly, an estimate of the difficulty of the
    regional spread. US evaluation of the cervical chain of nodes is planned surgical procedure.
    probably underutilized for head and neck tumours. Cervical lymph US is particularly useful is assessing portal or hepatic venous,
    nodes metastases are often difficult to identify on all modalities or biliary ductal, invasion by tumour. MRI is superior to CT in
    and a low threshold for biopsy is advisable. MRI is the modality of evaluating liver segmental involvement by tumour (Figure 2.4).
    choice for assessing genito-​urinary primaries, after initial assess- Surveillance imaging after completion of therapy has recently been
    ment with US (Figure 2.3). US of the liver should be performed in questioned as most relapses also result in a rise in serum alpha
    all patients to detect or exclude metastases. US of the retroperitoneal fetoprotein (AFP) levels. Hepatocellular carcinoma occurs in chil-
    area should be performed in all patients with pelvic, paratesticular, dren older than five years, often against a background of pre-​existing
    or lower limb primaries. Limb primary tumours (leg or arm) should cirrhosis.
    also have US of the inguinal region or axilla, and ideally the pop-
    liteal and epitrochlear areas too, to detect or exclude regional Germ-​cell tumours
    lymphadenopathy. Midline germ-​cell rests are thought to arise in arrested migration
    from the hind gut yolk sac to the genital ridge. Germ-​cell tumours

    Figure 2.3 Rhabdomyosarcoma. Sagittal T2W midline pelvis image Figure 2.4 Hepatoblastoma. Right-​lobe liver tumour (arrow) on a
    shows a cystic mass lesion with septations in the vagina distorting the fat-​suppressed T1-​weighted MRI after gadolinium enhancement. Poor
    cervix and uterus. The fluid-​filled bladder is seen anteriorly (arrow). Gas in enhancement in the central tumour was due to necrosis. The middle and
    the rectum returns dark signal. left hepatic veins were free of tumour which was classified as PRETEXT 2.
    16 CHAPTER 2 Imaging in Paediatric Oncology

    (GCTs) are thus commonly found in the pineal gland, neck, medi- sarcomas are rare in childhood. The Ewing’s family of tumours,
    astinum, retroperitoneum, and sacrococcygeal area. All GCTs show including primitive neuroectodermal tumours, tend to occur in chil-
    mixed cystic and solid areas, often with foci of fatty attenuation and/​ dren less than ten years of age in either a flat bone (pelvis, scapula)
    or calcification. A mass lesion that contains both fatty tissues and cal- or in the diaphysis of a long bone. Plain films are more reliable than
    cification is virtually always a GCT. Fat, and calcification in particular, MRI in assessing the aggressive nature of a lytic bone lesion. MRI
    are easily evaluated with CT, but fat in a mass lesion is easily identi- is best at depicting tumour extent, the associated soft-​tissue mass,
    fied on MRI too; and CT or MRI is equally good at defining tumour involvement of the neurovascular bundle, and spread of tumour
    extent. Rupture of a mediastinal teratoma may result in a pleural or across the growth plate into the epiphysis. Some centres report dy-
    pericardial effusion. Alternatively, pericardial invasion by tumour namic contrast enhanced MRI is useful at predicting percentage tu-
    may cause a pericardial effusion. Radiology cannot differentiate mour necrosis after initial chemotherapy, but the technique is not
    a benign tumour from a malignant GCT. Of the malignant GCTs, widely used.
    dysgerminoma and yolk sac tumours are the most common, with ele-
    vated alpha fetoprotein (AFP) and human chorionic gonadotropin
    (HCG) levels. The latter tumour markers are so reliable that post-​ Tumours within the central nervous system
    operative routine radiological surveillance is generally not required.
    The frequently used phrase ‘children are not small adults’ is pertinent
    Lymphomas
    for paediatric radiology in general and paediatric neuroradiology
    In the abdomen, a non-​Hodgkin lymphoma (NHL) may present as a in particular. Paediatric CNS tumours have multiple characteristic
    mass or intussusception, usually identifiable by US. NHL in the medi- features that differ significantly from their adult counterparts (see
    astinum may cause airway compression such that CT in the supine Chapters 22 and 24).
    position may be dangerous. Clinical airway assessment and a lateral Tumours of the paediatric central nervous system are typically
    chest X-​ray to evaluate the trachea are warranted prior to a routine diagnosed by emergent CT, followed by dedicated MRI. Correlating
    staging CT. NHL in Waldeyer’s ring manifests as tonsillar enlargement imaging findings (e.g. signal characteristics, primary tumour lo-
    and a nasopharyngeal mass. That mass tends to be associated with cation, and contrast enhancement, perfusion, and diffusion char-
    little if any adjacent osseous destruction, unlike rhabdomyosacoma acteristics) with the age and gender of the child allows to narrow
    in the nasopharynx which typically causes much skull base erosion. down the differential diagnosis. Treatment options and outcome
    At initial presentation, Hodgkin lymphoma merits high-​ depend on multiple factors including histological subtype of the tu-
    resolution US to evaluate for splenic infiltration and a chest CT is mour. Until recently, most tumours were primarily classified based
    necessary to evaluate the lung parenchyma. MRI of the neck and ab- upon their histological features. In the past decade, correct tumour
    domen is as sensitive as CT for lymphadenopathy, and would reduce classification and, more importantly, subclassification, has progres-
    the radiation burden in young patients. PET-​CT is now firmly estab- sively come to rely on molecular and genetic transcriptional data
    lished for staging of Hodgkin disease at diagnosis and for assessing analysis. Histologically similar tumours may have different mo-
    response to therapy at follow-​up. lecular characteristics linked to different outcomes. This advanced
    approach allows for a more reliable estimation of prognosis and
    Non-​rhabdomyosarcomatous soft-​tissue tumours survival rate, and consequently supports a tumour-​specific, tailored
    Approximately half of paediatric soft-​ tissue tumours are non-​ treatment. These advances have significantly influenced posterior
    rhabdomyomatous soft-​tissue sarcomas (NRSTSs). They are a het- fossa medulloblastoma treatment options, preventing under-​and
    erogeneous group of tumours, all of mesenchymal origin, that share overtreatment.
    some biological characteristics but differ histologically. The most Each age group has typical tumours, and certain tumours are more
    common NRSTSs are synovial cell sarcomas (17–​42%), followed frequent in boys than in girls, especially in the posterior fossa. The
    by fibrosarcomas (13–​15%), malignant fibrous histiocytomas (12–​ most frequent posterior fossa tumour is the pilocytic astrocytoma
    13%), and malignant peripheral nerve sheath tumours (MPNSTs) (PA), followed by the medulloblastoma, brainstem glioma, and
    (10%). The frequency of most of these tumours is age-​dependent. ependymoma. These four tumours may be considered to be the ‘big
    Fibrosarcomas occur more commonly in children under one year four’. If this ‘frequency list’ is combined with the age/​gender of the
    of age, with synovial sarcomas and MPNSTs being more common in patient, as well as the primary tumour location, the tumour type can
    children aged over ten years. Many MPNSTs occur against a back- frequently be predicted. In the supratentorial brain, the variety of
    ground of neurofibromatosis type 1. All these lesions generally have tumours is significantly greater. Hemispheric astrocytomas/​gliomas
    non-​specific radiological findings of a soft-​tissue mass. Like most are the most frequent tumours, followed by craniopharyngeomas
    paediatric tumours, they are hypovascular at Doppler US evalu- and hypothalamic/​optic pathway gliomas.
    ation, and more usually solid than cystic in appearance. The lesions The distribution of supra-​versus infratentorial tumours also
    are isointense (similar signal intensity) to muscle on T1W MRI, varies with age. In the first two years of life, supratentorial tumours
    hyperintense on T2W, and show variable enhancement after gado- are more frequent than infratentorial tumours; between two and
    linium administration. Locoregional adenopathy should be sought ten years, infratentorial tumours outweigh supratentorial tumours,
    or excluded in all cases. while in children aged ten years and older, supra-​and infratentorial
    tumours are equally frequent.
    Bone tumours Functional prognosis in children may be better because of the
    Osteosarcomas mainly occur around the knee (in the metaphyses of functional plasticity at a younger age. On the other hand, treatment
    the distal femur or proximal tibia) in adolescence. Radiation-​induced options may be limited because the brain is more vulnerable for the
    Tumours within the central nervous system 17

    potential deleterious effects of the various treatment options. For and subarachnoid space, at risk for leptomeningeal cerebrospinal
    example, radiotherapy is frequently contraindicated in very young fluid (CSF) seeding. Consequently, metastatic lesions should be
    patients. In addition, radiotherapy may induce secondary neoplasm searched for along the entire neuroaxis, including the most distal tip
    later in life, although it still plays an important role in the treatment of the dural sack. Unfortunately, grade III and IV astrocytomas have
    of CNS tumours. a poor prognosis.

    Infratentorial tumours Medulloblastomas


    As mentioned earlier, the ‘big four’ posterior fossa tumours in chil- Medulloblastomas are embryonal tumours of the posterior fossa
    dren are the cerebellar astrocytomas, medulloblastomas, brainstem that were, until recently, classified based upon their predominant
    gliomas, and ependymomas. Of course many less frequent tumours histological features into four major types—​classic; desmoplastic/​
    may be found, including atypical teratoid/​rhabdoid tumours (AT/​ nodular type; medulloblastoma with extensive nodularity (MBEN);
    RT), hemangioblastomas, cerebellar gangliocytomas. It would go large-​cell anaplastic (LCA) medulloblastoma. Prognosis and late-​
    beyond the scope of this chapter to discuss all these less frequent term outcome showed a significant heterogeneity between and
    tumours. within the subtypes. As mentioned earlier, the recent merging of
    the histological and molecular/​genetic transcriptional data has al-
    Cerebellar astrocytomas lowed a more prognostic-​relevant subclassification. Based upon the
    Cerebellar astrocytomas can be subdivided in the frequent pilocytic molecular findings, medulloblastomas are now grouped as sonic
    astrocytoma (PA) and the less common fibrillary astrocytoma. PA is hedgehog (SHH) activated, wingless (WNT) activated, group 3, and
    a unique form of astrocytoma which is rather benign and has a good group 4. These tumours have very different behaviours, primary lo-
    prognosis if the tumour is recognized early and completely resected. cations, and prognosis. Treatment options should consequently be
    PAs may be seen associated with neurofibromatosis (NF1). The so-​ tailored to maximize outcome while minimizing treatment-​related
    called ‘epicentre’ of the tumour is frequently within the cerebellar collateral brain damage. WNT tumours have an excellent prognosis
    vermis, followed in frequency by the cerebellar hemisphere. Most and arise from the lower rhombic lip in the dorsolateral brainstem in
    children present with symptoms related to an obstructive hydro- close proximity to cerebellopontine angle/​foramen of Luschka. SHH
    cephalus. PAs are usually well demarcated, have a solid, strongly medulloblastomas originate from the upper rhombic lip and are typ-
    contrast-​enhancing tumour component, and a cystic component ically seen outside of the fourth ventricle within the cerebellar hemi-
    which shows no, or minimal, contrast enhancement (Figure 2.5). spheres, but may occasionally originate from the cerebellar vermis.
    The cyst may be significantly larger than the solid tumour nodule. Prognosis is good/​intermediate depending on the age of the child.
    Calcifications are seen in approximately 20% of cases; vasogenic oe- Group 3 and 4 include both classic and LCA histology subtypes, with
    dema is usually mild or minimal. The tumour typically compresses a variable prognosis, being poor for group 3 and intermediate for
    or displaces adjacent structures, including the brainstem and fourth group 4.
    ventricle, resulting in a supratentorial hydrocephalus. The imaging Most medulloblastomas are hyperdense on CT due to the high
    features may resemble hemangioblastomas. Hemangioblastomas cellularity. On MR examination, the tumour is solid, predomin-
    are however significantly rarer, occurring usually as part of the von antly T1-​hypointense, and T2-​iso-​or hyperintense (Figure 2.6).
    Hippel Lindau syndrome; tumour nodules are usually multiple, and Depending on the subtype, various degrees of contrast enhance-
    dilated feeding and draining vessels may be seen. ment are noted. Exact identification of the epicentre of the tumour
    Next to the ‘benign’ PAs, each grade of astrocytomas (WHO I–​IV) may allow differentiation of tumour subtype. In advanced stages,
    may be encountered in the posterior fossa. Frequently, lesions start the tumour may invade the fourth ventricle making differentiation
    as lower-​grade lesions with de-​differentiation into higher-​grade less reliable. If the fourth ventricle is invaded, CSF seeding of the
    astrocytomas during disease progression. High-​grade or anaplastic neuroaxis must be excluded. Spinal imaging should preferably be
    astrocytomas may show diffuse infiltrative, ill-​defined tumour com- done preoperatively, because a mild post-​operative leptomeningeal
    ponents and are, due to their close proximity to the fourth ventricle enhancement may mimic tumour seeding.

    Figure 2.5 Axial T2-​, T1-​, and contrast-​enhanced T1-​weighted imaging of child with a partially solid, predominantly cystic pilocytic astrocytoma
    within the left cerebellar hemisphere and vermis. The fourth ventricle is compressed and displaced to the right.
    18 CHAPTER 2 Imaging in Paediatric Oncology

    Figure 2.6 Axial T2-​weighted MRI of three different children with a SHH (left), WNT (middle), and group 3 (right) medulloblastoma. The epicentre
    of the SHH subtype is midline and in a more cranial location (originating from the upper rhombic lip) compared to the WNT medulloblastoma
    (originating from the lower rhombic lip) which is consequently located in close proximity to the foramen of Luschka and CPA. The group 3
    medulloblastoma is in a typical midline location.

    Brainstem gliomas are more accessible to surgical resection. Tectal gliomas have a
    Brainstem gliomas can be classified according to their primary lo- good prognosis, may remain stable for many decades, and usu-
    cation into (a) diffuse infiltrative brainstem glioma (also known as ally only require symptomatic treatment. The proximity to the
    pontine glioma), (b) exophytic glioma of the medulla oblongata, and Sylvian aqueduct usually results in an obstructive hydrocephalus
    (c) tectal glioma. Clinical symptoms are determined by the primary lo- which can be treated by a ventriculo-​peritoneal shunt or a third
    cation and degree of infiltration or compression of adjacent functional ventriculostomy. Because of the stability of the lesions, several
    centres. Gliomas usually develop and grow slowly over a long time groups have discussed if a tectal glioma should be considered to
    period before they become symptomatic. A sudden increase in size is be a variant hamartoma.
    usually an indication for a significant change in biological behaviour or On MRI, most gliomas are T1-​hypointense, T2-​hyperintense, and
    malignancy grade. Patients may present with isolated or combined cra- show no or minimal contrast enhancement. Tectal gliomas are usu-
    nial nerve palsies as well as ataxia. Because these gliomas grow slowly, ally recognized as a ‘thick/​bulky’ quadrigeminal plate. The exophytic
    hydrocephalus is often encountered late in disease progression. glioma of the medulla oblongata is usually characterized by a large
    Most gliomas are recognized between three and ten years of age. exophytic component attached to the medulla oblongata. The MR
    Long-​term prognosis of the diffuse infiltrative brainstem glioma imaging features of the diffuse infiltrative brainstem glioma are
    remains poor. These brainstem gliomas have a low sensitivity for considered to be sufficient to establish diagnosis. Typically, the le-
    chemotherapy, and the young age of most patients, as well as the sion is centred in a diffuse enlarged pons, preserved fibre tracts may
    central location of the tumour, prevents high-​dose radiotherapy be seen as linear signal intensities within the lesion (seen especially
    or surgical exploration/​resection. However, radiotherapy is still well on T2-​weighted imaging), the basilar artery is embraced, and,
    the most important treatment modality. The exophytic glioma depending on the grade of malignancy, multifocal areas of contrast
    of the medulla oblongata has a more favourable prognosis; enhancement may be observed. These gliomas may extend into the
    histologically, they usually belong to the group of PAs and they lower brainstem as well as into the mesencephalon and diencephalon.

    Figure 2.7 Sagittal T1-​, axial T2-​, and contrast-​enhanced T1-​weighted MRI of a child with a glioblastoma multiforme. A large, inhomogeneous,
    partially multicystic, partially solid, contrast-​enhancing tumour is seen within the right parieto-​occipital region.
    Tumours within the central nervous system 19

    Figure 2.8 Sagittal T1, contrast-​enhanced sagittal and coronal T1-​weighted MRI of a child with a sellar/​suprasellar craniopharyngeoma. The lesion
    shows inhomogeneous contrast enhancement with a large cyst protruding into the third ventricle. The optic chiasm is pushed anteriorly.

    Ependymomas and are typically located in the sellar/​suprasellar region along the
    Ependymomas typically originate from the ependymal lining of the axis between the hypothalamus and pituitary gland. These tumours
    fourth ventricle (60–​70% of cases). In contrast to medulloblastomas, most likely originate from residual pluripotent cells after the devel-
    the ‘epicentre’ of ependymomas is within the fourth ventricle. opment of the hypothalamus/​pituitary gland. Clinical symptoms are
    Ependymomas rarely infiltrate the vermis, brainstem, or cerebellum. related to compression of the optic chiasm/​pathway, pituitary gland/​
    They typically respect the boundaries of the ventricular system and hypothalamus, or infiltration of the connection between the hypo-
    extend along the various outlet channels of the fourth ventricle (for- thalamus and pituitary gland. Craniopharyngeomas are easily iden-
    amen of Luschka and Magendie), along the brainstem into basal tified on CT and MRI. They are usually heterogeneous, with solid
    cisterns and the spinal canal. CSF seeding may occur early due to and cystic components, as well as variable degrees of calcification
    the intraventricular location. Clinically, most children present with and contrast enhancement (Figure 2.8). The sellar may be enlarged.
    signs of an obstructive hydrocephalus or compression of the brain- Optic pathway gliomas
    stem and cranial nerves. On imaging, ependymomas are solid, T1-​
    iso-​or hypointense, T2-​iso-​or hyperintense, with a variable degree Optic pathway gliomas represent a separate group of astrocytomas
    of contrast enhancement. Calcifications are seen in 50%, cysts in and are frequently seen in NF1 patients. Histologically, they repre-
    20%, and haemorrhage in 10% of cases. sent PAs. Non-​or minimally enhancing optic nerve gliomas can be
    seen next to T2-​hyperintense, strongly enhancing hypothalamic/​
    Supratentorial tumours chiasmatic tumours. These tumours may be stable for many years
    In contrast to the posterior fossa, the variation of tumours is higher but typically present with slowly progressive blindness and pitu-
    in the supratentorial region. itary/​hypothalamic dysfunction.

    Astrocytomas Supratentorial ependymomas and atypical teratoid/​


    rhabdoid tumours
    Astrocytomas (WHO I–​IV) are the most frequent supratentorial tu-
    mours in children (30%). Clinical symptoms are determined by the Supratentorial ependymomas and AT/​ RTs are rare tumours in
    primary location, degree of infiltration, and compression of func- childhood. Of all ependymomas, only 20–​40% are located within
    tional centres. Focal functional deficits may be seen next to epileptic the supratentorial brain. Supratentorial ependymomas are most
    seizures and more general signs of increased intracranial pressure. frequently located within the periventricular white matter;
    Most paediatric astrocytomas are low-​grade; however, in rare cases, intraventricular ependymomas are much rarer. Local compression
    grade IV glioblastoma multiforme occurs. PAs are much rarer in may result in trapped ventricles with asymmetrical hydrocephalus.
    the supratentorial brain compared to the posterior fossa. Next to Supratentorial AT/​RTs are also rare. These tumours are usually large
    white-​matter locations, astrocytomas may also be located within on initial presentation, heterogeneous, and may be located within the
    the thalami and basal ganglia. Depending on their WHO grade, white matter as well as in the ventricles. Differentiation from other
    the imaging appearance may differ on CT and MRI. Low-​grade rare tumours like choroid plexus malignancies or intraventricular
    astrocytomas are T1-​hypointense and T2-​hyperintense, with no or meningeomas may be challenging.
    minimal contrast enhancement or mass effect. High-​grade gliomas
    typically show an intense, inhomogeneous contrast enhancement, FURTHER READING
    with areas of necrosis within the centre of the tumour (Figure 2.7).
    Abdel Razek AA, et al. (2009) Characterization of pediatric head and
    Diffusion-​weighted imaging may be helpful to differentiate between
    neck masses with diffusion-​weighted MR imaging. Eur Radiol
    a necrotic tumour and an abscess.
    19, 201–​8.
    Craniopharyngeomas Arnold MA, et al. (2016) Histology, fusion status, and outcome in al-
    veolar rhabdomyosarcoma with low-​risk clinical features: a report
    Craniopharyngeomas make up about 15% of CNS tumours in chil- from the Children's Oncology Group. Pediatr Blood Cancer 63(4),
    dren. They are most frequently seen between 10 and 14 years of age 634–​9. doi: 10.1002/​pbc.25862
    20 CHAPTER 2 Imaging in Paediatric Oncology

    Bagatell R, et al. (2016) Assessment of primary site response in chil- Oberlin O, et al. (2015) Prognostic factors for outcome in localized
    dren with high-​risk neuroblastoma: an international multicenter extremity rhabdomyosarcoma. Pooled analysis from four inter-
    study. J Clin Oncol 34(7), 740–​6. doi: 10.1200/​JCO.2015.63.2042 national cooperative groups. Pediatr Blood Cancer 62(12), 2125–​
    Brisse HJ, et al. (2011) Guidelines for imaging and staging of 31. doi: 10.1002/​pbc.25684
    neuroblastic tumors: consensus report from the International Patay Z, et al. (2015) MR imaging characteristics of wingless-​type-​
    Neuroblastoma Risk Group Project. Radiology 261(1), 243–​57. subgroup pediatric medulloblastoma. Am J Neuroradiol 36,
    doi: 10.1148/​radiol.11101352 2386–​93.
    Dome JS, et al. (2015) Advances in Wilms tumor treatment and Pearce MS, et al. (2012) Radiation exposure from CT scans in child-
    biology: progress through international collaboration. J Clin Oncol hood and subsequent risk of leukaemia and brain tumours: a retro-
    33(27), 2999–​3007. doi: 10.1200/​JCO.2015.62.1888 spective cohort study. Lancet 380(9840), 499–​505.
    Giraudo C, et al. (2016) 18F-​ fluorodeoxyglucose positron emis- Roebuck DJ, Perilongo G (2006) Hepatoblastoma: an oncological re-
    sion tomography/​magnetic resonance in lymphoma: comparison view. Pediatr Radiol 36, 183–​6.
    with 18F-​ fluorodeoxyglucose positron emission tomography/​ Rojas Y, et al. (2014) Relapse surveillance in AFP-​ positive
    computed tomography and with the addition of magnetic res- hepatoblastoma: re-​evaluating the role of imaging. Pediatr Radiol
    onance diffusion-​weighted imaging. Invest Radiol 51(3), 163–​9. 44(10), 1275–​80. doi: 10.1007/​s00247-​014-​3000-​6
    doi: 10.1097/​RLI.0000000000000218 Taylor MD, et al. (2012) Molecular subgroups of medulloblastoma: the
    Monclair T, et al. (2009) The International Neuroblastoma Risk Group current consensus. Acta Neuropathol 123, 465–​72.
    (INRG) Staging System: an INRG task force report. J Clin Oncol
    27, 298–​303.
    3
    Chemotherapy and Other Anti-​Cancer
    Drugs
    Current Knowledge and New Perspectives

    C. Michel Zwaan, Gareth J. Veal, and Lucas Moreno

    Introduction Conventional chemotherapeutic agents

    In contrast to many types of cancer arising in adults, the majority Overview


    of those occurring in childhood are sensitive to cytotoxic chemo- Chemotherapy is a key element in the treatment of paediatric haem-
    therapy. This is generally attributed to the high proliferation rate of atological malignancies. In ALL treatment consists of multiple
    childhood malignancies and the propensity of the malignant cells to agents given in combination for a total treatment duration of two
    undergo apoptosis—​features that may be linked to the embryonal to three years. Central to this is multi-​agent induction treatment,
    origin of many of these tumour types. Chemotherapy, therefore, has followed by intrathecal and intravenous chemotherapy to avoid leu-
    a major role in treatment and contributes significantly to the gener- kaemia relapse in the central nervous system (CNS). Further blocks
    ally good outcome for most paediatric malignancies. The treatment of consolidation or reinduction therapy are introduced before the
    of childhood acute lymphoblastic leukaemia (ALL) provides an ex- start of a prolonged phase of maintenance therapy that aims to
    cellent example in this respect, with survival in more than 90% of eradicate minimal residual disease (MRD). In contrast, treatment
    patients. duration is short in acute myeloid leukaemia (AML), with most
    Chemotherapy is used either as the only treatment modality, or in protocols containing four or five blocks of intensive chemotherapy.
    combination with surgery and/​or radiotherapy. However, some dis- In both diseases, many treatment optimization studies have been
    eases remain or become refractory to chemotherapy, and evidence performed over several decades, fine-​tuning the application of con-
    for the long-​term toxicity of some chemotherapeutic compounds ventional chemotherapy. These have included studies implementing
    has emerged. Thus, there is a necessity to understand the mechan- new analogues of existing compounds with greater potency or better
    isms of both therapeutic failure and long-​term toxicity. Taking into pharmacological properties. Examples include asparaginase (E. coli
    account biological and pathological differences between paediatric derived asparaginase, PEGylated asparaginases, and, recently, re-
    and adult cancer, it is clear that specific drug development pro- combinant asparaginase) and corticosteroids (prednisone being re-
    grammes are needed to identify and test new drugs for children. placed by dexamethasone). Nevertheless, a small subset of patients
    Moreover, given the potential differences in drug metabolism at dif- still fail to achieve cure with current therapy and new agents are
    ferent ages, dosing may differ from adults. Young children may also needed to improve outcomes for these individuals.
    need specific formulations (e.g. oral solutions). Fortunately, some In paediatric solid tumours, chemotherapy is generally adapted to
    of these challenges are now being addressed by specific regulations the diagnosis and stage of the malignancy. It is often used to obtain
    facilitating paediatric drug development both in the United States regression of the primary tumour in order to facilitate local tumour
    and Europe. control by surgery and/​or radiotherapy. Preoperative chemotherapy
    The first part of this chapter will summarize the current know- of this type, referred to as ‘neoadjuvant’ chemotherapy, also al-
    ledge of conventional chemotherapeutic agents and their appli- lows evaluation of the chemosensitivity of a particular tumour in
    cation. The second part will focus on the development of novel a specific patient, which may influence the choice of postoperative
    compounds, which is clearly needed as cancer remains the leading treatment. Postoperative ‘adjuvant’ chemotherapy is used to treat
    cause of death due to disease in children. presumed microscopic metastases or remaining viable tumour after
    22 CHAPTER 3 Chemotherapy and Other Anti-Cancer Drugs

    incomplete local therapy. In other tumours, the primary approach is Routes of administration
    surgical removal of the localized primary tumour, after which adju- Most chemotherapy agents are given intravenously (IV), usually
    vant chemotherapy and/​or radiotherapy is given in order to achieve involving a central venous device. Some drugs are administered or-
    optimal local and systemic control. ally, either as tablets, capsules, or liquid preparations, with hospital
    Chemotherapy regimens usually consist of combinations of two pharmacies sometimes developing dose-​adapted capsules or liquid
    or more cytotoxic agents given as sequential courses in order to preparations for those not commercially available. Nasogastric or
    augment the probability of destroying the malignant cells. In most enterostomy tubes may be utilized for drug administration in mainly
    schedules, different classes of drugs with distinct mechanisms of ac- younger children. Subcutaneous and intramuscular dosing is less
    tion are used together in an attempt to achieve synergism and reduce acceptable in children due to the distress caused. Intrathecal admin-
    the likelihood of chemoresistance. The choice of chemotherapy de- istration of chemotherapy is used in leukaemia and most subtypes of
    pends on several factors such as tumour type, preclinical evidence non-​Hodgkin lymphoma (NHL), and is sometimes also applied in
    of in vitro cytotoxicity and in vivo activity of a specific drug, and treatment of brain tumours. Other specialized routes of administra-
    on the expected toxicity of the proposed combination. Sometimes tion which may be occasionally encountered include isolated limb
    preclinical data provide evidence of synergy between two or more perfusion for refractory sarcoma, or intra-​arterial administration in
    compounds, although most standard chemotherapy regimens util- hepatic tumours. A novel way of administration considered in brain
    ized during the last three decades have been developed empirically. tumours consists of intra-​tumoural infusion via specifically devel-
    oped catheters.
    Classification of chemotherapeutic agents
    There are several classes of anti-​cancer drugs that are defined by Toxicity of chemotherapy
    their mode of action. In general, drugs of different classes are Side effects of chemotherapy can be divided into acute and long-​
    chosen for combination regimens. The classification used for term toxicities. The latter become more relevant with increasing
    drugs most frequently used in paediatric oncology is summarized cure rates and because of the young age of most patients. Since most
    in Table 3.1. chemotherapeutics exert their mode of action on proliferating cells,

    Table 3.1 Summary of chemotherapeutic drugs

    Class Mode of action Examples


    Alkylating agents Form covalent bonds between alkyl groups and Mechlorethamine, cyclophosphamide, ifosfamide, melphalan,
    nucleotides in DNA busulfan, thiotepa
    Alkylation alters DNA and its replication Nitrosureas: carmustine (BCNU) and lomustine (CCNU)
    May induce mutations in DNA Procarbazine and dacarbazine (DTIC)
    Temozolomide
    Platinum compounds Form covalent bonds between platinum atoms and DNA, Cisplatin, carboplatin, oxaliplatin
    resulting in intrastrand adducts or interstrand crosslinks,
    leading to DNA damage
    Anti-​metabolites Interfere with synthesis of DNA/​RNA precursors Purine analogues (mercaptopurine, thioguanine, fludarabine,
    clofarabine)
    Enzyme inhibition (e.g. dihydrofolate reductase or Pyrimidine analogues (cytarabine, gemcitabine, 5-​fluorouracil)
    thymidylate synthetase)
    Anti-​folates (methotrexate, pemetrexed, trimetrexate)
    Anti-​microtubule compounds Interfere with the formation of microtubules during cell Vinca alkaloids (vincristine, vinblastine, vinorelbine)
    division
    Taxanes (paclitaxel, docetaxel)
    Ixabepilone
    Topoisomerase I inhibitors Inhibition of topoisomerase I Irinotecan (CPT-​11) and topotecan
    Epipodophyllotoxins Inhibition of topoisomerase II Etoposide (VP16), teniposide (VM26)
    Antibiotics: Inhibition of topoisomerase II, but act also through DNA Daunorubicin, doxorubicin, epirubicin, idarubicin
    Anthracyclines intercalation
    Anthracenediones Oxidative DNA intercalation Mitoxantrone
    Others DNA binding drug and inhibitor of RNA and protein Bleomycin, Dactinomycin
    synthesis
    Miscellaneous
    Corticosteroids Induction of apoptosis Prednisolone, dexamethasone
    Asparaginase Inhibits asparagine synthetase and depletes intracellular L-​Asparaginase, erwinase, PEGylated-​asparginase
    asparagine levels
    Differentiation-​inducing agents Inhibitor of ribonucleotide reductase Hydroxyurea
    Induces differentiation of promyeloctes or neuroblasts All-​trans retinoic acid, cis-​retinoic acid, fenretinide
    Conventional chemotherapeutic agents 23

    some normally dividing tissues are especially at risk of toxicity. These especially when the toxicity of these drugs is essentially haemato-
    commonly include the bone marrow (aplasia), gastrointestinal mucosa logical. The use of such high-​dose (myeloablative) chemotherapy
    (mucositis and diarrhoea), and hair follicles (alopecia), with most com- approaches requires autologous or allogeneic haematopoietic
    pounds causing transient toxicities on these organs. Some drugs may stem cell support, which assures reconstitution of haematopoiesis.
    also have a specific toxicity on one or more organ functions, often with Maximum therapeutic effect is obtained and non-​haematological
    implications in the mid to long term. For example, the anthracyclines toxicity becomes the dose-​ limiting factor. High-​
    dose chemo-
    are well known for their potential cardiac toxicity, but this is rarely therapy strategies are mostly applied in very good partial or com-
    evident at the time of treatment. Cisplatin and, to a lesser extent, plete remission in chemotherapy-​sensitive tumours with high risk
    carboplatin may be ototoxic and/​or nephrotoxic. Methotrexate may of relapse. The most common examples of chemotherapeutics used
    have acute renal, liver, or CNS toxicity, but this is generally not of long-​ for high-​dose chemotherapy are the alkylating agents. In paediatric
    term significance. The use of alkylating agents may result in secondary solid tumours, high-​dose chemotherapy with autologous stem cell
    malignancy and fertility damage in the long term. Epipodophyllotoxins rescue is most often used in neuroblastoma (NBL), Ewing sarcoma,
    may induce secondary myelodysplasia and/​or AML. and, more recently, also in medulloblastoma. In relapsed or high-​
    The risk of severe late effects may be avoided by reducing cumu- risk leukaemia, myeloablative regimens with allogeneic haemato-
    lative dose exposures, or by avoiding the use of compounds with poietic stem cell transplant may also include immunosuppressive
    unfavourable toxicity profiles. Such approaches are now being in- drugs such as fludarabine, monoclonal antibodies directed against
    vestigated, as exemplified by the current UK ALL 2011 study, which T cells, or total body irradiation. Where allogeneic stem cell sup-
    was designed with the primary objective to investigate the effect on port is used, a graft-​versus-​leukaemia effect may have an additional
    serious treatment-​related toxicity of receiving a higher dose but re- anti-​tumour effect besides the cytotoxic effects of chemotherapy
    duced number of days of dexamethasone treatment. and/​or TBI.
    The common incidence of toxicity associated with the use of
    cytotoxic chemotherapy reinforces the need to develop drugs with Drug resistance
    novel mechanisms of action, directed against specific biological tar- There are several important mechanisms of drug resistance, of
    gets characteristic of a given cancer type (also referred to as targeted which the multidrug resistance (MDR) phenotype, commonly ex-
    therapy). Targeted compounds may induce fewer side effects, al- pressed by cancer cells, is the best characterized. This is caused by
    though experience in children is still very limited and new toxicities enhanced expression of drug-​efflux pumps on the cancer cell mem-
    may emerge. The toxicity profiles of various compounds commonly brane. Several proteins are implicated including P-​glycoprotein,
    used in paediatric oncology are described in Table 3.2. The severity multidrug resistance proteins 1–​8 (MRP), breast cancer resistance
    of chemotherapy-​induced toxicity can be classified using the system protein (BCRP), and the lung resistance-​related protein (LRP).
    of toxicity developed by the National Cancer Institute, the Common High expression of these membrane proteins in tumour cells has
    Terminology Criteria for Adverse Events. been correlated with poor outcome in some malignancies, although
    Dose intensity the value of such observations remains controversial. Various pre-
    clinical models have been developed to circumvent the MDR-​
    Chemotherapeutics are usually delivered at a maximum tolerated related mechanism of drug resistance but clinical trials using MDR
    dose (MTD) in order to obtain the maximum anti-​tumour effect, blocking agents (e.g. verapamil, ciclosporin A, PSC 833) have so far
    largely due to the exhibited dose–​effect relationship and the narrow failed to demonstrate clear benefits. It is likely that other mechan-
    therapeutic window. The concept of dose intensity is defined as the isms are concurrently or sequentially involved in the development
    amount of drug administered per unit of time. Strategies which of drug resistance.
    incorporate high-​dose intensity have proven beneficial in several
    paediatric tumours. For example, in Burkitt lymphoma more than Dosing of drugs in paediatric oncology
    90% of the patients can now be cured by intensive multi-​agent chemo-
    Body surface area
    therapy regimens of relatively short duration in combination with
    an anti-​CD20 monoclonal antibody (rituximab). Dose-​intensive The principle of using body surface area (BSA) for dosing chemo-
    regimens demand the shortest intervals possible between sequen- therapy results from pharmacological research studying drug ab-
    tial courses of treatment and may cause severe toxicity. Improved sorption, distribution, metabolism, and excretion. Data show that
    supportive care protocols (e.g. anti-​emetic therapy, haematopoietic the most reliable method of comparing physiological variables (e.g.
    growth factors, antibiotic and antifungal prophylaxis, transfusion kidney function (glomerular filtration rate), cardiac output, and
    support) have been developed, which has allowed further dose in- basal metabolic rate) between species and across ages is by cor-
    tensification without increasing treatment-​related mortality. recting for BSA. Use of body weight instead of BSA results in an
    More recently there has been increased interest in metronomic unreliable interspecies correction of all mechanisms contributing
    chemotherapy. This is based on the chronic administration of to the exposure. Although the use of BSA is controversial in adult
    chemotherapeutic agents at relatively low, minimally toxic doses, oncology, variability in weight and height is so great in children that
    and without prolonged drug-​free breaks. The precise mechanism of prescriptions are almost always based on BSA. BSA can be calculated
    action is uncertain and is likely to be multifactorial. by formulas or the use of nomograms, with the ‘gold standard’ for-
    mula described by Dubois in 1916. Although very reliable, this for-
    High-​dose myeloablative chemotherapy mula is not easy to use and other formulas have since been validated,
    Several cytotoxic compounds have roughly linear dose–​effect re- taking into account either weight and height/​length or weight only
    lationships and are suitable for administration at high doses, (Table 3.3).
    24 CHAPTER 3 Chemotherapy and Other Anti-Cancer Drugs

    Table 3.2 Specific adverse effects of cytotoxic compounds categorized as short-​term (days to weeks) and long-​term (months to years) side
    effects. (General adverse effects such as bone marrow suppression, gastrointestinal toxicity, and alopecia are not included.)

    Drug Short-​term side effects Long-​term side effects


    Asparaginase Clotting disorders, anaphylactic reactions, pancreatitis, hyperglycaemia Sequelae from thrombosis/​bleeding (CNS or elsewhere) or
    pancreatitis
    Bleomycin Fever, malaise, skin rash Pulmonary fibrosis
    Busulfan Veno-​occlusive disease (VOD), seizures, hyperpigmentation Hyperpigmentation, pulmonary fibrosis, fertility disorders
    Carboplatin Ototoxicity, allergic reactions Ototoxicity
    Cisplatinum Renal toxicity, ototoxicity, radio-​sensitization Renal toxicity, ototoxicity
    Cyclophosphamide Haemorrhagic cystitis Fertility disorders, secondary leukaemia
    Cytarabine Mucositis, rash, conjunctivitis, fever, encephalopathy, seizures, Encephalopathy
    pancreatitis
    Dactinomycin Jaundice, veno-​occlusive disease, radio-​sensitization Unknown
    Daunorubicin Mucositis, cardiomyopathy, radio-​sensitization Cardiomyopathy, secondary leukaemia
    Dexamethasone Mood disorders, increased appetite, Cushingoid appearance, muscular Bone fractures, avascular femoral head necrosis, vertebral
    atrophy, bone demineralization, skin disorders collapse, adrenal suppression, metabolic syndrome
    Doxorubicin Mucositis, cardiomyopathy, radio-​sensitization Cardiomyopathy, secondary leukaemia
    Epirubicin Mucositis, cardiomyopathy, radio-​sensitization Cardiomyopathy, secondary leukaemia
    Etoposide (VP-​16) Allergic reactions, mucositis Secondary leukaemia
    Fludarabine Mucositis, fever, pneumonitis, neurotoxicity, hepatitis Unknown
    Idarubicin Mucositis, cardiomyopathy, radio-​sensitization Cardiomyopathy, secondary leukaemia
    Ifosfamide Haemorrhagic cystitis, tubulopathy, encephalopathy, seizures Tubular and glomerular renal toxicity, fertility disorders
    Irinotecan Abdominal pain, diarrhoea, sweating, hyper lacrimation, salivary excess Unknown
    Melphalan Mucositis, interstitial pneumonitis Pulmonary fibrosis, fertility disorders
    6-​Mercaptopurine Hepatitis Unknown
    Methotrexate Hepatitis, mucositis, encephalopathy, renal toxicity Liver fibrosis, encephalopathy
    Mitoxantrone Mucositis, cardiomyopathy Cardiomyopathy
    Prednisolone Mood disorders, increased appetite, Cushingoid appearance, muscular Bone fractures, avascular femoral head necrosis, vertebral
    atrophy, bone demineralization, skin disorders collapse, adrenal suppression, metabolic syndrome
    Procarbazine Allergic reactions, hepatic dysfunction, headache, paraesthesia, Fertility disorders
    hallucinations
    Temozolomide Lymphocytopenia Secondary malignancy
    Teniposide (VM-​26) Allergic reactions, mucositis Secondary leukaemia
    6-​Thioguanine Hepatitis, VOD Unknown
    Thiotepa Headache, encephalopathy, dizziness, allergic reactions, skin rash, fever Fertility disorders
    Topotecan Mucositis, radio-​sensitization Unknown
    Vinblastine Paraesthesia, neuralgia, sensory disorders, hypertension, Raynaud’s Raynaud’s phenomenon
    phenomenon
    Vincristine Paraesthesia, neuralgia, muscular weakness, sensory disorders, Neurotoxicity
    constipation, ileus, abdominal cramps, seizures, SIADH

    Table 3.3 Formulas for calculating body surface area (BSA), body Chemotherapy in infants
    mass index (BMI), and ideal body weight
    Tolerance to chemotherapy at a given dose is generally poorer in in-
    Type Formula fants when compared to older children and adults. Maturation of
    Dubois’ formula BSA = W0.425 × L0.725 × 0.007184 the physiological mechanisms that contribute to the pharmacokin-
    etics of drug handling occurs during the first year of life. For ex-
    Mosteller’s formula
    BSA = W × L / 3600 ample, the water content of the human body decreases from 75%
    at birth to 60% at age one and 55% in adults. The content of plasma
    Formula without using length BSA = (4 W + 7)/​( W + 90)
    proteins also changes during the first year of life and hepatic drug-​
    Body mass index BMI (kg/​m2) = W/​L2 metabolizing enzymes (cytochrome P450 isoenzymes, Uridine 5'-​
    Lawrence’s formula Ideal body weight diphospho (UDP)-​glucuronyltransferase, glutathione metabolizing
    (kg) = [L−100]−[(L−150)/​K] enzymes) achieve their full physiological activity between six and
    Key: BSA = body surface area in m2; W = body weight in kg; L = body length in cm; K = 4 twelve months of age. Glomerular filtration rate (GFR) attains values
    for males and 2 for females.
    Conventional chemotherapeutic agents 25

    28 Chemotherapy in obese patients

    26 Obesity is associated with modifications in body composition that


    may change the pharmacokinetics of cytotoxic compounds, po-
    24 tentially resulting in either inadequate drug exposure or increased
    BW/BSA (kg/m2) toxicity. Increased body fat mass will alter the distribution volume
    22
    of a drug depending on its affinity for fatty tissues and plasma pro-
    20 teins. Fatty degeneration of liver tissue may also modify hepatic
    metabolizing capacity in obese patients.
    18 The diagnosis of obesity is defined by the World Health
    Organization (WHO) as a body mass index (BMI) (Table 3.3)
    16
    of more than 30 kg/​m2. For children up to five years of age,
    14 overweight and obesity are defined as more than two or three
    0 1 2 3 4 5 6 7 8 9 10 standard deviations above the median. For older children, this
    Age (years) is defined as more than one or two standard deviations above
    the median.
    Figure 3.1 Development of body weight (BW) to body surface area
    (BSA) ratio during childhood.
    For most cytotoxic compounds, no pharmacokinetic data are
    available in obese patients and, thus, it is recommended that chemo-
    therapy is prescribed according to ideal body weight instead of
    actual body weight to avoid excessive toxicity. In adolescents and
    comparable to those in adults at the age of five months. All these fac-
    young adults, ideal body weight can be calculated according to the
    tors contribute to the poorer tolerance of infants to chemotherapy,
    formula by Lawrence (Table 3.3). In younger children, the ideal
    and treatment doses should be reduced.
    body weight is best determined by the body weight corresponding
    In general, chemotherapy in infants should be prescribed as mg/​
    to the centile of height.
    kg body weight instead of BSA, since the relationship between body
    For those drugs where clinical pharmacology studies have been
    weight and BSA is different in infants as compared to older children
    carried out, clinically significant variations in pharmacokinetics
    (Figure 3.1). A dose prescribed in mg/​m2 in an infant will lead to a
    have been observed in obese children for approximately two-​thirds
    higher dose than one prescribed in mg/​kg (Table 3.4). In neonates
    of drugs studied, with these differences resulting in marked dif-
    and very young children (less than three months old), even greater
    ferences in drug exposure in approximately one-​third of cases. It
    caution is required and additional dose reductions are frequently ap-
    is clear that more research needs to be carried out in this area, to
    plied, although many of these are empirical. It is important to appre-
    achieve a consensus as to the most appropriate approach to dosing
    ciate the balance between avoiding drug exposures that are likely to
    chemotherapy in obese children.
    lead to toxicity and ensuring that drug levels are achieved that will
    lead to effective anti-​tumour activity.
    In order to provide a more scientific rationale for the dosing of Prescribing chemotherapy in difficult situations
    neonate and infant cancer patients, it is important that more clinical Some patients may require chemotherapy in an acute situation
    pharmacology studies are carried out in this population. An example where organ failure (most typically hepatic or renal failure) limits
    is provided by the drug 13-​cis-​retinoic acid for NBL: pharmacokinetic the clearance of chemotherapy. In such situations, chemotherapy
    studies highlighted a trend towards lower exposures being observed in doses should be reduced, taking into account the type of drug and
    children less than 12 kg, receiving a mg/​kg dosing regimen, as com- the mechanisms of clearance for that drug. In clinical situations
    pared to standard treatment in mg/​m2 in older children, resulting in where dialysis is used, consideration should be given to whether
    removal of the mg/​kg dosing recommendation. The study also high- this contributes to the wash-​out of the drug. In addition, the poten-
    lighted the importance of the availability of appropriate drug formula- tial additional toxicity on the affected organ should be taken into
    tions, as children unable to swallow isotretinoin capsules experienced account. Drugs which have been shown to exhibit marked differ-
    lower drug exposures following extraction of the drug from capsules ences in disposition in children receiving dialysis include platinum-​
    prior to administration. This again represents an area where more work based drugs and cyclophosphamide. Detailed recommendations
    needs to be done to facilitate the optimal treatment of childhood cancer. about dose adaptations are available from other sources.

    Table 3.4 Comparison between dose prescribed in mg/​m2 versus dose in mg/​kg in children

    Child’s characteristics Dose calculated as Dose reduction from calculation using


    mg/​kg compared to BSA
    Age Weight BSA 100 mg/​m2 3.33 mg/​kg
    2
    10 years 30 kg 1m 100 mg 100 mg 0%
    1 year 10 kg 0.46 m2 46 mg 33 mg 28%
    3 months 6 kg 0.29 m2 29 mg 20 mg 31%

    Key: BSA = body surface area in m2.


    26 CHAPTER 3 Chemotherapy and Other Anti-Cancer Drugs

    Intrathecal chemotherapy dose–​response relationships and narrow therapeutic windows—​


    It is important to recognize that, compared to adults, the CNS com- characteristics which very much lend themselves to TDM ap-
    partment in a child is proportionally larger than the rest of the proaches. TDM approaches have been shown to be beneficial for
    body. For example, a child between four and six years of age has a several chemotherapeutics, including methotrexate, busulfan, and
    CNS volume of 80–​90% compared to the adult brain. This results carboplatin, alongside some of the newer targeted anti-​cancer drugs.
    in the prescription of intrathecal therapy in an absolute dose (mg), For these drugs and others, the use of TDM approaches should be
    depending on the age of the patient, rather than in mg/​m2. encouraged, particularly in relation to the treatment of patient
    subpopulations such as neonates and infants, anephric patients re-
    Pharmacogenetics and drug interactions quiring dialysis, and patients receiving high-​dose chemotherapy.
    Intracellular metabolism contributes to the clearance of drugs by
    transforming them to metabolites in order to facilitate renal or bil-
    iary excretion. Generally, drug metabolites have less or no thera- Drug development in paediatric oncology
    peutic activity, although some metabolites may still have a strong
    cytotoxic effect. In some cases prodrugs are administered to pa- Overview
    tients, which require metabolism to their active forms, as is the case There are several factors that influence the development of novel
    for cyclophosphamide. Many enzymes are implicated in the bio- compounds, analogues, or formulations for children with cancer:
    transformation of drugs, of which the cytochrome P450 isoenzymes,
    1. The need for appropriate safety and activity/​efficacy studies in
    the glucuronidation pathways, and detoxifying enzymes implicated
    a paediatric population with biologically different subtypes of
    in glutathione metabolism are most relevant. The activity of these
    cancer compared to adults.
    enzymes influences drug plasma concentrations and thus concen-
    trations in tumour tissue. 2. Reduction of side effects and studying the effect on growth and
    It is useful to know the metabolic pathways of drugs in order to development or other long-​term safety concerns.
    identify potential drug interactions. Hepatic mechanisms, such as 3. Development of age-​adapted paediatric formulations, and the
    CYP450 enzyme induction or inhibition, may dramatically alter ability to deliver oral compounds by nasogastric tube and/​or
    the metabolism and hence the efficacy of a cytotoxic drug. Anti-​ with food.
    epileptics are well known to induce the activity of the CYP3A4 and 4. Pharmacokinetic aspects, especially in children below the age of
    CYP2C9 isoenzymes, and thus increase the biotransformation and two years.
    clearance of many cytotoxic drugs. On the other hand, drugs such as 5. Market authorization/​regulatory factors.
    azoles may inhibit the activity of several CYP isoenzymes, resulting Potential pharmacokinetic differences between children and adults
    in higher plasma concentrations and increasing toxicity of cytotoxic need to be evaluated to ensure paediatric dosing schedules are safe
    drugs, such as vincristine. Drug interactions should be taken into and effective. The use of population pharmacokinetic models has
    account when prescribing both chemotherapeutic and other drugs enabled the feasibility of such studies with limited sampling, and is
    used for supportive care. being applied in paediatric oncology. However, in most studies the
    As the influence of these pharmacokinetic modifying enzymes has differences in pharmacokinetic properties between adults and older
    emerged, population-​based studies of pharmacogenetics have been children are limited, and there is a relatively high correlation be-
    undertaken. Genetic polymorphisms of these or other enzymes may tween the adult and paediatric MTD. This may be different in very
    cause variable phenotypes, resulting in different enzyme activities young patients.
    from patient to patient and, thus, differences in drug clearance. An Until approximately a decade ago, there was limited interest from
    example of pharmacogenetic variability is encountered in the me- the pharmaceutical industry to perform clinical trials in children. By
    tabolism of 6-​mercaptopurine (6-​MP). This drug is inactivated by 2003, of more than a hundred drugs that had been approved by the
    the enzyme thiopurine methyltransferase (TPMT) and patients with Federal Drug Agency (FDA) for the treatment of cancer, informa-
    TPMT mutations require treatment with lower dosages of 6-​MP to tion about paediatric use was available for less than 50% of those. In
    avoid excessive haematological toxicity. Another example consists 2007, following an earlier similar initiative in the USA, new legisla-
    of polymorphisms in the gene UGT1A1 that lead to large differences tion governing the development of medicines for use in children was
    in the glucuronidation of irinotecan metabolites and, thus, differ- introduced in the European Union. Regulation EC no. 1901/​2006
    ences in systemic exposure of the active metabolite SN38. Future obliges pharmaceutical companies to explore paediatric indications
    insights into genetic polymorphisms implicated in the metabolism when they want to have a new medicinal product authorized for an
    of chemotherapeutics may facilitate individualization of treatment. adult indication. Evidence shows that the regulation is positively
    influencing industry interest in paediatric drug development. Since
    Therapeutic drug monitoring
    the introduction of this legislation in Europe, more than eight hun-
    An important approach to the treatment of children with cancer dred paediatric investigational plans (PIPs) have been submitted, of
    is the utility of therapeutic drug monitoring (TDM). This relates which 83 are in oncology (as of December 2015). This has resulted in
    to the real-​time quantification of drug levels in an individual pa- four new drug approvals for cancer in children: everolimus, recom-
    tient, with a view to adapting the dose of drug administered to im- binant asparaginase, imatinib, and dintuximab. This has been sup-
    prove the likelihood of clinical response, whilst minimizing the ported in Europe through establishment of the European Network
    incidence of toxicity. Many well-​established cytotoxic drugs exhibit of Paediatric Research at the European Medicines Agency (Enpr-​
    significant inter-​patient pharmacokinetic variability and have steep EMA) group—​a consortium of research networks, investigators,
    Drug development in paediatric oncology 27

    and centres with recognized expertise in performing clinical studies radiotherapy, although its single-​agent activity against high-​grade
    in children. For paediatric oncology, this is the result of the achieve- glioma and brainstem glioma is limited (response rates of 5–​20%). It
    ments of major academic collaborative groups such as the European is possible that new metronomic dosing schedules may improve effi-
    Innovative Therapies for Children with Cancer (ITCC) consor- cacy and/​or may be more compatible with concurrent radiotherapy.
    tium, and the Phase I consortium of the Children’s Oncology Group Temozolomide also seems to be active against medulloblastoma
    (COG) in North America. and NBL, although it has not yet been integrated into first-​line
    chemotherapeutic regimens and is mostly used as salvage regimen.
    A revised drug development model Taxanes and gemcitabine have also been explored in children but,
    A revised drug development model for paediatric cancers was cre- so far, have not shown sufficiently promising results to be evaluated
    ated by the ITCC and others, in the era of molecularly targeted in frontline trials. Clofarabine is a novel deoxyadenosine analogue
    drugs, and includes the following steps: which has been registered in Europe for use in children with re-
    lapsed/​refractory ALL. Interestingly, clofarabine is one of the few
    1. Coordination of molecular profiling studies to identify new tar-
    drugs where a label was first obtained in paediatrics. Clofarabine
    gets and potential predictive (selection) biomarkers, and to iden-
    can either be used as a single agent or in combination with cyclo-
    tify paediatric clinical patterns of drug targets matching with
    phosphamide and etoposide. In AML, it is used in combination
    new (adult) compounds.
    with cytarabine, replacing fludarabine in the FLA (fludarabine and
    2. Biological and preclinical research to select the most promising
    high-​dose cytarabine) regimen. Nelarabine is a purine nucleoside
    drugs and combinations, and identification of the patient sub-
    analogue (ARA-​G) and is registered for use in paediatric relapsed/​
    groups that could most benefit by matching the mechanism of
    refractory T-​cell ALL and NHL. In COG, the use of nelarabine in up-
    action of the targeted agent with the biology of paediatric cancers.
    front protocols for T-​cell disease is currently being evaluated.
    3. First-​in-​child early phase I/​II combination clinical studies with
    strong biological rationale, incorporating pharmacokinetic,
    Agents targeting specific molecular aberrations that
    pharmacodynamic, and predictive biomarker studies where
    are oncogenic drivers
    possible. Studies designed to maximize efficiency, using the least
    number of patients to identify the recommended phase II dose, Tyrosine kinase inhibitors inhibiting the ABL gene
    incorporating expansion cohorts for preliminary efficacy evalu- Imatinib mesylate is a small molecule drug and the prime example
    ation to gain further knowledge on the tumour types or bio- of successful inhibition of the tyrosine kinase ABL, which is con-
    logical entities of interest. stitutively active as a result of a translocation between the BCR
    4. Drugs that are successful in first-​in-​child trials should be taken gene on chromosome 22 and the ABL gene on chromosome 9 (the
    forward into randomized multi-​arm trials in the relapse setting. Philadelphia chromosome) in chronic myelogenous leukaemia
    5. Drugs that have promising activity are then taken into the (CML) or Philadelphia-​chromosome positive (Ph+) ALL. Only
    frontline setting in larger randomized clinical trials conducted limited phase I and II studies have been performed in paediatrics
    by tumour-​specific cooperative groups. and approval for use in children was mainly granted on extrapo-
    lating adult data. In CML, imatinib has changed the existing para-
    Novel drugs can be broadly categorized as follows, although sig-
    digms of treatment with cytarabine/​interferon followed by stem-​cell
    nificant overlap can exist:
    transplantation and is now a first-​line treatment for this disease,
    1. Recently developed cytotoxic drugs. achieving high rates of remissions. Given the relatively recent intro-
    2. Agents targeting specific molecular aberrations that are onco- duction of BCR-​ABL tyrosine kinase inhibitors (TKIs) and the
    genic drivers. durable exposure to these agents, long-​term toxicities need to be
    3. Agents targeting signalling pathways relevant to paediatric carefully monitored. In Ph+ ALL, a recent COG study combining
    cancers. imatinib with chemotherapy showed a clear improvement in out-
    4. Epigenetic agents. come for children treated with the combination compared to histor-
    5. Agents against tumour microenvironment and angiogenesis. ical data. Current studies in Ph+ ALL aim to decrease chemotherapy
    while maintaining cure rates.
    6. Targeted inhibitors of cell cycle and DNA repair.
    Imatinib also inhibits the targets c-​ Kit, and PDGFR-​ α and​
    7. Immunotherapy and monoclonal antibodies.
    β—​ a finding which led to its successful introduction in spe-
    Over the following section, some examples are provided. A de- cific solid tumours, notably gastrointestinal stromal tumours and
    tailed description of agents and targets can be found in the ‘Further dermatofibrosarcoma protuberans, especially when radical surgery
    reading’ section. fails. Other disease indications are the Philadelphia-​like cases of
    ALL with fusions of ABL-​class kinase genes, including ABL1, ABL2,
    Recently developed cytotoxic drugs CSF1R, and PDGFRB gene fusions. Imatinib is also registered for
    Several new cytotoxics and improved analogues have been relatively hypereosinophilic syndrome.
    recently developed. Most new compounds have a slightly different Despite the success story of imatinib in CML, it was quickly ap-
    cytotoxic mechanism of action and/​or better pharmacological prop- parent that new drugs might also give rise to old problems, in terms
    erties as compared to the conventional chemotherapies of the class of development of resistance, in this case through mutations in ABL.
    to which they belong. These mutations prohibit effective binding of imatinib to the kinase
    The alkylating agent temozolomide now has an established thera- pocket and hence cause resistance. Several second-​generation kinase
    peutic role in the treatment of high-​grade glioma with concurrent inhibitors are now available, which can still bind despite the presence
    28 CHAPTER 3 Chemotherapy and Other Anti-Cancer Drugs

    of a mutation. The best-​studied second-​generation inhibitor in chil- Agents targeting signalling pathways relevant
    dren is dasatinib. In general, the drug was better tolerated than in to paediatric cancers
    adults and appeared effective as salvage treatment in CML and Ph+
    Sorafenib
    ALL, although responses were of short duration in the latter group.
    Studies with nilotinib and bosutinib are in progress. Sorafenib was initially developed as a targeted compound against
    melanoma given its selectivity against BRAF, especially its mutant
    Tyrosine kinase inhibitors inhibiting anaplastic lymphoma variant V600E. However, the compound proved to be a multi-​target
    kinase (ALK) TKI, with broad selectivity against the receptor tyrosine kinases
    VEGFR-​1 and -​2, PDGFR, c-​KIT, and RET. Sorafenib is active
    Anaplastic lymphoma kinase (ALK) translocations were ini-
    against renal cell carcinoma and hepatocellular carcinoma. In paedi-
    tially identified in anaplastic large-​cell lymphoma (ALCL), but,
    atrics, the recommended phase II dose of sorafenib administered
    more recently, point mutations and amplifications have been dis-
    every 12 hours continuously for children with solid tumours is 200
    covered in NBL, and translocations in patients with inflammatory
    mg/​m2/​dose and 150 mg/​m2/​dose for children with leukaemias.
    myofibroblastic tumours (IMT) (a rare type of sarcoma). Crizotinib
    Sorafenib is currently used off-​label as a FLT3-​inhibitor in combin-
    is a first-​in-​class ALK inhibitor, which also inhibits MET and ROS,
    ation with chemotherapy in various collaborative group paediatric
    and was developed in adults with ALK-​translocated non-​small-​cell
    AML studies, with comparable toxicities observed in children and
    lung cancer (NSCLC). In a paediatric study, the recommended phase
    adults.
    II dose was 280 mg/​m2 twice daily—​much higher than the approved
    dose for lung cancer in adults. Anti-​tumour activity was outstanding Agents targeting epidermal growth factor receptors
    for tumours with ALK translocations such as ALCL, NSCLC, and
    Several compounds targeting the epidermal growth factor receptors
    IMT, but disappointingly low for tumours with ALK mutations such
    (EGFRs)—​a member of the ErbB family of receptors (EGFR (ErbB-​
    as NBL. Preclinical evidence suggests that the resistance to ALK-​
    1), HER2/​c-​neu (ErbB-​2), Her 3 (ErbB-​3), and Her 4 (ErbB-​4))—​
    mutated NBL can be overcome by more potent inhibitors such as
    have been developed in adult oncology, and include monoclonal
    lorlatinib, or combination strategies with other targeted agents or
    antibodies and small molecules. FDA approval was obtained for the
    chemotherapy. Further studies of crizotinib and ceritinib (LDK378)
    treatment of patients with NSCLC (gefitinib, afatinib, and erlotinib),
    have confirmed the positive results for ALK-​translocated cancers.
    colorectal cancer (cetuximab and panitumumab), head and neck
    Important questions relating to long-​term safety, CNS penetrance,
    cancer (cetuximab), pancreatic cancer (erlotinib), and breast cancer
    and secondary resistance mutation development still need to be
    (lapatinib).
    answered.
    The potential role of these compounds in paediatric oncology
    seems limited to high-​ grade glioma (HGG) in which EGFR
    Tyrosine kinase inhibitors inhibiting BRAF and MEK
    overexpression, amplification, or mutation can be found. Erlotinib
    BRAF (serine/​threonine-​protein kinase B-​raf) inhibitors have been has been evaluated in a phase I study in children with brain tumours,
    successfully developed to treat adults with melanoma harbouring either as a single agent or with radiotherapy. However, a phase II
    the BRAF V600E mutation. While melanomas are very rare in study in patients with HGG failed to show clear benefit. In a phase
    paediatrics, other tumour types harbour the same mutation, which I paediatric study, a recommended phase II dose of 900 mg/​m2
    is an oncogenic driver, and these patients have experienced signifi- twice daily was established for lapatinib. High-​level expression of
    cant benefit from treatment with BRAF inhibitors. Approximately phosphorylated EGFR and ERBB2 receptors was observed particu-
    10% of cases of low-​and high-​grade glioma have these mutations, larly in ependymomas, which showed prolonged stable disease on
    as well as a significant proportion of children with Langerhans cell lapatinib. In a phase II study however, activity in medulloblastomas
    histiocytosis (LCH). Response rates above 60% in patients with and ependymomas was limited. The failure of the anti-​EGFR agents
    heavily pretreated relapsed or refractory V600E positive tumours to improve outcome in brain tumours is probably related to limited
    have been shown for dabrafenib, with an acceptable toxicity profile drug penetration, a low frequency of mutations, and the failure to
    mainly consisting of rash, fever, and arthralgia. significantly affect downstream signalling with activation of par-
    MEK (mitogen-​activated protein kinase kinase) is an intracel- allel signalling pathways. Currently, paediatric studies are initiated
    lular kinase downstream of BRAF, with MEK inhibitors developed with afatinib, which aims, in the expansion phase, to limit treatment
    to overcome resistance to BRAF inhibition in adult patients with to a biomarker-​selected group of patients and the antibody drug
    melanoma. In addition to paediatric patients with V600E mutant conjugate ABT-​414, which is an anti-​EGFR antibody linked to an
    cancers, MEK inhibitors have also proven promising in other RAS-​ anti-​microtubule agent.
    driven cancers such as plexiform neurofibroma related to NF1, and
    might have a role in other tumour types such as gliomas harbouring Agents targeting insulin growth factor receptors
    BRAF fusions, NBL, or other tumour types that acquire pathway Various compounds targeting the insulin growth factor receptor
    mutations at relapse. Various leukaemias also carry mutations in the IGF1R have been investigated in phase I/​II trials in paediatrics, since
    RAS pathway, most typically juvenile myelomonocytic leukaemia there is significant preclinical evidence that IGF1R signalling seems
    (JMML), and need to be studied. So far, three MEK inhibitors to contribute to the malignant behaviour of various sarcomas. At
    have been tested in paediatrics in early clinical trials: selumetinib, least five different antibodies targeting IGF1R have recently been
    trametinib, and cobimetinib. For selumetinib (the first with phase tested. However, most of these compounds have since been taken off
    I completed), most common toxic effects included acneiform rash, the market as results in adults were not sufficiently convincing. In pa-
    gastrointestinal effects, and asymptomatic creatine kinase elevation. tients with Ewing sarcoma family of tumours, the objective response
    Monoclonal antibodies and immunotherapy 29

    rate was around 10% across phase I/​II studies, although these pa- in combination with ICE (ifosfamide, carboplatin, etoposide). In a
    tients derived significant benefit since the responses were maintained phase II window study in 136 children with newly diagnosed Burkitt
    for several years. In rhabdomyosarcoma, osteosarcoma, synovial sar- lymphoma or diffuse large B-​cell NHL, rituximab was well toler-
    coma, and other sarcomas, the response rate was only 2.5%. ated, with a response rate of 41.4%. Recently, results of a phase III
    trial with rituximab added to chemotherapy in frontline high-​risk
    Mammalian target of rapamycin (mTOR) inhibitors B-​NHL became available, and randomization was stopped as the ex-
    Mammalian target of rapamycin (mTOR) inhibitors that are ana- perimental arm with rituximab showed a clinically meaningful im-
    logues of rapamycin—​ so-​
    called rapalogues such as sirolimus, provement for EFS of 84% in the standard chemotherapy arm versus
    everolimus, ridaforolimus, and temsirolimus—​ are being de- 95% in the rituximab plus chemotherapy arm.
    veloped in paediatric oncology in several phase II combination Gemtuzumab ozogamicin (GO), an anti-​CD33 antibody linked
    studies. Everolimus was recently approved for use in children with to the cytotoxic compound calicheamicin, has been tested as a single
    unresectable subependymal giant-​cell astrocytoma. agent or in combination with standard chemotherapy in relapsed/​
    refractory and upfront paediatric AML. Despite clear efficacy in
    Epigenetic agents early clinical studies, GO was recently taken off the market outside
    An area of active clinical research is the inhibition of histone Japan, but regained approval, given that studies in upfront adult
    deacetylases (HDACs), which is thought to re-​activate silenced AML have shown evidence of benefit using a fractionated schedule
    tumour suppressor pathways. HDAC inhibitors have been widely of GO in combination with chemotherapy. Of interest is the de-
    investigated preclinically in vitro and in vivo, and showed positive velopment of a combination of an anti-​CD22 antibody linked to
    results for leukaemia, CNS tumours, and NBL. Both vorinostat calicheamicin (inotuzumab ozogamicin), which will be evaluated in
    and panobinostat have entered paediatric phase I studies. Another paediatric ALL, following studies in adult ALL showing improved
    strategy comprises hypomethylating agents such as decitabine remission induction and survival rates compared to standard-​of-​
    and azacitidine, which are being tested in paediatric trials for care chemotherapy.
    myelodysplastic syndromes, AML, and some solid tumours. The Epratuzumab is a naked antibody directed against CD22, which is
    inhibition of BET (Bromodomain and extra-​terminal) proteins to being tested in a randomized fashion in standard-​risk relapsed ALL
    target MYC-​amplified tumours is another strategy. BET is an im- in the IntReALL study. This follows an initial single-​arm phase I/​II
    portant transcription factor of the MYC family of proteins. However, study in North America in which it was combined with reinduction
    BET inhibitors have still not entered paediatric clinical trials. chemotherapy in relapsed ALL, which showed tolerance and sug-
    gested an improved MRD response.
    Agents against tumour microenvironment Blinatumomab is an example of a BITE (bispecific T-​cell engaging)
    and angiogenesis antibody, linking CD19 expressed on B-​cell precursor ALL cells to
    Several strategies have been developed to interfere with the process CD3 expressed on T-​cells. Results of a phase I/​II study have recently
    of tumour angiogenesis. Such compounds may decrease signalling been reported: among the 70 patients who received the recom-
    through VEGF/​VEGFR by inhibiting either the ligand VEGF or one mended dosage, 39% achieved complete remission within the first
    of the receptors of VEGF (flt-​1, KDR, flt-​4), or inhibit the signalling two cycles, 52% of whom were MRD-​negative. The most frequent
    pathways through PDGF and its receptors (PDGFRα and β). grade 3 or higher adverse events were anaemia (36%), thrombocyto-
    The antibody bevacizumab (Avastin®) was the first to demon- penia (21%), and hypokalaemia (17%). Four patients had a grade
    strate clinical proof of principle, and has been approved for use in 3 or higher cytokine-​release syndrome. Two patients interrupted
    adults. A paediatric phase I trial was conducted by the COG, with treatment after grade 2 seizures. Blinatumomab is now being studied
    bevacizumab administered intravenously every two weeks in 28-​day in relapsed and upfront ALL trials.
    courses in children with solid tumours. No dose-​limiting toxicities Brentuximab vedotin is a conjugated antibody linking a chimeric
    were observed; adverse events included infusion reaction, rash, anti-​CD30 antibody to the antimitotic agent monomethyl auristatin
    mucositis, proteinuria, and lymphocytopenia. Randomized trials E (MMAE), and has been studied in paediatric Hodgkin lymphoma
    have shown that the addition of bevacizumab in combination with (HL) and ALCL, diseases in which CD30 is expressed. The first re-
    standard chemotherapy did not improve survival in metastatic soft sults in paediatric HL show that Brentuximab vedotin 1.8 mg/​kg
    tissue sarcoma and HGG. A randomized phase II trial in NBL con- every three weeks (RP2D) was generally well tolerated and demon-
    tinues recruiting and results will soon be available. Other more po- strated preliminary evidence of activity, with an objective response
    tent inhibitors—​VEGFR tyrosine kinase inhibitors such as axitinib, rate of 64%. Based on these data, brentuximab vedotin is now being
    pazopanib, and lenvatinib (as opposed to bevacizumab which is studied in newly diagnosed HL in a COG study, and brentuximab
    a monoclonal antibody against VEGF)—​are being evaluated in vedotin and crizotinib in combination with standard chemotherapy,
    paediatrics. in upfront ALCL.

    Anti-​GD2 monoclonal antibodies in neuroblastoma


    Monoclonal antibodies and immunotherapy The first monoclonal antibodies that specifically targeted the
    disialoganglioside GD2 on tumour cells of neuroectodermal origin
    Monoclonal antibodies in haematological malignancies entered clinical trials in 1992. In addition to being identified in vir-
    Several antibodies are being tested in various stages of develop- tually all NBL, GD2 expression has also been identified in other
    ment. Rituximab, a chimeric naked anti-​CD20 monoclonal anti- tumour types such as soft-​tissue sarcomas, Ewing sarcoma, osteo-
    body, has been used in salvage treatment for CD20-​positive NHL, sarcomas, small-​cell lung cancers, and melanoma. The chimeric
    30 CHAPTER 3 Chemotherapy and Other Anti-Cancer Drugs

    anti-​GD2 monoclonal antibody dinutuximab entered paediatric European Medicines Agency (2018) Annual Report 2017. Available
    phase I trials both in Europe and USA shortly after development. at: www.ema.europa.eu.
    The COG phase III study (ANBL0032) demonstrated signifi- Gore L, et al. (2013) Targeting developmental pathways in children
    cant improvement of overall survival with immunotherapy using with cancer: what price success? Lancet Oncol 14, e70–​8.
    dinutuximab in combination with IL2 and GM-​CSF compared with Haidar C, Jeha S (2011) Drug interactions in childhood cancer. Lancet
    Oncol 12, 92–​9.
    standard therapy. The SIOPEN group has developed another anti-
    Hanahan D, Weinberg RA (2011) Hallmarks of cancer: the next gener-
    body called dinutuximab-beta that is given in a long-​term infusion
    ation. Cell 144, 646–​74.
    and appears to have more tolerable side effects. The role of other
    Hirsch S, et al. (2015) Targeted approaches to childhood cancer: pro-
    cytokines such as IL2 or GM-​CSF has still not been established and gress in drug discovery and development. Expert Opin Drug Discov
    efforts continue to identify more tolerable regimens to administer 10, 483–​95.
    immunotherapy with anti-​GD2 monoclonal antibodies. Kearns GL, et al. (2003) Developmental pharmacology—​drug dispos-
    ition, action, and therapy in infants and children. N Engl J Med 349,
    Immunotherapy
    1157–​67.
    The CTLA4 inhibitor ipilimumab and the checkpoint inhibitors Lee DP, et al. (2005) Pediatric phase I trials in oncology: an analysis of
    (anti-​PD1 and anti-​PD-​L1) are discussed elsewhere in this book study conduct efficiency. J Clin Oncol 23, 8431–​41.
    (Chapter xx). Mossé YP, et al. (2013) Safety and activity of crizotinib for paedi-
    atric patients with refractory solid tumours or anaplastic large-​cell
    Miscellaneous lymphoma: a COG phase 1 consortium study. Lancet Oncol 14, 472–​80.
    Differentiation induction, resulting in a maturation of tumour cells, Norris RE, Adamson PC (2012) Challenges and opportunities in child-
    hood cancer drug development. Nat Rev Cancer 12, 776–​82.
    has been utilized as a strategy both in acute promyelocytic leu-
    Panetta JC, et al. (2003) The importance of pharmacokinetic limited
    kaemia (APL) and in NBL. In APL, the value of all-​trans retinoic acid
    sampling models for childhood cancer drug development. Clin
    (ATRA) is mainly in reducing early death by quickly diminishing
    Cancer Res 9, 5068–​77.
    the bleeding tendency that is characteristic of this disease. Current Pearson AD, et al. (2016) Implementation of mechanism of action
    studies in APL are focusing on the combination of ATRA with ar- biology-​driven early drug development for children with cancer.
    senic trioxide, which showed improved results compared to com- Eur J Cancer 62, 124–​31.
    bined ATRA and chemotherapy regimens. Moreover, data from Schultz KR, et al. (2009) Improved early event-​free survival with
    the use of 13-​cis-​retinoic acid in NBL, given in a MRD setting, now imatinib in Philadelphia chromosome-​ positive acute lympho-
    shows clear evidence of survival benefit. blastic leukemia: a COG study. J Clin Oncol 27, 5175–​81.
    Another interesting compound is the recombinant human Van den Berg H, et al. (2012) Cytostatic drugs in infants: a review on
    arginase BCT-​100, designed to interfere with the arginine addiction pharmacokinetic data in infants. Cancer Treat Rev 38, 3–​26.
    of tumour cells, which will be studied in paediatric malignancies in Vassal G, et al. (2013) New drugs for children and adolescents with
    the near future. cancer: the need for novel development pathways. Lancet Oncol 14,
    Finally, compounds inhibiting proteins involved in apoptosis e117–​24.
    Vassal G, et al. (2013) Is the European pediatric medicine regulation
    regulation (BCL-​2, survivin) are also being explored in preclinical
    working for children and adolescents with cancer? Clin Cancer Res
    paediatric tumour models. There is so far no firm paediatric clinical
    19, 1315–​25.
    experience with these compounds, but a paediatric phase I/​II basket Von Stackelberg A, et al. (2016) Phase I/​phase II study of blinatumomab
    trial with the BCL-​2 inhibitor venetoclax is imminent. in pediatric patients with relapsed/​refractory acute lymphoblastic
    leukemia. J Clin Oncol 34, 4381–​9.
    Yu AL, et al. (2010) Anti-​GD2 antibody with GM-​CSF, interleukin-​2,
    FURTHER READING
    and isotretinoin for neuroblastoma. N Engl J Med 363, 1324–​34.
    Adamson PC, et al. (2014) Drug discovery in paediatric on- Zwaan CM, et al. (2013) Dasatinib in children and adolescents with
    cology: roadblocks to progress. Nat Rev Clin Oncol 11(12), 732–​9. relapsed or refractory leukemia: results of the CA180-​018 phase
    Cheymol G (2000) Effects of obesity on pharmacokinetics: implica- I dose-​escalation study of the Innovative Therapies for Children
    tions for drug therapy. Clin Pharmacokinet 39, 215–​31. With Cancer Consortium. J Clin Oncol 31, 2460–​8.
    4
    Radiotherapy in Paediatric Oncology
    Mark Gaze and Tom Boterberg

    Introduction suggested because things have not gone according to plan. A frank
    discussion should help to allay any misunderstandings about radio-
    Radiotherapy is a key component of treatment for many childhood therapy. Early familiarity with the team will reduce subsequent anx-
    cancers. The role is usually to improve local control, which may often ieties, and will help to prevent the problems which can arise after
    enhance the probability of survival, but in some cases additionally it delayed referral.
    is part of a curative strategy for metastatic disease. Radiotherapy may An important part of the initial assessment is to decide whether it
    also be used in the salvage of relapsed patients and for palliation. is likely that the child will be able to cooperate with the radiotherapy
    team, be able to tolerate the requirements of immobilization while
    awake, and feel comfortable being left alone in the treatment room.
    The patient’s pathway through radiotherapy Some time spent with a radiotherapy play specialist can transform
    an initially scared and uncooperative child into one who happily
    Treatment centres complies with treatment. If not, arrangements will have to be made
    Children should only receive radiotherapy in centres suitably for preparations and treatment under general anaesthesia.
    equipped, staffed, and experienced in paediatric radiotherapy, sup-
    Informed consent
    ported by the necessary paediatric backup facilities. This includes
    radiation oncologists with training and experience in paediatric It is important that families receive accurate and complete informa-
    radiotherapy fully integrated into the core paediatric oncology tion about the indications for, practicalities of, and likely early and
    multidisciplinary team, the availability of play specialists and paedi- late side effects of the radiotherapy required. This should be given
    atric anaesthetic facilities, and sufficient on-​site paediatric support verbally by members of the radiotherapy team, and be supplemented
    to deal with emergencies. This may mean that some children cannot by written information. Children should be given relevant informa-
    receive treatment in the radiotherapy centre closest to their homes. tion in a sensitive and age-​appropriate manner, which may include
    It will need to be explained to families that this inconvenience is comics or videos. Families speaking other languages should have
    offset by improved safety and quality. the services of an interpreter available at every interaction with the
    radiotherapy team. Children and their families should have an op-
    Decision making portunity for their questions to be answered in full and in an un-
    It is essential that radiation oncologists meet regularly with paedi- hurried manner. Often, if there are anxieties about radiotherapy, it
    atric haematology and oncology colleagues, radiologists, patholo- is recommended that families have more than one opportunity to
    gists, and surgeons to discuss the management of all patients in the discuss the information and ask questions. Finally, the receipt of in-
    light of full diagnostic and prognostic information. formed consent for radiotherapy treatment should be documented
    While the decision to treat with radiotherapy is often straightfor- according to local practice.
    ward, in many cases it is not simple, and discussion is vital. Children
    Positioning and immobilization
    with cancer should be treated where possible in appropriate clinical
    trials, and outside trials, radiotherapy should be given according to It is essential for the accurate and safe delivery of radiotherapy that
    approved guidelines. the patient does not move during imaging for planning, or during
    treatment. As indicated above, this may require the use of daily gen-
    Assessment eral anaesthesia. For young children who are to be treated awake,
    The initial consultation will help to make it clear that radiotherapy the input of an experienced play specialist in the preparation can
    is an important part of treatment, and confirm that the radiotherapy be indispensable. To minimize movement, an immobilization de-
    team is part of the overall team caring for the child. An early intro- vice may be necessary, for example a head and neck shell fixed to a
    duction to the radiotherapy team is important as it may prevent the base board. This will ensure accurate reproducibility of set-​up on a
    family from feeling that radiotherapy is an afterthought, or has been daily basis.
    32 CHAPTER 4 Radiotherapy in Paediatric Oncology

    Imaging and target volume definition is usually recommended. The doctor then approves the best plan,
    The art of radiotherapy is to ensure that all areas involved by tumour and the DVHs document the doses received as a guide to follow-​up.
    get an adequate radiation dose, while uninvolved normal tissues re- Prospective radiotherapy quality assurance, with review of treat-
    ceive as little radiation as possible, in order to minimize the risk of ment plans by other clinicians, is increasingly being introduced in
    adverse effects. To achieve this, a series of targets is defined on the clinical trials.
    basis of imaging which is interpreted in the light of the known path- Dose prescription and fractionation
    ology and likely patterns of spread, and natural barriers. Usually,
    prior to radiotherapy, a CT scan is performed in the radiotherapy The clinician will specify the total dose to the tumour and the
    department with the patient immobilized in the treatment position. number of fractions in the course (hence the dose per fraction), the
    Intravenous contrast may be required. The images obtained can be course duration, and the type of radiation and its energy. Typically in
    fused with other imaging, for example MRI or PET scans, which paediatric radiotherapy, a dose per fraction of 1.5 to 1.8 Gy is used,
    define the extent of tumour at critical times such as at diagnosis, fol- smaller than the 2 to 2.67 Gy fraction sizes often used in adult prac-
    lowing induction chemotherapy, or following surgery. Sometimes, if tice. This is because smaller fraction sizes allow better sparing of
    the tumour or normal organs may move with the respiratory cycle, a vulnerable normal tissue from late effects. Normally, treatments are
    movie loop with four dimensional CT (4DCT) is performed to allow given once each day, five days per week.
    the variable position to be demonstrated and taken into account There is evidence that prolongation of the overall treatment time
    accurately. may compromise outcome. Compensation should therefore be
    The first step in target volume definition is usually to define the made for any gap so that the overall treatment time does not exceed
    gross tumour volume (GTV) by contouring it, slice by slice, on that intended. The simplest and best ways of doing this are to treat
    the planning CT scan. Even if the tumour has shrunk as a result of on days which are normally days off (for example, treat on Saturdays
    chemotherapy, it may still be helpful to outline the original GTV around public holidays) or to give two treatments per day, with a
    on earlier scans. When this volume is then transferred to the radio- minimum inter-​fraction interval of six hours.
    therapy planning CT scan, it indicates the area at risk of recurrence. In some cases, hyperfractionated (a larger number of smaller
    The GTV will then be expanded into a clinical target volume (CTV). fractions than standard) and accelerated (given in a shorter than
    The CTV includes the original tumour volume plus a margin to en- standard overall time) schedules are used to reduce late effects and
    compass any subclinical extension of disease. The CTV margin may overcome repopulation in highly proliferative tumours. This re-
    be cut back to natural barriers of spread, such as the skull in the case quires more than one treatment a day, and the inter-​fraction interval
    of brain tumours, or if previously displaced uninvolved normal or- should be at least six hours, preferably eight hours for radical brain
    gans like the liver have returned to their normal position following treatments. In other cases, hypofractionation (a smaller number of
    treatment of the tumour with chemotherapy and surgery. The CTV larger than standard fractions) is used for stereotactically guided
    is then expanded by another margin to create the planning target treatment of small volumes or in palliation.
    volume (PTV), which takes into account errors and uncertainties in Verification of treatment accuracy
    planning and set-​up, and internal and external motion. The margin
    It is important that radiotherapy is delivered to the highest possible
    between CTV and PTV will depend on the anatomical site and ac-
    standards. Quality assurance procedures for radiotherapy delivery,
    curacy of immobilization as determined by departmental audit.
    particularly geometric variation for set-​up errors, are essential. This
    It is also necessary to outline organs at risk so that dosimetrists
    is usually done with electronic portal imaging or integrated (cone
    and planning systems can limit the dose received by healthy normal
    beam) CT scanning. Dose verification may be done with in vivo
    tissues, especially those that are most sensitive to adverse effects
    dosimetry, and is recommended as an additional quality assurance
    from radiation.
    measure.
    It is good practice, if possible, for volumes to be checked and ap-
    proved by another experienced clinician. This may be done within Supportive care
    one centre, or remotely as part of a clinical trial quality assurance
    The blood count should be checked before treatment. Irradiation of
    programme.
    anaemic children should be avoided, as anaemia is known to com-
    promise the outcome of radiotherapy. If necessary, a blood transfu-
    Planning and dosimetry sion should be given to bring the haemoglobin up to 120 g/​l. Nausea
    Dosimetrists, working under the guidance of the responsible doc- and vomiting can usually be prevented by the use of antiemetics.
    tors, will prepare treatment plans, which indicate how the treatment Radiotherapy may cause loss of appetite, and nutrition is important,
    will be delivered, and the doses received by the tumour and normal especially in children receiving daily general anaesthesia. The advice
    tissues. Plans should take into account the need to give homogenous of a paediatric dietician and nutritional supplements, nasogastric
    irradiation to the PTV while minimizing exposure to organs at risk. or gastrostomy feeding, or even total parenteral nutrition may be
    Dose volume histograms (DVHs) that show the planned dose to needed.
    the target volumes and relevant organs at risk should be produced. Radiotherapy can be stressful for children, their parents, and
    These allow the doctor to check that the target is receiving full ir- other family members. This can have both a psychological and prac-
    radiation, and that doses to organs at risk are acceptable. Sometimes tical basis. Psychological supportive care for the family can be very
    achievement of these two things is not practicable. A compromise beneficial. Travelling for radiotherapy, or staying away from home,
    is then needed with part or all of the PTV receiving less than the can be expensive. Availability of (free) family accommodation in a
    ideal radiotherapy dose and/​or organs at risk receiving more than nearby hotel or dedicated hostel for children and young people with
    Types of radiotherapy 33

    cancer is ideal. Families should have access to support from social or seven, of beam alignments which may be non-​co-​planar. A step-​
    workers or other appropriate healthcare professionals who can ad- and-​shoot approach entails movements of the linear accelerator
    vise on benefits and other practical issues. gantry and multi-​leaf collimator between the delivery of each treat-
    ment field. Alternatively, a sliding-​window technique is used, where
    Documentation the leaves move during delivery of one beam, instead of being placed
    It is important that there is a summary of radiotherapy treatment, at several fixed positions before irradiation starts, as in step-​and-​
    which is accessible to treating paediatric oncologists and is also shoot mode. A more sophisticated delivery, intensity-​modulated arc
    available in long-​term follow-​up. This should include adequate pa- therapy, involves dynamic gantry and multi-leaf collimator move-
    tient identifiers and sufficient details of the diagnosis and other prin- ments with the beam on, and is becoming more widely available.
    cipal treatments (i.e. chemotherapy and surgery) to allow it to be Whatever form of IMRT is used, the result is to better conform
    read in context. the shape of the high-​dose volume to the shape of the target, and to
    diminish the volume of healthy normal tissue that is incidentally in-
    Follow-​up cluded in the high-​dose volume. The price of this is usually low-​dose
    It is important that the patient is followed up in the long term with irradiation of a much larger volume of normal tissue than is the case
    regard to late effects as well as for local tumour control and survival. with conventional radiotherapy.
    Periodic assessments of relevant normal organ function should There have been concerns that the low-​dose bath effect seen with
    be systematically documented. Appropriate psychosocial support IMRT may result in a greater number of second cancers than has
    should be available as the child grows up and reaches maturity, to been the case with conventional radiotherapy. These anxieties have
    answer questions that might arise and to address proactively other sometimes limited the use of IMRT in children. However, IMRT
    survivorship issues. does allow full irradiation of the tumour in some circumstances
    where, with conventional radiotherapy, it has been necessary to
    compromise the dose to the tumour in order to respect normal
    Types of radiotherapy tissue tolerance. Also, it has been recognized that most second can-
    cers occur in or adjacent to the high-​dose volume (which is usually
    Conventional external beam radiotherapy significantly smaller with IMRT compared with conventional radio-
    The technical aspects of radiotherapy are evolving continuously, so therapy) rather than in the low-​dose area, so the hazards of using
    it is difficult to define what might be considered ‘conventional’ or IMRT in children may have been overstated. Furthermore, emerging
    standard therapy compared with what is historic and what is innova- long-​term follow-​up data have been reassuring. As always in paedi-
    tive. Advances in diagnostic imaging and computer science have atric radiotherapy, it is important to weigh up the likely benefits and
    largely led this progress. potential risks of any proposed treatment very carefully, and make a
    Nowadays, conventional radiotherapy usually implies the use of: balanced judgement of what is in the child’s best interests.

    • megavoltage photons (or sometimes electrons for superficial le- Stereotactic radiotherapy
    sions) from a linear accelerator; Stereotactic radiotherapy has been used for quite some time to treat
    • target volume definition based on images acquired on a CT small intracranial lesions. Accuracy of treatment of small volumes
    simulator; with narrow margins is achieved with the use of a fixed headframe
    • three-​dimensional planning with two or more shaped beams in (for a single treatment or radiosurgery) or relocatable headframe
    the same plane or in several planes to allow conformal treatment (for fractionated courses of treatment).
    of an irregular volume. More recently, interest has developed in stereotactic body radio-
    therapy for small lesions in, for example, the lungs or liver. As ex-
    Intensity-​modulated radiotherapy ternal frames would not be helpful, image guidance is used to ensure
    Intensity-​modulated radiotherapy (IMRT) is a way of changing the accuracy of treatment delivery. Hypofractionated or single treat-
    dose distribution of radiation in tissue by using beams where the ments are used, rather than fully fractionated courses of treatment.
    amount of energy varies in different parts of the beam. The aim is So far there has been only limited use of stereotactic body radio-
    to achieve the best possible dose distribution, especially where the therapy in children.
    target has a complex shape and there are adjacent critical structures.
    Image-​guided and adaptive radiotherapy
    This can be achieved in a number of ways of increasing complexity.
    The use of additional smaller segment fields to supplement a The term image-​ guided radiotherapy (IGRT) is potentially
    principal field, or field-​in-​field boost, is a relatively straightforward misleading, as target volume definition has been based on imaging
    method for forward-​planned IMRT to reduce non-​homogeneity of for years. IGRT refers instead to the use of imaging equipment inte-
    the delivered dose within a target volume. Inverse-​planned IMRT grated into treatment units to ensure accurate delivery of treatment.
    requires the clinical oncologist to define target volumes and organs The position of the target may be indicated by implanted fiducial
    at risk, as with conventional conformal radiotherapy, and to specify markers or by normal anatomy. Imaging, which may simply be or-
    the desired target dose and constraints on the dose which an organ thogonal radiographs or more sophisticated cone beam CT images,
    at risk may be allowed to receive. The planning computer used by is used with each fraction of treatment to ensure that the irradiated
    the dosimetrist then calculates the best field arrangement to deliver volume covers the target volume, which may move during (as with
    this. This usually uses a larger than conventional number, often five respiration) or between treatments.
    34 CHAPTER 4 Radiotherapy in Paediatric Oncology

    Adaptive radiotherapy implies the ability to change the irradiated a radionuclide within tumour cells. For example, in patients with
    volume during a course of treatment in response to a change in the well-​differentiated thyroid cancer of follicular cell origin, radioactive
    target volume. It is dependent on image guidance, and is so far little iodine (131I) can be biologically targeted into residual normal thyroid
    used in paediatric radiotherapy. Linear accelerators with integrated tissue and metastases. In children, lymph node and distant metas-
    MRI are being developed, which may in due course bring greater tases are more common than in adults with thyroid cancer, yet thanks
    accuracy to IGRT and adaptive radiotherapy. to the use of radioactive iodine this is a highly curable disease.
    In neuroblastoma, the catecholamine analogue meta-​
    Proton beam radiotherapy iodobenzylguanidine (mIBG) can be used for diagnostic scintig-
    This is a form of particle radiotherapy using high-​energy protons pro- raphy (123I-​mIBG) and for therapy (131I-​mIBG). Therapeutic mIBG
    duced by a cyclotron or synchro-​cyclotron. The advantage lies in the is recognized as an established palliative treatment, and it may also
    dose distribution that can be achieved with protons. Unlike photons have value if used intensively in patients with refractory disease as
    which are exponentially attenuated as they pass through tissue, protons part of a multi-​modality strategy for cure.
    travel for a finite distance, and deposit most of their energy towards the Additional types of molecular radiotherapy are being evaluated
    end of their path, the Bragg peak. This means that it is possible to spare, in clinical trials.
    theoretically completely, normal tissues beyond the target.
    In the treatment of relatively radio-​resistant well-​defined skull
    base tumours, cure rates may be increased as higher doses can be Side effects of treatment
    delivered to the target with effective sparing of adjacent critical
    structures like the brainstem. In other circumstances, the dose Factors affecting radiation injury to normal tissues
    distributions that can be achieved with protons are advantageous
    because the dose to normal tissues for a given target dose can be re- The incidence and severity of adverse reactions to radiotherapy
    duced, thereby limiting normal tissue toxicity. among children treated for cancer are related to several factors,
    Although proton beam therapy has a long history, there is very knowledge of which allows new strategies to be devised that may
    little level-​one evidence of benefit in the patient’s outcome in com- reduce adverse effects without compromising the chances of cure.
    parison with photon radiotherapy. Most of the case for the use of These factors include whether or not radiotherapy is used at all,
    proton beam therapy has been made from theoretical predictions. the total dose used, the dose per fraction, the volume and type of
    However, with the increasing use of proton radiotherapy in selected tissue irradiated, the type of radiation used, the use of concomitant
    patients, more outcome data (in relation to both disease control and chemotherapy, and, not least, the age of the patient at the time of
    late effects) are emerging, and support its use, although also some treatment.
    unexpected toxicity has been described (e.g. brainstem necrosis
    Early side effects
    in patients treated for CNS tumours). Although the availability of
    proton beam therapy has been limited, an increasing number of Early or acute side effects occur during, or in the first weeks after,
    centres are being opened, and soon it should be available in most radiotherapy. They are usually the result of the depletion of stem
    large developed nations. As with photon-​based radiotherapy, tech- cells in the healthy tissue surrounding the tumour area to be irradi-
    nical advances in proton beam therapy equipment, including spot ated. In most cases they subside with time or supportive therapy
    scanning (as opposed to passive scattering) and improved image only, although some of them may be life-​threatening and require
    guidance, have allowed the development of intensity-​modulated early interruption of therapy.
    proton therapy, which should be more advantageous. The most common early side effects are skin reactions (like ery-
    thema and alopecia) and mucositis (like stomatitis, oesophagitis,
    Brachytherapy cystitis, and diarrhoea), all depending of the site of the body being
    Brachytherapy, with direct implantation of radioactive sources, was treated. Nausea and vomiting may occur in case of abdominal irradi-
    a valuable form of treatment decades ago. Interest waned with the ation. Fatigue is a general and frequently encountered early effect
    introduction of the linear accelerator, which made treatment of which may take several months to resolve. Many of these side effects
    deeply seated tumours much easier, with the increasing awareness can also be aggravated by the concurrent use of chemotherapy, es-
    of radiation protection hazards, and the difficulties of achieving a pecially actinomycin and anthracyclines. Subsequent chemotherapy
    perfect geometrical implant. With the development of machines for may even mimic the cutaneous side effects of radiotherapy at sites
    remote afterloading of implanted catheters, the radiation protection where radiation previously entered the body, the so-​called ‘radiation
    risks have largely been overcome. The ability to use CT-​and MRI-​ recall’ reaction, requiring appropriate care and explanation to the
    based planning with varying position and dwell times for the source parents.
    has reduced the need for perfect geometry. These technical advances
    have led to a resurgence of interest in brachytherapy in paediatric Late side effects
    oncology. In some childhood cancers, brachytherapy allows con- Late side effects are defined as toxicity occurring or persisting from
    formal irradiation of small target volumes with greater sparing of three months onwards after radiotherapy. They are usually irrevers-
    normal tissues than is possible with external beam radiotherapy. ible and sometimes worsen with time. Therefore, all care should be
    given to prevent this type of side effect. On the other hand, some
    Molecular radiotherapy of these side effects are very difficult to prevent and should be con-
    This is essentially the use of radiotherapy as a drug. It is possible sidered as ‘the price to be paid’ for cure. Finding the right balance is
    where a metabolic pathway can be used to allow accumulation of a huge challenge for doctors, patients, and their parents.
    Side effects of treatment 35

    Late central nervous system side effects kidney. In patients who have undergone nephrectomy, special atten-
    Although the number of neurons usually does not increase after tion is needed to spare the contralateral kidney as much as possible,
    birth, the brain matures by the development and branching of axons not only to preserve its function per se, but also not to hinder the
    and the formation of new synapses. This process mostly takes place compensatory hypertrophy.
    within the first three years of life and is almost complete by the age Late musculoskeletal and soft-​tissue side effects
    of six, although it continues at a much lower rate till the end of pu-
    berty. Radiotherapy may affect this process by killing astrocytes and In adults, muscles and bones are very radio-​resistant. In children,
    oligodendrocytes, thus causing demyelinization. Radiotherapy also muscles and bones are in full development and these dividing tissues
    affects vascular endothelial cells. Mostly these effects are irrevers- are therefore very sensitive to radiation, especially the epiphyseal
    ible, although sometimes a transient somnolence syndrome is seen. growth plates. A dose around 12–​15 Gy will impair the chondroblast
    This is characterized by drowsiness, apathy, and irritability, and usu- function and have an adverse effect on bone growth. It is clear that
    ally occurs one to two months after the end of treatment. It is self-​ there is not a threshold dose, and the degree of growth impairment is
    limiting within 12 weeks of its onset. related to the age of the child, the volume of bone irradiated, and the
    Even at rather low doses, whole brain irradiation may lead to dose given. Avascular necrosis is also possible, and the risk is higher
    functional disturbances in the brain, although no obvious struc- if concurrent steroids are to be given. The effect on growth is usually
    tural changes can be observed. The most common side effects are greatest following the onset of puberty, when the growth spurt starts.
    neurocognitive deficits: lower IQ, worse short-​term memory, lower Disproportion between body parts may not only have functional im-
    speed of processing new information, and concentration difficulties. plications but also leads to psychological and social problems.
    In medulloblastoma, safe replacement of craniospinal irradiation by Spinal irradiation will result in shorter height, but may also be
    chemotherapy has proved challenging. Despite its risk for late cen- accompanied by deformities. Even if all care has been taken to uni-
    tral nervous systems (CNS) side effects, craniospinal irradiation is formly irradiate the involved vertebrae, flank surgery and radiation-​
    still standard treatment for older children, and parents should be induced hypoplasia of the unilateral paraspinal and abdominal
    well informed about this. Chemotherapy can be used to allow post- muscles may contribute to the development of scoliosis. Partial
    poning radiotherapy until the child is of a more suitable age to re- irradiation of the head and neck region can lead to facial bone
    ceive radiotherapy to the brain. Proton beam therapy is of particular hypoplasia and consequent asymmetry of the face. When growth
    value in treating small localized brain tumours, with less irradiation is complete, plastic surgery may sometimes be used to improve
    of non-​target healthy brain tissue than is possible with photon tech- cosmetic appearances. Radiation-​induced impaired dental devel-
    niques. However, it should be noted that the neurocognitive deficits opment may be aggravated by an increased risk of dental caries if
    are usually limited if small volumes are treated, even at high doses xerostomia develops after irradiation of the salivary glands higher
    and in small children. than 25–​30 Gy.
    High doses or re-​irradiation may lead to brain necrosis. This Thoracic irradiation including the breast bud, or the developing
    may develop years after treatment and may even mimic tumour re- breast in pubertal girls, may lead to hypoplasia. Plastic surgery for
    currence. Steroids are commonly used for treatment, and in some breast augmentation to restore symmetry may be considered once
    cases resection may be necessary to stop the process. Concurrent growth is complete. Irradiation of the developing breast raises the
    radiosensitizing chemotherapy increases the risk of developing risk of breast cancer developing subsequently. Patients at risk should
    brain necrosis. Necrosis of the spinal cord is a very serious complica- have screening performed as part of their long-​term follow-​up from
    tion leading to irreversible paralysis. Fortunately it is only very rarely an earlier age than is standard for breast cancer screening.
    encountered as great care is taken to keep the dose to the spinal cord Late pulmonary side effects
    within safe limits.
    Radiotherapy may damage blood vessels, and this in turn can lead Radiation pneumonitis usually develops between three and six
    to neurotoxicity. Patients with neurofibromatosis type 1 are at in- months after radiotherapy. Cough and dyspnoea are the most
    creased risk. Moya moya may lead to strokes. Cavernous haemangi- common symptoms, and diffuse infiltrates can be seen on imaging. If
    omas or cavernomas are being recognized increasingly commonly no spontaneous remission occurs, steroid treatment may be needed.
    after brain radiotherapy. Usually these are incidental findings on In some cases, pneumonitis will lead to lung fibrosis, although fi-
    surveillance MRI scans. Sometimes a cavernoma may bleed into the brosis can also be seen without preceding pneumonitis. Whole-​lung
    brain. This may be either a minor bleed without overt brain damage irradiation to doses of 15 Gy is considered safe and without clinic-
    or, rarely, a catastrophic haemorrhage with resulting neurodisability ally important consequences, although lung function tests usually
    or death. reveal some degree of restrictive disease at adult age. Bleomycin,
    actinomycin, and busulfan are radiosensitizing chemotherapeutic
    Late renal side effects agents and should be avoided or only used with great care in com-
    bination with radiotherapy on the lungs. Anaesthetists should be
    Acute renal side effects are usually rare, but long-​term effects may aware of oxygen toxicity in patients treated with bleomycin and
    be important and include chronic renal failure, hypertension, and radiotherapy. Obviously, smoking should be firmly discouraged, but
    anaemia. The development of these conditions is insidious and especially following lung radiotherapy.
    takes several years after treatment. Careful follow-​up of patients
    who received abdominal radiotherapy is therefore very important. Late cardiac side effects
    Concurrent chemotherapy like cisplatin or previous surgery to the Severe cardiac damage, including cardiomyopathy, valvular disease,
    kidney may necessitate lowering the dose, at least to part of the and myocardial infarction, has been seen in the past, especially in
    36 CHAPTER 4 Radiotherapy in Paediatric Oncology

    patients treated for Hodgkin lymphoma. However, it should be noted with a low birthweight, due to the uterus or cervix being partially
    that these patients had large volumes (often nearly the whole heart) less well developed. Sometimes treatment does not lead to complete
    treated with high doses, up to 45 Gy or higher. Currently, the use of infertility, but female survivors of childhood cancer may have their
    radiotherapy has been reduced, as have volume and dose. Nowadays, reproductive period reduced by premature ovarian failure and an
    the main reason for cardiac toxicity is probably not the dose to the early menopause. This should be discussed with young women at-
    ventricles (which may, in part, tolerate doses of 50 Gy or more) but tending long-​term follow-​up clinics.
    to the valves and coronary arteries. Even much lower doses may still
    affect subsequent cardiac function, and the mean heart dose should Late ocular side effects
    be kept as low as reasonably achieveable. Following incidental cardiac The radiosensitivity of the eye depends on which part is irradiated.
    irradiation, patients should be encouraged to have a healthy life style The retina and optic pathways are relatively resistant and may tol-
    with enough physical exercise, regular blood pressure checks, and erate doses of 45 to 54 Gy without major toxicity. Dose constraints
    without smoking or excessive intake of fatty food. Finally, the use of of 45 Gy for the retina, 50 Gy for the optic nerves, and 54 Gy for
    anthracyclines also plays a role in the development of cardiomyopathy. the optic chiasm, are often specified. Above 54 Gy, optic neuritis
    and permanent blindness may develop. By contrast, the lens is very
    Late endocrine side effects radiosensitive and cataract formation may occur after doses as low as
    Most endocrine deficiencies result from irradiation of the 2–​6 Gy. However, lens replacement is an effective and safe treatment
    hypothalamic–​hypophyseal axis. They may develop very insidiously for these patients. Irradiation of the lacrimal gland may result in dry
    and years after radiotherapy, so close follow-​up is necessary. Doses eye syndrome, a very painful condition that may require enucleation.
    as low as 18 Gy can be sufficient to impair growth hormone secre- The dose constraints are not very clear: 40 Gy is considered as accept-
    tion after a period of 18 months or two years. Above 36 Gy, growth able, although less than 26 Gy to 50% of the volume is probably safer.
    hormone secretion may be totally stopped; and at doses above 40
    Gy, other hormones like adrenocorticotropic hormone (ACTH), Late hepatic side effects
    thyrotropin-​releasing hormone (TRH), and the gonadotrophins The tolerance of the liver for hepatitis and liver failure due to at-
    may be affected. Hyperprolactinaemia and early puberty may also rophy and the development of fibrosis depends on the total dose,
    develop. Hormone replacement therapy should be started as soon the volume, concurrent chemotherapy, and previous surgery.
    as it is judged safe. A mean dose of 30 Gy to the entire liver can be safe if part of it (at
    Radiotherapy above about 20 Gy to the thyroid gland increases least 25%) receives a dose of 10 Gy or less and no previous surgery
    the risk of primary hypothyroidism. In patients treated with doses was performed or concurrent chemotherapy is needed. After par-
    around 40 Gy, the risk of developing hypothyroidism is as high tial hepatectomy, irradiation of the regenerating liver should be
    as 50%. Even compensated hypothyroidism, where the thyroid-​ performed with great care, excluding as much liver as possible.
    stimulating hormone (TSH) is elevated but the T3 and T4 are within Combined with actinomycin, the upper dose limit may need re-
    the normal range, should be treated with thyroid hormone replace- duction to 15 Gy. Total body irradiation (TBI), or other hepatic
    ment therapy to reduce the risk of a raised TSH level over a pro- radiotherapy combined with chemotherapy, may result in veno-​
    longed period, stimulating the development of thyroid cancer. occlusive disease (VOD), currently usually called hepatic sinus-
    oidal obstructive syndrome (SOS). Caution should be exercised
    Late side effects on fertility in irradiating the liver if there is pre-​existing hepatic dysfunction.
    Spermatogenesis is especially radiosensitive. A dose as low as 15
    cGy (0.15 Gy) is enough to result in a reversible reduction of the Metabolic syndrome
    sperm count, and permanent sterilization has been observed at cu- A more recently recognized long-​term side effect of especially ab-
    mulative doses of 1–​2 Gy. Small fractionated doses seem to be more dominal radiotherapy is the metabolic syndrome. It is characterized
    toxic than a single fraction dose. Therefore adequate shielding of by higher blood pressure, increased levels of triglycerides, choles-
    the testicles from radiation, for example during pelvic radiotherapy, terol, and free fatty acids, and increased adiposity. Prevention being
    should be undertaken if possible. A dose above 10 Gy usually re- difficult, adequate follow-​up for early detection and treatment are
    sults in permanent azoospermia. If possible and indicated, semen important countermeasures.
    cryopreservation or testicle biopsy cryopreservation should be or-
    ganized in (post-​) pubertal boys before any radiotherapy starts. Second cancers
    Testosterone production by the Leydig cells is usually not affected All patients who have survived cancer are at risk of developing a
    by doses below 20 Gy. second cancer. This may be due to their genetic background, epi-
    In girls, permanent sterility can be seen at an ovarian dose of 8 genetic risk factors, and, last but not least, previous treatment with
    Gy or higher. Hormonal deficit develops from 12 Gy onwards and chemotherapy and radiotherapy. All of these factors may influence
    is usually permanent above 20 Gy. Until some years ago, avoiding each other, making it even more complicated to assess that risk and
    irradiation of the ovaries or oophoropexy (surgical relocation of the to counsel patients with regard to secondary prevention. Patients
    ovary outside the irradiated volume) were the only possibilities to with DNA repair deficiencies like xeroderma pigmentosum are par-
    protect the ovarian function in prepubertal girls who needed radio- ticularly sensitive to radiotherapy and the development of second
    therapy to the pelvic region. Partial ovarian cryopreservation and tumours after radiotherapy. Girls treated at or just before puberty for
    follicle cryopreservation are new techniques that may allow pres- mediastinal Hodgkin lymphoma are at higher risk to develop breast
    ervation of fertility. However, even partial irradiation of the uterus cancer and should therefore be screened earlier than the general
    results in a higher risk of intra-​uterine death, stillbirth, or a child population. Basal cell carcinomas are often seen on irradiated skin.
    Side effects of treatment 37

    Fossati P, et al. (2009) Pediatric medulloblastoma: toxicity of current


    treatment and potential role of protontherapy. Cancer Treat Rev
    External Contour 35, 79–​96.
    Planning Target Volume Gains JE, et al. (2013) Intensity-​modulated arc therapy to improve ra-
    Clinical Target Volume diation dose delivery in the treatment of abdominal neuroblastoma.
    Gross Tumour Volume Future Oncol 9, 439–​49.
    Gan HW, et al. (2015) Neuroendocrine morbidity after pediatric optic
    gliomas: a longitudinal analysis of 166 children over 30 years. J Clin
    Endocrinol Metab 100, 3787–​99.
    Haas-​Kogan D, et al. (2018) National Cancer Institute Workshop on
    Proton Therapy for Children: considerations regarding brainstem
    injury. Int J Radiat Oncol Biol Phys 101(1), 152–​68.
    Figure 4.1 Stylised axial CT slice of body to show the position of the Hodgson DC, et al. (2007) Pediatric Hodgkin lymphoma: maximizing
    gross tumour volume (GTV), expanded to cover areas of subclinical efficacy and minimizing toxicity. Semin Radiat Oncol 17, 230–​42.
    potential spread forming the clinical target volume (CTV), and expanded Indelicato DJ, et al. (2008) Definitive radiotherapy for Ewing tumors
    again to form the planning target volume, allowing for inaccuracies and of extremities and pelvis: long-​term disease control, limb function,
    uncertainties in set-up. and treatment toxicity. Int J Radiat Oncol Biol Phys 72, 871–​7.
    Ness KK, et al. (2010) Physical performance limitations among adult
    survivors of childhood brain tumors. Cancer 116, 3034–​44.
    Radiotherapy to the thyroid may not only result in hypothyroidism Puri DR, et al. (2006) The challenging role of radiation therapy for very
    but also in the development of thyroid cancer. young children with rhabdomyosarcoma. Int J Radiat Oncol Biol
    Phys 65, 1177–​84.
    Compromising the chance of survival by omitting radiotherapy
    Schoot RA, et al. (2015). Adverse events of local treatment in long-​
    in the hope of minimizing late effects, including second malignant
    term head and neck rhabdomyosarcoma survivors after external
    neoplasms, is not an attractive option. Therefore, patients should re-
    beam radiotherapy or AMORE treatment. Eur J Cancer 51,
    ceive radiotherapy where indicated. The knowledge that they are at 1424–​34.
    risk of developing a second tumour or other late effects justifies their Selo N, et al. (2010) Acute toxicity profile of radiotherapy in 690 chil-
    life-​long follow-​up by a multidisciplinary team. dren and adolescents: RiSK data. Radiother Oncol 97, 119–​26.
    The Royal College of Radiologists (2018) Good Practice Guide for
    Paediatric Radiotherapy, second edition. London: The Royal
    FURTHER READING College of Radiologists.
    Alvarez JA, et al. (2007) Long-​term effects of treatments for childhood Timmermann B, et al. (2007) Spot-​scanning proton therapy for malig-
    cancers. Curr Opin Pediatr 19, 23–​31. nant soft tissue tumors in childhood: first experiences at the Paul
    Bhandare N, et al. (2008) Hypopituitarism after radiotherapy for extra- Scherrer Institute. Int J Radiat Oncol Biol Phys 67, 497–​504.
    cranial head and neck cancers in pediatric patients. Am J Clin Oncol van Waas M, et al (2012) Abdominal radiotherapy: a major deter-
    31, 567–​72. minant of metabolic syndrome in nephroblastoma and neuro-
    Bölling T, et al. (2008) Late effects of thoracic irradiation in children. blastoma survivors. PLoS ONE 7(12), e52237.
    Strahlenther Onkol 184, 289–​95. Vivekanandan S, et al. (2015) The UK experience of a treatment strategy
    Bölling T, et al. (2010) Dose-​volume analysis of radiation nephrop- for pediatric metastatic medulloblastoma comprising intensive in-
    athy in children: preliminary report of the RiSK Consortium. Int J duction chemotherapy, hyperfractionated accelerated radiotherapy
    Radiat Oncol Biol Phys 80, 840–​4. and response directed high dose myeloablative chemotherapy or
    Boterberg et al. (Eds). (2020) Radiotherapy and cancers of children maintenance chemotherapy (Milan strategy). Pediatr Blood Cancer
    teenagers and young adults. Oxford: Oxford University Press. 62, 2132–​9.
    Carver JR, et al. (2007) ASCO Cancer Survivorship Expert Panel. Waber DP, et al. (2007) Neuropsychological outcomes from a ran-
    American Society of Clinical Oncology clinical evidence review on domized trial of triple intrathecal chemotherapy compared with
    the ongoing care of adult cancer survivors: cardiac and pulmonary 18 Gy cranial radiation as CNS treatment in acute lymphoblastic
    late effects. J Clin Oncol 25, 3991–​4008. leukemia: findings from Dana-​ Farber Cancer Institute ALL
    Castellino S, et al. (2010) Hepato-​biliary late effects in survivors of Consortium Protocol 95–​01. J Clin Oncol 25, 4914–​21.
    childhood and adolescent cancer: a report from the Children’s Wright KD, et al. (2009) Late effects of treatment for Wilms tumor.
    Oncology Group. Pediatr Blood Cancer 54, 663–​9. Pediatr Hematol Oncol 26, 407–​13.
    5
    Clinical Trials in Childhood Cancer
    Maria Grazia Valsecchi, Stefania Galimberti, and Pamela Kearns

    Introduction This chapter will review how the conventional approaches to reli-
    ably defining the benefit of a new drug or a new treatment strategy
    Childhood cancer was regarded as almost universally fatal until must be adapted.
    the first clinical trials performed by Sidney Farber in the late 1940s
    showed that it was possible to induce temporary remissions in
    childhood leukaemia. Over the ensuing decades, clinicians have Design and implementation of clinical trials
    organized themselves into national and then international co-
    operative groups, with the aim of conducting clinical trials to im- For a clinical trial to have the best chance of success, it must ask a
    prove outcomes in these rare diseases. Such collaborative efforts clear question that is seen as important both to the clinicians who
    have optimized the dose intensity and combinations of multi-​agent will implement the trial and the patients who will participate. For
    conventional chemotherapy, and have helped define the role of these reasons, early engagement with families affected by childhood
    surgery and radiotherapy to achieve maximum overall survival. cancer is key: if a trial question cannot be explained easily to those
    This means that in the early twenty-​first century, approximately who represent the patient group, then it will be even more difficult to
    80% of children newly diagnosed with cancer, who are treated in be understood by parents, especially at the distressing time of initial
    countries where participation in clinical trials is widely available, diagnosis of cancer in their child. The issue of study design is equally
    can expect to be long-​term survivors. However, these high overall crucial and, quoting Sir David R. Cox (2007): ‘A seriously defective
    survival rates present challenges for future trial design, to secure design may be incapable of rescue even by the most ingenious of
    further incremental improvements in survival or to demonstrate analyses; a good design may lead to conclusions so clear that simple
    survival equivalence from the promising new agents that may have analyses are enough.’ The design of a study basically aims at an-
    reduced side effects. swering, reliably, the research question by both minimizing biases
    Recent advances in understanding cancer biology have defined (i.e. systematic errors that may deviate the estimate of interest from
    even smaller subgroups of childhood cancers according to molecular its ‘true’ unknown value) and by maximizing precision (i.e. minim-
    signatures. This means that international recruitment is essential to izing variability of estimates, through control of random errors with
    conduct trials within a reasonable time frame. Such collaboration appropriate sample size).
    across linguistic and cultural boundaries presents not only legal
    Background definitions
    and regulatory hurdles but also challenges the childhood cancer
    research community to reappraise their individual therapeutic Clinical trials are traditionally classified into four phases, from Phase
    preferences. Paediatric regulations in North America and Europe I to Phase IV, following the stages of pharmacological development
    should encourage manufacturers of new anti-​cancer drugs to sup- of new drugs for human use. Phase I is the earliest stage, carried out
    port the whole process, possibly with an adequate structure of incen- to investigate pharmacokinetics and safety of the new product, so as
    tives. In the last decade, since the introduction of the EU Paediatric to establish the most tolerable dose. Phase II trials study the safety
    Regulation (EC) 1901/​2006 and Amendment Regulation (EC) 1902/​ and the activity profiles on a larger scale. They may define disease
    2006, very few new compounds among those available for adult can- types or subgroups with better response rates and, hence, lead to
    cers have completed a paediatric investigation plan (PIP) or have Phase III trials, in which the new treatment is compared to an alter-
    reached marketing authorization for children. It should be ensured, native (an established ‘standard of care’ or placebo). Here, the aim
    in particular, that a PIP is based on how a drug works and its capacity is to demonstrate improved efficacy or increased safety with equal
    to address unmet medical needs in children and adolescents, rather efficacy. Phase IV trials are performed, usually on an observational
    than on the type of disease in adults for which it was introduced. basis, after the new intervention has received regulatory approval,
    Mechanisms to prioritize drug development in relation to real needs and aim at obtaining additional information about risks (e.g. rare
    of children and to make a more efficient use of trial methodology but serious side effects, long-​term side effects), benefits (long-​term
    are required. outcomes), and optimal use. Of note, in these definitions, the new
    Design and implementation of clinical trials 39

    product’s beneficial property is either referred to as activity or effi- to children with cancer when there is already some knowledge of
    cacy, with two different meanings. Activity relates to the biological the dose tolerated by adults and the expected side effects. This re-
    activity and to disease control, while efficacy relates to an outcome duces the number of children who may otherwise be exposed to a
    measure that unequivocally reflects tangible benefits to patients. potentially subtherapeutic dose, and gives clinicians and families
    Examples are the impact of a drug/​therapeutic strategy on response, some reassurance about what toxicities to expect. However, there
    either morphological or molecular (activity), as compared to the im- may be unpredictable side effects in children, especially because the
    pact on event-​free survival or survival (efficacy). side effects may be related to growth, and there is also a need for
    This classification into Phase I–​IV trials has recently been criticized caution in very young infants/​neonates, whose physiology is more
    because it does not provide a general terminology able to accommo- immature and therefore may handle drugs differently in terms of
    date the development of medical products that do not necessarily fit pharmacokinetics. The second point that motivates studies in chil-
    that of traditional pharmaceuticals (e.g. biologically targeted agents). dren is that with the evolution of more targeted therapies, there are
    A more flexible and general approach, that classifies studies on the now examples where the target is unique to paediatric cancers (such
    basis of their principal aim, is as follows: (a) early-​development as GD2 in neuroblastoma).
    studies (translational or treatment-​ mechanism testing studies), A conventional phase I trial determines the maximum tolerated
    (b) middle-​development studies (assessing treatment tolerability), dose (MTD) of a drug in the patient group under investigation, and
    and (c) late-​development studies (including comparative studies as gives a preliminary indication of the toxicity profile and whether the
    well as expanded safety or post-​marketing studies). The traditional acute phase is reversible. In adults, blood sampling for pharmaco-
    classification is easily recognized within this new ‘descriptive’ one. kinetic analysis is mandatory, in order to determine if the dose and
    Early-​development studies include translational trials and the schedule achieves sustained blood levels of the active agent equiva-
    name reflects their purpose—​connecting laboratory findings to lent to those required for activity in preclinical models. A phase
    clinic and vice versa. A more traditional early-​phase study is the one I study defines a starting dose and a modality of dose escalation
    focused on understanding the treatment mechanism and the dose–​ based on the assumption that the greater the dose, the greater the
    safety profile, in a context guided by a biological model. At this stage, likely anti-​tumour activity. The MTD is obtained as the maximum
    biomarkers that are associated with the desired effect on the bio- dose at which the target toxicity level (TTL) defined as acceptable is
    logical target should be developed (pharmacodynamic studies). not exceeded.
    Assessment of the clinical outcome of the new treatment is formally Phase I designs can be broadly classified as ‘rule-​based’ or ‘model-​
    introduced in middle-​development studies, where its relationship to based’. Among the ‘rule-​based’, the conventional ‘3 + 3’ design is still
    safety and activity is studied. This stage is particularly important when most commonly used: three patients are entered at each dose level
    various new compounds/​interventions are available and it is neces- and if none of the patients experience the type of toxicity defined to
    sary to select out the more promising ones. Phase II trials are a typ- be relevant (the so-​called dose-​limiting toxicity, DLT) (i.e. grade 3
    ical example. Given the developmental purpose, they usually involve or 4 toxicity of given type), the dose is escalated to the next prede-
    a broader definition of activity, for instance utilizing biomarkers (e.g. termined level. If one patient experiences a DLT (corresponding to
    tumour or host genotype or marker in peripheral blood lymphocytes a fixed 33% TTL), then the cohort is expanded to six patients. This
    associated with target inhibition) or an early response (e.g. complete cycle continues until two patients (out of three or six) experience
    remission rate in leukaemia) to better identify the patient population DLT; the MTD is then specified as the previously tested dose level.
    who may benefit. In specific circumstances, middle-​development Sometimes, a predetermined maximum feasible dose is specified,
    studies can comprise a randomized question, when there are two or based on the pharmaceutical properties of the agent.
    more new promising treatments and the primary objective is to select Variations to the ‘3 + 3’ design have been proposed to accelerate
    the one to be considered for a successive comparative trial. patient recruitment and reduce the time periods for which a trial is
    Comparative trials are the most important among late-​development suspended for observation for potential toxicity. The ‘rolling 6’ de-
    studies and are focused on questions of comparison of treatment effi- sign permits continued enrolment of up to six patients at each dose
    cacy (and as such they correspond to classical Phase III trials). These level, pending toxicity evaluation of the first three patients enrolled.
    represent the gold standard for gathering evidence in medical research Among the ‘model-​based’ designs, the continual reassessment
    as they are designed to minimize bias through control over treatment method (first introduced in 1990) adopts a Bayesian approach (see
    assignment, systematic errors, and ascertainment of endpoints, and are the section ‘Novel strategies for clinical trial design’) with a para-
    completed by predefined analysis. Replication of trials and consolida- metric dose–​response curve (e.g. logistic) which is updated after
    tion of their results with meta-​analyses further increase the credibility each patient (or cohort) is enrolled at the current best estimate of the
    of conclusions. A key feature in comparative trials is equipoise, defined MTD using all information available after the first DLT is observed
    as the state in which any rational, informed person (patient, clinician, and avoiding waiting between subsequent cohorts. This design may
    or researcher) does not favour a treatment over the other(s) being be useful in shortening the dose determination phase but tends to
    studied. This condition of genuine uncertainty is fundamental in clin- expose more patients at MTD and higher doses. To reduce this risk,
    ical research comprising a randomization procedure. a methodological adaptation (escalation with overdose control) has
    been proposed. Other methodological adaptations have been de-
    Overview of clinical studies in childhood cancer veloped to include ‘time to event’ continual reassessment method
    (CRM) in modelling the dose–​toxicity relationship and summary
    Phase I trials toxicity scores that can take account of a multiplicity of organ
    Phase I trials are nearly always ‘first in child’ studies rather than toxicities and moderate-​grade events, all of which may be important
    ‘first in man’. In other words, a new anti-​cancer drug is only offered considerations in testing new drugs in young children.
    40 CHAPTER 5 Clinical Trials in Childhood Cancer

    Interval-​
    based designs have been recently proposed as a new subpopulations of childhood leukaemias would theoretically benefit
    option. The modified toxicity probability interval (mTPI) and the from tyrosine kinase inhibitors for BCR-​ABL + ALL or FLT3 inhibi-
    Bayesian optimal interval (BOIN) define an interval of accept- tors in acute myeloid leukaemia (AML) (or infants with ALL) whose
    able toxicity around the target DLT probability and two additional leukaemic cells have FLT3 internal tandem duplications. These ther-
    subintervals of low and high toxicity rates. Dose adaptation is made apies open a different scenario as the DLT approach is generally not
    on the posterior probabilities for the three subintervals using a appropriate and because the clinical development may be aimed pri-
    Bayesian model and observed data at each dose level in the mTPI de- marily at assessing the combination of the target agent with standard
    sign and on the observed toxicity rate in the BOIN design. This latter agents. When a phase II trial does not use the new drug as a single
    is a Bayesian design in that it selects optimal interval boundaries agent, activity data become difficult to interpret because there might
    minimizing the decision errors of dose assignment under a Bayesian be an unquantifiable response due to the underlying therapy. Even
    framework. Interval-​based designs, that are a mixture of rule-​based if historical data are available on the adopted underlying standard
    and model-​based designs, are attractive as they are flexible, well-​ therapy, these may not be fully appropriate for the patient mix on
    performing (they show operative characteristics that compete with which the phase II study is conducted. These considerations stimu-
    CRM), and represent a good compromise between 3+3 and CRM. lated the development of randomized phase II designs such as the
    screening design.
    Phase II trials Randomized phase II (or phase IIB) studies are smaller than
    A phase II trial is designed to obtain an initial estimate of the anti-​ phase III studies, as they look for large differences and relax the type
    tumour activity of an agent. The purpose is to determine, with a rea- I and II error rates, and they tend to be quicker, as they assess activity
    sonable degree of confidence, whether the new anti-​cancer agent has (or early intermediate endpoints) rather than efficacy. They may be
    sufficient activity against a specified tumour type as to warrant its of value in high-​risk patient populations (i.e. patients whose disease
    further development or evaluation in that tumour type. A further behaves aggressively and there is a high event rate in terms of disease
    consideration is to minimize the number of patients treated with a progression), but also in good-​prognosis diseases where random-
    drug of low activity. ization is used to have an unbiased evaluation with an internal con-
    The classical study design (phase IIA) comprises a single-​arm, trol. For these reasons they should not compromise the conduct of
    multi-​stage recruitment plan with the endpoint of anti-​tumour ac- a definitive phase III study on the promising new drug combination
    tivity generally based on categorical response criteria (complete and on the most appropriate patient population, which may be newly
    partial response). The presence of multiple stages, usually two, en- diagnosed patients. The definition of the endpoint is here a key issue
    sures that the study continues to the second stage of recruitment and, recently, a lot of attention has been stimulated on adopting early
    only if in the first stage, on fewer patients, a minimal response rate activity endpoints that could be regarded as surrogate biomarkers
    is reached according to a predefined statistical design. Such a design for the clinical outcome of efficacy that requires longer-​term evalu-
    presents practical problems in childhood cancer studies. First, the ation (event-​free survival or survival). Surrogacy requires a proper
    endpoint of response may take several weeks to manifest, leading to validation; without that a dangerous claim could be made on the
    periods of suspension of recruitment between the first and second long-​term advantage of a given treatment based on the early advan-
    stage, which can present difficulties for clinicians and patients tage seen on the biomarker. The debate is open on minimal residual
    waiting to be enrolled. Second, the fortunate reality is that there are disease (MRD) as a candidate, but evidence on its surrogate value is
    very few available patients who are eligible for early phase trials in not available.
    childhood cancer, due to the effectiveness of current therapies for Seamless phase I/​II trials can be designed which streamline the
    most newly diagnosed patients. Third, many of the patients with re- transition between the two phases, reducing patients’ exposure to
    lapsed or progressive disease do not have ‘measurable’ disease ac- ineffective doses as well as saving time and money. An example is
    cording to response evaluation criteria in solid tumours (RECIST) the so-​called ‘dose expansion-​cohort’ approach: after the MTD is
    (see Chapter 2 ‘Imaging in Paediatric Oncology’), which require a detected, additional patients are treated at MTD and toxicity is re-
    soft-​tissue lesion visible on cross-​sectional imaging. Instead, chil- corded through the whole trial (pre and post expansion), while ac-
    dren with relapsed neuroblastoma and sarcomas may have only tivity is evaluated on patients belonging to the cohort at MTD as a
    bone or bone marrow disease, reducing the number of eligible pa- secondary endpoint.
    tients. Hence, a classical phase II trial testing efficacy in a single
    or restricted range of childhood cancers inevitably requires inter- Phase III trials
    national collaboration in order to recruit in a reasonable time frame Phase III trials in childhood cancer are generally offered to newly
    of 18–​24 months. This presents practical and financial challenges for diagnosed patients or, for some diseases with standard effective sal-
    implementation. Fourth, the response criteria need to be reassessed vage therapies, to patients at first relapse. They are designed to be
    in the light of new biological findings and, for instance, considered ‘low risk’ to participants in terms of not compromising their chance
    as continuous variables rather than categorical variables. This could of cure. These trials usually compare the current best-​known treat-
    be the case in using the level of minimal residual disease as a measure ment (the standard arm) against a modification of that treatment
    of response in acute lymphoblastic leukaemia (ALL). Also, for some (the experimental arm), and are designed to test whether there is
    targeted treatments in paediatric tumours, especially tumours of the either an improvement in efficacy or the same efficacy with reduced
    central nervous system (CNS), prolonged stable disease could be a toxicity.
    valuable endpoint yet requiring a longer time of observation. In designing a trial, the number of children who can potentially
    Early drug development studies of targeted therapies in child- be recruited in order to ensure sufficient statistical power must be
    hood cancer present new challenges. For example, only rare anticipated. For most childhood cancers, the numbers that need to
    Design and implementation of clinical trials 41

    Table 5.1 Approximate number of newly diagnosed cases per year using existing chemotherapeutic agents. Even for the high-​risk can-
    and expected number of deaths per year for the more common cers, where patients have a dismal prognosis and new experimental
    subgroups of childhood cancer diagnosed before age 15 years, therapeutic approaches offer hope of more marked improvements,
    expected in a European country, total population approximately
    60 million. required sample sizes are in the hundreds, yet these high-​risk can-
    cers generally represent small subgroups. Also in this context, it is
    Diagnostic subgroup Expected no. of Expected no. of important to make every effort not to revert to the use of single-​arm
    cases per yr* deaths per yr* studies and historical controls in order to have sound and unbiased
    Acute lymphoblastic leukaemia 372 71 estimates.
    Acute myeloid leukaemia 69 26 Since the 1970s, many countries have developed national child-
    Hodgkin lymphoma 58 <2 hood cancer study groups which have stimulated the development
    of tumour-​specific committees to design and run clinical trials. High
    Non-​Hodgkin lymphoma 82 16
    rates of participation in national clinical trials have been achieved
    Medulloblastoma 70 31
    through the active engagement of clinicians and patients in specialist
    Astrocytoma 155 33 treatment centres and through national consensus. However, even
    Neuroblastoma 89 38 for the more common childhood cancers, achieving the numbers
    Retinoblastoma 43 2 required for a robustly powered trial within a reasonable time-​span
    Renal tumour 81 13 (normally five years of recruitment) ideally requires participation
    from several countries. This adds to the expense and the time re-
    Hepatoblastoma 11 3
    quired to design and implement such trials and can mean national
    Osteosarcoma 31 10
    compromise in terms of accepting what is defined as the ‘standard’
    Ewing sarcoma 22 6 arm. Nonetheless, expansion of recruitment from national to inter-
    Rhabdomyosarcoma 55 17 national communities has been largely successful, although it has
    Malignant germ-​cell tumour 31 2 sometimes required an innovative approach. For example, a study
    design that prospectively used the principles of meta-​analysis was
    *Estimated from population-​based registration and mortality figures for Great Britain
    (Stiller, 2007) and Europe (Steliarova-​Foucher et al., 2006). adopted in an international trial on intensification of maintenance
    Source: data from Stiller C (Ed), Childhood Cancer in Britain: Incidence, Survival, treatment in ALL.
    Mortality (2007), Oxford University Press, Oxford, UK; and Steliarova-Foucher E et al. In the rarer paediatric cancers or in specific subgroups, inter-
    (Eds), ‘Cancer in Children and Adolescents in Europe’, European Journal of Cancer
    Special Edition (2006), Volume 42, pp. 1913–2190. national cooperation is often the only way to produce solid evidence
    and to improve clinical practice. Thus, when the SIOPEL group of the
    International Society of Paediatric Oncology (SIOP) commenced
    be recruited into a randomized trial are large in comparison with clinical trials in the very rare tumour group, hepatoblastoma, this
    the annual numbers of children diagnosed in an average-​sized led to considerable improvements in survival in participating coun-
    European country per year (Tables 5.1 and 5.2). Optimal sample tries (Figure 5.1).
    size in clinical trials depends largely on the target difference the trial Very large numbers are also needed for non-​inferiority trials.
    aims to detect: the smaller the difference, the greater the number of These are typically used when the aim is to reduce treatment burden,
    participants that needs to be enrolled. The required sample size can toxicity, and long-​term side effects for patients who have a good
    easily reach thousands of patients when the study is powered to de-
    tect less than a 10% absolute increase in event-​free survival (EFS).
    Such relatively modest yet clinically important improvements are to 100
    be expected with more efficient definition of therapeutic strategies 1996–2000 N = 73
    1991–1995 N = 65
    80 1986–1990 N = 62
    1981–1985 N = 51
    Table 5.2 Number of subjects needed in different scenarios 1976–1980 N = 48
    of superiority and non-​inferiority trials with two parallel groups 60 1971–1975 N = 43
    % still alive

    of equal allocation and comparison of event-​free survival (EFS) based 1966–1970 N = 43


    on log-​rank test (one-​sided test, type I error 0.025 and 0.80 power).
    40
    Standard treatment Standard treatment Total number of
    5-​year EFS Δ in 5-​year EFS* subjects (events) 20
    25% +20% 167 (109)
    25% +15% 280 (189) 0
    60% +10% 711 (249) 0 5 10 15 20 25 30 35 40
    Years since diagnosis
    60% +5% 2,897 (1,086)
    90% –​4% 2,077 (249) Figure 5.1 Rapid improvement in survival of children with
    hepatoblastoma (average 11 cases per year) in Great Britain following
    90% –​3% 3,554 (409) the introduction of participation in the SIOPEL clinical trials in 1991. The
    *A positive Δ indicates the target improvement and a negative Δ indicates the target non-​ number of cases diagnosed per quinquennium is indicated.
    inferiority margin that the trial aims to detect. The number of subjects to be enrolled in Reproduced with permission from Stiller C (Ed), Childhood Cancer in
    each treatment group is half the total number reported in the last column (software PASS Britain: Incidence, Survival, Mortality, Table 5.26, p. 204, Oxford University Press,
    11. NCSS, LLC. Kaysville, Utah, USA). Oxford, UK, Copyright © 2007.
    42 CHAPTER 5 Clinical Trials in Childhood Cancer

    risk profile, without compromising overall survival rates. For in- adoption of an adaptive design for a phase III trial could be seen
    stance, in ALL, they are useful in patients with very good clinical as a main contradiction to the confirmatory nature of such late-​
    outcome (at least 90–​95% five-​year EFS), for whom the aim is to development studies, especially if these are pivotal trials aiming at
    optimize therapy under controlled conditions, de-​intensifying cer- approval of a new product.
    tain therapy elements with known short-​or long-​term side effects, Adaptive approaches are justified by the need to cope with difficult
    without compromising the EFS outcome. Up to 30% of children with experimental situations, where the paucity of available knowledge
    ALL treated in current protocols in North America and Western determines uncertainty about crucial aspects of the study. These dif-
    Europe may be candidates for this type of trial. In general, however, ficulties must be acknowledged in advance, so that any modification
    non-​inferiority ‘head-​to-​head’ trials, where a novel drug replaces an of the trial design (e.g. sample size re-​estimation) is anticipated and
    existing one, are often viewed with suspicion, as they may be used to justified in the study protocol. Adaptive designs cannot in fact be
    promote drugs of less therapeutic value. Switching the objective of a used as a remedy for poor trial planning, but actually require more
    trial from superiority to non-​inferiority can be much more seriously upfront planning, both statistically and in managing logistics and
    misleading than vice versa as, for instance, a poorly conducted or an implementation. In short, a proper adaptive design must control
    underpowered superiority trial may easily lead to a non-​significant potential sources of bias, first of all by ensuring that type-​I error is
    result, because treatment effect is diluted or it has wide confidence fully preserved and valid estimates and confidence intervals for the
    intervals. Whenever appropriate, based on the objective and trial re- treatment effect will be available as a summary of the totality of evi-
    cruitment, the design can include a preplanned schedule of interim dence. Additional measures to be taken operationally, in order to
    analyses on the primary endpoint and guidelines for stopping rules maintain the integrity of the trial, include the careful disclosure of
    on selected aspects (e.g. toxicities). interim analyses. In practice, access to the available data and related
    As a final consideration, trial design requires that experts in results must be limited (it is recommended that the trial’s clinical
    various fields (clinical, biological, pharmacological, biostatis- team be isolated from both) and communication of interim analysis
    tical) cooperate to formulate in advance, as clearly and precisely decisions is to be preferred over communication of interim results.
    as possible, based on up-​to-​date knowledge, the objectives and hy- A general concern about monitoring designs is heterogeneity in
    pothesis of the trial. Data from a trial where the results support a treatment effects estimated from the different stages of the trial.
    sound, prespecified hypothesis is more persuasive than one where This may be due to random variation, but in practice it cannot be
    the results are ‘surprising’, and unplanned testing for non-​inferiority excluded that a calendar time effect might be present, for instance
    in trials designed for superiority should be discouraged. Study de- because of a learning process or a change in important prognostic
    signs that incorporate both non-​inferiority and superiority testing factors occurring over time. Heterogeneity in stage results can be de-
    are possible, and may be useful and applicable in selected contexts. tected by applying a heterogeneity test, although its statistical power
    is generally limited. Should it indicate that a discrepancy exists, fur-
    ther investigation into possible causes is recommended.
    Novel strategies for clinical trial design An interesting example of adaptive design is the so-​called adap-
    tive ‘seamless’ design, in which two separate phases of drug develop-
    The issues raised by the more precise characterization of subgroups ment, that could be conducted as separate trials, are in fact combined
    in paediatric oncology continue to grow in importance. Advances into a single one, both operationally and inferentially. Data gathered
    in understanding the molecular basis of many childhood can- from the first, learning stage (e.g. dose selection) are used to adapt
    cers and the development of more sensitive methods for assessing the trial plan of the second, confirming stage, and data from both
    response mean that even the common childhood cancers such as are used to run the final analysis and answer study questions. This
    ALL are now being subdivided into ever smaller subgroups (e.g. adaptive ‘seamless’ design does not simply select out the most prom-
    infants, Philadelphia-​positive patients), each with their own dif- ising treatment option, but actually allows continuation of the pre-
    ferent survival profiles and therapeutic questions. New statistical ferred arm without the need to start a new trial (see Figure 5.2).
    approaches to design are therefore needed to address important Application of this type of study requires that the endpoint for the
    therapeutic questions. These include increased efficacy for those learning phase be immediately available for continuation with the
    subgroups where survival is still very poor. They also include opti- confirming stage and that it also be adopted as a primary endpoint,
    mizing the design of phase I and II studies, given the limited number to answer the study question. A multi-​arm, multi-​stage trial design
    of patients available and the growing number of new drugs that must was recently proposed to compare many new therapies simultan-
    be evaluated for toxicity and activity. These kinds of questions have eously against a control treatment and reject insufficiently active
    stimulated the development of adaptive designs. Such designs may therapies in a randomized pairwise comparison with the control,
    be most useful when drug activity is evaluated in terms of early re- performed at different stages, with a seamless move from one stage
    sponse to treatment, rather than in phase III studies, where efficacy is to the next.
    measured by long-​term evaluation of event-​free and overall survival.
    A broad range of design modifications—​for instance, sample Bayesian designs
    size re-​estimation, discontinuation of treatment arms, changes in Bayesian statistics is a methodology, used both in the design and
    the primary measure of treatment response—​is allowed for adap- in the analysis stage of a trial, that provides a coherent method for
    tive designs. Adaptive designs make use of known interim findings learning from evidence as it accumulates. This is performed by com-
    to modify the sampling plan and are therefore data-​driven. For this bining prior information with accumulating evidence through a
    reason, they raise concern about the unintended consequence of mathematical model, and leads to the so-​called ‘posterior distribu-
    introducing bias and imprecision in the estimates. In addition, the tion’ of the clinical endpoint of interest.
    Practical considerations in conducting clinical trials 43

    Treatment Traditional design


    options
    A

    C
    Control

    Phase II Time-gap Phase III

    Stage 1 (learning) Stage 2 (confirming) Time

    Control
    Phase II/III
    Treatment selection
    Adaptive seamless design

    Figure 5.2 Comparison of adaptive seamless and traditional design.

    The Bayesian approach may be useful especially when good prior the study aim, should the current observed trend continue and the
    information on clinical use of a device or treatment exists, so that trial be continued to the end of planned recruitment.
    incorporating it in the design and analysis of the future trial may re- A Bayesian approach can be used in phase I studies (as shown in
    sult in a smaller-​sized or shorter-​duration study. It can also be used the section ‘Design and implementation of clinical trials’). It can also
    when prior information is only partially (or not) informative, be- be used to design phase II trials more efficiently by thinking of them
    cause it can provide flexible methods for interim analyses and modi- as randomized selection trials comparing several therapies. Thus, in
    fications of on-​going trials (e.g. changes to sample size, changes in rare subgroups, it has been proposed to formally combine, at the de-
    the randomization scheme). sign stage, the expected outcome of the new small trial being planned
    Usually, the Bayesian approach for trial design does not determine with the prior information on treatment effect generated by earlier
    sample size in advance; instead, it may specify a particular criterion studies. A Bayesian approach of this kind provides an estimate of
    to stop the trial. Appropriate stopping criteria may be based on a the potential contribution to the demonstration of treatment effect
    specific amount of information about the clinical measure of interest of the new trial which, if seen in isolation, would be underpowered.
    (e.g. a sufficiently narrow credible interval, as defined below) or an Regulatory agencies have recently issued guidelines on clinical
    appropriately high probability for a prespecified hypothesis. trials in small populations that indicate the possibility of using non-​
    At any point during a Bayesian clinical trial, it is possible to update standard designs. These guidelines propose that there are no spe-
    the computation of the expected additional number of observations cial methods for designing, carrying out, or analysing clinical trials
    needed to meet the stopping criterion. Because the sample size is not in small populations. They note though that the need for statistical
    explicitly part of the stopping criterion, the trial can be terminated efficiency should be weighed against the need for clinically rele-
    whenever enough information has been collected to answer the trial vant/​interpretable results, the latter being more important. While
    questions. When sizing a Bayesian trial, it is recommended that the frequentist formulations of analysis are typically best suited for
    minimum sample size is prespecified according to safety and effect- careful assessment of the strength of evidence, Bayesian formula-
    iveness endpoints. It is also wise to anticipate a minimum level of tions may allow the insertion of additional information which may
    information from the current trial needed to verify model assump- open the route to bolder speculations, sometimes providing an alter-
    tions and appropriateness of prior information used. native to sensitivity analysis.
    Prespecified interim analyses can be part of a Bayesian trial, and
    various methods are available. The final analyses within the Bayesian
    approach include hypothesis testing and interval estimates. For Practical considerations in conducting
    Bayesian hypothesis testing, the posterior distribution can be used clinical trials
    to calculate the probability that a particular hypothesis regarding
    the endpoint is true, given the observed data. Interval estimates in The conduct of clinical trials is a highly regulated process, designed
    the Bayesian context are called credible intervals: a 95% credible to give maximum safety to patients whilst allowing them access to
    interval is the interval with 95% posterior probability of including innovation in a well-​designed, controlled scientific study that has
    the ‘true’ unknown treatment effect. Of note, Bayesian methods for a high chance of achieving its aims. Each country has its own legal
    interim monitoring of futility or efficacy can also be applied in the framework for ethical review procedures and regulatory approval
    setting of a traditional trial. For example, guidelines based on pre- processes. Ensuring consistent data quality when working across
    dictive probability can be used to calculate how likely it is to reach national, cultural, legal. and linguistic boundaries adds further to
    44 CHAPTER 5 Clinical Trials in Childhood Cancer

    the challenges. Hence, multinational trials require additional time entry into a phase III trial is generally considered ‘best practice’ in
    to design and implement. paediatric oncology. More research is needed, however, to under-
    The implementation of the Clinical Trial Directive 2001/​20/​EC stand how best to present the alien concept of a randomized clinical
    (EUCTD) and its translation into law throughout the European trial at this stressful time and to understand why there are differ-
    Union in 2004 aimed to provide a common set of regulations and ences in randomization rates between countries participating within
    defined standards of quality of clinical studies across member states. the same clinical trial. Also, given the success in treating various
    Although the overall aim of the EUCTD is to protect the safety and childhood cancers, attention must be paid to long-​term quality of
    rights of patients and the integrity of trial data, the implementation life and health challenges, with strategies in place to capture all rele-
    of the legislation has proved difficult, and it is well recognized that it vant late effects.
    significantly delayed academic-​driven clinical research and has sig- Engaging families as informed partners in clinical trials is a
    nificantly increased the cost and bureaucracy for non-​commercial key goal of sound practice, but is far from straightforward in its
    trials conducted by universities and public research institutions. achievement. One key issue concerns whether to involve a child
    A key problem is the Directive’s requirement for equal appli- with cancer in the decision about participating in a trial. The prin-
    cation of a high level of regulation and governance to all clinical ciple that children should be consulted, informed, and involved in
    trials that investigate the effects of drugs, regardless of the existing decisions affecting them, but not have the final authority over deci-
    knowledge and experience in using that drug. The EUCTD does sions, is enshrined in the UN Convention of the Rights of the Child
    not differentiate between trials investigating unlicensed medicinal (UNCRC). The guiding principle of the UNCRC is that the ‘best
    products, licensed products used within their marketing author- interests’ of the child must be promoted, but differences of opinion
    ization, or those used ‘off-​label’, as in many cancer indications. between parents, children, and professionals about the ‘best inter-
    For the majority of late-​development clinical trials in childhood ests’ of the child will sometimes need to be resolved. Involving a
    cancer, the drugs in the treatment regimens being investigated child in a decision about participation in a cancer trial may be es-
    have marketing authorization, although are not necessarily being pecially complex in those situations where parents, struggling to
    used according to their licensed indication (i.e. off-​label use). This come to terms with the diagnosis themselves, prefer to defer telling
    off-​label use is supported by considerable published experience the child the full diagnosis and its implications, perhaps until treat-
    and safety data relating to the use of these regimes in the paedi- ment is established and the parents are better able to cope with
    atric cancer population. their own emotions. Some parents may see a demand to involve the
    The need for a more proportionate, risk-​ adapted approach child in the decision—​and, by definition, disclose the diagnosis—​
    to regulation of clinical trials was recognized by the European as interfering with their autonomy as parents. Further complexity
    Commission and was a major factor that led to a revision of the EU’s arises because of varying definitions of minors and of the legal age
    clinical trial legislation. The new law, the Clinical Trial Regulation of consent between different countries. Additionally, some jurisdic-
    (CTR), is anticipated to be implemented in early 2020. The basic re- tions rely on the assessment of ‘competence’ rather than specifying
    quirement that each trial should have an overall sponsor—​taking the a particular age of consent.
    legal responsibility for the conduct of the trial including obtaining a Patient ‘misunderstandings’ of concepts such as randomiza-
    clinical trial authorization from the competent authorities, ensuring tion are often assumed to arise from incomplete, badly written, or
    the suitability of each participating institution and of investigators, misleading information. Much attention, particularly from research
    overseeing pharmacovigilance, and ensuring the integrity of clin- ethics committees, tends to focus on the detail of patient informa-
    ical trial data—​will continue. Nevertheless, the CTR introduces tion sheets (PIS). However, the role of written information in this
    the concept of proportionate regulation and oversight of clinical context remains poorly understood, and parents may find discus-
    trials according to risk. For example, it incorporates a category of sion with staff more helpful than the consent document.
    ‘low-​intervention trial’. This category will apply to clinical trials in
    which licensed drugs are used, either in their licensed indication or
    according to standard clinical practice, and in which the trial inter- Summary
    ventions pose no more than minimum additional risk or burden
    to the safety of the participants. For a trial that complies with the Despite the progress made in curing children with cancer, there
    definition of low intervention, the regulatory requirements will be is little room for complacency and, if progress is to be sustained,
    proportionately reduced, acknowledging that investigational medi- formidable challenges must be faced. These include the scientific
    cinal products in such trials are licensed products that are prescribed challenge of studies in small populations, the complexities of the
    routinely outside the context of clinical trials and for which exten- regulatory environment, ethical and organizational constraints,
    sively documented safety profiles exist. This, amongst other prag- the emotional context, and the attitudes of individual families and
    matic changes in the CTR, will have an important effect in reducing physicians. All efforts should be made to ensure that clinical effi-
    the unnecessary administrative and financial burdens for many late-​ cacy of a treatment is scientifically established for future generations
    development clinical trials in paediatric oncology, where the prin- of children and adolescents. Some of these issues may be addressed
    cipal aim is to optimize the use of currently available, licensed drugs by innovative trial design and implementation, but all will require
    to guide best clinical practice. engagement with the communities of families and physicians on
    Although there are many ethical considerations in undertaking whom trials depend. Clinical research must include a well-​designed
    clinical research on children, a ‘therapeutic alliance’ begins between social science and ethical evaluation of how best to inform and en-
    doctors and families when a child is first diagnosed with cancer. All gage the patient community who need clinical trials to continue in
    parents want their child to access the best treatment available, and order for progress to be made.
    Summary 45

    Lee DP, et al. (2005) Pediatric phase I trials in oncology: an analysis of


    FURTHER READING study conduct efficiency. J Clin Oncol 23(33), 8431–​41.
    Cox, DR (2007) Applied statistics: a review. Ann Appl Stat 1, 1–​16. Machin D, et al. (1997) Sample Size Tables for Clinical Studies (2nd
    Dupont JC, et al. (2016). Ethical issues of clinical trials in paediatric on- edn). Malden, MA: Blackwell Science.
    cology from 2003 to 2013: a systematic review. Lancet Oncol 17(5), Moroz V, et al. (2014) Comparison of anticipated and actual control
    e187–​97. group outcomes in randomised trials in paediatric oncology pro-
    Galimberti S, Devidas M, Lucenti A, Cazzaniga G, Möricke A, Bartram vides evidence that historically controlled studies are biased in fa-
    CR, Mann G, Carroll W, Winick N, Borowitz M, Wood B, Basso vour of the novel treatment. Trials 10(15), 481.
    G, Conter V, Zimmermann M, Suciu S, Biondi A, Schrappe M, Parmar MKB, et al. (2008) Speeding up the evaluation of new agents in
    Hunger S, Valsecchi MG (2018) Evaluation of minimal residual cancer. J Nat Can Inst 100(17), 1204–​14.
    disease as a potential surrogate endpoint for event free survival Piantadosi S (2005) Clinical Trials: A Methodologic Perspective (2nd
    in randomized childhood B-lineage acute lymphoblastic leu- edn). New York: Wiley.
    kemia clinical trials. Journal of the National Cancer Institute Cancer Pritchard-​Jones K, et al. (2008) Improving recruitment to clinical trials
    Spectrum, 2(4). for cancer in childhood. Lancet Oncol 9(4), 392–​9.
    European Medicines Agency: Committee for Medicinal Products Rubinstein LV, et al. (2005) Design issues of randomized phase II trials and
    for Human Use (CHMP) (2006) Guideline on clinical trials a proposal for phase II screening trials. J Clin Oncol 23(28), 7199–​206.
    in small populations. CHMP/​ EWP/​ 83561/​ 2005, July 2006. Scott M, et al. (2010) Bayesian Adaptive Methods for Clinical Trials.
    Available at: http://​www.ema.europa.eu/​docs/​en_​GB/​document_​ Florida: CRC Press.
    library/​Scientific_​guideline/​2009/​09/​WC500003615.pdf (accessed Simon CM, et al. (2004) Comparison of the informed consent process
    May 2017). for randomized clinical trials in pediatric and adult oncology. J Clin
    European Medicines Agency: Committee for Proprietary Medicinal Oncol 22(13), 2708–​17.
    Products (CPMP) (2000) Points to consider on switching between Simon R (1989) Optimal two-​stage designs for phase II clinical trials.
    superiority and non-​ inferiority. CPMP/​ EWP/​ 482/​ 99, July 2000. Control Clin Trials 10, 1–​10.
    Available at: http://​www.ema.europa.eu/​docs/​en_​GB/​document_​ Steliarova-​Foucher E, et al. (eds) (2006) Cancer in children and adoles-
    library/​Scientific_​guideline/​2009/​09/​WC500003658.pdf (accessed cents in Europe. Eur J Cancer Special Edn 42, 1913–​2190.
    May 2017). Stiller C (ed.) (2007). Childhood Cancer in Britain. London: Oxford
    Freedman, B (1987) Equipoise and the ethics of clinical research. New University Press.
    Engl J Med 317, 141–​5. Sullivan R, et al. (2013) New policies to address the global burden of
    Garattini S, Bertelè V (2007) Non-​inferiority trials are unethical be- childhood cancers. Lancet Oncol 14(3), e125–​35.
    cause they disregard patients’ interests. Lancet 370, 1875–​7. Valsecchi MG, Masera G (1996) A new challenge in clinical research
    Hearn J, Sullivan R (2007) The impact of the ‘Clinical Trials’ directive in childhood ALL: the prospective meta-​ analysis strategy for
    on the cost and conduct of non-​commercial cancer trials in the UK. intergroup collaboration. Ann Oncol 7, 1005–​8.
    Eur J Cancer 43, 8–​13. Valsecchi MG, et al. (2008) Web-​based international studies in limited
    Kumar A, et al. (2005) Are experimental treatments for cancer in chil- populations of pediatric leukemia. Pediatr Blood Cancer 50, 270–​3.
    dren superior to established treatments? Observational study of Vassal G, et al. (2013) New drugs for children and adolescents with
    randomised controlled trials by the Children's Oncology Group. cancer: the need for novel development pathways. Lancet Oncol
    BMJ 331, 1295. 14(3), e117–​24.
    6
    Evidence-​Based Paediatric Oncology
    Leontien C.M. Kremer, Erik A.H. Loeffen, and Robert S. Phillips

    History cancer. There is an overwhelming amount of information avail-


    able; however, this information is rarely summarized and the con-
    The term evidence-​based medicine was introduced by a research group clusions may be conflicting. Even in 1995, the scale of this problem
    led by David Sackett and Gordon Guyatt from the McMaster Medical was already estimated as requiring that a physician would need to
    School in Canada, in 1992, with one of the first publications about read 17 journal articles a day, all year round, in order to keep up
    evidence-​based medicine appearing in the Journal of the American to date with all research relevant to a particular area of clinical
    Medical Association. The work built on strong foundations, particu- practice. With the number of studies published annually having
    larly a book published in 1972 by a Scottish epidemiologist called more than doubled since then, one can imagine that this is an un-
    Archie Cochrane—​Effectiveness and Efficiency: Random Reflections attainable effort.
    on Health Services—​which outlined the concepts of evidence-​based In this chapter, we will give an overview how EBM can be used in a
    practice. The Cochrane Collaboration (based on these ideas) was non-​exhaustive but still comprehensive way in daily practice of care
    founded in 1993 and continues to grow. It is dedicated to making up-​ for children with cancer, and which tools are available for paediatric
    to-​date, accurate information about the effects of healthcare readily oncologists.
    available worldwide. The practice of evidence-​based medicine is very Three different roles for practice of EBM have been pro-
    important in delivering optimal patient care, and the terms evidence-​ posed: replicator, user, and doer. The replicator describes the
    based medicine, or evidence-​based practice, are used all around the doctor who copies, knowingly or unconsciously, the actions
    world. The British Medical Journal listed evidence-​based medicine as and practices of a trusted colleague. This is hopefully a mentor
    one of the 15 most important medical advances since 1840. whose work is centred on the delivery of EBM. The user is the
    commonest practice of taking evidence-​based resources such as
    randomized controlled clinical trials (RCTs), systematic reviews,
    What is evidence-​based medicine? or evidence-​based guidelines, and using them to guide clinical
    problem solving. When the resources that the doctor is using are
    Evidence-​based medicine (EBM) is the practice of identifying the solidly based on the best-​quality evidence that exists, this is an
    best available research evidence, utilizing a process of systematically extremely efficient way of practicing EBM. This is the most prac-
    finding, appraising, and interpreting research findings, and integrating tical and widely used way of putting evidence-​based paediatric
    these with clinical expertise and patient values to come to a clinical de- oncology into action, and the majority of this chapter will focus
    cision (see Figure 6.1). It has been strikingly defined by Sackett et al. as on how to learn to become a skilled user of EBM.
    ‘the conscientious, explicit, and judicious use of current best evidence The doer refers to performing EBM in detail; full expression of the
    in making decisions about the care of individual patients’. It describes structured question, broad search and detailed appraisal of the evi-
    a philosophy of care which integrates the highest quality scientific re- dence identified, summarizing the evidence, followed by a careful de-
    search, the treating clinician’s expertise, and the child’s (and his/​her cision and implementation plan. This is infrequently undertaken by
    family’s) specific medical, psychological, and social condition. It is a a paediatric oncologist in daily clinical practice, but more common
    heuristic way of thinking—​not a recipe or fixed set of instructions. in the development of local policies, systematic reviews, or guide-
    lines. It is the most thorough but also the most time-​consuming and
    The different roles for evidence-​based medicine technically demanding of the three roles, and learning this one re-
    quires a level of detail well beyond this chapter. We will only give
    in paediatric oncology
    some insight into the Grading of Recommendations, Assessment,
    Development, and Evaluation (GRADE) methodology for guideline
    There is a daily need for valid information about therapy, diag-
    development.
    nosis, prognosis, aetiology, and prevention for children with
    Steps in evidence-based paediatric oncology for a user of evidence-based medicine 47

    answer their question. In this effort, it might also be worthwhile to


    a priori decide which outcomes are most important for decision
    making, as there is a chance the identified evidence is incomplete or
    Science Clinician
    conflicting on various outcomes (i.e. improvement on one outcome,
    EBM worsening on another outcome).

    Second step: acquisition is not a vice


    Patient Timely supply of useful evidence from the scientific medical litera-
    (and family)
    ture is essential for the optimal care of children with cancer, and
    professionals have a responsibility to keep up to date with research
    Figure 6.1 The integration of the triad of patients, physicians, and findings to ensure that their practice remains effective. These needs
    publications. can be characterized by those of keeping up to date (‘just in case’
    information) and answering day-​to-​day clinical questions (‘just in
    time’ information).
    Steps in evidence-​based paediatric oncology for a Summary or alerting systems can be useful for keeping abreast
    user of evidence-​based medicine of new information. ‘EvidenceAlerts’, a service from DynaMed
    Plus and McMaster University, alerts subscribers by email to new
    The practical output of EBM in daily practice, as a user of EBM, can studies (https://​plus.mcmaster.ca/​EvidenceAlerts/​). EvidenceAlerts
    be viewed as a four-​step process: uses highly trained research staff to select papers based on meth-
    odological quality from over 110 premier clinical journals. At least
    1. Ask a structured clinical question. three members of a worldwide panel of practicing physicians then
    2. Acquire relevant research. rate the papers for clinical relevance and interest, and often leave
    3. Appraise the identified evidence for its validity, importance, and contextual comments which can be really useful to practicing clin-
    applicability. icians. In addition, in most electronic databases, such as PubMed, it
    4. Apply the results in practice. is also possible to create a periodical alert for new results of a user-​
    specified search.
    First step: ask . . . and it might be given unto you A widely described hierarchy of information resources to address
    the day-​to-​day clinical questions is that of the ‘5S’ pyramid (Figure
    Asking an answerable question can prove terribly difficult, but when
    6.2). In the ideal situation, evidence from applicable research is in-
    good questions are dissected, the physiology becomes clear. The first
    tegrated into computerized systems that aid in decision making,
    step in deconstructing the anatomy of inquiry is to ask ‘What sort of
    such as electronic care pathways. If this is not available, then sum-
    question is being asked?’ Clinical questions can be grossly categor-
    maries in a guideline or protocol could deliver the right answers in
    ized as ‘foreground’ and ‘background’. Background questions are
    an evidence-​based fashion. If these cannot be found, then specific
    generally broad, and can be caricatured as those asked by a medical
    questions can be effectively answered if others provide a resource
    student. They are often of the ‘what is’ or ‘what causes’ type: for ex-
    that finds, appraises, and summarizes the key evidence related to a
    ample, ‘What is a pilomyxoid astrocytoma?’, ‘What causes leukaemia
    specific topic. These synopses can be incredibly useful to both an-
    in childhood?’ Foreground questions are specific, have intent, and
    swer specific questions and to try to keep up to date with the new
    can be asked in the domains of therapy (efficacy or harm), diagnosis,
    knowledge that is flooding out of medical research. Without these,
    prognosis, aetiology, or prevention. They frequently have three or
    then a systematic review can provide a synthesis of the breadth of the
    four components, summarized in the acronym ‘PICO’:
    best evidence to answer specific questions. Lastly, and only rarely,
    Patient, problem, or population (the group to which the question should we need to rely on looking at single, individual studies to an-
    applies) swer our common queries.
    Intervention (or diagnostic test, or exposure, or prognostic marker) We highlight examples of the 5S pyramid in the rest of this section
    Comparison (not always necessary) and in Table 6.1. There are also various meta-​search engines that
    Outcome(s) of interest (which should include patient-​relevant, ad- search in multiple layers of the 5S pyramid: examples are TRIP
    verse, and beneficial measures) (www.tripdatabase.com) and SUMsearch2 (www.sumsearch.org).

    An example might be ‘In children with metastatic rhabdomyosar- Systems: computerized decision-​support and care pathways
    coma [population], does high-​dose therapy with stem-​cell rescue There are few opportunities in paediatric oncology to use compu-
    [intervention] compared with intensive conventional chemotherapy terized decision-​support systems where evidence-​based guidelines
    [comparison] increase five-​ year overall survival [outcome]?’ are included in computer systems for daily practice. But the idea of
    Questions of therapy can usually fit quite easily into this PICO frame. having a ‘pathway’ of care, based on the best available evidence, is
    As a final aside in the description of structured clinical questions, not new to paediatric oncologists. Every treatment protocol we use
    it is worth reflecting on how to choose which outcomes to look at. has, built into it, a plan of investigation, treatment, response assess-
    Do doctors really know what outcome their patients and families ment, and follow-​up. The use of care pathways has been demon-
    most value, or do physicians presume? If the question is highly par- strated to have good improvements in patient-​related outcomes in
    ticular to one patient and their family, it may be pertinent to ask the a number of adult and paediatric specialty areas, although no direct
    patients and their families directly, to frame the PICO to actually evidence exists for children with malignancies. In the (near) future,
    48 CHAPTER 6 Evidence-Based Paediatric Oncology

    Computerized decision support and care pathways


    Examples: decision support systems that search the evidence for a given
    Systems patient/problem

    Evidence-based clinical practice guidelines (CPGs)


    Summaries Examples: US National Clearinghouse, Guidelines International network,
    World health Organization, SIOP, IGHG, iPOG, etc.

    Evidence-based journals, structured abstracts, other summaries


    Synopses Examples: ACP Journal Club, BMJ Clinical evidence, EBM-Journal,
    ADC ‘Archimedes’, UpToDate

    Systematic reviews
    Syntheses
    Examples: The Cochrane Library, PubMed

    Single studies; RCTs, cohort studies, case series, etc


    Studies Examples: PubMed, EMBASE, Cochrane Central Register of Controlled
    Trials

    Figure 6.2 The ‘5S’ pyramid.

    we expect to see computerized decision-​support software, based on Clinical guidelines might require adaptation before use in a local
    the flow charts of clinical practice guidelines, being integrated more healthcare environment. The ADAPTE toolkit provides a helpful
    fully in electronic health record systems. approach to doing so (www.g-​i-​n.net/​working-​groups/​adaptation/​
    history). Some useful sites to find guidelines related to paediatric
    Summaries: evidence-​based clinical practice guidelines oncology are in Table 6.1.
    The Institute of Medicine defined clinical practice guidelines (CPGs)
    as ‘statements that include recommendations, intended to optimize Synopses: evidence-​based journals, structured abstracts,
    patient care, that are informed by a systematic review of evidence and other summaries
    and an assessment of the benefits and harms of alternative care Several journals present evidence-​based summaries of the know-
    options’. Guidelines are frequently produced by multidisciplinary ledge in healthcare: examples are ACP (American College of
    teams which include healthcare professionals of different types, Physicians) Journal Club, BMJ (British Medical Journal) Clinical
    guideline methodologists, health economists, and patient advocates. Evidence, and Evidence-​Based Medicine (EBM Online). These jour-
    The development of a CPG can be divided into five phases: nals focus on a general medical audience, sometimes with paedi-
    atric issues being included. The ‘Archimedes’ section of Archives of
    1. Preparation—​
    selecting topic, identification of existing CPGs,
    Diseases of Childhood provides short evidence-​based overviews of
    formulating clinical questions
    paediatric topics.
    2. Search and selection—​performing the systematic search, selecting
    A further resource is UpToDate (www.uptodate.com). This is
    citations based on predefined inclusion and exclusion criteria
    an internet-​based electronic textbook that offers clinical decision-​
    3. Extraction and appraisal—​extracting the data using standard- support information using current evidence to answer clinical ques-
    ized data extraction forms and appraising the methodological tions. All topics are peer reviewed. However, all the steps of EBM are
    quality of the included evidence not clearly described which makes it difficult to judge the influence
    4. Summarization and formulation of recommendations—​the in- of bias. UpToDate covers around 8,000 topics but only a few topics
    cluded evidence is summarized, whereafter the panel discusses important for the care of children with cancer are included.
    the benefits, harms, patient values, and other important factors,
    and formulates recommendations Syntheses: systematic reviews in childhood cancer
    5. Writing, review, and publication—​the CPG is written (with a
    The synthesized evidence in systematic reviews can help healthcare
    specific description of the process, and the choices made in for-
    professionals to keep up to date. In a systematic review, the three first
    mulating recommendations) and sent out for external review,
    steps of EBM are included—​formulating a clinical relevant ques-
    whereafter it is published.
    tion, performing an extensive search for literature, and assessing the
    Guidelines can have pitfalls, as can all forms of research. For ex- methodological quality of the selected studies. Furthermore, a meta-​
    ample, the high number of topics to be covered by a guideline in analysis can be included. A systematic review aims to minimize the
    a limited time may mean the evidence summaries are suboptimal, occurrence of bias through a systematic search, selection, and as-
    and guidelines may become ‘stale’ as it is difficult to refresh them sessment of methodological quality and to reduce random error
    frequently and keep the knowledge up to date. The term ‘evidence-​ by using quantitative methods (i.e. meta-​analysis) where appro-
    based’ should be reserved for guidelines that followed the structured priate. In paediatric oncology, 117 systematic reviews, including 18
    process of CPG development. It is important to realise that the term Cochrane reviews, had been published between 1988 and November
    ‘evidence-​based’ does not only apply to those recommendations 2006. Unfortunately, the methodological quality of most of the sys-
    where strong evidence supports the choice for a clinical decision but tematic reviews in paediatric oncology seemed to have serious
    also, rather, refers to the process of generating the guidance. methodological flaws, leading to a high risk of bias.
    Steps in evidence-based paediatric oncology for a user of evidence-based medicine 49

    Table 6.1 Examples for the ‘5S’ evolution of information services and, since 2006, Cochrane Childhood Cancer (CCC) has been regis-
    for evidence-​based healthcare decisions tered within the Cochrane Collaboration (www.ccg.cochrane.org).
    The aim of CCC is to conduct and maintain systematic reviews on
    5S Examples Websites
    interventions and diagnosis for cancer in children and young adults
    Systems Decision-​support systems
    that search the evidence
    with respect to prevention, treatment, supportive care, psychosocial
    for a given patient/​ care, palliative and terminal care, nursing care, and late effects of
    problem treatment. As of 2017, 75 author groups from all around the world
    Summaries General (examples) have been preparing and updating Cochrane systematic reviews in
    Clinical US National www.guidelines.gov collaboration with CCC. Those who prepare systematic reviews for
    practice Clearinghouse CCC are mostly volunteers.
    guidelines
    Guidelines International www.g-​i-​n.net
    Network Studies
    UK National Institute www.nice.org.uk If no other sources are available, the only option to get scientific
    for Health and Clinical information is to identify as many single studies as are relevant to
    Excellence
    the question in PubMed or similar bibliographical databases. Each
    Scottish Intercollegiate www.sign.ac.uk
    Guidelines Network
    type of question (i.e. therapy (efficacy or harm), diagnosis, prog-
    nosis, aetiology, or prevention) is best answered by a different study
    Australian Clinical www.clinicalguidelines.gov.au
    Practice Guidelines design.
    World Health www.who.int Third step: appraisals are more than criticism
    Organization
    The process of critical appraisal is to examine a study in detail, in
    Evidence-​based guidelines www.ebm-​guidelines.com
    order to determine if it provides a valid, important, and applicable
    Paediatric oncology (examples)
    answer to a specific clinical question. It is not the process of finding
    SIOP (Society of www.siop.nl flaws and destroying research: it is evaluating strengths and weak-
    Paediatric Oncology)
    nesses and coming to a conclusion in clear sight.
    Children’s Oncology www.childrensoncologygroup. There are a wide range of structured checklists, which differ ac-
    Group org
    cording to the study type that is being appraised. We favour the
    International Late Effects www.ighg.org
    Cochrane risk of bias tool for randomized studies and the ROBINS-​
    of Childhood Cancer
    Guideline Harmonization I tool for non-​randomized studies. For other study types, guidance
    Group (IGHG) is provided in the JAMA Users Guides (www.jamaevidence.com).
    International Paediatric www.sickkids.ca/​Research/​iPOG The methodological quality of synthesized evidence (like systematic
    Oncology Guidelines in reviews or guidelines) can be appraised in a systematic manner by
    Supportive Care Network
    using the AMSTAR tool or ROBIS tool for systematic reviews, and
    (iPOG)
    the AGREE instrument for guidelines.
    Synopsis ACP Journal Club annals.org/​aim/​journal-​club
    BMJ Clinical Evidence www.clinicalevidence.bmj.com Fourth step: apply the results in practice
    Evidence-​Based Medicine www.ebm.bmj.com
    (EBM Online)
    After the first three steps of EBM, a decision should be made on how
    to apply the results to your patient(s) or how to formulate recom-
    ‘Archimedes’/​‘Picket’ adc.bmj.com
    section of Archives of mendations for (a group of) patients.
    Diseases of Childhood For only 20% of the interventions in clinical practice do we know
    UpToDate www.uptodate.com that the treatment is beneficial or we know that there is a trade-​off
    between benefits and harms. Even in these cases, a weighting of the
    Synthesis The Cochrane Library www.cochranelibrary.com
    (Cochrane database of evidence, clinical expertise, patient values, and sociocultural values
    systematic and Dare; will be essential. When the evidence is uncertain, making these de-
    non-​Cochrane reviews) cisions and trade-​offs explicit and insightful becomes even more im-
    PubMed www.pubmed.com portant. Furthermore, the costs and resources used or saved should
    Single PubMed www.pubmed.com be taken into account.
    studies In the world of paediatric oncology, there are frequently times
    Embase www.embase.com when the rarity of the condition makes ‘gold standard’ or high-​
    Cochrane Central Register www.cochranelibrary.com/​ quality, directly applicable evidence unobtainable. For these sorts of
    of Controlled Trials central situations, Jeremy Howick, Paul Glazsiou, and Jeffrey Aronson have
    (CENTRAL) revisited the Revised Bradford-​Hill Guidelines for Causation and
    developed the concept of categories of evidence: a trio of the main
    category (direct evidence) and two subsidiary categories (mechan-
    The largest single provider of systematic reviews for healthcare is istic and parallel evidence).
    the Cochrane Collaboration, with approximately 6,000 systematic re-
    views included in the Cochrane Library. The Cochrane Collaboration Direct evidence is the data we rely on most securely: the RCT gives
    developed standardized methods to perform systematic reviews results of a therapy being effective, or when an exceptionally
    50 CHAPTER 6 Evidence-Based Paediatric Oncology

    effective intervention is considerably better than all others (like problem, anticipated benefits and harms (using the evidence and
    when insulin was first used in diabetic ketoacidosis). its determined quality level), resource use, equity, acceptability,
    Parallel evidence is used in paediatrics a great deal: similarities and feasibility. This leads to an overall balance of desirable and
    exist between the direct evidence in other populations and the undesirable consequences, from which a recommendation is
    ones we treat, and by replication of results across different areas. distilled. This recommendation to do or not do something can
    Mechanistic evidence is the shakiest: coherence with other theories either be strong (‘you should’, ‘we recommend’) or weak (‘you
    of how things work, and a plausible biological explanation. might’, ‘we suggest’). For each recommendation, the accom-
    panying quality of evidence is provided, where it should be noted
    Another reason why evidence could be difficult to translate dir- that high-​quality evidence does not automatically lead to a strong
    ectly to practice is when, for instance, the results of a systematic recommendation or low-​quality evidence to a weak recommen-
    review are focused on concepts. For example, a systematic review dation. A strong recommendation based on very low-​quality
    of antiemetics to prevent nausea and vomiting after chemotherapy evidence (e.g. small, observational studies with large effect on
    tells you ‘antiemetics work’. However, you cannot just prescribe pivotal outcome) or a weak recommendation based on high-​
    ‘antiemetics’, so how can you transfer the review result into clinical quality evidence (e.g. large, randomized trials with a marginal
    practice? A couple of suggestions have been offered, and both have beneficial effect) is also possible.
    clinical common sense: We would not advise to follow this EtD methodology strictly and
    1. Go for the strategy of the largest trial (it has the most weight in extensively when in need of an answer for a specific patient ques-
    the review, so has the ‘most’ evidence). tion, as this is very time-​consuming and more appropriate for true
    2. Go for the most often-​used dose/​drug/​approach (it was shown guideline development. However, it is beneficial to familiarize one-
    to be feasible the most often). self with this way of thinking about/​coming to a decision, as it might
    be extremely useful when weighing the options and the evidence for
    This clear conceptualization of the various guises in which evidence a specific patient question.
    exists can help us decide what works, what works best, and what is In conclusion, there are various methods for deriving a clinical
    likely to work for this patient. decision from the available evidence, of which critical appraisal of
    the identified evidence is always a pivotal step. For each clinical de-
    cision the (quality of) available evidence, values of patients, cultural
    Steps in guideline development as a doer values, clinical expertise, and the resource use should be taken into
    of evidence-​based medicine account.

    A more rigorous approach to answering a question is to complete a


    GRADE evidence to decision (EtD) framework for the clinical ques- Common criticisms of evidence-​based medicine
    tion at hand. This framework helps to systematically and transpar-
    ently assess things like benefits, harms, use of resources, acceptability, Since the introduction of EBM, criticisms of this paradigm shift have
    and feasibility of a treatment option, and thus aids in making a clin- been described which can be divided into ‘limitations universal to
    ical decision. This approach comprises two major steps: summar- the practice of medicine’, ‘misperceptions of EBM’, and ‘genuine
    izing the evidence and formulating one or more recommendations. limitations of EBM’.
    In the first step, the GRADE system is used to determine the so-​ Within the group of ‘limitations universal to the practice of medi-
    called ‘body of evidence’ per outcome. For each outcome of interest, cine’ fall criticisms such as ‘There is a shortage of coherent, con-
    a different combination of studies can be included. Therefore, for sistent scientific evidence’ or ‘There is no evidence available’, along
    each outcome the quality of evidence is determined separately on with complaints stating ‘It is difficult to translate evidence to the care
    a four-​level scale—​high, moderate, low, or very low. The initial of an individual patient’. In instances like this, the methods of EBM
    quality depends on the included type of studies: randomized con- can be used to assess these shortages and difficulties systematically,
    trolled trials are regarded as high-​quality, observational studies are and thus identify research gaps for further study.
    regarded as low-​quality. These levels of evidence can then be down- The second group of criticisms of EBM are based on mispercep-
    graded due to risk of bias (see step 3: appraisal), inconsistency, indir- tions. For example, criticisms that EBM ‘denigrates clinical expertise’
    ectness, imprecision, and/​or publication bias. Upgrading the level of or ‘ignores patients’ values’ or ‘promotes “cookbook” medicine’ arise
    evidence is also possible due to the magnitude of effect, identified through a failure to include all the important steps of EBM. EBM is
    dose–​response gradient, or identified plausible confounders that a guide in science and aids decision making; it is not a law, nor it is a
    would have reduced the effect. There is an excellent series of articles system that will make a decision for you. A further example of mis-
    in the Journal of Clinical Epidemiology teaching the reader how to perception is that ‘only randomized trials or systematic reviews’ con-
    use this system. (Details of the first of these articles, by Guyatt et al., stitute the ‘evidence’ in EBM. This is not true, and several sources of
    can be found in ‘Further reading’). The GRADE system is being in- best available evidence may inform clinical decision making. For in-
    creasingly adopted by organizations other than the GRADE working stance, if RCTs to answer your clinical questions are lacking, you can
    group: for instance, the Cochrane Collaboration recommends this choose to divert to RCTs in other populations, but in certain scenarios
    approach in assessing the quality of evidence in their reviews. it might be more suitable to divert to other study designs, such as non-​
    In the second step, the GRADE EtD framework is completed, randomized comparison studies or cohort studies, in your own target
    in which questions are answered regarding the priority of the population.
    The way forward 51

    Finally, some of the criticisms of EBM are genuine limitations of


    the process: there is a need to develop new skills and apply the pro- FURTHER READING
    cess during limited time and with limited resources. To overcome Alonso-​Coello P, et al. (2016) GRADE evidence to decision (EtD)
    this barrier, we believe an increasing number of prepared evidence-​ frameworks: a systematic and transparent approach to making well
    based products and training programmes will be available in the informed healthcare choices. 1: Introduction. BMJ 353, i2016.
    future. Caldwell PH, et al. (2004) Clinical trials in children. Lancet 364, 803–​11.
    Davidoff F, et al. (1995) Evidence based medicine. BMJ 310, 1085–​6.
    Evidence-​ Based Medicine Working Group (1992) Evidence-​ based
    medicine. A new approach to teaching the practice of medicine.
    The way forward JAMA 268, 2420–​5.
    Godlee F (2007). Milestones on the long road to knowledge. BMJ
    To take paediatric oncology further forward, in developing its evi- 334(Suppl 1), s2–​3.
    dence base and the implementation of that evidence, we need to con- Graham R, et al. (2011) Clinical Practice Guidelines We Can Trust.
    tinue doing lots of what we are doing: Washington DC: The National Academies Press.
    Guyatt G, et al. (2011) GRADE guidelines: 1. Introduction—​GRADE
    • We need to work as international collaborators, focused on the evidence profiles and summary of findings tables. J Clin Epidemiol
    cure and support of children with malignancies by engaging in 64, 383–​94.
    high-​quality research, including our patients, their families, and Haynes RB (2006) Of studies, syntheses, synopses, summaries, and sys-
    our colleagues in other specialties. tems: the ‘5S’ evolution of information services for evidence-​based
    • We need to continue to collaborate in developing high-​quality healthcare decisions. Evid Based Med 11, 162–​4.
    trials in paediatric oncology. As previously stated by Caldwell and Higgins JPT, et al. (eds) (2019) Cochrane Handbook for Systematic
    colleagues in the Lancet: ‘Participation in protocol-​driven clinical Reviews of Interventions, Version 6.0. Available from: www.training.
    research is clearly better than the ad hoc patterns of non-​protocol cochrane.org/​handbook
    Higgins JPT, et al. (eds) (2019) Cochrane Handbook for Systematic
    treatments, paediatric cancer trials offer a paradigm for paediatric
    Reviews of Interventions (2nd edn). Chichester, UK: Wiley.
    clinical research’.
    Howick J, et al. (2009) The evolution of evidence hierarchies: what can
    • We need to collaborate in summarizing the evidence. CCC offers Bradford Hill’s ‘guidelines for causation’ contribute? J R Soc Med
    a unique possibility to collaborate worldwide in preparing scien- 102, 186–​94.
    tific summaries of evidence. We should strive towards making the Lundh A, et al. (2009) Quality of systematic reviews in pediatric
    evidence easily accessible and usable, for instance with the devel- oncology—​a systematic review. Cancer Treat Rev 35, 645–​52.
    opment of summaries of findings tables. Phillips R, Glasziou P (2008) Evidence based practice: the practicalities
    • We need to collaborate in producing evidence-​based guidelines. of keeping abreast of clinical evidence while in training. Postgrad
    For this purpose, various international initiatives have been under- Med J 84, 450–​3.
    taken, such as the International Late Effects of Childhood Cancer Rotter T, et al. (2010) Clinical pathways: effects on professional prac-
    Guideline Harmonization Group (IGHG) and the International tice, patient outcomes, length of stay and hospital costs. Cochrane
    Paediatric Oncology Guidelines in Supportive Care Network Database Syst Rev 17, CD006632.
    Sackett DL, et al. (1996) Evidence based medicine: what it is and what
    (iPOG). International collaboration is of the utmost importance to
    it isn’t. BMJ 312, 71.
    reduce duplication of work and maximize results from the available
    Shea BJ, et al. (2009) AMSTAR is a reliable and valid measurement tool
    time and effort. to assess the methodological quality of systematic reviews. J Clin
    • We need to develop new systems and strategies to implement up-​ Epidemiol 10, 1013–​20.
    to-​date high-​quality scientific knowledge in daily practice. Sterne JAC, et al. (2016) ROBINS-​I: a tool for assessing risk of bias in
    • We need to develop training for researchers and clinicians in clin- non-​randomised studies of interventions. BMJ 355, i4919.
    ical epidemiology and EBM methods. Straus SE, McAlister FA (2000) Evidence-​based medicine: a commen-
    tary on common criticisms. CMAJ 163, 837–​41.
    In conclusion, we need interventions to bridge the gap between Whiting P, et al. (2016) ROBIS: a new tool to assess risk of bias in sys-
    science and practice in paediatric oncology worldwide and make tematic reviews was developed. J Clin Epidemiol 69, 225–​34.
    our daily practice as evidence-​based as possible.
    7
    Allogeneic Stem-​Cell Transplantation
    in Children and Adolescents
    with Malignancies
    Adriana Balduzzi, Giovanna Lucchini, Jean-​Hugues Dalle, and Rupert Handgretinger

    History of stem-​cell transplantation Recent developments


    The myeloablative effect of the conditioning regimen and the
    alloreactivity of donor cells, potentially eradicating residual malig-
    The incipit
    nant cells in the patient, play crucial role in transplantation.
    The first allogeneic haematopoietic stem-​cell transplantation (HSCT) Even though, in the current era, higher resolution of HLA typing—​
    was pioneered by E. Donnall Thomas and reported in the New allowing strict donor–​recipient matching—and better management
    England Journal of Medicine in 1957. Radiation and chemotherapy, of infection and graft-​versus-​host disease (GvHD) ​lessen—the inci-
    followed by intravenous infusion of bone marrow (BM) from a dence of complications, the risk of morbidity and, mainly, mortality
    normal donor, led to haematological reconstitution and leukaemia (defined as transplant-​related mortality, TRM) remain definitely
    eradication. Despite the fact that no patients transplanted in the late higher than with chemotherapy alone. The use of peripheral blood
    1950s and early 1960s survived—​due to graft failure, rejection, graft-​ stem cells (PBSC) or cord blood (CB), instead of BM, for HSCT, has
    versus-​host disease (GvHD), or opportunistic infections—​ three become a routine part of transplantation.
    major principles of allogeneic HSCT were demonstrated: In the mid 1990s, alongside myeloablative (MA) conditionings,
    1. the role of the preparative anti-​leukaemic regimen, non-​MA or reduced-​intensity (RI) conditionings were introduced.
    2. the ability of the new engrafted immune system to prevent leu- As expected, non-​MA and RI conditionings broadened the eligi-
    kaemia relapse, and bility for HSCT to include patients with comorbidities.
    3. the activity of the engrafted immune system against the recipient.

    It was a proof of principle. The bone marrow transplantation era Donor type
    had begun.
    After van Rood’s recognition of anti-​HLA antibodies in pregnant The human leukocyte antigen barrier
    women in 1958, the discovery of the role of human leukocyte antigen
    After van Rood’s pivotal discovery and the subsequent studies which
    (HLA) in tissue transplantation, and van Bekkum’s animal models,
    unravelled the genetics of HLA, allowing better donor selection, a
    three patients with congenital immunodeficiencies were trans-
    major concern in the 1970s was the limitation of allogeneic grafting
    planted from their HLA-​identical siblings, in the United States and in
    only to those patients who had an HLA-​identical sibling available.
    the Netherlands, in 1968. Thomas then reported the first allografting
    In 1979, Hansen performed the first successful marrow graft from
    experience in patients with severe aplastic anaemia in 1972, in 100
    an unrelated donor (URD) for a patient with leukaemia. Since then,
    advanced acute leukaemia patients in 1977, after which 13 patients
    the establishment of URD registries has facilitated HSCT worldwide
    were alive without disease, and in acute myeloid leukaemia (AML)
    and dramatically increased the odds for finding a match, which have
    patients in first remission in 1979, achieving a cure rate of 50%. In
    risen from 25% to 75% for Caucasian patients.
    1990, E. Donnall Thomas won a Nobel Prize ‘for his discoveries in
    cell transplantation in the treatment of human diseases’. Donor and patient registries
    During the past 50 years, the role of HSCT has changed from a In 1972, the International Bone Marrow Transplant Registry
    desperate therapeutic approach to a curative treatment modality for (IBMTR) and, in 1974, the European Group for Blood and Marrow
    thousands of patients with haematological diseases. Transplantation (EBMT) were established for documenting HSCT
    Donor selection 53

    outcome data and facilitating cooperative studies. Between 1989 of an individual, whereas class II genes are found primarily on cells
    and 1995, the use of volunteer URDs was extended and reported of the immune system, such as B lymphocytes and macrophages.
    in children affected with congenital disorders or malignancies by In contrast to previous serological methods, HLA specificities are
    many groups. TRM was a major obstacle; relapse remained the nowadays mostly identified by DNA sequencing. A locus name is
    most frequent cause of treatment failure in malignancies. followed by a one-​or two-​digit number, indicating a serological-​
    defined specificity (e.g. HLA-​A2). A DNA sequence registered with
    Human leukocyte antigen typing the World Health Organization (WHO) is assigned a number with
    Nowadays, HLA compatibility with the donor, either related or unre- the first digits usually related to the closer serological equivalent,
    lated, is defined by high-​resolution typing (four digits) for HLA-​A, -​ such as HLA-​A*02:01, and the following digits representing the
    B, -​C, -​DR, -​DQ HLA alleles. All the ten alleles are usually matched in order of sequence registration for alleles of that group (e.g. HLA-​
    paediatrics, but also matching for all but DQB1 alleles is commonly A*02:01:01:01). The asterisk indicates that data are assessed by mo-
    accepted. lecular typing.
    According to the Paediatric Diseases Working Party of the Individuals inherit a set of HLA, called a haplotype, from each
    European Bone Marrow Transplantation association (EBMT) parent; therefore, the probability that a child shares both haplotypes
    and the HSCT Committee of the International BFM (Berlin–​ with a full sibling is 25%, the chance of sharing one haplotype is
    Frankfurt–​Munster) Study Group (I-​BFM SG), donors are de- 50% and the chance of sharing neither haplotype is 25%. An HLA-​
    fined as matched siblings (MSD) if they are HLA-​identical to identical sibling is considered an optimal donor.
    their recipients, as each donor and recipient inherited the same The HLA system has proven to be the most polymorphic genetic
    parental haplotypes. Regardless of their relationship, donor–​ region known in the human genome; nevertheless, the HLA genes
    recipient pairs are defined as HLA-​matched donors (MD) if they are closely linked and the recombination rate is therefore limited.
    are fully matched (10/​10 or 8/​8) or have a single allelic or anti- Unfortunately, the likelihood of identifying a URD decreases with
    genic disparity (9/​10 or 7/​8). Donors are defined as mismatched the requested degree of matching.
    donors (MMD) if they have two (8/​10 or 6/​8) or more allelic or
    antigenic disparities, up to a different haplotype. Any donor who Donor-​related issues
    is not an HLA-​identical sibling or a MD, as defined above, is to be Besides HLA compatibility and stem-​ cell source, donor age,
    considered a MMD. By definition, both an MD and MMD could gender, female parity, weight, ABO blood group, and viral sero-
    be either related or unrelated to their recipient, regardless of cell logical status should also be considered in the decision-​making
    source. process for donor selection, whenever more than one donor is
    available (which may not often be the case). Most studies report
    Cord-​blood matching that a young donor is better than an older one, and a male donor
    As a matter of fact, the concept of ‘compatibility’ for CB donor–​ is better for a male recipient and better than a multiparous woman
    recipient pairs is still under debate. In the past, any CB unit which for any recipient, even though there is no consensus on this issue.
    was 6/​6 or 5/​6 matched was labelled HLA compatible (MD), as de- Unfavourable weight disparity, with donors weighing less than
    fined by low-​resolution typing at A and B loci and high-​resolution their recipient, should be avoided, when possible. CMV (cyto-
    at the DRB1 locus; more recently, high-​resolution typing for at megalovirus)-​IgG-​positive and EBV (Epstein-​Barr virus)-​positive
    least A, B, C, DRB1 loci has been requested and, progressively, the patients should be grafted from CMV-​and EBV-​positive donors,
    same criteria used for volunteer donors has come to define CB HLA respectively. ABO matching is usually preferred, especially in-
    matching. stead of a major or even minor incompatibility. Donor location
    might also be considered, as overseas deliveries increase the time
    elapsing between collection and infusion, thus reducing cell via-
    Donor selection bility and potentially jeopardizing engraftment. More recently,
    killer-​cell immunoglobulin-​like receptor (KIR) genotyping has
    It is known that multiple factors have an impact on transplantation allowed identification of alloreactive donors who may help to pre-
    outcome; the most crucial ones are disease-​related (e.g. disease re- vent relapse.
    fractoriness and phase, clonal abnormalities in malignancies; dis-
    ease type and associated rejection risk in non-​malignant diseases) Hierarchy
    or patient-​related (e.g. age, comorbidities, infectious diseases/​colon- Even though it is mainly clear which variant should be preferred
    ization), but many are donor-​related. within each feature, there is no consensus regarding the hierarch-
    ical order by which the aforementioned factors should be com-
    Human leukocyte antigen compatibility bined. In a recent survey within the Paediatric Diseases Working
    The outcome of HSCT depends on the matching between the donor Party of the EBMT, the features were generally listed in the following
    and the recipient for the HLAs, encoded by a group of genes on order of importance, although evaluations widely differed among
    chromosome 6; genes and products are labelled as major histocom- responders: HLA compatibility, CMV serological status in case of
    patibility complex (MHC). The most relevant genes for transplant- positive recipients, stem-​cell source, donor age, donor gender, ABO
    ation belong to class I (HLA-​A, -​B, and -​Cw) and class II (HLA-​DR, major incompatibility, donor-​centre location, ABO minor incom-
    -​DQ, and -​DP). Class I specificities are present on all nucleated cells patibility (unpublished data).
    54 CHAPTER 7 Allogeneic Stem-Cell Transplantation

    was similar to children who grafted from matched siblings or


    Haploidentical transplant: history and
    matched URDs. Promising results with the same approach were re-
    recent development
    ported in children with non-​malignant diseases and confirmed by
    a multicentre phase I/​II study recently performed in children with
    Unmanipulated bone marrow
    various diseases.
    The first haploidentical transplantations, performed in the 1980s
    with unmanipulated BM, were complicated by severe and often le- In vivo T-​cell depletion
    thal side effects, such as pulmonary oedema and multi-​organ failure, An alternative approach to the ex vivo T-​cell depletion is the use
    most likely caused by severe GvHD. Attempts were then made to of unmanipulated BM (T-​replete BM), with or without PBSCs with
    remove T-​cells from BM grafts by means of soybean agglutinin extensive GvHD prophylaxis. Huang and colleagues have pioneered
    and rosette formation with sheep red-​blood cells in children with the use of granulocyte colony-​stimulating growth factor (G-​CSF)-​
    immunodeficiencies. primed BM, together with PBSCs, in adult patients. Grade II–​IV
    acute and chronic GvHD were observed in 78% and 73% of the pa-
    Ex vivo T-​cell depletion
    tients, respectively. The use of (G-​CSF)-​mobilized PBSCs alone was
    Subsequently, Henslee-​Downey and colleagues introduced partial inferior compared to the use of G-​CSF-​primed BM together with
    in vitro T-​cell depletion of BM grafts by means of anti-​CD3 anti- PBSCs. G-​CSF-​primed BM was also used in paediatric patients with
    bodies (T10B9 or OKT-​3) with a 1–​1.5 log T-​cell reduction and haematological malignancies, with a two-​year disease-​free survival
    post-​transplant pharmacological GvHD prophylaxis. Out of the 201 of 53%.
    patients treated, 98% engrafted, and the five-​year cumulative inci-
    dence of relapses was 31% and TRM was 51%. Post-​transplant cyclophosphamide
    A more recently developed approach of T-​replete transplantation
    CD34+ positive selection
    of either BM or peripheral blood (PB) is the use of post-​transplant
    The introduction of CD34+ cell positive selection methods from mo- cyclophosphamide (PT-​Cy) to induce tolerance of donor-​derived
    bilized PBSCs allowed the transplantation of high numbers of CD34+ T-​lymphocytes. In an initial study in 68 adult patients with haem-
    stem cells and the efficient depletion of T-​cells. Haploidentical trans- atological malignancies, cyclophosphamide (50 mg/​kg) was given
    plantation was performed in adults and children without, or with on day three, or days three and four, after transplant. Graft failure
    only minimal, pharmacological GvHD prophylaxis. However, the was seen in 13% of the patients and cumulative incidence of grade
    immune reconstitution was delayed, resulting in viral and bacterial II–​IV, grade III–​IV acute GvHD, and chronic GvHD was 34%, 6%,
    infections. and 22%, respectively. The two-​year event-​free survival was 26%.
    With the CD34+ cell positive selection, no other cells are co-​ Based on the feasibility of this approach, PT-​Cy is increasingly used,
    transplanted, whereas with negative depletion strategies, large num- mainly in adult patients. Less experience exists in paediatric pa-
    bers of other cells (e.g. NK cells, monocytes, dendritic cells) are tients, but reports are promising so far. Whether the risk of relapse
    infused together with high numbers of CD34+ stem cells. Therefore, in children with haematological malignancies will increase due to
    innovative methods were developed to negatively deplete T-​cells heavier immune suppression needs to be assessed in future pro-
    (and also B-​cells) from PBSCs. spective clinical studies.
    CD34+ negative selection Summary
    Depletion of CD3+ T-​cells enabled a T-​cell depletion of 3.5–​4 log. Overall, haploidentical transplantation in children has evolved as a
    Various clinical studies using this method were performed in chil- realistic alternative for patients in need of a transplant lacking a suit-
    dren and in adults. The engraftment was rapid and an accelerated able donor. In children, haploidentical donors are mainly three-​loci
    recovery of natural killer (NK) cells was observed. The rate of GvHD mismatched parental donors, but haploidentical relatives can also
    was acceptable and the risk of infectious complications was lower in be included in the rare cases where parents are not medically fit to
    children compared to adult patients. donate.
    Since the efficacy of the CD34+ negative selection (i.e. CD3+ Currently, the choice between T-​depleted versus T-​repleted grafts
    depletion method was 1-​ log lower compared to the previous depends on the experience of the transplant centre with each dif-
    CD34+ positive selection methods), attempts were made to im- ferent method, while a prospective study comparing the two dif-
    prove the negative depletion efficacy by the removal of TCRαβ+ T-​ ferent approaches would be desirable.
    lymphocytes. TCRαβ+ depletion yields a 4–​4.5 log CD3+ depletion,
    comparable to the CD3 depletion which can be achieved with the
    CD34+ positive selection method. In addition, non-​alloreactive γδ Stem-​cell source
    T-​lymphocytes remain within the graft, which might induce anti-​
    infectious and anti-​leukaemic effects. As for adult patients, three different stem-​cell (SC) sources may be
    used for paediatric HSCT: BM, PBSC obtained after donor mobiliza-
    TCRαβ depletion
    tion with G-​CSF, and CB—​or a combination of the three.
    An improved immune recovery after TCRαβ-​depleted stem cells
    from haploidentical donors in paediatric patients was reported by Bone marrow
    Lang and colleagues. In another study in children with acute leu- For many years and for historical reasons, BM was the most
    kaemias, the outcome of TCRαβ-​depleted stem cell transplantation common source used in the allograft setting and still remains the
    Conditioning regimens 55

    ‘gold standard’ in paediatrics, as it offers the most consolidated re- Consortium recommendations, unrelated CB with two or less HLA
    sults. The graft should contain at least 3×108 nucleated cells/​kg re- disparities typed in low resolution (i.e. two-​digit) for class I (A and
    cipient body weight and 3×106 CD34+ cells/​kg body weight, in order B loci) and high resolution (i.e. four-​digit) for class II (DRB1 locus),
    to yield more than 95% neutrophil engraftment chances at a median and with more than 2.5×107 nucleated cell dose/​kg or 2×105 CD34+
    of 21 days in the setting of haematological malignant diseases. cells/​kg are suitable for engraftment. Recent studies from Eurocord,
    Netcord, EBMT, and the Center for International Blood and Marrow
    Peripheral blood Transplant Research (CIBMTR) recommend high-​resolution HLA
    Mobilized PB as a SC source has become more and more common typing for A, B, C, and DRB1, and a maximum of one or two mis-
    over the years (initially in adults, and eventually in children) mainly matched loci with a cellularity of 3×107 TNC/​kg or higher.
    due to its higher cell content, especially CD34, allowing faster Two prospective studies could demonstrate no benefit of double
    engraftment. CB in paediatric patients transplanted for malignant diseases.

    Haploidentical setting Minor sibling donors


    PBSCs are also used in the setting of T-​cell-​depleted haploidentical Paediatricians have to take into special consideration minor sib-
    transplantation, after the concept of mega-​cell dose, in order to over- ling donors. As for any transplantation from living donors, it is
    come HLA disparity and ensure engraftment. mandatory that the healthy donor’s safety should remain the most
    important issue of the procedure. However, minor donors are par-
    Impact of stem-​cell source on graft-​versus-​host disease ticularly fragile, as they do not really feel free regarding donation
    and relapse acceptance. Physicians have to ensure donor autonomy and both
    The possible increased risk of both acute and chronic GvHD after physical and psychological integrity as much as possible.
    PBSC transplantation, as compared to BM transplantation, is still The American Academy of Pediatrics Committee on Bioethics
    debated. In a recent European retrospective study, including 2,584 (AAPCOB) deems the use of a child BM donor ethically appropriate
    paediatric patients transplanted from 2003 to 2012 for acute lympho- when five conditions are met: no available matched adult donor;
    blastic leukaemia, both TRM and chronic GvHD appeared signifi- reasonable expected benefit for recipient; strong relationship be-
    cantly higher after PBSC, as compared with other SC sources, despite tween donor and recipient; medical, emotional, and psychological
    that overall survival (OS) was similar for both stem-​cell sources. In risk to the child donor minimal and reasonable in comparison to the
    the prospective ALL-​SCT-​BFM 2003 Study chaired by Peters, the benefit gained by the recipient; and, finally, parent consent and pa-
    same OS was reported, and no difference could be demonstrated in tient assent. This last point is the most difficult ethical issue. In order
    TRM, GvHD, and relapse, whichever the stem-​cell source in the two to ensure it, some countries or institutions provide independent
    cohorts of patients transplanted from HLA-​identical siblings and committees to interview and evaluate minor donors.
    other matched donors. Nevertheless, within patients transplanted
    from HLA-​identical siblings, the cumulative incidence of chronic
    GvHD was higher in PB compared with BM recipients.
    Nowadays, in the ongoing prospective ALL-​I-​BFM HSCT trial
    Conditioning regimens
    (FORUM), the algorithm for choosing stem-​cell source recom-
    mends BM as the first choice. Reinforced GvHD prophylaxis may be
    The struggle to reduce the use of radiotherapy
    recommended when PBSC are used, mainly when no serotherapy is in children
    included as for GvHD prophylaxis, as in most protocols in the HLA-​ Total body irradiation (TBI) has been at the heart of conditioning re-
    identical sibling setting. gimens for malignancies in both adults and children since the begin-
    To date, there is no demonstration for a better graft- ning of the HSCT era. Despite its efficacy, the burden of long-​term
    ​versus-​leukaemia effect after either BM or PBSC in the paediatric side effects, specifically in the paediatric population, has been a con-
    population. cern for transplanters all along. While fractionating radiation doses
    and lung shielding has proven efficacious in reducing, from 50% to
    T-​cell replete haploidentical graft 4%, the incidence of pulmonary complications, recent data confirms
    Recently, T-​cell repleted haploidentical transplantation was devel- that TBI exposure is implicated in the long-​term post-HSCT burden
    oped. So far, there are no available data to advocate for using BM or of diseases. Indeed, TBI significantly increases the cardiovascular
    PBSC as the preferred stem-​cell source, particularly in the paediatric risk in long-​term transplant survivors, by increasing the risk of
    population transplanted for malignant diseases. metabolic syndrome, specifically of hypertriglyceridaemia, hyper-
    cholesterolaemia, and insulin resistance. A strong association is also
    Cord blood known between cataracts and TBI, with 28–​78% of exposed patients
    From the first CB transplantation performed for a Fanconi anaemia presenting with this complication. In terms of endocrine dysfunc-
    patient in 1987, CB has appeared as a useful and efficient stem-​cell tion, a clear role for TBI versus alkylating agents such as busulfan
    source, due to two major features: high proliferative capacity (al- (Bu) could not be found to support long-​term thyroid impairment
    lowing engraftment despite 1-​log fewer cells) and immune plasticity as well as ovarian failure/​azoospermia, but TBI is definitely impli-
    (allowing a wider HLA disparity within each donor–​recipient pair). cated in growth failure. Indeed, not only is growth hormone (GH)
    The possibility to adopt less stringent HLA matching criteria en- production reduced post TBI but also the growth spurt is absent due
    larged the availability of grafts to at least 90% of the paediatric pa- to low GH in combination with low sex hormones, and growth plates
    tients in need of an allogeneic transplant. According to Eurocord in long bones undergoing premature closure after TBI exposure.
    56 CHAPTER 7 Allogeneic Stem-Cell Transplantation

    With regard to lung function, it is unclear whether fractionated study will be crucial to shape further HSCT ALL treatment, it has
    TBI plays a worse role than BU in causing restrictive lung func- to be remembered that the role of TBI in the context of mismatched
    tion. Moreover, the specific role of TBI in undermining the cogni- or haploidentical transplants is even more difficult to understand,
    tive functions of transplanted patients has been confirmed by some, with low patient numbers and numerous differences in conditioning
    but not by other, studies. In this regard, it is important to underline regimen and HSCT approach jeopardizing the possibility of mean-
    that all studies conclude that socioeconomic status, as well as pre- ingful comparisons.
    HSCT IQ, are the most important predictive factors for long-​term
    cognitive functions in patients undergoing HSCT for malignant dis- Reduced-​intensity conditioning to reduce toxicity post
    orders. Dramatically though, secondary malignancies appear to be stem-​cell transplantation
    the most serious long-​term effect of HSCT, with cranio-​spinal ir- The traditional aim of conditioning regimens in the malignancy set-
    radiation being a significant risk factor. With an incidence of second ting had been that of eradicating residual disease and guaranteeing
    tumour of up to 11% at 15 years post-HSCT and an age-​dependent complete marrow ablation to allow donor cell engraftment. Data
    factor (patients receiving HSCT aged 10 years or younger are at a showing the role of donor T-​cells in mediating graft-​versus-​tumour
    significantly higher risk of second tumour than older patients), this effect, and preclinical experiments documenting the possibility of
    alone represents a significant reason to try and avoid TBI as much as engraftment despite reduced myeloablation, paved the way for the
    possible in the preparation regimen of children. development of so-​called ‘reduced-​intensity conditioning’ (RIC).
    This definition underpins numerous conditioning regimens, the in-
    Acute myeloid leukaemia tensity of which varies considerably from minimal to little less than
    Thomas and colleagues were the first to describe the use of TBI for myeloablative. Overall, a consensus had been reached in 2009 with
    AML and reported a reduced toxicity by using a fractionated TBI regard to the definition of RIC, which is still valid, despite novel
    schedule. Santos firstly described the use of Bu-​based regimens asso- agents entering the chemotherapy field.
    ciated with Cy 200 mg/​kg. This regimen showed low cumulative inci- Since the 1990s, the use of RIC regimens has been increasingly
    dence of relapse (CIR) but increased TRM, which could be corrected popular for older recipients with comorbidities which prevent the
    by reducing the Cy dose to 120 mg/​kg as described by Tutschka safe use of conventional myeloablative regimens. The intrinsic nature
    and colleagues. Blaise and Ringden published two randomized of RIC regimens has seen them being allocated to more vulnerable
    trials, which compared TBI-​Cy120 versus Bu-​Cy120, and showed patients, making direct comparisons between RIC and myeloablative
    higher non-​relapse mortality (NRM) and higher CIR in the Bu conditioning (MAC) regimens difficult. As children conventionally
    arm. Nevertheless, further updates on these studies provided longer tolerate higher treatment exposures, the use of RIC in the paedi-
    follow-​up and failed to document a worse outcome in Bu-​treated pa- atric population has remained quite limited so far. The possibility
    tients. De Berranger and colleagues also concluded that there was no of expanding its use with the aim of limiting long-​term post-HSCT
    advantage of one regimen over the other when considering only a sequelae is nevertheless appealing. Early data which still need to be
    paediatric population with AML in CR1 at HSCT. A more recent pro- confirmed and have been gathered in the context of non-​malignant
    spective study from Ishida and colleagues confirmed no significant disorders, showed better gonadal function following RIC regimens,
    differences in outcome in children with AML in CR1/​CR2 receiving with hope for fertility preservation. It appears that thyroid dysfunc-
    TBI-​Cy or IV Bu-​Cy. Copelan and colleagues documented lower tion, as well as growth impairment, might remain significant, despite
    TRM and lower CIR in patients with AML in CR1 treated with IV avoiding myeloablation, but the substantial question will be whether
    Bu-​Cy as compared to those treated with TBI-​Cy, and lower CIR as secondary malignancies could be significantly reduced after RIC.
    compared to both TBI-​Cy and oral Bu-​Cy treatment. Overall, these Data obtained so far explore children mainly transplanted for
    data have resulted in a markedly reduced use of TBI-​based condi- non-​malignant diseases, which may be influenced by genetic factors
    tioning regimens in the AML paediatric population. as well. Data on the use of RIC regimens for malignancies are rare in
    paediatric patients.
    Acute lymphoblastic leukaemia
    In the context of ALL, it has been less easy to outplay TBI. Standard Reduced-​intensity versus myeloablative conditioning
    use of TBI combined with VP16 or Cy has been challenged by in paediatric acute myeloid leukaemia
    prospective and randomized studies comparing Bu-​ containing CIBMTR ran a case-​control study on the use of RIC versus MA
    regimens with no benefit in terms of relapse rate and increased protocols for children receiving their first HSCT for AML. Case
    transplant-​related toxicities in the non-​TBI arm. Moreover, some and controls were matched for age, disease status, cytogenetics,
    cooperative groups showed a correlation between increasing the graft type, and year of HSCT. Thirty-​nine patients receiving RIC
    TBI dose and reducing relapse rate in paediatric and adult patients. were matched with 141 receiving MAC protocols. Not surpris-
    Given the small proportion of ALL-​affected patients that need ingly, patients receiving RIC displayed a lower performance score.
    HSCT as a standard treatment, the implementation of an inter- Overall, the two groups showed no statistically significant differ-
    national randomized study was deemed necessary to finally an- ences in terms of count recovery (RIC 87% versus MAC 84%),
    swer the question of whether a TBI-​containing regimen is superior GvHD incidence (28% versus 39%), TRM (16% in both cohorts),
    to chemotherapy-​alone regimens in the treatment of ALL in chil- relapse rate (39% in both cohorts), leukaemia-​free survival (LFS)
    dren. Such a study is currently open for recruitment in 20 coun- (46% in both cohorts), and OS (45% versus 48%). After adjusting
    tries worldwide, and aims at comparing the standard VP16 TBI for disease status at HSCT, cytogenetic risk group, graft type, and
    12-​Gy conditioning regimen with Fludarabine/​Thiotepa/​IV Bu or HSCT year, the only characteristic associated with higher LFS and
    Fludarabine/​Thiotepa/​Treosulfan regimens. While results of this OS was the patient performance score. It has to be underlined that
    Acute lymphoblastic leukaemia 57

    these are retrospective, registry data, not accounting for MRD no active studies investigating the use of RIC as primary condi-
    status. It has also to be noted that these results are different from tioning for ALL in children.
    the ones repeatedly reported in adult cohorts, in which the use of
    RIC for AML has often resulted in increased relapse rate but re-
    duced TRM, thus achieving a similar OS to MAC-​based HSCT, Indications for stem-​cell transplantation
    where toxicity was higher but relapse rate lower.
    A recent meta-​analysis of available data in this setting correctly The definition of the indications for HSCT in childhood is a dy-
    underlined that in all previous studies, apart from a single random- namic process, requiring constant assessment of the likelihood of
    ized one, patients undergoing RIC regimens were older, and had a cure with chemotherapy only, in order to identify the subsets of
    more advanced stage of disease and potentially lower performance children who would benefit from transplantation. Eligibility criteria
    scores than those allocated to MAC conditionings, thus leaving have evolved over time, due to improvements in treatment outcome
    open the question of RIC versus MAC in a young cohort of patients. obtained with chemotherapeutic protocols and changes in trans-
    There are currently two ongoing prospective trials in chil- plantation outcome.
    dren assessing the use of RIC in AML (NCI NCT02724163 and Randomized controlled clinical trials represent the gold standard
    NCT02626715), and one of those is a randomized trial which aims by which treatment choices are made in modern medicine. Paediatric
    at comparing standard busulfan-​cyclophosphamide association to oncology has, par excellence, been one of the specialties that have
    busulfan-​fludarabine association in AML CR1 high risk for MRD adopted the approach of randomized controlled trials. Nevertheless,
    and/​or cytogenetics. to date, the evidence upon which a decision is made to transplant a
    child or not has rarely been based on randomized trials. Reliance on
    Reduced-​intensity versus myeloablative conditioning data collected by retrospective studies, observational databases (e.g.
    regimen in acute lymphoblastic leukaemia EBMT), or prospective non-​randomized studies mainly identifies
    Given the reduced evidence of a graft-​versus-​leukaemia effect in the current role of HSCT in paediatrics.
    rapid-​evolving diseases like ALL, less experience has been gathered The most frequent indication for HSCT in paediatrics is
    on the use of RIC in this setting. Specifically, for paediatric patients acute leukaemia. Other haematological malignancies, including
    receiving their first HSCT, two retrospective studies are available to myelodysplastic syndrome, non-​Hodgkin lymphoma, chronic mye-
    describe the use of RIC conditioning. Verneris and colleagues col- loid leukaemia, and other myeloproliferative neoplasms, may be
    lected data on 38 patients transplanted between 1995 and 2005 in treated with HSCT when they are refractory to conventional therapy.
    several American centres. Median age at HSCT was 12 years; 22% of
    the patients had active disease at HSCT and different RIC regimens
    were used. This study observed three-​year disease-​free survival Acute lymphoblastic leukaemia
    (DFS), in line with other paediatric studies using MAC, but signifi-
    cant three-​year TRM of 32%, which is higher than the TRM reported ALL is the most common neoplastic disease in childhood, with an
    in the only prospective study assessing the use of RIC in malignan- incidence of four new diagnoses per year out of 100,000 children
    cies for children not eligible to MAC transplant. The increased TRM below 18 years of age.
    in this study has been attributed to the possible non-​homogeneity Almost 50 years have passed since the first reports that childhood
    of RIC used and to transplants performed in an earlier era than the ALL was no longer incurable. The improved outcome of paediatric
    Pulsipher study. Moreover, no clear identification of the reasons for ALL is one of the greatest successes in modern medicine, with a cure
    allocating patients to RIC regimens is described in this study, though rate which has risen from 3% in the 1960s to the current 85%, at
    no patient was reported to have significant comorbidities at the time least in developed countries. These advances are the result of basic
    of HSCT. research, biological investigations, large clinical trials, and intensive
    The Japanese Society for HSCT reported recently on their retro- collaboration of cooperative groups. Unfortunately, the remaining
    spective experience on the use of RIC in paediatric ALL and its com- 15% of children relapse despite appropriate treatment.
    parison with MAC: 1,201 patients undergoing MAC were compared In 2005, a cooperative prospective study, including 357 patients,
    with 133 patients receiving RIC between the year 2000 and 2010, demonstrated the benefit of HCT from HLA-​identical siblings as
    with a median follow-​up of five years post-HSCT. Most patients compared to further intensified chemotherapy in improving the
    were in CR1 at the time of transplant, with disease phases being outcome of the commonly defined ‘very high risk’ (VHR) childhood
    equally represented in RIC and MAC cohorts. In the RIC setting, a ALL patients in first complete remission (CR1). At the time, VHR
    higher percentage of patients received HSCT from a MMD. In this features were defined by the presence of at least one of the following
    context, relapse rate was found to be higher in patients receiving RIC criteria:
    (43% versus 33%; p = 0.02), while TRM rate was superimposable in
    1. failure to achieve CR after the first four-​drug induction phase;
    both cohorts at 15%, and OS was 52% for RIC patients and 56% for
    2. t(9;22) or t(4;11) clonal abnormalities;
    MAC patients (p = ns). In the multivariate analysis—​adjusted for
    3. prednisone poor response (PPR) if associated with T
    age, gender, disease status, stem-​cell source, conditioning regimen,
    and HLA matching—​the conditioning regimen was no longer a sig- immunophenotype and/​or white blood-​cell count equal to or
    nificant prognostic factor for relapse. greater than 100×109/​L.
    Overall, while ALL-​affected paediatric patients can be allocated to Patients were randomized by ‘genetic chance’, according to the avail-
    RIC regimens in case of unsuitability for MAC, there are currently ability of a compatible related donor or not. According to intention
    58 CHAPTER 7 Allogeneic Stem-Cell Transplantation

    to treat, the five-​year DFS was 40.6% (SE 3.1) and 56.7% (SE 5.7), that outcomes of transplantation from HLA-​identical sibling and
    (hazard ratio 0.67, p = 0.02), and the five-​year survival was 50.1% matched URD are comparable in paediatric ALL, even though there
    (SE 3.1) and 56.4% (SE 5.9) (p = 0.12) in the ‘No donor’ and ‘Donor’ are no randomized trials nor studies on the ‘intention-​to-​treat’ con-
    arms, respectively. The availability of a compatible related donor was cept in this regards.
    associated with a better outcome, particularly for patients with the According to the prospective multicentre BFM ALL SCT 2003
    worst prognostic criteria (as the subset with ‘induction failure’), and Study, held in Austria, Germany, and Switzerland, the outcome of
    less evident for PPR-​only patients. transplantation from MD was not inferior to the outcome of trans-
    In the last decade, the detection of minimal residual disease plantation from MSD in terms of EFS (67% versus 71%, p = 0.05)
    (MRD) in childhood ALL turned out to be the most sensitive method and CIR (22% versus 24%, p = 0.4), with NRM (10% versus 3%,
    to evaluate treatment response and one of the strongest predictors of p = 0.017) being higher for the former, as compared with the latter.
    outcome. The inclusion of MRD into survival models decreased the The role of donor type and compatibility in a broader network of
    predictive value of any other prognostic feature, to the extent that it countries has been evaluated within the International BFM Study
    overtook most ‘biological’ high-​risk features. Group (I-​BFM-​SG) ALL SCT 2007 Protocol, run in ten countries,
    which confirms the non-​inferiority of MD versus MSD HSCT, in
    Haematopoietic stem-​cell transplantation in first terms of EFS, OS, CIR, and NRM. Therefore, URD search acti-
    complete remission vation and transplantation might be recommended in the future
    Nowadays, transplantation is recommended for less than 10% of for virtually every child for whom an allogeneic transplantation
    paediatric patients with ALL at presentation. Within the AIEOP-​ is indicated.
    BFM ALL 2009 Protocol, mainly patients having MRD levels greater Depending on the experience of each transplant centre, MMD
    than or equal to 5×10-​4 after the second induction element are con- might be preferred, carrying the advantage of prompt donor
    sidered VHR and are therefore eligible for HSCT. Patients who fail availability and flexible schedules, and bringing a higher degree
    to achieve CR after the first induction phase, carry the t(4;11) clonal of alloreactivity, potentially associated with lower relapse risk.
    abnormality, or experience PPR in T immunophenotype ALL are HSCT from MMD is widely recommended in advanced diseases
    currently considered VHR only if their MRD level is greater than or and relapses after transplant, when timing adjustment is crucial.
    equal to 5×10-​4 after the first induction element or whenever their Disparities within donor and recipient pairs are progressively ac-
    MRD is not available. cepted as the risk profile of the patient worsens.

    The t(9;22) translocation Outcome according to minimal residual disease


    Patients carrying the t(9;22) translocation are treated with chemo- at haematopoietic stem-​cell transplantation
    therapy associated with tyrosine-​kinase inhibitors. Within the up- MRD status before and after HSCT is strongly associated with the
    dated ESPHALL trial, ‘Ph-​positive’ patients are considered at VHR outcome of paediatric patients undergoing allogeneic transplant-
    when the MRD level is greater than or equal to 5×10-​4 after the ation for ALL. HSCT in MRD-​positive patients is significantly in-
    second element of induction or is positive at any level after the three ferior in terms of both EFS (around 25%) and cumulative incidence
    high-​dose blocks. of relapse (between 40% and 60%) compared with patients with
    MRD that is negative or less than 1×10-​4, for whom an EFS of more
    Infants than 70% and a CIR of less then 20% are reported.
    Infant patients are defined as VHR if they present with two or more Whether additional intensified treatment with either chemo-
    of the following features: six months old or younger at diagnosis; therapy (i.e. fludarabine-​cytarabin +/​–​daunorubicin association)
    carry mixed leukaemia lineage (MLL) gene rearrangement; present or innovative strategies (i.e. blinatumomab) given to patients with
    with a white blood-​cell count greater than 300×109/​L and/​or PPR; a pre-​transplant MRD greater than or equal to 1×10-​4 might be ef-
    MRD levels greater than or equal to 10-​4 at the start of consolidation. fective in improving outcome is still controversial. Early tapering of
    immunosuppression when HSCT is performed in MRD-​positive
    Haematopoietic stem-​cell transplantation in second patients may reduce the risk of relapse, and the timing of +55 to
    complete remission +200 days after transplant has been identified as the optimal window
    HSCT is recommended for all patients relapsing in the BM early to initiate intervention to prevent relapse. In a 2012 study, Pulsipher
    (less than 30 months) after diagnosis and generally for the majority demonstrated the combined effect of GvHD and MRD status in a co-
    of the patients who experience relapse, depending on early MRD hort of 144 patients, suggesting a role for the graft-​versus-​leukaemia
    response after initial relapse treatment. effect.
    Within the ongoing IntReALL Protocol, patients randomly allo- MRD positivity after transplantation does not necessarily imply
    cated to arm A (previous ALL-​REZ BFM 2002) are eligible for HSCT relapse, but it is well demonstrated that ALL patients experiencing
    if MRD after induction (day one of week five) is greater than or equal any MRD positivity after transplant have a worse outcome com-
    to 10-​3, while those allocated to arm B (previous UK ALLR3) are eli- pared with patients for whom no MRD could be detected in the
    gible for HSCT if MRD after induction (day 1 of week 6) is greater post-​transplant course. Whether the dismal prognosis of MRD-​
    than or equal to 10-​4. positive patients could be reset by immunomodulatory strategies
    to enhance a graft-​versus-​leukaemia effect, such as early tapering/​
    Donor type according to patient risk profile sudden discontinuation of immunosuppression and donor
    In the last decade, results of allogeneic HSCT from unrelated donors lymphocyte infusion (DLI) or other cell therapies, is still to be
    has improved to the extent that there is an increasing agreement assessed.
    Solid tumours 59

    MRD measurement by flow cytometry or molecular techniques the AML02 Study, HSCT offered an advantage over chemotherapy
    yields consistent impact on outcome. Ideally, a more sensitive and when high risk was defined by MRD rather than by cytogenetics.
    specific method for identification of patients with impending re- Nowadays, the current general approach to HSCT for AML in
    lapse is desirable; next-​generation sequencing might unravel the CR1 is based on a stratification that accounts for cytogenetics, mo-
    conundrum in the near future. lecular characteristics of the disease, and post-​induction MRD.
    A classification combining all these features was proposed by
    Relapse after transplant Rubnitz in 2017.
    The prognosis of patients who relapse after transplant is dismal. Relapsed AML remains the most common indication for allo-
    Nowadays, innovative tools such as chimeric antigen receptor T-​cells, geneic HSCT. After relapsing, approximately 65% of paediatric AML
    bispecific T-​cell engagers, and advanced conjugated immunotoxins patients will achieve a CR2 and their outcome, post HSCT, will yield
    offer further chances of cure. Hopefully, tools to foresee impending a survival ranging between 10% and 50% depending on the patient
    relapse will gain better sensitivity and specificity, so that identi- risk factors. Key prognostic features in relapsed AML remain the
    fied patients will be suitable candidates for MRD-​driven targeted time from diagnosis to relapse, cytogenetics risk group, disease re-
    therapy. sponse to re-​induction treatment, and previous HSCT.
    For HSCT in CR2, MDs (both family and unrelated) have been
    used with superimposable results, and even HSCT from MMDs
    Acute myeloid leukaemia has given good results in the absence of adequate MDs. Noticeably,
    recent developments in the haploidentical transplant setting have
    HSCT has been the cornerstone of consolidation treatment for AML provided extremely positive outcomes when compared to matched
    in children according to many collaborative groups. The 1990s were URD transplants, in prospective single-​centre trials. Further studies
    dominated by allocation to allogeneic HSCT versus autologous are required to confirm those promising reports.
    HSCT or further consolidation chemotherapy, depending on the The role of MRD pre HSCT in CR1 and CR2 has been investigated
    availability of a matched sibling donor and/​or the cytogenetic risk by few studies. In the St Jude AML02 trial, MRD pre HSCT showed a
    group of the patient. More recently, the detection of molecular ab- limited impact on the post-HSCT prognosis, not reaching statistical
    normalities in AML blasts allowed further refinement in the risk-​ significance. The contrary has been proven in adult studies, where
    group definition. MRD risk could stratify patients pre HSCT and define long-​term
    With this approach, data progressively showed no benefit of au- prognosis. Further paediatric data are warranted to understand the
    tologous HSCT over chemotherapy, thus causing the progressive impact of pre HSCT MRD.
    abandonment of autologous HSCT as a consolidation strategy. On Patients with primary induction failure have been documented to
    the contrary, indications for allogeneic HSCT in CR1 remained con- benefit from HSCT as opposed to chemotherapy in prospective col-
    troversial. Multiple studies suggested a reduction in the CIR for the laborative studies, with a potential EFS around 30%, depending on
    HSCT recipient, but this did not translate into better OS, given the associated risk factors. HSCT in patients with active disease, either
    high TRM. Moreover, the distinction between intermediate-​risk primary refractory or relapsed refractory disease, has been investi-
    and high-​risk groups, as per cytogenetics/​molecular abnormalities, gated by few studies. Blast counts and age at HSCT in these groups of
    remained non-​homogeneous among groups, hampering the possi- patients have proven to be helpful in selecting a subgroup of children
    bility of comparisons across studies. with extremely dismal prognosis who might therefore be allocated
    Kasper, in 2007, showed that by using cytogenetic-​based classifica- to experimental treatments in the attempt to reduce the blast burden
    tion, different study groups would allocate 15–​80% of their patients pre HSCT.
    to HSCT in CR1. When disease stratification according to cytogen- With regard to the type of conditioning regimen recommended
    etics was taken into account, few retrospective studies were able to for AML in CR1, the use of TBI has shown no benefit over chemo-
    document a real benefit of HSCT over chemotherapy, although not therapy in multiple studies and should be abandoned given the well-​
    all the risk groups were equally well represented in those studies. known, long-​term consequences of irradiation in children. Busulfan
    While HSCT techniques, as well as supportive therapy, has pro- remains the most commonly used agent in conditioning, and its
    gressively improved over time, allowing a reduction in TRM and im- association with cyclophosphamide has been documented as effi-
    provement of OS for HSCT recipients, the need of exposing children cacious by many studies. A further increase in the conditioning in-
    affected with AML to the long-​term toxicities of allogeneic HSCT is tensity by adding a third alkylating agent has been proven efficacious
    still debated. Moreover, a significant number of patients relapsing in reducing CIR, but often at the cost of increased TRM (specific-
    post the first AML diagnosis can achieve a second remission and ally if used in older patients or in CR2). The use of clofarabine and
    proceed to HSCT in CR2, while disease relapse post HSCT in AML fludarabine as alternative agents in AML conditioning is becoming
    remains very hard to address. The introduction of MRD monitoring, more frequent and deserves attention.
    either with flow cytometry or molecular polymerase chain reaction
    (PCR), might represent a key factor in harmonizing the use of allo-
    geneic HSCT for patients in CR1. Solid tumours
    So far, two prospective studies have shown MRD to be the single
    more significant prognostic factor to predict children at high risk Currently, there are no standard indications for allogeneic trans-
    of relapse. Both in the St Jude AML02 Study and in the NOPHO plantation in children with solid tumours. However, there are some
    AML 2004 Study, MRD positivity at the end of induction identified hints that allogeneic transplantation might play a future role in the
    a group of patients at high risk of disease recurrence. Significantly, in treatment of refractory/​relapsed paediatric solid tumours.
    60 CHAPTER 7 Allogeneic Stem-Cell Transplantation

    Allogeneic HSCT led to the cure of an incurable patient after a of the reconstituting donor-​derived NK cells. This approach is feas-
    non-​myeloablative preparative regimen and tumour regression was ible with low TRM (less than 5%) and shows promising clinical effi-
    associated with acute and chronic GvHD. A similar observation was cacy. However, longer follow-​up is needed to draw final conclusions.
    made in a patient with relapsed disseminated Ewing sarcoma, who Another interesting approach could be the exploitation of the
    received an haploidentical transplantation and experienced impres- anti-​tumour effect of γδ T-​lymphocytes in the setting of allogeneic
    sive tumour regression, even of a cerebral metastasis in association HSCT. These lymphocytes are non-​alloreactive and their spontan-
    with the development of chronic GvHD. Interestingly, this patient is eous cytotoxicity and their ADCC against solid tumours can be ac-
    still alive and tumour-​free 18 years after transplantation. tivated in vitro with bisphosphonates and Interleukin 2 (IL-​2). It has
    In a retrospective analysis of allogeneic HSCT in paediatric pa- been reported that those children with haematological malignancies
    tients with advanced rhabdomyosarcoma, the outcome of 30 pa- who received an haploidentical graft depleted of TCRαβ+ T-​cells
    tients who received llogeneic HSCT from a donor who was matched and three or more infusions post-​transplant of the bisphosphonate
    sibling (n = 17), matched unrelated (n = 6), or one or more allele zoledronate had better survival compared to those who received
    mismatch (n = 7) was analysed. Ten patients received MAC, whereas only one or two infusions, suggesting that the in vivo activation of
    20 patients were conditioned with a RIC regimen. Three-​year OS the γδ T-​lymphocytes improved survival. Whether similar results
    was 20% with a median survival time of 12 months. The cumulative could be obtained in patients with solid tumours should be the focus
    risk of progression and TRM was 67% and 11%, respectively. A RIC of future clinical trials.
    regimen was also given to paediatric and young adults with solid tu- Prompt donor availability in haploidentical transplantation
    mours (n = 20), who received haploidentical BM transplantation fol- would, especially, allow the post-​transplant adoptive transfer of
    lowed by post-​transplant cyclophosphamide. OS was 88%, 56%, and donor-​derived NK cells, cytokine-​induced killer (CIK) cells, γδ T-​
    21% at 6, 12, and 24 months, respectively. Paillard and colleagues lymphocytes, and also of tumour-​specific TCRαβ+ T-​lymphocytes.
    also transplanted 20 patients with solid tumours with a RIC regimen. Tumour-​associated antigens, such as NY-​ESO-​1 which is widely
    Three-​year EFS was 28% for neuroblastoma and 22% for sarcoma expressed on solid tumours, could be a suitable target for T-​cells,
    patients. Earlier studies of allogeneic HSCT in neuroblastoma did and large-​scale clinical methods have been established to isolate
    not show an advantage of allogeneic HSCT versus autologous HSCT. NY-​ESO-​1-​specific T-​cells from healthy donors. In the autologous
    With current conventional transplant approaches, the outcome setting, objective clinical responses were observed in four out of
    in disseminated sarcoma is still poor, and more research and pro- six patients with metastatic synovial cell sarcomas after adoptive
    spective clinical trials are necessary to fully exploit the graft-​versus-​ transfer of autologous T-​cells transduced with a T-​cell receptor
    tumour (GvT) effect of allogeneic transplantation. against NY-​ESO-​1. Since NY-​ESO-​1-​specific T-​cells can be found
    in healthy persons, adoptive transfer of donor-​derived NY-​ESO-​1-​
    Strategies to exploit the graft-​versus-​tumour effect specific T-​cells is possible, and future clinical trials are warranted
    in solid tumours using such an approach.
    Currently, no clear strategies are available to separate GvHD from It is currently not known whether checkpoint inhibitors (PD-​1/​
    GvTD. However, allogeneic HSCT could be a platform for further PD-​L1 blockade) would have an additional impact on such T-​cell-​
    post-​transplant immunotherapeutic approaches, as the use of cel- based approaches or whether they would, rather, increase the risk of
    lular therapy and/​or the application of antibodies against suitable GvHD as it has been observed in patients with relapsed lymphoma
    target antigens. Haploidentical transplantation with ex vivo T-​cell in whom PD-​1 blockade post-​allogeneic HSCT resulted in a high re-
    depleted grafts without post-​ transplant pharmacological GvHD sponse rate but also in rapid onset of severe and treatment-​refractory
    prophylaxis and the adoptive transfer of donor-​derived NK cells GvHD. Prospective clinical trials are necessary to investigate the
    has been performed in children with refractory solid tumours. In role of checkpoint inhibitors post transplant in patients with solid
    this preliminary study, donor NK cells were stimulated in vitro with tumours.
    Interleukin 15 (IL-​15) and adoptively transferred to the patient. Four In summary, allogeneic HSCT and the establishment of a donor-​
    patients died with progressive disease and two of TRM. It remains to derived immune system can provide a platform for various post-​
    be shown in future clinical trials whether the in vivo application of transplant immunotherapeutic strategies for patients with otherwise
    IL-​15, such as the more recently described IL-​15 superagonist com- non-​ curable solid tumours, and prospective clinical trials are
    plex, comprised of an IL-​15N72D mutation and IL-​15RαSu/​Fc fu- warranted.
    sion (ALT 803), might induce GvTD without GvHD in humans, as it
    has been reported in mice.
    These strategies aim at the activation of a newly established donor-​ Post-​transplant immunotherapy
    derived NK-​cell system, and favourable NK-​cell activity has been re-
    ported after haploidentical transplantation in patients with relapses Chimerism monitoring post HSCT has been established as a useful
    of disseminated Ewing sarcoma. The rapid recovery of NK cells after tool to identify recurrence of recipient haematopoiesis. Post-​
    T-​cell-​depleted haploidentical transplantation and the omission of myeloablative HSCT for malignant disorders, the detection of early
    a long and intensive pharmacological GvHD prophylaxis is also mixed chimerism has been associated with impending relapse,
    the rationale for an ongoing trial (NCT02258815) of haploidentical while its role following RIC must be taken cautiously depending on
    transplantation followed by the post-​transplant application of the regimen intensity. Post-HSCT chimerism assessment represents a
    monoclonal anti-​GD2 antibody ch14.18 in patients with refractory/​ standard monitoring to assess engraftment in most centres, and the
    relapsed high-​risk neuroblastoma. In this clinical study, ch14.18 is availability of reliable techniques on PB, as well as on BM, makes
    applied to induce antibody-​dependent cellular cytotoxicity (ADCC) its use practical. If assessed by means of short tandem repeat PCR,
    Importance of long-term follow-up 61

    a sensitivity of 10–​3 can be marginally increased if subsets of cells the description of both physical and psychological side effects ex-
    are sorted (typically CD3+ and CD34+ cells). The use of quantita- perienced by survivors after childhood cancer and/​or HSCT.
    tive PCR to detect donor/​recipient polymorphisms allows 2-​log in- The amount of produced data is extremely impressive and
    creased sensitivity. According to international recommendations, somehow difficult to summarize due to confounding factors like re-
    chimerism should always be assessed together with other more spe- cruitment bias, loss to follow-​up growing with time, time-​dependent
    cific disease monitoring strategies (flow or quantitative/​molecular evolution of both first and subsequent lines of therapy for primary
    PCR) to confirm the diagnosis of disease recurrence. malignancies as well as HSCT procedures, and, finally, methodology
    In the paediatric setting, data from both European and American for data collection (e.g. questionnaires only, register studies, studies
    centres are available with regard to the use of immunotherapy for with or without clinical examination). However, it is now clearly
    halting immunosuppression and/​or DLI to treat mixed chimerism demonstrated that transplanted patients experience more severe
    and efficaciously prevent relapse both in AML and ALL. These data and more frequent sequelae than non-​transplanted children and
    have been encouraging in showing limited GvHD post DLI, but the more often report worse quality of life.
    dosage and schedule of DLI administration differed significantly be- The most frequent physical sequelae are metabolic syndrome
    tween the three studies, thus preventing strong recommendations in (defined by dyslipidaemia, hypertension, glucose intolerance, and
    this regard. Crucial to all of these studies has been the intensity of abdominal obesity), thyroid dysfunction, growth retardation (with
    chimerism monitoring in the post-​transplant setting, with weekly low adult final height), hypofertility or sterility, cataract, heart in-
    assessments until day +100 in the American study and +200 in the jury, kidney-​function injury, lung-​function alteration, and second
    German studies. It can be speculated that this represents a key com- cancer. Moreover, transplanted patients describe less success in pro-
    ponent in the use of pre-​emptive DLI, as it has been described that fessional, social, and family life, with lower global well-​being and
    the quick kinetic of disease relapse often escapes the possibility of quality of life as compared to the general population or even to pa-
    DLI treatment. tients treated for cancer during childhood in the same years but
    This argument has prompted multiple studies in which DLI has without HSCT. Overall, it appeared that myeloablative conditioning
    been administered prophylactically post HSCT, specifically after regimens, especially TBI-​based ones, yielded the most severe and
    T-​cell-​depleted or RIC-​conditioned HSCT. While the rationale of frequent sequelae.
    prophylactic DLI proved correct in increasing the survival probability In 2002, the International BFM Study Group initiated a prospective
    of patients with high-​risk diseases in multiple studies, this came at the randomized study comparing a TBI-​based versus a chemotherapy-​
    high price of acute and chronic GvHD, specifically when DLI was based conditioning regimen for paediatric patients transplanted for
    administered in the first 100 days post HSCT. More recent experi- acute lymphoblastic leukaemia, with long-​term follow-​up being one
    ence with cautious DLI timing and dosing and/​or modified DLI has of the secondary endpoints of this study. An important achievement
    proven better results. Nevertheless, the choice of prophylactic DLI was the publication of guidelines for the population of patients trans-
    specifically in paediatric patients must be heavily considered against planted during childhood and the development of specific long-​
    the high risk of severe and potentially life-​limiting GvHD. term follow-​up clinics in many centres. Specifically trained, multiple
    professional caregivers should offer comprehensive and global care
    and deal with issues in endocrinology, dermatology, cardiology,
    Importance of long-​term follow-​up pulmonology, gynaecology, andrology, neurology, nephrology, in-
    ternal medicine, ophthalmology, orthopaedics, physiotherapy, and
    As other paediatricians taking care of patients with severe diseases psychology.
    requiring intensive or prolonged therapy, paediatric transplanters Recommendations may differ among countries and may de-
    are very concerned about the long-​term side effects on their patients. pend on the type of conditioning regimen, patient age at the time
    Thanks to the improvements achieved in the HSCT setting, approxi- of HSCT, and early and late complications occurring after HSCT.
    mately 70% of the children transplanted for malignant diseases sur- However, a minimal corpus of tests should include at least blood
    vive after transplantation and are virtually cured from their primary tests (haematology, immunology, metabolic tests, endocrinological
    disease. Then, besides the goal to improve OS, the most important functions, vitamin D level, ferritin); physical examination, with a
    challenge for the transplant community is to improve patient quality special emphasis on skin (especially for patients who received TBI
    of life by reducing sequelae without losing disease control. and/​or experienced chronic GvHD) and growth (before, during,
    Multiple national or international consortiums were developed and after puberty); as well as imaging and diagnostics, with ultra-
    in previous decades in order to better describe, understand, and sound echocardiography, lung-​function tests, dual X-​ray absorpti-
    treat the mid-​and long-​term complications of patients surviving ometry (DEXA) for bone metabolism, eye examination, and limb
    after childhood cancer. The first long-​term follow-​up structured MRI in case of avascular necrosis symptoms. This test set has to be
    programme was the Childhood Cancer Survivor Study in the USA performed regularly after HSCT, within one year after transplant,
    and Canada, reporting on more than 5,000 patients treated for acute and then repeated on a yearly basis up to adulthood.
    leukaemia between 1970 and 1986. Unfortunately, there were no As for any other chronic paediatric disease, transition from the
    specific data regarding transplantation in this cohort. A European paediatric to the adult medical environment represents a very crit-
    consortium created the Pancare programme dedicated to the same ical period and challenge, and has to be anticipated, well organized,
    population and, finally, a French group built the Leucémie Enfants and accompanied. Patients surviving after childhood HSCT have to
    et Adolescents (LEA, i.e. leukaemia children and adolescents) co- be well informed about the treatment they received and their poten-
    hort from 2004, including patients specifically treated for acute leu- tial late sequelae in order to play as primary actors in their subse-
    kaemia with or without transplantation. Specific cohorts allowed for quent medical life.
    62 CHAPTER 7 Allogeneic Stem-Cell Transplantation

    Nowadays, it is mandatory that prevention, early diagnosis, and Horan JT, et al. (2008) Impact of disease risk on efficacy of matched
    treatment of long-​term side effects are integrated in the care of trans- related bone marrow transplantation for pediatric acute myeloid
    planted paediatric patients, without which the whole work of any leukemia: the Children's Oncology Group. J Clin Oncol 26(35),
    transplanting physician would be valueless. 5797–​801.
    Hunger SP, Mullighan CG (2015) Acute lymphoblastic leukemia in
    children. N Engl J Med 373(16), 1541–​52.
    FURTHER READING Kato K, et al. (2015) Comparison of transplantation with reduced and
    myeloablative conditioning for children with acute lymphoblastic
    Abdul Wahid SF, et al. (2014) Comparison of reduced-​intensity and
    leukemia. Blood 125(8), 1352–​4.
    myeloablative conditioning regimens for allogeneic hematopoietic
    Kuhlen M, et al. (2018) Outcome of relapse after allogeneic HSCT in
    stem cell transplantation in patients with acute myeloid leukemia
    children with ALL enrolled in the ALL-​SCT 2003/​2007 trial. Br J
    and acute lymphoblastic leukemia: a meta-​analysis. Stem Cells Dev
    Haematol 180(1), 82–​9.
    23(21), 2535–​52.
    Kupst MJ, et al. (2002) Cognitive and psychosocial functioning of
    Ash RC, et al. (1990) Successful allogeneic transplantation of T-​cell-​
    pediatric hematopoietic stem cell transplant patients: a prospective
    depleted bone marrow from closely HLA-​ matched unrelated
    longitudinal study. Bone Marrow Transplant 30(9), 609–​17.
    donors. N Engl J Med 322(8), 485–​94.
    Lang P, et al. (2015) Improved immune recovery after transplantation
    Baker KS, et al. (2010) The burden of cure: long-​term side effects fol-
    of TCRalphabeta/​ CD19-​ depleted allografts from haploidentical
    lowing hematopoietic stem cell transplantation (HSCT) in chil-
    donors in pediatric patients. Bone Marrow Transplant 50(Suppl
    dren. Pediatr Clin North Am 57(1), 323–​42.
    2), S6–​10.
    Balduzzi A, et al. (1995) Unrelated donor marrow transplantation in
    Llosa NJ, et al. (2017) Reduced-​intensity haploidentical bone marrow
    children. Blood 86(8), 3247–​56.
    transplantation with post-​transplant cyclophosphamide for solid
    Bashey A, Solomon SR (2014) T-​cell replete haploidentical donor
    tumors in pediatric and young adult patients. Biol Blood Marrow
    transplantation using post-​transplant CY: an emerging standard-​
    Transplant 23(12), 2127–​36.
    of-​care option for patients who lack an HLA-​identical sibling
    Locatelli F, et al. (2017) Outcome of children with acute leukemia given
    donor. Bone Marrow Transplant 49(8), 999–​1008.
    HLA-​haploidentical HSCT after alphabeta T-​cell and B-​cell deple-
    Beck F et al. (2015) Survivorship after childhood cancer. PanCare: a
    tion. Blood 130(5), 677–​85.
    European Network to promote optimal long-​ term care. Eur J
    Mertens AC, et al. (2015) Conditional survival in pediatric malignan-
    Cancer 51(10), 1203–​11.
    cies: analysis of data from the Childhood Cancer Survivor Study
    Biondi A, et al. (2012) Imatinib after induction for treatment of chil-
    and the Surveillance, Epidemiology, and End Results Program.
    dren and adolescents with Philadelphia-​ chromosome-​ positive
    Cancer 121(7), 1108–​17.
    acute lymphoblastic leukaemia (EsPhALL): a randomised, open-​
    Panasiuk A, et al. (2015) Gonadal function and fertility after stem cell
    label, intergroup study. Lancet Oncol 13(9), 936–​45.
    transplantation in childhood: comparison of a reduced intensity
    Blaise D, et al. (1992) Allogeneic bone marrow transplantation for
    conditioning regimen containing melphalan with a myeloablative
    acute myeloid leukemia in first remission: a randomized trial of a
    regimen containing busulfan. Br J Haematol 170(5), 719–​26.
    Busulfan-​Cytoxan versus Cytoxan-​total body irradiation as pre-
    Pérez-​Martínez A, et al. (2015) A phase I/​II trial of Interleukin-​15—​
    parative regimen: a report from the Group d’Etudes de la Greffe de
    stimulated natural killer cell infusion after haploidentical stem cell
    Moelle Osseuse. Blood 79(10), 2578–​82.
    transplantation for pediatric refractory solid tumors. Cytotherapy
    Bonneville M, et al. (2010) Gammadelta T cell effector functions: a
    17(11), 1594–​603.
    blend of innate programming and acquired plasticity. Nat Rev
    Peter J, et al. (2017) Safety and efficacy of Tcralpha/​beta and CD19
    Immunol 10(7), 467–​78.
    depleted haploidentical stem cell transplantation following re-
    Conter V, et al. (2010) Molecular response to treatment redefines all
    duced intensity conditioning in children: results of a prospective
    prognostic factors in children and adolescents with B-​cell pre-
    multicenter phase I/​II clinical trial. Blood 130, 214.
    cursor acute lymphoblastic leukemia: results in 3184 patients of the
    Peters C, et al. (2015) Stem-​cell transplantation in children with acute
    AIEOP-​BFM ALL 2000 study. Blood 115, 3206–​14.
    lymphoblastic leukemia: a prospective international multicenter
    Copelan EA, et al. (2013) Better leukemia-​free and overall survival in
    trial comparing sibling donors with matched unrelated donors—​
    AML in first remission following cyclophosphamide in combin-
    The ALL-​SCT-​BFM-​2003 trial. J Clin Oncol 33(11), 1265–​74.
    ation with busulfan compared with TBI. Blood 122(24), 3863–​70.
    Powles RL, et al. (1983) Mismatched family donors for bone-​marrow
    de Berranger E, et al. (2014) Impact on long-​term OS of conditioning
    transplantation as treatment for acute leukaemia. Lancet 1(8325),
    regimen in allogeneic BMT for children with AML in first CR: TBI+CY
    612–​15.
    versus BU+CY: a report from the Societe Francaise de Greffe de Moelle
    Rettinger E, et al. (2011) Preemptive immunotherapy in childhood
    et de Therapie Cellulaire. Bone Marrow Transplant 49(3), 382–​8.
    acute myeloid leukemia for patients showing evidence of mixed
    Eapen M, et al. (2017) Allele-​level HLA matching for umbilical cord
    chimerism after allogeneic stem cell transplantation. Blood 118(20),
    blood transplantation for non-​malignant diseases in children: a
    5681–​88.
    retrospective analysis. Lancet Haematol 4(7), e325–​33.
    Robinson TM, et al. (2016) Haploidentical bone marrow and stem cell
    Gluckman E, et al. (1989) Hematopoietic reconstitution in a patient
    transplantation: experience with post-​transplantation cyclophos-
    with Fanconi's anemia by means of umbilical-​cord blood from an
    phamide. Semin Hematol 53(2), 90–​7.
    HLA-​identical sibling. N Engl J Med 321(17), 1174–​8.
    Rubnitz JE, et al. (2010) Minimal residual disease-​directed therapy
    Hasle H (2014) A critical review of which children with acute myeloid
    for childhood acute myeloid leukaemia: results of the AML02
    leukaemia need stem cell procedures. Br J Haematol 166(1), 23–​33.
    multicentre trial. Lancet Oncol 11(6), 543–​52.
    Hjorth L, et al. (2015) Survivorship after childhood cancer. PanCare: a
    Schlegel P, et al. (2015) Favorable NK cell activity after haploidentical
    European Network to promote optimal long-​ term care. Eur J
    hematopoietic stem cell transplantation in stage IV relapsed Ewing's
    Cancer 51(10),1203–​11.
    sarcoma patients. Bone Marrow Transplant 50(Suppl 2), S72–​6.
    Importance of long-term follow-up 63

    Socie G, et al. (2000) New malignant diseases after allogeneic marrow trans- Verneris MR, et al. (2015) HLA mismatch is associated with worse
    plantation for childhood acute leukemia. J Clin Oncol 18(2), 348–​57. outcomes after unrelated donor reduced-​intensity conditioning
    Thiel U, et al. (2013) Allogeneic stem cell transplantation for patients hematopoietic cell transplantation: an analysis from the Center for
    with advanced rhabdomyosarcoma: a retrospective assessment. Br International Blood and Marrow Transplant Research. Biol Blood
    J Cancer 109(10), 2523–​32. Marrow Transplant 21(10), 1783–​9.
    Thomas ED, et al. (1957) Intravenous infusion of bone marrow in
    patients receiving radiation and chemotherapy. N Engl J Med
    257, 491–​6.
    8
    Cancer Immunotherapy in Children
    Koichi Hirabayashi, Gianpietro Dotti, and Barbara Savoldo

    Introduction with haematological malignancies such as high-​risk acute lympho-


    blastic leukaemia (ALL) and acute myeloid leukaemia (AML), re-
    Chemotherapy and radiotherapy play a critical role in promoting a lapsed/​refractory lymphoma, juvenile myelomonocytic leukaemia,
    high cure rate in several paediatric malignancies. However, the long-​ and myelodysplastic syndrome. The inclusion of HSCT as a form
    term toxicities associated with these treatments remain significant. of cellular immunotherapy, called graft-​versus-​leukaemia (GvL),
    Furthermore, patients with tumours relapsing after chemotherapy or more generally graft-​versus-​tumour (GvT), stems from the ob-
    and/​or radiotherapy, or resistant to these treatments, do not have servation that tumour relapse rate is reduced in patients developing
    effective alternative therapeutic options. graft-​versus-​host disease (GvHD). Furthermore, lower relapse rates
    In addition to toxic agents, tumour cells can be eliminated through are observed in patients receiving non-​T-​cell-​depleted allografts and
    other mechanisms including the cytolytic effects mediated by cellular experiencing GvHD, as compared to those receiving T-​cell-​depleted
    components of the immune system, or by antibodies targeting surface allografts without occurrence of GvHD. The critical role of GvT ef-
    antigens expressed by tumour cells and activating either the comple- fects were further supported by the re-​induction of remission upon
    ment or immune cells, or by delivering conjugated toxins. Cellular donor lymphocyte infusion (DLI) in patients who relapsed after
    and humoral immunotherapy approaches are thus explored in cancer HSCT. Currently, HSCTs are safely, widely, and more effectively per-
    patients with the goal to offer less toxic therapies, overcome tumour formed by virtue of the diversification of donor sources, reduced tox-
    chemo/​radio resistance and prevent relapse, or treat minimal diseases. icity associated with conditioning regimens, and better supportive
    In this chapter, we will discuss the principle of cancer immuno- therapies. As a result, prognoses of paediatric patients with haem-
    therapy in paediatric patients starting with the most common form atological malignancies have significantly improved upon HSCT.
    of cellular immunotherapy: allogeneic haematopoietic stem-​cell
    Separating graft-​versus-​tumour versus graft-​versus-​host
    transplantation (HSCT). We will then discuss specific immuno-
    disease effects
    therapy strategies based on the administration of classic monoclonal
    antibodies (mAbs) targeting tumour-​ associated antigens, novel GvHD is mediated by both donor T-​cell responses to host alloantigens
    bispecific antibodies that simultaneously target tumour-​associated and the concomitant dysregulation of inflammatory cytokines cas-
    antigens and activate CD3+ T lymphocytes, and mAbs that block cade that follows the HSCT procedure. The most critical trigger
    key inhibitory molecules of the immune system (checkpoint for acute GvHD is found in disparities in major histocompatibility
    blockade). Finally, we will describe specific cellular immunotherapy complex (MHC) molecules. Minor histocompatibility antigens and
    approaches, such as tumour vaccine and adoptive transfer of im- donor/​recipient differences in the killer immunoglobulin-​like re-
    mune cells. Although only a few immunotherapies have so far been ceptor (KIR) and KIR-​ligand (KIRL) interactions also affect the risk
    incorporated into standard practice for paediatric cancers, their role of acute GvHD. Natural killer (NK) cells participate in the GvT ef-
    is enjoying a new revival, after the promising results obtained in fects through KIRL mismatches. NK cells exert their alloreactivity
    recent clinical trials. The great majority of clinical studies are con- by sensing missing-​self on leukaemia cells, especially in adult pa-
    ducted in patients with relapsed cancer, and consist of Phase I trials. tients with AML and paediatric patients with B-​cell precursor ALL.
    Here, we only outline the most innovative ones and/​or those that A major goal in the setting of HSCT is to suppress GvHD while
    have reached Phase II/​III (Table 8.1). maintaining or promoting GvT. In animal models, controlling the
    proliferation of alloreactive T cells is an effective strategy to uncouple
    GvT from GvHD. The clinical translation of these observations has
    Allogeneic haematopoietic stem-​cell been pursued for the past three decades by administering immuno-
    transplantation suppressive agents or by T-​cell depletion of the graft. In recent years,
    the most innovative approaches to control GvHD, especially in the
    Allogeneic haematopoietic stem-​ cell transplantation (HSCT) is setting of haploidentical transplant, consist of the administration of
    currently considered the standard of care for paediatric patients cyclophosphamide in the post-​transplant setting and/​or in selective
    Allogeneic haematopoietic stem-cell transplantation 65

    Table 8.1 Summary of the clinical studies in paediatric patients with anti-​tumour antibody agents

    Trial Phase Agents Disease Number of Age in years Outcome Reference


    patients (range)
    Goldman et al. III Rituximab Burkitt 40 11(3–​23) 90% (3-​year OS and EFS; Br J Haematol (2014), 167,
    lymphoma /​ 95% CI, 76–​96%) 394–​401.
    leukaemia
    Meinhardt et al. II Rituximab Mature B-​cell 87 10.4(1.5–​17.5) 41.4% (RR) (95% CI, J Clin Oncol (2015), 28,
    NHL and Burkitt 31–​52%) 3115–​21.
    leukaemia
    Raetz et al. II Epratuzumab ALL 54 8.4 49.3% ± 8.1% (2-​year OS), Pediatr Blood Cancer (2015),
    (weekly) 39.9 ± 8.0% (2-​year EFS), 62, 1171–​5.
    66% (CR)
    Epratuzumab 60 (4–​15) 60% (RR), 40% (CR)
    (twice-​weekly)
    Rytting et al. II Inotuzumab B-​cell ALL 5 (4–​15) 60% (RR), 40% (CR) Pediatr Blood Cancer (2014),
    ozogamicin 61, 369–​72.
    Pro et al. II Brentuximab ALCL 58 52(14–​76) 86% (objective J Clin Oncol (2012), 30,
    vedotin response = CR+PR, 57% 2190–​6.
    (CR), 29% (PR))
    Yu et al. III Dinutuximab Neuroblastoma 226 96% 86% ± 4% (2-​year OS), N Engl J Med (2010), 363,
    <18 months 66% ± 5% (2-​year EFS) 1324–​34.
    O'Hear et al. III Gemtuzumab AML 17 No data 14 had a reduction in their Cancer (2013), 119,
    ozogamicin alone MRD level and 13 became 4036–​43.
    MRD-​negative
    Gemtuzumab 29 55% ± 13.9% (5-​year OS),
    ozogamicin & 50% ± 13.4% (5-​year EFS)
    chemotherapy
    Ebb et al. II Trastuzumab Osteosarcoma 41 13.8 59% ± 14% (30-​month J Clin Oncol (2012), 30,
    OS), 32% ± 14% (30-​ 2545–​51.
    month EFS)
    Malempati et al. I/​II Cixutumumab Solid tumours 47 15(4–​28) 9% (PR Ewing sarcoma J Clin Oncol (2012), 30,
    and Ewing group) 256–​62.
    sarcoma
    Mossé et al. I Crizotinib Solid tumours 79 10.1(1.1–​21.4) 18% (RR) COG Phase 1 Consortium,
    and ALCL 2013.
    von Stackelberg I/​II Blinatumomab ALL 70 8(1–​17) 39% ± 12% (CR) within the J Clin Oncol (2016), 34,
    et al. first two cycles 4381–​9.
    Merchant et al. I Ipilimumab Solid tumours 33 13.4(2.3–​21) 18% (SD), 0% (CR and PR) Clin Cancer Res (2016), 22,
    1364–​70.
    Blumenthal et al. Retrospective Pembrolizumab Brain tumour 5 5(3–​7) 3.2 months (median OS), J Neurooncol (2016), 129,
    analysis 100% (PD) 453–​60.

    Abbreviations: NHL, non-​Hodgkin lymphoma; AML, acute myelogenous leukaemia; ALCL, anaplastic large-​cell lymphoma; ALL, acute lymphoblastic leukaemia; OS, overall survival;
    EFS, event-​free survival; RR, response rate; SD, stable disease; PR, partial response; PD, progressive disease; PFS, progression-​free survival

    T-​cell depletion of the graft. In haploidentical HSCT, the adminis- post transplant and selective removal of αβTCR+ T cells are both
    tration of cyclophosphamide post transplant, together with standard effective in controlling and preventing GvHD. However, the impact
    prophylaxis, reduces the incidence of GvHD to acceptable rates of these methodologies in promoting increased GvT effects require
    without the need for T-​cell depletion of the graft. Interestingly, the larger and controlled clinical studies.
    timing of cyclophosphamide administration on the third to fifth day
    after transplant mirrors the timing predicted by preclinical models Adoptive T-​cell therapy after haematopoietic
    in which alloreactive T cells are activated. stem-​cell transplantation
    As far as the selected T-​cell depletion of the graft is concerned, the DLI continues to be used in the attempt to re-​induce remission in
    most novel approach consists of the selective removal of αβTCR(T-​ patients relapsing after HSCT. While highly effective in patients with
    cell receptor)+CD3+ T cells, while preserving the γδTCR+CD3+ chronic myeloid leukaemia (CML), DLI is less efficient in control-
    subset. Alloreactive precursors that mediate GvHD are confined to ling other malignancies and also increases the rate of GvHD.
    T cells expressing the αβTCR, which represent the great majority of
    T cells. Between 2% and 10% of T cells express the γδTCR and these Donor lymphocyte infusion with safety switches
    cells do not cause GvHD, while maintaining protective effects against To control GvHD associated with DLI and to improve GvT effects, T
    a wide range of pathogens. Administration of cyclophosphamide cells can be manipulated ex vivo to include safety switches or suicide
    66 CHAPTER 8 Cancer Immunotherapy in Children

    genes. Insertion of the herpes simplex virus-​derived enzyme thymi- complement-​dependent cytotoxicity (CDC), or delivery of toxic
    dine kinase (HSV-​tk) and the inducible human caspase9 (iCasp9) agents upon internalization of the antigen/​mAb complex. ADCC
    have been tested clinically as DLI after allogeneic HSCT. In par- occurs when the fragment crystallizable (Fc) domains of the mAb
    ticular, iCasp9 has been safely administered in paediatric patients. engage the Fc receptors on the surface of immune effector cells.
    The iCasp9 is a human-​derived, chimeric caspase9 that activates the
    cell’s physiological apoptotic pathway when a specific small mol- Antibodies targeting B-​cell antigens
    ecule, known as chemical inducer of dimerization (CID) (AP1903/​ Rituximab, a CD20-​directed chimeric mAb, is the most validated and
    Rimiducid), binds to the FK-​ binding portion of the chimeric effective mAb, and is currently included in many clinical protocols to
    caspase9. In a small group of patients, iCasp9-​based DLI produced treat B-​cell lymphomas in adults. Safety and tolerability of rituximab
    improved immune reconstitution, elimination of opportunistic have been reported for paediatric patients with high-​risk mature B-​
    pathogens, and effective control of acute GvHD in paediatric pa- cell lymphomas in a pilot study conducted by the Children’s Oncology
    tients. Larger studies are currently ongoing and should provide de- Group (COG). In this study, rituximab was used in association with
    finitive conclusions concerning the impact of iCasp9-​based DLI in conventional chemotherapy. An ongoing international Phase III
    controlling GvHD and promoting GvL effects. trial will determine the efficacy of rituximab in paediatric B-​cell
    lymphoma and B-​cell ALL (NCT01595048 and NCT01516580).
    Natural killer cells as donor lymphocyte infusion Following promising results in Phase I/​II studies, epratuzumab,
    The adoptive transfer of NK cells is also an attractive option. The a humanized IgG1 mAb that targets CD22 (a B-​lineage restricted
    administration of these cells after HSCT does not cause GvHD. As antigen expressed on B-​precursor ALL), recently entered a Phase III
    NK cells recognize self MHC Class I molecules to prevent their ac- clinical trial in paediatric patients with standard-​risk relapsed ALL
    tivation, Rubnitz et al. investigated, in a pilot study, the safety and (NCT01802814). CD22 has also been targeted with inotuzumab
    feasibility of haploidentical NK cell infusions in children or young (CMC-​544), in which the mAb is conjugated to ozogamicin, with
    adolescents with AML. Ten patients who had completed chemo- promising results in a small pilot study.
    therapy and were in first complete remission (CR) of AML were
    enrolled. They received cyclophosphamide and fludarabine, fol- Antibodies targeting non-​classical B-​cell antigens
    lowed by KIR-​HLA mismatched NK cells and interleukin-​2 (IL-​2) Brentuximab vedotin is a highly successful example of conjugated mAb.
    to support NK-​cell expansion in vivo. All patients had transient Brentuximab targets CD30 and is conjugated to the toxin monomethyl
    engraftment and expansion of KIR-​mismatched NK cells. Non-​ auristatin E (vedotin). After binding via CD30 mAb, the toxin binds
    haematological toxicities were limited, with no GvHD. However, the to tubulin and disrupts the microtubule network with consequent
    efficacy of this approach remains to be determined. In another study, cell-​cycle arrest and apoptosis. Brentuximab is used in patients with
    haploidentical NK cells obtained from parental donors were infused Hodgkin's lymphoma and CD30+ anaplastic large-​cell lymphoma
    in 29 paediatric patients with relapsed or refractory leukaemia. (ALCL). CR rates in adults with relapsed/​refractory CD30+ ALCL re-
    Objective leukaemia regression was observed and patients success- ceiving brentuximab vedotin as a single agent are in the 57% range.
    fully proceeded to a second HSCT. Currently, an ongoing international Phase III trial is determining
    the efficacy of brentuximab vedotin in CD30+ ALCL and Hodgkin's
    Gene-​modified donor lymphocyte infusion lymphoma (NCT01979536, NCT01780662, and NCT02166463).
    The most innovative modified DLI is however represented by the
    infusion of donor T cells genetically modified to specifically target Antibodies targeting gangliosides
    tumour-​associated antigens via either chimeric antigen receptors Dinutuximab is a chimeric anti-​disialoganglioside-​2 (GD2) CAR in
    (CARs) or conventional T-​cell receptors (TCRs). These technologies Chimeric antigen receptor T cell section mAb, validated in a COG
    are described in specific sections of this chapter. However, relevant Phase III trial of post autologous HSCT in paediatric patients with
    for this section, in clinical studies so far reported using donor-​ neuroblastoma (NB). Patients were randomly assigned to receive
    derived T cells genetically manipulated ex vivo to express the CD19-​ dinutuximab administered with granulocyte macrophage colony-​
    specific CAR, no GvHD has occurred. In one of them, the lack of stimulating growth factor (GM-​CSF) and IL-​2 in conjunction with
    GvHD upon infusion of these cells was expected, since the specific isotretinoin, versus isotretinoin alone. Immunotherapy combined
    CAR was engrafted on virus-​specific cytotoxic T cells, which are well with isotretinoin (event-​free survival (EFS), 66%) was superior to
    known to not cause GvHD. Importantly, when donor polyclonal T standard isotretinoin maintenance therapy (EFS, 46%). As a result,
    cells were used as the CAR vehicle in another study, anti-​tumour ac- immunotherapy with dinutuximab after autologous HSCT is cur-
    tivity was confirmed without the occurrence of GvHD. The mechan- rently considered the standard of care for high-​risk NB.
    isms responsible for these effects remain elusive. However, a recent
    mouse model suggests that alloreactive T cells experience enhanced Antibodies targeting myeloid antigens
    stimulation due to the engrafted CAR, resulting in the progressive Gentuzumab ozogamicin is an anti-​ CD33 mAb conjugated to
    loss of both their effector function and proliferative potential, which the drug calicheamicin and designed to target AML. Upon initial
    likely prevents the development of GvHD. encouraging Phase I studies, gentuzumab was withdrawn due to liver
    toxicity and lack of significant clinical benefit in a Phase III trial con-
    Monoclonal-​based antibody therapies ducted in adult patients with AML. However, O'Hear et al. reported
    Monoclonal antibodies (mAbs) recognize antigens expressed more promising results in paediatric patients. In a multicentre trial,
    on the tumour cell surface and cause anti-​tumour activity via ei- paediatric patients with AML received gentuzumab, either alone
    ther antibody-​
    dependent cell-​ mediated cytotoxicity (ADCC), or in combination with chemotherapy, and achieved molecular
    Allogeneic haematopoietic stem-cell transplantation 67

    remission in both cohorts. All patients subsequently safely received II clinical trial which validated safety and efficacy of blinatumomab
    HSCT, without significant increase in treatment-​related mortality in paediatric patients with relapsed/​refractory B-​ALL. They treated
    after transplantation. Furthermore, this study showed that lower 49 patients in Phase I and 44 patients in Phase II. Four patients had
    doses of gentuzumab could be safely administered. dose-​limiting toxicities during cycle 1 (Phase I); three experienced
    grade 4 cytokine-​release syndrome (one attributed to grade 5 car-
    Antibodies targeting growth factors diac failure); and one had fatal respiratory failure. Among the 70 pa-
    Vascular endothelial growth factor (VEGF) is a signalling protein tients who received the recommended dosage, 27 (39%) achieved CR
    that is critical for solid tumour vascular proliferation. Bevacizumab, a (including molecular remission) within the first two cycles with 14
    VEGF inhibitor, has shown activity both as a single agent and in com- (52%) achieving molecular remission. These results demonstrated
    bination with other chemotherapy drugs for several tumours. Modak anti-​leukaemic activity of blinatumomab as a single agent, endorsing
    et al. reported a Phase II study that evaluated the combination therapy additional studies for the approval of this drug. New BiTE®s, such
    with bevacizumab, irinotecan, and temozolomide for refractory/​re- as CEA/​CD3 bispecific antibody (MEDI-​565) and EpCAM/​CD3
    lapsed NB. Although objective responses were achieved, the addition bispecific antibody (MT110) have also been developed and are being
    of bevacizumab did not improve response rates in resistant NB. evaluated for safety and efficacy in clinical trials.

    Antibodies targeting growth factor receptors Immune checkpoint blockade


    Cetuximab is an mAb directed against epidermal growth factor While endogenous immune responses to cancer cells are often de-
    (EGF) receptors and, although approved for the treatment of several tected in patients, their anti-​tumour effects are usually insufficient
    adult malignancies, its role in paediatric patients remains unclear. to prevent cancer progression. This immune tolerance of tumours is
    In a Phase I trial to determine the dose of cetuximab that can be induced through the expression of ligands by the tumour or tumour-​
    safely combined with irinotecan in children and adolescents with infiltrating cells that ultimately hamper T-​cell functions within the
    refractory solid tumours, stable disease was reported, especially in tumour microenvironment. To break these immune-​ inhibitory
    patients with brain tumours (38.5%), granting further investigation. pathways, immune checkpoint blockades using anti-​programmed
    The expression of human epidermal growth factor receptor 2 cell death protein -​1 (PD-​1) mAb, anti-​programmed death ligand
    (HER2) has been associated with poor outcome in osteosarcoma, (PD-​L1) mAb, and anti-​cytotoxic T lymphocytes-​associated protein
    suggesting that targeting this antigen may have a clinical impact. Ebb 4 (CTLA-​4) mAb have been developed.
    et al. reported a Phase II trial of the anti-​HER2 mAb trastuzumab in
    combination with cytotoxic chemotherapy in patients with meta- Anti-​CTLA-​4
    static HER2+ osteosarcoma. Forty-​one patients were treated with CTLA-​4 was the first immune checkpoint receptor to be clinically tar-
    chemotherapy and trastuzumab for 34 consecutive weeks, but no geted. After T-​cell activation, CTLA-​4 is upregulated, to specifically
    significant differences in EFS and overall survival (OS) rates were inhibit effector function through a variety of mechanisms, including
    observed compared to patients receiving chemotherapy only. preventing co-​stimulation by outcompeting CD28 for its ligand, B7,
    Insulin-​like growth factor-​1 receptor (IGF-​1R) promotes tumour and also by inducing T-​cell cycle arrest. Two antibodies targeting
    growth, especially in sarcomas. Cixutumumab, a fully human IgG1 CTLA-​4, ipilimumab and tremelimumab, entered clinical evaluation.
    mAb against IGF-​1R was used in a Phase I/​II trial in paediatric pa- A Phase I clinical trial of ipilimumab in paediatric patients with ad-
    tients with refractory solid tumours and Ewing sarcoma. The drug vanced solid tumours showed more toxicities than in adults.
    was well tolerated, but the number of responses observed did not
    meet the statistical criteria for single-​agent activity in Ewing Sarcoma. Anti-​PD-​1 and anti-​PD-​L1
    By interacting with its two ligands PD-L1 or PD-L2, PD-1 is also
    Antibodies targeting tumour-​specific mutation a negative regulator of T-cell activity. When engaged by its ligand,
    Various human cancers, including NB and ALCL, have gene trans- PD-1 inhibits kinase signalling pathways that normally lead to
    locations, amplifications, or oncogenic mutations of the anaplastic T-cell activation. PD-1 is expressed on many immunological cells,
    lymphoma kinase (ALK) tyrosine kinase receptor. The anti-​ALK including B cells and NK cells, and blocking anti-PD-1 mAb and
    mAb Crizotinib was tested in a Phase I study in children with re- anti-PD-L1 mAb influences the function of these cells. Anti-​PD-​1
    fractory solid tumours and ALCL. Objective tumour responses were and PD-​L1 mAbs have demonstrated unprecedented clinical efficacy
    documented in 14 of the 79 patients, including eight CR and five in more than 15 cancer types in adults, with very good safety profiles.
    partial remission (PR). Responses were more evident in patients In paediatric cancer, Phase I or II clinical trials are currently under
    with ALK aberrations. way. Blumenthal et al. retrospectively described outcomes in chil-
    dren with pontine gliomas, but no responses were documented. To
    Bispecific T-​cell engager antibodies enhance anti-​tumour efficacy of the immune checkpoint blockade,
    Bispecific T-​cell engager (BiTE®) antibodies use a technology that combination therapy such as chemotherapy, targeted therapy, radio-
    capitalizes on T-​cell cytotoxicity with direct cell-​killing capability. therapy, and other immunotherapy may be required.
    These bispecific mAbs bridge surface antigens on target cells with
    the αβTCR of effector T cells to facilitate targeted cytotoxic T-​ Chimeric antigen receptor T cells
    cell-​mediated cytolysis. The prototype of this new class of mAbs is Chimeric antigen receptors (CARs) are fusion proteins where the
    blinatumomab, which is a bispecific CD19-​directed CD3 T-​cell en- binding site of a monoclonal antibody is linked to intracellular
    gager. Blinatumomab has produced durable responses in patients with signalling molecules of the TCR and co-​ stimulatory molecules.
    various B-​cell malignancies. Von Stackelberg et al. reported a Phase I/​ These molecules are expressed in T cells via viral or non-​viral vector
    68 CHAPTER 8 Cancer Immunotherapy in Children

    delivery. When CAR molecules bind to the specific antigen expressed Paediatric patients with osteosarcoma also underwent CAR T-​cell
    on target cells through their single-​chain variable fragment (scFv) therapy using HER2-​specific CAR. In a Phase I/​II trial, 19 patients
    segment, they activate the lytic and co-​stimulatory pathways of the with HER2+ tumours received escalating doses of autologous HER2-​
    T cells, ultimately promoting MHC-​independent cytotoxic activity CAR T cells. Of the 17 evaluable patients, four had stable disease for 12
    and cell expansion. While first-​generation CAR T cells rely on the weeks to 14 months. Three of these patients had their tumour removed,
    exclusive intracellular CD3ζ chain-​signalling domain, second-​and with one showing extensive necrosis . The median OS of all 19 infused
    third-​generation versions have included one and two co-​stimulatory patients was 10.3 months (range, 5.1 to 29.1 months) (Figure 8.1).
    molecules to improve activation and persistence of engineered T cells.
    Cytokine release syndrome (CRS) is the most severe complica- Natural-​killer-​cell immunotherapy
    tion following the adoptive transfer of CAR T cells and is considered NK cells are innate lymphoid cells dedicated to the clearance of malig-
    to occur following a dysregulated activation of the immune system nant cells via direct cytotoxicity and production of immunoregulatory
    driven by the activated CAR T cells. CRS ranges from mild and cytokines. Although cytokine stimulation promotes NK-​cell activation
    self-​limiting symptoms such as high temperature and myalgia, to and greater cytotoxicity against malignant targets in vitro, clinical trials
    severe and life-​threatening scenarios characterized by multi-​organ based on the infusions of autologous NK cells have shown limited clin-
    failure. Corticosteroid and the interleukin-​6 receptor-​blocking mAb ical efficacy. Suboptimal results are likely to be found in NK-​cell dys-
    tocilizumab are useful for controlling CRS. functions produced by multiple rounds of chemotherapy that these
    patients undergo and in the expression of inhibitory receptors by tu-
    Clinical trials with chimeric antigen receptors mour cells. On the other hand, clear benefits are reported in patients
    for haematological malignancies with AML and ALL receiving allogeneic KIR-​mismatched NK-​cell in-
    Thus far, the most successful example of CAR T-​cell-​based therapy fusion (see the section ‘Adoptive T-​cell therapy after haematopoietic
    is achieved by targeting CD19. Clinical trials of CD19-​CAR T cells stem-​cell transplantation’). To overcome the KIR-​KIRL-​mediated in-
    have shown remarkable outcomes in B-​cell ALL, with CR rates of hibition and increase the anti-​leukaemic effects of NK cells, different
    70–​90% in relapsed/​refractory cases. Greater efficacy and tolerability strategies are being evaluated, including blocking of KIR-​KIRL inter-
    is observed in children. The Children’s Hospital of Philadelphia re- actions with a monoclonal antibody, using bispecific killer engagers and
    ported a 90% CR rate in 30 patients with relapsed/​refractory ALL (25 a trispecific killer engager to activate NK cells via the Fc-​receptor against
    treated on the paediatric trial) in Phase I trials with CTL019, a CD19-​ antigens expressed on leukaemia cells and via expression of CARs.
    targeted CAR T-​cell product containing the 4-​1BB molecule as a co-​
    stimulatory domain. EFS and OS at six months were 67% and 78%,
    Tumour vaccines and dendritic-​cell immunotherapy
    respectively. The National Cancer Institute (NCI) Pediatric Oncology Active vaccination of patients with cancer, including paediatric pa-
    Branch performed an intent-​to-​treat analysis of 21 children and young tients, has been extensively explored, although the rate of clinical re-
    adults with ALL treated with CD19-​CAR T cells expressing the CD28 sponses remains extremely low. Paediatric patients should in theory
    domain as a co-​stimulatory molecule, and reported a 70% CR rate. be more responsive to vaccination, since their immune system is
    At a median follow-​up of ten months, OS was 51.6%. EFS in the 21 largely composed of naive T cells. However, clinical studies based
    patients achieving molecular emission was 17.7 months, with all but on cancer vaccines remain limited in paediatric patients, likely re-
    three receiving subsequent HSCT. Despite high initial response rates, flecting the scepticism in the field due to the poor outcome of clin-
    some patients had however relapsed. Antigen escape (loss of CD19 by ical studies in adult patients. Kushner et al. reported a Phase I trial
    tumour blasts) and short CAR T-​cell persistence have been implicated targeting GD2 and GD3 antigens with OPT-​821 adjuvant in com-
    as causes for failure. Up to 60% of the relapsed cases after CD19-​CAR bination with oral β-​glucan for high-​risk NB in second or later re-
    T cells have indeed shown loss of CD19 antigen expression, which ren- mission. Vaccine and β-​glucan were well tolerated, with 80% EFS
    ders the malignant cells invisible to CD19-​specific immunotherapies. at 24 months. Antibody responses against GD2 and/​or GD3 were
    To overcome this tumour escape mechanism, CAR T cells targeting observed in patients achieving clinical responses. These data should
    alternative antigens are being tested of which CAR T cells targeting encourage the use of cancer vaccines in other malignancies in
    the CD22 molecule are the most advanced. Concerning issues related paediatric patients, also taking into account that several candidate
    to the limited persistence of CAR T cells, current clinical outcomes antigens have been identified and characterized.
    suggest that CAR co-​stimulated through the 41BB endodomain have Vaccination using dendritic cells (DCs) generated from peripheral
    prolonged survival (up to two years) over CD28 co-​stimulated CAR T monocytes following exposure to granulocyte macrophage colony-​
    cells (two to three months). stimulating factor (GM-​CSF) and IL-​4 is the most advanced type of
    cancer vaccine. Sipuleucel-​T (Provenge®) is an American Food and
    Clinical trials with chimeric antigen receptors Drug Administration-​ approved autologous DC-​ based immuno-
    for solid tumours therapy for metastatic hormone-​refractory prostate cancer. In the
    In addition to ALL, NB is another paediatric disease treated with paediatric setting, DC-​based vaccines have been used in Phase I/​
    CAR T cells. Actually, NB is the first paediatric solid tumour in II trials. Krishnadas et al. combined decitabine and DCs targeting
    which CAR T cells have been tested in clinical trials. The NB-​ MAGE-​A1, MAGE-​A3, and NY-​ESO-​1, and, although feasible and
    specific GD2 molecule was targeted in the first trial with CAR T safe, this strategy produced only modest anti-​tumour activity. Objective
    cells. Although challenges from the tumour microenvironment responses were documented in only one of the ten patients treated, and
    in solid tumours are expected to hinder CAR T cells, results were in two patients who had no evidence of disease at the time of treatment,
    encouraging, with three of the 11 patients with active disease one remained disease-​free two years post therapy, while the other re-
    achieving a complete response. lapsed ten months post therapy. Mackall et al. reported a pilot study
    Summary 69

    <Second-generation CAR>
    T cell or NK cell
    Cytokines and granules
    Cytokines and granules
    T cell or NK cell

    VH VL

    VL VH
    CD3 or CD16
    VH
    VL
    <BiTE and BiKE>

    Linker
    Tumour cell Tumour antigen

    VH
    VH
    Ectodomain
    VL <TriKE>
    VL
    Hinge
    VH VL

    Transmembrane domain VL
    VH CD16

    CD28 or 4-1BB
    Cytokines and granules
    Endodomain
    CD3ξ
    T cell or NK cell
    NK cell

    Figure 8.1 Cellular immunotherapy for tumours utilizing T cell/​NK cells. Chimeric antigen receptors (CAR) are recombinant proteins composed of
    an extracellular single-​chain variable fragment (scFv) derived from an antibody, joined to a hinge and a transmembrane domain, further linked to the
    intracellular signalling molecules of the T-​cell receptor complex (CD3ζ) and co-​stimulatory (CD28 or 4-​1BB) endodomains. CAR T cells combine the
    specificity of an antibody with the cytotoxic and memory functions of T cells. Bispecific/​trispecific T-​cell/​NK-​cell engagers (BiTE/​BiKE or TriKE) are
    unique bispecific/​trispecific antibodies that have two/​three linked scFv constructed to be flexible and have a 1+1 antigen-​binding valence. They bind
    CD3 on T cells or CD16 on NK cells, and one and two antigens on tumour cells, to activate T cells/​NK cells for effective elimination of tumour cells.

    of consolidative DC vaccines in patients with high-​risk paediatric patients. Considering the astonishing results of anti-​CD19 CAR
    sarcomas. Thirty patients with recurrent or metastatic Ewing’s sarcoma T-​cell therapy for ALL, CAR T cells have also the potential to rad-
    had an initial cell harvest to collect autologous cells, and then received ically change conventional cancer treatment. As conventional and
    chemotherapy, radiation therapy, and/​or surgery, followed by infusion high-​dose chemotherapy regimens maintain high toxicities, novel
    of autologous T cells and DCs pulsed with tumour-​specific peptides de- immune-​based strategies validated in adults should be promptly
    rived from tumour-​specific breakpoints and the E7 antigen. Treatment moved to applications in children, who urgently require effective
    was well tolerated and induced immune responses in 39% of patients anti-​tumour control with diminished toxicities.
    vaccinated with the translocation breakpoint peptides. Nevertheless,
    the clinical benefits of the treatment are difficult to ascertain. The re-
    cent introduction in the cancer arena of immune checkpoint inhibitors FURTHER READING
    may reinvigorate the interest in vaccine approaches. Ahmed N, et al. (2015) Human epidermal growth factor receptor 2
    (HER2)—​specific chimeric antigen receptor-​modified T cells for
    the immunotherapy of HER2-​positive sarcoma. J Clin Oncol 33,
    Summary 1688–​96.
    Appelbaum FR (2007) Hematopoietic-​cell transplantation at 50. N
    Engl J Med 357, 1472–​5.
    HSCT has been refined for over 60 years by incorporating the latest
    Blumenthal DT, et al. (2016) Pembrolizumab: first experience with
    clinical/​preclinical findings. Consequently, it represents the best
    recurrent primary central nervous system (CNS) tumours. J
    curative option for many diseases. New methods aimed at reducing Neurooncol 129, 453–​60.
    the rate of malignant relapse, preventing and treating GvHD, further Ciceri F, et al. (2007) Antitumour effects of HSV-​TK-​engineered donor
    reducing infectious complications, as well as increasing the avail- lymphocytes after allogeneic stem-​cell transplantation. Blood 109,
    ability of HSCT, continue to be investigated. 4698–​707.
    There have been several new developments in cancer immuno- Copelan EA (2006) Hematopoietic stem-​cell transplantation. N Engl J
    therapy over the past decade which have dramatically altered Med 354, 1813–​26.
    standard cancer therapy mainly in adults. In addition to mAbs, Cruz CR, et al. (2013) Infusion of donor-​derived CD19-​redirected
    immune checkpoint blockade, administered for a variety of can- virus-​specific T cells for B-​cell malignancies relapsed after allo-
    cers in adults, is one of the most promising drugs for paediatric geneic stem cell transplant: a phase 1 study. Blood 122, 2965–​73.
    70 CHAPTER 8 Cancer Immunotherapy in Children

    Daniele N, et al. (2012) Transplantation in the onco-​ hematology Merchant MS, et al. (2016) Phase I clinical trial of ipilimumab in pediatric
    field: focus on the manipulation of αβ and γδ T cells. Pathol Res Pract patients with advanced solid tumours. Clin Cancer Res 22, 1364–​70.
    208, 67–​73. Modak S, et al. (2017) Combination of bevacizumab, irinotecan, and
    Ebb D, et al. (2012) Phase II trial of trastuzumab in combination with temozolomide for refractory or relapsed neuroblastoma: results of a
    cytotoxic chemotherapy for treatment of metastatic osteosarcoma phase II study. Pediatr Blood Cancer 64(8). doi: 10.1002/​pbc.26448
    with human epidermal growth factor receptor 2 overexpression: a Mossé YP, et al. (2013) Safety and activity of crizotinib for paediatric
    report from the children's oncology group. J Clin Oncol 30, 2545–​51. patients with refractory solid tumours or anaplastic large-​ cell
    Geldres C, et al. (2016) Chimeric antigen receptor-​redirected T cells lymphoma: a Children's Oncology Group phase 1 consortium
    return to the bench. Semin Immunol 28, 3–​9. study. Lancet Oncol 14, 472–​80.
    Ghosh A, et al. (2017) Donor CD19 CAR T cells exert potent Mussetti A, et al. (2017) Post-​transplant cyclophosphamide, a prom-
    graft-​versus-​lymphoma activity with diminished graft-​versus-​host ising anti-​graft versus host disease prophylaxis: where do we stand?
    activity. Nat Med 23, 242–​9. Expert Rev Hematol 10(5), 479–​92.
    Goldman S, et al. (2014) Rituximab with chemotherapy in children O'Hear C, et al. (2013) Gemtuzumab ozogamicin can reduce minimal
    and adolescents with central nervous system and/​or bone marrow-​ residual disease in patients with childhood acute myeloid leukemia.
    positive Burkitt lymphoma/​ leukaemia: a Children's Oncology Cancer 119, 4036–​43.
    Group Report. Br J Haematol 167, 394–​401. Porter DL, et al. (2011) Chimeric antigen receptor-​modified T cells in
    Holtan SG, et al. (2014) Acute graft-​versus-​host disease: a bench-​to-​ chronic lymphoid leukemia. N Engl J Med 365, 725–​33.
    bedside update. Blood 124, 363–​73. Postow MA, et al. (2015) Immune checkpoint blockade in cancer
    Huang MA, et al. (2015) Cellular and antibody based approaches for therapy. J Clin Oncol 33, 1974–​82.
    pediatric cancer immunotherapy. J Immunol Res 2015, 675269. Pro B, et al. (2012) Brentuximab vedotin (SGN-​35) in patients with
    Jenq RR, van den Brink MR (2010) Allogeneic haematopoietic stem relapsed or refractory systemic anaplastic large-​cell lymphoma: re-
    cell transplantation: individualized stem cell and immune therapy sults of a phase II study. J Clin Oncol 30, 2190–​6.
    of cancer. Nat Rev Cancer 10, 213–​21. Pule MA, et al. (2008) Virus-​specific T cells engineered to coexpress
    Kantoff PW, et al. (2010) Sipuleucel-​T immunotherapy for castration-​ tumour-​specific receptors: persistence and antitumour activity in
    resistant prostate cancer. N Engl J Med 363, 411–​22. individuals with neuroblastoma. Nat Med 14, 1264–​70.
    Klingemann H (2015) Challenges of cancer therapy with natural killer Pulsipher MA, et al. (2014) New frontiers in pediatric allo-​SCT: novel
    cells. Cytotherapy 17, 245–​49. approaches for children and adolescents with ALL. Bone Marrow
    Krishnadas DK, et al. (2015) A phase I trial combining decitabine/​den- Transplant 49, 1259–​65.
    dritic cell vaccine targeting MAGE-​A1, MAGE-​A3 and NY-​ESO-​ Raetz EA, et al. (2015) Re-​ induction chemoimmunotherapy
    1 for children with relapsed or therapy-​refractory neuroblastoma with epratuzumab in relapsed acute lymphoblastic leukemia
    and sarcoma. Cancer Immunol Immunother 64, 1251–​60. (ALL): Phase II results from Children's Oncology Group (COG)
    Kushner BH, et al. (2014) Phase I trial of a bivalent gangliosides vac- study ADVL04P2. Pediatr Blood Cancer 62, 1171–​5.
    cine in combination with β-​glucan for high-​risk neuroblastoma in Rashidi A, et al. (2016) Post-​transplant high-​dose cyclophosphamide
    second or later remission. Clin Cancer Res 20, 1375–​82. after HLA-​matched vs haploidentical hematopoietic cell trans-
    Lee DW, et al. (2015) T cells expressing CD19 chimeric antigen re- plantation for AML. Bone Marrow Transplant 51, 1561–​4.
    ceptors for acute lymphoblastic leukaemia in children and young Rubnitz JE, et al. (2015) Natural killer cell therapy in children with re-
    adults: a phase 1 dose-​escalation trial. Lancet 385, 517–​28. lapsed leukemia. Pediatr Blood Cancer 62, 1468–​72.
    Louis CU, et al. (2011) Antitumour activity and long-​term fate of Rubnitz JE, et al. (2010) NKAML: a pilot study to determine the safety
    chimeric antigen receptor-​positive T cells in patients with neuro- and feasibility of haploidentical natural killer cell transplantation in
    blastoma. Blood 118, 6050–​6. childhood acute myeloid leukemia. J Clin Oncol 28, 955–​9.
    Luznik L, et al. (2008) HLA-​haploidentical bone marrow transplant- Rytting M, et al. (2014) Initial experience with CMC-​544 (inotuzumab
    ation for hematologic malignancies using nonmyeloablative con- ozogamicin) in pediatric patients with relapsed B-​ cell acute
    ditioning and high-​dose, posttransplantation cyclophosphamide. lymphoblastic leukemia. Pediatr Blood Cancer 61, 369–​72.
    Biol Blood Marrow Transplant 14, 641–​50. Sharma P, Allison JP (2015) The future of immune checkpoint therapy.
    Mackall CL, et al. (2008) A pilot study of consolidative immuno- Science 348, 56–​61.
    therapy in patients with high-​risk pediatric sarcomas. Clin Cancer Shimasaki N, et al. (2016) Expanded and armed natural killer cells for
    Res 14, 4850–​8. cancer treatment. Cytotherapy 18, 1422–​34.
    Malempati S, et al. (2012) Phase I/​II trial and pharmacokinetic study Suck G, et al. (2016) Natural killer cells for therapy of leukemia.
    of cixutumumab in pediatric patients with refractory solid tumours Transfus Med Hemother 43, 89–​95.
    and Ewing sarcoma: a report from the Children's Oncology Group. Tey SK, et al. (2007). Inducible caspase 9 suicide gene to improve the
    J Clin Oncol 30, 256–​62. safety of allodepleted T cells after haploidentical stem cell trans-
    Mallhi K, et al. (2015) Hematopoietic cell transplantation and cellular plantation. Biol Blood Marrow Transplant 13, 913–​24.
    therapeutics in the treatment of childhood malignancies. Pediatr Trippett TM, et al. (2009) Phase I and pharmacokinetic study of
    Clin North Am 62, 257–​73. cetuximab and irinotecan in children with refractory solid tu-
    Maude SL (2017) Future directions in chimeric antigen receptor T cell mours: a study of the pediatric oncology experimental therapeutic
    therapy. Curr Opin Pediatr 29, 27–​33. investigators' consortium. J Clin Oncol 27, 5102–​8.
    Maude SL, et al. (2014) Chimeric antigen receptor T cells for sustained von Stackelberg A, et al. (2016) Phase I/​Phase II Study of blinatumomab
    remissions in leukemia. N Engl J Med 371, 1507–​17. in pediatric patients with relapsed/​refractory acute lymphoblastic
    Meinhardt A, et al. (2010) Phase II window study on rituximab in newly leukemia. J Clin Oncol 34, 4381–​9.
    diagnosed pediatric mature B-​cell non-​Hodgkin's lymphoma and Yu AL, et al. (2010) Anti-​GD2 antibody with GM-​CSF, interleukin-​2,
    Burkitt leukemia. J Clin Oncol 28, 3115–​21. and isotretinoin for neuroblastoma. N Engl J Med 363, 1324–​34.
    9
    Supportive Care During Treatment
    Marianne van de Wetering and Robert S. Phillips

    Introduction geographically varied. Infections with fungal organisms such as


    Candida yeasts, Aspergillus moulds, or other opportunistic fungi,
    Reducing the toxic and unpleasant effects of cancer therapies is a and with viruses such as cytomegalovirus (CMV), varicella zoster
    critical component of cancer management in children. As the inten- virus (VZV), adenovirus, and infections with Pneumocystis jirovecii
    sity of primary treatment has escalated, so have side effects such as (pneumocystis pneumonia, PCP) may occur.
    myelosuppression and infection. Improved outcomes are dependent
    Prevention of infection
    on skilled supportive care, and advances in supportive care have par-
    alleled advances in cancer therapy. In this chapter we will highlight a A number of strategies are commonly used to reduce the risk of in-
    few key aspects, evidence-​based if possible. fection, although the evidence for some of these is lacking.

    The environment
    Infection Neutropenic patients are at increased risk of hospital-​acquired in-
    fections and the hospital environment should be avoided where
    Infection remains a common cause of hospital admission during possible. Children are encouraged to continue their normal lifestyle,
    cancer therapy and is the commonest cause of death after disease including school and hobbies. Teachers should be informed of the
    relapse/​progression. Children are at increased risk of bacterial, viral, risks to the child with cancer and asked to alert the child’s parents
    protozoal, and fungal infections, and the prevention and treatment when there is VZV or measles in the school.
    of infection are central to supportive care.
    The hospital environment
    Many factors influence the frequency and severity of infection in
    cancer patients. These include myelosuppression or immunosup- If hospitalized during neutropenia, children are often cared for in
    pression as a result of disease or therapy, and disruption of natural protective isolation in general hospitals, although many specialist
    barriers to infection. Neutropenia (often defined as an absolute neu- paediatric oncology units manage neutropenic children on the open
    trophil count of less than 500 cells/​mm3) is the most important risk oncology ward, with isolation being reserved for those with clearly
    factor, with both the depth and duration of neutropenia influencing transmissible infections (e.g. respiratory viral infections, diarrhoea).
    infection risk. Death has been reported in 1–​3% of episodes of fe- Careful handwashing is essential and prevents the spread of infection.
    brile neutropenia (FN) in children, with adolescent/​young adult pa-
    Nutrition/​diet
    tients at greater risk.
    Around 15–​20% of all episodes of FN in children are associated There are data refuting the role of special measures concerning food
    with positive blood-​culture isolates. Gram-​positive organisms cause products during neutropenia, but sensible food hygiene rules and
    up to 70% of proven bloodstream infections, Gram-​negative infec- avoiding potentially contaminated products such as unpasteurized
    tions cause 28%, and fungal infections, mainly yeasts, cause 2%. milk or cheese, seafood, and pâté should be followed. A Cochrane
    The commonest Gram-​positive organisms are coagulase-​negative review was performed (with an update in 2016) summarizing the
    Staphylococci, but enterococci and viridans group streptococcal evidence concerning this topic. As the evidence is very poor, no
    species have the potential for severe disease and antibiotic resist- definite recommendation could be given for clinical practice.
    ance. Among Gram-​ negative organisms, the most frequent are Intensive dietary restrictions often apply to patients undergoing
    Escherichia coli, Klebsiella species, Pseudomonas aeruginosa, Serratia haematopoetic stem-​cell transplantation (HSCT), though with little
    species, and Proteus species. Infections with Gram-​negative bacteria supportive data.
    producing extended spectrum beta-​lactamases are of increasing
    Invasive procedures
    concern in hospitals and the community, but are not yet widely re-
    ported in children with cancer. As with antibiotic resistance in other The requirement for any procedure that can disrupt the normal
    groups of patients, the nature of the local resistance pattern is very mucosa or skin barriers during neutropenic periods (e.g. dental
    72 CHAPTER 9 Supportive Care During Treatment

    procedures) should be carefully balanced against the increased pos- Children should be tested for VZV IgG at diagnosis, and VZV-​
    sibility of infection. negative patients should be given prophylaxis with varicella zoster
    immune globulin or aciclovir after significant contact. Little data is
    Antibiotic prophylaxis available to suggest which is better.
    Oral antibiotics may be used to preserve beneficial anaerobic or-
    ganisms while preventing colonization of the gut by pathogenic Prevention of Pneumocystis jirovecii pneumonia
    aerobic organisms. Historically, trimethoprim/​ sulfamethoxazole TMP/​SMZ is highly effective in preventing PCP and is important
    (TMP/​SMZ) was used but, more recently, quinolones have been for all patients undergoing treatment for leukaemia, lymphoma, or
    employed. Selective decontamination of the digestive tract (SDD) HSCT, and for patients with solid tumours with prolonged neutro-
    is effective in preventing (mainly Gram-​ negative) bacteraemia penic episodes. An alternative to TMP/​SMZ prophylaxis is aerosol-
    during neutropenic episodes, and reduces deaths from infective ized or intravenous pentamidine. Dapsone can be given if patients
    causes. Quinolones are slightly more effective than TMP/​SMZ. cannot tolerate pentamidine.
    Some groups support the use of antibiotic prophylaxis in selected
    patients, such as those with acute leukaemia, HSCT patients, and Granulocyte colony-​stimulating factor
    relapsed patients. The risk of emergence of resistant organisms as Granulocyte colony-​ stimulating factor (G-​ CSF) is widely used
    the result of routine prophylaxis has not been significant to date, but in adult and paediatric practice to reduce the depth and duration
    requires ongoing study. of chemotherapy-​associated neutropenia. Data support its use in
    preventing hospitalization, but not episodes of confirmed infection
    Fungal prophylaxis or death. The choice to give colony-​stimulating factors is usually de-
    The need for fungal prophylaxis is less clear than for bacterial termined by protocol, to support the administration of dose-​intense
    prophylaxis, but is recommended by many as standard of care in chemotherapy, or in the setting of progenitor cell mobilization.
    high-​risk patients (i.e. HSCT patients, high-​risk phases of acute Secondary prophylactic G-​CSF may be used if previous episodes of
    leukaemia therapy). Early diagnosis of invasive fungal infection neutropenia have led to severe infections, delay, or dose reduction of
    is expected to improve outcomes, but prevention remains the chemotherapy. If given, G-​CSF should be started 24–​72 hours after
    ultimate goal. The nature of the prophylactic agent chosen re- chemotherapy and continued until the neutrophil count is greater
    mains under considerable debate. Fluconazole provides effective than 1000/​mm3 on two occasions.
    prophylaxis against most Candida infections, although some spe-
    cies such as Candida krusei may be resistant. Aspergillus prophy- Immunization in prevention of infection
    laxis requires a mould-​active, agent and co-​administration of Some aspects of vaccine-​acquired immunity are lost during cancer
    azoles. Monitoring of children at increased risk for Aspergillus therapy. Booster vaccinations are recommended for all children six
    infection is essential. months after completing therapy, and a complete re-​vaccination
    programme is recommended for children at a year or more after
    Viral prophylaxis HSCT (Table 9.1). Off-​treatment immunization with HPV (human
    Aciclovir is used in patients at high risk of CMV and VZV activation, papilloma virus) vaccines is recommended, at an appropriate age.
    namely patients on very intensive protocols and those undergoing There is a balance between the most significant time for infection
    HSCT. Clinical VZV (shingles) is common after viral prophylaxis is risk and the likelihood of response to therapy. Live attenuated vac-
    discontinued in HSCT patients, and it is currently recommended to cines should not be administered until at least six months after
    continue for up to a year post HSCT. stopping chemotherapy, but if herd immunity for measles is low,

    Table 9.1 Vaccination recommendations during and after chemotherapy treatment (not applicable to children during/​after autologous or
    allogeneic bone-​marrow transplant)

    Timing in relation to chemotherapy Vaccination* Recommendation


    During chemotherapy MMR If low-​herd immunity, then use single-​antigen measles vaccine
    OPV Not allowed. If needed, then eIPV
    DTaP Prefer to wait until stopping chemotherapy
    Hepatitis B High-​risk areas—​Recombivax HB
    After chemotherapy DTaP/​MMR/​HiB/​pneumococcal/​HPV Restart schedule 6 months after stopping chemotherapy
    (age-​related)
    Special considerations Influenza vaccine Not contraindicated during chemotherapy
    Varicella zoster Degree of protection uncertain
    Pneumococcal No evidence-​based recommendation
    In maintenance chemotherapy, safe and immunogenic, not registered
    yet for use in children with cancer. Indicated for splenic dysfunction e.g.
    splenectomy in Hodgkin lymphoma, post radiotherapy (radiation spleen)

    * DTaP: diptheria, tetanus, acellular pertussis vaccine; eIPV: enhanced inactivated polio vaccine; HiB: haemophilus influenzae b conjugate vaccine; MMR: measles, mumps, rubella
    vaccine; OPV: oral polio vaccine; HPV: human papilloma virus
    Infection 73

    single-​antigen measles vaccine may be given during or soon after neutropenia. Laboratory evaluation should include a complete
    chemotherapy, with measles vaccination repeated six months after blood count, liver enzymes, renal function, and blood cultures from
    stopping chemotherapy. Ideally, hepatitis B vaccination is given after both central lines and peripherally. Urine culture is routinely ad-
    completing chemotherapy, but in high-​risk groups or areas it can be vised, as dipstick screening and symptoms are less prevalent in this
    given during chemotherapy, with reduced immunogenic response. population. Stool culture testing for Clostridioides (formely known as
    Periodic booster doses are usually necessary, with the timing deter- Clostridium) toxin, cerebrospinal fluid examination, and chest X-​ray
    mined by serological testing at 12-​month intervals. There are no data should be done only if clinically indicated. Based on the patient his-
    on changes in immune status or re-​vaccination against HPVs after tory and initial clinical findings, patients may be assessed as at low
    chemotherapy or HSCT. or high risk of complications of FN. Multiple risk-​assessment tools
    Most countries recommend yearly influenza vaccination of have been used in paediatric FN; geographical variation in their per-
    household contacts and patients during chemotherapy and after formance means local validation should be undertaken prior to their
    HSCT. A Cochrane systematic review emphasized the paucity of use in practice.
    data on this vaccine. Serological responses are lower than expected
    in healthy controls, and antibody levels considered protective in Empirical antibiotic therapy
    healthy individuals may not prevent clinical infection in an immuno- The progression of infection in neutropenic patients can be rapid, so
    compromised individual. There are no data on whether vaccination empirical intravenous antibiotic therapy should be started promptly.
    of paediatric cancer patients protects against clinical infection. The initial goal is to provide broad-​spectrum cover against both
    However, as the vaccine is well tolerated, it is not contraindicated. Gram-​negative and Gram-​positive organisms. In areas with ap-
    Active VZV vaccination has been given to VZV IgG negative pa- propriate resistance patterns, this is best accomplished with mono-
    tients during leukaemia maintenance therapy. The vaccine is im- therapy with an extended-​spectrum beta-​lactam, fourth-​generation
    munogenic, but some patients develop vaccine-​related infection cephalosporin, or carbapenem such as meropenem or imipenem.
    requiring aciclovir treatment. In addition, chemotherapy must be Through resistance, either prior isolation from the patient or
    omitted for one week before and one week after the time of vaccin- knowledge of the local pattern, it may be that an aminoglycoside
    ation, and steroids should not be given for two weeks after vaccin- is required. Low-​risk patients should be treated as part of a clinical
    ation. The vaccine has not been widely studied in other groups of trial or according to an established institutional low-​risk protocol,
    children with malignancy. For these reasons, geographical differ- which may recommend commencement of oral antibiotics instead.
    ences in practice exist with respect to VZV vaccination of suscep- However, we should realize that this is not routine practice. The em-
    tible children with cancer. pirical antibiotic regimen should be reviewed regularly in each in-
    Live or attenuated vaccines such as measles, mumps, rubella stitution, taking account of changes in the spectrum of infections,
    (MMR), or oral polio are contraindicated during cancer therapy. antimicrobial susceptibility patterns, and the underlying aetiology
    Siblings or household contacts may safely receive MMR, but en- of the neutropenia.
    hanced inactivated polio vaccine (eIPV) should be given in place of
    oral polio vaccine (see Table 9.1). ‘Planned progressive’ therapy
    The therapeutic plan should be modified according to clinical re-
    Treatment of infection sponse and microbiological findings. Children who are clinically
    FN is defined differently across the world but usually includes a well, lack a significant clinical focus of infection other than upper
    neutrophil count of less than 500 cells/​mm3 (or above 500 cells/​ respiratory tract infection, have no positive cultures, and whose
    mm3 and expected to fall) and significant fever. Fever definitions fever settles quickly, can discontinue antibiotics once afebrile after
    vary but include a single temperature of more than 38.5°C, or a 24–​48 hours. Some protocols allow for step down to oral anti-
    temperature of more than 38°C twice in 12 hours. Defining the op- biotics in patients at low risk of significant complications. If the
    timal limits is subject to ongoing study. Management is a medical patient without a positive clinical focus or cultures has persistent
    emergency because of the potential risk of overwhelming sepsis. fever after three to five days of treatment, reassess carefully, take
    An international group has published evidence-​based guidelines on further cultures, and look for a focus. If the child is clinically well,
    the topic. Children who are neutropenic and significantly unwell, the same antibiotics can be continued. If the child is deteriorating
    but without fever, should also be managed as for FN; occasionally, clinically, consider changing antibiotics and/​or adding anaerobic
    Gram-​negative sepsis presents with abdominal pain and diarrhoea, cover. In patients with a high risk of fungal infection, empirically
    but without fever, especially in children receiving steroid therapy. add antifungal therapy after five days of persistent fever where
    Febrile, neutropenic children with cancer require careful assess- there is no other explanation, and consider early computed tomog-
    ment. Treatment is divided into empirical therapy and ‘planned pro- raphy (CT) of the chest and abdominal ultrasound scan, along with
    gressive therapy’ stages. directed investigation if other symptoms suggest localized disease.
    The duration of therapy for patients with positive blood cultures
    Assessment is not well defined, with many clinicians choosing to give seven to
    A careful history and thorough physical examination is needed, ten days of intravenous antibiotics. If fever settles quickly, therapy
    including emphasis on the mucosal membranes, the chest, soft for the isolated organism should be optimized, but maintain broad-​
    tissues, central venous catheter site, and perianal area. The assess- spectrum antibiotic cover to prevent breakthrough bacteraemia. It
    ment phase is ongoing: careful and repeated evaluation for specific is not usually necessary to continue antibiotics until the neutrophils
    symptoms and signs or a focus of infection is critical, but there recover. Invasive fungal infection requires a prolonged course of
    may be minimal signs of inflammation at an infected site during antifungal therapy.
    74 CHAPTER 9 Supportive Care During Treatment

    If fever persists in a child with a positive culture, optimize antibac- respiratory, or other symptoms, and without severe lymphopenia
    terial therapy before making any other changes, and consider adding (less than 100 cell/​mm3), an oral treatment may be used.
    antifungal therapy after four to five days of fever. Rarely, fever may CMV can result in fever, rash, hepatosplenomegaly, pneumonia,
    be due to the medication, but stopping antibiotics can be dangerous neurological symptoms, and retinitis. Treatment is with ganciclovir
    and should only be considered after a very careful, thorough exclu- intravenously or foscarnet. Prolonged courses of therapy are neces-
    sion of other causes of fever. sary to eradicate the infection.

    Central venous catheter infections


    Complications related to the long-​term use of central venous catheters Blood-​product transfusion
    are minimized by recommended protocols for catheter placement
    and care, but there is the potential for infection of the bloodstream Anaemia and thrombocytopenia can occur as a consequence of
    and/​or device, the subcutaneous pocket, and the tunnel or exit site. chemotherapy or the underlying disease.
    The overall incidence of catheter-​related infections is approximately
    2 in 1,000 catheter days. Decreasing the number of infections using Red-​cell transfusion
    preventative techniques with ethanol lock or taurolidine lock is The haemoglobin concentration should be maintained at safe levels
    promising but not implemented as standard care yet. but normal physiological levels are not required. There is little evi-
    Most infections are with Gram-​ positive organisms (mainly dence to guide the transfusion threshold but most centres choose a
    coagulase-​negative Staphylococcus), but cover for Gram-​negative level of less than 7 g/​dl and, with radiotherapy, an even higher level
    organisms is necessary until an organism is identified. More than of more than 10 g/​dl is maintained because of the theoretical benefit
    80% of documented catheter-​related infections can be treated suc- of cellular oxygenation during radiotherapy. Leukodepleted, packed
    cessfully by giving antibiotics through the catheter, but treatment red cells are given. Leukodepletion reduces the chance of human
    failures may result from infections with multiple organisms, fungi, leukocyte antigen (HLA) sensitization and infection with prions
    Pseudomonas aeruginosae, resistant Gram-​negative organisms, and and CMV. Irradiated products are not routinely required but im-
    tunnel infections. In Staphylococcus aureus infection, antibiotics portant exceptions are before HSCT, after HSCT until lymphocytes
    should be administered for at least two weeks if the catheter is left have recovered (six months after transplant), in severe combined
    in place. Persistent bacterial catheter infections and most fungal immunodeficiencies, in neonates, and in Hodgkin lymphoma. The
    (Candida) catheter infections require catheter removal. Catheter volume requirement can be calculated variably, with some using 10–​
    removal should also be considered early in the acutely unwell, 15 ml/​kg and others using Target –​Current Hb (in g/​dL) × weight
    septic patient whose condition is not responding rapidly to broad-​ (kg) × 5. There are insufficient data to recommend the routine use of
    spectrum antibiotics and fluid resuscitation. erythropoetin in paediatric oncology.

    Fever without neutropenia Platelet transfusions


    The non-​neutropenic, febrile child should be evaluated very care- Platelet transfusions are indicated if a thrombocytopenic patient is
    fully. Laboratory evaluation includes a blood count and blood cul- actively bleeding. Prophylactic platelet transfusions are indicated
    tures from the central venous catheter. Management is dictated by in patients who are septic, have a known bleeding disorder, or are
    the clinical findings. If the child is clinically well other than fever, undergoing an invasive procedure such as central venous catheter
    with no clinical focus of infection, blood-​culture results can be insertion or lumbar puncture. There is little evidence about the ap-
    awaited before starting antibiotics. If there is a clinical focus of bac- propriate transfusion threshold. Platelets are usually kept more than
    terial infection (e.g. ear, throat, skin), consider appropriate oral anti- 10 × 109/​l, except for brain tumour patients (where levels of 20–​40 ×
    biotics such as amoxicillin or co-​amoxiclav until the blood-​culture 109/​l are often used) and for invasive procedures (where the level is
    result is known. If the non-​neutropenic child is unwell with fever, kept more than 50 × 109/​l). If patients develop allo-​immunization,
    commence empirical antibiotics as for a neutropenic child and give which reduces the therapeutic effect of transfused platelets, single-​
    appropriate supportive care, pending further results. donor transfusions can be used.

    Viral infections
    The most common viral pathogens that affect the immunocom- Nausea and vomiting
    promised child are the herpes viruses including herpes simplex virus
    (HSV), VZV, CMV, and Epstein–​Barr virus (EBV). Herpes viruses Nausea and vomiting remain an important concern in cancer treat-
    can result in mucosal lesions, skin lesions, and neurological symp- ment but adequate control can often be achieved. Techniques such
    toms. Systemic treatment with aciclovir is needed for at least five as relaxation, distraction, and explanation of procedures should be
    days. Aciclovir is not routinely indicated for chemotherapy-​induced considered along with antiemetic medication. Paediatric-​specific
    mucositis. guidelines have been developed to guide practitioners.
    Primary infection with VZV results in chickenpox. In the im- A validated nausea and vomiting tool such as the PENAT score
    munocompromised child, severe complications may be seen, leading (comparable with the face pain rating scale for assessment of pain)
    to a fulminating illness with visceral dissemination of the virus. should be used to assess the severity of nausea and vomiting at
    Untreated VZV pneumonitis can be fatal in up to 7% of affected chil- regular intervals. Management is adjusted according to findings.
    dren. In unwell children, treatment should be intravenous acyclovir. Chemotherapeutic agents can be grouped into four classes: min-
    For those without significant illness, who have a rash but no fever, imal (less than 10% of patients experience nausea and vomiting), low
    Constipation 75

    emetogenicity (10–​30% of patients experience nausea and vomiting), Table 9.2 Antiemetic therapy in relation to different types
    moderate emetogenicity (30–​90% of patients experience nausea and of chemotherapy
    vomiting), and high emetogenicity (more than 90% of patients ex-
    Emetogenic Drug Antiemetic Delayed
    perience nausea and vomiting). Starting medication is tailored to potential therapy emesis
    the drug of highest emetogenicity within the chemotherapy com- Minimal Asparaginase None None
    bination (Table 9.2). In minimal emetogenic chemotherapy, no Bleomycin
    antiemetic therapy is needed. In low emetogenic chemotherapy, a Bevacizumab
    Mercaptopurin
    serotonin receptor antagonist, usually ondansetron, is indicated; in
    Methotrexate (<50 mg/​m2)
    moderate emetogenicity, corticosteroids can be added as the drugs Rituximab
    work synergistically. In highly emetogenic chemotherapy, the com- Vinca alkaloids (except
    bination of a serotonin receptor antagonist plus steroids plus an NK1 vinoralbine)
    antagonist (e.g. aprepitant) should be used unless there are contra- Low Busulphan (oral) 5 HT3 antagonist None
    Cytarabine <1 g/​m2 (e.g. ondansetron/​
    indications. Radiotherapy, especially of the abdomen, can also lead Doxorubicin (liposomal) granisetron)
    to nausea and vomiting, and a serotonin receptor antagonist is often Etoposide
    given 30 minutes before treatment. Steroids can also be useful, espe- Topotecan
    cially for radiotherapy of the brain. Moderate Busulfan 5 HT3 antagonist None
    Those with breakthrough or refractory nausea or vomiting may Cyclophosphamide (e.g. ondansetron/​
    < 1g/​m2 granisetron)
    be difficult to manage. Increasing the intensity of medication is re- Doxorubicin Dexamethasone
    commended and, if necessary, olanzapine, levomepromazine, or Irinotecan
    metoclopramide may be used. An aggressive antiemetic plan at Ifosfamide
    the start of therapy minimizes the initial experience of nausea and Epirubicin
    Idarubicin
    probably reduces the risk of anticipatory nausea and vomiting. If Methotrexate >1 but <12
    anticipatory vomiting does occur, hypnotherapy, cannabinoids, or g/​m2
    banzodiazepines may be effective. High Carboplatin 5 HT3 antagonist Continue
    Carmustine (e.g. ondansetron/​ ondansetron
    Cisplatin granisetron) for 72 hours
    Cyclophosphamide and
    Nutrition support >1 g/​m2 dexamethasone
    Cytarabine >3 g/​m2 (adjusted dose)
    Good nutrition is important during cancer therapy. Patients may Actinomycin and
    Dacarbazine NK1 antagonist
    present undernourished at diagnosis and cancer treatment may pro- Doxorubicin >60 mg/​m2 (aprepitant)
    foundly impact their nutritional status. Melfalan (iv) If an NK1
    Children with malnutrition or at high risk of developing malnu- Methotrexate >12 g/​m2 antagonist is not
    Mitoxantrone >15mg/​m2 used, usual dose
    trition should be reviewed by a paediatric dietician for advice about Procarbazine of dexamethasone
    healthy eating and vitamin supplementation. For children requiring
    Source: data from Dupis L. et al., Guideline for Classification of the Acute Emetogenic
    nutritional intervention, enteral nutrition is the management of Potential of Antineoplastic Medication in Pediatric Cancer Patients. Guideline release date:
    choice if gut function is maintained. Patients may be offered nutri- 11 August 2010. Pediatric Oncology Group of Ontario (POGO). Available from:
    tional supplements such as high-​calorie, nutritional drinks, if able ​https://www.pogo.ca/healthcare/practiceguidelines/ pogoemetogenicitycla/

    to swallow normally and tolerate such supplements. Alternatively,


    nasogastric tube feeding or gastrostomy feeding may be indicated.
    Gastrostomy feeding is particularly suited to older children or young
    people who wish to avoid the cosmetic associations or irritation of a with oral sponges moistened with water or diluted chlorhexidine.
    nasogastric tube or children who cannot safely tolerate a nasogastric Regular assessment of the oral mucous membranes during inpatient
    tube. When gut function is poor for a prolonged period or where therapy using one of the available scoring systems will allow early
    nasogastric feeds cannot be tolerated, intravenous total parenteral detection of, and intervention for, mucositis and other problems.
    nutrition is used. Children who develop mucositis need ongoing basic mouth care and
    adequate analgesia.

    Mouth care
    Constipation
    Good mouth care plays an important role in the prevention of in-
    fection and is essential for children developing complications such Constipation is a common problem in children and young people
    as mucositis, oral candidiasis, xerostomia, or herpes simplex infec- with cancer, and appropriate therapeutic intervention can greatly
    tions. The evidence for mouth-​care interventions is incomplete but improve quality of life. The aetiology is mostly multifactorial: these
    guidelines based on available evidence are now available. patients have reduced activity, reduced oral fluid and food intake,
    At diagnosis, an assessment of the mouth and teeth is essential, and and take medications such as opioids, vinca alkaloids, carboplatin,
    children should be referred to a dentist or dental hygienist promptly. and ondansetron which contribute to constipation. There are no ran-
    During cancer therapy, children should brush their teeth twice daily domized controlled studies performed in this age group to inform
    using a soft toothbrush and fluoride toothpaste, or clean the mouth laxative choice, so management is directed towards the most likely
    76 CHAPTER 9 Supportive Care During Treatment

    cause of constipation in the individual child. Stool softeners and os-


    motic laxatives (such as polyethelene glycol sachets) are most often FURTHER READING
    used, and bowel evacuants are rarely needed. Baskin JL, et al. (2009) Management of occlusion and thrombosis as-
    sociated with long-​term indwelling central venous catheters. Lancet
    374, 159–​69.
    Tumour lysis syndrome Cairo MS, et al. (2010) Recommendations for the evaluation of risk
    and prophylaxis of tumour lysis syndrome (TLS) in adults and chil-
    Tumour lysis syndrome (TLS) is a set of complications that can arise dren with malignant diseases: an expert TLS panel consensus. Br J
    Haematol 149, 578–​86.
    from treatment of rapidly proliferating, drug-​sensitive neoplasms.
    Dupuis LL, et al. (2006) Development and validation of the pediatric
    These are typically haematological malignancies (particularly Burkitt
    nausea assessment tool for use in children receiving antineoplastic
    lymphoma and the acute leukaemias) but, rarely, also solid tumours. agents. Pharmacotherapy 26, 1221–​31.
    Chemotherapy causes rapid destruction of tumour cells leading to Dupuis LL, et al. (2013) Guideline for the prevention of acute nausea
    release of intracellular substances into the bloodstream. Metabolic and vomiting due to antineoplastic medication in pediatric cancer
    disturbances include hyperuricaemia, hyperphosphataemia, hypo- patients. Pediatr Blood Cancer 60, 1073–​82.
    calcaemia, and hyperkalaemia. Clinical manifestations occur 12–​72 Dupuis LL, et al. (2013) Guideline for the prevention and treatment of
    hours after initial start of therapy and can include renal failure (from anticipatory nausea and vomiting due to chemotherapy in pediatric
    precipitation of uric acid crystals and calcium phosphate salts), seiz- cancer patients. Pediatr Blood Cancer 61, 1506–​12.
    ures, and cardiac arrhythmias. Dupuis LL, et al. (2017) Updated MASCC/​ESMO consensus recom-
    Prevention of TLS relies on identification of at-​risk patients. Such mendations: prevention of acute chemotherapy-​induced nausea
    patients should have laboratory and clinical TLS parameters moni- and vomiting in children. Support Care Cancer 25(1), 323–​31.
    tored every 4–​6 hours until stabilized. Therapy is best commenced Flank J, et al. (2016) Guideline for the treatment of breakthrough and
    the prevention of refractory chemotherapy-​induced nausea and
    in a unit with paediatric intensive-​care facilities and nephrology
    vomiting in children with cancer. Pediatr Blood Cancer 63, 1144–​51.
    input for advice and, potentially, dialysis.
    Gafter-​Gvili A, et al. (2012) Antibiotic prophylaxis for bacterial in-
    Standard supportive care to prevent TLS consists of vigorous
    fections in afebrile neutropenic patients following chemotherapy.
    hyperhydration (2–​3 L/​m2/​day) and avoidance of hyperuricaemia. Cochrane Database Syst Rev 18(1), CD004386.Gaynor EP, Sullivan
    The primary treatment for hyperuricaemia has been allopurinol PB (2015) Nutritional status and nutritional management in chil-
    (50–​100 mg/​m2 every eight hours orally, maximum 300 mg/​m2/​ dren with cancer. Arch Dis Child 100, 1169–​72.
    day) combined with hyperhydration. The use of rasburicase (recom- Gibson BE, et al. (2004) Transfusion guidelines for neonates and older
    binant urate oxidase dose 0.2 mg/​kg IV once a day) is preferred for children. Br J Haematol 124(4), 433–​53.
    high-​risk patients, and alkalinization becomes unnecessary. Urate Glenny AM, et al. (2010) The development of evidence-​based guide-
    oxidase catalyses conversion of uric acid to water-​soluble allantoin lines on mouth care for children, teenagers and young adults treated
    which is readily excreted in urine. It decreases uric acid levels more for cancer. Eur J Cancer 46, 1399–​1412.
    than allopurinol, with a rapid drop occurring within four hours of Goossen GM1, et al. (2013) Influenza vaccination in children being
    administration. treated with chemotherapy for cancer. Cochrane Database Syst Rev
    Management is tailored according to risk of TLS. Low-​ 1(8), CD006484.
    Jones L, et al. (2010) Nutritional support in children and young people
    risk patients (solid tumours and AML with white blood-​cell
    with cancer undergoing chemotherapy. Cochrane Database Syst
    count (WCC) less than 25×109/​l) only need monitoring and
    Rev 7(7), Review. [2015 update in Cochrane Database Syst Rev 8,
    hyperhydration. ALL with WCC less than 100×109/​l, AML with CD003298]
    WCC more than 25×109/​l, and large-​cell anaplastic lymphoma Lehrnbecher T (2015) Antifungal prophylaxis in pediatric patients
    are classified as intermediate-​risk disease for which monitoring, undergoing therapy for cancer: drugs and dosing. Curr Opin Infect
    hyperhydration, and allopurinol is given. High-​risk disease (ALL Dis 28(6), 523–​31.
    WCC more than 100×109/​l, bulky Burkitt lymphoma, and other Lehrnbecher T, et al. (2012) Guideline for the management of fever
    stage IV lymphomas) should be managed in specialized paedi- and neutropenia in children with cancer and/​or undergoing hem-
    atric oncology units where patients need careful monitoring, atopoietic stem-​cell transplantation. J Clin Oncol 30, 4427–​38.
    hyperhydration, and rasburicase. Loeffen EAH, et al. (2017) The importance of evidence-​based sup-
    portive care practice guidelines in childhood cancer—​a plea for
    their development and implementation. Support Care Cancer
    Conclusion 25, 1121.
    Mermel LA, et al. (2001) Guidelines for the management of intravas-
    cular catheter-​related infections. Clin Infect Dis 32, 1249–​72.
    This chapter highlights the most important physical aspects of sup-
    Patel SR, et al. (2008) Vaccinations in children treated with standard-​
    portive care for the child with cancer. Pain management is discussed
    dose cancer therapy or hematopoietic stem cell transplantation.
    in Chapter 11 (‘Palliative Care for Children with Advanced Cancer’). Pediatr Clin North Am 55, 169–​86, xi.
    Other aspects, such as psychological support, social interventions, Phillips RS, Gibson F (2008) A systematic review of treatments for
    and clinican–​family–​patient communication, have not been dis- constipation in children and young adults undergoing cancer treat-
    cussed here. The critical importance of supportive care continues ment. J Pediatr Hematol Oncol 30, 829–​30.
    to be the relief of cancer and treatment-​related symptoms, less mor- Pittiruti M, et al. (2016) Evidence-​based criteria for the choice and
    bidity, and reduced mortality. the clinical use of the most appropriate lock solutions for central
    Conclusion 77

    venous catheters (excluding dialysis catheters): a GAVeCeLT con- Sartori AM (2004) A review of the varicella vaccine in immunocom-
    sensus. J Vasc Access 17(6), 453–​64. promised individuals. Int J Infect Dis 8, 259–​70.
    Rheingans JI (2007) A systematic review of nonpharmacologic ad- Sung L, et al. (2007) Meta-​analysis: effect of prophylactic hematopoi-
    junctive therapies for symptom management in children with etic colony-​stimulating factors on mortality and outcomes of infec-
    cancer. J Pediatr Oncol Nurs 24, 81–​94. tion. Ann Intern Med 147, 400–​11.
    Robenshtok E, et al. (2007) Antifungal prophylaxis in cancer patients van Dalen EC, et al. (2016) Low bacterial diet versus control diet to pre-
    after chemotherapy or hematopoietic stem-​cell transplantation: sys- vent infection in cancer patients treated with chemotherapy causing
    tematic review and meta-​analysis. J Clin Oncol 25, 5471–​89. episodes of neutropenia. Cochrane Database Syst Rev 4, CD006247.
    10
    Psychosocial Care
    Martha Grootenhuis, Momcilo Jankovic, Esther van den Bergh, and Femke Aarsen

    Introduction beginning of treatment (e.g. study randomization). Children may


    anxiously ask ‘Am I going to die?’ in a phase where their parents
    The diagnosis of cancer in a child is one of the most stressful events are in emotional turmoil and experiencing loss of control. In the
    that can happen to a family. It confronts the child and parents with a past, it was not common to communicate frankly with children
    range of stressors, including the most feared: the death of the child. about cancer, survival, and death. Fortunately, it is now argued that
    Although treatment has greatly improved over the last decades and enabling children to assimilate information and feelings allows them
    survival rates have increased dramatically, parents and children still to cope better. Research has shown that providing information to the
    face an extensive period of intensive treatment which fundamen- child with cancer about diagnosis and prognosis is beneficial for the
    tally impacts their lives. The whole family has the difficult task of child’s emotional experience of their situation. In this phase, parents
    adjusting to a situation dominated by the stresses of long-​lasting un- may feel guilty about late identification of symptoms related to the
    certainty and uncontrollability, in a situation with a risk for experi- disease. Parents may also be angry about delay in referral to the hos-
    encing medical traumatic stress. pital. Learning to cope with these feelings takes time. Furthermore,
    Children differ from adults. From birth until adulthood, children parents have to cope with the reactions of other family members
    are in a process of physical, cognitive, emotional, and social develop- (e.g. siblings, grandparents), which sometimes can give a profound
    ment. In developing their abilities to cope with their environment, extra burden.
    children are dependent on adults. Together they are in a child-​rearing A second stressor at this stage is learning to deal with the ef-
    relationship, in which an important role is given to their parents. fects of medical treatment. Chemotherapy usually starts shortly
    Younger children need their parents to help them cope with basic after diagnosis. The child rapidly comprehends what cancer treat-
    fears and basic desires, and also with the demands of socialization. ment is about: painful medical procedures (e.g. port-​ a-​cath,
    Older children need their parents in learning to cope with questions lumbar punctures, bone marrow aspirations), sickness because of
    related to physical and cognitive growth, development of personal chemotherapeutic agents (nausea, vomiting, mycosis), or fatigue be-
    identity, and difficulties of functioning in peer groups. Children cause of radiotherapy, and side effects such as hair loss. As well as the
    also differ from adults in their understanding and experience of chemotherapeutic treatment in the hospital, parents have to admin-
    health, illness, and medical care. Depending on the child’s develop- ister cancer treatment to their child at home (e.g. dexamethasone).
    mental level, special needs have to be met in the context of the family Managing this is challenging and makes parents feel extremely
    (parents, siblings), education, and the enlarged social environment responsible. After a tumour resection, children have to cope with
    (friends, school, healthcare workers). Each phase of treatment has physical disabilities, and for those with a brain tumour, they must
    its own characteristics that contribute to the reactions of parents and also deal with neuropsychological deficits.
    children. In this chapter, psychosocial care in these different phases A third stressor, especially at the beginning of treatment, is
    will be described, focusing on both children and parents. informing the social environment (other family members, school,
    employers). Ensuring social support is of major importance because
    of the long duration of treatment. It is well known that social sup-
    Acute phase port is a protective factor in adaptation to the disease. Families need
    ongoing support and to stay in touch with everyday life as much as
    For all children and their parents, the diagnostic phase and start possible. Many families receive social support (e.g. for the care of
    of treatment is very stressful. It commences with invasive medical siblings, household), but it is also important for a social worker to
    procedures necessary to obtain a diagnosis. The child has to en- support the family in all the emotional, practical, or financial con-
    dure pain and the parents have to witness their child in pain and sequences of the disease and treatment (including contact with em-
    fear. The challenge is to accomplish a realistic and, for the child, an ployers or welfare insurance companies). Special care is needed for
    age-​appropriate understanding of diagnosis and treatment implica- non-​native families. Both language and cultural background can
    tions. This is especially necessary for decisions to be made at the make it very difficult to communicate properly about the cancer and
    After treatment 79

    its treatment. Using an interpreter in breaking the bad news, but also decline in health or cognitive functions. As a group, however, they
    in subsequent conversations about changes in the course of treat- hardly seem different from their healthy peers in terms of their class-
    ment, contributes to their understanding. room behaviour and social acceptance. Research shows that chil-
    To help parents adjust to the circumstances after diagnosis can be dren with cancer who receive social support from their teachers,
    demanding (e.g. administering medication, timing of routine hos- parents, and friends report fewer psychosocial problems and have
    pital visits, learning to evaluate symptoms and to decide whether higher self-​esteem.
    or not it is necessary to go and see the doctor, maintaining balance A common problem in childhood cancer treatment is the reduced
    inside the family). Thus, to help parents adjust, psychological inter- immunity due to therapy. This means that parents must be especially
    vention programmes can be useful in teaching them both practical alert for risk of infections. Guidance in the realistic management of
    and emotional problem-​solving skills. these risks and of not becoming too obsessive and frightened is one
    of the issues in which psychological care can make a firm contribu-
    tion. Unscheduled hospital visits due to reduced immunity prob-
    During treatment lems (e.g. fever) are largely disruptive for everyday family life. Daily
    life becomes unpredictable and difficult to plan.
    Modern childhood cancer treatment, in spite of tremendous im- Overall, parents also experience many stress reactions during
    provements in supportive care, still has many unwanted side ef- treatment. Studies on parental stress indicate that both fathers and
    fects. Especially older children have difficulties with losing their hair mothers, six and twelve months after diagnosis, report increased
    and changes in their physical appearance. These can be reversible feelings of anxiety and depression which only return to ‘normal’
    changes such as the use of a feeding tube and substantial weight loss, levels after about 20–​24 months. It is important to deal with the par-
    but also irreversible changes such as stretch marks due to rapid in- ental stress and adaptation to the illness and treatment, because of
    crease in weight because of corticosteroid medication. Adolescents the relationship of parental functioning with the adjustment of the
    typically suffer from feelings of shame and inconvenience. The child child. For some parents, additional guidance is needed from either
    or adolescent should be given information about the side effects of a social worker or a psychologist. Other parents turn to their own
    medication and treatment. Good psychosocial support may include social network for help.
    teaching the child what he or she can do to cope with pain and other Parents of children in treatment for cancer may find it hard to set
    negative feelings. Cognitive behavioural techniques such as relax- limits on their sick child or to be separated from their child. One
    ation, visualization, distraction, and the use of ‘helping thoughts’ are reason is that the stress of parenting a child can be enhanced by the
    beneficial. effects of certain medications, including dexamethasone, which
    In the first months after diagnosis, many children show adjust- may cause behavioural, mood, or cognitive problems in the child.
    ment towards treatment and being ill, but they also start to realize Furthermore, in this stage, often one of the parents needs to go back
    the impact of the cancer experience on their daily lives. Reactions to work. This can increase the burden on the other parent. Going to
    to this adjustment can involve grief and mourning over the fact that school is often encouraged and promoted by the hospital staff, but
    their life as it was before being diagnosed is ‘lost’ or the development often parents are reluctant, as they are scared about the risk of infec-
    of a negative self-​image due to changed appearance or loss of cogni- tions. For the cognitive and social-​emotional development of chil-
    tive abilities. The way a child deals with his/​her illness and treatment dren, however, it is very important to stay in touch with the school.
    has a major impact on psychological well-​being. It is essential that a psychosocial approach is integrated in care, in
    Most studies in this field indicate that children with cancer, as terms of systematic attention to all quality-​of-​life domains (phys-
    a group, are not very different to control groups or comparison ical, emotional, social, and cognitive functioning). Multidisciplinary
    groups in terms of adaptation or adjustment and psychopathology meetings with all team members involved (physician, nurses, psych-
    in all stages of the disease and treatment. However, there are chil- ologists, social worker, teacher)is of utmost importance.
    dren who are at risk for difficulties in adapting to the disease, such
    as those who react with withdrawal or social isolation, or those with
    premorbid psychopathology. These children need intensive psycho- After treatment
    logical support during and after treatment. Possible types of treat-
    ment, to help them get their life back on track, include contact with The end of the treatment in paediatric oncology can be considered
    fellow patients (through special camps or meetings), group training, a tremendous transition in care and can be very stressful. Children
    or individual therapy aimed at dealing with the disease and its con- and parents are supported during the entire treatment by the multi-
    sequences. For younger children, play therapy is a common way to disciplinary team. When treatment ends, this multidisciplinary
    work through emotions, while older children are often treated with guidance quickly diminishes and families must regain their own
    cognitive behavioural therapy techniques. Another group at risk lives. In this phase, studies show that children of different ages report
    for severe negative (neuro)psychological sequelae are the children problems with physical functioning as a consequence of loss of gen-
    treated with global central nervous system irradiation. eral condition and strength or ongoing tiredness. In this new phase,
    Children with cancer often feel anger and disappointment be- returning to school, picking up the ‘ordinary life’, and being released
    cause of restrictions on their participation in important areas of life, from hospital care are the main goals for the children, but these goals
    such as going to school or engaging in sports and other social ac- might become hampered because of physical impairments. Next to
    tivities. Sometimes it is difficult to stay in touch with friends due to physical functioning, adolescents often experience a sense of dis-
    poor clinical conditions. Children who attend school during treat- tance towards their peers, which sometimes makes it difficult to re-
    ment often feel ‘different’because of their changed appearance or integrate. Individual counselling can help children and adolescents
    80 CHAPTER 10 Psychosocial Care

    to express feelings of loneliness or of being misunderstood, and help significant difference. A dose of hydrocortisone seems to reduce the
    them to bridge the gap to ‘normal life’. serious side effects of dexamethasone on sleep, mood, and cognition.
    When chemotherapy or radiotherapy ends, parents typically Survivors who have been treated for brain tumours are more
    realize that cure only lasts if there is no relapse, and that there are less likely to have neuropsychological and learning difficulties as well
    treatment options if the disease recurs. In this period, parents have as psychosocial difficulties. Important risk factors for brain-​tumour
    to deal with feelings of immense uncertainty and with diminishing patients are hydrocephalus, dose of irradiation, time since cranial
    social support, which can be very stressful. Besides the fear of re- irradiation, effects of multiple treatments, female sex, and also
    lapse, parents are being informed about late effects of the cancer younger age at diagnosis. Improvements of the understanding of
    treatment, which also can give rise to post-​treatment stress. At the the exact neuropsychological deficits that accompany both disease
    same time, they have to pick up their lives, or are confronted with and treatment factors remain important goals. Prolonged intensive
    demands of the outside world such as employers who expect them rehabilitation and monitoring of neuropsychological functioning is
    to return to work quickly. This increases the pressure on parents at needed for this subgroup of patients to offer timely treatment inter-
    a time when they need to pay some attention to themselves after ventions for neuropsychological sequelae.
    such a long and exhausting period. Parents often describe this as the Studies on the long-​term psychosocial consequences of treatment
    first period in which they pay attention to their own functioning. on survivors show mixed results. Several studies report a reduced
    Physicians should be sensitive to symptoms of severe exhaustion in quality of life in physical and social functioning. Young adult sur-
    parents and refer them to psychologists or social workers if needed. vivors of childhood cancer achieve less developmental milestones,
    Early monitoring of psychosocial distress and pre-​existing prob- or at a later age, than their peers. They value their functioning lower
    lems is an important first step to identify families at risk for prob- in the areas of independence, and social and psychosexual devel-
    lems, and this is internationally endorsed by paediatric oncology opment. In comparison with healthy controls, survivors are less
    psychosocial standards of care. likely to be married or living together, and seem to have problems
    finding a job that provides financial independence. Some of the
    survivors report mood swings or post-​traumatic stress disorder
    Long-​term follow-​up (PTSD), a cluster of symptoms related to trauma such as intrusions,
    hypervigilance, and avoidance. However, the numbers of survivors
    Almost all children and adolescents who have been successfully with these problems are considerably fewer than for their parents.
    treated for cancer have to deal with negative health outcomes. They Parents of survivors can suffer from anxiety, tension, and un-
    may develop health problems as a result of treatment (e.g. second certainty. Also a large percentage of parents report PTSD symp-
    cancers, cardiac conditions, leukoencephalopathy, endocrine prob- toms, including intrusions and avoidance behaviour. Sometimes
    lems, infertility), but may also suffer cognitive or social disadvan- these avoidance behaviours compromise the follow-​up of the child
    tages in terms of academic achievement, finding a job or a partner, (parents are reluctant to revisit the hospital) and can be considered
    getting insurance. Regular screening of survivors concerning all the a risk factor. Raising a child who has had cancer is a challenging task
    important areas of functioning, in so-​called ‘late-​effects clinics’, is for parents. It is sometimes difficult to modify habits consolidated
    very important. Apart from screening for late effects, prevention is during treatment, but that now limit the child’s functioning. Finding
    another issue that needs attention in paediatric survivorship pro- a balance in caring for and protecting the child and, at the same time,
    grammes and should be aimed at addressing lifestyle issues such as promoting healthy development towards young adulthood is a very
    smoking, sunbathing, and diet. hard task.
    Neuropsychological consequences as a result of the treatment for
    childhood cancer can be severe and long-​lasting. Children treated
    for brain tumours, treated with high-​dose chemotherapy or with End-​of-​life issues
    cranial radiotherapy, are especially at risk. Treatment has both short-​
    term and long-​term impact on neuropsychological functioning due Despite the increased survival and progress in treating children with
    to acquired brain injury. Early brain damage has a cumulative effect cancer, about one quarter of parents and children have to face the
    on brain development, with increasing deficits in skills that have yet fact that cure is not possible. The treatment goal shifts from cure
    to emerge. Common cognitive deficits are in the domains of intel- to quality of life and dying, as quoted by Veronica, aged 13: ‘What
    ligence, attention, speed, and memory. Long-​term neuropsycho- really matters in life is not so much being able to value it. What really
    logical deficits for survivors of childhood acute lymphoblastic counts, is being able to “embrace” the moment when it finishes.’
    leukaemia (ALL) after treatment with chemotherapy appear only to What is a terminally ill child? It is not a dying person; it is a child
    be subtle. These mainly involve processes of attention and execu- who has arrived at a phase of life where death is close and cure is no
    tive functioning, while global intellectual function is relatively pre- longer possible. This phase could last for months. The approach to
    served. Younger age at diagnosis, female sex, and treatment intensity it does not need to be either ‘therapeutically renouncing’ or display
    have emerged as risk factors for ALL survivors. ‘ruthless obstinacy’. It is essential to build up the most appropriate
    With the replacement of prednisolone in many recent treat- modality to assist a child and the family in the palliative and ter-
    ment protocols by dexamethasone, the interest in the long-​term minal phase. The main purpose in this phase is to keep control over
    side effects of this treatment has increased. Studies report more both physical and psychological pain and discomfort, either with
    neuropsychological side effects with dexamethasone than with medical, psychological, or spiritual interventions, and to turn any
    prednisonolone, but randomized controlled trials do not reveal a experience into something positive.
    Conclusion 81

    Today, many children who die of cancer die at home, although always be in the best interest of the child. The parents may wish to
    this is highly variable between countries even within Europe. shield the child, which may only make the patient’s fear worse. Also,
    Psychosocial care should be focused on helping the family with their a conflict may arise when the family wishes to continue pursuing an
    daily life, supporting parents in dealing with the dying child, their unrealistic possibility of cure or, on the contrary, when the family
    own emotions, and also with any siblings. The hospital psychologist wishes to stop curative treatment prematurely. If members of the
    can play an important role in this situation, either in an advisory or healthcare team take time to try to understand and discuss the
    consultative role for the professionals visiting the families at home reasons for the family’s wishes, the inevitably painful conflicts can
    (e.g. general practitioner, palliative nursing team) or as caregiver. usually be overcome. Advanced Care Planning guidelines (ACP)
    The paediatric oncologist can have an important role as well, al- could be helpful in overcoming these difficulties. The continuation of
    though again, this differs from country to county. curative treatment beyond the point when cure is no longer possible,
    For some children, talking about death and the fear of death is the so-​called ‘ruthless obstinacy’ treatment, should be avoided.
    easier with a professional than a parent. The way in which children In their desperation to avoid the upcoming death, parents often
    think about death, and their understanding of it, depends on their look for alternative treatments. This can be interpreted as a way of
    age and cognitive and psychological development. Young children dealing with what is perceived as unacceptable. Many complemen-
    do not yet have a fully developed sense of death. They are cognitively tary treatments give parents feelings of control and the perception
    unable to grasp the sense of irreversibility and permanence of death. that they can contribute to the quality of life of their child. Many
    On the contrary, older schoolchildren have greater awareness of the parents experience anticipatory grief in the shape of depressive re-
    irreversibility of death, and teenagers generally have a fully devel- actions or anger. The identification of these feelings as normal re-
    oped awareness. Teachers and classmates at the child’s school should actions to abnormal circumstances can be helpful. Once a child is
    be informed about the child’s impending death. In this phase, the deceased, parents often seek assistance from professionals such as a
    classmates and the school will be an important help to the teacher. psychologist or grief counsellor. Furthermore, after a child dies, that
    The continuous participation of the child in school activities can be individual child’s medical history should be evaluated. This evalu-
    of great importance. Most children want to participate in normal ation should be made by the healthcare team as a group. It is critical
    life, including school life, for as long as possible, and being with their to reflect on all events, even minor ones, that occurred during the
    peers distracts them from being sick. Siblings of the sick child need course of the child’s treatment, and also on the choices that were
    good guidance as well. made and why. This will help the staff come to terms with their own
    Many parents struggle with whether or not they should tell their grieving and to learn from the experience in order to help future
    child that he/​she is dying. In a large Swedish study, parents of de- families. Furthermore, this could help parents to better understand
    ceased children were asked retrospectively whether or not they told and clarify conflicting final decisions, and to start to work through
    their child about their impending death, and how they experienced their bereavement.
    and looked back on their decision. This study included a group of
    500 parents: 34% of the parents talked with their child, and none of
    these parents had regretted it afterwards. In the other group, where Conclusion
    parents did not talk to the child, the majority (73%) also did not
    regret this afterwards. The main criterion was the age of the child. Even with the increased survival rates of children with cancer, uncer-
    Most parents chose to talk in terms of ‘not getting better’ or used tainties about the nature and extent of late effects remain. To achieve
    metaphors, thus creating space for the child to ask questions. This optimal support for children and adolescents who grow up with
    study shows that the perception and appreciation of the situation childhood cancer, physicians should have knowledge about their
    by parents themselves should be included in the guidance for these developmental trajectory and pay attention to this during treatment
    end-​of-​life matters. For many parents, it is confusing, after the fatal and follow-​up. Therefore, periodical evaluation of their ongoing psy-
    news, to see that their child wants to continue living as if nothing chosocial, educational, neuropsychological, and vocational needs
    has changed. Children often want to go to school and play in spite of during their development should be an integral component of the
    their impending death. comprehensive care of children and adolescents with cancer.
    Communication remains the basic issue when the diagnosis of
    cancer has to be reported to the patient and the family. How best to
    do it becomes the first step in a communicative process that involves FURTHER READING
    the medical team and the family, and allows for growth and change Aarsen FK, et al. (2009) Cognitive deficits and predictors 3 years after
    over time until the final stage. Parents most often manage what and diagnosis of a pilocytic astrocytoma in childhood. J Clin Oncol
    how their children are told about cancer. Adolescents or young 27, 3526.
    adults vary in their preferences as to how much information should Annet RD, et al. (2015) Monitoring and assessment of neuropsycho-
    logical outcomes as a standard of care in pediatric oncology. Pediatr
    be disclosed to them. Some adolescents and young adults with
    Blood Cancer 62, S460.
    cancer prefer to be fully informed about their disease, but approxi-
    Barakat LP, et al. (1997) Families surviving childhood cancer: a com-
    mately one third among those who survived preferred not to know.
    parison of posttraumatic stress symptoms with families of healthy
    Many parents find it difficult to talk to young children about issues children. J Pediatr Psychol 22, 843.
    as serious as the diagnosis of cancer or impending death. Without Eiser C, et al. (2000) Examining the psychological consequences of sur-
    proper communication, starting at diagnosis, the child’s reaction to viving childhood cancer: systematic review as a research method in
    the disease, treatment, and new or altered life situation can cause pediatric psychology. J Pediatr Psychol 25, 449.
    additional stress for the parents at any time. Parental wishes may not
    82 CHAPTER 10 Psychosocial Care

    Geenen MM, et al. (2007) Medical assessment of adverse health out- Sahler OJ, et al. (2005) Using problem-​solving skills training to reduce
    comes in long-​term survivors of childhood cancer. JAMA 297, 2705. negative affectivity in mothers of children with newly diagnosed
    Hinds PS, et al. (2007) Dexamethasone alters sleep and fatigue in pedi- cancer: report of a multisite randomized trial. J Consult Clin Psychol
    atric patients with acute lymphoblastic leukemia. Cancer 110, 2321. 73, 272.
    Hoekstra-​Weebers JE, et al. (2001) Psychological adaptation and social Skinner R, et al. (2007) Long-​term follow-​up of children treated for
    support of parents of pediatric cancer patients: a prospective longi- cancer: why is it necessary, by whom, where and how?. Arch Dis
    tudinal study. J Pediatr Psychol 26, 225. Child 92, 257.
    Jankovic M, et al. (1994) Association of 1800 cGy cranial irradiation Stam H, et al. (2005) The course of life of survivors of childhood cancer.
    with intellectual function in children with acute lymphoblastic leu- Psycho-​Oncology 14, 227.
    kaemia. Lancet 344, 224. Stam H, et al. (2001) Social and emotional adjustment in young sur-
    Kars MC, et al. (2010) Factors influencing parental readiness to let vivors of childhood cancer. Support Care Cancer 9, 489.
    their child with cancer die. Pediatr Blood Cancer 54, 1000. Temming P, Jenney ME (2010) The neurodevelopmental sequelae of
    Kazak AE, et al. (2006) An integrative model of pediatric medical trau- childhood leukaemia and its treatment. Arch Dis Child 95, 936.
    matic stress. J Pediatr Psychol 31, 343. Trask PC, et al. (2003) Parent and adolescent adjustment to pediatric
    Kreicbergs UC, et al. (2007) Parental grief after losing a child to cancer: associations with coping, social support, and family func-
    cancer: impact of professional and social support on long-​term out- tion. J Pediatr Oncol Nurs 20, 36.
    comes. J Clin Oncol 25, 3307. Van der Geest IM, et al. (2013) Emotional distress in 652 Dutch
    Landolt MA, et al. (2003) Incidence and associations of parental and very long-​term survivors of childhood cancer, using the Hospital
    child posttraumatic stress symptoms in pediatric patients. J Child Anxiety and Depression Scale (HADS). J Pediatr Hematol/​Oncol
    Psychol Psychiatry 44, 1199. 35, 525.
    Langeveld NE, et al. (2002) Quality of life in young adult survivors of Varni JW, et al. (1994) Perceived social support and adjustment of chil-
    childhood cancer. Support Care Cancer 10, 579. dren with newly diagnosed cancer. J Develop Behav Pediatr 15, 20.
    Last BF, et al. (2007) Positive effects of a psycho-​educational group Vrijmoet-​Wiersma CM, et al. (2008) Assessment of parental psy-
    intervention for children with a chronic disease: first results. Patient chological stress in pediatric cancer: a review. J Pediatr Psychol
    Educ Counsel 65, 101. 33, 694.
    Lotz J, et al. (2013) Pediatric advance care planning: a systematic re- Warris LT, et al. (2016) Hydrocortisone as an intervention for
    view. Pediatrics 131(3), e873–880. dexamethasone-​induced adverse effects in pediatric patients with
    Martel D, et al. (2005) Use of alternative and complementary therapies acute lymphoblastic leukemia: results of a double-​blind random-
    in children with cancer. Pediatr Blood Cancer 44, 660. ized controlled trail. J Clin Oncol 34, 2287.
    Moore BD (2005) Neurocognitive outcomes in survivors of childhood Wiener L, et al. (2015) Standards for the psychosocial care of children
    cancer. J Pediatr Psychol 30, 51. with cancer and their families: an introduction to the special issue.
    Nazemi KJ, Butler RW (2011) Neuropsychological rehabilitation for Pediatr Blood Cancer 62, S419.
    survivors of childhood and adolescent brain tumors: a review of the Zebrack BJ, et al. (2002) Psychological outcomes in long-​term survivors
    past and a vision for a promising future. J Pediatr Rehab Med 4, 37. of childhood leukemia, Hodgkin’s disease, and non-​ Hodgkin’s
    Patenaude AF, Kupst MJ (2005) Psychosocial functioning in pediatric lymphoma: a report from the Childhood Cancer Survivor Study.
    cancer. J Pediatr Psychol 30, 9. Pediatrics 110, 42.
    11
    Palliative Care for Children
    with Advanced Cancer
    Michelle Koh, Finella Craig, and Joanne Wolfe

    Introduction face the uncertainty that surrounds their child’s prognosis. This un-
    certainty impacts the child and family’s experience, their decision
    The World Health Organization (WHO) defines palliative care for making on how intensively to pursue cancer-​directed therapy, care
    children as ‘the active, total care of the child's body, mind and spirit, planning, and prognostic communications. Acknowledging the un-
    and also involves giving support to the family’ and recommends that: certainty of prognosis and developing an awareness of the child and
    family’s hopes and fears may lead to a greater understanding of the
    • It begins when illness is diagnosed, and continues regardless of challenges faced and promote more open and honest communica-
    whether or not a child receives treatment directed at the disease. tion at this critical period.
    • The child's physical, psychological, and social distress should be For many families, maintaining their belief that there might be a
    addressed. ‘miracle round the corner’ is crucial, yet can coexist alongside emo-
    • It requires an interdisciplinary approach which includes the family tional and practical preparation for their child’s death. Healthcare
    and utilizes available community resources. professionals tend to view hope as primarily related to a positive
    • It can be provided in tertiary-​care facilities, in community settings outcome, many finding it difficult to maintain hope in the face of
    (school, respite care, primary care), and in the child's home. a poor prognosis. Parents are more likely to report that physician
    Generalist palliative care involves care delivered by the interdiscip- communication always makes them feel hopeful when they are re-
    linary paediatric oncology team. Specialist palliative care is delivered ceiving more prognostic information and high-​quality communi-
    by interdisciplinary clinicians with additional training and expertise cation. Therefore, contrary to how many professionals may feel, an
    in palliative care. All children with cancer should receive generalist open discussion about prognosis can support hope, even when the
    palliative care from the time of diagnosis. Specialist palliative care is prognosis is poor.
    especially important for children with more complex symptom and/​ Palliative care services in this situation can often help in the
    or quality-​of-​life needs and at more advanced stages of the illness. following ways:
    At any illness stage, palliative care is delivered alongside disease-​ • Assessing physical and psychosocial symptoms
    directed therapy in concordance with family goals of care. • Supporting patients and families in establishing goals of care and
    In this context, the palliative care team can be involved in symptom with decision making with regards to ongoing treatment
    management and provide outreach home support for the family. The • Coordinating patient care based on the individual needs of the
    level of support provided by the service should be flexible, according patient.
    to the child and family’s requirements, which, to some extent, de-
    pends on the child’s diagnosis, treatment, and prognosis. At times Decision making about cancer-​directed therapies
    of crises (such as relapse) or when cure is no longer possible, the
    Bluebond-​Langner et al. studied parental decision making regarding
    palliative care team may take a more active lead in the child’s care
    further cancer-​ directed treatments for children when standard
    and management, as a seamless continuation of the care they have
    cancer treatment was unsuccessful. Parents bring to these discus-
    already been receiving.
    sions both the knowledge they have acquired throughout their
    child’s illness as well as their emotions as parents. They sought meet-
    Communication and decision making ings with the oncologist regarding further investigative procedures,
    and about half the parents looked for options beyond what was
    When a child with cancer relapses, it is typical for a family to fluc- offered. No parent initiated discontinuation of cancer-​directed or
    tuate between hoping for a cure and contemplating death as they symptom-​directed therapies.
    84 CHAPTER 11 Palliative Care

    Parents do not see cancer-​directed therapy and symptom-​directed Box 11.1 Features of ongoing oncological treatments
    care as mutually exclusive. Physicians and parents negotiate care
    throughout the child’s treatment regardless of whether it is directed. Chemotherapy
    Indeed, experience suggests that labels such as ‘curative’ or ‘palliative • Low-​toxicity oral, intravenous, or intrathecal chemotherapy to halt
    intent’ establish a false dichotomy, since ultimately parents hope that or slow disease progression and manage symptoms of disseminated
    their child lives as long as possible. disease.
    • Examples:
    Advanced-​care planning — Oral etoposide in children with metastatic solid tumours, or oral
    temozolamide in children with relapsed brain cancers.
    End-​of-​life care planning requires effective communication to guide — Intrathecal methotraxate for symptoms due to CNS leukaemia.
    parents through clarifying the goals of care and establishing agree- • Monitor the efficacy of treatment closely as treatment needs to be
    ment on what treatments are appropriate to achieve these goals, reconsidered or stopped if symptoms progress or side effects out-
    including resuscitative and palliative measures. Providing choice for weigh potential benefits for quality of life.
    place of end-​of-​life care is crucial and may impact on family deci- Radiotherapy
    sions about hospital-​based interventions. Among families who felt • Useful for controlling distressing symptoms from a focal site and can
    they had time to plan where their child would die, an overwhelming prophylactically treat sites of disease before they cause symptoms
    majority of parents preferred to have their child die at home. (e.g. radiation to bulky bony disease before the bone fractures).
    Advanced-​care planning requires time and often needs to be • Given as hypofractionated regimens (i.e. fewer fractions with bigger
    done in stages, with discussions around the appropriate level of doses per fraction).
    cardio-​pulmonary resuscitation, life-​sustaining treatment, and in- • Symptomatic gain should be greater than the burden of any side
    effects or negative impact on quality of life (e.g. attending hospital,
    tensive care, and about the child and family’s preferred place of care need for anaesthetic).
    and place of death. This should be accompanied by assurance that • Examples:
    symptom management and supportive care will be proactive in
    — Pain caused by bulky disease, nerve compression, bone pain.
    whichever setting they choose. — Obstruction of any hollow passage or viscus (e.g. spinal cord com-
    There should be an understanding and acknowledgment of the pression, SVC or airway obstruction).
    past experiences as well as the cultural values and social set-​up of — Bleeding from rupture or breakdown of tumour: haemoptysis, va-
    the child and family. There must be clarity on the practicalities of ginal or gastrointestinal bleeding, haematuria.
    what the different options of home, hospice (also called palliative — Fungating or disfiguring tumours, which can cause immense dis-
    tress by their appearance and discharge. Tumour-​volume reduction
    care unit), or hospital care entail, so that the families have realistic can help the surrounding skin and soft tissues to heal.
    expectations to inform their decision making.
    Surgery
    The care plan must be communicated to the wider multidiscip-
    • Considered early in situations where symptoms can be anticipated,
    linary team, including out-​of-​hours services and the ambulance as the child needs to be well enough to tolerate surgery.
    service. Patient-​held records of discussions, decisions, and contacts • Examples:
    are useful communication tools. — Debulking of tumour can help with pain control.
    — Drains for effusions/​ascites.
    Cancer-​directed therapy — Decompression (e.g. for spinal cord compression, relief of bowel
    The purpose of ongoing cancer-​directed treatments is to improve the obstruction).
    quality of life, relieve symptoms, and prolong the symptom-​free period. — Stenting of intraluminal lesions can often be done endoscopically
    (e.g. biliary stents, nephrostomy stents, urinary catheterization).
    As a principle, side effects and toxicities must be minimal and manage-
    — Stabilization of fractures, prophylactically or post fracture.
    able, but long-​term toxicity is unlikely to be an issue. Decisions around
    these treatments are taken by the oncologist, in conjunction with the
    palliative care team, child, and family (Box 11.1).
    Although the parents often provide vital information regarding
    the child’s symptoms, it is important to focus on what the child may
    Common and difficult symptoms be able to tell us, as well as clinical signs. This is crucial for managing
    subjective symptoms such as pain, nausea, and breathlessness. It val-
    Many studies have shown that children with advanced cancer ex- idates the child’s symptom experience, and builds trust between the
    perience substantial suffering from symptoms amenable to palli- child, family, and professionals.
    ation. Increased availability of palliative care services is associated The child’s symptom experience is affected by the pathophysio-
    with improvement in symptom control. The approach to symptom logical processes, and other biological and psychosocial factors
    management involves: (Box 11.2).

    • Assessment of the child’s symptoms Pain


    • Evaluation of potentially reversible causes of the symptoms It is crucial to elicit what the cause of pain is so that targeted onco-
    • Planning and initiation of treatment logical treatments may be considered. Assessment of the possible
    • Reassessment following intervention. pain mechanism (nociceptive or neuropathic) will inform the treat-
    ment approach and drug selection. Nociceptive pain is caused by
    This is an ongoing process as the disease progresses and symptoms superficial and/​ or deep-​ tissue injury (somatic pain), or organ
    change over time. damage (visceral pain). Neuropathic pain is caused by nerve injury,
    Common and difficult symptoms 85

    Box 11.2 Principles of symptom management The dose is then gradually increased until effective. If regular back-
    ground analgesia is required, the use of a long-​acting opioid prepar-
    • Symptom management should be planned in advance. ation is preferable. It is important to continue the use of short-​acting
    • It requires a holistic approach, not just drug management. morphine as needed, to manage any breakthrough pain. There is no
    • Parents and patients should be empowered to have some control upper dosage limit for opioid analgesics because there is no ‘ceiling’
    over the situation. They should:
    analgesic effect. The appropriate dose is the dose that produces pain
    — Know what to expect
    — Know how to deal with it relief for the individual child.
    — Know who to call for help. Morphine is currently the first-​line opioid of choice as it is inex-
    • Assessment should include the use of developmentally appropriate pensive and a wide range of formulations are available. Its efficacy,
    instruments when available. Uncontrolled distressing symptoms con- pharmacokinetics (caution in renal or hepatic impairment), and side
    stitute a medical emergency and should be managed intensively. effects (constipation, nausea, urinary retention, itch, drowsiness)
    are well understood. Oxycodone’s efficacy and side-​effect profile are
    similar to that of morphine. It has a different molecular structure
    either in the peripheral nerves or the central nervous system. Pain and, therefore, can be an effective alternative to morphine in patients
    may arise from a combination of nociceptive and neuropathic mech- with intolerable side effects or inadequate analgesia from morphine.
    anisms, and a combined management approach is often required. Fentanyl is a synthetic opioid, useful in patients in renal failure as
    In assessing pain, as much information as possible should be de- it metabolizes to inactive metabolites in the liver and has a shorter
    rived directly from the child. Self-​reporting using pain scales remain half-​life. It is available in transdermal or transmucosal prepar-
    the ‘gold standard’ in measuring pain, but the child needs to be com- ations. It can simplify analgesic management in patients with poor,
    petent in using these. Parents and familiar carers can be good surro- deteriorating, or even absent renal function, but is not suitable for
    gates as they know the child’s normal baseline behaviour and will be the initiation or titration phases in managing escalating pain be-
    familiar with the child’s pain behaviour. However, there is a risk that cause of the time lag to achieve steady state of the transdermal patch.
    they may under-​report pain, for fear of acknowledging a symptom Methadone is an opioid with N-​methyl-​D-​aspartate (NMDA)
    that suggests disease progression. Physiological changes such as antagonistic properties and, therefore, effective in difficult pain
    heart rate and blood pressure may help in identifying pain, but in with neuropathic qualities. Methadone should only be commenced
    a dying child these are likely to be affected by many confounding by practitioners experienced in its use. This is due to wide inter-​
    factors. individual variation in response, variable conversion ratios with
    The WHO three-​step analgesic ladder introduced in 1986 has now other opioids, complex pharmacokinetics (long half-​ life, large
    been modified for children to a two-​step approach. The principles volume of distribution, and risk of accumulation), and side effects
    of dosing at regular intervals (‘by the clock’), administering by the (prolongation of QT interval). It is available as a liquid and can be
    appropriate route (‘by the mouth’), and individualizing treatment to very useful in countries where liquid long-​acting morphine is not
    each child (‘by the child’) remain central. available.
    For Step 1—​mild pain—​ paracetamol and ibuprofen are the
    medicines of choice. Other non-​steroidal anti-​inflammatory drugs Routes of administration
    (NSAIDs) may be considered if appropriate to the clinical situation. Medications should always be given by the most acceptable route
    NSAIDs (e.g. ibuprofen, diclofenac) have analgesic, antipyretic, and for the child. Practical considerations, such as place of care and ease
    anti-​inflammatory properties. They are especially useful for mus- of administration, as well as patient preference, may also influence
    culoskeletal and liver pain. NSAIDs may cause gastric irritation, route of administration and, as a consequence of this, drug of prefer-
    so their regular use should be accompanied by gut protection (e.g. ence. The intramuscular route is to be avoided in children.
    Ranitidine, Omeprazole). As they cause platelet dysfunction, their
    • Oral medications: the route of choice as long as the child is able
    potential benefit must be balanced against the child’s risk of bleeding.
    to take and tolerate it. Enteral medications can often be continued
    For Step 2—​moderate to severe pain—​an opioid should be ad-
    throughout the child’s terminal phase, particularly if they already
    ministered in addition to paracetamol and ibuprofen. Morphine is
    have a nasogastric or gastrostomy tube.
    the first-​line opioid but alternative opioids may be needed in the
    event of poor analgesic response or significant side effects. Codeine • Transdermal opioid patches: appropriate for use in patients who
    was previously part of the recommendations but as a significant pro- are unable to take or comply with oral medications and who have
    portion of children are unable to effectively convert codeine to its stable pain. They are unsuitable for patients with escalating pain as
    active metabolites or may be hypermetabolizers, its use in children the onset of effect of a higher-​dose patch is too slow.
    is restricted in many countries. Furthermore, codeine has a high • Continuous intravenous infusion: may be the preferred route
    toxicity-​to-​efficacy ratio; children experience substantial side effects of choice for a child who already has a central line and cannot
    even at very low doses. It is preferable to initiate lower doses of safer, manage oral medications.
    more effective opioids such as morphine, and titrate as appropriate. • Continuous subcutaneous infusion: should be considered in
    preference to intravenous medication if the child does not have
    Opioids: morphine, oxycodone, fentanyl, methadone pre-​existing venous access.
    The aim in using opioids is to obtain a dose that provides adequate • Buccal/​sublingual preparations: useful in children who cannot
    analgesia with a minimal or acceptable degree of side effects. manage oral medications, particularly for breakthrough symp-
    A short-​acting form of opioid, usually morphine, may be started toms. Most, however, will have a short duration of action, so are
    on an ‘as needed’ basis or, if the pain is severe, on a regular basis. not ideal for sustained symptom relief.
    86 CHAPTER 11 Palliative Care

    Neuropathic pain morphine preparations can be tried in the first instance, al-
    When nerves are invaded or compressed by tumour, it results in though some patients respond better to a continuous infusion
    changes to normal pain transmission, causing neuropathic pain. of morphine. A benzodiazepine can be used in conjunction as
    Often, there is considerable overlap between neuropathic and noci- an anxiolytic.
    ceptive pain. Some children report sharp pain associated with tin- Nausea and vomiting
    gling and numbness or a burning sensation or a deep ache. Others
    may not be able to articulate this type of pain clearly. Knowledge The following simple, practical measures can often improve nausea
    of the disease pathology, location, and invasive nature should alert significantly:
    physicians to the potential for neuropathic pain. It should be con- • The child should be given small, manageable amounts of food they
    sidered when the patient has a poorer than expected response to enjoy, at the time they feel most able to eat. Several small meals are
    escalating doses of opioids. preferable to three large meals a day.
    As neuropathic pain is mostly mediated through NMDA recep- • Strong smells should be avoided, such as cooking smells and
    tors, a single class of analgesic is not usually adequate; a combination perfumes.
    of opioids and adjuvant analgesics benefit patients most. • Constipation must be actively managed.
    Adjuvant analgesics • The number of medications and drug volumes should be reduced
    as much as possible. Medication should be timed so that large
    Adjuvant analgesics can be opioid-​sparing, and it is important to
    amounts are not being given together.
    consider if the child’s pain is likely to be neuropathic in origin, or if
    it is not responding to escalating opioid therapy. Although there is The key to successful management of nausea and vomiting is to elicit
    no evidence for the use of these drugs for pain management in chil- the causes, as this informs the antiemetic choice (Table 11.1).
    dren, there is a large body of clinical experience. The following are
    Constipation
    the most commonly used adjuvant neuropathic analgesics:
    Constipation in children with advanced cancer may be secondary to
    • Antidepressants: amitriptyline, nortriptyline
    decreased mobility and oral intake, as a direct result of tumour ob-
    • Anticonvulsants: gabapentin, carbamazepine struction or as a side effect of opioids or anti-​cholinergic drugs. As
    • NMDA antagonists: ketamine, methadone a general rule, a laxative should be prescribed when prescribing any
    Other alternative agents may also be helpful for specific situations, opioid for a patient. Constipation should be proactively prevented as
    such as bisphosphonates in the management of pain associated with it is often difficult to manage once established; it is usually multifac-
    metastatic bone disease. torial and a combination of laxatives is often required. Three classes
    Any patient with intractable pain despite optimal management of of laxatives are commonly used:
    systemic analgesia should be discussed with an anaesthetics team, • Osmotic laxatives (macrogol)
    for consideration of an appropriate intervention such as epidural • Softening agents (lactulose, docusate sodium)
    analgesia or a nerve block. Patients most likely to benefit include
    • Stimulant agents (senna, sodium picosulphate, bisacodyl)
    patients with significant locally advanced disease, neuropathic pain,
    or marked movement-​related pain.

    Breathlessness Table 11.1 Choice of antiemetics for different causes of vomiting


    Breathlessness may be secondary to disease, causing airway ob- Cause Drug Class/​Action
    struction, pulmonary metastases, pleural effusion or infection, and (site of action)
    pooling of secretions. This may be compounded if there is bone Raised intracranial Cyclizine Antihistamine
    marrow disease causing anaemia. pressure Hyoscine Antimuscurinic
    (vomiting centre) Levomepromazine 5HT2-​antagonist
    Breathlessness is a subjective symptom: its severity is what the
    patient says it is. It can be frightening. General measures such as Decreased intestinal Metoclopramide D2-​antagonist
    motility or dysmotility Domperidone D2-​antagonist
    reassurance, optimizing the child’s position, and, in older children, Erythromycin Motilin-​receptor
    relaxation techniques, may help. A stream of cool air on the face (e.g. agonist
    by opening a window or having a fan on) stimulates the trigeminal Metabolic or drug-​induced Haloperidol D2-​antagonist
    nerve and can reduce the perception of breathlessness. Oxygen may (chemoreceptor trigger Metoclopramide D2-​antagonist
    be recommended if it reduces the child’s sensation of breathlessness, zone)
    rather than aiming to improve the child’s oxygen saturation levels. Chemotherapy-​induced Ondansetron 5HT3-​antagonist
    (gut wall/​chemoreceptor
    As the child becomes less conscious, they become less able to
    trigger zone)
    clear their secretions effectively. Glycopyrronium or a hyoscine
    ‘Broad-​spectrum’ agent for Levomepromazine 5HT2-​antagonist
    hydrobromide patch reduces the volume of secretions, but can also resistant nausea
    make them thicker and cause anti-​cholinergic side effects (e.g. dry (vomiting centre)
    mouth or eyes). They should be used judiciously, with appropriate Chemotherapy-​induced Aprepitant NK1-​antagonist
    suctioning, gentle chest physiotherapy, and mouth care. nausea and vomiting (used in combination
    Opioids reduce the sensation of breathlessness by reducing with a corticosteroid or
    5HT3-​antagonist)
    the ventilatory response to hypercapnia and hypoxia. Oral
    Use of steroids for symptomatic relief 87

    Rectal enemas (phosphate enemas) and suppositories (glycerine sup- used: cyclizine as the first line, with haloperidol or levomepromazine
    positories) can be tried if laxatives are ineffective but may only empty added in if required. Laxatives (combination of senna and docusate)
    the rectum, and can cause significant pain if the stool is not softened. should be used to maintain a comfortable stool without colic.
    Their use should be limited in children with profound neutropenia. Octreotide can be effective in reducing the amount of gastrointes-
    For patients with opioid-​induced constipation, when the response tinal secretions in children with severe outlet obstruction.
    to laxatives is inadequate, methylnaltrexone (a peripheral opioid re- Parenteral hyoscine butylbromide can be rapidly effective for
    ceptor antagonist) may be helpful. It acts on the bowel but does not colic. If the coeliac or superior mesenteric plexus is involved, there
    reverse the central analgesic effects of opioids. It is contraindicated may be severe neuropathic pain that will require opioids and an anti-​
    in bowel obstruction and is available as a subcutaneous or intra- neuropathic pain agent.
    venous bolus injection.
    Bone marrow failure
    Fatigue Decisions surrounding the appropriate use of transfusions need to be
    Fatigue is the most common symptom experienced by children with addressed sensitively as families who have been through intensive onco-
    terminal cancer and is often under-​recognized and undertreated. It logical treatments will be used to having anaemia or thrombocytopenia
    is an overwhelming tiredness, out of proportion to activity and ex- managed aggressively. Continued transfusions are usually warranted
    ertion, and not adequately relieved by rest. It is important to address when the child with advanced cancer remains active. In the context
    underlying aetiologies (e.g. pain, nutrition, anaemia) including psy- of end-​of-​life care, transfusions are recommended only if the child’s
    chological symptoms such as depression and anxiety. Increasingly, symptoms are likely to be ameliorated by the transfusion. The burden
    methylphenidate or dextroamphetamine is recommended for of transferring the child and spending several hours in hospital must be
    symptomatic relief of fatigue, regardless of the suspected aetiology. considered, as well as the potential fluid shifts the transfusion will cause.
    Physiotherapy and sports activities may be effective in relieving fa- Bleeding is a major distressing symptom of bone marrow failure.
    tigue for children, even near the end of life. The child may bleed from fungating tumours, nose and gums, and
    pressure areas. Management includes:
    Agitation
    • Preparing the family adequately; dark towels and sheets can help
    Agitation and restlessness are common at the end of life and can be
    moderate the anxiety of the child and witnesses.
    distressing for parents and carers. Symptomatic causes of agitation,
    such as pain or vomiting, should be assessed and treated. Fear and • Positioning the child: for example, in acute haemoptysis or a se-
    unaddressed anxieties can be a major factor and early, open discus- vere nose bleed, lie them on their bleeding side, or upright/​re-
    sion with the child, allowing them to express their fears and con- clined with head tilted forward.
    cerns, may reduce agitation as death approaches. Calm reassurance • Topical adrenaline 1:1000 or tranexamic acid intravenous prep-
    is crucial, and other measures, such as careful attention to lighting, aration, soaked in gauze, applied to bleeding point. A mouthwash
    particularly at night or if a child’s vision is failing, as well as relax- preparation of tranexamic acid is also available.
    ation techniques, may be helpful. If medication is necessary, a short-​ • An anxiolytic such as buccal midazolam can help calm the
    acting benzodiazepine such as buccal midazolam or sublingual distressed child.
    lorazepam is often effective. For more severe agitation, haloperidol
    Anxiety and depression
    or levomepromazine can be very effective.
    Anxiety and depression are common emotions in children with life-​
    Gastrointestinal complications: malignant ascites/​ limiting conditions. It is important to assess if the symptoms are
    intestinal obstruction appropriate reactions to the difficult situation, or more severe and
    These complications are less common in children than adults, and entrenched emotions that require intervention.
    tend to happen late and close to end of-​life. They can be very difficult The child psychiatry team should be consulted on the med-
    to manage. ical management of psychological symptoms. Serotonin-​reuptake
    Malignant ascites most commonly occurs in patients with liver inhibitors such as fluoxetine can be effective; however, they re-
    involvement or peritoneal disease. Although, in children, draining quire several weeks to achieve maximum benefit. When indi-
    ascites will usually require sedation or a general anaesthetic, it is an cated, such medications should be started early. Psychostimulants
    effective way to relieve the pain and distension. Importantly, the as- (dextroamphetamine and methylphenidate) deserve special con-
    cites is likely to recur, requiring repeated procedures unless a cath- sideration in treating depression near the end of life, because they
    eter is left in situ. The risk of hypoalbuminaemia and hypovolaemia take effect quickly. Non-​medical interventions, such as counsel-
    from the draining of the peritoneal fluid also needs to be considered. ling, cognitive behavioural therapy, drama, art, or play therapy, are
    Intestinal obstruction can be caused by intraluminal, intramural, or also important. Acute anxiety can be managed with a short-​acting
    extramural obstruction, motility disorders, or severe constipation. benzodiazepine such as buccal midazolam or sublingual lorazepam.
    Surgery should only be considered if there is clear benefit. Medical
    management with antisecretory and antiemetic medication should
    aim to control nausea and reduce the frequency and severity of Use of steroids for symptomatic relief
    vomiting to an acceptable level. The route of drug delivery will
    need to be parenteral, and a nasogastric tube would only be recom- The rationale for the use of steroids in relieving symptoms of raised
    mended to decompress the stomach if there is faeculant vomiting or intracranial pressure is the effect of reducing tumour-​associated
    gastric outlet obstruction. ‘Broad-​spectrum’ antiemetics should be oedema. There are also other indications for the use of steroids in
    88 CHAPTER 11 Palliative Care

    cancer management, such as spinal cord and/​or nerve compression, loss of skills and abilities; social losses such as social isolation and
    bone pain, superior vena caval obstruction, and airway obstruction. pressures on the parents’ marriage; financial losses as parents often
    The initiation of steroids must be carefully considered and the have to reduce or give up their jobs to care for their child, in addition
    benefits be weighed against the many potential side effects. For ex- to the financial cost of care; and the emotional loss of the parents’
    ample, steroids frequently stimulate appetite but, if the tumour has hopes and dreams for their child.
    caused the patient to have a poor or unsafe swallow, they may be- Families need to be allowed to grieve and adapt to these changes
    come highly distressed, being unable to satisfy their hunger. Other at their own pace. Siblings have to cope with different relationships
    side effects include gastritis or peptic ulceration, behavioural prob- with their parents and the unwell child; grandparents grieve not
    lems, agitation, and extreme exacerbation and lability of mood only for their grandchild, but also the heartbreak that their child is
    (tearfulness, physical aggression). Longer-​term side effects such as enduring.
    diabetes, osteoporosis, and muscle wasting may or may not be rele- Support for the children may be available through schools and
    vant. The use of steroids should always be accompanied by gastric hospices, music, drama, or play therapy; voluntary sector organiza-
    protection such a proton-​pump inhibitor. tions such as children’s hospices provide sibling support; and adults
    If the patient is on a short course of steroids (less than seven days), and older children may find pre-​bereavement counselling valu-
    the treatment can be stopped abruptly. If the course is longer, a able, with therapy continuing through and after their bereavement.
    weaning regimen may help prevent rebound symptoms. Social-​work input is invaluable in helping the family obtain their
    financial entitlements and funding for extra care or home adapta-
    tions. In the UK, there are voluntary organizations (e.g. Rainbow’s
    Issues around feeding at the end of life Trust) which provide practical help with such as household chores
    or care of the other children.
    Loss of appetite and loss of interest in food is common in children who The sick child should be encouraged to maintain their inter-
    are dying. Fluid and caloric requirements decrease as their underlying ests and have things to look forward to. This often includes con-
    illness deteriorates and activity levels decrease. Artificial hydration and tinuing access to an appropriate level of educational input, play,
    feeds need to be given judiciously. If the child is already on artificial and social interaction with their peers. It is frequently useful, with
    hydration, reassess the child’s likely fluid requirement and reduce the the family’s permission, to provide relevant information for the
    amount given appropriately. If the child had previously been feeding school.
    orally, it is not usually beneficial to start artificial hydration. Close at- After the child’s death, the family may need to be supported
    tention should also be paid to mouth care and care of pressure areas. through the activities of mourning, be given information about the
    At the end of life, artificial hydration could alleviate signs of de- sources of support, and referred to specialist bereavement counsel-
    hydration but worsen peripheral oedema, ascites, pleural effusions, ling if needed.
    respiratory distress, and frequency of urination. The potential bene-
    fits of artificial hydration therapy need to be considered against the
    distress of worsening fluid retention. FURTHER READING
    Bennett M (2010) Palliative medications: drugs for neuropathic pain.
    Eur J Palliat Care 17(4), 167–​9.
    Spiritual distress Bluebond-​Langner M, et al. (2007) Understanding parents' approaches
    to care and treatment of children with cancer when standard
    Healthcare professionals often feel uncomfortable and inadequate in therapy has failed. J Clin Oncol 25(17), 2414–​19.
    Bruera E, et al. (2003) Patient-​controlled methylphenidate for the
    addressing the spiritual needs of the dying child and their family. It is
    management of fatigue in patients with advanced cancer: a prelim-
    important to access the spiritual and faith leaders in the interdiscip-
    inary report. J Clin Oncol 21, 4439–​43.
    linary team or in the child’s community. Often, a multi-​faith chap- De Graves S, Aranda S (2008) Living with hope and fear-​-​the uncer-
    lain will be expert in exploring spiritual distress in a non-​religious tainty of childhood cancer after relapse. Cancer Nurs 31(4), 292–​301.
    context with families who may not have a specific religious belief. Dussel V, et al. (2009) Looking beyond where children die: deter-
    The main spiritual need described by parents is maintaining con- minants and effects of planning a child’s location of death. J Pain
    nection with their child. Parents maintain connection at the time of Symptom Manage 37(1), 33–​43.
    death by physical presence and, after the death, through memories, Friedrichsdorf SJ, et al. (2015) Improved quality of life at end of life re-
    mementos, memorials, and altruistic acts such as organ donation and lated to home-​based palliative care in children with cancer. J Palliat
    charitable work. The other spiritual needs include the need for truth, Med 18(2), 143–​50.
    compassion, prayer, and ritual; connection with others; bereavement Goldman A (1998) ABC of palliative care: special problems of chil-
    support; meaning and purpose; anger, blame, and dignity. Professionals dren. BMJ 316, 49–​52.
    can support parents by ensuring that the child and family remain a unit Jalmsell L, et al. (2006) Symptoms affecting children with malignan-
    at the time of death, able to create memories and say their goodbyes. cies during the last month of life: a nationwide follow-​up. Pediatrics
    117(4), 1314–​20.
    Jennings A (2010) Palliation of breathlessness: a combined approach is
    needed. Eur J Palliat Care 17(4), 162–​6.
    Family support and bereavement care Kreicbergs UC, et al. (2007) Parental grief after losing a child to
    cancer: impact of professional and social support on long-​term out-
    The child and family face a series of losses even before the child ap- comes. J Clin Oncol 25(22), 3307–​12.
    proaches the terminal stages of the illness. These include the physical
    Family support and bereavement care 89

    Meert KL, et al. (2005) The spiritual needs of parents at the time of Ullrich CK, et al. (2010) Fatigue in children with cancer at the end of
    their child’s death in the pediatric intensive care unit and during life. J Pain Symptom Manage 40(4), 483–​94.
    bereavement: a qualitative study. Pediatr Crit Care 6(4), 420–​7. Wolfe J et al. (2000) Understanding of prognosis among parents of
    Ripamonti C, et al. (2000) Role of octreotide, scopolamine children who died of cancer: impact on treatment goals and inte-
    butylbromide, and hydration in symptom control of patients with gration of palliative care. JAMA 284(19), 2469–​75.
    inoperable bowel obstruction and nasogastric tubes: a prospective Wolfe J, et al. (2008) Easing of suffering in children with cancer at the
    randomized trial. J Pain Symptom Manage 19(1), 23–​34. end of life: is care changing? J Clin Oncol 26(10), 1717–​23.
    Rosenberg AR, et al. (2016) Quality of life in children with advanced Wolfe J, et al. (2015) Symptoms and distress in children with ad-
    cancer: a report from the PediQUEST Study. J Pain Symptom vanced cancer: prospective patient-​reported outcomes from the
    Manage 52(2), 243–​53. PediQUEST Study. J Clin Oncol 10(33:17), 1928–​35.
    Temel JS, et al. (2010) Early palliative care for patients with metastatic World Health Organization (2012) WHO Guidelines on the
    non-​small-​cell lung cancer. N Engl J Med 363, 733–​42. Pharmacological Treatment of Persisting Pain in Children with
    Thompson A, et al. (2006) Feeding in palliative care. In: A Goldman, Medical Illnesses. Geneva: World Health Organization.
    et al. (eds) Oxford Textbook of Palliative Care for Children, pp. 374–​ Yee JD, Berde CB (1994) Dextroamphetamine or methylphenidate as
    86. Oxford: Oxford University Press. adjuvants to opioid analgesia for adolescents with cancer. J Pain
    Twycross R, Wilcock A (2007) Anti-​emetics. In: R Twycross (ed.) Palliative Symptom Manage 9,122–​5.
    Care Formulary (3rd edn), pp. 175–​9. Nottingham: palliativedrugs.com.
    12
    Cancer Treatment in Low-​and
    Middle-​Income Countries
    Elizabeth Molyneux, Trijn Israels, and Scott C. Howard

    Introduction care are paid by a non-​profit foundation, Ayúdame a Vivir (Help


    Me to Live, www.ayuvi.org.gt). In many other countries, the greatest
    The World Bank divides economies (according to their gross na- barrier to cure is the lack of access to treatment, because families
    tional income per capita in US dollars) into low income ($1,025 or must cover the costs of diagnosis, staging, treatment, and follow-​up
    less), lower middle income ($1,026–​$4,035), upper middle income in addition to the non-​medical costs associated with missed work,
    ($4,036–​$12,475), and high income ($12,476 or more), calculated travel, and lodging.
    using the World Bank Atlas method (http://​data.worldbank.org/​ Even patients who attain access to diagnosis and treatment may
    about/​country-​classifications/​world-​bank-​atlas-​method). Cancer not survive, since socioeconomic and cultural factors in LMIC make
    treatment in low-​and middle-​income countries, referred to here- abandonment of treatment a common occurrence and death rates
    after as LMIC, differs substantially from that in high-​income coun- due to the toxicity of chemotherapy are frequently high. For example,
    tries (HIC), and these differences are the focus of this chapter. in the 1980s in Recife, Brazil, 8.4% of children with acute lympho-
    blastic leukaemia (ALL) died of toxicity and 16% abandoned treat-
    ment, versus 1.5% toxic death and no abandonment with the same
    Healthcare disparities between low-​and middle-​ chemotherapy regimen in Memphis, USA. Similarly, in Honduras
    income countries and high-​income countries a decade ago, 21% of children with ALL died of toxicity and 23%
    abandoned treatment.
    Cure rates for children with cancer differ greatly between LMIC and Government spending on healthcare per capita is the strongest
    HIC. Cancer cure depends on a series of steps which can be concep- predictor of childhood cancer survival, because it is associated with
    tualized as links in the chain of care, from suspicion by the patient healthcare infrastructure at all levels (Figure 12.2), and correlates
    and family to primary care, secondary care, tertiary care, and, finally, strongly with access to diagnosis, treatment, and event-​free survival
    correct diagnosis and treatment (Figure 12.1). The cure rate is the since well-​funded health systems are more likely to have tertiary care
    product of access to diagnosis, treatment, and event-​free survival, centres equipped with immunohistochemistry and flow cytometry,
    each of which can be problematic in LMIC. Late diagnosis and mis- access to necessary chemotherapy and supportive care, and to be
    diagnosis are common in healthcare systems with inadequate infra- staffed with trained pathologists, surgeons, oncologists, radiation
    structure for early referral, inadequate pathology laboratories that therapists, and nurses. Unfortunately, most paediatric oncologists
    often lack even basic tools like immunohistochemistry, and high have little influence over public health infrastructure, so the focus
    costs of cancer diagnosis and staging. These and other obstacles de- of their efforts must be to establish functional PCUs and to progres-
    crease access to diagnosis in many LMIC. For example, in Mali and sively develop them into centres of excellence, as discussed in the
    Malawi, the reported incidence of acute leukaemia is less than 10% section ‘Importance of paediatric cancer units’. Fortunately, there are
    of the incidence in HIC (Table 12.1). many proven strategies to overcome each cause of preventable treat-
    Poverty and inadequate public health infrastructure prevent or ment failure.
    delay access to even primary care for many of the world’s children,
    and many LMIC have few or no paediatric cancer units (PCUs),
    so that patients must sometimes travel long distances for cancer Importance of paediatric cancer units
    care. For example, only an estimated 55% of children with cancer
    in Guatemala are referred to the National Pediatric Oncology Unit A PCU is a centre at which children with cancer can receive services
    (the only PCU in the country). However, all patients who reach the necessary to make a correct diagnosis of cancer, perform a staging
    PCU in Guatemala City have access to treatment, since the costs of evaluation, and deliver protocol-​based therapy. Establishing a unit
    Importance of paediatric cancer units 91

    Denmark
    Finland
    Iceland
    Norway
    Sweden
    Iceland
    UK England and Wales
    UK Northern Ireland
    UK Scotland
    Austria
    Belgium
    France
    Germany
    Netherlands
    Switzerland
    Italy
    Malta
    Portugal
    Slovenia
    Spain
    Czech Republic
    Poland

    Europe
    0 10 20 20 40 50 60 70 80 90 100

    Figure 12.1 Links in the chain of care for children with cancer.
    SES, socioeconomic status
    Reproduced with permission from Howard SC et al. 'Improving outcomes for children with cancer in low-​income countries in Latin America: a report on the recent meetings
    of the Monza International School of Pediatric Hematology/​Oncology (MISPHO)–​Part I'. Pediatric Blood Cancer, Volume 48, Issue 3, pp. 364–​9, Copyright © 2007, John Wiley
    and Sons. DOI: https://​doi.org/​10.1002/​pbc.21003

    with its own physical space, dedicated personnel, and self-​contained salary in the PCU to develop professionally and feel they have a
    administrative structure allows the multidisciplinary team of doc- future career path.
    tors, nurses, pharmacists, social workers, psychologists, phys- Education and training of healthcare professionals at all levels is
    ical therapists, and data managers to work together to improve all paramount to developing and improving a PCU, and many profes-
    aspects of patient care. sional organizations have made educational materials and training
    scholarships available on their websites (e.g. International Society
    Recruitment, training, and retention of personnel of Pediatric Oncology www.siop.nl, American Society of Pediatric
    Delivery of anti-​cancer therapy requires trained personnel, but Hematology/​Oncology www.ASPHO.org, International Network
    healthcare providers with expertise in paediatric oncology in for Cancer Treatment and Research www.inctr.org, American
    LMIC are far fewer per capita than in HIC. In some cases, no Society of Hematology www.hematology.org). Online meetings can
    paediatric oncologist is available and medical oncologists with provide not only education but also a forum to discuss individual
    little paediatric experience or paediatricians with little oncology patients, adapted protocols, and quality-​improvement initiatives
    experience manage children with cancer. Oncology nurses in with local, regional, and global collaborators.
    LMIC are rare, and in many hospitals, nurses are required to rotate Until recently there were few training opportunities in oncology
    among all the hospital services, such that none gains expertise in for clinicians and nurses working in low-​income settings. Those
    oncology nursing and the effect of training programmes is limited. that were available often took the applicant away from home for
    Even if trained personnel are recruited, they must have a stable long periods of time to study and gain experience in settings very

    Table 12.1 Determinants of cure for children with cancer: estimated access to diagnosis, treatment, and event-​free survival by country-​
    income category

    Country or region Approximate rate of access Approximate rate of access Approximate event-​free Estimated cure rate
    to diagnosis to treatment survival of treated patients
    High-​income countries 100% 100% 83% 83%
    Upper middle-​income countries 95% 95% 75% 68%
    Lower middle-​income countries 80% 70% 50% 28%
    Low-​income countries 60% 40% 30% 7%

    Reproduced courtesy of Scott C. Howard, University of Tennessee College of Health Sciences, USA.
    92 CHAPTER 12 Treatment in Low- and Middle-Income Countries

    Lack of diagnosis

    treatment failure
    No diagnosis Causes Strategies
    Lack of awarencess
    Misdiagnosis Community

    Causes of
    by the patient or education and
    Refusal parents about awarences programs
    cancer symptoms
    Abandonment
    Healthcare Training of
    Toxic death providers unaware of community health
    early signs of cancer workers and primary
    care providers
    Relapse Lack of diagnostic
    capabilities Equipment and
    training for
    CURE Inability to afford a
    diagnostic
    pathology facilities
    Universal healthcare
    evaluation
    coverage

    Incorrect diagnosis Refusal and abandonment


    Causes Strategies Causes Strategies
    Logistical and financial Subsidized transportation
    Insufficient Public-private partnerships
    barriers (travel time or to and from clinic
    pathology for training, funding, and
    distance, costs of care) appoinments, local
    infrastructure sustainability housing for out-of-town
    Lack of Miscommunication by patients, subsidized food
    immunohisto- Implementation of
    specialized testing for the healthcare providers Pratice messaging,
    chemistry, flow
    cytometry, and other most common cancers Missed appoinments consistent messages
    specialized testing Pathology training due to non-logistical Patient tracking system
    Insufficient numbers programs, telepathology factors appoinment scheduling
    of trained for remove support Health beliefs and notification system
    pathologists Technology innovation, (curability of cancer, Patient education
    Lack of funding for pooled procurement of need for post- Family education
    specialized testing regentes and services remission theraphy, Peer/parent support
    toxicities of therapy) groups
    Toxic death Avoidable relapse
    Causes Strategies Causes Strategies
    Delayed treatment Golden Hour quality Protocols from high- Depoly locally adapted
    of febrile improvement income countries protocols
    neutropenia programme not suitable for local Personal training,
    conditions monitoring drugs
    Less intense
    chemotherapy, Lack of knowledge prescribed and
    Multi-drug
    probiotics about use of adapted administered
    resistant bacteria
    protocols Prevent, detect, and
    Limit use of mitgate drug shortages
    antibiotics, fungal Gaps in access to by forecasting needs,
    Fungal infections prophylaxis, essential establishing inventory,
    probiotics chemotheraphy and measuring patient
    access and adherence
    Haemorrhage Blood bank quality Gaps in adherence
    improvement Information
    to chemotheraphy
    system/patient tracking

    Figure 12.2 Strategies to reduce each cause of preventable treatment failure for children with cancer in low-​and middle-​income countries.
    Reproduced courtesy of Scott C. Howard, University of Tennessee College of Health Sciences, USA.

    different from his/​her own. On returning, it was difficult to align Continuous quality improvement
    what had been taught with what was feasible and locally appro- The process of continuous quality improvement requires a data man-
    priate. Now there are several opportunities to train appropriately on agement programme, which consists of a data manager, a database,
    short courses, and on Master’s, Bachelor’s, and Certificate courses and data analysis. Training for data managers can be conducted
    that are designed for busy people working in less resourced parts online, and participation in patient discussion meetings with the
    of the world. Some courses are taught by distance learning (on- supervising doctor can improve collegiality, data quality, and the
    line), others by rotations in countries with well-​developed educa- utility of data collection for improving patient care. Many free soft-
    tional programmes, and others are a combination of the two, with ware tools are available, including spreadsheets (www.OpenOffice.
    short periods of training away from home and work. Many are org, Google documents) and EpiInfo (www.cdc.gov/​epiinfo/​),
    open to nurses and clinicians, while others are specifically for one which allow the user to quickly configure databases with any type
    healthcare cadre. of data for any purpose. EpiInfo includes security features, double-​
    Table 12.2 highlights some useful websites to help find appro- data entry to improve accuracy, and analytical tools. Both are easy
    priate training. For (paediatric) radiotherapy, the International to use and ideal for single-​centre research. For centres with reliable
    Atomic Energy Agency (IAEA) supports LMICs in the development internet access, that work within multi-​centre groups, the Resonance
    of infrastructure and training of human resources needed to provide Patient Center (http://​rpcprod.manuhsolutions.com:9090/​RPC/
    high-​quality radiation treatment. or www.ResonanceOncology.org​) is available at no cost. It includes
    Protocol-based care 93

    Table 12.2 Training opportunities for paediatric oncology professionals from low-​and middle-​income countries

    Programme Description
    Cure4Kids (www.cure4kids.org) Excellent online resource for healthcare professionals worldwide, which includes disease-​and treatment-​
    specific seminars, articles, image-​based instruction, and online education and collaboration tools.
    Funded by St Jude Children’s Research Hospital
    CancerPointe (www.cancerpointe.com) POINTE has compiled information about clinical and research training courses in paediatric oncology
    for healthcare workers from resource-​limited locations. This excellent website was developed and is
    maintained by the Pediatric Oncology in Developing Countries (PODC) group of the International
    Society of Pediatric Oncology (SIOP).
    Groupe Franco–​Africain d'Oncologie Pédiatrique GFAOP offers a one-​year intensive course for nurse educators, a one-​year diploma in paediatric
    (GFAOP; French–​African Paediatric Oncology Group) oncology for paediatricians, and an intensive course in paediatric surgical oncology. There are many
    (www.gfaop.org) Francophone courses developed through the GFAOP. These are listed on the POINTE website.
    The African Pediatric Fellowship Program The Program offers training opportunities to African paediatricians seeking subspecialty training. The Red
    Cross Children’s Hospital in Cape Town offers this programme through recognized institutions in several
    African countries.
    Uganda training programmes (www.fredhutch.org/​en/​ The Uganda Cancer Institute in Kampala and University of Washington offers a wide range of training
    labs/​vaccine and www.oncology/​uganda/​training.html) opportunities including PhDs and Masters degrees to Ugandan scientists and doctors. Funded by the
    Fred Hutchinson Cancer Research Center.
    Hospice training (www.hospice-​africa.merseyside.org/​ Training in palliative care is offered by Hospice Africa International Programmes which are based in
    Internat.html) Kampala and run distance learning courses open to all.
    International Paediatric Oncology Society (SIOP) Asia The website provides information about training opportunities in Asia including www.siop.asia.com/​
    Continental Branch (www.siop.asia.com) index.php/​information/​fellowship.html, which describes the SIOP international outreach fellowships
    scheme.
    CaseHippo (www.CaseHippo.com) Paediatric oncology content, connection to paediatric professional societies, and networking capabilities.

    customizable data-​entry forms and reports, anonymized data sharing


    Protocol-​based care
    among sites that choose to do so, high-​level security, automated off-​
    site back-​ups, and an integrated clinical trials management system
    Protocol-​based care standardizes treatment regimens and improves
    for sites that wish to conduct clinical research. It also has a large base
    results for every cancer in which its effect has been measured.
    of existing users with whom new users can collaborate.
    However, clinicians who apply protocols in LMIC face uncertain-
    Data analysis to improve patient care is the most important com-
    ties about disease stage and risk classification, medication shortages,
    ponent of any data-​management programme, the utility of which
    excess toxicity from comorbid illnesses, and preventable deaths due
    depends far more on the quality of the data collected, regular and
    to poor transportation or other social factors. Furthermore, doc-
    critical analysis of results, and the quality-​improvement strategies
    tors in LMIC manage large numbers of patients with little time,
    implemented than on the software tool selected. The top priority
    rarely have the luxury of specializing in a subset of childhood can-
    in LMIC (after prompt and correct diagnosis) is to prevent toxic
    cers, and often lack access to experts. Information systems provided
    deaths, reduce abandonment, and monitor compliance with therapy
    at no cost to centres in LMIC, like Resonance Oncology (www.
    and, thus, reduce relapse. This clinical research is best undertaken
    ResonanceOncology.org), facilitate adherence to the protocol by
    with the support and encouragement of regional colleagues in the
    context of regional and global networks, as discussed in the next
    section.
    Development of individual cancer units (depth)
    Development of paediatric cancer units and national
    paediatric oncology programmes
    Pilot Paediatric Centre of Super-specialty
    Some PCUs began as pilot projects, the goal of which was to simply
    unit cancer unit excellence centre
    demonstrate that some children with cancer can be cured with
    simple therapy that can be delivered locally, like the simplified Development of national coverage (breadth)
    Burkitt lymphoma treatment used in Malawi. A successful pilot
    project can lead to development of a PCU, which can be further de-
    veloped into a centre of excellence, eventually with its own satellite
    centres for shared care, and, finally, into a regional or national net-
    work of PCUs (Figure 12.3). For example, the Pediatric Oncology
    Unit of Guatemala began as a pilot unit, developed into a centre
    of excellence, and recently launched its first satellite in an effort to Pilot units and Paediatric Centres of Collaborative
    eventually provide national coverage and to improve event-​free satellite centres cancer units excellence network
    survival by promoting early diagnosis and referral for locally-​ Figure 12.3 Phases of development of paediatric oncology and the
    adapted protocol-​based care. Implementing such care is discussed central role of paediatric cancer units.
    in the next section. Reproduced courtesy of Scott C. Howard.
    94 CHAPTER 12 Treatment in Low- and Middle-Income Countries

    providing patient-​specific chemotherapy roadmaps, patient treat- and global networks of collaborators have kept the momentum by
    ment diaries, and automated calculation of drug doses based on the sharing protocols, projects, funding sources, and supportive-​care
    patient’s height and weight. Such systems can also reduce abandon- guidelines. Online meetings via www.Cure4Kids.org have trans-
    ment by including a scheduling system with call-​back functionality formed twinning programmes and regional collaborations. For ex-
    for patients who miss appointments and are at risk to abandon. ample, since 2005 professionals from centres in Morocco, St Jude
    Finally, information systems form the backbone of any quality-​ Children’s Research Hospital, and guests from several continents
    improvement effort designed to address specific causes of prevent- have been meeting weekly to discuss patients, review supportive care
    able treatment failure (Figure 12.2). issues, and develop and implement treatment protocols adapted to
    local conditions. An Asia Pacific Leukemia Lymphoma meeting held
    Adapted-​treatment regimens and clinical research every other Friday, via Cure4Kids, currently draws healthcare pro-
    Protocols that work well in HIC may cause disastrous toxicity in viders from 5–​10 countries to present and discuss patients, protocols,
    LMIC, as occurred when a regimen developed at St Jude Children’s and supportive-​care issues. The Global Neuroblastoma Network
    Research Hospital in Memphis, USA, was adopted in El Salvador (www.globalneuroblastoma.net) meets weekly on Thursdays to im-
    without modification. Regimens must be adapted for use in LMIC, prove care for this disease.
    and the rate of toxic death monitored to determine whether the
    supportive care in the hospital is adequate to support the proposed
    treatment regimens and to identify preventable causes of treat- Treatment challenges in low-​and
    ment failure. If toxic death occurs too frequently, the regimen must
    middle-​income countries
    be further adapted pending improvements in supportive care and
    hospital infrastructure. Practical research to address causes of treat-
    Toxicity of treatment
    ment failure in the local setting is best conducted with colleagues
    who practice in similar circumstances. For example, centres in Chemotherapy is especially toxic in malnourished children with ad-
    Central America (AHOPCA) and Latin America (MISPHO) have vanced disease. In HIC, more intense and toxic therapies, matched
    shared protocols and compared results for the past two decades, by intense and costly supportive therapy, have achieved good out-
    with great benefits to all. Collaborating centres in French-​speaking comes. In LMIC, children present with advanced cancers and large
    Africa, India, Malaysia, and elsewhere have also developed adapted-​ tumour loads. For example, in children with Burkitt lymphoma in
    treatment regimens and locally feasible supportive-​care guidelines. Africa, the tumour burden is high, with frequent kidney involve-
    In addition to local adaptation and implementation of protocol-​ ment and high risk of tumour lysis syndrome even with low-​intensity
    based care, adapted-​ treatment regimens have been developed for chemotherapeutic regimens. The risk to the patient is increased even
    Burkitt lymphoma, Wilms tumour, Kaposi sarcoma, low-​grade glioma, more by lack of rasburicase and inadequate access to laboratory
    Hodgkin lymphoma, medulloblastoma, and acute lymphoblastic leu- testing to follow potassium levels and prevent cardiac dysrhythmias.
    kaemia (ALL). In the case of ALL, the regimens establish a backbone Malnutrition and underlying infections must also be managed prior
    of therapy with low toxicity and cost that will cure a large proportion of to starting chemotherapy (to the extent possible) to minimize treat-
    children with ALL while minimizing toxic death. If there are less than ment toxicity. HIV-​infected children can be expected to have more
    two toxic deaths among the first 25 patients, the PCU is ready to ‘step up’ episodes of fever, neutropenia, and treatment delays.
    to a more intense treatment regimen to try to reduce the rate of relapse. Malnutrition has been shown to be associated with more
    If two or more patients die of toxicity, the PCU can ‘step down’ to a safer, chemotherapy-​induced neutropenia and fever episodes with conse-
    less intense regimen and work to improve supportive care. Rational quent delays in therapy, but few PCUs in LMIC can provide paren-
    movement between steps requires monitoring of patient outcomes in teral feeding, which is unacceptable to many parents who equate it
    real time, so a data-​management programme is a prerequisite to suc- with death. Nausea and vomiting, and especially mucositis, become
    cessful implementation of an adapted-​treatment strategy. The adapted difficult to manage when oral feeding is required. Pre-​treatment
    treatment for Wilms tumour is currently being field-​tested in many fluids (oral or intravenous) and allopurinol (in children with haem-
    centres in Africa with prospective data collection and practical support atological cancers) should be given, as well as an antiemetic such
    including training for pathologists, sonographers, and surgeons. as metoclopromide, which is inexpensive and widely available.
    Enriched milk (F100) or ready-​to-​use therapeutic foods such as
    Regional and global networks Plumpynut or Chiponde can rapidly improve nutritional status.
    Twinning partnerships between centres in HIC and LMIC can Nursing staff should be empowered to act quickly and independ-
    greatly speed the development of PCUs and centres of excellence, but ently to initiate a fever protocol and start antibiotics as soon as fever
    such partnerships are even more effective when associated with other is recorded and malaria (in malaria-​endemic areas) is ruled out.
    PCUs in the same region. Interaction with colleagues who practice Treatment should be modified to suit local facilities and expertise, to
    in a similar setting and face similar problems affords an opportunity use doses and drugs that can be safely given to achieve the best out-
    to learn from each other’s mistakes, share guidelines, and optimize come possible. As experience is gained with protocols and as local
    adapted-​treatment regimens for each centre. In this regard, cooper- health provision improves, the protocols can be intensified in a step-​
    ating centres in Argentina, Brazil, India, Colombia, and many other wise fashion.
    LMIC have developed shared protocols and joint advocacy pro-
    Abandonment of treatment
    grammes to improve outcomes at the national and regional level.
    Improving outcomes for children with cancer by developing and When treatment is initiated but not completed, it is described as
    improving PCUs has proven effective all over the world, and regional abandonment. In children with cancer, abandonment significantly
    Advocacy 95

    decreases the probability of cure. The reasons for abandonment of gradually develop their own support networks. The International
    treatment are various and individual, but there is an overarching Society of Pediatric Oncology (SIOP), with its regional organiza-
    pattern of causes reported from LMIC. The most commonly cited tions in Africa, Asia, Oceania, and Latin America, provides pro-
    reason for abandonment is financial. Families from LMIC cannot fessional support and expertise. The Children’s Oncology Group
    afford transportation to a distant care centre. The long stays in the (COG, www.childrensoncologygroup.org) collaborates with South
    hospital drain family funds and keep them away from any gainful American centres for training and shared protocols, and the Groupe
    employment they had previously. In Recife, Brazil, a social support Franco-​ Africaine d’Oncologie Pediatrique (GFAOP, www.igr.fr/​
    programme for children with ALL and their families included trans- gfaop) has a long tradition of international collaboration between
    portation, accommodation, financial support, and job training. This PCUs in France and those in Africa.
    reduced the number who abandoned treatment from 16% to 0.5%. The World Health Organization (WHO) has increasingly become
    Patient and family education are essential, in addition to logistical involved in improving paediatric cancer care, and has joined forces
    and financial support. Keeping an open ward or providing accom- with SIOP, Childhood Cancer International, St Jude Children’s
    modation for those who live far away and cannot easily get to and Research Hospital, World Child Cancer, and others to strengthen
    from home between courses of chemotherapy is necessary. local health systems and support cancer planning and strategy de-
    Assisting on the ward with extra food and basic necessities such velopment in selected LMIC. It has endorsed an essential drug list
    as soap and laundry detergent also helps prevent abandonment. The for paediatric cancer treatment, which should be made available to
    hospital is a bewildering place to people new to the city, so a system all children in all countries (http://​www.who.int/​selection_​medi-
    to welcome newcomers and provide amenities such as washing lines cines/​committees/​expert/​17/​second_​children_​list_​en.pdf). Parent
    and cooking facilities is essential. In Blantyre, Malawi, the mothers support organizations can provide a strong voice to influence polit-
    are given a small amount of money to spend in the market every ical decisions and raise the profile of cancer in LMIC. Wherever we
    weekend on simple foodstuffs and firewood so that they can cook work and however meagre our tools, there is always something that
    in a shelter close to the ward. This programme improves nutritional we can do to improve the lives of children with cancer.
    status, and also bonds the women together in a shared and enjoy-
    able task. Other reasons for abandonment of therapy include an
    extended family’s reluctance or inability to help with home chores FURTHER READING
    while the mother is absent in the hospital with one child; death or Arora RS, et al. (2007) The problem of treatment abandonment in chil-
    illness in other close relatives; anxiety about leaving the remaining dren from developing countries with cancer. Pediatr Blood Cancer
    children in inadequate care at home; and concerns about the fidelity 49, 941–​6.
    of a spouse during long absences at the hospital. Arya LS, et al. (2010) Pattern of relapse in childhood ALL: challenges
    and lessons from a uniform treatment protocol. J Pediatr Hematol
    Palliative care Oncol 32, 370–​5.
    Cavalli F (2008) The World Cancer Declaration: a roadmap for change.
    Sometimes pragmatic decisions must be made. Surgical options may
    Lancet Oncol 9, 810–​11.
    be unavailable, and if chemotherapy does not improve the patient’s Gupta S, et al. (2009) Incidence and predictors of treatment-​related
    disease, palliative care may be the only way to help a child. Palliative mortality in paediatric acute leukaemia in El Salvador. Br J Cancer
    care offers symptom control and emotional and spiritual support. 100, 1026–​31.
    One of the most significant recent advances in LMIC is the recog- Harif M, et al. (2008) Treatment of B-​cell lymphoma with LMB modi-
    nition that pain control is a fundamental human right and that oral fied protocols in Africa—​report of the French-​African Pediatric
    morphine must be made available for children with cancer who Oncology Group (GFAOP). Pediatr Blood Cancer 50, 1138–​42.
    experience pain. Hesseling P, et al. (2005) The 2000 Burkitt lymphoma trial in Malawi.
    Pediatr Blood Cancer 44, 245–​50.
    Hesseling P, et al. (2009) Endemic Burkitt lymphoma: a 28-​day treat-
    Advocacy ment schedule with cyclophosphamide and intrathecal metho-
    trexate. Ann Trop Paediatr 29, 29–​34.
    Given the rarity of cancer in children compared to infectious dis- Howard SC, et al. (2004) Establishment of a pediatric oncology pro-
    gram and outcomes of childhood acute lymphoblastic leukemia in
    eases and other health problems, it is not surprising that cancer is
    a resource-​poor area. JAMA 291, 2471–​5.
    not a priority for health systems in LMIC. Therefore, other ways
    Howard SC, et al. (2007) Improving outcomes for children with
    must be found to support cancer care. Twinning between PCUs in cancer in low-​income countries in Latin America: a report on the
    HIC and LMIC has been very successful. The gain is bilateral. Funds, recent meetings of the Monza International School of Pediatric
    expertise, drugs, and specialized diagnostic tests flow from high-​ Hematology/​Oncology (MISPHO)-​Part I. Pediatr Blood Cancer
    income to low-​income centres, while experience, thinking outside 48, 364–​9.
    the box, valuing what is possible, and using what is available are les- Howard SC, et al. (2008) Childhood cancer epidemiology in low-​
    sons learned by those who work in HIC centres. income countries. Cancer 112, 461–​72.
    Charitable organizations and professional organizations can Hunger SP, et al. (2009) Treatment strategies and regimens of gradu-
    support PCUs in LMIC with expert advice, assistance with modi- ated intensity for childhood acute lymphoblastic leukemia in low-​
    fying protocols, and moral support. World Child Cancer (www. income countries: a proposal. Pediatr Blood Cancer 52, 559–​65.
    worldchildcancer.org and www.worldchildcancer.us) was formed Israels T, et al. (2008) The guardians’ perspective on paediatric cancer
    to channel support to cancer units in low-​income countries. It treatment in Malawi and factors affecting adherence. Pediatr Blood
    Cancer 51, 639–​42.
    offers long-​term support to PCUs to help them get established and
    96 CHAPTER 12 Treatment in Low- and Middle-Income Countries

    Israels T, et al. (2009) Malnutrition and neutropenia in children treated Mostert S, et al. (2010) Comparing childhood leukaemia treatment be-
    for Burkitt lymphoma in Malawi. Pediatr Blood Cancer 53, 47–​52. fore and after the introduction of a parental education programme
    Israels T, et al. (2010) Strategies to improve care for children with in Indonesia. Arch Dis Child 95, 20–​5.
    cancer in Sub-​Saharan Africa. Eur J Cancer 46, 1960–​6. Norman K, et al. (2008) Prognostic impact of disease-​related malnutri-
    Leander C, et al. (2007) Impact of an education program on late diag- tion. Clin Nutr 27, 5–​15.
    nosis of retinoblastoma in Honduras. Pediatr Blood Cancer 49, 817–​9. Ribeiro RC, et al. (2008) Baseline status of paediatric oncology care
    Magrath I, et al. (2005) Treatment of acute lymphoblastic leukaemia in in ten low-​income or mid-​income countries receiving My Child
    countries with limited resources; lessons from use of a single protocol Matters support: a descriptive study. Lancet Oncol 9, 721–​9.
    in India over a twenty year period. Eur J Cancer 41, 1570–​83. Sinfield RL, et al. (2007) Spectrum and presentation of pediatric ma-
    Masera G (2009) Bridging the childhood cancer mortality gap be- lignancies in the HIV era: experience from Blantyre, Malawi, 1998–​
    tween economically developed and low-​income countries. J Pediatr 2003. Pediatr Blood Cancer 48, 515–​20.
    Hematol Oncol 31, 710–​12. Sitaresmi MN, et al. (2010) Treatment refusal and abandonment in
    Masera G, et al. (1998) North-​South twinning in paediatric haemato-​ childhood acute lymphoblastic leukemia in Indonesia: an analysis
    oncology: the La Mascota programme, Nicaragua. Lancet 352, of causes and consequences. Psycho-​oncology 19, 361–​7.
    1923–​6. Wilmshurst JM, et al. (2016) The African Pediatric Fellowship
    Metzger ML, et al. (2003) Outcome of childhood acute lymphoblastic Program: Training in Africa for Africans. Pediatrics 137(1).
    leukaemia in resource-​poor countries. Lancet 362, 706–​8. doi: 10.1542/​peds.2015-​2741.
    13
    Cancer in Adolescents and Young Adults
    Lucy J.W. Jones, Emmanouil Saloustros, Andrea Ferrari, and Dan Stark

    Epidemiology and outcomes environmental factors appear more significant; increased age in-
    creases individual risk of exposure to pathogens, such as ultraviolet
    sunlight, smoking, viruses (e.g. HPV, HIV, EBV), and iatrogenic
    Incidence causes, such as the treatment of childhood cancers (see the section
    Cancer in the adolescent and young adult (AYA) population is un- ‘Late effects’). However, the aetiology for the majority of cancers in
    common: 2–​4% of all invasive malignancies overall depending upon this age group remains unknown, and unrelated to environmental
    precise definitions. However, the annual incidence of cancer in the exposures or inherited genetic disorders.
    15-​to 39-​year age group is estimated to be 50–​70,000 in Europe, and
    Outcome
    one million worldwide. The incidence of cancer increases with age,
    from childhood to adulthood; between the ages of 15 and 30, cancer Cancer is the most common cause of disease-​related death in the
    is 2.7 times more frequent than in the under 15 age group. 15-​to 24-​year age group in high-​income countries, accounting for
    The overall incidence of cancer in the AYA population is 10% of deaths in AYAs annually in Europe. Survival varies with the
    increasing. Testicular cancer has seen substantial increases in the specific cancer diagnosed. Malignancies with the highest survival
    recent era, while in the USA, increasing diagnosis of early thyroid rates include thyroid carcinoma, Hodgkin lymphoma, germ-​cell
    cancers may be partly responsible. The incidence of some cancers tumours, malignant melanoma, chondrosarcomas, and fibrosar-
    seen in AYAs is decreasing in the Western world (such as Kaposi’s comas. Conversely, poorer outcomes are associated with liver and
    sarcoma, melanoma, lung carcinoma, cervical carcinoma) as a result lung carcinomas, rhabdomyosarcomas, astrocytomas, Ewing’s sar-
    of public health measures directed at young people. coma, and acute myeloid leukaemia (AML), whose relative five-​year
    The spectrum of cancer types diagnosed in AYAs aged 15–​39 survival is less than 55%.
    is distinct from children and older adults. The most frequently In high-​income countries, the overall five-​year survival rate of all
    occurring cancers in the AYA population are: cancers in the AYA population is now greater than 80%. Survival
    outcomes in AYA malignancies are improving after a prolonged
    • carcinomas (in particular breast, melanoma, gynaecological, and period of stagnation or deterioration going back to the 1990s when
    thyroid carcinomas, but also colorectal tumours), there was little if any focus upon the field. A recent study of five-​
    • haematological malignancies, notably Hodgkin and non-​Hodgkin year survival trends in the UK demonstrated an overall increase in
    lymphoma (HL and NHL), five-​year survival from 75.5% in 1992–​1996, to 82.2% in 2002–​2006.
    • germ-​cell tumours. An improvement was observed in all cancer types, except for osteo-
    In contrast to children under 12, embryonal tumours are much rarer sarcoma, rhabdomyosarcoma, non-​gonadal and ovarian germ-​cell
    in the AYA population. Within AYAs, the incidence of certain can- tumours, and thyroid carcinomas; comparable with that observed
    cers varies with age, and some cancer types, such as osteosarcoma, in other countries. However, despite this improvement in survival
    peak in incidence. The incidence of carcinomas gradually increases outcomes, in comparison to younger age groups, AYAs still have
    with age, while HL and germ-​cell tumours peak at ages 15–​24 years worse outcomes. In Europe, between 2000 and 2007, patients under
    and 20–​29 years respectively. The incidence of acute lymphoblastic 15 years had a better five-​year relative survival compared to AYAs for
    leukaemia (ALL) declines with increasing age. ALL (85.8% versus 55.7%), AML (60.5% versus 49.7%), NHL (83%
    The causes of cancer in AYAs are poorly understood, and appear versus 77.4%), Ewing’s tumours (66.6% versus 49.3%), rhabdomyo-
    to differ from adults and children. Fewer than 5% of all cancers in the sarcoma (66.6% versus 37.7%), and astrocytomas (61.9% versus
    AYA population are attributed to currently known inherited cancer 46.5%). Some differences can be understood to some extent by
    syndromes, although AYA-​ specific studies of this are planned, examining the detail; astrocytoma is associated with a higher pro-
    such as SPECTA-​AYA across Europe and INFORM specifically in portion of pilocytic pathology in children (25%) and adolescents
    Germany. While prenatal factors may have some effect on the oc- (35%) than at 20–​24 years (17%), 25–​29 years (8%), or 30–​39 years
    currence of AYA malignancies, in contrast to childhood cancer, (4%). The reason for each of these poorer outcomes is not clearly
    98 CHAPTER 13 Cancer in Adolescents and Young Adults

    elucidated but is undoubtedly multifactorial and should be subject diagnosis reported a significant increase in GP attendance across
    to more research. all cancers in the interval proceeding a cancer diagnosis. This time
    The five-​year relative survival in all cancers is 78% in men, com- period varied according to malignancy: 17 months for central ner-
    pared to 83% in women. There is some data suggesting this may vous system (CNS) tumours, 12 months for sarcoma, nine months
    relate in part to clinical management choices and drug pharma- for lymphomas, five to six months for leukaemia, bone tumour,
    cology during chemotherapy treatment administration in AYA can- or germ-​cell tumour, and three months for malignant melanoma.
    cers. There are also differences in cancer biology (see the section Researchers also observed an increase in the number of blood tests
    ‘Biology’). performed by the GP, which began on average 11 months before a
    These disparities highlight the need to better understand and char- cancer diagnosis.
    acterize what it is that makes AYA cancer management distinct from It has been hypothesized that young people’s cancers result in
    other age groups. Throughout this chapter we will explore some of them reporting non-​specific symptoms. However, pain and lumps
    these differences in terms of characteristic biology, patient factors, are the most prevalent in population-​based studies of presenting
    and the methods of treatment delivery within current healthcare in- symptoms in primary care. If reassured initially, some believe AYAs
    stitutions, and how we can work towards improving outcomes. may be more loathe to return for a further review than older people
    or the parents of a symptomatic small child. The challenge is to dif-
    ferentiate patients with malignancy from the majority with a benign
    Biology illness. The comparatively rare occurrence of AYA malignancies also
    may result in a high threshold (and therefore low likelihood) for in-
    There is growing evidence that the biology of AYA cancers is distinct vestigation for cancer in this age group. This highlights the need to
    from children and older adults, enabling physicians to target therapy improve awareness and understanding of cancer in the AYA popu-
    more specifically. Examples to highlight the clinical importance of lation among both the general public and healthcare professionals,
    this include: as well as establishing clear diagnostic pathways. A number of coun-
    tries, including the UK and Denmark have implemented a fast-​track
    • ALL: the TEL-​AML1 translocation is favourable for response to
    service for suspected cases of cancer, in order to facilitate prompt
    therapy, and occurs in only 10% of AYA cases, compared to 50% of
    diagnosis and treatment.1 These pathways may require particular at-
    childhood cases. In contrast, the poor prognostic indicator, Ph-​1,
    tention in making them optimal for AYA.
    peaks in incidence in the AYA age group.
    • Malignant melanoma: BRAF-​positive disease is much commoner
    in AYAs than in older adults. This increases treatment options. Age-​specific psychosocial aspects
    • Breast cancer: more younger women have tumours that are ‘triple
    negative’ for hormonal and epithelial growth factor receptors While the biology and range of malignancies in the AYA popu-
    compared to older women, therefore making them resistant to lation is distinct from that of children and older adults, there are
    many targeted systemic therapies. many other age-​specific issues that can form barriers for AYAs.
    • Rhabdomyosarcoma (RMS): outcomes are also significantly worse The large, distinct, and very specific impact of cancer at this age
    in the AYA population, with more young people found to have the can affect patients’ experience, quality of life, adherence to treat-
    alveolar subtype, which is associated with a worse survival out- ment, and survival outcomes. Many of the developmental goals
    come. Additionally, the incidence of MYOD1, a recently identified of life in this phase are not just impaired but directly placed into
    mutation associated with an aggressive and more advanced form reverse by cancer: independence, financially and personally, al-
    of embryonal RMS, peaks at age 25. ready being delayed in many modern Western societies, can go
    into ‘reverse gear’.
    AYAs are in the process of developing their own thoughts and
    Access to care and delay in diagnosis perceptions of the world, forming new relationships, and pushing
    boundaries. This stage of life is often formative in terms of educa-
    A credible causal candidate for poorer outcomes in AYA cancer pa- tion, career planning, and for some, family planning. There is often
    tients is that AYA patients frequently report a prolonged period with a desire to challenge the rules set by their parents, and experiment
    symptoms (before presenting for care) and then after presenting (be- with risk-​taking behaviours such as smoking, alcohol consumption,
    fore receiving a cancer diagnosis and treatment). Such prolonged or recreational drug use. With cancer, there are commonly periods
    time to diagnosis may impact the chance of cure. However, the im- of returning to reliance on parents, and moving away from inde-
    pact of this, clinically, may not be uniform between cancer types, pendent thinking and decision making.
    appearing to be particularly important in sarcomas and germ-​cell During adolescence, many young people also begin to explore
    tumours. sexual activity. Sexual identity development may regress, due to
    In many countries, the general practitioner (GP) is the first physical and psychological challenges, and issues relating to body
    healthcare professional to see these patients, and therefore plays a image, psychosexual identity, and fertility (see the section ‘Late ef-
    key role in the diagnostic pathway. It has been suggested that AYAs fects’). Young people have to find their own ‘new normal’.2 These
    may be ‘embarrassed’ to discuss their bodily symptoms, and there- issues, therefore, must be addressed during their treatment by
    fore might avoid seeing their GP in the first place. However, this is health professionals skilled at delivering clear, sensitive, and age-​
    not the case—​AYAs do consult. A retrospective study investigating appropriate information, which is not what AYAs frequently report
    the use of primary care by AYA patients preceding their cancer they experienced. Preserving fertility or safe menstrual periods with
    Models of care 99

    low platelets may require medically-​induced amenorrhea. This is communication channels to their peers and family may not be
    also a physiological regression towards childhood. used with authenticity and at the ideal frequency. The adult MDT
    Throughout their diagnosis and treatment, AYA patients should is focused upon clinical diagnosis and decision making, and there
    be empowered to live as normal a life as possible, and be able to is less access to non-​medical specialists such as psychological,
    achieve as many developmental milestones as possible, despite their social, and educational support.
    disease. Even when life is short, young people often benefit from • An AYA model of care must be patient-​centred, and support fra-
    a project or some personal development. While this may require gile patient autonomy and informed decision making. There is
    treating teams to demonstrate focus, creativity, and flexibility, the also a need to adapt the involvement of parents and other sig-
    frequent result is the development of a trusting rapport with the pa- nificant parties, depending on the maturity and life stage of the
    tient, better treatment adherence, and the best chance of the best patient.
    outcome.
    A broad MDT is required to meet the holistic needs of these pa-
    Maintaining normality includes the importance of peer support,
    tients. Experience of working as a team with AYAs enables, for ex-
    which is easily lost during long periods of treatment and hospital-
    ample, any challenges of concordance to be managed proactively,
    ization. This can be facilitated by a welcoming, age-​appropriate en-
    using flexibility but also clear, consistent information. Then, ‘battle-
    vironment, with flexible visiting times, and opportunities for young
    grounds’ between young people and their carers (such as over nutri-
    people to connect with both healthy peers and fellow AYA patients.
    tion or the balance between social and clinical priorities) should be
    Social isolation is a problem, but work on quality of life specifically
    expected, identified, and minimized.
    for AYAs also indicates that simple boredom is a large but remedi-
    AYAs have been described as existing in a ‘no man’s land’ be-
    able problem during their cancer treatment. Providing support is a
    tween children and adult services. A key aspect of AYA care is the
    key component of any AYA cancer service, from the point of diag-
    integration of paediatric and adult oncologists and haematologists.
    nosis to beyond the completion of treatment, and not only in re-
    Fundamental to providing quality, personalized care to AYA cancer
    lation to care for individual patients but also as one considers the
    patients is the bringing together of paediatric and adults services.
    design of AYA-​specific services and models of care.
    However, clinically, departments operate in dissimilar ways; for ex-
    Therefore, healthcare professionals caring for these patients need
    ample, they apply different staging and grading of diseases. The chal-
    to have not only expert clinical knowledge about AYA malignancies
    lenge of AYA healthcare providers is to act as a bridge between these
    but also of the broad and varied psychosocial needs and the fight by
    interfaces.
    some young people against a return to childhood. This knowledge
    Table 13.1 describes the models of care currently adopted by
    is acquired through regular experience of ‘walking with patients’
    various counties, and the projects that are currently in place. At pre-
    throughout their cancer journey, and also through national and
    sent, no ‘one-​size fits all’ model of care for AYA oncology services has
    international forums where experiences and ideas can be shared,
    been agreed, although there are detailed descriptions of several ef-
    and current practices reviewed and developed.3
    fective models. The collaborative work of national and international
    professional, government, and charitable organizations over the last
    Models of care ten years has seen much development in the design and implemen-
    tation of specialist oncology services for AYAs. The variation may
    be a good thing, enabling the services to meet the needs of young
    The clinical interface
    people in their local healthcare and societal context.
    Historically, and still in many parts of the world, AYAs were ei-
    ther treated by paediatric oncologists or adult specialists, neither Age range
    of which had specific expertise in their age group. This resulted in There is much variation between countries as to the age at which
    young people feeling isolated from their peers, at a time of life when AYAs reach this interface between children and adult care. In some
    friendship with those of a similar age is particularly important. Both southern European countries, this starts at age 14, while in other
    the model of care tailored to children’s services and the model of care countries, such as Germany, it does not begin until age 18. In the
    adopted in adult services fail to cater to the complex specific needs UK, AYA patients are teenagers (13–​18 years) and young adults (19–​
    of the AYA group. This was also the period when clinical outcomes 24 years). Different age ranges are also applied depending on the
    stagnated. purpose. In general, the upper age limit for clinical facilities is in the
    • Paediatric services are often based on a ‘family-​centred model’: in- mid-​twenties. This extends to age 29 or 30 for epidemiology, and 39
    formation shared with the patients is limited to what is considered or 40 years for clinical follow-​up. European Network for Teenagers
    by parents to be essential, and much of the decision making is and Young Adults with Cancer (ENTYAC) guidance for AYA care
    done by their parents/​carers. Both medical and non-​medical spe- suggests that rather than defining specific age ranges, departments
    cialists in the multidisciplinary team (MDT) work to ensure that should treat according to the needs of the community they serve,
    all the needs of the patient and their family are addressed, but the ensuring that those who fall in the interface between children and
    patient’s personal involvement may be less than they want (to ‘pro- adult service have their needs met.
    tect’ them), increasing problems with challenge to authority and
    Care environment
    poor treatment concordance.
    • Adult services aim to be patient-​centred (not family-​centred), Hospitals in many countries now have specified care environments
    and clinical management is structured according to disease site. for AYA cancer patients. These comprise inpatient wards and/​or out-
    Again, AYAs are isolated by this from other young people, and patient facilities which provide an environment, staff, and facilities
    100 CHAPTER 13 Cancer in Adolescents and Young Adults

    Table 13.1 The models of adolescent and young adult care currently adopted by various countries, and the projects that are currently
    in place

    UK France Germany Italy Spain The Denmark The Czech


    Netherlands Republic
    Model of Collaboration National integrated Inpatient treatment Developed within Developed Health No nation
    care between cooperative of adolescent the national by local professionals model
    paediatric and programme, joining cancer patients paediatric paediatric from regional
    adult services; paediatric and < 18 years takes oncology oncologists. centres started
    specalization in adult oncology/​ place within a community. a national
    AYA care. haematology. German Society Currentlydeveloping cooperative
    National cancer for Paediatric a national AYA project in
    plans have an AYA Oncology and programme, 2013, initiated
    element. Haematology through by medical
    (GPOH)-​certified collaboration oncology,
    unit. between paediatric focused on
    and adult those in the 18-​
    oncologists. 35 age range.
    Age range 15–​24 years 15–​25 years >18 years 15–​29 years 14–​30 years 18–​35 years Not defined 15–​19 years
    (13–​24 years
    in several
    departments)
    Local 25 AYA units 8 AYA programmes, 4 centres with 2 TYA units, 1 1 TYA unit 2 AYA centres 2 AYA units Collaboration
    projects with 3 AYA units interdisciplinary regional project and local on leukaemia
    programmes initiatives in trials
    paediatric
    centres

    Adapted with permission from Stark D et al. ‘Teenagers and young adults with cancer in Europe: from national programmes to a European integrated coordinated project’. European
    Journal Cancer Care, Volume 25, Issue 3, pp. 419–​427, Copyright © 2015, John Wiley and Sons. DOI: https://​doi.org/​10.1111/​ecc.1236

    tailored to the holistic needs of AYA patients. These also raise the services in local hospitals.4 This requires robust healthcare profes-
    profile and act as physical focal points for the AYA MDT. Often these sional networks across different care centres, and may also be helpful
    environments display colourful, contemporary, and creative décor, in managing the late effects of cancer.
    and cater for the variable needs of AYAs. This may include games France has an established national cooperative programme,
    areas, music facilities, study zones, relaxation areas, and places to integrating paediatric and adult oncology/​haematology services. The
    socialize and spend time with other young people. AYAs have em- national cancer plans involve AYA malignancies, and describe the
    phasized the importance of access to youth workers, psychologists, establishment of regional care pathways and an increased amount of
    and therapists. These environments are fundamental in facilitating research in AYA oncology. The Institut National du Cancer (INCa)
    peer support, which is hugely important in this age group, and have has set out clear guidelines for how each of their eight AYA centres
    a positive effect on patient experience and engagement in their and teams should be structured, including the integration of medical
    healthcare. They also bring their own specific challenges, such as be- and nursing staff with psychologists and social workers. In the USA,
    reavement of other patients, although on balance they are reported the number of specialist centres is growing.
    by young people to be highly beneficial in overcoming some of the
    social stigma of cancer. The need for adolescent and young adult-​specific
    professional training
    Centralization of services versus local hospital care The distinct population and needs highlights a training gap in both
    In order to provide experienced, specialized teams, treating a large paediatric and adult oncology. Professionals need to be educated in
    number of patients with the variety of AYA malignancies, the care the distinct biology and clinical management of AYA cancers, as well
    of these patients needs to be led from larger centres. Related to this as the psychosocial and interpersonal issues. AYAs have commented
    is the challenge of conducting AYA-​friendly clinical trials (see the on the importance of personal qualities, and the ability to com-
    section ‘Accrual in clinical trials’). However, the negative conse- municate effectively with young people and to be able to develop
    quence of this is the requirement for some patients and their fam- a good rapport. Some of this ability is cancer-​specific and some is
    ilies to travel significant distances in order to receive care. This also generic to staff working with young people with any chronic disease.
    affects the patient’s social and vocational development, as they may This training is reflected in adolescent health programmes created
    be moved far from their social network, educational institution, or through collaboration between the Royal Colleges of Paediatrics
    workplace. and Child Health, Nursing, General Practice, and Obstetricians and
    In 2005, the UK National Institute of Clinical Excellence (NICE) Gynaecologists in the UK, and intended for all health professions
    Improving Outcomes Guidance for Cancer in Children and Young working with children aged 12–​18 years. It is openly available as an
    People specified the need for all AYA cancer patients to be overseen online e-​learning module.5
    by the regional principal treatment centre (PTC) team. Patients may There is a growing recognition of the need for professional training
    request to receive their care at their local designated hospitals. More and qualifications for staff working specifically in AYA oncology.
    recently, some AYA PTCs have provided specialist AYA outreach The aim is that, in collaboration with universities and professional
    Late effects 101

    societies, healthcare staff will be able to achieve an accredited quali- A comprehensive, multi-​level, coordinated strategy specifically
    fication in AYA oncology. International courses include: addressing the poor clinical trial enrolment among AYAs is neces-
    sary to overcome barriers: for example, regulatory and practice-​
    • http://​ w ww.coventry.ac.uk/​ study-​ at-​ c oventry/​ f aculties- ​ and-​
    related challenges; that in some trials, regulatory guidelines are
    schools/​health-​and-​life-​sciences/​nursing-​midwifery-​and-​health/​
    implemented to protect ‘minors’; the fact that AYAs with cancer are
    teenager-​young-​adults-​cancer-​care/​
    not referred to institutions where clinical trials are offered; the age
    • http://​university.asco.org/​focus-​under-​forty
    eligibility criteria of protocols or the age cut-​off for the admission
    • https://​www.youtube.com/​playlist?list=PLD962FF429F3E4F4A to paediatric or adult medical wards; and the insufficient collabor-
    • http://​mdhs-​study.unimelb.edu.au/​degrees/​graduate-​certificate-​ ation between paediatric and adult medical oncology cooperative
    in-​adolescent-​health-​and-​wellbeing/​overview groups. But there are also patient-​related challenges that can affect
    • http:// ​ o ncologypro.esmo.org/ ​ Tumour- ​ Portal/ ​ C ancer- ​ i n- an adolescent’s decision to enrol and adhere to a clinical trial: young
    ​Young-​Adults people may feel shocked at the diagnosis, can experience denial, and
    Some have suggested the establishment of the ‘AYA oncologist’. might be overwhelmed by a long-​term commitment to adhering to
    Currently, systems are not mature enough for specific training a protocol.
    schemes and whole departments for AYA. More achievable, how- The low referral rate of adolescents as compared to children
    ever, is the formation of further AYA-​specialized MDTs, in which all has been discussed in a recent study developed by the European
    have expertise and qualifications. paediatric Soft-​Tissue Sarcoma Study Group (EpSSG), that com-
    pared the number of 15–​19 year olds (and 0–​14 year olds) treated
    in EpSSG trials with the number of cases expected to occur in
    European countries according to the incidence rates estimated
    Accrual in clinical trials through population-​based cancer registries. The ‘observed to ex-
    pected’ ratio (in the period 2008–​2015) was 0.64 for children and
    The best possible access to care remains a challenge for AYA with 0.30 for adolescents (with differences according to sarcoma subtypes
    cancer; in particular, it has been variously reported that patients and the different countries considered). That study emphasizes that
    in this age category are enrolled in clinical protocols at far lower the AYA-​dedicated national programmes developed across Europe
    rates than children. The accrual rate gap has been suggested as should increase the exchanges with the disease-​specific cooperative
    one of the possible reasons why AYAs with certain tumour types groups running clinical trials.
    have worse survival rates than children with the same disease.
    On the one hand, there is no question that clinical trials are a
    fundamental part of cancer research, benefitting subsequent Late effects
    generations of patients and furthering the scientific knowledge;
    meanwhile, on the other hand, it is difficult to demonstrate Cancer and its treatment are toxic, and AYAs are potentially exposed
    whether participating in clinical trials improves survival on an to a wide range of chemotherapy agents, surgery, radiotherapy, and
    individual level. However, there is indirect evidence suggesting other systemic agents such as immunological or hormonal ther-
    a positive effect of trials on participants’ outcome. Participation apies, each with distinct toxic effects that may occur acutely but also
    in trials may be associated with a better quality of care thanks to disrupt health and well-​being many years after completion of cancer
    the involvement of a broader group of highly-​specialized profes- treatment. Some of these effects are distinct in AYAs compared to
    sionals and/​or to the stricter organization and process control older or younger patients. These result in significant burden upon
    demanded by clinical protocols. young people and health services. Most late effects are linked to the
    A review of various studies showed that the rates of AYAs entering therapy received, so AYAs should be specifically informed as soon as
    clinical trials ranged from 5% to 34%. For example, a French they are receptive to this complex information; this may benefit from
    population-​based study underlined that most teenagers (82%) were the expertise found in AYA centres in counselling young people.
    treated on adult units, and only a small proportion of them (27%)
    were recruited in clinical trials. A study conducted by the Italian Cardiovascular disease
    national paediatric oncology society AIEOP (Associazione Italiana Chemotherapy is associated with delayed cardiovascular complica-
    Ematologia Oncologia Pediatrica) demonstrated that only 10% of tions including hypertension, accelerated atherosclerosis, coronary
    15-​to 19-​year-​old cancer patients were treated at AIEOP-​affiliated artery disease, and heart failure. In a Danish cohort of more than
    paediatric oncology centres (between 1989 and 2006), as opposed 43,000 AYA cancer survivors, an absolute excess risk of up to 400
    to 77% of children (0–​14 years); in addition, only half of adoles- extra hospitalizations due to cardiovascular disease per 100,000
    cents were enrolled in clinical trials. However, the proportion of person years in survivors was found. The mediastinal radiotherapy
    15-​19 yearolds treated at Italian paediatric oncology referral centres commonly used for the treatment of lymphomas and anthracycline
    thankfully improved significantly in the subsequent years, reaching chemotherapy have been linked to increased risk. Regardless of
    28% for the years 2007–​2012. Improvement in the adolescent accrual therapy received, all AYA survivors should be offered lifestyle advice
    rate was also recently observed in the UK (from the 25% reported to reduce their risk of cardiovascular disease.
    in the 2005–​2007 period, with variations across cancer types), as a
    result of a global healthcare policy including a strategic approach Second malignant neoplasms
    to improve AYAs’ recruitment to clinical trials based on five key AYAs are at greater risk of developing new, secondary cancers
    words: ‘available, accessible, aware, appropriate, and acceptable’. due to predisposition or treatment, when compared to older
    102 CHAPTER 13 Cancer in Adolescents and Young Adults

    survivors. AYAs should be supported to follow the appropriate


    Future national and international priorities
    evidence-​based cancer screening guidelines applicable to them-
    in this area
    selves as members of the general population, but also additional
    procedures. For example, young women who had thoracic radio-
    Internationally, various recent priorities have been set in this field.
    therapy should receive mammogram and magnetic resonance
    In the USA, in 2013, an update meeting was held to review evidence,
    imaging (MRI) due to their excess breast-​cancer risk. In all AYA
    scientific data, and priorities for AYAs aged 15–​39 at cancer diag-
    survivors it is essential to support the young person to undertake
    nosis. A group of 60 experts identified the large increase in pub-
    lifestyle modifications that reduce cancer risk—​smoking cessa-
    lished research in seven years and prioritized next steps, including:
    tion, alcohol moderation, moderate sun exposure, a balanced diet,
    and regular exercise. • Epidemiology—​subgroup analyses in incidence and outcomes,
    population-​based data, and treatment data
    Fertility
    • Biology—​ translational studies on ‘difference’ in AYA cancer
    Treatment-​induced reduction in fertility, either in terms of lowered biology, tissue banking, and data sharing
    immediate fertility or reduction in the duration of fertility after • Trials—​ population trial-​entry data, age of eligibility, trial
    treatment, is primarily due to dose-​dependent gonadal toxicity, intergroups and collaborations, and consent processes
    although disruption of reproductive endocrine or reproductive • Health services—​place of care, pathways to care, financial toxicity,
    anatomy should also be considered. The risk depends on the age at protocols, adherence, and insurance coverage
    treatment (higher with older age) and the treatment selected. Those
    • Patient-​reported outcomes—​prospective, comparative, interven-
    containing radiotherapy and/​or alkylating agents such as cyclophos-
    tion, and psychometric study designs
    phamide and procarbazine are considered as the most toxic to the
    gonad. In all AYA patients planning to receive any treatment that In Canada, a group in Toronto have prioritized symptom manage-
    may impact upon future fertility, expert discussion regarding fer- ment, return to work, and other aspects of survivorship such as
    tility preservation early in treatment planning is recommended, in fatigue. Fernandez et al. identified, as priorities, age-​specific psy-
    both men and women, followed by referral to specialized services if chosocial, survivorship, palliative, and cancer care, and research
    the patient wishes. to redress inequities in care for both the younger and older cancer
    patient. A pan-​European group, funded by the European Union,
    Fatigue prioritized work on professional networks and professional educa-
    Fatigue is prevalent in AYA cancer survivors; Spathis et al. define tion, clinical trials, international comparative epidemiology, health
    it as a ‘distressing, persistent, subjective sense of physical, emo- lifestyles for young people to minimize cancer risk, and public and
    tional, and/​or cognitive tiredness or exhaustion related to cancer patient involvement in cancer services and research.
    or cancer treatment that is not proportional to recent activity Globally, the importance of collaborative working and a network
    and interferes with usual functioning’ (see ‘Further reading’). approach, bringing together the patient, charities, professionals, and
    When reversible medical causes have been excluded (such as an- their organizations is growing in the AYA field, including:
    aemia or depression), structured aerobic exercise routines (such
    • Professional societies with specific groups working on AYAs
    as moderate walking exercise) may be helpful, as may educational
    through SIOPE ESMO and ASCO
    interventions.
    • Professional networking groups such as CPAC (Canada),
    Psychosocial late effects ENTYAC (EU), Go-​ AJA (France), AjeT (Germany), SIAMO
    (Italy), Space4AYA (Netherlands), AAA (Spain), TYAC (UK), and
    Anxiety, depression, and stress are more prevalent in AYAs than
    Critical Mass (USA)
    some other patient groups living after cancer, particularly soon after
    treatment. Financial issues, difficulties with work or education, and • Charities such as Teenage Cancer Trust (UK), CanTeen (Australia),
    maintaining or forming close or sexual relationships with peers and Canteen (Eire), and Teen Cancer America (USA)
    life partners are common issues reported by AYAs over a prolonged Working together, the next few years appear promising for further im-
    period. It is important to offer support, to highlight reliable sources provements in outcomes for young people living with and after cancer.
    of information to ameliorate these adverse effects of cancer and its
    treatment, and also to manage AYAs in a multi-​professional team
    with occupational health and social services that have expertise in NOTES
    managing these specific issues in AYAs. 1. See http://​www.sst.dk/​~/​media/​A7052DCF93A641508A48A5B60
    A933A7D.ashx
    Transition 2. See http://​www.cancerworld.org/​Articles/​Issues/​66/​May-​June-​2015/​
    Cancer care services can only provide expert specialist care for sub- e-​Grand-​Round/​719/​Teenagers-​and-​young-​adults-​with-​cancer-​
    groups of the population. Therefore, as children living with or after addressing-​the-​most-​important-​care-​needs.html
    cancer grow up into young people, it can be very good practice to 3. See https://​www.teenagecancertrust.org/​about-​us/​what-​we-​do/​
    professional-​leadership/​international-​conference and http://​
    transition their care to AYA services who specialize in that aspect
    criticalmass.org/​.
    of care. Later, they will move on to adult services and, eventually,
    4. See https://​www.teenagecancertrust.org/​about-​us/​what-​we-​do/​
    if care has been successful, elderly care services. These transition support-​outside-​our-​units
    processes will not succeed if they are not carried out proactively 5. See http://​www.e-​lfh.org.uk/​programmes/​adolescent-​health/​
    and expertly.
    Future national and international priorities in this area 103

    Felgenhauer J, Hooke MC (2014) Regulatory barriers to clinical trial


    FURTHER READING enrollment of adolescent and young adult oncology patients.
    Abrams AN, et al. (2007) Psychosocial issues in adolescents with Pediatrics 133, S119–​22. doi: 10.1542/​peds.2014-​0122H
    cancer. Cancer Treat Rev 33, 622–​30. Fern L, et al. (2008) Rates of inclusion of teenagers and young adults in
    Abrams AN, et al. (2007) Psychosocial issues in adolescents with England into National Cancer Research Network clinical trials: re-
    cancer. Cancer Treat Rev 33, 622–​30. port from the National Cancer Research Institute (NCRI) Teenage
    Ahrensberg JM, et al. (2016) Primary care use before cancer diagnosis and Young Adult Clinical Studies Development Group. Br J Cancer
    in adolescents and young adults—​a nationwide register study. PLoS 99, 1967–​74.
    One, 11, e0155933. Fern LA, et al. (2013) Why can't we improve the timeliness of cancer
    Barnett M, et al. (2016) Psychosocial outcomes and interventions diagnosis in children, teenagers, and young adults? BMJ 347, f6493.
    among cancer survivors diagnosed during adolescence and young Fern LA, et al. (2014) Available, accessible, aware, appropriate, and ac-
    adulthood (AYA): a systematic review. J Cancer Surviv 10(5), ceptable: a strategy to improve participation of teenagers and young
    814–​31. adults in cancer trials. Lancet Oncol 15, e341–​50.
    Barr RD, et al. (2016) Cancer in adolescents and young adults: a nar- Fernandez C, et al. (2011) Principles and recommendations for the
    rative review of the current status and a view of the future. JAMA provision of healthcare in Canada to adolescent and young adult-​
    Pediatr 170(5), 495–​501. aged cancer patients and survivors. J Adolesc Young Adult Oncol
    Barr RD, et al. (2016) Incidence and incidence trends of the most fre- 1, 53–​9.
    quent cancers in adolescent and young adult Americans, including Ferrari A, et al. (2012) Upper age limits for accessing pediatric on-
    ‘nonmalignant/​noninvasive’ tumors. Cancer 122, 1000–​8. cology centers in Italy: a barrier preventing adolescents with cancer
    Bennett S, et al. (2016) Educational interventions for the management from entering national cooperative AIEOP trials. Pediatr Hematol
    of cancer-​related fatigue in adults. Cochrane Database Syst Rev 11, Oncol 29, 55–​61.
    CD008144. Ferrari A, Bleyer A (2007) Participation of adolescents with cancer in
    Bleyer A, Barr R (2009) Cancer in young adults 20 to 39 years of clinical trials. Cancer Treat Rev 33, 603–​8.
    age: overview. Semin Oncol 36, 194–​206. Ferrari A, et al. (2009) Adolescents with cancer in Italy: entry into the
    Bleyer A, et al. (2005) National survival trends of young adults with national cooperative paediatric oncology group AIEOP trials. Eur
    sarcoma: lack of progress is associated with lack of clinical trial par- J Cancer 45, 328–​34.
    ticipation. Cancer 103, 1891–​7. Ferrari A, et al. (2010) Starting an adolescent and young adult pro-
    Bleyer A, et al. (2006) Cancer in 15-​to 29-​year-​olds by primary site. gram: some success stories and some obstacles to overcome. J Clin
    Oncologist 11, 590–​601. Oncol 28, 4850–​7.
    Bleyer A, et al. (2012) Children, adolescents, and young adults with Ferrari A, et al. (2015) Clouds of oxygen: adolescents with cancer tell
    leukemia: the empty half of the glass is growing. J Clin Oncol 30, their story in music. J Clin Oncol 33, 218–​21.
    4037–​38; author reply 4038–​9. Ferrari A, et al. (2016) The sooner the better? How symptom interval
    Bleyer WA, et al. (eds) (2006) Cancer Epidemiology in Older Adolescents correlates with outcome in children and adolescents with solid tu-
    and Young Adults 15 to 29 Years of Age, Including SEER Incidence mors: regression tree analysis of the findings of a prospective study.
    and Survival, 1975–​2000. Bethesda: National Cancer Institute. Pediatr Blood Cancer 63, 479–​85.
    Bleyer WA, et al. (1997) National cancer clinical trials: children have Ferrari A, et al. (2016) Adolescents with cancer in Italy: improving ac-
    equal access; adolescents do not. J Adolesc Health 21, 366–​73. cess to National Cooperative Pediatric Oncology Group (AIEOP)
    Butow P, et al. (2010) Review of adherence-​related issues in adolescents centers. Pediatr Blood Cancer 63, 1116–​19.
    and young adults with cancer. J Clin Oncol 28, 4800–​9. Ferrari A, et al. (2017) Access to clinical trials for adolescents with soft
    Campbell F, et al. (2016) Transition of care for adolescents from paedi- tissue sarcomas: enrollment in European pediatric Soft tissue sar-
    atric services to adult health services. Cochrane Database System coma Study Group (EpSSG) protocols. Pediatr Blood Cancer 64(6).
    Rev 4, CD009794. doi: 10.1002/​pbc.26348
    Chiavari L, et al. (2015) Difficult choices for young patients with Frazier AL, et al. (2008) Application of the adult international germ
    cancer: the supportive role of decisional counseling. Support Care cell classification system to pediatric malignant non-​seminomatous
    Cancer 23, 3555–​62. germ cell tumors: a report from the Children's Oncology Group.
    Coccia PF, et al. (2014) Adolescent and young adult oncology, version Pediatr Blood Cancer 50, 746–​51.
    2. J Natl Compr Canc Netw 12, 21–​32; quiz 32. Frazier AL, et al. (2015) Revised risk classification for pediatric extra-
    Collins M, et al. (2013) Benefits and adverse events in younger versus cranial germ cell tumors based on 25 years of clinical trial data from
    older patients receiving neoadjuvant chemotherapy for osteosar- the United Kingdom and United States. J Clin Oncol 33, 195–​201.
    coma: findings from a meta-​analysis. J Clin Oncol 31, 2303–​12. FREYER DR, Brugieres L (2008) Adolescent and young adult on-
    Desandes E, Stark DP (2016) Epidemiology of adolescents and cology: transition of care. Pediatr Blood Cancer 50, 1116–​19.
    young adults with cancer in Europe. Prog Tumor Res 43, 1–​15. FUNG C, et al. (2013) Solid tumors after chemotherapy or surgery for
    doi: 10.1159/​000447037 testicular nonseminoma: a population-​based study. J Clin Oncol 31,
    Desandes E, et al. (2012) Pathways of care for adolescent patients with 3807–​14.
    cancer in France from 2006 to 2007. Pediatr Blood Cancer 58, 924–​9. Gatta G, et al. (2014) Childhood cancer survival in Europe 1999–​
    Dieluweit U, et al. (2011) Educational and vocational achievement 2007: results of EUROCARE-​5—​a population-​based study. Lancet
    among long-​ term survivors of adolescent cancer in Germany. Oncol 15, 35–​47.
    Pediatr Blood Cancer 56, 432–​8. doi: 10.1002/​pbc.22806. Geraci M, et al. (2007) Cancer mortality in 13 to 29-​year-​olds in
    Dommett RM, et al. (2013) Features of cancer in teenagers and young England and Wales, 1981–​2005. Br J Cancer 97, 1588–​94.
    adults in primary care: a population-​based nested case-​control Gibson F, et al. (2012) A scoping exercise of favourable characteris-
    study. Br J Cancer 108, 2329–​33. tics of professionals working in teenage and young adult cancer
    104 CHAPTER 13 Cancer in Adolescents and Young Adults

    care: 'thinking outside of the box'. Eur J Cancer Care (Engl) NICE(UK) (2005) Improving Outcomes in Children and Young People
    21, 330–​9. with Cancer. Cancer Service Guidance. London: NICE.
    Gupta AA, et al. (2010) Clinical outcome of children and adults with O’Hara C, et al. (2015) Trends in survival for teenagers and young
    localized Ewing sarcoma: impact of chemotherapy dose and timing adults with cancer in the UK 1992–​2006. Eur J Cancer 51, 2039–​48.
    of local therapy. Cancer 116, 3189–​94. doi: 10.1002/​cncr.25144 Partridge AH, et al. (2016) Subtype-​dependent relationship between
    Gupta AA, et al. (2012) Patterns of chemotherapy-​induced toxicities young age at diagnosis and breast cancer survival. J Clin Oncol 34,
    in younger children and adolescents with rhabdomyosarcoma: a 3308–​14. doi: 10.1200/​JCO.2015.65.8013
    report from the Children's Oncology Group Soft Tissue Sarcoma Peccatori FA, et al. (2013) Cancer, pregnancy and fertility: ESMO
    Committee. Cancer 118, 1130–​7. Clinical Practice Guidelines for diagnosis, treatment and follow-​
    Gupta AA, et al. (2016) Reimagining care for adolescent and young up. Ann Oncol 24(Suppl 6), vi160–​70.
    adult cancer programs: moving with the times. Cancer 122, Peppercorn JM, et al. (2004) Comparison of outcomes in cancer pa-
    1038–​46. tients treated within and outside clinical trials: conceptual frame-
    Hayes-​Lattin B, et al. (2010) Adolescent and young adult oncology work and structured review. Lancet 363, 263–​70.
    training for health professionals: a position statement. J Clin Oncol Pini S, et al. (2013) The impact of a cancer diagnosis on the education
    28, 4858–​61. engagement of teenagers—​patient and staff perspective. Eur J Oncol
    Hollis R, Morgan S (2001) The adolescent with cancer—​at the edge of Nurs 17, 317–​23. doi: 10.1016/​j.ejon.2012.08.005
    no-​man's land. Lancet Oncol 2, 43–​8. Pini S, et al. (2012) What effect does a cancer diagnosis have on the
    Jakes AD, et al. (2014) Critical review of clinical practice guidelines for educational engagement and school life of teenagers? A systematic
    fertility preservation in teenagers and young adults with cancer. J review. Psycho-​oncology 21, 685–​94. doi: 10.1002/​pon.2082
    Adolesc Young Adult Oncol 3, 144–​52. Richardson DP, et al. (2015) Hospitalization rates among survivors of
    Jones LJ, et al. (2016) How do teenagers and young adults with cancer young adult malignancies. J Clin Oncol 33, 2655–​9.
    experience their care? A European survey. J Adolesc Young Adult Robert C, et al. (2015) Improved overall survival in melanoma with
    Oncol 17, 17. combined dabrafenib and trametinib. N Engl J Med 372, 30–​9.
    Keegan TH, et al. (2016) Comparison of cancer survival trends in the doi: 10.1056/​NEJMoa1412690
    United States of adolescents and young adults with those in chil- Robison LL, Hudson MM (2014) Survivors of childhood and ado-
    dren and older adults. Cancer 122, 1009–​16. lescent cancer: life-​long risks and responsibilities. Nat Rev Cancer
    Kelly D, et al. (2004) 'Being in the same boat': ethnographic insights 14, 61–​70.
    into an adolescent cancer unit. Int J Nurs Stud 41, 847–​57. Rosen DS, et al. (2003) Transition to adult health care for adolescents
    Kohsaka S, et al. (2014) A recurrent neomorphic mutation in MYOD1 and young adults with chronic conditions: position paper of the
    defines a clinically aggressive subset of embryonal rhabdomyosar- Society for Adolescent Medicine. J Adolesc Health 33, 309–​11.
    coma associated with PI3K-​AKT pathway mutations. Nat Genet 46, Rugbjerg K, Olsen JH (2016) Long-​term risk of hospitalization for
    595–​600. somatic diseases in survivors of adolescent or young adult cancer.
    Kondryn HJ, et al. (2011) Treatment non-​adherence in teenage and JAMA Oncol 2(2), 193–​200.
    young adult patients with cancer. Lancet Oncol 12, 100–​8. Rugbjerg K, et al. (2014) Cardiovascular disease in survivors of adoles-
    Kumar A, et al. (2005) Are experimental treatments for cancer in chil- cent and young adult cancer: a Danish cohort study, 1943–​2009. J
    dren superior to established treatments? Observational study of Natl Cancer Inst 106(6), dju110.
    randomised controlled trials by the Children's Oncology Group. Smith AW, et al. (2013) Unmet support service needs and health-​
    BMJ 331(7528), 1295. related quality of life among adolescents and young adults with
    Kwak M, et al. (2013) Trajectories of psychological distress in adoles- cancer: the AYA HOPE Study. Front Oncol 3, 75.
    cent and young adult patients with cancer: a 1-​year longitudinal Smith AW, et al. (2016) Next steps for adolescent and young adult on-
    study. J Clin Oncol 31, 2160–​6. cology workshop: An update on progress and recommendations for
    Lambertini M, et al. (2016) Cancer and fertility preservation: inter- the future. Cancer 122, 988–​99.
    national recommendations from an expert meeting. BMC Spathis A, et al. (2015) Teenage and young adult cancer-​related fa-
    Med 14, 1. tigue is prevalent, distressing, and neglected: it is time to intervene.
    Lee JS, et al. (2016) Increased risk of second malignant neoplasms in A systematic literature review and narrative synthesis. J Adolesc
    adolescents and young adults with cancer. Cancer 122, 116–​23. Young Adult Oncol 4, 3–​17.
    Michelagnoli MP, et al. (2003) Adolescent oncology—​a homeland for Stark D, et al. (2015) Teenagers and young adults with cancer in
    the ‘lost tribe’. Eur J Cancer 39, 2571–​2. Europe: from national programmes to a European integrated co-
    Moricke A, et al. (2005) Prognostic impact of age in children and ado- ordinated project. Eur J Cancer Care (Engl) 25(3), 419–​27.
    lescents with acute lymphoblastic leukemia: data from the trials Stiller CA, et al. (2012) Population survival from childhood cancer
    ALL-​BFM 86, 90, and 95. Klin Padiatr 217, 310–​20. in Britain during 1978–​2005 by eras of entry to clinical trials. Ann
    Mulder RL, et al. (2013) Recommendations for breast cancer surveil- Oncol 23, 2464–​69. doi: 10.1093/​annonc/​mds183
    lance for female survivors of childhood, adolescent, and young Sultan I, et al. (2009) Comparing adult and pediatric rhabdomyosar-
    adult cancer given chest radiation: a report from the International coma in the surveillance, epidemiology and end results program,
    Late Effects of Childhood Cancer Guideline Harmonization Group. 1973 to 2005: an analysis of 2,600 patients. J Clin Oncol 27, 3391–​7.
    Lancet Oncol 14, e621–​9. Tai E, et al. (2014) Understanding and addressing the lack of clinical
    Munhoz RR, et al. (2016) Gonadotropin-​releasing hormone agon- trial enrollment among adolescents with cancer. Pediatrics 133,
    ists for ovarian function preservation in premenopausal women S98–​103. doi: 10.1542/​peds.2014-​0122D
    undergoing chemotherapy for early-​stage breast cancer: a system- Taylor RM, et al. (2013) Developing a conceptual model of teenage
    atic review and meta-​analysis. JAMA Oncol 2, 65–​73. doi: 10.1001/​ and young adult experiences of cancer through meta-​synthesis. Int
    jamaoncol.2015.3251 J Nurs Stud 50, 832–​46.
    Future national and international priorities in this area 105

    Teenage Cancer Trust (2016) The Blueprint of Care: For Teenagers and Warner EL, et al. (2016) Social well-​being among adolescents and
    Young Adults with Cancer (2nd edn). London: Teenage Cancer Trust. young adults with cancer: a systematic review. Cancer 122(7),
    Tian L, et al. (2016) Effects of aerobic exercise on cancer-​related fa- 1029–​37.
    tigue: a meta-​analysis of randomized controlled trials. Support Care Weller D, et al. (2012) The Aarhus statement: improving design and
    Cancer 24, 969–​83. reporting of studies on early cancer diagnosis. Br J Cancer 106,
    Trama A, et al. (2016) Survival of European adolescents and young 1262–​7.
    adults diagnosed with cancer in 2000–​07: population-​based data Woodward E, et al. (2011) Late effects in survivors of teenage and
    from EUROCARE-​5. Lancet Oncol 17, 896–​906. young adult cancer: does age matter? Ann Oncol 22, 2561–​8.
    Tricoli JV, et al. (2016) Biologic and clinical characteristics of adoles- Zebrack B, et al. (2010) To foster healing among adolescents and young
    cent and young adult cancers: acute lymphoblastic leukemia, colo- adults with cancer. What helps? What hurts? Support Care Cancer
    rectal cancer, breast cancer, melanoma, and sarcoma. Cancer 122, 18, 131–​5.
    1017–​28. Zebrack B, Isaacson S (2012) Psychosocial care of adolescent and
    Veal GJ, et al. (2010) Clinical pharmacology in the adolescent on- young adult patients with cancer and survivors. J Clin Oncol 30,
    cology patient. J Clin Oncol 28, 4790–​9. 1221–​6. doi: 10.1200/​JCO.2011.39.5467
    14
    Late Effects of Therapy
    and Survivorship Issues
    Lars Hjorth, Riccardo Haupt, Gisela Michel, Maria Luisa Garrè,
    Leontien C.M. Kremer, and Rod Skinner

    Introduction persons is a CCS. It has been estimated that there are now about
    400,000 CCSs in the USA, and 300,000–​500,000 in Europe. Based on
    The high survival rate in children and adolescents with malignancy a conservative estimate that 75% of those treated on current proto-
    is accompanied by a substantial risk of late adverse events (LAEs). cols for childhood cancer will become long-​term survivors (more
    Treatment may interfere with physiological growth and psychosocial than5 years), each year about 8,000 new CCSs are added to each of
    development in children and adolescents, and have an important the USA and European populations. Their median age is estimated
    impact on health status later in life, whilst some late toxicities may at 29 years, some are already more than 50 years old, and more than
    even cause premature death. 20% have already survived more than 30 years since cancer diag-
    One study in the USA revealed that 62% of adult survivors of nosis. These data demonstrate the need for adult physicians to be
    childhood malignancy have at least one treatment-​induced chronic interested in the care of CCSs.
    health condition (and 38% have two or more), whilst 28% have a
    severe or life-​threatening problem. Similar data were reported in a
    Dutch study. Survivors of central nervous system (CNS) tumours or Late mortality
    haematopoietic stem-​cell transplantation (HSCT) are at particularly
    high risk. Much of the survivorship data has come from the North Cumulative mortality rates rise amongst CCSs as the duration of
    American Childhood Cancer Survivor Study (CCSS), the British their follow-​up continues. The BCCSS recently showed that CCSs
    Childhood Cancer Survivor Study (BCCSS), the Dutch Childhood had a 9.1-​fold increased risk of dying prematurely, which equals 64.2
    Oncology Group LATER (DCOG LATER), and the Nordic Adult excess deaths per 10,000 person years. Similar data have been re-
    Life after Childhood Cancer in Scandinavia (ALiCCS) study. ported by the Swiss population-​based registry, with an 11.0 times
    In general, frequently recognized risk factors include patient age increased risk of premature death and 45 excess deaths per 10,000
    at treatment, site and type of the tumour, type and cumulative dose person years. However, a decline in late mortality has been observed
    of chemotherapeutic agents, as well as dose and site of radiotherapy. with more recent treatment periods. Data from the CCSS found
    In addition, the patient’s pre-​existing clinical status and previous or 15-​year cumulative mortality for CCSs diagnosed between 1970
    concurrent exposure to other toxic treatments may predispose to and 1979 to be 10.7%, while it was 5.8% for those diagnosed be-
    greater late toxicity. tween 1990 and 1999. A Nordic study found the overall standardized
    The importance of moving from cure at any cost to cure at least mortality ratio (SMR) to be 8.3, with an absolute excess risk of 6.2
    cost is now well recognized, especially for patients with high cure per 1,000 person years. This study also found that after ten years,
    rates. For all patients, higher cure rates and the risk of late complica- the mortality changed very little with time period. During the first
    tions make long-​term follow-​up (LTFU) a vital component of care 10–​15 years after diagnosis, excess mortality is mainly due to the
    (Figure 14.1). primary cancer, but beyond this point there is increasing mortality
    from second malignant neoplasms (SMNs) and circulatory and pul-
    monary causes.
    Epidemiology There is good evidence that exposures to radiotherapy, alkylating
    agents, or epipodophyllotoxins carry the greatest relative risks (RRs)
    Recent estimates from several western countries put the propor- for late mortality from SMNs. The risk factors for long-​term cir-
    tion of long-​term childhood cancer survivors (CCSs) at 0.10–​0.15% culatory and cardiovascular risk have been less well determined.
    of the general population, implying that one in every 650–​1,000 A French and UK study evaluated this in the context of data on
    Secondary malignant neoplasms 107

    Northern Europe UK and Ireland Central Europe Southern Europe Eastern Europe All Europe

    Denmark
    Finland
    Iceland
    Norway
    Sweden
    Ireland
    UK (England and Wales)
    UK (Northern Ireland)
    UK (Scotland)
    Austria
    Belgium
    France
    Germany
    Switzerland
    Netherlands
    Croatia
    Italy
    Malta
    Portugal
    Slovenia
    Spain
    Bulgaria
    Estonia
    Hungary
    Latvia
    Lithuania
    Poland
    Slovakia
    Europe*

    0 10 20 30 40 50 60 70 80 90 100
    5-year survival (%)

    Figure 14.1 Five-​year survival for all cancers combined (CNS tumours excluded) diagnosed in 2000–​2007, by country, in European children (0–​
    14 years) of both sexes. The data are adjusted by age, sex, case mix, and period of diagnosis using a Cox proportional hazards model.
    Reproduced with permission from Gatta G. et al. 'Childhood cancer survival in Europe 1999-​2007: results of EUROCARE-​5—​a population-​based study', Lancet Oncology,
    Volume 15, Issue 1, pp. 35–​47. Copyright © 2014 Elsevier Ltd. DOI: https://​doi.org/​10.1016/​S1470-​2045(13)70548-​5

    both chemotherapy exposure and radiation dose to the heart. The cancers per 1,000 person years. After 20 years’ follow-​up, the cu-
    overall SMR for survivors was 8.3-​fold higher in relation to the gen- mulative risk is 3–​7%. The risk remains elevated even in the fifth
    eral populations in France and the UK, and there was a five-​fold and sixth decades of life. A recent study from the CCSS found that
    increase in death from cardiovascular disease. The risk of dying the risk of a second cancer remained higher for CCSs diagnosed in
    was significantly higher in individuals who had received a cumu- the 1990s than for the age-​matched general population, but that the
    lative anthracycline dose greater than 360 mg/​m2 and in individ- scale of increased risk was lower than that observed in CCSs diag-
    uals who received an average cardiac radiation dose that exceeded nosed in the 1970s. Use of lower treatment doses was considered to
    5 Gy. The largest analysis of late mortality yet is being performed by be the likeliest reason for this reduction in risk.
    the PanCareSurFup Consortium and includes 77,279 five-​year sur- The most common SMNs are breast cancer, myelodysplasia/​acute
    vivors with a mean follow-​up of 23 years. myeloid leukaemia (MDS/​AML), bone and soft-​tissue sarcoma, CNS
    These findings highlight the need for adult healthcare providers tumours (e.g. meningiomas), and thyroid cancer. The latency period
    to understand the risks confronting CCSs and for strategies to de- varies greatly, with an average of 12 years between first and second ma-
    tect serious complications early in an attempt to prevent excess late lignancies, with leukaemia occurring earlier and solid tumours later.
    deaths. Screening for secondary breast cancer is one example of such Patients with a genetic tumour predisposition have an increased risk
    an approach. of developing a SMN. Radiotherapy, even at doses of less than 1 Gy,
    is the main risk factor for developing SMNs, whilst alkylating agents
    and epipodophyllotoxins are the most commonly implicated cyto-
    Secondary malignant neoplasms toxic drugs. Recently, the LATER study showed that doxorubicin ex-
    posure increases the risk of subsequent solid cancers and breast cancer,
    Several studies have shown that CCSs have an increased risk of whereas cyclophosphamide increases the risk of subsequent sarcomas.
    developing SMNs. The RR is high (5–​10) but, in absolute num- Measures that CCSs can take to decrease their risk of developing
    bers, the individual risk is low, ranging from one to three additional SMNs include avoidance of tobacco smoking and protection
    108 CHAPTER 14 Late Effects of Therapy and Survivorship Issues

    from excessive exposure to sunlight. The International Guideline and working memory, and slower processing speed, perhaps due to
    Harmonization Group (IGHG) has been developing evidence based disruption of white-​matter circuitry in the developing brain.
    guidelines concerning surveillance for SMNs as well for other clin- Alternative options to CRT, such as high-​dose chemotherapy or in-
    ical relevant late complications. So far recommendations have been tensification of systemic chemotherapy (with or without intrathecal
    published for secondary breast cancer screening in female survivors chemotherapy), have been investigated in young children with ma-
    treated with radiotherapy in the breast region, as well for secondary lignant brain tumours, in order avoid irradiation or to delay it until
    thyroid cancer in survivors irradiated to the neck. Further guide- after three years of age. However, these treatment strategies have
    lines are under development addressing surveillance for secondary been successful for some tumour types such as medulloblastoma and
    CNS tumours and colon carcinoma. Moreover, within the recently supratentorial CNS embryonal tumours, but not for ependymomas.
    granted PanCareFollowUp project general consensus based recom- In children less than five to seven years old with midline low-​grade
    mendations are being developed using a more pragmatic approach gliomas not suitable for radical surgery, ‘gentle chemotherapy’ has
    to address secondary skin, bladder, bone, lung, and oral cancers. been successful in avoiding or delaying CRT.
    Other aspects of neurotoxicity are related more to specific treat-
    ment components. For example, vasculopathy and radiation ne-
    Neurological and neuropsychological toxicity crosis are associated with previous CNS radiotherapy, myelopathy
    with spinal radiotherapy, and peripheral neuropathy with cisplatin
    Long-​term neurological toxicity is the commonest LAE reported in or vincristine chemotherapy. In CNS tumour survivors, neuro-
    the CCSS, occurring in 27% of survivors, with an RR of 3.3 com- logical and neuropsychological status can be impaired not only by
    pared to a sibling control group. Cognitive impairments are fre- radiotherapy, chemotherapy, and age at treatment, but also by con-
    quently found in processing speed, attention, working memory, and sequences of the tumour itself and surgery. Surgical resection can
    executive function. Survivors of CNS tumours are at the highest risk, damage white-​matter integrity, reduce intellectual ability, and cause
    followed by survivors of Hodgkin lymphoma or acute leukaemia. deficits in attention, processing speed, and memory.
    However, not only tumour and treatment type, but also psycho- The management of neurotoxicity comprises both general and
    logical distress, social support, and long absences from school play specific measures. Physiotherapy and/​or occupational therapy as-
    an important role in the development of neurocognitive toxicity. sessment is very important in survivors with physical disabilities,
    There are several manifestations of CNS toxicity, including whilst children with leucoencephalopathy and neuropsychological
    leucoencephalopathy, vasculopathy (from stenosis or occlusion), ra- toxicity benefit from careful educational and neuropsychological as-
    diation necrosis, myelopathy, and secondary CNS tumours, with a sessment, with provision of extra support at school and subsequent
    wide variety of clinical sequelae threatening life or greatly impairing employment guidance. Psychotherapy and specific drug therapy can
    the survivor’s quality of life. In contrast, peripheral nervous system be used to treat depression and anxiety, resulting in a better quality
    toxicity has fewer manifestations, usually presenting as peripheral of life. Surgery may be considered in selected cases of vasculopathy
    neuropathy, although long-​term functional difficulties are increas- and drug treatment may be beneficial in painful neuropathy.
    ingly recognized in CCSs treated with vincristine (e.g. for acute In recent years, the doses of radiotherapy used in the treatment
    lymphoblastic leukaemia (ALL)) or cisplatin. of several childhood malignancies have been reduced (especially
    Leucoencephalopathy may present with focal motor signs, spasti- those used in CRT), as have the volumes of normal brain treated.
    city, seizures, and ataxia. Neuropsychological dysfunction may occur, Radiotherapy has been omitted in many malignancies (especially
    being apparent by cognitive impairment with considerable adverse ALL) or delayed until the child is older (e.g. in infants with brain
    impact on intelligence, attention and learning skills, memory, and tumours). Improvement of CRT techniques has been obtained with
    verbal and visual–​spatial skills. Vasculopathy may cause seizures, intensity-​modulated radiotherapy (IMRT), intensity-​modulated arc
    stroke, and dementia. Radiation necrosis may result in neurocognitive therapy (IMAT), and proton therapy. These techniques allow sparing
    and/​or neuropsychological dysfunction, whilst myelopathy may lead of normal brain with the aim of minimizing neurocognitive sequelae.
    to paraplegia or quadriplegia, as well as sphincter dysfunction. A var-
    iety of secondary CNS tumours are reported including malignant
    gliomas, neuroectodermal tumours, and meningiomas. Endocrine toxicity
    Most neurological/​neuropsychological toxicity is related to CNS
    radiotherapy and/​or CNS-​directed chemotherapy (i.e. intrathecal Endocrine complications are the second commonest group of
    chemotherapy, or certain chemotherapy regimens such as high-​ chronic conditions in CCSs, occurring in 20–​50% of survivors fol-
    dose methotrexate or cytarabine). Leucoencephalopathy and neuro- lowed into adulthood. They include hypothalamic–​pituitary, thyroid,
    psychological toxicity occur predominantly in survivors of CNS and gonadal dysfunction, bone disease, and metabolic disorders.
    tumours or acute leukaemia treated with CNS radiotherapy or CNS-​ Sequelae may be associated both with the tumour type and loca-
    directed chemotherapy, and are commoner and more severe in pa- tion, and with treatment. In addition to the effects of surgery, direct
    tients treated at a younger age. endocrine gland involvement by the malignancy (e.g. germinal tu-
    Cranial radiotherapy (CRT) can decrease the white-​matter volume mours, CNS tumours involving the hypothalamic–​pituitary axis
    and has been associated with reduced cognitive abilities across do- (HPA)), and hydrocephalus caused by some brain tumours, both
    mains including global intelligence, attention, working memory, radiotherapy and chemotherapy may increase the risk of endo-
    and executive dysfunction, translating into a higher risk of educa- crine complications. Survivors treated with radiotherapy to the
    tional and social difficulties. Intrathecal chemotherapy has been as- head, neck, or pelvis, and those treated with total body irradiation
    sociated with reduced academic achievement, deficits in attention (TBI) or with alkylating agents, are at increased risk.
    Gonadal dysfunction 109

    Hypothalamic–​pituitary dysfunction younger age. However, delayed puberty is more common. The risk
    The total dose of CRT is the strongest risk factor for anterior pitu- of developing gonadotrophin deficiency is dose-​related, especially
    itary dysfunction. However, the threshold dose for specific endo- with CRT doses of more than 30 Gy.
    crine dysfunction varies for different pituitary hormones. Patients at risk should be closely monitored during the peripubertal
    years (from nine years in females and ten years in males) or even be-
    Growth hormone fore in survivors at risk for precocious puberty. Growth and pubertal
    status assessed by Tanner scoring should be monitored at least yearly
    Growth hormone (GH) deficiency is the most common anterior pi-
    until completion of puberty; follicle-​stimulating hormone/​lutein-
    tuitary deficiency observed after CRT, and if it occurs, it is usually
    izing hormone (FSH/​LH) and 17β-​oestradiol or testosterone should
    permanent. It may become manifest after a CRT dose ≥18 Gy; after
    be evaluated in children with suspected early or delayed onset of
    more than 24 Gy, the deficiency usually becomes apparent within
    puberty. This allows identification of hypo-​or hypergonadotrophic
    five years, while for doses in the range of 18–​24 Gy, GH deficiency
    hypogonadism, or precocious puberty, and the initiation of appro-
    may develop ten years or even later after treatment, with increased
    priate replacement (delayed puberty) or antagonist (precocious pu-
    risk persisting throughout adult life.
    berty) treatment.
    Patients treated for brain tumours, sarcomas, or ALL are at higher
    risk of GH deficiency. The frequency of GH deficiency in CCSs Other hypothalamic trophins
    treated in the last two decades has fallen since leukaemic patients
    Both thyroid-​stimulating hormone (TSH) and adrenocorticotropic
    treated with recent protocols are less likely to have received CRT.
    hormone (ACTH )deficiency may occur after radiotherapy doses
    Growth and pubertal development should be monitored closely
    to the brain of more than 30 Gy in up to 10% of patients at risk.
    in children at risk of GH deficiency to facilitate early diagnosis and
    Hyperprolactinaemia is rare and usually subclinical, occurring only
    treatment. Untreated GH deficiency will impair growth velocity
    in women treated with high-​dose radiotherapy.
    and eventually final height. However, other factors such as spinal
    radiotherapy (which may impair vertebral growth), age at treat- Hypothalamic obesity syndrome
    ment, and timing of puberty contribute to the individual’s final
    Hypothalamic obesity may occur in patients with tumours infiltrating
    height. Importantly, precocious puberty (see below) may mask GH
    the medial hypothalamic region, like craniopharyngiomas. Severe
    deficiency, which then becomes evident only after cessation of the
    obesity is seen in about 50% of patients, leading to a high rate of
    growth spurt. CNS tumour survivors are at higher risk of developing
    cardiovascular mortality. Structural damage to the hypothalamic
    precocious puberty, thus careful examination of such individuals
    nuclei may lead to hyperphagia, rapid weight gain, central insulin
    may maximize their potential to achieve adequate final height.
    and leptin resistance, decreased sympathetic activity, low energy ex-
    The most sensitive and specific test to diagnose GH deficiency is
    penditure, and increased energy storage in adipose tissue.
    the insulin tolerance test (ITT). If abnormal, another stimulation
    test (either with GNRH-​arginine or glucagon) should be performed
    to confirm the diagnosis. Measurement of insulin-​like growth factor Gonadal dysfunction
    (IGF-​1) and IGFBP-​3 alone is not a good predictor of GH deficiency.
    If GH deficiency is confirmed, GH replacement therapy should be Pubertal growth and reproductive function are controlled by hor-
    considered. Concerns have arisen about the potential impact of GH mones secreted by the testis or ovary. The main difference between
    treatment on tumour recurrence, but the available data do not dem- males and females is that in males, two distinct testicular cell types
    onstrate a significantly increased risk of tumour relapse in children control sexual development (Leydig cells) and fertility (Sertoli cells),
    treated with GH compared to those with the same type of tumour while in females, the oocyte controls both pubertal development and
    not treated with GH. Although a small but significant increased fertility.
    risk of SMNs seems to exist for survivors during GH treatment, the
    second tumours observed are the same as those occurring in those Ovarian dysfunction
    CCSs not treated with GH. Germ-​cell failure and loss of ovarian endocrine function occur con-
    comitantly in females. They may be due either to pelvic, abdom-
    Gonadotrophins inal, or spinal radiotherapy, TBI, or alkylating-​agent chemotherapy.
    Timely and appropriate gonadotrophin release from the HPA trig- There is no clear threshold dose for either radiotherapy or alkylating
    gers hormonal release from the testis or ovary. This allows a linear agents, but there is evidence that women exposed to higher doses
    growth spurt, development of secondary sexual characteristics, have a higher risk of premature ovarian insufficiency. Doses above
    fertility, and maintenance of gonadal function during adulthood. 20 Gy to the ovaries will almost always lead to permanent hormonal
    Except in a few cases of CNS tumours directly affecting the HPA, deficit. Age at treatment is important in predicting ovarian failure;
    CRT is the main risk factor for altered gonadotrophin release. In females treated at a younger age are less likely to develop ovarian
    these cases, gonadotrophins may be secreted either earlier or later failure, probably because of the higher number of primordial fol-
    (precocious or delayed puberty, respectively) than the usual physio- licles at the time of treatment.
    logical timing. The clinical picture depends on the age at onset and severity of the
    Precocious puberty has been associated with intracranial tu- ovarian damage. In pre-​or peripubertal girls, delay, arrest, or failure
    mours, raised intracranial pressure, cranial surgery, or radiation of pubertal onset may occur; in post-​pubertal women, the clinical
    doses to the HPA as low as those given for CNS prophylactic treat- picture may comprise secondary amenorrhoea, irregular menstrual
    ment in ALL (range 18–​24 Gy), in particular in females treated at a periods, or premature menopause. Laboratory testing demonstrates
    110 CHAPTER 14 Late Effects of Therapy and Survivorship Issues

    elevated gonadotrophins and reduced or absent 17β-​oestradiol. lesions (primarily adenomas), multinodular goitre, and malignant
    Replacement oestro-​progestogen therapy should be considered in lesions, with an average 5–​69-​fold increased risk compared to the
    order to induce puberty in females without onset and/​or mainten- normal population, depending on the radiotherapy dose. The effect
    ance of regular periods. Ovarian insufficiency may be permanent of radiotherapy on the thyroid gland seems to be both dose-​and
    or transient, thus replacement treatment should be discontinued at age-​dependent, with younger children at higher risk. A linear rela-
    intervals to determine if spontaneous periods occur. Most of these tionship with radiotherapy doses and thyroid late complications has
    females are infertile or at high risk of premature menopause, thus been observed, but after doses of more than 30 Gy, the cancer risk
    appropriate counselling should be given concerning the likely need decreases, probably because of a cell-​killing effect.
    for assisted reproduction. Thyroid hormones and TSH should be measured routinely in
    Uterine morphology (shape, dimensions, and structure, e.g. fi- patients at risk. Overt (clinical) primary hypothyroidism results in
    brosis or reduced endometrial thickness) should be evaluated in fe- low thyroid hormone values with high TSH concentrations, whilst
    males treated with pelvic or abdominal radiotherapy to assess the subclinical hypothyroidism is characterized by normal thyroxine
    likelihood of embryonal implantation or successful completion of but high TSH levels. Replacement therapy must be prescribed in
    foetal development. case of overt hypothyroidism and should be considered in case of
    multinodular goitre.
    Testicular dysfunction Surveillance for thyroid cancer should be recommended to all
    The testis is sensitive both to chemotherapy and radiotherapy. patients at risk. Careful clinical examination of the neck to detect
    Among chemotherapeutic agents, cumulative doses of alkylating nodules or lumps should be performed regularly, and some centres
    agents and timing of treatment influence the risk of oligo/​azoo- supplement this with ultrasound imaging, although it is uncertain
    spermia. Damage to germinal epithelium is estimated to occur after which, if any, strategy is superior in terms of reducing morbidity
    a testicular radiotherapy dose of more than 2–​3 Gy (and occasion- and mortality from thyroid cancer. Therefore no strong recommen-
    ally even lower doses) either as direct testicular radiotherapy or dation can be made about the type and timing of screening modal-
    during abdominal or spinal radiotherapy or an inverted Y field for ities. This aspect has been summarized by the IGHG, addressing the
    Hodgkin lymphoma treatment. Leydig cells are much more radio-​ balance between benefits and harms of each screening modality.
    resistant and only doses ≥20 Gy or TBI are likely to cause complete Needle biopsy may be considered for suspicious abnormalities.
    primary hypogonadism. In contrast to females, age at treatment has
    minimal impact on testicular function.
    The clinical picture of testicular insufficiency depends on the
    severity and timing of damage. Low testosterone levels with high Metabolic syndrome
    gonadotrophins indicate decreased Leydig cell function, resulting
    in delayed or absent pubertal onset in younger patients, or clin- Metabolic syndrome (MS) is characterized by a clustering of hyper-
    ical symptoms of testosterone deficiency such as reduced muscle tension, dyslipidaemia (low HDL cholesterol, hypertriglyceridaemia,
    strength, irritability, and decreased libido in post-​pubertal survivors. hypercholesterolaemia), type 2 diabetes, or preclinical conditions
    CCSs with evidence of testosterone deficiency should be referred (insulin resistance or glucose intolerance) and obesity. This condi-
    to an appropriate specialist for consideration of testosterone replace- tion is associated with a pro-​inflammatory and pro-​thrombotic state
    ment treatment. This should be given for delayed or absent pubertal that may lead to atherogenic dyslipidaemia. There is evidence that
    onset, or in adults with documented testosterone deficiency (less survivors exposed to CRT, prolonged steroid treatment, TBI, or ab-
    than 250 ng/​dl; less than 8.7 nmol/​l) with or without symptoms. dominal irradiation, and those with hypogonadism or limitations in
    Damage to the Sertoli cells is more frequent and it is nearly always physical performance (e.g. due to cardiac or pulmonary toxicity or
    permanent. A definitive diagnosis of male infertility may only be functional impairments caused by major surgery) are at increased
    made after semen analysis. For pre-​pubertal boys, studies of tes- risk of glucose intolerance and MS.
    ticular tissue cryopreservation are ongoing. Diagnosis of MS in adults is defined (International Diabetes
    Federation) by the presence of central obesity (waist circumfer-
    ence ≥94 cm in males, ≥80 cm in females), and at least two of the
    Thyroid toxicity following:
    • Raised blood pressure: more than 130/​85 mmHg
    The thyroid gland is sensitive to radiotherapy given either exter- • Raised triglycerides: more than 150 mg/​dl (1.7 mmol/​l)
    nally to the neck (as in Hodgkin lymphoma treatment, CRT, or
    • Reduced HDL cholesterol: less than 40 mg/​dl (1.03 mmol/​l) in
    TBI) or targeted via thyroid metabolism (131I treatment for thy-
    males, or less than 50 mg/​dl (1.29 mmol/​l) in females
    roid diseases like carcinoma or hyperthyroidism, sometimes after
    • Raised fasting glucose: ≥100 mg/​ml (5.6 mmol/​l).
    metaiodobenzylguanidine (MIBG) treatment in neuroblastoma due
    to uncoupling of 131I from 131I-​MIBG). The functional changes after The International Diabetes Federation does not recommend the diag-
    external-​beam radiotherapy sometimes occur as early as six months nosis of MS in children under ten years, but the criteria for the diagnosis
    after treatment, but may only become evident up to 20 years later, of MS in children from 10 to 16 years of age are broadly similar to those
    and comprise clinical or subclinical hypothyroidism with a com- used in adults. Survivors, and in particular those with predisposing or
    bined incidence of 20–​30%. Hyperthyroidism and autoimmune treatment-​related risk factors for MS, should be monitored with waist
    manifestations (Hashimoto thyroiditis) have been described in a circumference and markers of MS, and counselled about appropriate
    smaller proportion. The morphological changes consist of benign diet and/​or physical activity and avoidance of smoking.
    Other toxicities 111

    and chest-​wall abnormalities, with a cumulative incidence of 21%


    Other toxicities
    after 35 years. Chronic pulmonary health conditions were reported
    in 11.8% (RR 2.8) of survivors in the overall CCSS population. In
    Cardiac toxicity
    survivors of CNS malignancies, the CCSS found rates of pulmonary
    Cardiovascular disease, including heart failure, coronary artery fibrosis, chest-​wall abnormalities, chronic cough, and the need
    disease, pericarditis, arrhythmia, and valvular disease, is an im- for extra oxygen to be increased compared to siblings. There was
    portant health problem for CCSs. Survivors have a three-​to seven-​ a strong correlation between craniospinal irradiation and these ad-
    fold increased risk of premature death due to cardiac causes. Van verse outcomes.
    der Pal et al. concluded that one in eight survivors treated with Both restrictive and obstructive lung disease may occur, whilst
    anthracyclines and radiotherapy to the heart region will develop a bronchiectasis may result from previous infections of the lower re-
    life-​threatening cardiovascular event by 30 years after treatment. spiratory tract. Restrictive pulmonary disease may be subclinical,
    Heart failure is the most common cardiovascular disease. The manifest only by abnormalities in pulmonary function tests (PFTs),
    risk factors for heart failure are treatment with anthracyclines, or may present with progressive dyspnoea and respiratory failure
    mitoxantrone and/​or radiotherapy to the heart region, and younger due to pulmonary fibrosis, and is usually a consequence of either
    age at treatment. However, there is an individual variability for heart chemotherapy (particularly busulfan, carmustine, or lomustine) or
    failure which may be caused by genetic susceptibility, although it is radiotherapy that includes large volumes of normal lung, which is
    too early to use genetic assessment in clinical practice. Radiotherapy less and less the case with current techniques. Obstructive lung dis-
    to the heart is a major risk factor for coronary artery disease, myo- ease occurs most commonly in survivors of HSCT and is associated
    cardial infarction, pericardial disease, and valvular abnormalities. At with chronic graft-​versus-​host (GvH) disease. Previous thoracic
    doses ≥15 Gy, the CCSS study found a two-​to six-​fold increased risk surgery may also impair lung function.
    for the above complications in CCSs compared to healthy siblings. PFTs should be performed at the completion of treatment in
    Early detection of cardiovascular disease in CCSs is important. at-​risk patients and at regular intervals thereafter in symptomatic
    Surveillance recommendations exist for early detection of heart patients or those with abnormal test results. It is advisable to rec-
    failure. Echocardiography is widely used to identify asymptomatic ommend pneumococcal and annual influenza immunization in pa-
    cardiac dysfunction before symptoms occur in survivors treated tients with established lung disease.
    with cardiotoxic chemotherapy or radiotherapy to the heart. Routine
    surveillance is recommended for survivors at high risk for cardio- Renal toxicity
    myopathy, defined as those with anthracycline exposure ≥250 mg/​ Nephrotoxicity may be due to glomerular or tubular damage, or
    m2, or ≥35 Gy chest radiotherapy, or a combination of ≥100 mg/​m2 both. Chronic kidney disease (CKD) or the requirement for dialysis
    of anthracyclines and ≥15 Gy chest radiotherapy. It may also be rea- is an uncommon but serious late complication, occurring in 0.5% of
    sonable for survivors treated with lower doses, as is the case in more survivors in the CCSS (RR 8.9). Glomerular toxicity is commonly
    recent years. Although some studies have shown elevated levels of subclinical and revealed only by a high-​serum creatinine concen-
    blood biomarkers (troponin and NT-​proBNP), monitoring of bio- tration, but may occasionally lead to CKD or, rarely, end-​stage renal
    markers during follow-​up is not recommended due to insufficient failure. Likewise, tubular damage may be subclinical, but clinic-
    evidence for added diagnostic value. For the other cardiovascular ally overt proximal tubulopathy may occur in 25–​30% of children
    diseases, no routine surveillance is recommended currently in the treated with potentially nephrotoxic chemotherapy. Chronic tubular
    follow-​up of survivors. toxicity after ifosfamide may result in hypophosphataemia and even
    Prevention of cardiovascular diseases for patients treated for hypophosphataemic rickets (HR), whilst cisplatin (and less com-
    childhood cancer is essential. Primary prevention (before and during monly carboplatin) nephrotoxicity may cause hypomagnesaemia,
    treatment) can include limiting anthracycline and radiotherapy which when severe may lead to convulsions, tetany, or cardiac ar-
    doses. Other strategies include liposomal encapsulated doxorubicin rhythmias. Ifosfamide may also rarely cause nephrogenic diabetes
    or longer infusion time of anthracyclines. However, data in children insipidus (NDI) due to distal tubular damage, or proximal or distal
    are lacking. Concomitant treatment with a cardioprotective agent, renal tubular acidosis (RTA). Renal damage may also cause hyper-
    such as dexrazoxane, has also been suggested for primary preven- tension and proteinuria in CCSs.
    tion. Randomized controlled trials (RCTs) in children and adults The main risk factors for nephrotoxicity are the specific chemo-
    showed a cardioprotective effect. One observational study suggested therapy drugs received and their doses, radiotherapy including renal
    an association between dexrazoxane and a higher risk of secondary tissue, and previous nephrectomy, whilst tumour-​related urinary
    malignancies, but these concerns have not been substantiated. tract obstruction or pre-​existing renal dysfunction may exacer-
    Secondary prevention strategies include treatment with angiotensin-​ bate renal impairment. Ifosfamide nephrotoxicity is more common
    converting-​enzyme (ACE) inhibitors and beta-​blockers, and studies in patients treated with a cumulative dose of more than 80 g/​m2.
    have suggested this may prevent or delay deterioration of cardiac The increased frequency and severity of ifosfamide nephrotox-
    function in survivors of cancer, although this remains to be con- icity in children younger than five years old at treatment reported
    firmed in CCSs. Furthermore, it is important for survivors to have a in several studies is not a universal finding, especially with longer
    healthy lifestyle and to seek early treatment of comorbidities. follow-​up. Platinum nephrotoxicity is more common in children
    treated with a higher cisplatin dose rate (more than 40 mg/​m2/​
    Pulmonary toxicity
    day) or higher cumulative carboplatin doses, and those treated at
    A recent Swiss study looked at self-​reported pulmonary disease an older age. Other cytotoxic drugs that may cause or contribute
    compared to siblings and found an increased risk of pneumonias to nephrotoxicity in children include melphalan and high-​dose
    112 CHAPTER 14 Late Effects of Therapy and Survivorship Issues

    methotrexate (both of which usually present with acute kidney in- CNS tumour survivors. The consequences of deafness in children
    jury which may not recover completely, leading to CKD), and the include delayed speech development and impaired educational/​
    nitrosoureas (carmustine, lomustine, semustine) which may cause social functioning, especially in younger children.
    CKD. In addition, supportive-​care drugs including aminoglycoside The major cause of chronic deafness in survivors of childhood
    antibiotics and amphotericin, or immunosuppressive agents such as malignancy is platinum chemotherapy, particularly cisplatin and,
    ciclosporin, may contribute to or represent a primary cause of renal less commonly, high-​dose carboplatin. Both higher cumulative
    damage. cisplatin dose and younger age at treatment predict a higher risk
    Patients with suspected or established nephrotoxicity should of deafness, with a total dose of more than 400 mg/​m2 and an age
    undergo regular surveillance including measurement of blood pres- of under five years being associated with development of bilateral
    sure (especially also for those who received radiotherapy on the sensorineural hearing loss in 40% of children. High-​dose radio-
    kidneys) and growth, urinalysis for proteinuria, and biochemical therapy to the middle ear may cause mixed sensorineural and
    monitoring. Electrolyte supplementation or specific treatment of conductive hearing loss, but clinically significant deafness is un-
    HR, RTA, and NDI may be required. common in children who have not also received platinum treat-
    ment. Hearing loss may also be exacerbated or caused by other
    Lower urinary tract toxicity ototoxic treatment (e.g. aminoglycosides). Finally, CNS surgery,
    Chronic lower urinary tract toxicity may manifest as haemor- in particular for posterior fossa or brainstem tumours, may cause
    rhagic cystitis (HC) which may follow either radiotherapy or hearing problems.
    oxazaphosphorine (cyclophosphamide or ifosfamide) chemo- Children who may have hearing loss should be assessed by pure-​
    therapy. Although most acute episodes of HC resolve fully, a few tone audiography, or by behavioural audiometry or otoacoustic
    patients suffer from persistent or recurrent urinary symptoms, emissions (or occasionally auditory brainstem responses) in
    including frequency, dysuria, irritability and/​or urgency, incon- younger children, and referred for ear, nose, and throat (ENT) or
    tinence, and, occasionally, retention due to bladder fibrosis and audiometric/​speech-​therapy assessment if hearing loss is identified.
    dysfunction. Bladder damage due to the original malignancy or Education and community paediatric services should be alerted to
    its treatment (with surgery or radiotherapy) may result in a small-​ children with significant hearing impairment. Early evaluation, even
    volume bladder or sometimes a urethral stricture. Bladder malig- before confirmation of hearing loss, is important in infants treated
    nancy is an uncommon but well recognized late complication of HC. with cisplatin or high-​dose carboplatin.
    Urinary tract symptoms in CCSs should be evaluated by urin-
    alysis, microscopy, and culture, and referral to a urologist may be Craniofacial and dental toxicity
    required for cytoscopy or urodynamic studies. Regular urine cy- Craniofacial and dental complications are more prevalent in pa-
    tology has been recommended in patients with a history of previous tients treated at a young age. The most important risk factor is
    severe HC. radiotherapy to the face or brain, although chemotherapy may be
    an added risk factor. Radiotherapy to the growing skull and face will
    Visual toxicity lead to hypoplasia of the irradiated area; the younger the child and
    A wide variety of late ocular and visual toxicities may occur, ran- the higher the radiotherapy dose, the more pronounced the growth
    ging from dry eyes due to impaired lacrimal gland tear production, impairment. As normal growth occurs, the deformity becomes pro-
    which predisposes to corneal ulceration and scarring, to visual im- gressively more pronounced.
    pairment due to posterior subcapsular cataracts. Severe visual im- Correction of skeletal complications may be problematic due
    pairment was reported in 3% of survivors in the CCSS (RR 5.8) and to the nature of the irradiated bone which can make surgery dif-
    may also be a consequence of the underlying disease (e.g. by direct ficult. Damage to rudimentary teeth may result in the absence of
    invasion of optic structures or by intracranial hypertension) in sur- permanent teeth, whilst that occurring later in dental development
    vivors of CNS tumours, 13% of whom reported legal blindness in may result in small, brittle teeth. Damage to dental roots may lead
    one or both eyes in the CCSS. Rarer adverse effects include keratitis, to tooth loss, and caries is accelerated by xerostomia from radiation
    uveitis, chorioretinopathy, and optic neuritis. effect on salivary glands.
    Lacrimal gland dysfunction and cataract are related to radio- Due to these potential dental problems, preventive measures are
    therapy including the eye or the use of TBI in children with haem- very important when receiving radiotherapy and/​or chemotherapy,
    atological malignancies, up to a third of whom may develop later and guidelines for this have been published.
    cataracts. Prolonged steroid treatment may also lead to cataract
    formation even in the absence of radiotherapy. Very rarely, chemo- Gastrointestinal and hepatic toxicity
    therapy may cause severe visual toxicity (e.g. visual loss related to Surgery (including permanent stomas) and radiotherapy to the
    high-​dose fludarabine). gastrointestinal tract may cause late complications including stric-
    Treatment of ocular/​visual toxicities includes artificial tear re- tures, malabsorption, and vitamin B12 deficiency. Large radiotherapy
    placement for dry eyes and surgical removal of the lens affected by volumes and/​or high doses are associated with a higher incidence of
    cataract (with insertion of a lens implant), but established visual loss gastrointestinal problems. Hepatitis due to viral infection after blood
    is usually irreversible. transfusion remains a problem, although the frequency is now much
    lower due to better blood-​product surveillance. Long-​term follow-​
    Auditory toxicity up (LTFU) for those at risk should include liver function tests and
    Chronic hearing loss not corrected by a hearing aid was noted in 2% examination for signs of liver disease, such as hepatosplenomegaly
    of CCSS survivors (RR 6.3), and any hearing impairment in 12% of or jaundice.
    Survivors of haematopoietic stem-cell transplantation 113

    Skin toxicity DXA examination should be performed two years after treat-
    Late skin complications after treatment for childhood cancer in- ment completion and at the end of puberty in at-​risk survivors.
    clude non-​melanomatous skin cancer (NMSC) (most commonly Vitamin D supplementation and appropriate dietary calcium intake
    basal cell carcinoma, but also squamous cell carcinoma), melanoma, should be advised. After puberty, and in severe cases, treatment with
    alopecia and sparse/​patchy hair, scarring, atrophy, and pigmentation bisphosphonates may also be considered.
    disorders. Although rarely fatal, NMSC can cause substantial cos- Avascular necrosis of bone may occur either during or after
    metic problems if not treated in a timely fashion. These tumours also therapy, particularly with radiotherapy and/​or prolonged steroid
    occur at a much younger age in CCSs than in the rest of the popula- treatment, and usually effects joints in long bones, causing pain and
    tion, sometimes even below the age of ten. functional impairment.
    Radiotherapy is the main culprit (RR 4–​6); 90% of these tumours
    occur within the irradiated skin area and young age at treatment
    confers increased risk. Higher radiotherapy dose is a risk factor for
    Immunological dysfunction
    malignant melanoma, as is treatment with alkylating agents and
    vinca alkaloids. Chemotherapy may also increase the frequency Immunological function usually recovers satisfactorily within
    of pigmented naevi which may be precursors to a more malignant 6–​12 months of standard chemotherapy but is slower after HSCT
    transformation. Avoiding excess ultraviolet (UV) radiation and (possibly up to 18 months after mismatched or unrelated donor
    careful examination of irradiated areas is of vital importance during allogeneic transplants), especially in the presence of chronic GvH
    follow-​up. disease.
    Patients who have undergone splenectomy or received high-​dose
    Skeletal toxicity splenic radiotherapy (more than 40Gy), and survivors of HSCT
    Bone growth involves both linear growth and increase in width and (especially those with functional hyposplenism following condi-
    density. The process is controlled by a complex interaction between tioning with TBI), are at increased lifelong risk of potentially life-​
    many endocrine signals including GH, IGF-​1, thyroid hormone, threatening encapsulated bacterial infections. These individuals
    oestrogens, androgens, glucocorticoids, parathyroid hormone, and should be immunized with conjugate haemophilus influenzae type b
    vitamin D, and other factors such as nutrition and physical activity. (Hib), meningococcal and pneumococcal vaccines, and given anti-
    In cancer patients, disease itself or reduced physical activity or in- biotic prophylaxis with phenoxymethylpenicillin or amoxicillin.
    adequate nutrition, as well as radiotherapy or chemotherapy, may
    interfere with normal bone metabolism.
    The most evident effect on linear growth is that of radiotherapy
    Survivors of haematopoietic
    when given before bone maturation has completed. Bone irradi- stem-​cell transplantation
    ation of more than 20 Gy (although there is no consensus on a clear
    threshold), especially when long bone growth plates are included, Survivors of HSCT are at particularly high risk of LAEs, with more
    leads to reduced bone growth and potentially to asymmetric limb than 90% suffering from at least one and more than 70% from at least
    growth. If the spine is involved, vertebral bodies will display im- three. Patients conditioned with TBI are at the highest risk of late
    paired growth leading to a reduced final height with disproportion toxicity, and high-​dose conditioning chemotherapy is an additional
    between standing and sitting height. Scoliosis may result from soft-​ potent risk factor. Additive and potentially synergistic damage re-
    tissue fibrosis secondary to radiotherapy to paravertebral tissues/​or- sults from numerous other factors, including previous treatment
    gans and surgery in that area. Finally, facial asymmetry may result given before transplant, the development of other serious compli-
    from radiotherapy including the facial bones. cations after HSCT, potentially toxic supportive-​care drugs, and es-
    It is vital to maximize the peak bone mineral density (BMD), pecially chronic GvH disease which may affect any organ, tissue, or
    which normally occurs on completion of puberty, with subse- body system.
    quent physiological decay over decades. Dual X-​ray absorptiometry It is important to undertake regular and detailed surveillance in
    (DXA) is used to assess BMD. view of the wide range, frequency, and severity of potential compli-
    CRT may indirectly affect BMD by interfering with GH and sex cations. For example, SMNs are reported in up to 10–​15% of HSCT
    hormone secretion; in addition, prolonged steroid treatment and survivors 15 years post HSCT (skin, oral, thyroid, and CNS tumours
    antimetabolites may have a dose-​dependent adverse effect on BMD. and post-​transplant lymphoproliferative disease being amongst the
    Survivors of HSCT are at increased risk of skeletal late complica- commonest diagnoses), whilst reproductive toxicity with germ-​cell
    tions due to the multiple effects of TBI and high-​dose chemotherapy failure (leading to azoospermia in males and ovarian failure in fe-
    on the neuroendocrine and gonadal systems, and the effects of im- males) is virtually inevitable after TBI and occurs in at least 90%
    paired physical activity. After treatment discontinuation, recovery after busulfan/​cyclophosphamide.
    usually occurs but cumulative prednisone-​equivalent doses of more Chronic GvH disease occurs in up to 30% of HSCT patients, with
    than 9 g/​m2 or methotrexate doses of more than 40 g/​m2 are asso- multiple potential sequelae including organ and tissue damage (e.g.
    ciated with a higher risk of failure to regain normal BMD. Finally, pulmonary, gastrointestinal, hepatic, renal, musculoskeletal, skin,
    bone metabolism may be indirectly affected in case of urinary cal- serosal effusion), as well as functional impairment (e.g. immune-​
    cium and/​or phosphate wasting in patients with ifosfamide-​induced mediated cytopenias, delayed immune reconstitution) and the po-
    tubulopathy. tential adverse effects of immunosuppressive drugs.
    114 CHAPTER 14 Late Effects of Therapy and Survivorship Issues

    clinical trials are including quality of survival as one of the main


    Survivors of central nervous system tumours
    outcome measures.
    The management of CNS tumour survivors requires LTFU, with
    LAEs are reported with higher frequency and usually higher severity
    appropriate imaging and a multidisciplinary approach including spe-
    among survivors of CNS tumours as compared to CCSs from other
    cialists not always involved with other childhood solid tumours (e.g.
    tumour types. The reported frequencies vary between studies due
    (paediatric) endocrinologists, ophthalmologists, ENT specialists,
    to methodological differences and length of follow-​up. Neurological
    neuropsychologists, social workers). The transition to adult life may
    and neuropsychological toxicity, hypothalamic–​pituitary dysfunc-
    require special facilities and collaborations with adult specialists, as
    tion, and visual and auditory toxicity can occur after treatment of
    already mentioned, and the adoption of rehabilitation programmes
    other tumour types, but also occur frequently in survivors of CNS
    or specific support measures through adult life. Survivors of CNS tu-
    tumours. They have already been discussed earlier in this chapter;
    mours may require additional help in finding an appropriate job and
    this section deals with some other toxicities and deficits that are
    for independent living or support to live semi-​independently. The
    more specific in survivors of CNS tumours.
    way social and governmental services deal with these problems may
    More than one functional deficit often occurs in the same survivor
    differ from country to country, but all specialists involved should be
    and these may be due either to the tumour itself (e.g. seizures, focal
    aware of the possibilities that are available.
    nerve palsies, hemiparesis, hemiplegia, visual impairment, cerebellar
    dysfunction, endocrine deficits) or to treatment. Surgery may cause
    cerebellar mutism and posterior fossa syndrome (usually regressing Surgical complications
    in the months after surgery), shunting sequelae, stroke, oculo-
    motor deficits, ataxia, vision and hearing loss, and hypothalamic Surgical complications and late physical effects may be obvious, and
    damage. Radiotherapy may lead to stroke, leucoencephalopathy, although the quality of life of these patients in domains other than
    neurocognitive impairment, radiation necrosis, endocrinopathies, physical functioning may be rated as high as that in controls, there
    and secondary tumours. Chemotherapy may be associated with is a lifelong need for many patients to have continued contact with
    ototoxicity, peripheral neuropathy, and gonadotoxicity. Recently health services (e.g. for prosthetic reasons after limb-​salvage surgery
    introduced novel biological agents (e.g. bevacizumab and mTOR or for post-​surgical complications). Other complications vary in lo-
    inhibitors) have also been associated with complications such cation and may be subtle, depending on the anatomy and physiology
    as chronic hypertension and amenorrhoea, and their long-​term of the affected part of the body. Spinal and neurological surgery may
    toxicity (currently only partially known about) requires further lead to complications affecting the central and peripheral nervous
    investigation. and musculoskeletal systems. Thoracic surgery may affect cardiac
    Age at treatment and irradiation remain the most relevant in- and pulmonary function. Abdominal surgery may affect the gastro-
    dependent risk factors for neurocognitive, endocrine, and hearing intestinal system (adhesions, strictures, and obstruction), liver, or
    changes. Late neurological impairment can also occur because of tu- kidneys, and this risk is, as for most of the surgical complications, in-
    mour recurrence, secondary CNS tumours, or bleeding of cavernous creased with additional radiotherapy. Surgery to the head and neck
    angiomas (cavernomas). may lead to facial deformities and thyroid disease.
    Neurocognitive and neuropsychological impairment has been re- Although surgeons always try to avoid mutilating or functional
    ported, with frequencies ranging between 8% and 57% depending damage, this may be unavoidable for some diagnoses and tumour
    on the measured skill. Endocrine late effects have been reported at locations when cure is the ultimate aim. Late complications that may
    frequencies ranging between 43% and 61%, and in up to 95% of chil- require surgical intervention later on include prosthetic surgery for
    dren treated for a midline tumour. Other complications are reported osteonecrosis of major joints and the occurrence of SMNs.
    with different frequencies: seizures in 27% of subjects, motor deficits
    in 20–​35%, stroke in 29.3% of irradiated cases, social impairment in
    50%, hearing loss in 9–​14%, visual changes in up to 40%, and SMNs Psychosocial outcomes
    in 10.7% of cases at 25 years from diagnosis. As mentioned earlier,
    the tumour itself may also be responsible for these problems, next In addition to medical complications, CCSs may also suffer psy-
    to the consequences of treatment. Patients and parents should be chological or social consequences. Survivors may go through con-
    given a fair appraisal of all factors involved, including to what extent trasting experiences: on the one hand, enhancing their appreciation
    recovery may be expected or not. of life (post-​traumatic growth) by promoting maturity and resili-
    Cancer predisposition syndromes are associated with a higher ence, while at the same time increasing psychological vulnerability
    frequency of CNS tumours, which occur in 15% of patients with because of the previous cancer history or the presence of chronic
    these syndromes. This is particularly the case in infants with health conditions. While many survivors fare well, a considerable
    medulloblastoma (20% of cases with Gorlin syndrome) and in chil- number report psychological distress such as depression or anx-
    dren with choroid plexus tumours (40% of cases with Li Fraumeni iety, post-​traumatic stress symptoms, and fatigue. Psychological and
    syndrome). Furthermore, patients with these syndromes have an social problems are often interrelated: distress may lead to problems
    increased risk of chemotherapy-​or radiotherapy-​ related organ in school, employment, or relationships, and problems in the social
    damage or SMNs, and should be followed up carefully. field may, in return, influence psychological well-​being. The provi-
    Strategies to improve the quality of survival of CNS tumour sion of medical, psychological, and social care to survivors in need is
    survivors and to prevent LAEs as far as possible are mandatory in therefore paramount. The term ‘Damocles syndrome’ has been used
    planning treatment for these patients, and an increasing number of to illustrate their psychological condition of aiming to achieve full
    The future 115

    integration as active members of the society whilst living with the


    fear that possible treatment complications or cancer recurrence may FURTHER READING
    interfere with their aspirations. Ajithkumar T, et al. (2017) Prevention of radiotherapy-​ induced
    It is important that the general public recognizes and accepts the neurocognitive dysfunction in survivors of paediatric brain tu-
    reality of cure of childhood cancer, and ensures that survivors have mours: the potential role of modern imaging and radiotherapy
    equal access to education, jobs, insurance, and medical care. techniques. Lancet Oncol 18, e91–​100.
    Allodji RS, et al. (2019) Risk of subsequent primary leukaemias among
    69,460 five-year survivors of childhood cancer diagnosed from
    1940 to 2008 in Europe: A cohort study within PanCareSurFup. Eur
    Follow-​up care
    J Cancer 117, 71–83.
    Armstrong GT (2009) Long-​term survivors of childhood central ner-
    LTFU of CCSs is essential to identify and treat LAEs early, increase vous system malignancies: the experience of the Childhood Cancer
    knowledge about specific groups at higher risk of long-​term mor- Survivor Study. Eur J Paediatr Neurol 14, 298–​303.
    bidity and mortality, and improve patient care. Research indicates Armstrong GT, et al. (2016) Reduction in late mortality among 5-​year
    that chronic LAEs increase in frequency and severity with time, and survivors of childhood cancer. N Engl J Med 374, 833–​42.
    may interact adversely with the normal ageing process, resulting in Bass JK, et al. (2016) Hearing loss in patients who received cranial ra-
    increasing and clinically significant impairment of vital organ sys- diation therapy for childhood cancer. J Clin Oncol 34(11), 1248–​55.
    tems during adulthood at a younger age than normal and an in- doi: 10.1200/​JCO.2015.63.6738
    creased risk of premature major illness or death. This highlights the Bhakta N, et al. (2016) Cumulative burden of cardiovascular morbidity
    need to develop effective models of LTFU care which will optimize among pediatric, adolescent and young adult Hodgkin lymphoma
    efficient use of health-​service resources to deliver maximal benefit survivors: an analysis from the St Jude Lifetime Cohort Study.
    to survivors. Lancet Oncol 17, 1325–​34. doi:10.1016/​S1470-​2045(16)30215-​7.
    Bitsko MJ, et al. (2016) Psychosocial late effects in pediatric cancer sur-
    LTFU should be lifelong and transition programmes should be
    vivors: a report from the Children's Oncology Group. Pediatr Blood
    organized to successfully transfer survivors from paediatric to
    Cancer 63, 337–​43.
    adult healthcare. Furthermore, care should be based on surveil-
    Bright CJ, et al. (2018) Risk of Soft-Tissue Sarcoma Among 69 460
    lance guidelines. The IGHG (www.ighg.org) is a collaboration of Five-Year Survivors of Childhood Cancer in Europe. J Natl Cancer
    paediatric late-​effect guideline groups, the PanCare network (www. Inst 110, 649–660.
    pancare.eu), and other professionals to formulate harmonized Brown MC, et al. (2015) The views of European clinicians on guide-
    recommendations. lines for long-​ term follow-​ up of childhood cancer survivors.
    Each CCS should be provided with a paper or electronic docu- Pediatr Blood Cancer 62, 322–​8.
    ment providing information on their cancer-​related medical history, Byer L, et al. (2016) Clinical trials in pediatric neuro-​oncology: what is
    including treatment received, and offering risk-​adapted suggestions missing and how we can improve. CNS Oncol 5, 233–​9.
    for LTFU care which can then be tailored to each survivor, based on Clement SC, et al. (2018) Balancing the benefits and harms of thy-
    the type and dose of treatment received, and age at treatment and at roid cancer surveillance in survivors of childhood, adolescent and
    follow-​up. Survivors should be able to decide if, when, and where young adult cancer: recommendations from the International Late
    they would like their LTFU to be conducted, and the treating centre Effects of Childhood Cancer Guideline Harmonization Group in
    collaboration with the PanCareSurFup Consortium. Cancer Treat
    should be able to respond to any queries that may arise over time.
    Rev 63, 28–​39.
    El-​Fayechh C, et al. (2017) Cerebrovascular diseases in childhood
    cancer survivors: role of the radiation dose to Willis circle arteries.
    The future Int J Radiat Oncol Biol Phys 97, 278–​86.
    Essig S, et al. (2012) Follow-​up programs for childhood cancer sur-
    Increased understanding of the pathogenesis of LAEs of treatment vivors in Europe: a questionnaire survey. PLoS One 7, e53201.
    should ultimately reduce their frequency and severity. In particular, Fidler MM, et al. (2016) Long term cause specific mortality among
    it should inform the design of newer and less toxic agents or proto- 34,489 five year survivors of childhood cancer in Great Britain: popu-
    cols and facilitate the development of protective drugs to reduce lation based cohort study. BMJ 354, i4351. doi: 10.1136/​bmj.i4351
    or prevent damage to normal tissues. Genetic risk factors may be Fidler MM, et al. (2018) Risk of subsequent bone cancers among
    identified that predict a higher risk for the development of LAEs in 69 460 five-year survivors of childhood and adolescent cancer
    individual patients, hence explaining why some patients develop se- in Europe. J Natl Cancer Inst 110, 183–194.
    vere LAEs whilst others, with apparently similar clinical risk factors, Friedman DL, et al. (2010) Subsequent neoplasms in 5-​year survivors
    do not. of childhood cancer: the Childhood Cancer Survivor Study. J Natl
    Cancer Inst 102, 1083–​95.
    Finally, newer agents may have as yet unknown LAEs, so clinical
    Garwicz S, et al. (2012) Late and very late mortality in 5-​year sur-
    research is vital to document the nature of and risk factors for such
    vivors of childhood cancer: changing pattern over four decades—​
    toxicities. In particular, it is important that phase III randomized experience from the Nordic countries. Int J Cancer 131, 1659–​66.
    trials should evaluate LAEs of treatment. The knowledge gained Gawade PL, et al. (2014) A systematic review of dental late effects in
    from LTFU of CCSs should be used not only to facilitate design of survivors of childhood cancer. Pediatr Blood Cancer 61, 407–​16.
    new treatment protocols but also to provide important information Geenen MM, et al. (2007) Medical assessment of adverse health out-
    for survivors and their families to enable them to become advocates comes in long-​term survivors of childhood cancer. JAMA 297,
    for their own health. 2705–​15.
    116 CHAPTER 14 Late Effects of Therapy and Survivorship Issues

    Gibson TM, Robison LL. (2015) Impact of cancer therapy-​related ex- Pierson C, et al. (2016) A meta-​analysis of the neuropsychological ef-
    posures on late mortality in childhood cancer survivors. Chem Res fects of chemotherapy in the treatment of childhood cancer. Pediatr
    Toxicol 28, 31–​7. Blood Cancer 63, 1998–​2003.
    Glenny AM, et al. (2010) The development of evidence-​based guide- Reulen RC, et al. (2010) Long-​term cause-​specific mortality among
    lines on mouth care for children, teenagers and young adults treated survivors of childhood cancer. JAMA 304, 172–​9.
    for cancer. Eur J Cancer 46, 1399–​412. Roth CL (2015) Hypothalamic obesity in craniopharyngioma pa-
    Gunn ME, et al. (2016). Use of endocrinological and neurological tients: disturbed energy homeostasis related to extent of hypothal-
    medication among 5-​year survivors of young onset brain tumours. amic damage and its implication for obesity intervention. J Clin
    J Neurooncol 128, 473–​9. Med 4(9), 1774–​97. doi: 10.3390/​jcm4091774
    Henderson TO, Nathan PC. (2015) Childhood cancer survivors: con- Shah SS, et al. (2015) Long-​term psychiatric outcomes in pediatric
    siderations for surgeons in the transition from pediatric to adult brain tumor survivors. Childs Nerv Syst 31, 653–​63.
    care. Semin Pediatr Surg 24, 93–​9. Sharma D, et al. (2017) Need for improved skin cancer surveillance in
    Huang TT, et al. (2014) Pulmonary outcomes in survivors of childhood pediatric cancer survivors. Am J Clin Dermatol 18, 165–​8.
    central nervous system malignancies: a report from the Childhood Skinner R. (2018) Late renal toxicity of treatment for childhood
    Cancer Survivor Study. Pediatr Blood Cancer 61, 319–​25. malignancy—​risk factors, long-​term outcomes and surveillance.
    Hutchinson AD, et al. (2017) Cancer-​related cognitive impairment in Pediatr Nephrol 33(2), 215–​25. doi: 10.1007/​s00467-​017-​3662-​z
    children. Curr Opin Support Pall Care 11, 70–​5. Skinner R, Leiper AD. (2004) Bone marrow transplantation. In: WHB
    Lund LW, et al. (2011) A systematic review of studies on psychosocial Wallace, DM Green (eds) Late Effects of Childhood Cancer, pp. 304–​
    late effects of childhood cancer: structures of society and methodo- 20. London: Arnold.
    logical pitfalls may challenge the conclusions. Pediatr Blood Cancer Taiwo Z, et al. (2017) The Neurological Predictor Scale: a predictive
    56, 532–​43. tool for long-​term core cognitive outcomes in survivors of child-
    McCurdy D, et al. (2016) Associations among treatment-​ related hood brain tumors. Pediatr Blood Cancer 64, 172–​9.
    neurological risk factors and neuropsychological functioning in Tallen G, et al. (2015) Strategies to improve the quality of survival for
    survivors of childhood brain tumor. J Neurooncol 127, 137–​44. childhood brain tumours survivors. Eur J Paediatr Neurol 19, 619–​39.
    Michel G, et al. (2019) Evidence-based recommendations for the or- Tukenova M, et al. (2010) Role of cancer treatment in long-​term
    ganization of long-term follow-up care for childhood and adoles- overall and cardiovascular mortality after childhood cancer. J Clin
    cent cancer survivors: a report from the PanCareSurFup Guidelines Oncol 28, 1308–​15.
    Working Group. J Cancer Surviv 13(5), 759–772. Turcotte LM, et al. (2017) Temporal trends in treatment and subse-
    Mulder RL, et al. (2009) Prevalence and risk factors of radiation-​ quent neoplasm risk among 5-​year survivors of childhood cancer,
    induced growth hormone deficiency in childhood cancer sur- 1970–​2015. JAMA 317, 814–​24.
    vivors: a systematic review. Cancer Treat Rev 35, 616–​32. Turcotte LM, et al. (2015) Risk of subsequent neoplasms during the
    Mulrooney DA, et al. (2009) Cardiac outcomes in a cohort of adult sur- fifth and sixth decades of life in the CCSS. J Clin Oncol 33, 3568–​75.
    vivors of childhood and adolescent cancer: retrospective analysis Wells EM, et al. (2017) Longitudinal assessment of late onset neuro-
    of the Childhood Cancer Survivor Study cohort. BMJ 339, b4606. logic conditions in survivors of childhood central nervous system
    Oeffinger KC, et al. (2006) Chronic health conditions in adult sur- tumors: a Childhood Cancer Survivor Study report. Neuro Oncol
    vivors of childhood cancer. N Engl J Med 355, 1572–​82. 20(1), 132–​42. doi: 10.1093/​neuonc/​nox148
    Olsen JH, et al. (2009) Lifelong cancer incidence in 47,697 patients Whelan KF, et al. (2010) Ocular late effects in childhood and adoles-
    treated for childhood cancer in the Nordic countries. J Natl Cancer cent cancer survivors: a report from the Childhood Cancer Survivor
    Inst 101, 806–​13. Study. Pediatr Blood Cancer 54, 103–​9.
    15
    Acute Lymphoblastic Leukaemia
    Denis Schewe, Mignon Loh, Rob Pieters, and Martin Schrappe

    Epidemiology abnormalities (cytopenias, atypical cells) detected in a blood draw


    for workup of non-​specific symptoms.
    Acute lymphoblastic leukemia (ALL) is the most common type of
    cancer in childhood. It accounts for 20–​25% of all childhood can-
    cers. The peak incidence is at 2–​5 years of age, which is mainly due Diagnosis
    to a peak of ALL cases with hyperdiploid ALL and ETV6/​RUNX1
    gene-​rearranged ALL. The incidence rate is higher in high-​income The blood-​cell counts often show anaemia and/​or thrombocytopenia
    countries and in Hispanic children, and lower in black children as and granulocytopenia. The leucocyte count is increased to more
    compared to white children. Males are affected more often than fe- than 10e9/​l in 50% of cases. Hyperleukocytosis more than 50e9/​l is
    males, except in infants. seen in more than 50% of infants, 15–​20% of children, and 20–​30%
    The aetiology of ALL is characterized by the fact that imma- of adolescents, and occurs in T-​ALL more frequently. Leukocyte
    ture lymphoid cells acquire different consecutive genetic lesions counts may also be normal or even leukopenic. Morphological
    of which the first lesion can occur before birth. For example, the examination of peripheral blood smears shows lymphoblasts in the
    most common translocation that occurs in childhood ALL, the majority, but not all, cases.
    ETV6/​RUNX1 translocation, can be traced back in the Guthrie Nowadays, the diagnosis of ALL is made by a combination of
    cards of the patients. However, only few children born with this morphology, immunophenotyping, cytogenetics, and molecular
    translocation will develop ALL, suggesting that leukemogen- genetic techniques in bone marrow and peripheral blood samples.
    esis is multifactorial and depends on multiple consecutive gen- Even though the French–​American–​British (FAB) classification has
    etic events. Less than a few percent of children with ALL have an become historical due to a lack of information regarding treatment
    underlying genetic predisposition, such as Down syndrome, even decisions or prognosis, classic cytomorphology still plays an im-
    though germline variants in certain driver genes such as PAX5 or portant role for the differential diagnosis with other diseases such
    p53 can lead to leukaemia development. Leukaemia initiation by as myelodysplastic syndrome, aplastic anaemia, or cytopenias due
    environmental exposures or common infectious pathogens is fre- to infections. Although most protocols still require the classic cut-​
    quently hypothesized but cannot yet be confirmed based on the off level of 25% or greater lymphoblasts in the bone marrow for the
    available data. diagnosis of ALL, lower percentages will be accepted in some clin-
    ical trials if this is confirmed by immunophenotyping or the pres-
    ence of typical genetic abnormalities.
    Clinical presentation Immunophenotyping by flow cytometry is considered standard
    of care and is used to determine the type of ALL (B-​lineage or T-​
    Most symptoms at diagnosis are secondary to leukaemic infiltration lineage) and the differentiation stage of the leukaemic cells. B-​cell
    of the bone marrow: pallor, petechiae, and haematomas. General precursor ALL is usually CD79a+, CD10+, or CD19+, and often
    symptoms such as fatigue and fever often occur. Bone or muscu- CD22+ and CD20+. Mature B-​cell ALL cases are usually CD19+,
    loskeletal pain is sometimes misinterpreted as a rheumatological CD20+, CD22+, or CD25+, and express surface immunoglobulin.
    condition. Refusal to walk in young children is always an alarming Co-​ expression of myeloid markers (e. g. CD13 and CD33) is
    symptom. Physical examination may show hepatosplenomegaly and common. T-​lineage ALL is positive for CDs 2, 3, 4, 5, 7, and 8, un-
    lymphadenopathy. In rare cases, especially in T-​ALL, dyspnoea or less they express a CD5dim, CD8-​profile typical for early T-​cell pre-
    signs of vena cava superior compression can be found due to the cursor ALL (ETP-​ALL). Genetic or fluorescence in situ hybridization
    presence of an anterior mediastinal mass. Other rare symptoms are (FISH) techniques are used for the detection of recurrent chromo-
    one-​sided testicular enlargement, headaches, or focal neurological somal or molecular abnormalities, the most common of which are
    abnormalities due to leukaemic involvement of the testes or the cen- the ETV6/​RUNX1 translocation in 20–​25%, hyperdiploidy in 20–​
    tral nervous system. Often, ALL is diagnosed upon peripheral blood 25%, rearrangements involving the KMT2A (MLL) gene in 5%, and
    118 CHAPTER 15 Acute Lymphoblastic Leukaemia

    the BCR/​ABL1 and TCF3/​PBX1 translocations, each in 3–​4% of Table 15.1 Most important prognostic factors currently used
    BCP-​ALL cases. for therapy stratification (prevalence in % of all cases)
    Recently, the molecular features of TCF3/​HLF-​positive BCP-​
    Risk factor Favourable Less favourable
    ALL have been described, which occurs in less than 1% of the cases
    Gender Female (~45%) Male (~55%)
    and is a disease that is almost always fatal. Roughly 2% of patients
    have hypodiploid BCP-​ALL, encompassing near-​haploid (24–​31 White blood-​cell Low (e.g. <50e9/​L) High (e.g. >50e9/​L)
    count (~80%) (~20%)
    chromosomes) and low-​hypodiploid (32–​39 chromosomes) cases.
    The former is characterized by alterations in IKZF3 while the latter Age 1–​9.99 years (~75%) <1 year (~4%)
    15–​18 years (~7%)
    very frequently harbours TP53, RB1,and IKZF2 mutations. In low-​
    Central nervous CNS1 (~75%) CNS3 (~3–​5%)
    hypodiploid ALL, the TP53 mutations are germline in nearly half
    system TLP–​ (~90%) TLP+ (~10%)
    of patients, indicative of Li Fraumeni syndrome. Both entities can
    Immunophenotype Common/​preB ALL ProB ALL (~4%)
    have activated RAS and PI3K signalling and are associated with a (~80%) T-​ALL (~15%)
    very poor prognosis. Ten to fifteen percent of children with BCP-​
    Genetic ETV6/​RUNX1 (~25%) KMT2A
    ALL suffer from ‘BCR/​ABL-​like’ or ‘Ph-​like’ ALL, which has a gene-​ abnormalities Hyperdiploidy (~25%) translocations (~5%)
    expression profile similar to BCR/​ABL1-​positive ALL but lacks the E2A/​PBX1 (~3%) BCR/​ABL (~3%)
    BCR/​ABL1 fusion. This disease entity is characterized by deletions Hypodiploidy (1%)
    TCF3/​HLF (<1%)
    in B-​cell development genes such as IKZF1, TCF3, or PAX5, and IKZF del (~12%)
    has been associated with a poor prognosis. However, when therapy BCR-​ABL-​like (~13%)
    was stratified based on minimal residual disease (MRD), these pa- Response to therapy
    tients had survival rates comparable to other BCP-​ALL patients in a
    Day 8 prednisone <1×10e9 blasts/​l (~90%) ≥1×10e9 blasts/​l (~10%)
    small cohort and with the use of allogeneic haematopoetic stem cell response in
    transplantation. peripheral blood
    ‘BCR/​ABL1-​like’ patients partly harbour ABL1-​class transloca- Day 15 bone M1 marrow <5% blasts M3 marrow >25% blasts
    tions which are targetable. The therapeutic targeting of JAK2 and marrow response (~60%) (~15%)
    CRLF2 abnormalities in ‘BCR/​ABL1-​like’ patients is the subject of Morphological Yes (~98%) No (~2%)
    further investigation. Recurrent chromosomal or molecular ab- complete remission
    after induction
    normalities in T-​ALL such as TAL/​LMO, TLX1/​HOX11, TLX3/​ therapy
    HOX11L2, and HOXA are usually present in only small patient sub-
    Bone marrow <10e-​4 after 5 weeks ≥10e-​3 after 12 weeks
    groups, and their effect on patient prognosis is less pronounced. minimal residual (~25–​40%) (~5%)
    Central nervous system (CNS) involvement is present in less than disease by PCR
    5% of children at the onset of ALL. It is diagnosed when more than Bone marrow <0.1% at day 15 of ≥10% at day 15, or
    five WBC/​µL cerebrospinal fluid (CSF) are measured upon cytospin minimal residual induction, or <0.01% at ≥0.1% at the end of
    and lymphoblasts are morphologically identifiable (CNS3 status), or disease by FCM end of induction induction
    if focal neurological symptoms are present due to leukaemic infiltra-
    tion. CNS2 status is defined as the presence of leukaemic cells upon
    microscopy but ≤five WBC/​µl in the CSF. CNS1 means that nei-
    ther leukaemic cells nor leukocytes are detected. Traumatic lumbar therapy which is defined as less than 5% blasts in the bone marrow
    puncture (TLP) potentially leads to leukaemic contamination of the and the absence of other evidence of leukaemia. Measurement of
    CSF and, similar to CNS3 status, has been associated with an inferior minimal residual disease (MRD) is incorporated as a major stratifi-
    prognosis. In the case of TLP, some protocols advocate extra intra- cation factor in many protocols. MRD is measured by flow cytometry
    thecal chemotherapy. or by polymerase chain reaction (PCR) at the end of induction and
    during the first months of post-​induction therapy.

    Prognostic factors and risk classification


    Treatment
    Many clinical and biological factors predict patient outcome.
    However, many of these factors are related to or dependent on each The backbone of every ALL treatment protocol consists of four
    other, so that proper analyses establishing the independent and phases: induction, consolidation, intensification (re-​induction), and
    quantitative values of each factor are challenging. Moreover, therapy maintenance (continuation) therapy. All treatment phases contain
    is often adjusted based on these factors influencing the contribution CNS-​directed therapy elements. Only a few patients receive CNS
    of a factor on outcome. Table 15.1 presents an overview of clinical irradiation or allogeneic haematopoietic stem-​cell transplantation
    and biological risk factors currently used for treatment stratification (SCT) in addition to chemotherapy. Current treatment protocols re-
    in paediatric ALL protocols. sult in five-​year event-​free survival (EFS) rates of up to 85% and five
    In most protocols, early response to therapy is used for therapy year overall survival rates of up to 90%. However, relapses and sec-
    stratification. Many European protocols use the absolute peripheral ondary malignancies may also occur more than five years after diag-
    blast count after one week of systemic prednisone and one dose of nosis. Also, a significant number of relapsed patients may be salvaged
    intrathecal methotrexate. A classic parameter still used in all proto- for many years before they ultimately die. Therefore, the final EFS and
    cols is the achievement of morphological response after induction survival rate are lower than the five-​year figures usually presented.
    Treatment 119

    Induction levels and better CNS penetration or to the fact that despite equiva-
    The goal of induction therapy is to achieve complete morphological lent doses, dexamethasone is still more potent than prednisone, as
    remission and restore normal haematopoiesis. This is achieved after some in vitro studies suggest. Unfortunately, dexamethasone also
    4–​6 weeks of chemotherapy in about 98% of all children. Almost leads to more side effects. The largest study thus far examining
    all protocols use a glucocorticoid (prednisone or dexamethasone), this issue found that dexamethasone was superior to prednisone
    vincristine, and L-​asparaginase as a so-​called three-​drug backbone regarding five-​year EFS rates, mainly due to improved prevention
    plus intrathecal chemotherapy for this course. The addition of an of extramedullary relapses, and a particularly good outcome of
    anthracycline as a fourth systemic drug is a matter of debate. Some dexamethasone-​treated T-​ALL patients with good response to the
    studies have made clear that at least about 50% of children can be prednisone pre-​phase. However, this was at the expense of a signifi-
    cured without the use of anthracyclines, but definitive answers to cantly higher rate of induction-​related deaths with dexamethasone.
    this question require further evidence from clinical trials. About 2% Reduction of therapy intensity
    of patients do not achieve remission after induction therapy, which
    may be due to early death often caused by infection or bleeding or to Measurements of MRD during the first months of ALL treatment
    chemotherapy-​resistant leukaemic cells. In this latter group, very in- predict patient outcome. Five-​year overall survival rates for MRD-​
    tensive treatment can result in improved survival and about 30–​40% negative patients come close to 100%, which raised the question
    long-​term survivorship. of whether therapy intensity could safely be reduced in order to
    Induction therapy involves therapy with asparaginase. Several lower toxicity without jeopardizing outcome. A non-​randomized
    studies have shown that patient outcome with low L-​asparaginase trial recently showed that reduced-​intensity delayed intensifica-
    doses or suboptimal L-​asparaginase serum levels are unfavour- tion was feasible in patients with undetectable MRD at the end of
    able. L-​asparaginase derived from E. coli has a longer half-​life than induction. It could also be demonstrated that in patients with un-
    the preparation derived from Erwinia chrysanthemi (Erwinase), detectable MRD levels, one or two courses of delayed intensification
    which can result in differences in asparagine depletion. Erwinase therapy were equally effective, with lower rates of toxicity in patients
    is usually reserved for patients intolerant to these preparations or receiving only one course. It is important to consider that the low
    showing silent inactivation of L-​asparaginase due to neutralizing number of relapses occurring in MRD-​negative patients can be sal-
    antibodies. PEGylated L-​asparaginase (PEG-​asparaginase) is less vaged in a vast majority of the cases, especially if the patient has been
    immunogenic, results in longer periods of asparagine depletion due exposed to only low levels of chemotherapy during initial treatment.
    to its longer half-​life, and has a comparable toxicity profile to other It may thus be an alternative to accept slightly higher relapse rates in
    L-​asparaginase preparations. The drug can be conveniently admin- a few standard-​risk patients, whilst sparing substantial toxicities for
    istered once every 2–​4 weeks, instead of once every 2–​3 days, to en- the majority, by reducing the intensity of chemotherapy protocols.
    sure complete continuous asparagine depletion. Clinical studies on treatment reduction in the MRD-​negative risk
    group are currently ongoing in various settings.
    Consolidation and intensification Maintenance
    Consolidation and intensification courses aim at eradicating re- The aim of maintenance treatment is to further reduce residual dis-
    sidual leukaemic cells. Berlin–​Frankfurt–​Munster (BFM) protocols ease and to suppress the outgrowth of these cells. The importance of
    use a post-​induction course consisting of low-​dose cytarabine, 6-​MP, maintenance therapy is illustrated by the fact that outcome depends
    and cyclophosphamide, which reduces the MRD load significantly. on the intensity and duration of therapy. Most protocols continue
    Other groups use different combinations of drugs in this phase with the maintenance for up to two years after diagnosis or achievement
    the same results. Many study groups use high-​dose methotrexate of morphological remission. Other protocols include a time period
    (MTX) and 6-​mercaptopurine (6-​MP) in combination with frequent of two years’ maintenance treatment for girls and three years’ main-
    intrathecal therapy in the consolidation phase. The use of high-​dose tenance for boys. A meta-​analysis showed that longer maintenance
    MTX significantly reduces the risk of bone marrow relapse, in par- leads to lower relapse rates but also increased toxic death rates.
    ticular if combined with dexamethasone. Reduction of the duration of maintenance to less than two years
    Drugs used in intensification courses are the same as in induc- led to an increased relapse risk but also showed that a significant
    tion therapy. Several randomized studies have proven the value of proportion of patients survive without two years of maintenance. It
    intensification treatment. Intensified and double-​delayed intensifi- is yet completely unclear how to identify patients who profit from
    cation improves outcome for patients with a slow initial response longer phases of maintenance or in whom therapy can be shortened
    to treatment. In patients with rapid early response, a more intensive without any risk.
    (but not a longer) intensification treatment was of benefit. Omission The backbone of every maintenance course is weekly (oral or
    of the intensification led to a significant rise in the relapse rate, al- intravenous) methotrexate and daily oral 6-​MP. Doses of these drugs
    though the same study showed that 50% of patients could be cured are usually adapted according to leukocyte count or to lymphocyte,
    without the use of intensification. neutrophil, and platelet counts. At an international expert meeting,
    consensus was reached to use a target leukocyte count of 1.5–​3×
    Dexamethasone or prednisone? 10e9/​L. There are large intra-​and inter-​individual differences in
    Several, but not all, randomized studies have shown that the use of MTX and 6-​MP doses that are tolerated or required to achieve the
    dexamethasone at a dose of 6 mg/​m2 results in a lower rate of bone target leukocyte count. These may be due to infections or underlying
    marrow and CNS relapses than prednisone at a dose of 40 mg/​m2. pharmacogenetic factors such as polymorphisms in the thiopurine
    This benefit of dexamethasone might be due to higher free plasma methyltransferase gene. Maintaining the highest dose of 6-​MP and
    120 CHAPTER 15 Acute Lymphoblastic Leukaemia

    MTX leads to a better outcome. Treatment adherence of doctors and remission after induction therapy may also have a better outcome
    patients (especially adolescents) therefore has an impact on out- with alloHSCT, especially if they have T-​ALL. As a stem-​cell source,
    come. 6-​MP remains the drug of choice in maintenance as patients matched sibling or unrelated donors perform equally well in chil-
    receiving 6-​thioguanine (6-​TG) may more frequently suffer from dren with ALL. Given the serious acute and late risks of alloHSCT,
    liver toxicity, with features of veno-​occlusive disease. its role in first-​line treatment for childhood ALL is currently limited
    The addition of steroid and vincristine pulses to maintenance is and offered to less than 5% of patients.
    a matter of debate, and data from randomized studies and meta-​
    Immunotherapy
    analyses are conflicting. Taken together, the use of pulses in main-
    tenance therapy is probably not beneficial but may contribute to a Antibody-​based and cellular immunotherapy is increasingly im-
    better outcome for specific risk groups or for patients treated on a plemented into current clinical study protocols in childhood BCP-​
    protocol without an intensification course. ALL. CD19 and CD22, but also CD20, have been used as target
    antigens in preclinical and, partly, in clinical phase I/​ II trials.
    Central nervous system directed therapy These include naked antibodies modified for improved effector-​
    CNS-​directed therapy is a prerequisite for successful treatment of cell binding, immunotoxins/​immunoconjugates, and the bispecific
    ALL. In the 1960s, more than half of the patients suffered from CNS T-​cell-​engaging antibody blinatumomab. Blinatumomab recruits
    relapses. Several treatment elements were effective in reducing this T-​cells into close proximity of CD19-​positive ALL cells, thereby
    high CNS recurrence rate to currently below 5%. First, cranial or causing cytotoxicity. It has been shown to be highly efficient in an
    craniospinal irradiation prevents CNS relapses, but has major side MRD-​positive situation in relapsed/​refractory adult ALL patients,
    effects such as long-​term neurocognitive deficits and secondary but evidence in children is scarce (Table 15.2). An alternative to
    malignant neoplasms. Because secondary neoplasms often occur blinatumomab is the use of autologous chimeric antigen receptor
    decades after radiation therapy, it is important to implement long T-​cells (CAR T-​cells) engineered to express a CD19-​specific T-​cell
    follow-​up into studies comparing the outcomes of radiation and receptor. CAR T-​cells have been found to be efficient in several
    other CNS-​directed therapies. Second, intrathecal chemotherapy trials in relapsed/​refractory ALL patients in the post-​HSCT setting,
    delivered at regular intervals reduces CNS relapse. Finally, systemic including children (Table 15.2). Both therapeutic modalities bear
    application of drugs that penetrate the blood brain barrier (dexa- the risk of acute cytokine-​release syndrome and encephalopathy, as
    methasone, high-​dose MTX, and cytarabine) all contribute to de- well as CD19-​negative disease relapse. Inotuzumab ozogamicin, an
    creased CNS relapses. anti-​CD22 antibody conjugated to calicheamicin has been shown to
    Current data favours the omission of cranial radiotherapy in pa- outperform standard therapy in adult patients with relapsed/​refrac-
    tients on contemporary treatment protocols. However, although the tory BCP-​ALL. This has, however, not been shown yet in paediatric
    majority of CNS relapses can be prevented by adequate systemic and patients, but trials are ongoing. The coming years will clarify how
    intrathecal chemotherapy, this does not exclude that specific high these therapies can further be developed so that they are most effi-
    risk groups may benefit from radiation. Most clinical protocols still cient and usable not only at relapse but also in the upfront setting.
    use radiation for a minority of high risk patients such as patients with
    a CNS3 status at diagnosis, and T-​ALL patients with a high initial leu-
    kaemia burden. Elimination of cranial radiation in the latter group Specific patient groups
    led to an increase in systemic but not CNS relapses. Most protocols
    advise extra intrathecal injections for patients with a CNS3 or CNS2 Adolescents
    status or in case of a traumatic lumbar puncture at diagnosis. Finally, The prognosis of paediatric ALL deteriorates with increasing pa-
    radiotherapy has still a role in CNS and testicular relapse. tient age. Studies in four different countries have shown that out-
    comes for adolescents with ALL from the same age are better when
    Allogeneic haematopoietic stem-​cell transplantation these patients are treated on paediatric or paediatric-​inspired ra-
    Improved outcomes with more intense chemotherapeutic regimens ther than adult protocols, with approximately 30% higher EFS. This
    has reduced the need for allogeneic haematopoietic stem-​cell trans- effect was mainly explained by differences in the dose intensity of
    plantation (alloHSCT) in childhood ALL. In addition, only a few the treatments. The paediatric protocols contained more gluco-
    studies have proven the efficacy of alloHSCT in specific high-​risk corticoids, vincristine, L-​asparaginase, MTX, and 6-​MP. Longer
    groups. A collaborative study showed that BCR/​ABL-​positive ALL delays between courses in adolescents treated according to adult
    cases benefit from alloHSCT after having received intensive chemo- protocols might also have played a role. It is possible that adult
    therapy courses. However, even in this patient group, the role of haematologists are less used to glucocorticoid-​and asparaginase-​
    alloHSCT has been debated since the introduction of ABL tyro- induced toxicities, and that delays between courses are due to the
    sine kinase inhibitors. Another collaborative analysis showed that fact that they generally treat older patients who do not tolerate in-
    alloHSCT was of no benefit for children with KMT2A-​rearranged tensive therapy as well. In a Dutch study, for instance, use of the
    ALL. A further publication made clear that only 25% of infant adult ALL treatment protocol resulted in both a significantly higher
    KMT2A-​rearranged ALL cases, namely those that carry two other relapse rate and in a significantly higher toxic death rate for adoles-
    high-​risk features (age younger that six months at diagnosis and cents due to the use of alloSCT in many patients. Adolescents and
    WBC more than 300×10e9/​L or poor prednisone response) may young adults display ALL with the ‘BCR-​ABL-​like’ gene-​expression
    benefit from alloHSCT. Some evidence suggests a benefit from signature more frequently, while ETV6/​RUNX1 and hyperdiploidy
    alloHSCT in patients with high-​ risk T-​
    ALL, but contradictory are less common, which likely leads to an inferior prognosis in this
    data exists. Some patients not achieving morphological complete population.
    Specific patient groups 121

    Table 15.2 Clinical trials using immunotherapeutic approaches (blinatumomab and CAR T-​cells) including adult and paediatric BCP-​ALL
    patients

    Citation Trial type Patient number Therapy/​Dosage Response Response duration Severe toxicities
    (adult/​paediatric) requiring
    permanent therapy
    discontinuation*
    Topp et al., Single arm 21/​0 BCP-​ALL with Blinatumomab (15 Complete 16/​21 RFS 78% @ 405 days 1/​21 (seizure)
    J Clin Oncol, 2011 phase II MRD persistence µg/​m2 × 24 h × 4 w) (MRD negativity) RFS 61% @
    33 months Topp
    et al., Blood, 2012
    Topp et al., Single arm 36/​0 BCP-​ALL, Blinatumomab (5–​30 Complete 25/​36 Median RFS 1/​36 (death, possibly
    J Clin Oncol, 2014 phase II + dose-​ relapsed/​refractory µg/​m2 × 24 h × up to Partial 2/​36 7.6 months related to treatment)
    finding cohorts (>5% blasts in bone five 4-​week cycles) MRD response 22/​36 (7.9 months if 2/​36 (other CNS)
    marrow) censored at HSCT) 1/​36 (cytokine
    Median OS release syndrome)
    9.8 months
    10/​36 long-​term
    survival >30 months,
    6 of which were
    relapse-​free Zugmaier
    et al., Blood, 2015
    Topp et al., Single arm 189/​0 BCP-​ALL, Blinatumomab (28 Complete 81/​189 Median RFS 5/​189 (death due to
    Lancet Oncol, 2015 phase II relapsed/​refractory µg/​m2 × 24 h × up MRD response 60/​73 5.9 months disease progression)
    (>10% blasts in to 57 days), dose Median OS 23/​189 (death due to
    bone marrow) reduction to 9 µg/​m2 6.1 months infection)
    × 24 h × 7 days in the 29/​189 (seizure and
    first cycle other CNS)
    Schlegel et al., Compassionate use 0/​9 BCP-​ALL, Blinatumomab (5–​15 Complete 6/​9 EFS/​OS 30% at 4/​9 (death due to
    Haematologica, programme relapsed/​refractory µg/​m2 × 24 h × up to MRD response 5/​9 243 and 398 days, disease progression
    2014 (>5% blasts in bone 18 4-​week cycles) (MRD negativity at respectively and relapse)
    marrow) one point) 2/​9 (death due to
    infection)
    Cruz et al., Phase I 6/​2 BCP-​ALL, 4 Autologous CD19. Complete 3/​8 (1 of n/​a Not reported
    Blood, 2013 adults with CLL, CAR-​VSTs, 1or 2 which is causally
    variable disease infusions uncertain)
    remission stages Partial 1/​8
    Stable disease 1/​8
    Grupp et al., Phase I 0/​2 BCP-​ALL, Autologous CTL019 Complete 2/​2 1/​2 continuous 1/​2 (severe CRS)
    N Engl J Med, 2013 relapsed/​refractory MRD response 2/​2 molecular remission 1/​2 (encephalopathy)
    in remission (<0.01% by FACS and for 9 months
    deep sequencing) 1/​2 CD19-​negative
    relapse after
    2 months
    Brentjens et al., Phase I 5/​0 BCP-​ALL, Autologous 19-​28z Complete 5/​5 4/​5 received 4/​5 (likely CRS and/​
    Science Transl Med, relapsed/​refractory, CAR MRD response 5/​ alloHSCT, 3/​4 are in or neurological side
    2013 variable disease 5 (negative by deep continuous remission effects of varying
    remission stages sequencing) for variable times degrees)
    1/​5 CD19-​positive
    relapse after 90 days
    Davila et al., Phase I 16/​0 BCP-​ALL, Autologous 19-​28z Complete 14/​16 7/​16 received 7/​16 (severe CRS)
    Science Transl Med, relapsed/​refractory, CAR MRD response 12/​ alloHSCT 2/​16 (seizure)
    2014 variable disease 16 (negative by FACS, 2/​16 (other CNS)
    remission stages PCR, and/​or deep
    sequencing)
    Lee et al., Phase I 6/​15 BCP-​ALL, Autologous Complete 14/​21 OS 51.6% at 6/​21 (severe CRS)
    Lancet, 2015 relapsed/​refractory, CD19-​CAR Stable disease 3/​21 9.7 months 1/​21 (cardiac arrest
    variable disease MRD response 12/​21 RFS of MRD-​negative with successful
    remission stages; (MRD negativity) patients 78.8% at resuscitation)
    adults were young 4.8 months 1/​21 (other CNS)
    adults <30 years 2/​12 CD19-​negative
    relapses after 3 and
    5 months
    Brudno et al., Phase I 20/​0 mixed B-​cell Allogeneic CD19-​ Complete 6/​20 CR durations 12/​20 ( grade 3 or 4
    J Clin Oncol, 2016 malignancy, mainly CAR from donor-​ Partial 2/​20 3–​30 months toxicities)
    B-​NHL, some derived T-​cells Stable disease 8/​20 SD durations No severe GVHD
    BCP-​ALL MRD response 4/​ 1–​31 months
    6 CR patients (MRD
    negativity)

    *Discontinuation in case of blinatumomab; Grade 3/​4 SAE for CAR T-​cells


    122 CHAPTER 15 Acute Lymphoblastic Leukaemia

    Infants than ten years) experience more toxicities (e.g. avascular necrosis
    Infants have an inferior outcome mainly due to the fact that around of bones, hyperglycaemia, thromboembolic complications, liver
    80% of them have an aggressive type of ALL characterized by the function disturbances) than younger patients. Late effects include
    presence of KMT2A gene rearrangements. Several specific treatment cardiomyopathy due to anthracyclines, neurocognitive deficits and
    protocols are used by the Children’s Oncology Group (COG), Japan, secondary brain tumours due to cranial radiation, and avascular
    and the Interfant Collaborative Study Group. The survival rate with necrosis of bone (although often reversible) due to glucocortic-
    current treatment is 50–​60%. Infant ALL cells have been shown to oids. An increasing body of evidence suggests that ALL treatment
    be relatively resistant to glucocorticoids and L-​asparaginase but is associated with a number of long-​term neurobehavioural and
    highly sensitive to cytarabine, which is therefore more intensively neurocognitive deficits. Late effects of ALL treatment on fertility are
    used in these protocols. Within infant ALL, all types of KMT2A-​ subject to ongoing investigations.
    rearrangements are associated with the same poor prognosis, and
    MRD diagnostics is equally important to predict prognosis as in
    childhood ALL. The younger the infant, the worse the outcome, with Perspectives
    the survival rate of congenital ALL being only 20%. AlloHSCT in
    infant ALL causes many late side effects and its efficacy is uncertain. Improvements in therapy are still being made by studying dose
    Its role may be limited to a small, very high-​risk group of infant ALL. schedules of known drugs, especially glucocorticoids and L-​
    asparaginase. Many groups study reductions (e.g. anthracycline
    BCR/​ABL-​positive acute lymphoblastic leukaemia dose) in specific risk groups (e.g. ETV6/​RUNX1 rearranged ALL),
    The translocation t(9;22) fuses the BCR gene on chromosome 22 to the while other groups have intensified therapy with L-​asparaginase
    ABL gene on chromosome 9 causing an abnormal ABL tyrosine kinase and dexamethasone.
    activity. The incidence of BCR/​ABL increases with age and is seen in Genome-​wide screening and molecular analyses have revealed
    ~3% of children with ALL but in more than 25% of adult ALL. Overall, genetic abnormalities with prognostic significance. Examples are
    children with BCR/​ABL ALL have a survival rate of about 60% with in- the discovery of the ‘BCR/​ABL-​like’ genetic subtype, IKZF1 ab-
    tensive chemotherapy and alloHSCT. Initial good responders to therapy normalities, and CRLF2 overexpression that predict a poor out-
    have a better outcome, and outcome is worse in initial poor responders. come. In the coming years, we will learn how to implement these
    In the last decade, targeted therapy of the ABL tyrosine kinase has be- markers for risk stratification and discover novel molecular
    come available. Data from two study groups suggest that the incorpor- mechanisms that underpin the high-​risk features of the disease.
    ation of imatinib into the backbone of intensive chemotherapy improves Immunotherapies with monoclonal antibodies, bispecific T-​cell-​
    patient outcome. In the near future, favourable therapy response will engaging molecules, and ‘living drugs’ (chimeric antigen recep-
    lead to a further reduction in the use of alloHSCT in these patients. tors) are exciting therapies with promising results. Considering the
    excellent survival rates currently obtained with traditional chemo-
    Relapse therapy, it is always a challenge to investigate how new drugs and
    Treatment outcome for relapsed patients mainly depends on the time immunotherapy can be incorporated in existing paediatric treat-
    and the site of relapse as well as biological characteristics. Poor prog- ment regimens in ALL.
    nostic features include a short duration of initial remission, bone
    marrow involvement at relapse, a T-​cell phenotype, the presence of
    FURTHER READING
    BCR-​ABL or KMT2A abnormalities in infancy, and, of course, MRD
    response measured after the first courses of chemotherapy. Genetic Armstrong GT, et al. (2016) Reduction in late mortality among 5-​year
    risk factors such as TP53 alterations or Ras pathway mutations can survivors of childhood cancer. N Engl J Med 374, 833–​42.
    Biondi A, et al. (2012) Imatinib after induction for treatment of chil-
    be predictive of outcome in the relapsed setting. Overall, about one
    dren and adolescents with Philadelphia-​ chromosome-​ positive
    third of patients with relapsed ALL will be cured. Paediatric patients
    acute lymphoblastic leukaemia (EsPhALL): a randomised, open-​
    with favourable characteristics (e.g. late relapse or isolated CNS re- label, intergroup study. Lancet Oncol 13, 936–​45.
    lapse) can even be cured without alloHSCT, especially those initially Boer JM, et al. (2017) Tyrosine kinase fusion genes in pediatric BCR-​
    defined as standard-​risk patients. ABL1-​like acute lymphoblastic leukemia. Oncotarget 8, 4618–​28.
    Boissel N, et al. (2003) Should adolescents with acute lymphoblastic
    leukemia be treated as old children or young adults? Comparison of
    Toxicity and late effects the French FRALLE-​93 and LALA-​94 trials. J Clin Oncol 21, 774–​80.
    Cario G, et al. (2010) Presence of the P2RY8-​CRLF2 rearrangement is
    Almost all side effects of treatment are temporary. The most im- associated with a poor prognosis in non-​high-​risk precursor B-​cell
    portant cause of toxic death are infections: 0.5% to 1.5% of patients acute lymphoblastic leukemia in children treated according to the
    ALL-​BFM 2000 protocol. Blood 115, 5393–​7.
    die from infections during induction therapy, and between 1% and
    Cheung YT, et al. (2016) Leukoencephalopathy and long-​ term
    3% while in complete remission. The most frequent drug-​specific
    neurobehavioural, neurocognitive, and brain imaging outcomes in
    toxicities include neuropathy and constipation caused by vincris- survivors of childhood acute lymphoblastic leukaemia treated with
    tine; mucositis caused by MTX; diabetes, behaviour disturbances, chemotherapy: a longitudinal analysis. Lancet Haematol 3, e456–​66.
    Cushingoid appearance, osteoporosis, and avascular necrosis of Den Boer ML, et al. (2009) A subtype of childhood acute lympho-
    bone caused by glucocorticoids; and allergic reactions, pancreatitis, blastic leukaemia with poor treatment outcome: a genome-​wide
    and thrombosis caused by asparaginase. Older patients (i.e. older classification study. Lancet Oncol 10, 125–​34.
    Perspectives 123

    Fischer U, et al. (2015) Genomics and drug profiling of fatal TCF3-​ Pieters R, et al. (2016) Successful therapy reduction and intensification
    HLF-​positive acute lymphoblastic leukemia identifies recurrent for childhood acute lymphoblastic leukemia based on minimal re-
    mutation patterns and therapeutic options. Nature Gen 47, 1020–​9. sidual disease monitoring: study ALL10 from the Dutch Childhood
    Hof J, et al. (2011) Mutations and deletions of the TP53 gene pre- Oncology Group. J Clin Oncol 34, 2591–​601.
    dict nonresponse to treatment and poor outcome in first relapse Pieters R, et al. (2007) A treatment protocol for infants younger than
    of childhood acute lymphoblastic leukemia. J Clin Oncol 29, 1 year with acute lymphoblastic leukaemia (Interfant-​99): an observa-
    3185–​93. tional study and a multicentre randomised trial. Lancet 370, 240–​50.
    Holmfeldt L, et al. (2013) The genomic landscape of hypodiploid acute Pui CH, et al. (2009) Treating childhood acute lymphoblastic leukemia
    lymphoblastic leukemia. Nature Gen 45, 242–​52. without cranial irradiation. N Engl J Med 360, 2730–​41.
    Hunger SP, Mullighan CG (2015) Acute lymphoblastic leukemia in Pui CH, et al. (2002) Outcome of treatment in childhood acute lympho-
    children. N Engl J Med 373, 1541–​52. blastic leukaemia with rearrangements of the 11q23 chromosomal
    Irving JA, et al. (2016) Integration of genetic and clinical risk factors region. Lancet 359, 1909–​15.
    improves prognostication in relapsed childhood B-​cell precursor Roberts KG, et al. (2014) Targetable kinase-​activating lesions in Ph-​
    acute lymphoblastic leukemia. Blood 128, 911–​22. like acute lymphoblastic leukemia. N Engl J Med 371, 1005–​15.
    Larsen EC, et al. (2016) Dexamethasone and high-​dose methotrexate Roberts KG, et al. (2014) Outcomes of children with BCR-​ABL1-​like
    improve outcome for children and young adults with high-​risk B-​ acute lymphoblastic leukemia treated with risk-​directed therapy
    acute lymphoblastic leukemia: a report from Children's Oncology based on the levels of minimal residual disease. J Clin Oncol 32,
    Group Study AALL0232. J Clin Oncol 34, 2380–​8. 3012–​20.
    Loh ML, et al. (2013) Tyrosine kinome sequencing of pediatric acute Schrappe M, et al. (2012) Outcomes after induction failure in child-
    lymphoblastic leukemia: a report from the Children's Oncology hood acute lymphoblastic leukemia. N Engl J Med 366, 1371–​81.
    Group TARGET Project. Blood 121, 485–​8. Schultz KR, et al. (2014) Long-​term follow-​up of imatinib in pediatric
    Mann G, et al. (2010) Improved outcome with hematopoietic stem Philadelphia chromosome-​positive acute lymphoblastic leukemia:
    cell transplantation in a poor prognostic subgroup of infants with Children's Oncology Group study AALL0031. Leukemia 28, 1467–​71.
    mixed-​lineage-​leukemia (MLL)-​rearranged acute lymphoblastic Stork LC, et al. (2010) Oral 6-​mercaptopurine versus oral 6-​thioguanine
    leukemia: results from the Interfant-​99 Study. Blood 116, 2644–​50. and veno-​occlusive disease in children with standard-​risk acute
    Moricke A, et al. (2016) Dexamethasone vs prednisone in induction lymphoblastic leukemia: report of the Children's Oncology Group
    treatment of pediatric ALL: results of the randomized trial AIEOP-​ CCG-​1952 clinical trial. Blood 115, 2740–​8.
    BFM ALL 2000. Blood 127, 2101–​12. Vora A, et al. (2016) Influence of cranial radiotherapy on outcome in
    Mullighan CG, et al. (2009) Deletion of IKZF1 and prognosis in acute children with acute lymphoblastic leukemia treated with contem-
    lymphoblastic leukemia. N Engl J Med 360, 470–​80. porary therapy. J Clin Oncol 34, 919–​26.
    Peters C, et al. (2015) Stem-​cell transplantation in children with acute Vora A, et al. (2013) Treatment reduction for children and young
    lymphoblastic leukemia: a prospective international multicenter adults with low-​risk acute lymphoblastic leukaemia defined by
    trial comparing sibling donors with matched unrelated donors–​the minimal residual disease (UKALL 2003): a randomised controlled
    ALL-​SCT-​BFM-​2003 trial. J Clin Oncol 33, 1265–​74. trial. Lancet Oncol 14, 199–​209.
    16
    Acute Myeloid Leukaemia
    Gertjan Kaspers and Dirk Reinhardt

    Introduction AML is the most frequent secondary malignancy in children


    treated for ALL, lymphomas, or other solid tumours, induced by in-
    Acute myeloid leukaemia (AML) is the second most frequent type of tensified therapy with exposure to alkylating agents such as cyclo-
    leukaemia in children, after acute lymphoblastic leukaemia (ALL). phosphamide, ifosfamide, or melphalan, and to topoisomersase II
    AML encompasses not only the myeloid leukaemias but also acute inhibitors.
    eythroblastic and acute megakaryoblastic leukaemia. Therefore, the The cause of AML is unclear in most cases. Unlike ALL, there is
    term acute non-​lymphoblastic leukaemia (as used formally in the no indication that infections play a role in the development of AML.
    past) is better, but AML is used in clinical practice. Improvement In a minority of cases, a predisposing condition to develop AML
    in the prognosis of AML in children came somewhat later than for is present, such as a chromosal breakage syndrome (e.g. Fanconi
    ALL, but has definitely occurred. Applying contemporary chemo- anaemia), a stem-​cell disease (congenital neutropenia, also known
    therapy, 70–​75% of children with AML are cured in high-​income as Kostmann syndrome), and, most frequently, Down syndrome.
    countries. In contrast to ALL, treatment is relatively short and usu- Nowadays, a clear causative factor such as ionizing irradiation,
    ally consists of four or five courses of very intensive combination previous chemotherapy, and certain chemicals is evident in even
    chemotherapy given over four to six months. A subset of the pa- fewer cases.
    tients may benefit from allogeneic haematopoietic stem-​cell trans- Although the aetiology is usually unknown, the pathophysiology
    plantation (HSCT). Experience with the intensive treatment and of AML is becoming more and more clear. AML is a clonal disorder
    high-​quality supportive care is of utmost importance in this setting. of haematopoietic stem cells, common myeloid-​lymphoid or early
    Therefore, the treatment of childhood AML should be centralized in myeloid progenitors. The hierarchical differentiation of AML in
    a limited number of paediatric oncology centres. part resembles normal haematopoiesis. Like normal haematopoietic
    stem cells (HSC), the leukaemic stem cell (LSC) is characterized by
    the capacity of self-​renewal, proliferation, remnant differentiation
    Epidemiology and aetiology programmes, and maturation. Several stages of leukaemic blast dif-
    ferentiation could be defined.
    The incidence of AML in children is about 7.6 per million chil- One model for AML development describes the coincidence of
    dren per year, and it accounts for about 20% of acute leukaemias type II and type I mutations as indispensable (Figure 16.1). The
    in this age group and for about 5% of all types of childhood cancer. type II (or class II) mutations are typically recurrent translocations
    An increased frequency of AML, especially of acute monoblastic ((t(8;21), inv(16), t(15;17), etc.) which are involved in impaired dif-
    and megakaryoblastic leukaemia, occurs within the first two years ferentiation and maturation of early haematopoietic progenitors.
    of life (1.5 in 100,000). The incidence remains relatively low during The isolated occurrence in a cell clone does not cause overt leu-
    childhood (0.5 in 100,000) and starts to increase during adoles- kaemia directly; however, it may define a pre-​leukaemic stem cell.
    cence and in young adults (0.9 in 100,000). There is a very high Only if a second activating mutation (type I) occurs as an additional
    concordance rate for leukaemia in monozygous twins, probably event, the uncontrolled proliferation and decreased rate of apop-
    the result of a common prezygotic or intrauterine genetic event or tosis cause the survival advantage of the leukaemic blasts, displace
    transplacental passage of a single leukaemic clone. The incidence of normal haematopoiesis, and result in a frank leukaemia. Type I mu-
    AML seems rather similar all over the world, with one clear excep- tations are typically related to cytokine receptors or other factors in
    tion, which is acute promyelocytic leukaemia (APL). APL is a sub- signal transduction pathways such as FLT3-​receptor, c-​kit-​receptor,
    type of AML, and occurs more frequently among the Hispanic and ras, PTPN11, and WT1.
    Mediterranean populations, and genetic susceptibility seems to play Although these events seem to be sufficient to cause leukaemia
    a role. Occasionally it is suggested that AML is more frequent in in some combinations, in most subtypes, additional events, specific
    certain Asian countries, but underreporting of other types of cancer conditions of the bone marrow microenvironment, or dysfunctions
    such as ALL, as an explanation, cannot yet be excluded. of the immunological system might contribute to the probability of
    Symptoms and signs: clinical presentation 125

    Epigenetic factors

    Type I Type II
    abnormalities AML abnormalities
    This can lead to This can lead to
    increased self- Acute Myeloid impaired
    renewal capacity Leukaemia differentiation
    and proliferation

    Unknown factors

    Figure 16.1 A model for leukaemogenesis in acute myeloid leukaemia involving type I abnormalities that lead to increased self-​renewal capacity and
    proliferation, type II abnormalities that result in impaired differentiation, and other partly unknown factors.

    leukaemia. These molecular abnormalities become more important extramedullary disease. The children usually have malaise and often
    in this era of targeted therapy, but they also provide opportunities weight loss. Major symptoms are caused by the leukaemic infiltra-
    for monitoring minimal residual disease during and after treatment. tion of the bone marrow, resulting in reduced normal haematopoi-
    Recent studies, based on next-​generation sequencing, revealed a esis. This translates into anaemia (resulting in paleness, fatigue, and
    more complex clonal hierarchy, with about three to seven subclones headache) and to thrombocytopenia (resulting in skin bleeding, epi-
    at diagnosis. Figure 16.2 illustrates current knowledge on chromo- staxis, and sometimes other signs of increased bleeding tendency).
    somal and molecular abnormalities in paediatric AML. Fever is a frequent symptom, either reflecting the inflammation
    caused by leukaemia or an infection as a consequence of the insuffi-
    cient number of normal white blood cells. The bone marrow infiltra-
    Symptoms and signs: clinical presentation tion often gives bone pain, which can be very severe. Enlargement of
    the liver and/​or spleen has been reported in about 50% of patients,
    The presentation of a child with AML is not fundamentally different and so is lymphadenopathy. A well-​known extramedullary localiza-
    from a child with ALL, although children with AML tend be older tion of AML is the central nervous system (CNS), most typically pre-
    (on average seven to eight years of age) and they more often have senting as asymptomatic disease detected at routine examination of

    NPM1
    n.a.
    20%
    Other 7.5%
    Complex 29% FLT3-ITD
    0.7% 13.1%
    FL T3-T K
    Trisomy 21 3.7%
    0.4% Normal WT1
    18% 16.0%
    MLL
    17.2% ras
    t(15;17) 24%
    5.5%
    t(8;21) inv(16) c-kit 7%
    8.8% PTPN11
    12.6%
    4%

    Figure 16.2 Percentages of acute myeloid leukaemia (AML) karyotypes; frequency of type I mutations.
    126 CHAPTER 16 Acute Myeloid Leukaemia

    the cerebrospinal fluid. However, CNS involvement can also present should be postponed because of the very high risk of bleeding com-
    with, for example, headache, intracerebral bleeding, and seizures. plications. Another reason to withhold a lumbar puncture is when
    Certain subtypes of AML (discussed later) have characteristic localized CNS disease is suspected, with the risk of herniation. The
    clinical problems. In acute promyelocytic and acute monoblastic abdominal ultrasound should look not only for hepatosplenomegaly
    leukaemia, bleeding tendency is often more severe than expected and lymphadenopathy, but also for renal enlargement caused by leu-
    as based on the platelet count. This is caused by diffuse intravas- kaemic infiltration. The latter should make the clinician aware of the
    cular coagulation. Acute myeloblastic leukaemia with the so-​called risk of renal dysfunction in the first days of treatment.
    t(8;21) is often associated with chloromas. Also called granulocytic Of course, the diagnosis of leukaemia requires laboratory investi-
    or myeloid sarcomas, chloromas are extramedullary manifestations gations. However, the importance of a detailed history and physical
    of AML presenting as solid masses. Acute myelomonocytic leu- examination cannot be underestimated. For example, the history
    kaemia in general, and especially the patient group with eosino- may reveal the suspicion of a preceding myelodysplastic syndrome.
    philia and the chromosal abnormality inv(16) in their AML cells, Physical examination is important both to detect features compat-
    has a very high frequency of extramedullary involvement, in up to ible with syndromes such as Down syndrome and Fanconi anaemia,
    60% of patients. Finally, acute monoblastic leukaemia is often char- and also to correctly determine extramedullary disease. For ex-
    acterized by skin involvement and gingival swelling, and sometimes ample, testicular involvement, gingival hyperplasia, and leukaemia
    with eye involvement. cutis will only be detected at physical examination. Physical exam-
    A significantly increased white blood cell (WBC) count may ination may also reveal serious bleeding tendency and a comprom-
    result in so-​called sludging, with hyperviscosity of the blood and ised circulation because of anaemia or sludging, which should result
    leukostasis. This can give rise to, for example, dyspnoea and to in immediate and appropriate supportive care and anti-​leukaemic
    confusion. treatment.

    Morphology
    Diagnosis and biology For the diagnosis of leukaemia, bone marrow aspiration is always
    attempted. Abnormal cells may also be present in sufficient number
    A correct and rapid diagnosis of AML is of utmost importance for in the peripheral blood. However, if this is not the case and if bone
    optimal treatment. First, the presence of AML should be established marrow aspiration does not provide enough cells (‘dry tap’), a bone
    without doubt, which occasionally is difficult when it resembles biopsy should be done and smears should be made, ‘rolling’ the bi-
    ALL or, for example, in case of a low blast count. Morphology and, opsy over the slide. Normally, enough abnormal cells from bone
    if available, immunophenotyping are still the fastest methods for marrow and peripheral blood are available for extensive charac-
    distinguishing between AML and ALL, or as an emergency means terization. First, this is done by morphological examination of
    to identifying APL immediately. Next, the extent of the disease the smears after staining. Additional enzymatic stainings are also
    should be determined, as CNS disease especially has therapeutic helpful to diagnose AML (Table 16.1). It requires experience to dis-
    consequences. tinguish leukaemic cells from normal precursors and, for example,
    from non-​malignant blasts in case of an infection, to identify a low
    Initial work-​up percentage of leukaemic cells, and to distinguish AML from ALL.
    Examinations required to determine the extent of AML are a chest Similarly, experience is required for diagnosing the correct FAB type
    X-​ray, lumbar puncture for investigation of the cerebrospinal fluid, according to the French–​American–​British (FAB) classification, de-
    and an ultrasound of the abdomen. The chest X-​ray should be done scribed in Table 16.2.
    the same day a leukaemia is suspected, because of the possibility of Although the most recent World Health Organization (WHO)
    mediastinal enlargement and compression on the trachea and upper classification relies more on chromosomal abnormalities than on
    airways. If that is the case, and depending on the severity, general morphology, a correct morphological diagnosis remains the basis
    anaesthesia may be contraindicated, as might investigations for and is essential for correct initial treatment. This can be simply
    which the patient must lie down flat. It also should lead to prompt illustrated by two examples. First, it is essential to quickly distin-
    treatment. A diagnostic lumbar puncture should only be done when guish AML from ALL in case of a high WBC count and the need
    there is no significant risk of bleeding and thereby introduction of to install anti-​leukaemic treatment immediately. A glucocorticoid
    peripheral leukaemic cells into the cerebrospinal fluid. If necessary, such as prednisone or dexamethasone is indicated in case of ALL,
    platelets should be transfused first. In case of APL, a lumbar puncture but is contraindicated in AML because it may actually increase the

    Table 16.1 The usefulness of enzymatic staining in the diagnosis of acute myeloid leukaemia

    M0 M1–​M3 M4 M5 M6 M7 ALL
    MPO − + + − − − −
    ANAE − − + + − +/​− −
    SB-​B − + + − − − −
    PAS − − +/​− +/​− + − +

    MPO = myeloperoxidase, ANAE = acid alpha-​naphthyl acetate esterase, SB-​B = Sudan-​Black B, PAS = Periodic Acid Schiff positive.
    Diagnosis and biology 127

    Table 16.2 Different types of acute myeloid leukaemia according to the contemporary treatment. Finally, sometimes additional investiga-
    French–​American–​British (FAB) classification, with percentage of the tions for immunophenotypical and chromosomal abnormalities fail
    total number of cases per FAB type, and the relation with karyotypic or are not available. In these cases, morphology may indicate the
    abnormalities
    presence of a specific chromosomal abnormality that is associated
    FAB subtype Full name (% of total number of Associated with a better prognosis. This is especially the case for AML FAB type
    cases) chromosomal M2 with auer rods, which is associated with t(8;21), and for AML
    abnormalities
    FAB type M4 with eosinophilia, which is associated with inv(16).
    M0 Acute myeloblastic leukaemia with (The other example, AML FAB type M3 or APL which is associated
    minimal differentiation (3%)
    with t(15;17), was described earlier.) Advantages of the WHO clas-
    M1 Acute myeloblastic leukaemia sification are that myeloid proliferations related to Down syndrome
    without maturation (10–​15%)
    are now recognized entities, as well as that patients with specific re-
    M2 Acute myeloblastic leukaemia with t(6;9) current genetic abnormalities are diagnosed with AML, even if the
    maturation (25–​30%)
    leukaemic blast percentage in the bone marrow is below 20%.
    M2au+ AML M2 with auer rods t(8;21)(q22;q22)
    In conclusion, paediatric haemato-​oncologists and those in training
    M3 Acute promyelocytic leukaemia t(15;17) for that specialty should be capable of diagnosing AML based on mor-
    (5–​10%)
    phological examination of bone marrow or peripheral blood smears.
    M4 Acute myelomonocytic leukaemia inv(16); del(16q) The official WHO classification requires at least 20% of leukaemic cells
    (5–​15%)
    in the bone marrow for the diagnosis of AML. In addition, the presence
    M4Eo Acute myelomonocytic leukaemia inv(16); t(16;16) of specific recurrent chromosomal abnormalities, independent from
    with abnormal eosinophils (5–​15%)
    the percentage of leukaemic cells, already leads to the diagnosis of AML
    M5 Acute monocytic leukaemia t(9;11); t(11;19);
    (15–​25%) del(11q)
    in the WHO classification, such as for t(8;21), inv(16), and t(15;17).
    It seems that many paediatric collaborative groups do not follow the
    M6 Acute erythroid leukaemia (1–​3%)
    WHO classification completely. Especially in case of 20–​30% leukaemic
    M7 Acute megakaryocytic leukaemia t(1;22) cells in the bone marrow, additional features such as organ involvement
    (5–​10%)
    and extramedullary disease are taken into account before the diagnosis
    Adapted with permission from Bennett JM et al., ‘Proposal for the recognition of of AML or, alternatively, myelodysplastic syndome (MDS) is established
    minimally differentiated acute myeloid leukaemia (AML-​M0)’. British Journal of
    Haematolgy, Volume 78, Issue 3, pp. 325–​29, Copyright © 2008, John Wiley and Sons. (Figure 16.3). If no additional features are present and doubt remains
    DOI: https://​doi.org/​10.1111/​j.1365-​2141.1991.tb04444 about the diagnosis being AML or MDS, bone marrow aspiration and
    biopsy should be repeated after several weeks, to look for progressive
    disease resulting in clear AML. However, the presence of specific cyto-
    WBC count (especially if given without chemotherapy). Second, the
    genetic abnormalities, such as those mentioned here, leads to the diag-
    early morphological recognition of AML FAB type M3 (APL) dir-
    nosis of AML independent from the blast percentage, even in children.
    ects therapy. The diagnosis (or even the suspicion) of APL should
    prompt the clinicians to start all-​trans retinoic acid (ATRA) and/​
    Immunophenotyping
    or arsenic trioxide (ATO) immediately. ATRA and ATO are es-
    sential in the initial treatment of APL because these drugs quickly Additional characterization of AML is based on immunophenotype
    reduce the risk of bleeding complications. Early bleeding is life-​ and chromosomal abnormalities. Nowadays, immunophenotyping
    threatening in patients with APL and fatal in 5–​10% of children with is done with multi-​colour flow cytometry in high-​income coun-
    APL. In addition, continued use of ATRA and ATO makes a very tries, but can be done on slides if such facilities are lacking.
    important contribution to the cure of most children with APL with Immunophenotyping may profit from the asynchronous or aberrant

    BM ≥30% Specific cytogenetic BM ≥30%


    myeloid abnormalities and/or organ myeloid
    blasts: AML involvement: AML blasts: AML

    AML BM <30% No other Repeat BM


    suspected myeloid blasts abnormalities after 2–3 weeks

    BM <30%
    myeloid
    blasts: MDS

    Figure 16.3 A systematic approach for dissecting acute myeloid leukaemia (AML) from myelodysplastic syndrome (MDS).
    128 CHAPTER 16 Acute Myeloid Leukaemia

    antigen expression on the leukaemic blasts. Asynchronous expres- of transcription factors, including PML-​RARα (t(15;17)(q22;q21)),
    sion is the simultaneous expression of stem-​cell antigens (CD34/​ AML1-​ETO (t(8;21)(q22;q22)), CBFB-​MYH11 (inv(16)(p13q22)),
    CD117) and antigens indicating myeloid differentiation (CD13, MLL rearrangements, or mutations in NPM1 or CEBPA. Aside from
    CD14, CD15, CD33, myeloperoxidase (MPO)). Aberrant expres- possibleinvolvement in leukaemogenesis, molecular abnormalities
    sion includes the simultaneous expression of myeloid antigen have other potential clinical applications. These include a prognostic
    and lymphoid or natural killer cell antigens (CD2, CD7, CD19, significance (see the section ‘Prognostic factors’) and use for the de-
    CD56). Typical antigens indicative for AML are CD33, CD13, tection of minimal residual disease. Perhaps most importantly, they
    CD14, CDw41, CD15, CD11b, CD36, and CD61. In the diag- have the potential to provide leukaemia-​specific treatment targets.
    nostic setting, immunophenotyping is especially important for the
    diagnosis of acute undifferentiated leukaemia (FAB type M0; no
    expression of MPO but reactivity with CD13 and/​or CD33) and Prognostic factors
    acute megakaryoblastic leukaemia (FAB type M7, expression of
    CD36, CD41, CD42b, or CD61)), because these subtypes are not- The previously described characterization of AML and its extension
    ably difficult to recognize and distinguish morphologically. Specific has prognostic relevance. Chromosomal abnormalities dominate
    immunophenotypes could also be defined for acute erythoid leu- clinical features (e.g. age, sex, WBC count), disease extension, and
    kaemia (FAB M6, expression of CD235a/​glycophorin) and for acute morphology of the AML cells, as prognostic factors. In addition,
    monoblastic leukaemia (AML FAB M5; CD33/​CD56/​CD36). Some early clinical response—​as determined by either morphological
    immunophenotypes are associated with genetic aberration, such as examination of bone marrow one, two, or four weeks after the start
    the combination of CD33/​CD19/​CD56 positivity that is indicative of treatment, or by more sophisticated minimal residual disease
    of t(8;21). Immunophenotyping is also instrumental in the measure- measurements—​is an important predictor of long-​term outcome.
    ment of minimal residual disease by the detection of leukaemic cells Clinically, a high WBC count remains an adverse prognostic
    with a leukaemia-​aberrant immunophenotype. factor. Otherwise, extramedullary involvement does not seem to im-
    pact prognosis with current intensive therapy, including high-​dose
    Cytogenetic and molecular diagnostics cytarabine. If information on chromosomal abnormalities is lacking,
    Chromosomal abnormalities can be detected with a number of morphological subtypes can be used to predict outcome. Favourable
    techniques. Conventionally, karyotyping is used, but, more and are FAB type M2 with auer rods, FAB type M3 (APL), and FAB type
    more, fluorescent in situ hybridization (FISH) is routinely applied M4, especially (but not exclusively) if associated with eosinophilia.
    as well, while polymerase chain reaction (PCR) techniques are add- Moreover, myeloid leukaemia of Down syndrome (usually FAB type
    itionally being used to detect molecular abnormalities. In the case M7, occasionally M6) has a favourable prognosis and will be de-
    of karyotyping, at least 20 metaphases are needed in order to reli- scribed later as a separate disease entity. Immunophenotype is not
    ably exclude an abnormal karyotype. If karyotyping and other tech- of significant prognostic relevance in AML.
    niques looking for chromosomal abnormalities fail, another attempt More and more, cytogenetic abnormalities are being used as a
    to obtain leukaemic cells should be considered, in view of the clin- prognostic factor, and most groups agree with three cytogenetic risk
    ical relevance of these investigations. FISH is a reliable and robust groups. The low-​risk group has core-​binding factor AML (i.e. either
    technique to detect, for example, gene rearrangements. PCR tech- inv(16), t(16;16), or t(8;21)) or has APL with t(15;17). Patients with
    niques and other assays, such as the multiplex ligation polymerase these low-​risk cytogenetic features, but a less than favourable early
    assay, are useful for the detection of somatic molecular abnormal- treatment response, should most likely remain low-​risk patients, if
    ities occurring in the leukaemic cells exclusively. they achieve complete remission (CR) after two induction courses;
    Recently, next-​generation sequencing (NGS), applied as targeted and similarly for t(8;21)-​positive patients, if there is no coinciding
    or whole-​exome sequencing, has become routine in some labora- gain of chromosome 4. Patients with other molecular abnormal-
    tories. Targeted NGS analyses predefined genes known to carry ities, such as mutations in NPM1 or GATA1s and double mutations
    AML-​associated mutations. In addition to the detection of AML-​ in CEPBA, and with t(1;11), also qualify for the low-​risk group. The
    specific mutations in one or more genes, NGS also provides insight intermediate-​(i.e. standard) risk group has no specific cytogenetic
    into the clonal hierarchy of the AML at diagnosis. Comparative abnormalities as defined for low-​or high-​risk groups. The high-​risk
    studies between diagnosis and relapse samples have elucidated the group includes patients with AML cells that harbour FLT3-​ITD but
    clonal evolution of the AML. without a NPM1 mutation, monosomy 7, an abnormal chromosome
    There are a wide range of other techniques being used to char- 5 (monosomy 5, 5q-​), inv(3) or t(3;3)/​t(6;9)/​t(7;12)/​t(9;22)/​t(4;11)/​
    acterize leukaemic cells, such as microarrays, comparative genomic t(5;11)/​t(6;11)/​t(10;11), abnormalities of 12p, or a complex karyo-
    hybridization, and proteomics. However, the results are not being type with three or more cytogenetic aberrations.
    applied in clinical practice yet and are therefore not discussed here, The use of prognostic factors for risk-​group-​adapted therapy is
    except in relation to the detection of molecular abnormalities. Such described in the section ‘Risk classification’.
    abnormalities are detected mainly with PCR techniques (as already
    mentioned regarding leukaemogenesis). Often these molecular ab-
    normalities are distinguished in type I and type II aberrations. Type Risk classification
    I aberrations lead to increased self-​renewal and increased prolifera-
    tive capacity, and well-​known examples are mutations in the RAS, According to the WHO classification, AML subgroups are de-
    KIT, FLT3, WT1, and PTPN11 genes. Type II aberrations cause dif- fined by the history of the disease—​a preceding myelodysplastic
    ferentiation arrest, and often concern chromosomal rearrangements syndrome or treatment for a malignancy, by recurrent cytogenetic
    Minimal residual disease 129

    abnormalities such as t(8;21), inv(16), t(15;17), and MLL rearrange- treatment itself remains the most important prognostic factor. A lot
    ments, and by the lack of such features (not otherwise character- of other prognostic factors have been proposed. These include the
    ized). This last category largely reflects the former FAB classification, WBC count at diagnosis, age, the occurrence of extramedullary leu-
    which is based on morphology and cytochemistry. kaemia, or the immunophenotype, although the prognostic rele-
    AML is a heterogeneous disease, and the prognosis of the sub- vance seems to be associated to intensity and schedule of therapy.
    types varies widely. In addition, the prognostic relevance of an Apart from genetic characteristics, treatment response is an im-
    identified factor does not only depend on the biological character- portant prognostic factor. Assessed by bone marrow morphology,
    istics but also on the applied therapy. The occurrence of balanced CR—​defined as less than 5% blasts and sufficient haematological re-
    translocations—​t(8;21)/​inv(16)—​affecting core-​binding factors generation (neutrophils higher than 1000/​µl; platelets higher than
    (CBF-​AML) is associated with a relatively good prognosis. This has 80.000/​µl)—​is a strong predictor for outcome. More difficult is the
    been shown in both children and adults. The event-​free and overall measurement of early response. Morphology is impeded by a low
    survival in this particular group of children is more than 70% and sensitivity and specificity, but has been successfully used at day 14
    85%, respectively. Even in cases of relapse, these karyotypes remain or 15 and/​or day 28, after the first induction course to stratify pa-
    a favourable prognostic factor, and these patients have achieved an tients to risk groups. However, measurements of so-​called minimal
    overall survival of more than 60% from relapse. Further, APL can be residual disease (MRD) have become more important.
    treated successfully in most cases if the frequently occurring early
    complications such as severe bleedings or infections can be treated
    or, better, prevented. The introduction of ATRA and ATO therapy Minimal residual disease
    in particular has prevented early deaths and enables survival rates of
    about 90% in APL. Several attempts have been made to establish more sensitive methods
    Nucleophosmin-​1 (NPM1) mutations, occurring in about 8% of to measure response and MRD. Quantitative reverse transcription
    childhood AML, seems to be associated with a favourable prognosis. PCR (RT-​PCR) allows the sensitive detection and quantification
    In the subgroup of AML with a normal karyotype, the incidence is up of known fusion genes such as AML1/​ETO or CBF/​MYH11. As
    to 20%. Interestingly, the non-​type A mutations are more frequent the frequency of these aberrations is about 40%, only a minority of
    in children than in adults. Further, a correlation with increasing age children can be monitored based on these leukaemia-​specific abnor-
    has been observed. malities. More critical is the fact that a relevant portion of patients
    A special entity is myeloid leukaemia of Down syndrome (ML-​ remains ‘MRD-​positive’ throughout the complete course of the dis-
    DS) with GATA1s mutation. If adapted treatment is applied, the ease, with persisting cells with the fusion gene. Up to 40% of the
    prognosis is good. The intensity of the treatment schedule should be children show detectable AML/​ETO1 even after intensive therapy
    significantly reduced as compared to standard protocols. By contrast without predicting relapse. By contrast, the monitoring of PML/​
    to other AML subtypes, the immediate continuation of treatment is RARα in APL has generally been shown to be very predictive and is
    not time-​sensitive; a treatment element should be started only if the used to stratify treatment intensity, if PML/​RARα remains after sev-
    child is in a good general condition. eral cycles of therapy. This even includes the recommendation to act
    In the past, MLL rearrangements were generally recognized as upon molecular refractory disease and to treat a molecular relapse.
    a poor prognostic factor; however, at least two studies in children In MLL-​gene rearranged AML the situation is even more compli-
    found the t(9;11) as a favourable marker. These results, based on a cated. Persistence of the rearranged MLL gene is clearly associated
    relatively low number of patients, have not been confirmed by larger with a very high risk of relapse; however, in most cases the MLL ar-
    study groups. The prognosis of childhood AML with t(9;11) seems rangement is not detectable after the first or second chemotherapy
    to be insignificantly better than in other MLL-​positive leukaemia; induction course, which does not exclude relapse. Further, re-​
    it is still within the range of intermediate-​and high-​risk patients. occurrence of the MLL rearrangement is strongly associated with re-
    Several aberrations that are associated with a very poor prognosis lapse within two to four weeks. Whether this very short pre-​warning
    in adults do not discriminate in children. For example, a complex phase has any clinical meaning is unclear.
    karyotype (three or more aberrations) or 7q–​do not predict a par- Another approach to monitoring MRD is immunophenotyping. In
    ticularly poor outcome in children. However, isolated monosomy 7 almost all cases of AML, a leukaemia-​associated immunophenotype
    is associated with an impaired CR rate and a poor outcome. (LAIP) can be defined based on an asynchronous expression of dif-
    Several additional molecular events, mostly type I mutations, ferential, stage-​dependent antigens (simultaneous expression of
    have been identified as a prognostic marker in childhood AML. stem-​cell antigens and antigens of myeloid maturation) or based on
    Most data are available about the poor prognostic value of internal lineage-​aberrant antigen expression. Although some studies have
    tandem duplications in the FLT3 gene (FLT3-​ITDs). Especially in demonstrated the possibility of monitoring the treatment response,
    children with normal karyotype, FLT3-​ITD is associated with poor the independent value of the MRD diagnostic is so far hampered
    prognosis. While FLT3-​ITD and NPM1 are frequently associated by the limited specificity. First, the LAIP of the bulk of leukaemic
    with AML with normal karyotype, the relevance of other aberra- blasts could differ from the immunophenotype of the leukaemic
    tions such as RAS or c-​KIT mutations remains unclear. Recently, stem cell. Second, the expression pattern of leukaemic blasts changes
    WT1 mutations have been identified as a poor prognostic factor in during treatment and at relapse. Third, especially in a regenerating
    childhood AML. bone marrow, the leukaemic blasts may not be distinguishable from
    Interestingly, there are recent reports that have described a possible normal progenitors.
    prognostic relevance of certain polymorphisms in AML; however, The discovery of a leukaemic stem-​cell-​specific antigen (CLL-​
    this is probably dependent on age and treatment regimen. Indeed, 1) within the CD34+/​CD38–​compartment might improve the
    130 CHAPTER 16 Acute Myeloid Leukaemia

    specificity of immunophenotyping and provide a more reliable overall survival (OS) have been within similar ranges. The most
    method in the future. This might also be achieved by improved relevant treatment differences concern the cumulative dose of
    techniques (8 colour-​flow) and better methods of analysis. Indeed, anthracyclines and the percentage of patients that undergoes allo-
    nowadays, most collaborative groups have implemented MRD geneic HSCT. Table 16.3 summarizes the EFS and OS of several
    monitoring by flow cytometry to stratify consolidation treatment. large clinical trials in paediatric AML.
    Recently, MRD has also been evaluated by using NGS. This seems
    promising, especially because additional information about the Induction
    residual clone might become available. However, the clinical and Anthracycline-​and cytarabine-​based regimens are generally ac-
    prognostic relevance of MRD monitoring by NGS have not yet been cepted to induce remission in AML. Sequential studies proved the
    proven. efficacy of intensive, high-​dose chemotherapy. Double-​induction
    schedules are now used by all collaborative groups in high-​income
    countries. Today, on average, 90% of the children with AML achieve
    Treatment and prognosis CR after induction, although this does depend on the risk group.
    Cytarabine is used in different dose regimens. Whereas in first
    Over the last three decades, the prognosis of childhood AML induction 3×60 mg/​m2 daunorubicin is frequently administered in
    has markedly improved (Figure 16.4). In children, the first ap- combination with lower-​dose cytarabine (and often etoposide), in
    proaches with intensive chemotherapy were initiated in the 1970s. second induction, several studies schedule high-​dose cytarabine,
    These protocols were adapted from ALL treatment schedules. In with 1–​3 g/​m2 combined with mitoxantrone or an anthracycline.
    the 1980s, the benefit of intensified elements, including cytarabine Newer analogues of cytarabine (such as fludarabine, cladribine, or
    and anthracyclines, became obvious. Several other drugs such as clofarabine) are frequently used in re-​induction in relapsed AML,
    etoposide (a topoisomerase II inhibitor), amsacrine, thioguanine, or but are rarely included in frontline protocols. The addition of chlor
    cyclophosphamide have been combined. The mechanism of action or fluor to the basic cytosine-​arabinoside impairs the intracellular
    of these drugs has been described elsewhere. degradation by purine nucleoside phosphorylase or hydrolysis. The
    Most of the treatment protocols recommend four or five elem- latest development in this field is clofarabine, which might combine
    ents of intensive chemotherapy. At least with regard to children with the advantages of fludarabine and cladribine, and which has shown
    favourable cytogenetics, the results of the UK Medical Research promising results in relapsed AML. However, a randomized com-
    Council (MRC) group and the AML-​ BFM (Berlin–​ Frankfurt–​ parison to currently used drugs is still lacking.
    Munster) group did not provide strong evidence that a fifth intensive
    element improves outcome. However, the subgroup of AML with Post remission
    t(8;21) might benefit from high-​dose cytarabine during induction In most studies, post-​ remission treatment includes intensive
    and/​or consolidation. chemotherapy such as high-​dose cytarabine-​based regimens and,
    In spite of the different treatment regimens of several different additionally, an anthracycline or etoposide. After two courses of in-
    study groups, the final results for disease-​free survival (DFS) and duction, between two to four intensive post-​remission courses have

    0,8
    AML-BFM 98
    0,6 AML-BFM 93
    AML-BFM 87
    P

    AML-BFM 83
    0,4
    AML-BFM 78
    0,2

    0
    0 5 10 15 20 25 30
    overall survival (years)

    AML-BFM 78 n = 151 5yrs-OS 42 ± 4% 10yrs-OS 39 ± 4%


    AML-BFM 83 n = 182 5yrs-OS 52 ± 4% 10yrs-OS 50 ± 3%
    AML-BFM 87 n = 322 5yrs-OS 51 ± 3% 10yrs-OS 49 ± 3%
    AML-BFM 93 n = 400 5yrs-OS 58 ± 3% 10yrs-OS 57 ± 3%
    AML-BFM 98 n = 345 5yrs-OS 65 ± 3% 10yrs-OS 64 ± 3%
    Figure 16.4 Improvement of overall survival in childhood acute myeloid leukaemia (AML): example, AML-​BFM Studies 78 to 98.
    Side effects and late effects 131

    Table 16.3 Results from paediatric acute myeloid leukaemia studies

    Study group Trial name Event-​free survival (%) Overall survival (%)
    (period)
    AIEOP AML2002/​01 (2002–​2011) 8-​yr (55±3) 8-​yr (68±2)
    AML-​BFM AML-​BFM 2004 (2004-​2010) 5-​yr (55±2) 5-​yr (74±2)
    COG AAML0531 3-​yr (53±4) 3-​yr (69±4)
    DB (2006–​2010) 3-​yr (50) 3-​yr (74)
    DB-​AML01 (95%CI 40–​59%) (95%CI 64–​82%)
    (2010–​2014)
    JPLSG AML05 (2006–​2010) 3-​yr (54±2) 3-​yr (73±2)
    MRC MRC AML12 (1995–​2002) 10-​yr (54) 10-​yr (63)
    NOPHO NOPHO AML 2004 3-​yr (57±5) 3-​yr (69±5)
    (2004–​2009)
    PPLLSG PPLLSG AML-​98 (1998–​2002) 5-​yr (47±5) 5-​yr (50±5)
    SJCRH AML02 3-​yr (63±4) 3-​yr (71±4)
    (2002-​2008)

    been scheduled in paediatric AML studies. Although not uniformly in APL, the recent introduction of ATO in addition to ATRA seems
    defined, at least one to two courses followed by allogeneic stem-​cell to have abolished the need for maintenance treatment.
    transplantation (SCT), are given in high-​risk AML.
    Acute promyelocytic leukaemia
    Stem-​cell transplantation APL is the first subtype of acute leukaemia which nowadays can be
    The value of allogeneic HSCT in first CR is controversial. Whereas cured without conventional cytostatic drugs, but with ATRA and
    some groups still recommend allogeneic HSCT for all AML sub- ATO. The combination of these differentiating agents, which appar-
    groups (except for APL) if a matched sibling donor (MSD) is ently also eliminates the leukaemic stem cells, has shown very fa-
    available, other groups have recently tended to restrict HSCT to vourable cure rates and, at the same time, reduced side effects and
    intermediate-​or high-​ risk AML patients. Children’s Oncology improved quality of life.
    Group (COG) studies, which previously recommended allogeneic
    HSCT, could not prove the advantage of allogeneic HSCT in patients
    with favourable or unfavourable cytogenetics. Side effects and late effects
    Although, as confirmed in a recent meta-​analysis, the relapse rate
    of intermediate-​and high-​risk AML could be lowered by allogeneic Side effects
    SCT in most studies, the higher treatment-​related mortality coun- Each treatment option in childhood AML have considerable
    terbalanced this advantage. New specific prognostic factors need acute and long-​term adverse effects, and almost all organ sys-
    to be identified to define those patients who will benefit from allo- tems are involved. Of the acute side effects and toxicities, expected
    geneic HSCT. Preliminary but yet unpublished results suggest that myelosuppression, mucositis, and the consecutive high risk of severe
    allogeneic SCT in high-​risk patients, as defined by a poor early treat- infections and septicaemia are the most relevant, since infections are
    ment response or specific biological features of the AML cells, might responsible for most treatment-​related deaths.
    improve EFS and OS. Improved supportive and intensive care can reduce early deaths
    By contrast, there is a clear indication to perform allogeneic HSCT by infections to less than 5%. This includes guidelines for sup-
    in refractory or relapsed AML, after the attempt to achieve a second portive care that cover prophylactic and pre-​emptive antibiotic
    CR with two cycles of intensive combination chemotherapy, usually and antifungal therapy, recent development of effective and orally
    including fludarabine, high-​dose cytarabine, and liposomal dauno- applicable antifungal drugs, and the availability of antiviral com-
    rubicin or another anthracycline or mitoxantrone. Even more experi- pounds. Improved intensive care has also enabled the management
    mental approaches, such as haploidentical SCT or cord blood SCT, may of severe complications and continuation of treatment.
    be indicated in these cases, if a matched donor is lacking. Interestingly, The most impressive effect has been achieved by the introduction
    recent data suggest that children with very late relapse and favourable of ATRA and/​or ATO into the initial treatment of APL, which has
    prognostic features can also be rescued by chemotherapy only. reduced the early deaths in this particular subgroup substantially.
    Efforts to reduce the duration of myelosuppression by adminis-
    Maintenance tering the granulocyte growth factor G-​CSF resulted in a limited
    With the possible exception of APL, there is no evidence of a benefit shortening of neutropenia but did not change the rate of severe
    of maintenance therapy in childhood AML. Only the AML-​BFM infections or the probability of survival. In fact, in those children
    group applied low-​dose maintenance therapy until recently, adapted treated with G-​CSF, there is increased incidence of relapse in those
    from adult AML studies, because the combination of thioguanine with favourable cytogenetics and an increased expression of the G-​
    and cytarabine has shown improved results in some adult groups. CSF receptor isoform IV. In conclusion, G-​CSF should not be ad-
    However, this has not been prospectively proven in children. Even ministered routinely to children with AML. Other options, such as
    132 CHAPTER 16 Acute Myeloid Leukaemia

    the application of supportive granulocyte infusions, are controver- is used. Although most paediatric groups worldwide are now using
    sial and efficacy has not been proven. four or five courses of intensive chemotherapy, there are still many
    Another open question concerns the management of differences in drugs, doses, and schedules, and concerning the in-
    hyperleukocytosis. It is undisputed that rasburicase and dications for allogeneic HSCT. Another main challenge is the per-
    antiproliferative therapy should be started immediately; however, formance of clinical trials under the EU directive and good clinical
    because of the risk of tumour lysis, the best starting dose cannot practice guidelines. Although this EU directive has had a positive
    easily be defined. Children with severe hyperleukocytosis (more effect on the number of clinical studies on new agents in children
    than 100,000 leukocytes/​µl blood; APL more than 10,000 leuko- with cancer, it has become a very complicated process in which, un-
    cytes/​µl) should be treated on an intensive-​care unit, especially if fortunately, not all countries apply it similarly. Investigator-​initiated
    they show any additional abnormal parameters such as those related studies have become much more cumbersome and there is a risk
    to coagulation, blood-​gas analysis, renal parameters, and increased that such studies will be done less often. That would be a major
    lactate. If severe tumour lysis is expected, the effect of the first dose problem, because company-​driven studies are still rare in children
    of cytostatics should be carefully monitored. There is no consensus with cancer, including AML.
    on whether such first doses should be lower than normal in patients Despite these challenges, the paediatric community has always
    with hyperleukocytosis above 50,000–​100,000 WBC/​µl blood. been able to improve the treatment and outcome of all types of ma-
    Another option that is sometimes considered in the case of a crit- lignancies. Techniques are rapidly improving to identify subgroups
    ically ill child due to hyperleukocytosis, is an exchange transfusion that should be treated differently. More and more novel drugs are
    or even leukapheresis, although the efficacy of this latter procedure emerging that could be used in different subgroups. Therefore, it
    is not proven. seems realistic to expect long-​term survival in more than 90% of
    paediatric AML patients, with an acceptable quality of life during
    Late toxicity treatment and with limited late effects.
    Late sequelae of therapy (see also Chapter 14) can be attributed to
    the initial disease and to the complications of its treatment. The lit-
    erature offers a range of perspectives on the incidence of significant FURTHER READING
    late effects in survivors of AML. In one retrospective study from the Arber DA, et al. (2016) The 2016 revision to the World Health
    AML-​BFM study group, only about 5% of all long-​term survivors Organization classification of myeloid neoplasms and
    were reported to experience severe late toxicities (such as blindness/​ acute leukemia. Blood 127(20), 2391–​ 405. doi: 10.1182/​
    cataracts, severe osteonecrosis, cardiomyopathy, or secondary malig- blood-​2016-​03-​643544.
    nancies). In comparison to reports from other studies, this incidence Bonnet D, Dick JE (1997) Human acute myeloid leukemia is organized
    as a hierarchy that originates from a primitive hematopoietic cell.
    is relatively low, and there is much greater morbidity in patients who
    Nat Med 3, 730–​7.
    undergo allogeneicHSCT. The importance of anthracycline drugs in
    Creutzig U, et al. (2006) Less toxicity by optimizing chemotherapy, but
    schedules for the treatment of AML raises particular concern about not by addition of granulocyte colony-​stimulating factor in chil-
    the risk of cardiotoxicity. dren and adolescents with acute myeloid leukemia: results of AML-​
    BFM 98. J Clin Oncol 24, 4499–​506.
    Creutzig U, et al. (2007) Longitudinal evaluation of early and late
    Challenges and perspectives anthracycline cardiotoxicity in children with AML. Pediatr Blood
    Cancer 48, 651–​62.
    Prognosis of paediatric AML has improved significantly, to even Creutzig U, et al. (2012) Diagnosis and management of acute myeloid
    above 70% long-​term survival in some recently published clinical leukemia in children and adolescents: recommendations from an
    studies. The biggest challenge is to further improve outcome to a international expert panel. Blood 120, 3187–​205.
    cure rate of well above 90% and ultimately 100%, with acceptable Creutzig U, et al. (2013) Randomized trial comparing liposomal dauno-
    side effects and limited late effects of treatment. Current conven- rubicin with idarubicin as induction for pediatric acute myeloid
    tional chemotherapy cannot be intensified further because of tox- leukemia: results from Study AML-​BFM 2004. Blood 122, 37–​43.
    Creutzig U, et al. (2017) First experience of the AML-​Berlin–​Frankfurt–​
    icity. Some novel but conventional (non-​targeted) drugs, such as
    Münster group in pediatric patients with standard-​ risk acute
    liposomal daunorubicin and clofarabine, will likely make an im-
    promyelocytic leukemia treated with arsenic trioxide and all-​trans ret-
    provement, but main steps forward are more likely to be achieved inoid acid. Pediatr Blood Cancer 64(8). doi: 10.1002/​pbc.26461
    with really novel drugs and novel treatment modalities. Novel agents Ehlers S, et al. (2010) Granulocyte colony-​stimulating factor (G-​CSF)
    are usually targeted, which has the implication that they will be used treatment of childhood acute myeloid leukemias that overexpress
    only in patients whose AML cells harbour the treatment target. An the differentiation-​defective G-​CSF receptor isoform IV is associ-
    example is tyrosine kinase inhibitors targeting mutated FLT3 or c-​ ated with a higher incidence of relapse. J Clin Oncol 28, 2591–​7.
    KIT. Another example is monoclonal antibodies that, either alone or Gibson BE, et al. (2005) Treatment strategy and long-​term results in
    by being linked to a cytotoxic agent, exert an anti-​leukaemic effect. paediatric patients treated in consecutive UK AML trials. Leukemia
    A good example of the latter is gemtuzumab ozogamicin—​an anti-​ 19, 2130–​8.
    CD33 antibody linked to the highly toxic calcheamicin. Gibson BES, et al. (2011) Results of a randomized trial in children with
    Studies on such novel agents in subgroups of paediatric AML will acute myeloid leukemia: Medical Research Council AML12 trial. Br
    require international collaboration because of the rarity of these J Haematol 155, 366–​76.
    Gilliland DG (2001) Hematologic malignancies. Curr Opin Hematol
    subgroups. An important challenge in such international collabora-
    8, 189–​91.
    tive studies is that, ideally, a common chemotherapeutic backbone
    Challenges and perspectives 133

    Greaves MF, Wiemels J (2003) Origins of chromosome translocations Leung W, et al. (2000) Late effects of treatment in survivors of child-
    in childhood leukaemia. Nat Rev Cancer 3, 639–​49. hood acute myeloid leukemia. J Clin Oncol 18, 3273–​9.
    Harrison CJ, et al. (2010) Cytogenetics of childhood acute myeloid leu- Lipshultz SE, et al. (1995) Female sex and drug dose as risk factors
    kemia: United Kingdom Medical Research Council treatment trials for late cardiotoxic effects of doxorubicin therapy for childhood
    AML 10 and 12. J Clin Oncol 28, 2674–​81. cancer. N Engl J Med 332, 1738–​43.
    Hasle H, et al. (2007) Monosomy 7 and deletion 7q in children and Perel Y, et al. (2002) Impact of addition of maintenance therapy to in-
    adolescents with acute myeloid leukemia: an international retro- tensive induction and consolidation chemotherapy for childhood
    spective study. Blood 109, 4641–​7. acute myeloblastic leukemia: results of a prospective randomized
    Hasle H, Kaspers GJL (2017) Strategies for reducing the treatment-​ trial, LAME 89/​91 for the group LAME of the French Society of
    related physical burden of childhood acute myeloid leukaemia—​a Pediatric Hematology and Immunology. J Clin Oncol 20, 2774–​82.
    review. Br J Haematol 176, 168–​78. Rubnitz JE, et al. (2002) Favorable impact of the t(9;11) in childhood
    Jeha S, et al. (2004) Clofarabine, a novel nucleoside analog, is active in acute myeloid leukemia. J Clin Oncol 20, 2302–​9.
    pediatric patients with advanced leukemia. Blood 103, 784–​9. Rubnitz JE, et al. (2010) Minimal residual disease-​directed therapy
    Kaspers GJ, Zwaan CM. (2007) Pediatric acute myeloid leukemia: to- for childhood acute myeloid leukaemia: results of the AML02
    wards high-​quality cure of all patients. Haematologica 92, 1519–​32. multicentre trial. Lancet Oncol 11, 543–​52.
    Kaspers GJL (2014) How I treat paediatric relapsed acute myeloid leu- Testi AM, et al. (2005) GIMEMA-​AIEOPAIDA protocol for the treat-
    kaemia. Br J Haematol 166, 636–​45. ment of newly diagnosed acute promyelocytic leukemia (APL) in
    Kaspers GJ, et al. (2013) Improved outcome in pediatric relapsed children. Blood 106, 447–​53.
    acute myeloid leukemia: results of a randomized trial on liposomal Zwaan CM, et al. (2003) FLT3 internal tandem duplication in 234 chil-
    daunorubicin by the International BFM Study Group. J Clin Oncol dren with acute myeloid leukemia: prognostic significance and re-
    31, 599–​607. lation to cellular drug resistance. Blood 102, 2387–​94.
    Langebrake C, et al. (2006) Residual disease monitoring in childhood Zwaan CM, et al. (2010) Salvage treatment for children with re-
    acute myeloid leukemia by multiparameter flow cytometry: the fractory first or second relapse of acute myeloid leukaemia with
    MRD-​AML-​BFM Study Group. J Clin Oncol 24, 3686–​92. gemtuzumab ozogamicin: results of a phase II study. Br J Haematol
    Lehrnbecher T, et al. (2007) Prophylactic human granulocyte colony-​ 148, 768–​76.
    stimulating factor after induction therapy in pediatric acute mye- Zwaan CM, et al. (2015) Collaborative efforts driving progress. in pedi-
    loid leukemia. Blood 109, 936–​43. atric acute myeloid leukemia. J Clin Oncol 33, 2949–​62.
    17
    Myelodysplastic Syndrome, Myeloid
    Leukaemia of Down Syndrome, and Juvenile
    Myelomonocytic Leukaemia
    Henrik Hasle and Charlotte M. Niemeyer

    Introduction Epidemiology
    Incidence, sex, age, and subtype distribution
    Myeloid malignancies in children are subdivided into four main
    Combined population-​ based data from Denmark and British
    groups: acute myeloid leukaemia (AML), myelodysplastic syn-
    Columbia in Canada showed an annual incidence of MDS of 1.8 and
    drome (MDS), juvenile myelomonocytic leukaemia (JMML),
    of JMML of 1.2 per million children, corresponding to a total of 6% of
    and the myeloid leukaemias of Down syndrome (TAM and
    all haematological malignancies in children, although data from the
    ML-​DS). The rarity of MDS in children and the lack of definite
    UK suggest a lower incidence of MDS of 0.8 per million. The UK study
    morphological and cytogenetic markers have contributed to
    excluded secondary MDS, partly explaining the lower incidence. The
    the paucity of MDS in the paediatric literature. Most protocol-​
    male/​female distribution in paediatric MDS is equal, with a median
    based studies on MDS and JMML have been performed by the
    age at presentation of 6.8 years. The exclusion of Down syndrome from
    European Working Group of MDS in Childhood (EWOG-​MDS;
    patient series of MDS has decreased the number of cases diagnosed as
    www.ewog-​mds.org).
    MDS-​EB by almost 50%. Most paediatric MDS presents as RCC.

    Associated abnormalities
    Myelodysplastic syndrome MDS has been reported in a number of constitutional cytogenetic
    abnormalities other than Down syndrome, but there is only solid
    Classification evidence for an association between trisomy 8 mosaicism and MDS.
    Reports of MDS and AML in patients with Klinefelter and Turner
    Starting in the early 1980s, the French–​American–​British (FAB) syndrome have appeared sporadically, but no increased risk has
    group developed the first widely accepted MDS classification, been documented in larger cohort studies.
    with five subgroups: refractory anaemia (RA), RA with ringed With careful examination, the percentage of children with MDS
    sideroblasts (RARS), RA with excess of blasts (RAEB), RAEB in and abnormalities of other organ systems (e.g. skeleton, heart, cen-
    transformation (RAEB-​ T), and chronic myelomonocytic leu- tral nervous system) and urogenital tract malformations, or signs
    kaemia (CMML). The World Health Organization (WHO) clas- and symptoms of immunodeficiency, rises sharply. Besides the clas-
    sification of 2001 elaborated on this and prompted a paediatric sical, inherited bone-​marrow failure disorders, several autosomal
    approach. Myelodysplastic and myeloproliferative disorders in dominant predisposition syndromes have recently been discovered
    children were separated into three main groups: JMML, MDS, and which often display a non-​haemological phenotype.
    ML-​DS. MDS was further subdivided into refractory cytopenia
    (RC), RAEB, and RAEB-​T. The change in nomenclature from RA to Myelodysplastic syndrome in inherited bone-​marrow
    RC reflects that anaemia is not a prerequisite for the diagnosis. The failure disorder
    revised WHO classification of 2008 recognized ML-​DS as a unique Inherited bone-​marrow failure disorders (IBMFDs) have an in-
    group and describes refractory cytopenia of childhood (RCC) as a creased risk of MDS/​AML. The risk varies greatly and is highest
    new provisional entity (Box 17.1). The most recent WHO classifi- in Fanconi anaemia and severe congenital neutropenia (SCN).
    cation of 2017 substitutes the term RAEB by MDS with excess of Diagnosing low-​grade MDS in a patient with IBFD is a challenge
    blasts (MDS-​EB; Box 17.1). because cytopenia and dysplasia may be part of the bone-​marrow
    Myelodysplastic Syndrome 135

    Box 17.1 Diagnostic categories of myelodysplastic and predisposition syndromes have recently been identified among
    myeloproliferative diseases in children according to the patients with ‘primary’ MDS. The most common of these con-
    paediatric approach to the World Health Organization (WHO) ditions is GATA2 deficiency. The phenotype of germline GATA2
    classification and revised classification mutations may lead to immunodeficiency with recurrent warts
    and monocytopenia (previously named MonoMAC syndrome),
    Myelodysplastic/​myeloproliferative disease
    or congenital deafness and lymphoedema (described as Emberger
    • Juvenile myelomonocytic leukaemia ( JMML) syndrome), but also to urogenital malformation or autism-​like
    Myeloid proliferations related to Down syndrome (DS) psychological problems. GATA2 deficiency can be diagnosed
    • Transient abnormal myelopoiesis (TAM) in 7% of all paediatric MDS and 15% of MDS-​EB. It is associ-
    • Myeloid leukaemia of Down syndrome (ML-​DS) ated with monosomy 7 and present in 70% of adolescents with
    Myelodysplastic syndrome (MDS) monosomy 7.
    • Refractory cytopenia (RCC) (PB blasts <2% and BM blasts <5%) Most recently, germline mutations in the paralogue genes SAMD9
    • MDS with excess blasts (MDS-​EB) (PB blasts 2–​19% or BM and SAMD9L have been described in children with pancytopenia
    blasts 5–​19%) with ataxia or a complex developmental disorder. These patients
    • MDS-​EB in transformation (PB or BM blasts 20–​29%)/​acute myeloid frequently develop MDS with monosomy 7 or del(7q) which may
    leukaemia (AML) with myelodysplasia-​related changes (PB or BM be transient but can progress to advanced MDS. Germline SAMD9/​
    blasts >20%) SAMD9L mutations can also underlie MDS in children without syn-
    PB: peripheral blood; BM: bone marrow dromic features.
    Source: data from Arber DA et al. ‘The 2016 revision to the World Health Organization There is a group of predisposition syndromes with germline mu-
    classification of myeloid neoplasms and acute leukemia’. Blood, Volume 217, Issue 20,
    pp. 2391–2405, (2016) American Society of Hematology. tations in RUNX1, ANKRD26, or ETV6 which present with isolated
    thrombocytopenia/​bleeding tendency prior to MDS development.
    Like GATA2 deficiency and SAMD9/​SAMD9L disease, the inheritance
    failure, and cytogenetic or molecular aberrations marking small pattern is autosomal dominant with de novo mutations. Collectively,
    clones may be temporary. The hierarchy of clonal events indicating these disorders cause many but not all cases of familial MDS.
    progression to MDS is quite specific for the individual IBMFDs.
    Clinical and laboratory features
    In Fanconi anaemia, myeloid neoplasia develops in a large frac-
    tion of patients during childhood or early adult life. The risk varies The presenting features in MDS are those of pancytopenia. Single-​
    according to genetic subgroup and associated abnormalities. Studies lineage cytopenia may occasionally be the predominant char-
    from the SCN International Register show a 15-​year accumulated acteristic. Foetal haemoglobin (HbF) is frequently moderately
    risk of MDS of 15%. Partial or complete loss of chromosome 7 is elevated, mean corpuscular volume (MCV) of red cells is high,
    found in more than half the patients who develop MDS. There is white blood-​cell (WBC) count is low to normal, and leukocy­
    no direct cause and effect relationship between the development tosis is generally not a feature of MDS. Some patients present with
    of MDS and growth factor therapy using granulocyte colony-​ moderate hepatosplenomegaly, but most have no organomegaly.
    stimulating growth factor (G-​CSF), but the risk of MDS is highest in Extramedullary myeloid tumour may be the presenting feature of
    patients with a poor response to G-​CSF. MDS, but blasts in the cerebrospinal fluid is not.
    MDS/​ AML develops in approximately 30% of those with
    Shwachman–​ Diamond syndrome (SDS) and is often associated Bone-​marrow features
    with chromosome 7 abnormalities of which isochromosome 7q may The bone marrow is hypocellular in most cases of RCC but
    re­present a separate entity with a long, stable clinical course and a low normocellular or hypercellular in most cases of MDS-​EB. Both per-
    risk of progression. Most SDS patients can be diagnosed by clinical ipheral blood and bone marrow display characteristic dysplastic
    features: screening for SBDS gene mutations in children diagnosed features with macrocytic and/​or megaloblastic erythropoiesis, bi-
    with primary hypoplastic RCC identifies only very few (less than 1%) zarre small or unusual large megakaryocytes, and dysgranulopoiesis
    SDS patients. (Figure 17.1). These characteristic dysplastic features are suggestive
    Patients with dyskeratosis congenita (DC) carry mutations of MDS but not diagnostic. A diagnosis of MDS should be confirmed
    in genes encoding components of the telomerase complex like by a repeat bone-​marrow examination after two to four weeks.
    DKC1, TERC, or TERT. These mutations impair telomerase ac-
    tivity, causing excessive telomere shortening and eventually Cytogenetics
    inducing cellular senescence and apoptosis. The cumulative in- An abnormal karyotype is found in 55% of children with primary
    cidence of MDS/​AML in DC may be similar to what has been MDS-​EB and in 76% with secondary MDS. The numerical abnor-
    reported in Fanconi anaemia. Many patients with DC cannot be malities dominate, and structural abnormalities are frequently part
    diagnosed by clinical features alone. Screening of a large cohort of of a complex karyotype with numerical abnormalities. This is in
    patients with RCC indicates that up to 5% of children with RCC contrast to AML where structural abnormalities are by far the most
    have unrecognized DC. frequent findings. Monosomy 7 is the most common cytogenetic
    abnormality in childhood MDS. It is seen in 25% of patients and
    Other predisposition syndromes
    not considered an unfavourable feature. Trisomy 8 and trisomy 21
    While MDS in IBMFD is conventionally named ‘secondary’ are the most common numerical abnormalities after monosomy
    MDS (like MDS following cytotoxic therapy), a number of 7. AML specific translocations (e.g. t(8;21)(q22; q22), t(15;17)(q22;
    136 CHAPTER 17 Myelodysplastic Syndrome

    Dysgranulopoiesis

    Normal Pseudo-Pelger- Macrocytosis Chromatin Hypo-, agranulation Asynchr. maturation


    segmented neutrophil Hüet anomaly clumping of cytoplasm nucleus - cytoplasm
    Dyserythropoiesis

    Normal Nuclear bridging Nuclear Multiple Cytoplasmic Macrocytic / megalo-


    erythroblast lobulation nuclei granules blastic changes
    Dysmegakaryopoiesis

    Normal Separated single Mikromega- Small binucleated Round, non-lobulated


    megakaryocyte nuclei karyocyte megakaryocyte megakaryocyte

    Figure 17.1 Cytological features of myelodysplasia.


    Reproduced courtesy of Irith Baumann, Institute for Pathology Kaufbeuren-Ravensburg, Germany

    q12), or inv(16)(p13q22)) should be considered as AML regardless autoimmune disease, and drug therapy. It can usually be separated
    of the blast count. from the typical histological pattern of RCC by islets of immature red
    cells, dysplastic megakaryocytes, and hypoplastic granulopoiesis.
    Molecular genetics Although dysplasia is essential for the diagnosis of MDS, it consti-
    MDS is a clonal disease arising in a stem or progenitor cell giving rise tutes only one aspect of the morphological diagnosis and may be
    to myelopoiesis, erythropoiesis, and megakaryopoiesis. The genes discrete. Since RCC cannot be distinguished from IBMFD, a diag-
    often mutated in adult MDS, such as TET2, DNMT3A, TP53,and nosis of MDS in IBMFD requires an increased blast count, a cyto-
    the spliceosome complex, are not involved in the pathogenesis in genetic or molecular abnormality known to be indicative for disease
    children. Instead, somatic driver mutations in SETBP1, ASXL1, progression, or an increase in bone-​marrow cellularity despite per-
    RUNX1, and the RAS oncogenes contribute to the genomic land- sistent peripheral blood (PB) cytopenia.
    scape of MDS in children. While one to four mutations are generally Ringsideroblasts in hypoplastic bone-​marrow disorders are ex-
    present in MDS-​EB, most cases of RCC do not exhibit any of these tremely rare in children. They may indicate a diagnosis of congenital
    somatic changes. ringsideroblastic anaemia or mitochondrial disease like Pearson
    syndrome. In MDS, they can indicate the presence of spliceosome
    Immunophenotype mutations.
    Immunophenotyping by flow cytometry has not the same diagnostic
    yield in MDS as in acute leukaemia, but may be of value in the quan- Refractory cytopenia of childhood versus
    titative and qualitative assessment of immature progenitor cells, aplastic anaemia
    especially when bone-​marrow smears are of suboptimal quality. As bone-​marrow cellularity is decreased in most cases of RCC, a tre-
    Flow-​cytometry immunophenotyping may be a useful addition to phine biopsy is fundamental for the evaluation of a child with sus-
    histopathology in the diagnosis of RCC. pected aplastic anaemia or MDS. In contrast to aplastic anaemia,
    hypoplastic RCC shows scarcely scattered granulopoiesis, patchy
    Differential diagnosis islands of immature erythropoiesis and micromegakaryocytes, or
    The two main diagnostic challenges are to distinguish RCC from absent megakaryopoiesis. Of note, the morphological picture of re-
    aplastic anaemia and other non-​clonal disorders including IBMF covering aplastic anaemia (e.g. under immunosuppressive therapy)
    syndromes, and also to differentiate MDS-​EB from de novo AML. cannot be distinguished from RCC. The morphology group of EWOG-​
    The traditional classification has been based on pure morphology, MDS have shown that after training of dedicated pathologists and con-
    but a number of additional factors need to be considered. sensus on dysplastic features, the inter-​observer variation is very low.

    Myelodysplastic syndrome versus non-​clonal disorders Separating myelodysplastic syndrome from acute
    Dysplasia of haematopoietic cells can be present in a variety of myeloid leukaemia
    disorders of very different aetiologies including infections, meta- The major differential diagnosis of MDS-​EB is AML. There are sig-
    bolic disorders and vitamin deficiencies, rheumatological and nificant differences between MDS and AML in clinical features,
    Myeloid leukaemia and Down syndrome 137

    Table 17.1 Major differences between myelodysplastic syndrome azacitidine in paediatric MDS-EB is limited. Immunosuppressive
    (MDS) and acute myeloid leukaemia (AML) in children therapy with antithymocyte globulin and cyclosporine in hypoplastic
    RCC has produced complete or partial response in 70%; the salvage
    Parameter MDS AML
    rate for those who fail is high with HSCT.
    WBC Low–​normal Low–​normal–​high
    Hepatomegaly Infrequent Common Acute myeloid leukaemia-​type chemotherapy
    Cytogenetic Numerical (-​7) Structural Conventional intensive chemotherapy without HSCT is unlikely
    aberrations to eradicate the primitive pluripotent cells involved in MDS, ren-
    Dysplasia Multi-​lineage Infrequent dering the therapy non-​curative in most patients. Induction chemo-
    Haematopoiesis Clonal (including CR) Non-​clonal in CR therapy is associated with significant morbidity and mortality, with
    Cell of origin Stem cell Lineage-​restricted a complete remission (CR) rate of less than 60%, treatment-​related
    Response to Poor Moderate
    mortality rate of 10–​30%, many relapses, and overall survival less
    chemotherapy than 30%. Children with monosomy 7, diagnosed as AML, have a
    Iatrogenic model Alkylating agents Epipodophyllotoxins poor response to induction chemotherapy as in MDS patients but,
    in contrast to MDS, those who respond well to chemotherapy have
    an outcome similar to AML patients. The EWOG-​MDS experiences
    cytogenetics, and in response to therapy. The latter reflect funda- showed poor response to chemotherapy in MDS-​EB and no benefit
    mental biological differences, thus making the morphologically from chemotherapy before HSCT.
    based classification a surrogate marker for the distinction between
    Haematopoietic allogeneic stem-​cell transplantation
    biological entities (Table 17.1). The morphologically defined MDS-​
    EB-​T group is heterogeneous and blast count in a single specimen Children with hypoplastic RCC may be successfully treated with a
    is insufficient to differentiate MDS from AML. Biological features, reduced-​intensity conditioning regimen, but for RCC with mono-
    rather than any arbitrary cut-​off in blast count, may be more im- somy 7 or MDS-​EB, a myeloablative regimen (e.g. with busulfan,
    portant in distinguishing MDS from (chemosensitive) AML. cyclophosphamide, and melphalan) is necessary and has cured more
    In borderline cases, with bone-​marrow blasts around 20% and than half of the children with MDS-​EB, both after matched family
    no cytogenetic clues to the diagnosis, the bone-​marrow examin- donor (MFD) and matched unrelated donor (MUD) HSCT. Total
    ation should be repeated after about two weeks. The major diag- body irradiation (TBI) has no superior anti-​leukaemic effect com-
    nostic pitfall may be associated with undue haste in starting therapy. pared with busulfan and is associated with more long-​term effects
    Significant organomegaly or increased WBC is suggestive of a diag- in children. Data from EWOG-​MDS on children with primary ad-
    nosis of AML. It should be emphasized that most children with mye- vanced MDS showed no benefit of intensive AML-​type therapy pre-
    loid malignancies have clear-​cut AML, some have MDS with low ceding HSCT. Considering the significant morbidity and mortality
    blast count, and only a small percentage have borderline features. of induction chemotherapy and the high rate of transplant-​related
    mortality following HSCT, highest in adolescents, most children
    Prognosis and natural course with MDS may benefit from HSCT as first-​line therapy, sparing the
    Children with RCC and MDS-​EB or even MDS-​EB-​T may show a long toxicity related to induction chemotherapy. Updated guidelines for
    and stable clinical course without treatment. Blood transfusions may the HSCT procedure may be found at www.ewog-​mds.org.
    only be required infrequently and severe infections are rarely seen. The Relapse following HSCT for advanced MDS is associated with
    condition may smoulder, with unchanged cytopenia, for months or a very grave outcome. Successful donor leukocyte infusions have
    even years. RCC with monosomy 7 is associated with a higher risk of occasionally been reported. Especially early relapse, detected by
    progression; once progression has occurred, the outcome is inferior. increasing mixed chimerism, may benefit from withdrawal of im-
    The International Prognostic Scoring System (IPSS), or other munosuppressive therapy and donor leukocyte infusion. Close
    scoring systems for adult MDS, weigh data on bone-​marrow blasts analyses of chimerism status post HSCT may allow initiation of pre-​
    count, cytopenia, and cytogenetics, to separate patients into prog- emptive immunotherapy.
    nostic groups. Overall, these scoring systems provide little diag- In contrast to advanced MDS, relapse following HSCT is rare in
    nostic information in children, but presence of a complex karyotype RCC. HSCT early in the course of the disease has therefore been re-
    is associated with a poorer outcome. commended for all children and adolescents with MDS. However, in
    Spontaneous regression of MDS has occasionally been reported in children with RCC and absence of profound cytopenia, postpone-
    the literature. The frequency of spontaneous remission is unknown, ment of HSCT, with a ‘watch and wait’ strategy, may be justified,
    but estimated to occur in well below 5% of patients, and may be re- especially in patients with a normal karyotype. A fludarabine-​based
    lated to SAMD9/​SAMD9L mutations. reduced-​intensity conditioning regimen has been piloted in a few
    children with RCC and normal karyotype, resulting in survival rates
    Treatment comparable to those of patients treated with myeloablative HSCT.
    MDS is a clonal disorder of an early stem or progenitor cell with limited
    residual non-​clonal stem cells. Therefore, haematopoietic stem-​cell
    transplantation (HSCT) is the main curative therapy approach for Myeloid leukaemia and Down syndrome
    children with all types of MDS. Hypermethylation may occur at a
    similar frequency in children and adults, thus making children po- Neonates with Down syndrome (DS) may have transient abnormal
    tential candidates for epigenetic therapy, however, the response to myelopoiesis (TAM), and DS individuals later have a strongly
    138 CHAPTER 17 Myelodysplastic Syndrome

    age-​related increased risk of myeloid leukaemia. The recognition 1st event 2nd event 3rd event
    of the unique biological features of GATA1-​mutated diseases in DS
    is reflected in the inclusion in the WHO classification of TAM and
    myeloid leukaemia of Down syndrome (ML-​DS).
    25% Myeloid
    TAM Regression
    leukaemia
    Transient abnormal myelopoiesis
    GATA1
    Increased WBC and circulating blasts with cell-​surface antigens mutation
    5–20%
    characteristic of megakaryoblasts, sometimes accompanied by an-
    1%
    aemia and thrombocytopenia, are seen in up to 10% of neonates Normal
    Myeloid
    leukaemia
    with DS. The percentage of blasts is often higher in blood than in
    Liver haematopoiesis Bone marrow haematopoiesis
    bone marrow, and bone-​marrow aspiration has limited additional
    Conception Birth Age 1–3 years
    diagnostic value. Clonal abnormalities are observed in 35%. The
    condition is referred to as transient abnormal myelopoiesis (TAM), Figure 17.2 Pathogenetic model for myeloid leukaemia in Down
    transient leukaemic reaction, or transient myeloproliferative dis- syndrome.
    order. The presentation is indistinguishable from leukaemia and
    some have therefore favoured the name ‘transient leukaemia’.
    TAM involves, selectively, the trisomic cells in individuals with DS megakaryocytic differentiation in accordance with the selective in-
    mosaicism. volvement of these two lineages in ML-​DS.
    Life-​threatening complications (mainly progressive hepatic dys- A model of the pathogenic steps in ML-​DS is presented in Figure
    function, coagulopathy, or pulmonary problems) may occur in 10–​ 17.2. Trisomy 21 is the first event that may predispose the cells to
    20% of patients with TAM, but spontaneous remission appears in the a proliferative advantage or further mutations. GATA1 mutation is
    majority within one to three months. Generally, no chemotherapy is found in TAM and is present in at least 5% of neonates with DS and
    indicated in TAM. However, in those with progressive hepatic or normal haematology. With more sensitive methods, GATA1 muta-
    pulmonary problems, a short course of low-​dose cytarabine may be tion is found in 20% of the neonates, in what may be called ‘silent
    very effective and life-​saving. Myeloid leukaemia develops one to TAM’. The mechanisms of the spontaneous regression of TAM re-
    three years later in 25% of children who recover from TAM. main unexplained but may be associated with the natural switch of
    haematopoiesis from foetal liver to bone marrow.
    Myeloid leukaemia of Down syndrome
    ML-​DS develops in 1% of the children with DS. A history of TAM is Clinical and laboratory features
    known in some 30%. The age distribution of ML-​DS is very remark- Isolated thrombocytopenia is often the presenting feature of ML-​
    able, with half being one year of age at diagnosis, one third being two DS. At diagnosis, both platelet count and WBC are lower than in
    years of age, and only 2% being more than four years of age. Only non-​DS patients and in contrast to the very high WBC seen in TAM.
    very few present before one year of age. There appears to be no age The blast cells have, in most cases, morphological and antigen fea-
    overlap between ML-​DS and TAM. tures of megakaryoblasts, although other morphological variants,
    ML-​DS may have a smouldering presentation with thrombocyto- especially erythroblasts, may occur. Many patients have a relatively
    penia and low blast count, and has been classified as MDS, while indolent course, characterized by a period of thrombocytopenia and
    others presenting with higher blast counts have been labelled AML. dysplasia, with relatively few symptoms and low blast count in the
    However, in contrast to non-​DS children, there are no biological or bone marrow.
    therapeutic differences between those presenting with a low or high
    blast count. Furthermore, the bone marrow in ML-​DS is often fi- Cytogenetics
    brotic and assessment of the blast count may be difficult. ML-​DS Numerical aberrations, mainly trisomy 8 and an extra chromosome
    is preferred to acute megakaryoblastic leukaemia (AMKL) because 21 (tetrasomy 21), are the most common acquired cytogenetic ab-
    other phenotypes in DS are observed sharing the same biological normalities. The recurrent structural aberrations seen in AML are
    and clinical characteristics. It is not appropriate to use the terms not a feature of ML-​DS. Karyotype has no strong prognostic impact.
    MDS or AML (or acute megakaryoblastic leukaemia) in young chil-
    dren with DS. There may be rare cases of DS patients older than four Treatment
    years of age who present with spontaneous MDS and AML without In contrast to TAM, ML-​DS is fatal if untreated. The prognosis
    GATA1 mutation. These older patients have a worse prognosis. of ML-​DS was considered very poor before 1990, when reports
    from the Nordic Society of Paediatric Haematology and Oncology
    Pathobiology (NOPHO) and the Pediatric Oncology Group (POG) showed a sur-
    Leukaemia in children with trisomy 21 mosaicism selectively in- prisingly high survival rate for DS patients receiving AML treat-
    volves the trisomic cells, pointing at the aetiological role of the ment. However, intensive induction therapy is associated with an
    additional chromosome 21 as the first hit in the multistep process increased risk of death during induction and consolidation. Less
    leading to leukaemia. An acquired mutation in GATA1 is found in time-​intensive therapy, allowing recovery prior to initiation of the
    both TAM and ML-​DS, and the identical mutation is found in TAM next chemotherapeutic course, is fundamental. HSCT in first remis-
    and ML-​DS in the same patient. GATA1 mutation is not found in sion is not indicated in children with ML-​DS. Several recent studies
    AKML or other AML patients without DS. The GATA1 gene en- have successfully lowered the cumulated doses of chemotherapy and
    codes a transcription factor essential for the normal erythroid and reported survival of 90%. Response evaluated by flow cytometry
    Juvenile myelomonocytic leukaemia 139

    may help identify those with the highest risk of relapse. The outcome bound state (Ras-​GTP) and an inactive guanosine diphosphate
    following relapse is also poor after HSCT. bound state (Ras-​GDP). Somatic point mutations in codons 12, 13,
    DS myeloblasts are ten-​fold more sensitive to cytarabine in vitro and 61 of NRAS or KRAS, impairing the intrinsic GTPase activity
    than non-​DS cells. The increased sensitivity of DS blasts may be of RAS and thereby causing high constitutive Ras-​GTP levels, are
    related to the expression of chromosome 21 localized genes like noted in leukaemic cells of about 25% of JMML patients.
    cystathionine-​β-​synthetase and superoxide dismutase. In rare cases, a developmental disorder with germline KRAS or
    NRAS mutations is accompanied by a myeloproliferative disorder
    clinically indistinguishable from JMML which, however, spontan-
    Juvenile myelomonocytic leukaemia eously regresses. Interestingly, a few cases of JMML associated with
    acquired somatic NRAS mutations have also been reported to re-
    JMML is a unique paediatric disorder previously named chronic gress spontaneously over time.
    myelomonocytic leukaemia (CMML) in childhood or juvenile
    chronic myeloid leukaemia (JCML), recognizing the distinction from Neurofibromatosis
    CML occurring in older children and adults. JMML is considered Neurofibromatosis (NF1) is associated with a more than 200-​fold
    a separate bridging disorder between MDS and myeloproliferative increased risk of JMML. A clinical diagnosis of NF1 is made in 10–​
    disorders in the WHO classification. 15% of children with JMML. NF1 is caused by mutations of the NF1
    tumour suppressor gene, which encodes the Ras-​GTPase activating
    Epidemiology protein (Ras-​GAP) neurofibromin. Children with NF1 and JMML
    Incidence studies from Denmark and British Columbia showed a have biallelic inactivation of the normal NF1 allele by uniparental di-
    similar JMML incidence of 1.2 per million children per year, corres- somy, resulting in duplication of the mutant NF1 allele or compound
    ponding to 2.4% of all haematological malignancies. However, an heterozygous mutations.
    incidence of only 0.6 per million was reported from the UK. The me-
    dian age at presentation is 1.8 years; 35% are below one year of age SHP-​2
    at presentation and only 4% are more than five years of age. JMML Somatic mutations in PTPN11, the gene encoding for the non-​
    displays a male predominance, with a male:female ratio of 2:1. receptor tyrosine phosphatase SHP-​2, are noted in about 35% of
    children with JMML. SHP-​2 relays growth signals from activated
    Clinical and haematological features growth factor receptors to other signalling molecules, including
    Patients present with pallor, fever, infection, bleeding, or symptoms Ras. Most PTPN11 mutations are predicted to disrupt the auto-​
    from the hepatomegaly, splenomegaly, or generalized lymphaden- inhibition of the catalytic phosphatase (PTPase) domain by the N-​
    opathy. A macular-​papular skin rash is seen in 35% of patients. An terminal src-​homology 2 (N-​SH2) domain, thereby promoting the
    elevated WBC with absolute monocytosis, anaemia, and thrombo- active conformation of the protein.
    cytopenia is almost universal. WBC at presentation exceeds 50×109/​l A transient myeloproliferative disorder is occasionally seen in
    in 30% of cases and is above 100×109/​l in 7%. Blood film appear- young infants with Noonan syndrome (NS). NS is an autosomal
    ance with immature myeloid and/​or erythroid precursors as well as dominant disorder characterized by short stature, distinct facial
    monocytosis is characteristic, while bone-​marrow monocytosis may anomalies, a typical spectrum of congenital heart defects, and de-
    be discrete. Blast count in the bone marrow is generally less than velopmental delays. Approximately 50% of patients with NS carry
    20%. Increased HbF is one of the main characteristics of JMML, with germline mutations of PTPN11. In contrast to JMML with somatic
    the notable exception being those with monosomy 7, who generally mutations in PTPN11, most cases with myeloproliferation in NS
    have a normal HbF for their age. resolve spontaneously. Normalization of myeloproliferation may
    take several months or even years, especially the monocytosis and
    Cytogenetics splenomegaly, which may persist for more than ten years. Fatal pro-
    Monosomy 7 (mostly as the sole abnormality) is present in about gressive disease has only been reported in a few patients with NS.
    25% of JMML; 10% have other aberrations, and 65% have a normal
    karyotype. Data from the EWOG-​MDS did not show any major CBL
    clinical differences between JMML in patients with and without CBL is an E3 ligase which promotes ubiquitylation and thereby
    monosomy 7. de­gradation of a large number of tyrosine kinases. CBL mutations
    are acquired somatically in adults with myeloproliferative disorders.
    Molecular diagnostics and pathophysiology In JMML, 10–​15% of the children have a developmental germline
    JMML is a clonal disorder that arises from a pluripotent stem cell. disorder with heterozygous CBL mutations and a high percentage
    The pathophysiology is characterized by hyperactivation of the Ras/​ of developmental delay, cryptorchidism, and impaired growth.
    MAPK signal transduction pathway. Germline and somatic muta- CBL appears to function as a classic tumour suppressor, with loss
    tions that deregulate Ras signalling are implicated as key initiating of heterozygosity in leukaemic cells. Clinical data reveals that most
    events in JMML. Approximately 90% of patients have disease-​ patients with CBL syndrome and JMML experience spontaneous
    initiating mutations in PTPN11, NF1, CBL, NRAS,or KRAS. resolution of their disease but are at risk to develop vasculopathies.

    Ras Differential diagnoses


    Members of the Ras family of signalling proteins regulate cellular JMML may mimic infections like Epstein–​Barr virus, cytomegalo-
    proliferation by cycling between an active guanosine triphosphate virus, herpes virus-​6 and parvovirus, Wiskott–​Aldrich syndrome,
    140 CHAPTER 17 Myelodysplastic Syndrome

    leukocyte adhesion defects, or infantile malignant osteoscleosis.


    However, since the discovery of the underlying lesions, more than FURTHER READING
    80% of children with JMML can be diagnosed by a typical clinical Aalbers AM, et al. (2015) Bone marrow immunophenotyping by flow
    picture and molecular markers. cytometry in refractory cytopenia of childhood. Haematologica
    100, 315–​23.
    Natural course and prognostic factors Arber DA, et al. (2016) The 2016 revision to the World Health
    For patients carrying somatic RAS and PTPN11 mutations or NF1, Organization classification of myeloid neoplasms and acute leu-
    kemia. Blood 127, 2391–​405.
    JMML is a rapidly fatal disorder if left untreated. Blastic transform-
    Baumann I, et al. (2012) Morphological differentiation of severe
    ation is infrequent with JMML and most untreated patients die from
    aplastic anaemia from hypocellular refractory cytopenia of child-
    organ failure due to infiltration of the leukaemic cells. Low platelet
    hood: reproducibility of histopathological diagnostic criteria.
    count, age above two years, and high HbF at diagnosis are the main Histopathology 61(1), 10–​17.
    factors predicting a short survival. Acquisition of secondary muta- Caye A, et al. (2015) Juvenile myelomonocytic leukemia displays mu-
    tions in SETPB1 and JAK3, or an additional acquired lesion of the tations in components of the RAS pathway and the PRC2 network.
    RAS pathway, generally infers a poor outcome. Gene-​expression Nat Genet 47(11), 1334–​40.
    studies and aberrant DNA methylation separate JMML into sub- Chang TY1, et al. (2014) Bedside to bench in juvenile myelomonocytic
    groups with distinct prognoses. leukemia: insights into leukemogenesis from a rare pediatric leu-
    Patients with germline mutations in PTPN11 often show kemia. Blood 124(16), 2487–​97.
    spontaneous regression of their JMML-​like myeloproliferation. Gohring G, et al. (2010) Complex karyotype newly defined: the
    Therefore, mutational studies have to be carried out in haemato- strongest prognostic factor in advanced childhood myelodysplastic
    poietic and non-​haematopoietic tissue (fibroblasts or cells from syndrome. Blood 116, 3766–​99.
    hair bulbs). Hasegawa D (2016) The current perspective of low-​ grade
    myelodysplastic syndrome in children. Int J Hematol 103(4), 360.
    Treatment Hasle H, Niemeyer CM (2011) Advances in the prognostication and
    management of advanced MDS in children. Br J Haematol 154,
    Intensive chemotherapy is mostly unsuccessful in JMML, resulting
    185–​95.
    in long-​term survival of less than 10%. While there is no evidence Hasle H, et al. (2008) Myeloid leukemia in children 4 years or older
    that cyto-​reduction prior to HSCT increases long-​term survival, with Down syndrome often lacks GATA1 mutation and cytogen-
    some investigators propose high-​dose ara-​C and fludarabine for etics and risk of relapse are more akin to sporadic AML. Leukemia
    children with a large tumour burden. Purine analogues, etoposide, 22, 1428–​30.
    and cytarabine as single agents may produce short term clinical re- Klusmann JH, et al. (2008) Treatment and prognostic impact of tran-
    sponse and azacitidine seems to provide clinical benefits in a consid- sient leukemia in neonates with Down syndrome. Blood 111(6),
    erable number of patients pre HSCT. 2991–​8.
    Allogeneic HSCT is the only curative approach for children with Lipka DB, et al. (2017) RAS-​pathway mutation patterns define epigen-
    JMML and somatic mutations in PTPN11 or NF1, and most chil- etic subclasses in juvenile myelomonocytic leukemia. Nat Commun
    dren with KRAS or NRAS mutations. It results in overall survival 8(1), 2126.
    in more than half the patients after both MFD and MUD HSCT or Locatelli F, Niemeyer CM (2015) How I treat juvenile myelomonocytic
    leukemia. Blood 125(7), 1083–​90.
    cord blood transplantation. Busulfan-​based myeloablative therapy
    Locatelli F, et al. (2005) Hematopoietic stem cell transplantation
    offers similar or greater anti-​leukaemic efficacy than TBI, which is
    (HSCT) in children with juvenile myelomonocytic leukemia
    to be avoided because of deleterious radiation-​induced late effects (JMML): results of the EWOG-​ MDS/​ EBMT trial. Blood 105,
    in these young patients. The current EWOG-​MDS study uses a pre- 410–​19.
    parative regimen with busulfan, cyclophosphamide, and melphalan, Niemeyer CM, et al. (2015) Criteria for evaluating response and out-
    and has shown event-​free survival (EFS) of around 50%, with no dif- come in clinical trials for children with juvenile myelomonocytic
    ference between related and unrelated donors. Splenectomy before leukemia. Haematologica 100, 17–​22.
    HSCT, as well as spleen size at time of the allograft, did not appear Niemeyer CM, et al. (2010) Germline CBL mutations cause develop-
    to have an impact on post-​transplantation outcome. Monosomy 7 is mental abnormalities and predispose to juvenile myelomonocytic
    associated with an outcome comparable to or even better than that leukemia. Nat Genet 42, 794–​800.
    of patients with normal karyotype. Pastor V, et al. (2017) Mutational landscape in children with
    Disease recurrence, generally within the first year, remains the myelodysplastic syndromes is distinct from adults: specific somatic
    main cause of treatment failure. Age of above four years is the drivers and novel germline variants. Leukemia 31, 759–​62.
    major significant risk factor for failure in multivariate analysis. Roberts I, Izraeli S (2014) Haematopoietic development and leukaemia
    in Down syndrome. Br J Haematol 167, 587–​99.
    Reduced intensity and duration of graft-​versus-​host (GvH) disease
    Roberts I, et al. (2013) GATA1-​mutant clones are frequent and often
    prophylaxis may significantly contribute to successful leukaemia
    unsuspected in babies with Down syndrome: identification of a
    control, and both acute and chronic GvH disease is associated with population at risk of leukemia. Blood 122(24), 3908–​17.
    a lower risk of relapse. Early detection of donor cells by increasing Sakaguchi H, et al. (2013) Exome sequencing identifies secondary mu-
    mixed chimerism may be successfully eradicated by reducing tations of SETBP1 and JAK3 in juvenile myelomonocytic leukemia.
    ongoing immunosuppressive therapy. Donor lymphocyte infu- Nat Genet 45(8), 937–​41.
    sion (DLI) in JMML relapse is largely unsuccessful, but a second Stieglitz E, et al. (2015) Subclonal mutations in SETBP1 confer a poor
    or even third transplantation gives a relatively high chance of long-​ prognosis in juvenile myelomonocytic leukemia. Blood 125(3),
    term survival. 516–​24.
    Juvenile myelomonocytic leukaemia 141

    Strahm B, et al. (2011) Hematopoietic stem cell transplantation for and non-​Down syndrome blast cells and relationship to in vitro
    advanced myelodysplastic syndrome in children: results of the sensitivity to cytosine arabinoside and daunorubicin. Blood 94(4),
    EWOG-​MDS 98 study. Leukemia 25(3), 455–​62. 1393–​400.
    Taga T, et al. (2016) Preserved high probability of overall survival with Uffmann M, et al. (2017) Therapy reduction in patients with Down
    significant reduction of chemotherapy for myeloid leukemia in syndrome and myeloid leukemia: the international ML-​DS 2006
    Down Syndrome: a nationwide prospective study in Japan. Pediatr trial. Blood 129(25), 3314–​21.
    Blood Cancer 63(2), 248–​54. Wlodarski MW, et al. (2016) Prevalence, clinical characteristics and
    Taub JW, et al. (1999) Expression of chromosome 21-​localized genes prognosis of GATA2-​related MDS in children and adolescents.
    in acute myeloid leukemia: differences between Down syndrome Blood 127, 1387–​97.
    18
    Non-​Hodgkin Lymphomas in Children
    and Adolescents
    Véronique Minard-​Colin and Catherine Patte

    Introduction Incidence and epidemiology

    Non-​ Hodgkin’s lymphoma (NHL) is a heterogeneous group of The overall incidence and frequency of the different histological sub-
    lymphoid malignancies. The histological classification of these dis- groups varies according to age at diagnosis. Data from the American
    eases has changed many times over the years due to a better under- National Cancer Institute’s Surveillance, Epidemiology, and End
    standing of lymphomagenesis and to the development of new Results (SEER) programme show a steady increase in NHL with
    diagnostic tools. The World Health Organization (WHO) classi- age. The annual incidence per million inhabitants ranges from 5.9
    fication is now widely used, providing clinicians with a common in children below five years of age to about 10 in children between
    language, and its revision in 2016 allowed identification of new/​pro- the ages of five and fourteen, and to 15 in adolescents (and about 150
    visional entities in paediatric lymphomas. in adults). The increase in adolescents is essentially related to the
    NHL is the fourth most common malignancy in children (and higher incidence of large-​cell lymphomas at that age. The vast ma-
    the most common in sub-​Saharan Africa) and can arise at any jority of childhood NHL is high-​grade lymphoma, mostly of B-​cell
    age. NHLs in children are rarer than in adults and exhibit signifi- origin. Burkitt lymphoma (BL) is by far the most frequent (50–​60%,
    cant differences in terms of the distribution of histological subtypes depending on the geographic region), followed by lymphoblastic
    and clinical features, with virtually exclusively diffuse high-​grade lymphoma (LL) (20–​25%), diffuse large B-​cell lymphoma (DLBCL)
    lymphomas and frequent extranodal involvement. Dramatic pro- (10–​15%), and anaplastic large-​cell lymphoma (ALCL) (10–​12%)
    gress has been achieved in the treatment of NHL in children and (Table 18.1). Other lymphomas, such as peripheral T-​cell lymphoma,
    adolescents, yielding an overall survival rate exceeding 80%. The extranodal natural killer (NK)-​cell/​T-​cell lymphoma, or paediatric
    progress in outcome is the consequence of a more accurate diag- follicular lymphomas may arise in children, but represent less than
    nostic process, of better risk-​adapted chemotherapy approaches, 5% of cases. Both the incidence and relative frequency of the dif-
    and of the ability to prevent and control life-​threatening treatment-​ ferent subtypes of NHL also vary considerably in different parts
    related toxicities. Due to the rarity of each subtype, such progress of the world; in equatorial Africa, lymphomas account for almost
    could not have been achieved without international collaborations, half of the childhood cancers, reflecting the very high incidence of
    especially through the European Intergroup for Childhood NHL Burkitt lymphoma but a relative paucity of LL in this region.
    (EICNHL) comprising most European countries, Japan, and Hong The aetiology of NHL is largely unknown. Immunodeficiency,
    Kong, as well as collaborations between Europe and North America. whether inherited or acquired, is clearly related to the development
    Conversely, the outcome of patients with NHL treated in countries of NHL. More recently described is the increased risk of NHL in
    with limited resources is far worse, due to the unavailability of cer- the constitutional mismatch repair MMR d ​ eficiency syndrome, due
    tain anti-​cancer medicines and to the suboptimal diagnostic and to biallelic germline mutations of MMR genes and characterized
    therapeutic management of patients. by multiple cancers including NHL (mainly LL), brain tumours,
    Increased knowledge in the field of biology of paediatric NHL and colorectal cancers in children with café au lait spots. The re-
    will improve outcome through the use of more efficient multi-​agent lationship between Epstein–​Barr virus (EBV) and the pathogenesis
    therapies and the introduction of novel agents, including targeted of NHL began with the elegant investigations by Denis Burkitt in
    therapies. Nevertheless, conditions that allow the application of the late 1950s. Burkitt demonstrated that the disease was associated
    standard treatments on a global or international scale are needed, to with climate and rainfall—​in short, the regions where malaria is
    provide a better chance of cure in less privileged areas of the world. holoendemic. This observation led to a collaboration with Anthony
    These challenges will represent the new frontiers in paediatric NHL Epstein and resulted in the discovery of a herpesvirus from cultures
    for the next decades. of endemic BL tissue, the first human tumour-​associated virus.
    Pathology and biology 143

    Table 18.1 Immunophenotypical, cytogenetic, and molecular markers of paediatric non-​Hodgkin lymphoma*

    Histology (frequency) BL (50–​60%) LBCL (10–​15%) LL (20–​25%) ALCL (10–​12%)


    Subtype DLBCL PMBL B−​LL T−​LL
    Immunohistochemistry
    MIB1 ~100% 40–​90% 30–​90% Not informative Not informative Not informative
    CD10 + +/​−​ −​/​+ +/​−​ +/​−​ −​
    CD19 + + + + −​ −​
    CD20 + + + +/​−​ −​ −​
    CD79a + + + + −​ −​
    sIg + +/​−​ −​ −​ −​ −​
    Bcl−​6 + +/​−​ +/​−​ −​ −​ +/​−​
    MUM1 −​ −​/​+ + −​ −​ −​
    MAL −​ −​/​+ + −​ −​ −​
    TdT −​ −​ −​ + + −​
    cCD3 −​ −​ −​ −​ + −​/​+
    CD4 −​ −​ −​ −​ +/​−​ −​/​+
    CD8 −​ −​ −​ −​ +/​−​ −​/​+
    CD7 −​ −​ −​ −​ + −​/​+
    CD5 −​ −​/​+ −​ −​ +/​−​ +/​−​
    CD30 −​ § +/​−​ −​ −​ +
    ALK −​ £ −​ −​ −​ +
    Cytogenetic t(8;14)(q24;q32) R8q24 (~30%) Few data Few data Translocations t(2;5)(p23;q35) >90%
    t(2;8)(p12;q24) involving 14q11–​13 or variants involving
    t(8;22)(q24;q11) Few other data 2p23
    Molecular biology MYC/​IGH NF−​kappa IGH/​TCR IGH/​TCR NPM/​ALK >90%
    IGK/​MYC B pathway rearrangements rearrangements or variants
    MYC/​IGL dysregulation NOTCH/​FBXW7,
    PTEN

    BL, Burkitt lymphoma; LBCL, large B−​cell lymphoma; DLBCL, diffuse large B−​cell lymphoma; PMBL, primary mediastinal (thymic) large B−​cell lymphoma; B−​LL, B lymphoblastic
    lymphoma; T−​LL, T lymphoblastic lymphoma; ALCL, anaplastic large−​cell lymphoma, ALK−​positive; +, >90% of cases; +/​−​​, >50% of cases; −​​/​+,<50% of cases; −​, <10% of cases.
    §, positive especially in the rare anaplastic variant; £, positive in the ALK−​positive LBCL
    * Other non−​Hodgkin lymphomas, such as peripheral T−​cell lymphoma, extranodal NK−​/​T−​cell lymphoma, nasal type, and paediatric follicular lymphoma, represent <5% of
    paediatric NHL and are not indicated in the table.
    Reproduced with permission from Minard−​Colin, V. et al. ‘Non−​Hodgkin Lymphoma in Children and Adolescents: Progress Through Effective Collaboration, Current Knowledge, and
    Challenges Ahead.’ Journal of Clinical Oncology, Volume 33, Issue 27, pp. 2963–​74. Copyright © 2015 American Society of Clinical Oncology. doi:10.1200/​JCO.2014.59.5827

    monoclonal surface immunoglobulin (sIg). Because BL is a very


    Pathology and biology
    highly-​proliferating human malignancy, with doubling time of less
    than 24 hours, analysis of proliferation markers such as Ki-​67 or
    Mature B-​cell non-​Hodgkin lymphomas
    MIB-​1 will show reactivity in more than 95% of the cells. BL cells
    Burkitt lymphoma (BL) accounts for more than 80% of childhood harbour characteristic chromosomal translocations involving the
    B-​cell NHL (Table 18.1). Morphologically, it is characterized by cells MYC oncogene locus on chromosome 8q24. The most common
    of intermediate size with nuclei containing prominent nucleoli and translocation, t(8;14)(q24;q32), which involves the IgH gene locus,
    limited amount of basophilic cytoplasm in which vacuoles are al- occurs in 80% of cases. The consequence of MYC rearrangements in
    most invariably present. Tissue sections often show the ‘starry-​sky’ BL is the deregulation of MYC expression, which promotes cell-​cycle
    morphology resulting from clear macrophages scattered among the progression and cell transformation. The remaining 20% of BL cases
    rather intensely stained population of blasts. The current revised have either a t(2;8)(p12;q24) or, less frequently, a t(8;22)(q24;q11)
    WHO classification of lymphoid malignancies also includes, under translocation involving MYC and either the kappa or lambda im-
    the term BL, the subtypes previously defined as Burkitt-​like and munoglobulin light-​chain gene loci, respectively, on chromosomes
    atypical Burkitt lymphomas, since it is not possible to differentiate 2 or 22. While MYC is involved in both sporadic and endemic BL,
    them based solely on morphology. Those tumours which present the breakpoints within the MYC gene and in the IgH loci differ be-
    with features intermediate between BL and DLBCL are now placed tween the two BL subtypes. The most common translocation, t(8;14)
    into the category of unclassifiable B-​cell lymphoma. (q24;q32), can be detected by long-​distance polymerase chain re-
    Immunophenotypical studies of BL demonstrate mature ma- action (LD-​PCR) assay. Based on this approach, the MYC-​IgH
    lignant B cells expressing CD19, CD20, CD22, CD10, CD79, and rearrangements can be used as markers to analyse minimal BM
    144 CHAPTER 18 Non-Hodgkin Lymphomas in Children and Adolescents

    infiltration at diagnosis and its response to chemotherapy with a paediatric-​type follicular lymphoma (FL) and other DLBCLs. It
    sensitivity of 10-​4. occurs most commonly in children and young adults, and typic-
    One controversial issue, not fully resolved, is whether true BL ally develops in Waldeyer’s ring and/​or cervical lymph nodes and
    without MYC translocations really exists. Some recent studies is low stage. Usually, strong IRF4/​MUM1 expression is seen, with
    have identified a subset of lymphomas that resemble BL morpho- BCL6 and a high proliferative fraction. Most cases have IG/​IRF4 re-
    logically, to a large extent phenotypically, but which lack MYC re- arrangements, sometimes together with BCL6 rearrangements, but
    arrangements. Instead, they have a chromosome 11q alteration they uniformly lack BCL2 rearrangements. This lymphoma is con-
    characterized by proximal gains and telomeric losses. Compared to sidered to be more aggressive than other paediatric-​type FLs, but
    BL, these lymphomas have more complex karyotypes, lower levels patients, at least when treated, have done very well.
    of MYC expression, a certain degree of cytological pleomorphism, A rare (1–​2% of all paediatric NHL) and clinically distinct entity
    occasionally a follicular pattern, and, frequently, a nodal presenta- among large B-​cell lymphomas is primary mediastinal large B-​cell
    tion. The revised 2016 WHO classification considered this a new lymphoma (PMBL). It originates from thymic B cells and is char-
    provisional entity designated as ‘Burkitt-​like lymphoma with 11q acterized by diffuse proliferation of large cells and evident scler-
    aberration’. osis. Due to both the almost exclusive mediastinal localization and
    Recent findings with new biological tools—​ such as next-​ the morphological appearance, PMBL may be difficult to distin-
    generation sequencing, microRNA and gene expression pro- guish from Hodgkin lymphoma. Cell surface markers in PMBL are
    filing, and single nucleotide polymorphism (SNP) array similar to those seen in DLBCL and include CD19, CD20, CD22,
    analysis—​could also lead to novel and more effective therapies. CD79, and PAX5, but, very often, only cytoplasmic Ig and not SIg
    In BL, genetic alterations other than the 8q24/​MYC translocation are present. CD30 is commonly expressed. Studies of gene expres-
    participate in the deregulation of the MYC pathway and may be sion show overlap of gene-​expression profiles between PMBL and
    essential for disease initiation and progression. Recurrent mu- HL suggesting that, biologically, PMBL is more similar to HL than
    tations in the PI3K pathway suggest a cooperative role between to DLBCL.
    MYC over-​expression and deregulation of PI3K signalling. Other
    studies highlight the pathogenic and clinical relevance of muta- Lymphoblastic lymphoma
    tions affecting the transcription factor TCF3 (E2A) or its nega- Lymphoblastic lymphoma (LL) accounts for 1520% of paediatric
    tive regulator ID3, a novel tumour-​suppressor gene, with TCF3 or NHL (Table 18.1). Conventionally, LL is distinguished from acute
    ID3 mutations affecting nearly 70% of sporadic BL. Other muta- lymphoblastic leukaemia (ALL) based on the extent of bone-​
    tions in genes including CCND3, GNA13, TP53, and SMARCA4 marrow (BM) involvement (ALL is defined when BM blasts are
    have also been described. more than 25%), but gene expression and genetic studies would
    Interestingly, a comparison between the frequency of mutations suggest that they may differ biologically. LL includes both pre-
    in BL and adult germinal centre B-​cell (GCB) DLBCL revealed a cursor T (T-​LL) and B (pB-​LL) NHL. Unlike ALL, with predom-
    stronger overlap, further supporting the notion that BL and GCB inant pB-​cell diseases, most LLs are of T-​cell origin (80–​90% versus
    DLBCL both originate from germinal centre B cells. Altogether, 10–​20%).
    these findings also support the development of drugs targeting the Morphologically, the malignant cells of LL appear indistinguish-
    PI3K kinase pathway, antigen-​independent B-​cell receptor (BCR) able from ALL, being of small or intermediate size, with high nu-
    signalling, and cyclin D3/​CDK6 pathways in childhood B-​NHL. clear/​cytoplasm ratio and finely dispersed chromatin. Nucleoli may
    Similarly to BL, DLBCL is characterized by a mature B-​cell pheno- be evident and, occasionally, blasts may have cytoplasmic vacuoles
    type, but cells display varying morphological and cytological appear- morphologically resembling L3 blasts (a feature that is not exclusive
    ances. Malignant cells are rather large in size and show more abundant of BL). LL blasts in the lymph nodes show a high mitotic rate and
    cytoplasm compared to BL cells. Paediatric DLBCLs have a high cause diffuse absence of nodal architecture, with occasionally spared
    proliferation rate (although lower than BL) as determined by Ki-​67 reactive follicles.
    or MIB-​1 (i.e. less than 90%). They express cell sIg and a number of Immunophenotypically, LL blasts do not express sIg but, char-
    B-​cell markers, including CD19, CD20, CD22, CD79, and PAX5. In acteristically, are positive for terminal deoxynucleotidyl trans-
    contrast to adult DLBCL, paediatric cases are less frequently positive ferase (TdT) in more than 95% of cases and often for CD99,
    for BCL2, but can express high levels of the germinal centre markers independently of their T-​or B-​cell lineage. TdT expression is the
    CD10 and BCL6. No consistent cytogenetic abnormalities have been best immunohistochemical marker for determining the precursor
    associated with DLBCL of children and adolescents, but more complex origin of the proliferation. This is very useful for differentiating
    karyotypes than BL have been reported. Chromosomal translocations pB-​LL from mature B-​cell NHL, since they can express other
    involving the MYC oncogene are more frequent in paediatric com- similar markers (CD79a, CD10, CD19, +/​-​CD20) (Table 18.1). T-​
    pared to adult DLBCL, whereas BCL2 translocations are rarely seen LLs express, in different combinations, T-​cell markers typical of
    in paediatric cases. Recent gene-​expression profile studies, together middle to late stages of thymocyte differentiation, including CD1a,
    with specific immunostaining using markers of germinal centre, have CD2, CD5, CD7, CD4, and/​or CD8. T-​cell receptor antigens are
    demonstrated that more than 80% of the paediatric DLBCLs belong frequently not expressed and CD3 is more often detected in the
    to the germinal centre B-​cell-​like subgroup, suggesting that DLBCLs cytoplasm than on the cell surface. pB-​LLs usually express CD19
    of children and adolescents are biologically more similar to BL than to and CD10, and have variable expression of HLA-​DR and CD20,
    DLBCL seen in older patients. but not sIg.
    Large B-​cell lymphoma with IRF4 rearrangement is a new pro- Although specific data on LL are scarce, chromosomal transloca-
    visional entity in the WHO classification to distinguish from tions described in ALL have been reported also in LL. Translocations
    Staging 145

    often involve rearrangements of T-​cell receptor (TCR) or Ig gene and cytotoxic molecules like TIA1, granzyme B, or perforin. Most
    rearrangements with proto-​oncogenes, resulting in the juxtapos- cases express at least one T-​cell antigen and exhibit clonal rearrange-
    ition of promoter and enhancer elements of TCR or Ig genes, and ment of the T-​cell receptor.
    transcription factor genes including HOX11, TAL1, and LYL1. T-​ Accumulating evidence indicates that the immune system plays
    LL displays early TCR gene rearrangements that can be detected a major role in both the pathogenesis and final control of ALK-​
    by molecular assays, whereas B-​cell precursor LL usually shows Ig positive ALCL. Antibodies against ALK and cytotoxic T-​cell (CTL)
    gene rearrangements. They represent a unique marker that can be and CD4 T-​helper responses to ALK have been detected in patients
    exploited through the use of molecular assays to determine clonality, with ALK-​positive ALCL, both at diagnosis and during remission,
    thus assisting in the differential diagnosis between LL and atypical with a significant inverse correlation between ALK-​antibody titres
    or reactive lymphadenopathy. Unlike ALL, the prognostic value of and the incidence of relapses. The presence of germline monoallelic
    minimal disseminated disease (MDD) and minimal residual dis- mutations of the perforin gene has been demonstrated in patients
    ease (MRD) using either polymerase chain reaction (PCR)-​based with ALCL, suggesting that impairment of the cytotoxic machinery
    methods, molecular techniques, or flow cytometry, has not yet been may predispose to ALCL.
    proven in LL (unlike ALL).
    Several biological characteristics considered of prognostic rele- Rare non-​Hodgkin lymphomas in paediatrics
    vance in T-​ LL have been described: in the Berlin–​ Frankfurt–​ Paediatric FL has become a definite entity in the WHO 2016 classi-
    Munster (BFM) series, loss of heterozygosity (LOH) on chromosome fication but is already known as paediatric-​type FL because similar
    6q, found in 12% of the patients, was significantly associated with lymphomas may occur in adults. It is a nodal disease characterized by
    a poorer outcome (event-​ free survival (EFS): 27%±9% versus large, expansile, highly-​proliferative follicles that often have prom-
    86%±3% if no LOH6q16), while the presence of NOTCH1 muta- inent blastoid follicular centre cells, rather than classic centroblasts
    tions (60% of patients) was associated with a favourable outcome (or centrocytes). Some have reported a moderate number of cases
    (EFS 84%±5% versus 66%±7%). NOTCH1 mutations and LOH6q16 as grade 1–​2 (of 3). BCL2 rearrangements must not be present, but
    were almost mutually exclusive. In a French series studying FLASH there may be some BCL2 protein expression. They also lack BCL6
    deletion on chromosome 6q, TCR rearrangements and NOTCH1/​ and MYC rearrangements, with ongoing investigations of their gen-
    FBXW7 mutations (the last events found in 55% of cases) were the etic/​molecular landscape. Nearly all cases are localized and may not
    only prognostic factors in a multivariate analysis, with an EFS rate require treatment other than excision. The criteria for paediatric-​
    of 96%±0.04% versus 45%±0.13% (p<0.01). Finally, the prognostic type FL, however, must be strictly applied to avoid underdiagnosing
    value of an RAS/​PTEN alteration and of the absence of biallelic TCR conventional grade 3 FL, with particular caution required before
    gamma deletion also warrants evaluation in T-​LL. making this diagnosis in an adult. This category also excludes cases
    with diffuse areas (i.e. foci of DLBCL).
    Anaplastic large-​cell lymphoma Some studies have raised the possibility that paediatric-​type FL
    Anaplastic large-​cell lymphoma (ALCL) accounts for 10–​15% of might be a ‘benign clonal proliferation with low malignant potential’.
    childhood/​adolescent NHL (Table 18.1). More than 90% of sys- Recently, genome analysis of paediatric-​type FL revealed unique
    temic ALCLs in children carry a translocation resulting in the fu- molecular features (i.e. low genetic complexity, recurrent alterations
    sion of the anaplastic lymphoma kinase (ALK) gene to one of several of TNFRSF14 gene, and high prevalence of MAPK pathway muta-
    partners. The most common ALK fusion protein is nucleophosmin tions) that define it as a biologically and clinically distinct indolent
    (NPM)-​ALK resulting from the t(2;5) translocation. The fusion pro- lymphoma of children and adults.
    tein can be detected using fluorescent in situ hybridization (FISH), Other rare B-​cell lymphomas are nodal and extranodal (mucosa-​
    reverse transcription-​ polymerase chain reaction (RT-​ PCR), or associated) marginal-​ zone lymphomas (MZLs). These entities
    immunohistochemistry. Detection of the NPM-​ALK transcript is histologically overlap with the adult lymphomas and may be difficult
    increasingly applied in the diagnosis of ALCL but, most import- to differentiate from reactive lymphocytic proliferations. Another
    antly, is exploited in the assessment of MDD that has been shown to rare group of paediatric NHL is represented by the peripheral T-​cell
    be one of the most relevant prognostic parameters in ALK-​positive lymphomas (PTCLs). Except for ALCL, the not otherwise specified
    ALCL of childhood. Indeed, less than 15% of patients exhibit cyto- (NOS) category of PTCL is the most common. Morphologically, they
    logically detectable BM disease, whereas RT-​PCR for NPM-​ALK have variable appearances with medium-​to large-​size neoplastic
    can detect MDD in blood or BM in around 50% of patients at diag- cells, with irregular nuclei and basophilic cytoplasm, infiltrating the
    nosis. A number of variant partners of ALK have been reported in lymph node. Cells show aberrant expression of T-​cell markers (CD2,
    the ALK gene rearrangements and they are often characterized by a CD3, CD5, or CD7) and may co-​express or lack both CD4 and CD8.
    specific pattern of cellular ALK expression, as detected by anti-​ALK TCR rearrangements are almost invariably detectable.
    antibodies.
    Characteristically, ALCL cells are large, pleomorphic,
    multinucleated, or horse-​shoe-​shaped cells (hallmark cells). They Staging
    are present in differing proportions in the various morphological
    subtypes of ALCL. The most frequent morphological subtype (more The Ann Arbor classification is not adapted to childhood NHL be-
    than 70% of cases) is the ‘common type’ in which typical anaplastic cause of predominantly extranodal primary sites. The usual staging
    and hallmark cells predominate. The lymphohistiocytic and small-​ classification is the St Jude’s classification in which mediastinal and
    cell variants account for about 20% of cases and appear to be clinic- extensive abdominal lesions are classified as stage III, and stage IV is
    ally more aggressive. ALCL express CD30 and, in most cases, EMA restricted to bone marrow (BM) and central nervous system (CNS)
    146 CHAPTER 18 Non-Hodgkin Lymphomas in Children and Adolescents

    involvement regardless of involvement of other visceral sites. The Box 18.1 International paediatric non-​Hodgkin lymphoma
    International Prognostic Index (IPI) is not used in childhood NHL staging system
    due to a difference in staging and because the performance status
    index does not reflect quality-​of-​life deterioration but, rather, the Stage I
    rapidity of growth and the tumour volume. • Single tumour with exclusion of mediastinum and abdomen (N; EN;
    The role of fluorodeoxyglucose (FDG)-​positron emission tom- B or S: EN-​B, EN-​S)
    ography (PET)/​computed tomography (CT) in the management Stage II
    of childhood NHL has not been assessed as yet. FDG-​PET may be • Single EN tumour with regional node involvement
    more accurate than conventional imaging to assess disease exten- • Two N areas on same side of diaphragm
    sion since high FDG uptake has been shown in most subtypes, but • Primary GI tract tumour (usually in ileocecal area), involvement of
    the impact on therapeutic stratification has not been evaluated. The associated mesenteric nodes, completely resectable (if malignant
    role of FDG-​PET for the assessment of remission has only been in- ascites or extension of tumour to adjacent organs, it should be re-
    garded as stage III)
    vestigated in small series and most reports indicate high rates of false
    positive results requiring histological examination of all residual Stage III
    masses before any changes in treatment. • Two EN tumours (including EN-​B or EN-​S) above and/​or below
    Several groups are working on the prognostic impact of MDD diaphragm
    in the blood, BM, and/​or CNS, identified by molecular techniques • Two N areas above and below diaphragm
    or flow cytometry. Recently, a group of paediatricians worked on a • Any intrathoracic tumour (mediastinal, hilar, pulmonary, pleural, or
    thymic)
    proposal for a Revised International Paediatric NHL Staging System
    • Intra-​abdominal and retroperitoneal disease, including liver, spleen,
    Classification (IPNHLSS), to add these results to the final staging of kidney, and/​or ovary localizations, regardless of degree of resec-
    childhood NHL (Box 18.1). tion (except primary GI tract tumour (usually in ileocecal region)
    with involvement of associated mesenteric nodes that is completely
    resectable)
    Clinical presentation and treatment • Any paraspinal or epidural tumour, regardless of whether other sites
    are involved
    • Single B lesion with concomitant involvement of EN and/​or non-​
    Mature B-cell non-Hodgkin lymphomas regional N sites
    BL generally arises in the abdomen and head and neck region,
    Stage IV
    and presents as advanced stage disease, with BM and CNS in-
    • Any of the above findings with initial involvement of CNS (stage IV
    volvement in about 20–​25% of cases. However, the predominant CNS), BM (stage IV BM), or both (stage IV combined) based on con-
    sites of involvement and patterns of spread are different in pa- ventional methods
    tients with sporadic compared to endemic BL. Jaw involvement Note: For each stage, type of examination and degree of BM and CNS involvement
    is the most common site of disease in endemic BL. In endemic should be specified.
    cases of BL, lesions in multiple quadrants of the jaw or orbital B, bone; BM, bone marrow; EN, extranodal; N, nodal; S, skin.
    Reproduced with permission from Rosolen A, et al. ‘Revised International Pediatric
    tumours (with or without maxillary disease) are observed in the Non-​Hodgkin Lymphoma Staging System’. Journal of Clinical Oncology, Volume 33,
    majority of children, especially those younger than five years. Issue 18, pp. 2112–​2118. Copyright © 2015, American Society of Clinical Oncology.
    Doi: 10.1200/​JCO.2014.59.7203
    Abdominal disease is also very common. In contrast to sporadic
    BL, BM infiltration at diagnosis and at relapse is uncommon, but
    CNS disease occurs in almost one third of patients. CNS disease
    may present as headache and increased intracranial pressure masses but with fever, pallor, and/​or bleeding secondary to pan-
    from meningeal infiltration, cranial nerve palsies, or an isolated cytopenia from extensive marrow replacement by tumour cells.
    epidural tumour and paraplegia. The most common site of dis- These children may also have peripheral lymphadenopathy and
    ease in sporadic cases of BL is the abdomen. Abdominal dis- hepatosplenomegaly; however, mediastinal primaries are more
    ease may present with pain or distention, nausea and vomiting, rarely seen than in other paediatric NHL. BL can involve almost any
    gastrointestinal bleeding, or, rarely, intestinal perforation. organ including testes, breasts, thyroid gland, skin, epidural space,
    Approximately 25–​30% of children present with a right lower bone, or pancreas.
    quadrant mass or acute abdominal pain caused by an ileocecal Children with DLBCL have a more heterogeneous range of clinical
    intussusception, which is often confused with acute appendicitis. presentations compared to patients with BL. In contrast to BL, the
    The majority of patients with abdominal BL have massive dis- BM and CNS are unusual sites of disease. DLBCL usually presents
    ease involving the mesentery, retroperitoneum, kidneys, ovaries, with nodal disease, though bone involvement (single or multiple
    and peritoneal surfaces (often associated with malignant ascites). sites) is relatively common. In immunocompromised individuals,
    Therefore, surgical debulking is neither feasible nor appropriate extranodal disease is common, and primary DLBCL of the brain is
    in these patients. not uncommon. Primary mediastinal B-​cell lymphoma is a distinct
    The head and neck region is the second most common site of dis- entity. It is typically diagnosed in young adult females in the third
    ease in sporadic BL, but in contrast to patients with endemic BL, to fourth decades of life. The tumour is locally invasive (e.g. peri-
    jaw involvement is observed in fewer than 10% of cases of sporadic cardial and lung extension) andoften associated with superior vena
    disease. BM involvement occurs in approximately 20% of cases of cava (SVC) syndrome. Additionally, there is often extrathoracic in-
    sporadic BL. Almost 20% of patients present without lymphomatous volvement of the kidney.
    Clinical presentation and treatment 147

    Outcomes of paediatric BL have improved dramatically due to the lower, with EFS around 70%, but is similar to that obtained with
    implementation of the results of several prospective national and adult protocols. Its biology is similar in both age groups (but grey-​
    multinational studies (Table 18.2). Clinical prognostic factors are zone lymphomas are not seen in children) indicating that rituximab
    already known in B-​cell NHL: stage, lactate dehydrogenase (LDH) should improve outcomes, as seen in adults.
    level, CNS involvement (especially the presence of blasts in the cere- The anti-​ CD20 chimeric monoclonal antibody rituximab
    brospinal fluid (CSF) and treatment-​related factors such as early re- (Mabthera®, Rituxan®) improves survival of adult B-​cell malignan-
    sponse), initial dose intensity, and complete response (CR). Biological cies. In childhood B-​NHL, pilot studies have yielded responses for
    characteristics could also have a prognostic value: the presence of ab- single-​drug rituximab use in newly diagnosed BL and immediate
    normalities on 13q, the presence of MYC rearrangement in DLBCL, good tolerance when combined with induction chemotherapy.
    and a high level of MDD and MRD in BL. However, these findings However, considering the very high cure rates obtained in children
    should still be confirmed in larger and prospective studies. and adolescents with no/​minimal long-​term toxicity, the benefit of
    Treatment of BL is based on two to six months of intensive pulse adding rituximab to conventional chemotherapy in BL should be
    polychemotherapy. Relapses occur early, within the first year. The evaluated. The international ‘Inter-​B NHL ritux 2010’ trial, with the
    French (Lymphomes Malins B) (LMB) studies and the international participation of several countries in Europe and of the Children’s
    FABLMB96 trial demonstrated that: Oncology Group (COG) of North America, combined a phase III
    study testing the impact of adding rituximab to the LMB regimen
    1. high-​dose (HD) methotrexate (MTX) was a very effective drug for advanced-​stage B-​cell lymphoma (excepting PMBL, for which
    for CNS prophylaxis, this trial includes a phase II study of the DA-​EPOCH-​R regimen
    2. patients with leukaemic presentation (mature B-cell acute leu- (http://​clinicaltrials.gov/​ct2/​show/​record/​NCT01760226)). The
    kaemia, B-​ AL) and CNS-​ positive disease benefitted from a randomized phase III trial compared the standard LMB chemo-
    higher dose of HDMTX and the introduction of HD cytarabine therapy to the same chemotherapy with rituximab in patients less
    (Ara-​C) (CYVE courses), than 18 years of age with high-​risk B-​NHL (stage III with LDH level
    3. treatment intensity should not only be adapted to the stage and more than 2N, stage IV) and mature B leukaemia. Rituximab was
    resection, but also to response to chemotherapy (at D7 and after given (375 mg/​m²) on days minus two and one during the first two
    three courses), chemotherapy courses, and on day one of the two following courses
    4. early dose intensity is essential: in the FABLMB96 trial, EFS in (in total, six infusions). The first interim analysis was based on 27
    the intermediate-​risk group was 10% lower when the second in- events (37.5% of information) occurring in 310 patients (155/​arm)
    duction course started later than 21 days after the first one, (85% BL). With a median follow-​up of 11.5 months, patients on
    5. the total dose of cyclophosphamide and doxorubicin could be the rituximab arm achieved better EFS than those on the control
    reduced to 3.3 g/​m² and 120 mg/​m² respectively for more than arm: one-​year EFS (95%CI) was 94.2% (88.5–​97.2%) versus 81.5%
    70% of the patients, allowing clinicians to anticipate fertility (73.0–​87.8%). The hazard ratio (HR) was 0.33 (90%CI: 0.16–​0.69),
    preservation and the absence of long-​term cardiac toxicity. one-​sided p-​value = 0.006. This p-​value was higher than the nom-
    inal alpha (0.0014) of this first interim analysis. However, the con-
    The German BFM studies showed that treatment intensity can also ditional power analysis demonstrated high likelihood of getting a
    be adapted to the LDH level and that HDMTX is a very effective positive study in later interim/​final analyses (from 100% to 87% for
    systemic drug, and more prolonged drug exposure is required for HR = 0.33 to 0.75).
    more advanced disease. They also confirmed that HDAra-​C is im- Thus, rituximab, in addition to standard LMB therapy, improves
    portant in advanced diseases. Although stratification and the weight EFS of children/​adolescents with high-​risk B-​NHL and mature B
    of therapy are not exactly comparable in both strategies, the results leukaemia. However, the use of rituximab in addition to current
    are very similar. CNS prophylaxis should comprise HDMTX +/​ chemotherapy backbone in standard-​risk B-​NHL is still question-
    –​HDAra-​C and intrathecal (IT) injections of MTX +/​–A ​ ra-​C, but able. Indeed, survival in standard-​risk B-​NHL (stages I/​II and III
    no cranial irradiation. With all these current strategies in Western LDH < 2N) is very high, about 97–​98%, with chemotherapy only.
    countries, EFS of BL attains 80–​90%, but immediate treatment-​ This chemotherapy is associated with acute toxicity but exceptional
    related toxicities require adequate supportive care and protocol toxic deaths and no expected long-​term sequelae. In this risk group,
    management experience. the use of rituximab is questionable, and more data are needed on
    BL frequently occurs in countries with fewer resources, especially its long-​term safety when used in addition to chemotherapy in chil-
    in sub-​Saharan countries (where Burkitt first described the disease). dren before introducing it (and reducing chemotherapy intensity) in
    Treatment should be adapted to find a compromise between suffi- standard-​risk B-​NHL.
    cient effectiveness and tolerable toxicity. Several groups have pub- Relapses in mature B-​cell NHL are rare, but their outcome is
    lished interesting results showing cure in about half of the children dismal, with a cure rate below 30%. The first step is to obtain a second
    with adapted protocols. These cure rates will improve with experi- CR. CYVE or ifosfamide, carboplatin, and etoposide (ICE) regimens
    ence, better resources, and less family abandonment of treatment. are effective salvage chemotherapy. Consolidation with HD chemo-
    DLBCL accounts for 10–​20% of B-​cell lymphoma in children and therapy and haematopoietic support is needed after a second CR. An
    occurs more frequently in adolescents. Outcomes are similar in chil- allograft does not afford more benefit than an autograft but could be
    dren and adolescents to that of BL, with the same protocols, except considered to maintain dose intensity and/​or in case of leukaemic
    that relapses can occur up to three years later. The subgroup of pri- involvement. Adverse prognostic factors at relapse are early relapse
    mary mediastinal large B-​cell lymphoma (PMBL) has to be separated (within six months), but also multiple sites of recurrence (versus one
    from other DLBCLs. In paediatric protocols, treatment outcome is site), BL histology (versus DLBCL), and initial high-​risk features (e.g.
    Table 18.2 Characteristics and results of main studies in childhood (less than 18 years of age) B-​cell lymphoma

    Protocol Study Ran Criteria for stratification and risk No. 3-​yr or 5-​yr EFS No. courses CNS prophylaxis TD TD TD
    groups domi groups patients MTX g/​m² (infusion Cyclo Doxo VP16
    za duration) g/​m² mg/​m² g/​m²
    tion (Ifo g/​m²)
    LMB89 SFOP No Stage, resection, 561 91%
    Patte C, et al. (2001) CNS, response at D7
    A Stage 1 & 2 resected 52 98% 2 No 3 120 0
    (9%)
    B Not A, not C 386 (69%) 92% 5 5 HDMTX=3 g/​m²(3h) 5.3 180 0
    10 DIT
    C CNS+ and/​or BM >70% 123 (22%) 84% 8 5 HDMTX=8 g/​m²(4h) 6.8 240 2.5
    CNS 90% 2 HDAra-​C (CYVE)
    CNS+79% 10 TIT
    FAB SFOP Yes Idem LMB89 1111 88%
    LMB96 UKCCSG
    Gerrard M, et al. CCG
    (2008) Patte C, et al.
    (2008) Cairo MS, et al.
    (2007)
    A Idem LMB89 132 (12%) 99% 2 No 3 120 0
    Yes B Not A, not C 744 (67%) 89% 4 Idem LMB89 3.3* 120* 0
    Yes C CNS+, BM >25% 235 (21%) 79% 8 Idem LMB89 6.8* 240* 2.5
    BFM90 BFM No Resection, site, 413 88%
    Reiter A, et al. (1999) LDH (500), BM, CNS
    R1 Complete resection 71 (17%) 100% 2 2 MTX 0.5 g/​m² (24h) 2 50 0.2
    3 TIT (4)
    R2 Extra abdominal only 167 (40%) 96% 4 4 HDMTX 5 g/​m²(24h) 3 100 0.4
    or abdominal & LDH 5 TIT (8)
    <500
    R3 Abdominal & LDH >500 175 (43%) 79% 6 6 HDMTX 5 g/​m²(24h) 4 150 0.6
    and/​or CNS+, and/​or CNS+ 65% 7 TIT (12)
    BM+, and/​or multifocal
    bone
    BFM95 BFM Yes Resection, stage, CNS 505 89%
    Woessmann W, et al. LDH (<500,500–​1000, >1000)
    (2005) Random: HDMTX 4h vs 24h
    Yes R1 Stage 1 & 2 resected 48 (10%) 94% 2 2 HDMTX 1 g/​m² (4h)* 2 50 0.2
    3 TIT (4)
    Yes R2 Stage 1 & 2 non-​resected 233 (46%) 94% 4 4 HDMTX 1 g/​m² (4h)* 2.4 100 0.4
    Stage 3 & LDH <500 5 TIT (8)
    Yes R3 Stage 3 & LDH <1000 82 (16%) 85% 5 4 HDMTX 5 g/​m² (24h) 2.4 100 0.9
    Stage 4/​B-​AL & LDH 1 HDAra-​C (CC) (8)
    <1000 6 TIT
    Yes R4 LDH > 1000 142 (28%) 81% 6 4 HDMTX 5 g/​m² (24h) 2.4 100 1.4
    CNS+ CNS+ 69% 2 HDAra-​C (8)
    7 TIT
    B-​NHL03 JPLSG No Stage, resection, BM, CNS 321 87%
    Tsurusawa M, et al.
    (2014)
    G1 Stage 1 & 2 resected 17 94% 2 2 DIT 1.5 120^ 0
    G2 Stage 1 & 2 non-​resected 103 99% 4 2 TIT, 2 DIT 1.5 120^ 0
    2 HDMTX 3 g/​m² (24h)
    2 MTX 0.5 g/​m² (6h)
    G3 Stage 3 & 4 CNS-​ 111 84% 6 4 TIT, 2 DIT 3 240^ 0.6
    4 HDMTX 3 g/​m² (24h)
    2 MTX 0.5 g/​m² (6h)
    G4 Stage 4 CNS+ & B-​AL 90 78% 6 6 TIT, 2 DIT 5.5 240^ 0.9
    4 HDMTX 5 (24h)
    2 HDAra-​C
    Inter-​B-​NHL ritux Part Yes Stage, LDH /​x2N, CNS, CSF 600 pts First interim
    2010 of EICNHL Randomization: with or without planned analysis
    & COG rituximab (R) (n=310 pts)57: 1-​
    year EFS
    (95%CI)
    Pts on the R is
    94.2% (88.5–​
    97.2) vs. 81.5%
    (73.0–​87.8) on
    the control arm.
    HR=0.33.
    Yes B high Stage 3 & 4 and LDH 4 5 HDMTX=3 g/​m²(3h) 3.3 120 0
    >Nx2 10 DIT
    Yes C1 Stage 4 CNS-​ & B-​AL 6 3–​4 HDMTX=8 g/​m²(4h) 5.8 180 1.6
    and CSF-​ 2 HDAra-​C (CYVE)
    10–​12 TIT
    Yes C3 CSF+ 6 4 HDMTX=8 g/​m²(24h) 5.8 180 1.6
    2 HDAra-​C (CYVE)
    12 TIT
    No PMBL PMBL CNS-​ 40 pts planned Ended 2016. 6 >6 >240 >1.2
    DA-​EPOCH-​R regimen Analysis in
    progress

    SFOP: Société Francaise d’Oncologie Pédiatrique; UKCCSG: United Kingdom Children’s Cancer Study Group; CCG: Cancer Children Group (USA); JPLSG: Japanese Pediatric Leukemia/​Lymphoma Study Group; COG: Children Oncology
    Group (North America, Australia, New Zealand); LMB: Lymphoma Malin B; FAB Franco–​American–​British; BFM: Berlin–​Frankfurt–​Munster; EICNHL: European Intergroup for Childhood NHL; Inter-​B-​NHLritux2010: clinical trial.gov
    identifier: NCT01760226; B-​AL: mature B-​cell acute leukaemia; PMBL: primary mediastinal B-​cell lymphoma; TD: total (cumulative) dose; HDMTX: high dose methotrexate (with folinic rescue); DIT: double intrathecal injection (two drugs
    injected, either MTX or cytarabine with corticosteroid); TIT: triple intrathecal injection (corticosteroids, methotrexate, cytarabine); HDAra-​C : high-​dose cytarabine; CYVE course: combination of cytarabine in continuous infusion and
    HDAra-​C with VP16; CC course: combination of HDAra-​C with VP16, vindesine; CNS+: central nervous system involvement; CNS-​: without central nervous system involvement; BM >70%: bone-​marrow involvement with more than 70%
    blasts; LDH > 2N: LDH level more than twice the institutional normal value; DA-​EPOCH-​R: dose adjusted of VP16, prednisone, vincristine, cyclophosphamide, doxorubicin with rituximab.
    * As results of the randomizations, these are the doses and the MTX infusion durations which are the standards in the current protocols.
    ^ pirarubicin instead of doxorubicin.
    150 CHAPTER 18 Non-Hodgkin Lymphomas in Children and Adolescents

    elevated LDH and advanced-​stage disease versus a low LDH and pre-​phase of corticosteroids. The international Euro-​LB02 trial was
    stage I and II disease). designed by the EICHNL group to compare dexamethasone (known
    Innovative therapies are difficult to explore in childhood B-​NHL to better penetrate the CNS) to prednisone in the induction phase,
    due to the small numbers of cases of refractory/​relapsed disease and to evaluate the possibility of decreasing the duration of treat-
    (in France, for example, there are about five relapses per year of ment from 24 to 18 months (second randomization). The reference
    BL and DLBCL). The most promising therapies include new anti- arm was the NHL-​BFM90 regimen without cranial irradiation, and
    body therapies, along with immunomodulatory agents that improve only T-​LLs were randomized. The study had to be closed prema-
    antibody-​coated tumour-​cell-​killing, cytotoxic T cells genetically turely because of excess toxic deaths. Final analysis showed an EFS
    engineered to express a chimeric antigen receptor (CAR) recog- rate of 821%+2% for the 319 eligible patients (75% T-​LL), with a me-
    nizing CD19 expressed by B cells, and BTK inhibitors. Additionally, dian follow-​up of 6.8 years. Although no CNS relapses occurred in
    for some lymphomas, it has been demonstrated that PDL1 expres- the dexamethasone arm, there was no trend towards better survival
    sion is driven by constitutive oncogenic signalling pathways in the in this arm.
    tumour cell (‘intrinsic’ immune resistance). In PMBL, as in Hodgkin With intensive frontline treatment such as the BFM regimen, re-
    lymphoma, chromosome 9p24.1/​PDL1/​PDL2 alterations are fre- fractory disease/​relapses are rare and have a dismal outcome, with a
    quent and anti-​PD1 pembrolizumab has been recently reported to cure rate generally below 30%. In LL, there is a clear indication for
    have promising activity in heavily pre-​treated PMBL (overall re- an allograft if a second CR is achieved. However, the difficulty of
    sponse rate 41% and, with a median follow-​up of 11.3 months, the obtaining a second CR emphasizes the need for new treatments. So
    median duration of response was not reached). far, the more recent drugs for T-​LL such as nelarabine or clofarabine
    have proved disappointing. Blinatumumab, a bispecific CD19/​CD3
    Lymphoblastic lymphoma antibody, may be promising for pB-​LL.
    T-​LLs generally present as mediastinal masses with advanced-​stage
    disease (>III in more than 90% of cases), may involve the BM at diag- Anaplastic large-​cell lymphoma
    nosis (around 30%), but less often the CNS (around 5%). Symptoms Two forms of ALCL have been described: systemic ALCL and
    often include cough, wheezing, shortness of breath, orthopnoea; cutaneous ALCL. Primary cutaneous ALCLs usually lack ALK
    though swelling of the neck, face, and upper extremities from SVC translocations and belong to the spectrum of primary cutaneous
    obstruction can occur. Haemodynamic compromise due to peri- CD30-​positive lymphoproliferations, a group that also comprises
    cardial effusions may also occur. Subdiaphragmatic disease is often lymphomatoid papulosis. They remain confined to the skin and
    present and may include hepatosplenomegaly, kidney infiltration, usually have an excellent prognosis after surgical resection without
    and retroperitoneal nodal disease. In contrast to the clinical fea- systemic therapy.
    tures associated with precursor T-​cell LL, children and adolescents The clinical presenting features of systemic ALCL in children are
    with pB-​LL tend to have limited disease in sites including skin, soft quite variable but often include constitutional symptoms of fever
    tissue, bone, testes, and peripheral lymph nodes. The cutaneous le- and weight loss. About two thirds of cases present with dissemin-
    sions are typically in the scalp and appear as enlarging, discoloured ated disease. The most frequent sites of involvement in systemic
    masses. Morphologically detectable BM involvement is uncommon ALCL are peripheral nodes, mediastinal adenopathy, and extra-​
    at presentation. nodal sites including skin, soft tissue, and bone. Skin involvement
    The best treatment approaches for childhood advanced LL are is much more frequent than in other childhood lymphomas and is
    ALL-​based therapeutic regimens (e.g. LSA2L2 and the BFM re- a most common site of extranodal disease in ALCL. Spontaneous
    gimens) (Table 18.3). These protocols were modified by adding regression or waxing and waning of skin disease has been observed
    HDMTX to the original LSA2L2 protocol to improve CNS prophy- in systemic ALCL, albeit less frequently than in primary cutaneous
    laxis, and by gradually excluding local and CNS radiotherapy. They ALCL. Bone disease is also common and it is multifocal in as many
    are based on semi-​ continuous intensive polychemotherapy fol- as 10% of patients. The CNS and BM are occasionally involved,
    lowed by maintenance therapy for a total duration of two years. and in several cases a leukaemic presentation has been described.
    Numerous drugs are given including corticosteroids, vincristine, Diagnosis of marrow involvement may be difficult and require
    anthracyclines, cyclophosphamide, cytarabine, asparaginase, and immunostaining with CD30 and/​or ALK or molecular analysis to
    methotrexate. The major role of asparaginase was shown in several identify rare tumour cells.
    studies in which outcome was correlated with the dose intensity of In recent trials with very diverse first-​line chemotherapy regimens
    the drug. CNS prophylaxis is generally based on HDMTX and intra- (in terms of the duration of treatment, and the number and cumu-
    thecal injections. lative doses of drugs), similar EFS rates of about 65–​75% have been
    However, outcomes have not changed significantly since the achieved in children and adolescents (Table 18.4). In the largest
    1980s, with EFS around 75–​85% in most clinical trials in children multinational trial, ALCL99, comprising more than 400 children,
    and adolescents (Table 18.2). The only exception was the BFM90 a 73% five-​year EFS rate was achieved using NHL-​BFM-​B chemo-
    study, with EFS attaining 90%, but this result was not reproduced in therapy as the reference arm. This trial showed that the methotrexate
    the following BFM95 study. Despite minimal differences between schedule of the BFM protocol, including IT therapy, can be safely
    the two protocols, the divergent results obtained may be explained replaced by a less toxic schedule of methotrexate 3 g/​m² in a three-​
    by a different dose and type of asparaginase. To date, except in the hour infusion without IT therapy, and that adding vinblastine to
    case of localized versus advanced stages, no clinical prognostic fac- each course and as maintenance for a total treatment duration of
    tors have been identified. A trend towards a better prognosis was 12 months significantly delayed the occurrence of relapses, but did
    shown in patients with a very good response after the seven-​day not reduce the absolute risk. Similarly, the COG group ANHL0131
    Table 18.3 Characteristics and results of main series of paediatric lymphoblastic lymphoma*

    Protocol Study group Study patients Treatment strategy Treatment RTH CNS Number 3-​5y EFS 3-​5y OS
    (period of the duration Prophy patients
    study) in months laxis with LL
    CCG-​551 CCG NHL LSA2L2 18 local IT 124 64% 67%
    (COMP vs LSA2L2) (1977–​82) st 3&4
    Anderson, J. R. et al.1993
    LSA2L2 NY LL LSA2L2 24-​36 local IT 95 75% 79%
    Mora, J. et al. 2003 (1971–​90) all stages (80%T)
    LMT81 IGR LL LSA2L2 +HDMTX 24 cranial HDMTX 84 75% 76%
    Patte, C. et al.1992 (1981–​89) all stages if CNS+ IT
    CCG 502 CCG LL LSA2L2 24 local IT 143 74% 77%
    (LSA2L2 vs ADCOMP) (1983–​90) craniospinal if
    Tubergen, D. G. et al.1995 CNS+
    POG 9404 POG T-​ALL & T-​LL Dana Farber 24 cranial IT 137 85% NA
    (HDMTX vs no) (1996–​2000) backbone +/​-​ HDMTX
    Asselin, B. L. et al. 2011 cranial RTX
    COG 5941 COG LL “condensed” 12 cranial IT 85 78% 85%
    Abromowitch, M. et al. 2008 (1994–​97) advanced if CNS+ HDMTX (91%T)
    stages
    BFM86 BFM LL NHL-​BFM-​nonB 24 cranial IT 63 84% 88%
    Reiter, A. et al. 1995 (1986–​1990) all stages HDMTX (91%T)
    cranial RTX (st
    III/​IV)
    BFM90 BFM LL NHL-​BFM-​nonB 24 cranial IT 136 90% 92%
    Reiter, A. et al.2000 (1990–​95) all stages HDMTX (81%T)
    cranial RTX (st
    III/​IV)
    BFM95 BFM LL NHL-​BFM-​nonB 24 cranial IT 198 80% 86%
    Burkhardt, B. et al. 2006 (1995–​2001) all stages if CNS+ HDMTX (73%T)
    CLG 58881 EORTC T-​LL NHL-​BFM-​nonB 24 no IT 119 77.5% 86%
    Uyttebroeck, A. et al. 2008 (1989–​98) HDMTX
    LMT96 SFOP T-​LL Modified BFM 18-​24 cranial IT 79 85% 89%
    Bergeron, C. et al. 2015 (1997–​2003) if CNS+ HDMTX
    AIEOP92 AEIOP LL LSA2L2 24 cranial IT 55 69% 74%
    Pillon, M. et al. 2009 (1992–​97) all stages if CNS+ HDMTX (85%T)
    COG A5971 COG LL Modified BFM 24 cranial IT +/​-​ HDMTX 266 82% 85%
    1/​ “CCG/​ALL” vs NHL/​BFM95 (2000–​05) advanced if CNS+ (86%T)
    2/​Intensification: yes vs no stages
    Termuhlen, A. M. et al. 2013
    EuroLb02 EICNHL LL NHL-​BFM90 24 vs 18 cranial IT 319 82±2% 87±2%
    dexamethasone vs prednisone (2003–​08) all stages if CNS+ HDMTX (73%T)
    Landmann et al. 2017

    * Series of localized diseases or B-​LL are not presented


    See Table 18.2 for CCG, COG, SFOP, BFM, EICNHL, HDMTX, IT, CNS+
    RTH: radiation therapy; ref: reference; NA: not available; LMT: lymphomes malins T; POG: Pediatric Oncology Group (USA); EORTC: European Organization for Research and Treatment of Cancer; AEIOP: Italian Association of Pediatric
    Hematology and Oncology; NY: New York Memorial Sloan-​Kettering Cancer Center; IGR: Institut Gustave Roussy, France
    152 CHAPTER 18 Non-Hodgkin Lymphomas in Children and Adolescents

    Table 18.4 Results of the main series of paediatric anaplastic large-​cell lymphoma

    Protocol Study group Treatment strategy Treatment No. of patients 3–​5-​yr EFS 3–​5-​yr OS
    (period of the duration
    study)
    HM89-​91 SFOP B-​cell regimen 7–​8 m 82 66% 83%
    Brugières L, et al. (1998) (1989–​97) (COPADM + maintenance)
    NHL-​BFM90 BFM B-​cell regimen (BFM-​B) 2–​5 m 89 76% -​
    Seidemann K, et al. (2001) (1990–​95)
    NHL 9000 & 9602 UKCCSG B-​cell regimen (LMB) 4–​5 m 72 59% 65%
    Williams DM, et al. (2002) (1990–​98)
    LNH92 AIEOP T-​cell regimen 24 m 34 65% 85%
    Rosolen A, et al. (2005) (1993–​97)
    POG9315 POG APO + randomization of 12 m 86 72% 88%
    Laver JH, et al. (2005) (1994–​2000) HDMTX & HDAra-​C
    CCSG5941 CCSG Compressed T-​cell regimen 11 m 86 68% 80%
    Lowe EJ, et al. (2009) (1996–​2001)
    ALCL99 EICNHL B-​cell regimen (BFM-​B) + 4–​12 m 352 73% 92%
    Brugières L, et al. (2009) (1999–​2006) randomization of vinblastine
    ANHL0131 COG APO + randomization of 12 m 125 74% 84%
    Alexander S, et al. (2014) (2004–​08) vinblastine

    See abbreviations of Box 18.1 and Table 18.2 for the study groups
    APO: doxorubicin, prednisone, vincristine.

    trial showed that replacing vincristine by weekly vinblastine, as part five-​year progression-​free survival (PFS) of 28%); low risk (MDD-​
    of multi-​agent maintenance therapy from the adriamycin, pred- negative and high antibody titre (31% of patients with five-​year EFS
    nisone, and vincristine (APO) regimen, did not improve EFS or of 93%); and intermediate risk: all the remaining patients (48% with
    overall survival (OS). five-​year PFS of 68%).
    One of the unique features of ALCL compared to other paediatric Several new drugs have recently been implemented in the treat-
    NHLs is its sensitivity to chemotherapy after recurrence, leading to a ment of ALCL. Brentuximab vedotin (BV) is an anti-​CD30 mono-
    survival rate of more than 90% in the recent ALCL99 study. Several clonal antibody conjugated to the antimicrotubule cytotoxic
    strategies, including re-​induction chemotherapy followed by autolo- monomethyl auristatin-​E (MMAE). A study on 58 patients with
    gous or allogeneic haematopoietic stem-​cell transplantation (SCT) relapsed/​refractory ALCL (16 ALK-​positive) treated with BV re-
    or weekly vinblastine, can be successful. The efficacy of vinblastine, ported an objective response rate of 86%, with a median duration
    initially shown in a small series of patients with multiple relapses, of 12.6 months. Following these results, BV was approved by the
    was confirmed in the European ALCL relapse study. An interim ana- Federal Drug Agency (FDA) and European Medical Agency (EMA)
    lysis in August 2011 confirmed the efficacy of vinblastine, showing for the treatment of systemic ALCL after failure of at least one
    an 87% two-​year EFS rate in a small series of patients with a late re- chemotherapy regimen in adults.
    lapse (median follow-​up 34 months). ALK inhibitors such as crizotinib, an ALK/​MET inhibitor, are
    Interestingly, vinblastine is one of the most potent drugs capable also promising drugs. Crizotinib is now approved by the FDA and
    of enhancing the anti-​cancer immune response by stimulating den- EMA for the treatment of ALK-​dependent lung cancers. The results
    dritic cell (DC) function. In vitro, vinblastine was found to induce in ALCL are also promising with seven out of nine CRs in patients
    the production of IL-​1α, IL-​6, and IL-​12, increase surface expression with relapsed/​refractory ALCL included in a paediatric phase I trial
    of CD40, CD80, CD86, and MHC class II, and to enhance T-​cell of crizotinib. The same results were obtained, with a response rate of
    stimulatory capacity of DCs. One of the mechanisms of DC matur- 91%, in 11 patients with ALK+ resistant/​refractory adult lymphoma.
    ation induced by vinblastine is thought to be the release of the high Following these results, the COG is currently investigating the
    mobility group box 1 protein (HMGB1), a major molecule involved feasibility of BV or crizotinib combined with ALCL99 chemo-
    in DC stimulation and immunogenic cell death. therapy in children with newly diagnosed stage II–​IV ALCL (http://​
    Prognostic factors associated with an increased risk of failure have clinicaltrials.gov/​ct2/​show/​record/​NCT01979536/​). The EICNHL,
    been identified in ALCL: the presence of mediastinal or visceral in- however, is planning a randomized trial with risk stratification, based
    volvement and skin lesions, and histological lymphohistiocytic and on MDD and the antibody level, aimed at evaluating the efficacy of
    small-​cell variant subtypes. More recently, new parameters such as adding crizotinib to ALCL99 in frontline therapy for intermediate-​
    the detection of MDD in BM and/​or blood, the persistence of posi- and high-​risk ALCL, and comparing weekly vinblastine to ALCL99
    tive MRD after four weeks of treatment, and low production of in low-​and intermediate-​risk disease.
    anti-​ALK titres were also identified. Using MDD and antibody titre NPM-​ALK has also been shown to promote immune evasion
    results in a series of 128 patients allowed them to be classified into by inducing the expression of the immunosuppressive cell-​surface
    three biological risk groups, each with a different prognosis: high protein, programmed death ligand 1 (PD-​L1), through the acti-
    risk: MDD-​positive and low antibody titre (20% of patients with vation of the STAT3 pathway. Remarkably, ALK+ ALCL cell lines
    Conclusion 153

    express PD-​L1 (both mRNA and protein). Furthermore, all pa- randomized phase III trial of APO versus a modified regimen with
    tient tissue samples reported to date have exhibited strong PD-​L1 vinblastine. A report from the Children’s Oncology Group. Pediatr
    immunostaining. Indeed, two refractory ALK+ ALCL patients Blood Cancer 61(12), 2236–​42. doi:10.1002/​pbc.25187
    demonstrated durable response with anti-​PD1 therapy after failure Anderson JR, et al. (1993) Long-​term follow-​up of patients treated
    to chemotherapy and ALK inhibitors (in one patient, also after allo- with COMP or LSA2L2 therapy for childhood non-​Hodgkin’s
    lymphoma: a report of CCG-​ 551 from the Children’s Cancer
    geneic haematopoietic stem-​cell transplantation). Thus, the devel-
    Group. J Clin Oncol 11, 1024–​32.
    opment of PD-​1 inhibitors in refractory/​relapsed ALCL, either in
    Anoop P, et al. (2012) Outcome of childhood relapsed or refractory
    case of progression after targeted therapies or as consolidative im-
    mature B-​cell non-​Hodgkin lymphoma and acute lymphoblastic
    munotherapy in patients in CR after ALK inhibitors or brentuximab leukemia. Leuk. Lymphoma 53, 1882–​8.
    vedotin, may be promising. Asselin BL, et al. (2011) Effectiveness of high-​dose methotrexate in
    T-​cell lymphoblastic leukemia and advanced-​stage lymphoblastic
    lymphoma: a randomized study by the Children’s Oncology Group
    Conclusion (POG 9404). Blood 118, 874–​83.
    Bergeron C, et al. (2015) Treatment of childhood T-​cell lymphoblastic
    Thanks to better recognition of the diseases, and to national and lymphoma—​long-​term results of the SFOP LMT96 trial. Pediatr
    international collaborations, dramatic improvements of cure rates Blood Cancer 62, 2150–​6.
    have been obtained in childhood/​adolescent NHL, with chemo- Bhojwani D, et al. (2015) The role of FDG-​PET/​CT in the evaluation
    therapy adapted to subtypes and to disease extension. Most of these of residual disease in paediatric non-​Hodgkin lymphoma. Br J
    treatments trigger immediate toxicities, commonly aplasia and Haematol 168, 845–​53.
    Bienemann K, et al. (2011) Promising therapy results for lymphoid
    mucositis, which not only hamper quality of life during treatment
    malignancies in children with chromosomal breakage syndromes
    but can also be life-​threatening. However, most of the current treat-
    (Ataxia teleangiectasia or Nijmegen-​breakage syndrome): a retro-
    ments have few expected long-​term toxicities. Furthermore, the
    spective survey. Br J Haematol 155, 468–​76.
    therapeutic strategy and treatment intensity vary considerably be- Bonn BR, et al. (2013) Incidence and prognostic relevance of genetic
    tween paediatric and adult approaches for similar subtypes. The lack variations in T-​cell lymphoblastic lymphoma in childhood and
    of uniform criteria for staging and the limited data concerning age-​ adolescence. Blood 121, 3153–​60.
    related biological characteristics of each subtype prevent us from Bouda GC, Traore F, Couitchere L, et al. (2019) Advanced Burkitt
    knowing the impact of these differences on patient outcomes. The Lymphoma in Sub-Saharan Africa Pediatric Units: Results of the
    use of excellent chemotherapy combination has gradually reduced Third Prospective Multicenter Study of the Groupe Franco-Africain
    the role of radiotherapy. However, radiotherapy can still be useful as d’Oncologie Pediatrique. J Global Oncol in press.
    a form of urgent treatment in children with large mediastinal masses Brugières L, et al. (1998) CD30(+) anaplastic large-​cell lymphoma in
    compromising respiration, in selected relapse cases, or as palliation. children: analysis of 82 patients enrolled in two consecutive studies
    Defining new prognostic factors will help better develop risk-​ of the French Society of Pediatric Oncology. Blood 92, 3591–​8.
    adapted strategies. Major progress is expected, with new treatments Brugières L, et al. (2000) Relapses of childhood anaplastic large-​cell
    lymphoma: treatment results in a series of 41 children—​a report from
    including antibodies and/​or targeted therapies. It is improbable that
    the French Society of Pediatric Oncology. Ann Oncol ESMO 11, 53–​8.
    they will cure NHL on their own, but their subtle combinations
    Brugières L, et al. (2009) Impact of the methotrexate administration
    with chemotherapy may allay some of the chemotherapy-​related dose on the need for intrathecal treatment in children and adoles-
    toxicities. Except for high-​risk patients who require further thera- cents with anaplastic large-​cell lymphoma: results of a randomized
    peutic improvements, the challenge over the coming years will not trial of the EICNHL Group. J Clin Oncol 27(6), 897–​903.
    be to increase already high cure rates but, rather, to strike a balance Burkhardt B, et al. (2005) The impact of age and gender on biology, clin-
    between immediate toxicity and no long-​term toxicity versus poten- ical features and treatment outcome of non-​Hodgkin lymphoma in
    tially decreased immediate toxicity but unknown long-​term toxicity. childhood and adolescence. Br J Haematol 131, 39–​49.
    This requires further paediatric international collaboration along Burkhardt B, et al. (2006) Impact of cranial radiotherapy on cen-
    with partnerships with adult lymphoma groups. tral nervous system prophylaxis in children and adolescents with
    central nervous system-​negative stage III or IV lymphoblastic
    lymphoma. J Clin Oncol 24, 491–​9.
    FURTHER READING Burkhardt B, et al. (2009) Poor outcome for children and adolescents
    Abromowitch M, et al. (2008) Shortened intensified multi-​ agent with progressive disease or relapse of lymphoblastic lymphoma: a
    chemotherapy and non-​ cross resistant maintenance therapy report from the Berlin-​Frankfurt-​Muenster group. J Clin Oncol 27,
    for advanced lymphoblastic lymphoma in children and adoles- 3363–​9.
    cents: report from the Children’s Oncology Group. Br J Haematol Burkhardt B (2010) Paediatric lymphoblastic T-​cell leukaemia and
    143, 261–​7. lymphoma: one or two diseases? Br J Haematol 149, 653–​68.
    Ait-​Tahar K, et al. (2006) B and CTL responses to the ALK protein in Burkitt D (1962) Determining the climatic limitations of a children’s
    patients with ALK-​positive ALCL. Int J Cancer 118, 688–​95. cancer common in Africa. BMJ 2, 1019–​23.
    Ait-​Tahar K, et al. (2010) Correlation of the autoantibody response Cairo MS, et al. (2007) Results of a randomized international study of
    to the ALK oncoantigen in pediatric anaplastic lymphoma kinase-​ high-​risk central nervous system B non-​Hodgkin lymphoma and
    positive anaplastic large cell lymphoma with tumor dissemination B acute lymphoblastic leukemia in children and adolescents. Blood
    and relapse risk. Blood 115, 3314–​19. 109, 2736–​43.
    Alexander S, et al. (2014) Advanced stage anaplastic large cell Cairo MS, et al. (2012) Advanced stage, increased lactate dehydro-
    lymphoma in children and adolescents: results of ANHL0131, a genase, and primary site, but not adolescent age (≥ 15 years), are
    154 CHAPTER 18 Non-Hodgkin Lymphomas in Children and Adolescents

    associated with an increased risk of treatment failure in children Landmann E, Burkhardt B, Zimmermann M, et al. (2017) Results
    and adolescents with mature B-​cell non-​Hodgkin’s lymphoma: re- and conclusions of the European Intergroup EURO-LB02 trial
    sults of the FAB LMB 96 study. J Clin Oncol 30, 387–​93. in children and adolescents with lymphoblastic lymphoma.
    Callens C, et al. (2012) Clinical impact of NOTCH1 and/​or FBXW7 Haematologica 102, 2086–96. doi:10.3324/haematol.2015.139162
    mutations, FLASH deletion, and TCR status in pediatric T-​cell Lavoine N, et al. (2015) Constitutional mismatch repair deficiency
    lymphoblastic lymphoma. J Clin Oncol 30, 1966–​73. syndrome: clinical description in a French cohort. J Med Genet
    Ciambotti B, et al. (2014) Monoallelic mutations of the perforin gene 52, 770–​8.
    may represent a predisposing factor to childhood anaplastic large Laver JH, et al. (2005) Advanced-​stage large-​cell lymphoma in chil-
    cell lymphoma. J Pediatr Hematol Oncol 36, e359–​65. dren and adolescents: results of a randomized trial incorporating
    Coustan-​Smith E, et al. (2009) Minimal disseminated disease in child- intermediate-​dose methotrexate and high-​dose cytarabine in the
    hood T-​cell lymphoblastic lymphoma: a report from the Children’s maintenance phase of the APO regimen: a Pediatric Oncology
    Oncology Group. J Clin Oncol 27, 3533–​9. Group phase III trial. J Clin Oncol 23, 541–​7.
    Damm-​Welk C, et al. (2007) Prognostic significance of circulating Le Deley MC, et al. (2008) Prognostic factors in childhood anaplastic
    tumor cells in bone marrow or peripheral blood as detected by large cell lymphoma: results of a large European intergroup study.
    qualitative and quantitative PCR in pediatric NPM-​ALK-​positive Blood 111, 1560–​6.
    anaplastic large-​cell lymphoma. Blood 110, 670–​7. Le Deley MC, et al. (2010) Vinblastine in children and adolescents with
    Ducassou S, et al. (2011) Clinical presentation, evolution, and prog- high-​risk anaplastic large-​cell lymphoma: results of the random-
    nosis of precursor B-​cell lymphoblastic lymphoma in trials LMT96, ized ALCL99-​vinblastine trial. J Clin Oncol 28, 3987–​93.
    EORTC 58881, and EORTC 58951. Br J Haematol 152, 441–​51. Louissaint A, et al. (2012) Pediatric-​type nodal follicular lymphoma: an
    Epstein MA, et al. (1964) Virus particles in cultured lymphoblasts from indolent clonal proliferation in children and adults with high pro-
    Burkitt’s lymphoma. Lancet 1, 702–​3. liferation index and no BCL2 rearrangement. Blood 120, 2395–​404.
    Gambacorti Passerini C, et al. (2014) Crizotinib in advanced, Louissaint A, et al. (2016) Pediatric-​type nodal follicular lymphoma: a
    chemoresistant anaplastic lymphoma kinase-​positive lymphoma biologically distinct lymphoma with frequent MAPK pathway mu-
    patients. J Natl Cancer Inst 106, djt378. tations. Blood 128, 1093–​100.
    Gerrard M, et al. (2008) Excellent survival following two courses of Love C, et al. (2012) The genetic landscape of mutations in Burkitt
    COPAD chemotherapy in children and adolescents with resected lymphoma. Nat Genet 44, 1321–​5.
    localized B-​cell non-​Hodgkin’s lymphoma: results of the FAB/​LMB Lowe EJ, et al. (2009) Intensive chemotherapy for systemic anaplastic
    96 international study. Br J Haematol 141, 840–​7. large cell lymphoma in children and adolescents: final results
    Gerrard M, et al. (2013) Outcome and pathologic classification of of Children’s Cancer Group Study 5941. Pediatr Blood Cancer
    children and adolescents with mediastinal large B-​cell lymphoma 52, 335–​9.
    treated with FAB/​ LMB96 mature B-​ NHL therapy. Blood 121, Magrath IT (1991) African Burkitt’s lymphoma. History, biology,
    278–​85. clinical features, and treatment. Am J Pediatr Hematol Oncol 13,
    Greenough A, Dave SS (2014) New clues to the molecular patho- 222–​46.
    genesis of Burkitt lymphoma revealed through next-​generation Marzec M, et al. (2008) Oncogenic kinase NPM/​ALK induces through
    sequencing. Curr Opin Hematol 21, 326–​32. STAT3 expression of immunosuppressive protein CD274 (PD-​L1,
    Goldman S, et al. (2014) Rituximab with chemotherapy in children B7-​H1). Proc Natl Acad Sci USA 105, 20852–​7.
    and adolescents with central nervous system and/​or bone marrow-​ Meinhardt A, et al. (2010) Phase II window study on rituximab in
    positive Burkitt lymphoma/​ leukaemia: a Children’s Oncology newly diagnosed pediatric mature B-​cell non-​Hodgkin’s lymphoma
    Group report. Br J Haematol 167(3), 394–​ 401. doi:10.1111/​ and Burkitt leukemia. J Clin Oncol 28, 3115–​21.
    bjh.13040 Miles RR, et al. (2008) Pediatric diffuse large B-​cell lymphoma dem-
    Goldman S, et al. (2013) Rituximab and FAB/​LMB 96 chemotherapy onstrates a high proliferation index, frequent c-​Myc protein ex-
    in children with stage III/​IV B-​cell non-​Hodgkin lymphoma: a pression, and a high incidence of germinal center subtype: report
    Children’s Oncology Group report. Leukemia 27, 1174–​7. of the French–​American–​British (FAB) international study group.
    Gross TG, et al. (2010) Hematopoietic stem cell transplantation for Pediatr Blood Cancer 51, 369–​74.
    refractory or recurrent non-​Hodgkin lymphoma in children and Minard-​Colin V, et al. (2015) Non-​Hodgkin lymphoma in children
    adolescents. Biol Blood Marrow Transplant 16, 223–​30. and adolescents: progress through effective collaboration, current
    Haddy TB, et al. (1991) CNS involvement in small noncleaved-​cell knowledge, and challenges ahead. J Clin Oncol 33, 2963–​74.
    lymphoma: is CNS disease per se a poor prognostic sign? J Clin Minard-​Colin V, et al. (2016) Results of the randomized intergroup trial
    Oncol 9, 1973–​82. Inter-​B-​NHL Ritux 2010 for children and adolescents with high-​risk
    Hebart H, et al. (2016) Nivolumab for refractory anaplastic large cell B-​cell non-​Hodgkin lymphoma (B-​NHL) and mature acute leukemia
    lymphoma: a case report. Ann Intern Med 165, 607–​8. (B-​AL): evaluation of rituximab (R) efficacy in addition to standard
    Hesseling P, et al. (2013) Practical recommendations for the manage- LMB chemotherapy (CT) regimen. J Clin Oncol 34(15), suppl 10507.
    ment of children with endemic Burkitt lymphoma (BL) in a re- Mora J, et al. (2003) Lymphoblastic lymphoma of childhood and the
    source limited setting. Pediatr Blood Cancer 60, 357–​62. LSA2-​L2 protocol: the 30-​year experience at Memorial–​Sloan–​
    Jourdain A, et al. (2015) Outcome and prognostic factors of relapse in Kettering Cancer Center. Cancer 98, 1283–​91.
    children and adolescents with mature B-​cell lymphoma and leukemia Mossé YP, et al. (2013) Safety and activity of crizotinib for paedi-
    treated in three consecutive prospective LMB protocols. Study by atric patients with refractory solid tumours or anaplastic large-​
    the SFOP/​SFCE (Société Française d’Oncologie Pédiatrique/​Société cell lymphoma: a Children’s Oncology Group phase 1 consortium
    Française des Cancers de l’Enfant). Haematologica 100(6), 810–​17. study. Lancet Oncol 14, 472–​80.
    Lamant L, et al. (2011) Prognostic impact of morphologic and pheno- Murphy SB (1980) Classification, staging and end results of treat-
    typic features of childhood ALK-​ positive anaplastic large-​cell ment of childhood non-​Hodgkin’s lymphomas: dissimilarities from
    lymphoma: results of the ALCL99 study. J Clin Oncol 29, 4669–​76. lymphomas in adults. Semin Oncol 7, 332–​9.
    Conclusion 155

    Mussolin L, et al. (2003) Prospective analysis of minimal bone marrow Reiter, A, et al. (1999) Improved treatment results in childhood B-​cell
    infiltration in pediatric Burkitt’s lymphomas by long-​distance poly- neoplasms with tailored intensification of therapy: a report of the
    merase chain reaction for t(8;14)(q24;q32). Leukemia 17, 585–​9. Berlin–​Frankfurt–​Münster Group Trial NHL-​BFM 90. Blood 94,
    Mussolin L, et al. (2011) Minimal disseminated disease in high-​risk 3294–​306.
    Burkitt’s lymphoma identifies patients with different prognosis. J Reiter A, et al. (2000) Intensive ALL-​type therapy without local radio-
    Clin Oncol 29, 1779–​84. therapy provides a 90% event-​free survival for children with T-​cell
    Mussolin L, et al. (2013) Use of minimal disseminated disease and im- lymphoblastic lymphoma: a BFM group report. Blood 95, 416–​21.
    munity to NPM-​ALK antigen to stratify ALK-​positive ALCL pa- Rigaud C (2018) Efficacy of nivolumab in a patient with systemic
    tients with different prognosis. Leukemia 27, 416–​22. refractory ALK+ anaplastic large-​cell lymphoma. Pediatr Blood
    Neth O, et al. (2000) Precursor B-​cell lymphoblastic lymphoma in Cancer, 65(4). doi: 10.1002/​pbc.26902
    childhood and adolescence: clinical features, treatment, and results Rohde M, et al. (2017) Relevance of ID3-​TCF3-​CCND3 pathway mu-
    in trials NHL-​BFM 86 and 90. Med Pediatr Oncol 35, 20–​7. tations in pediatric aggressive B-​cell lymphoma treated according
    Oschlies I, et al. (2006) Diffuse large B-​cell lymphoma in pediatric pa- to the non-​Hodgkin lymphoma Berlin–​Frankfurt–​Münster proto-
    tients belongs predominantly to the germinal-​center type B-​cell cols. Haematologica 102, 1091–​8.
    lymphomas: a clinicopathologic analysis of cases included in the Rosenwald A, et al. (2003) Molecular diagnosis of primary mediastinal
    German BFM (Berlin–​Frankfurt–​Munster) multicenter trial. Blood B cell lymphoma identifies a clinically favorable subgroup of diffuse
    107, 4047–​52. large B cell lymphoma related to Hodgkin lymphoma. J Exp Med
    Oschlies I, et al. (2011) Clinical, pathological and genetic features of 198, 851–​62.
    primary mediastinal large B-​cell lymphomas and mediastinal gray Rosolen A,et al. (2005) Anaplastic large cell lymphoma treated with a
    zone lymphomas in children. Haematologica 96, 262–​8. leukemia-​like therapy: report of the Italian Association of Pediatric
    Patte C, et al. (1991) High survival rate in advanced-​stage B-​cell Hematology and Oncology (AIEOP) LNH-​92 protocol. Cancer
    lymphomas and leukemias without CNS involvement with a short 104, 2133–​40.
    intensive polychemotherapy: results from the French Pediatric Rosolen A, et al. (2015) A revised international pediatric non-​Hodgkin
    Oncology Society of a randomized trial of 216 children. J Clin Oncol lymphoma (NHL) staging system classification (IPNHLSS): a re-
    9, 123–​32. port from the Fourth International Childhood, Adolescent and
    Patte C, et al. (1992) Results of the LMT81 protocol, a modified LSA2L2 Young Adult NHL Symposium. J Clin Oncol 20, 33(18), 2112–​18.
    protocol with high dose methotrexate, on 84 children with non-​B-​ Salaverria I, et al. (2011) Translocations activating IRF4 identify a sub-
    cell (lymphoblastic) lymphoma. Med Pediatr Oncol 20, 105–​13. type of germinal center-​derived B-​cell lymphoma affecting pre-
    Patte C, et al. (2001) The Société Française d’Oncologie Pédiatrique dominantly children and young adults. Blood 118, 139–​47.
    LMB89 protocol: highly effective multiagent chemotherapy tailored Salaverria I, et al. (2014) A recurrent 11q aberration pattern charac-
    to the tumor burden and initial response in 561 unselected children terizes a subset of MYC-​negative high-​grade B-​cell lymphomas re-
    with B-​cell lymphomas and L3 leukemia. Blood 97, 3370–​9. sembling Burkitt lymphoma. Blood 123, 1187–​98.
    Patte C, et al. (2007) Results of the randomized international FAB/​ Sander S, et al. (2012) Synergy between PI3K signaling and MYC in
    LMB96 trial for intermediate risk B-​cell non-​Hodgkin lymphoma Burkitt lymphomagenesis. Cancer Cell 22, 167–​79.
    in children and adolescents: it is possible to reduce treatment for Sandlund JT, et al. (1994) Clinical features and treatment outcome for
    the early responding patients. Blood 109, 2773–​80. children with CD30+ large-​cell non-​Hodgkin’s lymphoma. J Clin
    Pelicci PG, et al. (1986) Chromosomal breakpoints and structural al- Oncol 12, 895–​8.
    terations of the c-​myc locus differ in endemic and sporadic forms of Sandlund JT, et al. (2014) Pilot study of modified LMB-​based therapy
    Burkitt lymphoma. Proc Natl Acad Sci USA 83, 2984–​8. for children with ataxia-​telangiectasia and advanced stage high-​
    Percy C, et al. (1999) Lymphomas and reticuloendothelial neoplasms. grade mature B-​cell malignancies. Pediatr Blood Cancer 61, 360–​2.
    In: L. Ries, et al. (eds) Cancer Incidence and Survival Among Schmidt J, et al. (2016) Genome-​wide analysis of pediatric-​type fol-
    Children and Adolescents: United States SEER Program 1975–​1995, licular lymphoma reveals low genetic complexity and recurrent al-
    pp. 35–​50. Bethesda, MD: National Cancer Institute. terations of TNFRSF14 gene. Blood 128, 1101–​11.
    Pillon M, et al. (2009) Long-​term results of AIEOP LNH-​92 protocol Seidemann K, et al. (2001) Short-​pulse B-​non-​Hodgkin lymphoma-​
    for the treatment of pediatric lymphoblastic lymphoma: a report type chemotherapy is efficacious treatment for pediatric anaplastic
    of the Italian Association of Pediatric Hematology and Oncology. large cell lymphoma: a report of the Berlin–​Frankfurt–​Münster
    Pediatr Blood Cancer 53, 953–​9. Group Trial NHL-​BFM 90. Blood 97, 3699–​706.
    Poirel HA, et al. (2009) Specific cytogenetic abnormalities are associ- Swerdlow SH, et al. (2016) The 2016 revision of the World Health
    ated with a significantly inferior outcome in children and adoles- Organization classification of lymphoid neoplasms. Blood 127,
    cents with mature B-​cell non-​Hodgkin’s lymphoma: results of the 2375–​90.
    FAB/​LMB 96 international study. Leukemia 23, 323–​31. Tanaka H, et al. (2009) Dual therapeutic efficacy of vinblastine as a
    Pro B, et al. (2012) Brentuximab vedotin (SGN-​35) in patients with unique chemotherapeutic agent capable of inducing dendritic cell
    relapsed or refractory systemic anaplastic large-​cell lymphoma: re- maturation. Cancer Res 69, 6987–​94.
    sults of a phase II study. J Clin Oncol 30, 2190–​6. Termuhlen AM, et al. (2013) Disseminated lymphoblastic lymphoma
    Reiter A, Klapper, W (2008) Recent advances in the understanding in children and adolescents: results of the COG A5971 trial—​a
    and management of diffuse large B-​cell lymphoma in children. Br J report from the Children’s Oncology Group. Br J Haematol 162,
    Haematol 142, 329–​47. 792–​801.
    Reiter A, et al. (1995) Non-​Hodgkin’s lymphomas of childhood and Tsurusawa M, et al. (2014) Improved treatment results of children
    adolescence: results of a treatment stratified for biologic subtypes with B-​cell non-​Hodgkin lymphoma: a report from the Japanese
    and stage—​a report of the Berlin–​Frankfurt–​Münster Group. J Clin Pediatric Leukemia/​ Lymphoma Study Group B-​ NHL03 study.
    Oncol 13, 359–​72. Pediatr Blood Cancer 61, 1215–​21.
    156 CHAPTER 18 Non-Hodgkin Lymphomas in Children and Adolescents

    Tubergen DG, et al. (1995) Comparison of treatment regimens for Woessmann W, et al. (2005) The impact of the methotrexate admin-
    pediatric lymphoblastic non-​Hodgkin’s lymphoma: a Children’s istration schedule and dose in the treatment of children and ado-
    Cancer Group study. J Clin Oncol 13, 1368–​76. lescents with B-​cell neoplasms: a report of the BFM Group Study
    Uyttebroeck A, et al. (2008) Treatment of childhood T-​cell lympho- NHL-​BFM95. Blood 105, 948–​58.
    blastic lymphoma according to the strategy for acute lymphoblastic Woessmann W, et al. (2011) Relapsed or refractory anaplastic large-​
    leukaemia, without radiotherapy: long-​term results of the EORTC cell lymphoma in children and adolescents after Berlin–​Frankfurt–​
    CLG 58881 trial. Eur J Cancer 44, 840–​6. Muenster (BFM)-​type first-​line therapy: a BFM group study. J Clin
    Williams DM, et al. (2002) Anaplastic large cell lymphoma in child- Oncol 29, 3065–​71.
    hood: analysis of 72 patients treated on the United Kingdom Woessmann W, et al. (2012) Risk-​adapted therapy for patients with
    Children’s Cancer Study Group chemotherapy regimens. Br J relapsed or refractory ALCL—​interim results of the prospective
    Haematol 117, 812–​20. EICNHL-​trial ALCL-​relapse. Br J Haematol 159(S1), 41.
    Wimmer K, et al. (2014) Diagnostic criteria for constitutional mismatch Zinzani PL, et al. (2017) Safety and tolerability of pembrolizumab in
    repair deficiency syndrome: suggestions of the European consor- patients with relapsed/​refractory primary mediastinal large B-​cell
    tium ‘Care for CMMRD’ (C4CMMRD). J Med Genet 51, 355–​65. lymphoma. Blood 130, 267–​70.
    19
    Hodgkin Lymphoma
    Modern Management of Children and Young Adults

    Judith Landman-​Parker and Françoise Montravers

    Introduction distinct nodular lymphocyte predominant Hodgkin lymphoma


    (NLPHL) subtype with unique clinico-​pathological features and de-
    In three decades, Hodgkin lymphoma (HL) has become a highly cur- tection of histiocytic ‘pop corn’ cells. cLH is subdivided into four
    able lymphoid malignancy with an expected survival rate at ten years subtypes: nodular sclerosis, 70–​80%; mixed cellularity, 20–​30%;
    of over 90%. The number of patients cured of their disease is estimated lymphocyte-​rich, 2–​5%; and lymphocyte depletion, less than 1%.
    at 100,000 in Europe, and long-​term management of treatment-​related Immunostaining of HRS is typically CD30+, CD15+, PAX5+, ALK-​,
    subsequent morbidity is now a relevant health question. Moreover, new CD45-​, and CD 20+/​-​.
    drugs have recently emerged targeted at Hodgkin/​Reed–​Sternberg (H/​
    RS) cells and their environment, leading to new therapeutic opportun-
    ities in patients with relapsed or refractory disease Biology

    Recent insights into the biology of H/​RS cells are beginning to


    Epidemiology provide a better understanding of HL pathogenesis. H/​RS cells
    are clonal, arise from germinal centre B-​cell lymphocytes in virtu-
    Exceptional before five years of age, HL is the most common cancer ally all cases, and present crippling mutations of immunoglobulin
    in adolescence. In the European Union, the crude incidence of HL is genes. Their failure to express a normal B-​lymphocyte phenotype
    2.3–​3 cases per 100,000 people per year. Sex ratio varies with age, with is related to multiple molecular and epigenetic mechanisms leading
    a higher incidence in boys when HL is diagnosed below 12 years. HL to survival and proliferation of H/​RS cells. Most H/​RS cells con-
    displays a bimodal distribution in adults, with an initial peak in young stitutively express the anti-​apoptotic NF-​kB pathway as a result of
    adults and a later peak among patients over 55 years. Childhood and somatic mutation or amplification of REL, TNFα-​induced protein3,
    adolescent HL represents around 10% of the cases. Patterns of incidence and NFKBIA genes and upregulate the JAK-​Stat pathway by somatic
    are variable worldwide: lower in Asia and higher in the Middle East and mutations in 40% of cases.
    Mediterranean area. In economically less developed regions, the initial Moreover, H/​RS cells suppress anti-​tumour immunity through ex-
    HL peak is earlier among children and teenagers. pression of CD30, CD40, CD80, and PD1. They secrete numerous
    cytokines, such as TARC (CCL17), CCL5, CCL22, and interleukin-​7,
    attracting T-​helper 2 and regulatory T (Treg) cells to the tumour micro-
    Pathology environment. In 2016, a report by Roemer et al., combining fluorescent
    in situ hybridization (FISH), single nucleotide polymorphism (SNP) ar-
    HL is characterized by the presence of multinucleated giant H/​RS rays, and immunohistochemical analysis in 108 biopsy specimens from
    cells or large mononuclear cell variants (lymphocytic and histiocytic patients with newly diagnosed cHL, identified rearrangements of the
    cells) accounting for 1% of the cells surrounded by polymorphic in- PDL1/​PDL2 genes locus in 97% of cases and, therefore, confirmed a
    flammatory tissue reaction with lymphoid cells, macrophages, eo- potential major role of this mechanism in cLH development.
    sinophilia, and fibrosis.

    World Health Organization classification Hodgkin lymphoma and Epstein-​Barr virus


    Since 2001, the World Health Organization (WHO) classification
    has defined the classical variant of Hodgkin lymphoma (cHL): 90% Epstein–​Barr virus (EBV) is associated with cHL development in
    of cases characterized histologically by HRS cells and a rare and 30–​50% of cases in children and adolescents, depending on age,
    158 CHAPTER 19 Hodgkin Lymphoma

    socioeconomic context, and histopathology. In these cases, HR/​S rate (ESR), haemogram, white blood cell (WBC) count, immuno-
    cells express latency type II EBV genes: LMP1, LMP2, EBNA1, and globulin dosage, human immunodeficiency virus (HIV) serology,
    BamH1A. Transfection of viral genes in lymphocytes reproduces the and renal and hepatic tests, as well electrocardiogram (ECG) and
    HR/​S phenotype. High levels of antibodies against EBV are associ- echocardiography, have to be performed before treatment.
    ated with HL, and a population-​based study has confirmed a four-​
    times higher risk to develop EBV-​associated HL up to five years after Imaging in Hodgkin lymphoma
    a mononucleosis infection. Initial work-​up comprises X-​ray chest with measure of the medi-
    astinal diameter, cervico–​ thoracic–​
    abdominal pelvis contrast-​
    enhanced computed tomography (CT) or magnetic resonance
    Predisposition to Hodgkin lymphoma: a link imaging (MRI), and abdominal ultrasonography. Diagnostic CT
    with immune system control? and MRI reliably detect enlarged lymph nodes; intravenous (IV)
    contrast is mandatory to distinguish lymphadenopathy from other
    Familial Hodgkin lymphoma structures in the mediastinum and abdomen. Diagnostic CT allows
    Familial HL has been described since the 1970s. There is an esti- detailed evaluation of pulmonary parenchyma, pleura, and pericar-
    mated seven-​fold risk to develop HL in siblings and first-​degree dium. MRI is superior to contrast CT for evaluation of bone marrow,
    relatives. GWAS (genome-​wide association) analysis and more clas- liver, soft tissue, and central nervous system, and is currently per-
    sical candidates’ gene approaches have reported a possible locus of formed as a complement to other imaging examinations.
    interest in HLA and DNA repair genes.
    [18F] fluorodeoxyglucose-​positron emission tomography/​
    Paediatric immune deficiencies computed tomography in Hodgkin lymphoma
    Paediatric immune deficiencies are associated with HL develop- Metabolic scintigraphy by [18F] fluorodeoxyglucose-​positron emis-
    ment. Although rare, reports concern common variable immuno- sion tomography/​computed tomography (FDG-​PET/​CT) is indi-
    globulin deficiency, X-​LP, Ataxia telangiectasia, Wiskott–​Aldrich cated in cHL and highly recommended before any treatment as part
    syndrome, CD27/​CD70 deficiency, DNA repair disorders, and of initial staging. In 97–​100% of HL, FDG-​PET/​CT is positive (at
    other exceptional entities. In oncology practice, any familial his- least one [18F]FDG-​avid HL site). Low-​dose CT, usually performed
    tory or unusual clinical or histopathological presentation such with PET, is more accurate than PET alone. However, contrast-​
    as very young age, consanguinity, Ig-​deficiency, or interfollicular enhanced CT or MRI with thoracic CT remain mandatory, at least
    subtype has to lead to immunological investigations. Reports of at diagnosis, and can be performed either simultaneously with PET
    HL cases in the context of defined acquired immune deficiency or separately.
    such as HIV, organ transplantation, or immunosuppressive treat- [18F]FDG-​PET detects nodal and extranodal disease sites that are
    ment are important as providing clinical information about risk missed by CT/​MRI staging methods, and improves the character-
    and outcome. However, few of these concern children and ado- ization of lesions (e.g. bone/​bone marrow disease) that are equivocal
    lescents. In those cases, HL is mostly EBV-​associated and in an on other types of imaging. Bone marrow biopsy indicated previously
    advanced stage. in stage III or above HL can now be replaced by [18F]FDG-​PET. In
    contrast, [18F]FDG-​PET/​CT can under-​stage disease, especially in
    Autoimmune disease and Hodgkin lung localizations and because some lesions, even large, may not be
    lymphoma predisposition [18F]FDG-​avid. Therefore, careful comparison of [18F]FDG-​positive
    Such as acute autoimmune cytopenias, glomerulopathy with lesions and morphological imaging is essential for accurate staging,
    nephrotic syndrome, and neurological inflammatory disease have according to the Ann Arbor revised Cotswolds classification. In
    been reported concomitant with, before, or after HL. GWAS ana- children, [18F]FDG-​PET modified the stage in 10–​23% of patients as
    lysis, as well as epidemiological studies, have described common a consequence. Patient stage distribution in the Euronet paediatric
    polymorphisms in autoimmune disease and cHL. From a practical Hodgkin trial is stage I–​II, 55%; stage III, 20%, and stage IV, 25%, if
    point of view, treatment of HL in these patients should follow the FDG-​PET/​CT is integrated in the staging modalities.
    actual guidelines as much as possible, except in case of organ failures
    that require therapeutic modifications.
    Imaging during treatment and
    response-​adapted strategy
    Initial [18F]FDG-​PET/​CT provides a reference for subsequent treat-
    Clinical management and work-​up in classical ment evaluation in cHL. Since the first reports on inferior outcome
    variant of Hodgkin lymphoma of patients with [18F]FDG-​PET positivity during treatment after
    two ABVD (doxorubicin, bleomycin, vinblastine, dacarbazine)
    Initial peripheral node enlargement is described in 85% of patients cycles or escalated BEACOPP (bleomycin, etoposide, doxorubicin,
    with cHL. Indolent progression of the disease leads frequently to de- cyclophosphamide, vincristine, prednisone, procarbazine), con-
    layed diagnosis (median six months between first reported symp- sensus about standard definition of [18F]FDG-​PET positivity in HL
    toms and diagnosis). Systemic symptoms such as unexplained has been reported and the five-​point Deauville scale is used to de-
    fever, drenching night sweats, or weight loss have to be collected, fine criteria for treatment intensification or reduction (Box 19.1); a
    as well as detailed clinical node involvement and splenic or hepatic Deauville score of three or less is considered as metabolic complete
    enlargement. Usual blood tests such as erythrocyte sedimentation remission (Figure 19.1).
    Treatment 159

    Box 19.1 Semi-​quantitative Deauville five-​point scale on previous studies, the Euronet Paediatric Hodgkin Lymphoma
    Group currently defines localized disease as:
    1 no residual positivity
    2 fixation < mediastinum blood pool • stage I/​IIA—​ESR less than 30 and no bulky mass more than 200 ml
    3 fixation > mediastinum < liver • intermediate disease: stage I/​IIA—​ESR more than or equal to 30 or
    4 fixation moderately > liver bulky disease more than or equal to 200 ml; stage IIB—​no E-​lesion
    5 fixation higher > foie • advanced disease: stages IIIA, IIIB, and IV—​stage IIB with E-​lesion
    Adapted with permission from Meignan M., et al. ‘Report on the First International
    Workshop on interim-​PET scan in lymphoma’. Leukemia & Lymphoma, Volume Treatment strategy
    50, Issue 8, pp. 1257–​1260. Copyright © 2009 Informa UK Limited. DOI: 10.1080/​
    10428190903040048 With an actuarial survival of over 95% at ten years in children and
    adolescents with cHL, major efforts are undertaken to provide treat-
    ment with a low rate of adverse long-​term events such as secondary
    malignancies (notably breast cancer), fertility impairment, cardiac
    and pulmonary toxicity, thyroid dysfunction, musculoskeletal ab-
    Treatment normalities, and fatigue. These events are largely related to radiation
    therapy, cumulative doses of alkylating agents and anthracyclines,
    Stratification and steroid treatment.
    Worldwide, treatment groups based on previous trial results and Led by the EuroNet Paediatric Hodgkin Lymphoma Group, a
    prognostic factors are not identical between adult and paediatric large study was conducted in Europe, between 2007 and 2013, re-
    collaborative groups, and a harmonization process is ongoing to spe- cruiting 2,111 patients with a median age of 14.8 years. Objectives of
    cify the definition of bulky disease, E-​lesion (organ and parietal in- the EuroNet PHL-​C1 trial (Box 19.2) were to limit the indication for
    volvement), as well as biological criteria of inflammatory HL. Based radiotherapy (RT) to patients with an inadequate response defined at

    (a)

    (b)

    Figure 19.1 cLH stade IV. (a) Diagnostic; (b) Complete metabolic response after 2 OEPA.
    Reproduced courtesy of Françoise Montravers, Hôpital Tenon APHP, Paris, France.
    160 CHAPTER 19 Hodgkin Lymphoma

    Box 19.2 Chemotherapy in classical variant of Hodgkin Commonly used salvage regimens include ifosfamide, carboplatin,
    lymphoma: Euronet paediatric Hodgkin lymphoma C1 trial and etoposide (ICE), the combination of dexamethasone, high-​dose
    cytarabine, and cisplatin (DHAP), and other gemcitabine-​based com-
    Two OEPA cycles every 28 days bination regimens, including gemcitabine, prednisone, ifosfamide, and
    • Prednisone: 60 mg/​m2; days 1–​15 vinorelbine (IGEV). As many drug regimens are efficient in RR cHL,
    • Vincristine: 1.5 mg/​m2; days 1, 8, and 15 and in absence of randomized studies, the choice depends on age, first-​
    • Adriamycine: 40 mg/​m2; days 1 and 15 line treatment, and response. There is consensus for high-​dose chemo-
    • VP-​16: 125 mg/​m2; days 1–​5 therapy (BEAM) with haematopoietic stem-​cell support in most cases
    COPDAC cycles every 21 days of relapse occurring within one year after first treatment.
    • Prednisone: 40 mg/​m2; days 1–​15 Recently, two new drugs have been registered for RR cHL:
    • Dacarbazine: 250 mg/​m2; days 1, 2, and 3 brentuximab vedotin, an anti-​ CD30 coupled with monomethyl
    • Vincristine: 1.5 mg/​m2; days 1 and 8 auristatin-​E (MMAE), and nivolumab, an anti-​PDL1 monoclonal
    • Endoxan: 500 mg/​m2; days 1 and 8 antibody showing promising results in RR cHL, with complete remis-
    sion (CR) rates in monotherapy and overall response rates of 15–​20%
    and 60–​80%, respectively. Their place in RR strategy, alone or in com-
    bination with chemotherapy, is currently being explored in numerous
    an intermediate [18F]FDG-​PET and CT evaluation after two OEPA trials, along with other immunotargeted compounds and antibodies.
    (vincristine, etoposide, prednisone, adriamycin) cycles, and to replace
    procarbazine by dacarbazine in order to limit the impact on fertility. At
    the latest interim analysis, 48-​month overall survival (OS) and event-​ Clinical management and work-​up in nodular
    free survival (EFS) were 98% and 88%, respectively. The EFS in patients
    lymphocyte predominant Hodgkin lymphoma
    with or without RT was 88% and 87%. EFS did not differ between the
    COPP (vincristine, cyclophosphamide, procarbazine, prednisone)
    NLPHL is a rare form of HL accounting for 5–​10% of HL cases.
    and the COPDAC (vincristine, cyclophosphamide, dacarbazine,
    Clinical characteristics of patients with NLPHL are different, with
    prednisone) arm. In this trial, RT could be avoided in about 50% of
    a vast majority of boys, localized disease with peripheral lymph
    patients without impairing treatment results, and COPDAC is now
    nodes, and rarity of clinical or biological inflammatory symptoms.
    considered as standard consolidation. The EuroNet PHL-​C2 trial will
    Histopathology and immunophenotype shows typical CD20-​
    further focus on reducing RT indication by proposing a moderate in-
    positive ‘popcorn’ cells. A strategy based on limited treatment as-
    crease of chemotherapy in a randomized fashion.
    sociated to monoclonal anti-​CD20 antibodies in rare advanced
    Meanwhile, the current standard of care in Europe is induction
    cases, or ‘watch and wait’ after initial lymph node resection in stage
    treatment by two OEPA cycles in all patients, followed by one to four
    I limited disease, is currently proposed. Long-​term follow-​up is re-
    cycles of COPDAC in localized, intermediate, and advanced disease
    commended as subsequent high-​grade lymphomas have been de-
    groups; 20 Gy radiotherapy to involved sites is restricted to patients
    scribed in adult cohorts of patients with previous NLPHL.
    with inadequate response after two cycles of OEPA. Alternative
    chemotherapy strategies are being developed by the Children’s
    Oncology Group (COG), mainly with an ABV-​PC and DBVE-​PC
    Long-​term outcome and follow-​up after Hodgkin
    regimen and restricted use of RT in advanced stages, based on re-
    lymphoma during childhood and adolescence
    sponse after two or three cycles.
    In young adults with cHL, the ABVD regimen followed by 20–​30
    Secondary malignancies
    Gy RT to involved sites is the standard of care in localized dis-
    ease. In intermediate and advanced stages, ABVD and escalated The risk of secondary RT-​induced cancers is well documented
    BEACOPP provide excellent results, but cumulative doses of in cHL and has led to consensus guidelines for follow-​up and
    alkylating agents or anthracyclines, with or without RT, are not screening. Breast cancer is the most frequent secondary cancer
    considered adaptable to children or adolescents. Nevertheless, and needs screening from a median time of 8–​13 years after RT for
    studies are ongoing, with response-​adapted strategies leading to cHL. Other cancers are mainly gastric cancer and lung carcinoma.
    reduction of cumulative chemotherapy dose and restriction in in- The cumulative incidence of secondary cancers increases with time
    dications for RT. and reaches 25–​30% at 25-​year follow-​up. Leukaemia after cHL
    is reported mainly during the five first years following treatment.
    Treatment in relapsed and refractory classical variant Leukaemia has an incidence of less than 1% and is associated with
    of Hodgkin lymphoma alkylating agents or inhibitors of topoisomerase.
    Cumulative relapse rates in cHL are around 10–​15%. More than 90%
    Cardiovascular events
    arise within three years of diagnosis. In reports restricted to children
    and adolescents, the only prognostic factor associated with adverse Cardiovascular events are the second major cause of significant long-​
    outome in relapsed and refractory (RR) cHL is a time lag of less than term morbidity and premature deaths. Information to patients and re-
    12 months from the end of primary treatment. This has nevertheless to commendations about lifestyle, including avoiding smoking and other
    be tempered by initial stage and initial treatment, as well as stage at re- risk factors of cardiovascular morbidity, is essential. Echocardiography
    lapse. Recently, several publications have demonstrated [18F]FDG-​PET every three years, to detect valvular dysfunction and impaired car-
    response before intensification as an independent prognostic factor diac function, is recommended a fortiori in cases of mediastinal RT
    in RR cHL, leading to subsequent adapted strategies in this context. in association with anthracycline treatment. Significant premature
    Long-term outcome and follow-up after Hodgkin lymphoma during childhood and adolescence 161

    coronary disease in young adults after RT on the heart is possible, but Gorde-​Grosjean S, et al. (2012) Outcome of children and adolescents
    there is no consensus for systematic screening yet. with recurrent/​refractory classical Hodgkin lymphoma, a study
    from the Société Française de Lutte contre le Cancer des Enfants et
    Impaired fertility des Adolescents (SFCE). Br J Haematol 158, 649–​56.
    Impaired fertility in boys is related to cumulative doses of Jaffe ES, et al. (2008) WHO Classification of Tumors: Pathology and Genetics
    of Tumors of Haematopoietic and Lymphoid Tissues. Lyon: IARC Press.
    procarbazine and cyclofosphamide. In adolescents and young
    Hjalgrim H, et al. (2003) Characteristics of Hodgkin’s lymphoma after
    adults, sperm cryopreservation has to be proposed before treatment.
    infectious mononucleosis. N Engl J Med 349, 1324–​32.
    In girls and young women, fertility damage is mainly linked to pre-
    Kharazmi E, et al. (2015) Risk of familial classical Hodgkin lymphoma
    mature menopause or pelvic RT. Cryopreservation of oocytes before by relationship, histology, age, and sex: a joint study from five
    or after treatment should be discussed, particularly in cases of ad- Nordic countries. Blood 126, 1990–​5.
    vanced stage or relapse. Küppers R (2009) Molecular biology of Hodgkin lymphoma.
    Hematology Am Soc Hematol Educ Program 491–​6. doi: 10.1182/​
    Thyroid dysfunction
    asheducation-​2009
    Thyroid dysfunction occurs in 5–​10% of cases after five years and can Küppers R (2012) New insights in the biology of Hodgkin lymphoma.
    be detected by annual thyroid-​stimulating hormone (TSH) dosage. Hematology Am Soc Hematol Educ Program 328–​34. doi: 10.1182/​
    Thyroid echography is recommended every three to five years for asheducation-​2012.1.328
    screening of thyroid carcinoma. Lee AI, LaCasce AS (2009) Nodular lymphocyte predominant
    Hodgkin lymphoma. Oncologist 14, 739–​51.
    Late effects Mauz-​Körholz C, et al. (2007) Resection alone in 58 children with
    limited stage, lymphocyte-​ predominant Hodgkin lymphoma-​
    Late-​effect clinics are currently under development in some American
    —​ experience from the European network group on pediatric
    and European countries. Major challenges are to provide accurate
    Hodgkin lymphoma. Cancer 110, 179–​85.
    patient information and promote adherence to recommendations.
    Mauz-​Körholz C, et al. (2015) Pediatric Hodgkin lymphoma. J Clin
    Social and psychological distress and fatigue after cancer in child- Oncol 33, 2975–​85.
    hood and adolescence is better documented, with published results Moskowitz CH, et al. (2015) Brentuximab vedotin as consolida-
    from long-​term cohorts of patients treated for HL. Improvements in tion therapy after autologous stem-​ cell transplantation in pa-
    the care and management of these adverse effects are necessary. tients with Hodgkin’s lymphoma at risk of relapse or progression
    (AETHERA): a randomised, double-​ blind, placebo-​ controlled,
    phase 3 trial. Lancet 385, 1853–​62.
    FURTHER READING Nachman JB, et al. (2002) Children’s Cancer Group: randomized
    Ansell SM, et al. (2015) PD-​1 blockade with nivolumab in relapsed or comparison of low-​dose involved-​field radiotherapy and no radio-
    refractory Hodgkin’slymphoma. N Engl J Med, 372, 311–​19. therapy for children with Hodgkin’s disease who achieve a complete
    Bhatia S, et al. (2003) High risk of subsequent neoplasms con- response to chemotherapy. J Clin Oncol 20, 3765–​71.
    tinues with extended follow-​ up of childhood Hodgkin’s Ng AK, Mauch PM (2009) Late effects of Hodgkin’s disease and its
    lymphoma: report from the Late Effects Study Group. J Clin treatment. Cancer J 15, 164–​8.
    Oncol 21, 4386–​94. Pellegrino B, et al. (2003) Lymphocyte-​ predominant Hodgkin’s
    Bhakta N, et al. (2016) Cumulative burden of cardiovascular morbidity lymphoma in children: therapeutic abstention after initial lymph
    in paediatric, adolescent, and young adult survivors of Hodgkin’s node resection—​a study of the French Society of Pediatric Oncology.
    lymphoma: an analysis from the St Jude Lifetime Cohort Study. J Clin Oncol 21, 2948–​52.
    Lancet Oncol 17, 1325–​34. Purz S1, et al. (2011) [18F]fluorodeoxyglucose positron emission tom-
    Cerhan JR, Slager SL (2015) Familial predisposition and genetic risk ography for detection of bone marrow involvement in children and
    factors for lymphoma. Blood 126, 2265–​73. adolescents with Hodgkin’s lymphoma. J Clin Oncol 29, 3523–​8.
    Chen R, et al. (2017) Phase II study of the efficacy and safety of Roemer MG, et al. (2016) PD-​L1 and PD-​L2 genetic alterations define clas-
    pembrolizumab in relapsed/​refractory Hodgkin lymphoma J Clin sical Hodgkin lymphoma and predict outcome. J Clin Oncol 34, 2690–​7.
    Oncol 35, 2125–​32. Shankar A, et al. (2012) Treatment outcome after low intensity chemo-
    Daw S, et al. (2011) Management of relapsed and refractory classical therapy [CVP] in children and adolescents with early stage nodular
    Hodgkin’s lymphoma in children and adolescents. Br J Haematol lymphocyte predominant Hodgkin’s lymphoma—​an Anglo-​French
    152, 249–​60. collaborative report (2012). Eur J Cancer 48, 1700–​6.
    Friedman DL, et al. (2014) Dose-​intensive response-​based chemo- Sklar C, et al. (2000) Abnormalities of the thyroid in survivors of
    therapy and radiation therapy for children and adolescents with Hodgkin’s disease: data from the Childhood Cancer Survivor Study.
    newly diagnosed intermediate-​risk Hodgkin lymphoma: a report J Clin Endocrinol Metab 85, 3227–​32.
    from the Children’s Oncology Group Study AHOD0031. J Clin Steliarova-​Foucher E, et al. (2017) International incidence of child-
    Oncol 32, 3651–​58. hood cancer, 2001–​10: a population-​based registry study. Lancet
    Furth C, et al. (2009) Early and late therapy response assessment with Oncol 18, 719–​31.
    [18F]fluorodeoxyglucose positron emission tomography in pedi- Thorley-​Lawson DA, Gross A (2004) Persistence of the Epstein–​Barr
    atric Hodgkin’s lymphoma: analysis of a prospective multicenter virus and the origins of associated lymphomas. N Engl J Med 350,
    trial. J Clin Oncol 27, 4385–​91. 1328–​37.
    Gallamini A, et al. (2014) The predictive role of interim positron emis- Wallace WH, et al. (2005) Fertility preservation for young people with
    sion tomography for Hodgkin lymphoma treatment outcome is cancer: who is at risk and what can be offered? Lancet Oncol 6, 209–​18.
    confirmed using the interpretation criteria of the Deauville five-​ Younes A, et al. (2010) Brentuximab vedotin (SGN-​35) for relapsed
    point scale. Haematologica 99, 1107–​13. CD30-​positive lymphomas. N Engl J Med 363, 1812–​21.
    20
    Childhood Histiocytoses
    Maurizio Aricò, Cor van den Bos, and Sheila Weitzman

    Langerhans cell histiocytosis any area may be affected, including the nails. Cutaneous LCH can
    present with a wide range of clinical manifestations, from reddish
    Langerhans cell histiocytosis (LCH) is a rare clonal inflammatory papules and nodules to blisters, tumour-​like lesions, and scaling or
    myeloid neoplasm with a reported incidence in children of between purpuric macules.
    two and eight per million per year. In the recently revised classifica- Hepatomegaly is common in patients with disseminated dis-
    tion of the histiocytoses, it is placed in the L-​group, together with ease. Evidence of altered liver function (e.g. hypoalbuminaemia)
    Erdheim–​Chester disease and indeterminate cell histiocytosis. The is less commonly observed. A common complication of liver LCH
    diagnosis is based on pathological findings in the correct clinical is the development of sclerosing cholangitis (SC), which may be
    setting. Lesions are characterized by infiltration by cells originating present at diagnosis. In patients with SC, the progression of liver
    from the myelomonocytic lineage—​the LCH cells. Morphologically, damage may become independent of LCH activity, and progression
    LCH cells are mononuclear, express CD1a and/​or CD207 (Langerin), to end-​stage liver failure can occur. Spleen involvement presents as
    have coffee-​bean-​shaped or kidney-​shaped nuclei, and are often ac- splenomegaly.
    companied by abundant eosinophils and multinucleated giant cells. Anaemia, leukocytopenia, and thrombocytopenia are manifest-
    It is a disorder with a striking clinical heterogeneity, ranging from ations of ‘bone-​marrow involvement’. These cytopenias most often
    spontaneously healing single-​system bone and skin lesions to a life-​ occur in the absence of morphological bone-​marrow infiltration.
    threatening multisystem disease. Pathological confirmation of LCH in bone marrow is unnecessary
    Recently, considerable progress has been made in the under- if bi-​or tri-​lineage cytopenia is present that is not explained by an-
    standing of LCH biology. In 2010, the constitutive activation of the other cause. Haemophagocytosis can be observed in severe forms of
    RAS–​RAF–​MEK–​ERK pathway in LCH was first described. It was LCH, reflecting the hyperinflammatory state of the patient.
    shown that, in approximately half the cases, the BRAF-​V600E muta- In children, pulmonary LCH is usually part of multisystem dis-
    tion is present. Other mutations in BRAF (including small in-​frame ease, although in adolescents, isolated lung LCH is sometimes ob-
    deletions) and in other pathway genes such as ARAF, MAP2K,as well served. Clinically, lung involvement leads to respiratory distress with
    as FAM73A-​BRAF fusions have subsequently been described. On tachypnoea, retractions, and persistent cough. Rupture of superficial
    the basis of further studies in patients and murine models, a model cysts may lead to pneumothorax. As in the liver, progression of lung
    has been postulated whereby somatic mutations in an early myeloid fibrosis may become independent of LCH activity.
    progenitor cell lead to multisystem LCH, while mutations in more Diabetes insipidus (DI) is the most commonly observed endo-
    differentiated cells lead to localized, single-​system LCH. crine complication in LCH, involving up to 35% of children with
    LCH. DI may present as the first manifestation of LCH, or it may
    Clinical manifestations of Langerhans cell histiocytosis occur later, months or even many years after diagnosis. Patients with
    LCH can manifest itself in many tissues and organs. Bone is most DI may progress to multiple anterior pituitary hormone deficiencies,
    frequently affected, being observed in 80–​100% of cases in large the first usually being growth hormone deficiency.
    series, most commonly presenting as painful swellings. The skull, In the central nervous system, LCH can manifest as cerebral mass
    long bones, and then flat bones are most frequently involved. In the lesions. In a minority of patients, most often with preceding DI, a
    spine, typical vertebra plana may be observed. In the case of iso- progressive neurodegeneration (ND) can develop with ataxia, as well
    lated vertebra plana without a soft-​tissue component, the risk of as cranial nerve palsies and neuropsychological defects. Biopsies of
    biopsy may outweigh the need for tissue diagnosis, but the patient LCH ND lesions do not show CD1a positive cells, but more recent
    should be closely followed to exclude malignancy. Clinical features biopsies have demonstrated presence of the BRAF-​V600E mutation
    and possible complications of osseous lesions depend on the bone suggesting that inhibitor therapy might be effective.
    involved. Involvement of the gastrointestinal tract can manifest as intestinal
    Skin involvement is observed in over one third of children with malabsorption, (bloody) diarrhoea, or protein-​losing enteropathy.
    LCH. Most frequently affected sites are scalp and nappy areas, but Lymphadenopathy is seen in a minority of patients.
    Langerhans cell histiocytosis 163

    Table 20.1 Clinical classification of Langerhans cell histiocytosis Table 20.2 Response evaluation in paediatric Langerhans cell
    histiocytosis
    Main category Subcategories Definition
    Single-​system Unifocal bone One bone involved Main category Subcategories Definition
    LCH (SS LCH) Non-​active disease (NAD) Resolution of all
    Multifocal bone More than one bone
    involved signs/​symptoms (no
    evidence of disease)
    Skin
    Active disease (AD) Regressive disease Regression of signs/​
    Lymph node Better symptoms, no new
    lesions
    Central nervous system
    Active disease (AD) Stable disease Persistence of signs
    Isolated lung
    Stable or symptoms, no new
    Other organs lesions
    Multisystem MS LCH RO negative (RO–​) Two or more organs/​systems Source: data from ‘LCH-​IV, International Collaborative Treatment Protocol for Children
    LCH (MS LCH) involved without liver, and Adolescents with Langerhans Cell Histiocytosis’, NIH US Library of Medicine, https://​
    spleen, or bone marrow clinicaltrials.gov/​show/​NCT02205762
    MS LCH RO positive (RO+) Two or more organs/​systems
    involved with liver and/​or
    spleen and/​or bone marrow DI is confirmed by a water deprivation test and imaging of the pi-
    tuitary stalk. MRI may show absence of the posterior bright spot on
    Source: data from ‘LCH-​IV, International Collaborative Treatment Protocol for Children
    and Adolescents with Langerhans Cell Histiocytosis’, NIH US Library of Medicine, https://​ T2-​weighted imaging and thickening of the pituitary stalk.
    clinicaltrials.gov/​show/​NCT02205762 Evaluation of the efficacy of treatment is not always easy. Criteria
    have been developed by the Histiocyte Society (Table 20.2), but
    the determination of disease activity in the most common lesions
    (i.e. bone lesions) can be especially problematic. For patients with
    Clinical classification and evaluation of Langerhans MS LCH, a disease activity score has been developed that includes
    cell histiocytosis both clinical and laboratory parameters such as albumin levels. The
    The diagnosis of LCH is made pathologically, together with the clin- recent studies reporting measures of levels of circulating mutated
    ical picture. LCH can present as single-​system (SS) disease, at one DNA (circulating tumour DNA; ctDNA) need further confirmation
    site (unifocal), or at multiple sites (multifocal). In multisystem (MS) to allow determination of clinical relevance.
    LCH, the disease involves multiple tissues or organs. MS LCH is Progression of skeletal lesions is defined as unequivocal enlarge-
    subdivided into ‘risk-​organ positive (RO+)’ and ‘risk-​organ negative ment of the size of existing lesions and/​or appearance of new lesions
    (RO–​)’. MS-​RO+ LCH is defined by involvement of ‘risk’ organs—​
    liver, spleen, and haematopoetic system. Involvement of risk organs Treatment of paediatric Langerhans cell histiocytosis
    is associated with the risk of mortality (Table 20.1). Treatment of paediatric LCH should be directed by the extent and
    Thorough clinical evaluation at diagnosis is obligatory to de- quality of disease involvement. While for patients with involvement
    termine disease extent and treatment. For bony lesions, plain of vital organs (liver, spleen, haematopoietic system)—​which bears a
    radiography remains the first-​line approach for detection. In the risk of mortality—​systemic therapy is warranted, for single-​bone le-
    current LCH-​IV protocol of the Histiocyte Society, only the bone sions and isolated skin LCH, a ‘wait and see’ policy is often sufficient.
    lesions detected by plain radiography are considered for stratifi-
    cation. Lesions detected by alternative imaging, like Tc-​bone scan Single-​system skin Langerhans cell histiocytosis
    or fluorodeoxyglucose-​ positron emission tomography/​ computed The only indication for therapy in skin-​only LCH is bleeding, ulcer-
    tomography (FDG-​PET/​CT), will be documented but, according ation, or pain. However, skin-​only LCH in very young infants can
    to the LCH-​IV protocol ‘are not considered for stratification unless progress within weeks to months to high-​risk MS LCH, and these
    proven by plain radiography or MRI or pathology’. Total-​body mag- infants should be closely watched. For patients who deserve therapy
    netic resonance imaging (MRI), as well as FDG-​PET/​CT, have been for skin LCH, numerous approaches have been published: topical
    shown to detect lesions not otherwise visualized, but their clinical corticosteroids, nitrogen mustard, or tacrolimus; narrow band UVB
    relevance is unproven. Nonetheless, some studies have upstaged or PUVA. If systemic therapy is required, low-​dose methotrexate,
    patients based on these results. Results of the LCH-​IV study might alone or with the addition of 6 mercaptopurine, can be tried first
    allow insight into the clinical relevance of these total-​body imaging before going onto the more standard chemotherapy approaches
    techniques. Only a well-​designed prospective study, however, will described below. Oral etoposide or azathioprine, which have been
    yield definitive results. An MRI of the spine is indicated in the case shown to be useful therapy in adults, are rarely used in paediatrics.
    of vertebra plana to exclude spinal cord compression from an asso-
    ciated soft-​tissue mass. Single-​system bone Langerhans cell histiocytosis
    Abdominal ultrasound and physical examination are used to de- For unifocal disease, curettage alone allows for pathological diagnosis
    tect hepatosplenomegaly. Liver function is evaluated by standard and is often followed by spontaneous healing. If necessary, intralesional
    laboratory tests, including albumin levels and coagulation param- methylprednisolone can be safely used but, to date, no study has ever
    eters. A chest X-​ray is routine at diagnosis. In case of suspected pul- shown a clear superiority of intralesional steroids over observation
    monary involvement, a CT scan of the lungs is mandatory. If age only. The exception to observation alone for unifocal bone disease in-
    appropriate, lung function studies should be performed. cludes a vertebral body lesion with a soft-​tissue mass, with risk to the
    164 CHAPTER 20 Childhood Histiocytoses

    spinal cord. Craniofacial bone lesions have been reported to be asso- potentials, as well as cerebrospinal fluid (CSF) cytokine patterns or
    ciated with increased risk of developing DI and LCH-​associated ND continued BRAF mutation positivity in blood, are being studied for
    (‘CNS-​risk lesions’), thus deserving systemic chemotherapy. However, their ability to identify patients in need of treatment. A major obs-
    this ‘CNS-​risk lesion’ is still a matter of debate. Multifocal bone disease tacle is the absence of a clearly effective therapy. Intravenous im-
    is usually an indication for systemic treatment. munoglobulin (IVIG), cytarabine, and retinoic acid have been used.
    The efficacy of those treatments appears to depend on early diagnosis
    Multisystem Langerhans cell histiocytosis
    and therapy, which generally results in prevention of progression at
    Multisystem LCH is always treated systemically. The backbones most best. Based on the finding of positive immunostaining for the BRAF-​
    broadly used are either vinblastine–​prednisone (the mainstay of the V600E mutation in ND LCH lesions, targeted therapy has been given
    four consecutive LCH I–​IV trials), or the combination of vincristine–​ to a limited number of patients. Results of these early-​phase studies
    cytarabine–​prednisone. Recently, single-​agent cytarabine has been and the development of specific trials are eagerly awaited. The neuro-
    suggested but remains to be proven, particularly in high-​risk young psychological consequences of ND LCH may be severe.
    infants. Over the years, clinical studies performed by different study
    groups have led to the following conclusions:
    • In patients with MS-​RO+ LCH, early response is an important
    Haemophagocytic lymphohistiocytosis
    predictor of survival, necessitating a timely switch to intensified
    therapy in non-​responders. Haemophagocytic lymphohistiocytosis (HLH) is a disorder character-
    ized by a hyperinflammatory syndrome with fever, hepatosplenomegaly,
    • In partial responders or non-​responders without risk organ, a second
    cytopenia, liver dysfunction, and sometimes CNS involvement, which
    induction treatment has been shown to increase the NAD rate.
    may lead to multi-​organ failure and death. The disease may be in-
    • Prolongation of treatment to 12 months decreases the reactiva-
    herited (so-​called ‘primary’) or apparently sporadic, and in these
    tion rate in high-​and low-​risk MS LCH patients from 50% to
    cases is usually triggered by malignancies, rheumatological disorders,
    around 30%.
    or infections (so-​called ‘secondary’). The familial form of the disease
    To date, none of the treatment backbones has been shown to be (FHL) is genetically heterogeneous, involving genes important to the
    clearly superior, and the optimal treatment duration for both multi- normal function of cytotoxic T cells and natural killer (NK) cells. The
    focal bone LCH and MS LCH is still being studied (e.g. in the cur- recent discovery of sporadic HLH patients with monoallelic mutations
    rent Histiocyte Society LCH-​IV protocol,12-​month versus 24-​month or polymorphisms in typical FHL-​associated genes, as well as non-​
    duration). There is also an ongoing need for better understanding of cytotoxicity-​related genes, has led to the blurring of the distinction be-
    the disease, since it is impossible to distinguish, at diagnosis or early tween “primary” and “secondary” HLH. Nonetheless, it remains true
    in treatment, the MS LCH patients who will benefit from long main- that patients with familial disease (FHL) have an earlier onset, a more
    tenance from those for whom treatment prolongation is unnecessary. aggressive clinical course, and need allogeneic transplantation for cure.
    For non-​responding MS-​RO+ LCH patients, the combination of Otherwise, patients who lack biallelic mutations in FHL-​related genes
    high-​dose cytarabine and cladribine has been shown to be an effective usually have a later onset, a less aggressive clinical course, and do not
    but toxic salvage therapy, as has allogeneic stem-​cell transplantation. require haematopoietic stem-​cell transplantation for cure.
    More recently, inhibitors to mutated BRAF have been shown to lead The disease manifestations are due to excessive inflammatory
    to a quick response in those patients, but after cessation of short-​ response caused by hypersecretion of pro-​inflammatory cytokines
    term treatment, reactivation of the disease has been observed. Based such as interferon-​γ, tumour necrosis factor-​α, interleukin(IL)-​6,
    on the present knowledge, the use of targeted therapy with inhibi- and macrophage-​colony-​stimulating factor (M-​CSF). These medi-
    tors appears very useful to speed up initial disease control, but the ators are secreted by activated T lymphocytes and macrophages that
    combination with standard LCH-​directed chemotherapy appears infiltrate all tissues, and lead to tissue necrosis and organ failure.
    still warranted. Future cooperative studies addressing the combin- Differential diagnosis of HLH may be difficult and diagnostic
    ation of inhibitors (such as BRAF and MEK inhibitors) and the use guidelines for HLH have been established by the Histiocyte Society
    of limited chemotherapy appear justified. (Box 20.1).
    There are ongoing efforts among specialists to replace these cri-
    Permanent consequences of Langerhans teria with others more specific for the diagnosis. The clinical rele-
    cell histiocytosis vance of items such as hypertriglyceridaemia, hypofibrinogenaemia,
    There is a long list of potential permanent consequences of LCH, but haemophagocytosis, and NK activity testing is questionable, and for dif-
    the commonest and most important relate to the endocrine system ferent reasons, including low sensitivity or limited access. Bone-​marrow
    and the CNS. haemophagocytosis, for example, is neither specific nor sensitive for
    The commonest late effect, DI, can be managed with oral or HLH and, contrary to common belief, is not by itself a useful criterion.
    intranasal desmopressin. Other hormonal deficiencies such as The combination of fever, splenomegaly, thrombocytopenia, and
    growth-​hormone deficiency or secondary hypothyroidism require hyperferritinaemia (and/​or high levels of sCD25) may positively
    substitution therapy. identify the population deserving an immediate genetic and immuno-
    Management of ND remains a major problem and represent a really logical work-​up such as the granule release assay (see below). This is
    unmet need for patients with LCH. On MRI, the characteristic pat- only useful in centres where these tests are readily available, however.
    tern of usually bilateral and symmetrical demyelination starting from Additional laboratory tests that are not ‘diagnostic criteria’ but
    the cerebellum is seen. Not all patients with MRI signs of ND will de- that may help in diagnosis and in evaluation of therapeutic re-
    velop clinical ND, and neurophysiological tests such as basal evoked sponse, include measurement of transaminases and bilirubin,
    Haemophagocytic lymphohistiocytosis 165

    Box 20.1 Clinical and laboratory criteria for haemophagocytic with cases documented from all continents. Age at disease onset is
    lymphohistiocytosis* usually very young, median three months, but with a wide range
    reaching the third decade. Mutations of UNC13D have been asso-
    • Fever ciated with FHL3. It encodes the Munc13–​4 protein, which results
    • Splenomegaly
    in defective cellular cytotoxicity and a clinical picture very similar
    • Cytopenia affecting at least two of three lineages in the
    peripheral blood:
    to FHL2. Genotype/​phenotype correlations do not identify ethnic-​
    • Haemoglobin <9 g/​dl (below 4 weeks <12 g/​dl) specific mutations; CNS involvement is more frequent and age at
    • Neutrophils <1 × 103/​mm3 diagnosis is significantly higher than FHL2. FHL4 is due to defi-
    • Platelets <100 × 103/​mm3 ciency of syntaxin 11, a protein that contributes to cytotoxic cell de-
    • Hypertriglyceridaemia (fasting, ≥265 mg/​dl) and/​or granulation; most patients are of Turkish/​Kurdish descent. FHL5 is
    hypofibrinogenaemia (<150 mg/​dl) due to mutations in the Munc18–​2 gene encoding syntaxin-​binding
    • Ferritin ≥500 µg/​l protein-​2.
    • Soluble CD25 (that is, soluble IL-​2 receptor) ≥2400 U/​ml
    A few additional genetic conditions may have a clinical picture
    • Decreased or absent NK-​cell activity
    overlapping HLH. Chédiak–​ Higashi syndrome (CHS), Griscelli
    In patients with familial disease or known genetic defect (i.e. biallelic
    syndrome type II (GSII), and X-​linked lymphoproliferative syn-
    pathogenic mutations in genes related to the autosomal recessive form
    of the disease or single mutations in males with the X-​linked forms of drome (XLP) types 1 and 2 are immune deficiencies with distinctive
    disease), the diagnosis can be established even with incomplete criteria. clinical features in which the development of HLH is sporadic,
    * Five of the criteria are considered diagnostic though frequent. HLH is often the presenting symptom, but may
    Adapted with permission from Henter JI et al. ‘HLH-​2004: Diagnostic and Therapeutic also occur later in the disease. Patients with CHS and GSII show
    Guidelines for Hemophagocytic Lymphohistiocytosis’. Pediatric Blood and Cancer, partial albinism as melanin granules move by the pathway de-
    Volume 48, Issue 2, pp. 124–​ 131. Copyright © 2006, John Wiley and Sons.
    DOI: https://​doi.org/​10.1002/​pbc.21039 scribed earlier. XLP1 is mainly characterized by a defect of specific
    effective immune responses to EBV. Following exposure to EBV,
    these subjects are prone to develop HLH; if they survive, they may
    lactate dehydrogenase (LDH), and D-​dimers (even when INR, PTT, develop lymphoma or dysgammaglobulinaemia. Similarly, patients
    and fibrinogen are normal), and CSF pleocytosis. Hyponatraemia with XLP2 may develop HLH as part of their clinically heterogenous
    and hypoalbuminaemia may also be present. The utility of requiring phenotype.
    multiple clinical and laboratory abnormalities, none of which are Prompt identification of FHL in the absence of a positive family
    specific for diagnosis, is that they reflect the uncontrolled hyper- history is an ongoing challenge, even in young children. The ap-
    immune status which is damaging the body—​a criterion which has proach to this problem has changed over the years. Initially, the
    been called ‘immunopathology’ by some investigators. diagnosis was based on the identification of HLH in very young
    patients with recurrence of disease after cessation of therapy
    Pathogenesis of haemophagocytic lymphohistiocytosis and/​or a positive family history. Identification of the genetic de-
    The pathogenic mechanisms of FHL are based on insufficient con- fects has allowed investigation of novel tools for rapid screening
    trol of target antigens by cellular cytotoxicity. As a result, excessive of FHL. For example, perforin protein can be measured by flow
    cytokine production induces lymphocyte overstimulation. In this cytometry, providing a reliable and rapid identification of FHL-​
    vicious loop, the patient is unable to get rid of the target, often a 2 patients. Other proteins measurable by flow cytometry include
    common viral pathogen such as Epstein–​Barr virus (EBV) or cyto- UNCD (MUNC13.4) and SAP (XLP1). In addition, flow cytometry
    megalovirus (CMV). Failure to terminate the immune response (de- has been shown to be a useful screening tool for absence of de-
    fect in immune homeostasis) is also believed to be important. granulation of the PRF/​granzyme-​containing granules. CD107a
    Within the cytoplasm of NK cells and cytotoxic lymphocytes (LAMP-​1) lines the cytotoxic granule to protect the host cell from
    (CTLs) are cytolytic granules containing perforin and granzyme. damage by granule content and is expressed on the surface of the
    Upon contact between the effector killer cells and their target, an NK/​T cell after granule exocytosis. Lack of surface CD107a expres-
    immunological synapse is formed and cytolytic granules traffic to sion represents a rapid tool for identification of patients with FHL
    the contact site, dock, and fuse with the plasma membrane and and degranulation defects, in contrast to healthy control subjects
    release their contents into the synapse. Perforin, a pore-​forming or perforin-​deficient NK cells where degranulation is normal but
    protein, produces a channel in the target-​cell membrane allowing perforin is absent. Thus, this assay rapidly became the standard
    entry of granzymes, activation of caspases, and apoptosis of target for identification of patients with degranulation defects who need
    cells. Most known defects in FHL seem to be involved in this pro- formal genetic sequencing studies.
    cess, although there remains a very small minority of patients New methodologies, such as next-​generation sequencing (NGS),
    with unassigned defects. An increasing number of other con- are now available for genetic testing, allowing all HLH-​associated
    genital immunodeficiencies, such as X-​linked SCID and X-​line genes to be evaluated at significantly lower cost.
    hypogammaglobulinaemia, have been described as being associ- The aforementioned tools are extremely useful in the diagnostic
    ated with HLH. approach to a child with HLH, allowing discrimination of those pa-
    tients with a genetic defect who require aggressive therapy followed
    Genetic forms of haemophagocytic lymphohistiocytosis by haematopoietic stem-​cell transplantation (HSCT), from those in
    The identification of the causative gene(s) of FHL has been a long whom HLH may develop as a temporary complication of another
    process. In 1999, the first defect in the perforin gene (PRF1) was underlying condition, who nonetheless require anti-​HLH therapy
    described (FHL2). Patients with FHL2 have different ethnic origins, but who do not generally require HSCT.
    166 CHAPTER 20 Childhood Histiocytoses

    Therapy of haemophagocytic lymphohistiocytosis The juvenile xanthogranuloma family


    In most cases, the natural course of FHL is rapidly fatal within a The JXG family comprises a number of the UCH disorders with
    few weeks, unless appropriate treatment, including corticosteroids, an identical immunophenotype, with lesional cells positive for
    etoposide, or anti-​thymocyte globuline can obtain transient disease vimentin, factor XIIIa, CD68, CD163, fascin, CD68, and CD14, and
    control. Therapy duration for good responders is usually around negative for CD1a, S100, and langerin. Pathologically, they show
    eight weeks. well-​circumscribed nodules with a dense infiltrate of histiocytes.
    In FHL, disease reactivation is inevitable, leading to a rapidly Touton giant cells, seen in 85% of cases of JXG, but not limited to
    fatal outcome unless the immune system cells with a defective JXG, are characterized by a wreath of nuclei around a homogenous
    gene are replaced by allogeneic HSCT. The very high transplant-​ eosinophilic cytoplasmic centre, while the periphery shows prom-
    related mortality associated with HSCT, for both LCH and HLH, inent xanthomatization. Mutations in BRAF have been shown in a
    has led to current trials of reduced-​intensity conditioning for minority of JXG patients.
    these diseases. The commonest members of the JXG family in childhood are be-
    For patients with reactivated FHL as well as patients with insuffi- nign cephalic histiocytosis (BCH) and JXG. BCH presents with mul-
    cient response to standard front-​line therapy, recent introduction of tiple cutaneous lesions confined to the head and neck area in a young
    a novel anti-​gamma interferon agent (emapalumab) may represent a child. These lesions either spontaneously disappear or progress to
    novel therapeutic opportunity, which deserves prospective random- JXG clinically or pathologically. No therapy is necessary. Cutaneous
    ized evaluation versus standard therapeutic approach. JXG is a benign proliferative disorder that usually resolves spontan-
    Gene replacement therapy, although very appealing as in other eously within one to five years but may leave residual scarring. Long-​
    monogenic diseases, and shown to be effective in mouse models, is term sequelae of JXG are rarely reported. Diagnosis is made on
    unlikely to be a reality soon for most patients, although human trials biopsy and immunohistochemistry. Extensive work-​up should be
    are starting in several countries. reserved for those with clinical suspicion of systemic involvement.
    Median age of onset of JXG is two years, but lesions may be pre-
    Macrophage activation syndrome sent at birth. Most JXG presents with solitary lesions which vary in
    The term ‘macrophage activation syndrome’ (MAS) has been used size. Infants less than six months old tend to present with multiple le-
    to describe patients with rheumatological diseases, most com- sions, with a predisposition for the head and neck. The male prepon-
    monly sJIA, who develop a clinical syndrome similar to HLH. derance is much higher (12:1) in young infants with multiple skin
    Many patients with sJIA-​associated MAS have been shown to have lesions. JXG is associated with other diseases including the triad of
    monoallelic defects in the FHL genes. Different and less stringent JXG, neurofibromatosis type 1 (NF1), and juvenile myelomonocytic
    diagnostic criteria have been established for sJIA MAS, and the pa- leukaemia (JMML).
    tients tend to respond to less aggressive therapy such as corticoster- Extracutaneous involvement (systemic JXG) occurs in around
    oids and IVIG. Anti-​cytokine therapy, such as the inhibition of the 4% of children and 30% overall, with half of the patients having no
    IL1 receptor by anakinra and others, is being investigated in these skin involvement. Median age at presentation is 0.3 years. The com-
    patients. monest extracutaneous site is a solitary mass in the subcutis and/​
    or deeper soft tissues, followed by the liver, spleen, CNS, and lungs.
    Malignancy-​associated Like cutaneous JXG, most systemic lesions undergo spontaneous
    haemophagocytic lymphohistiocytosis involution, but ocular and CNS involvement may cause signifi-
    HLH can be triggered by malignancy, seen most often in mature cant problems and fatality. CNS involvement may result in seizures,
    T-​cell malignancies including anaplastic large-​cell lymphoma, but ataxia, increased intracranial pressure, subdural effusions, develop-
    also described in B-​cell disease including Hodgkin lymphoma and mental delay, DI, and other neurological deficits. In one series, two
    diffuse large B-​cell lymphoma. Therapy consists of treatment of the out of thirty-​six children died of progressive CNS disease, while in
    underlying malignancy, but specific HLH therapy may also be re- another, two neonates died of hepatic failure.
    quired. HLH in cancer patients may also occur later, in patients in Ocular JXG occurs in the very young child (92% are less than two
    remission, and is thought to be due to infection and chemotherapy-​ years of age), and may be without skin involvement. When skin le-
    induced immunosuppression. Treatment of this latter form includes sions coexist, they are always multiple. Eye involvement is usually
    therapy of the underlying infection and possibly delay in chemo- unilateral and commonly presents with an asymptomatic iris tu-
    therapy until recovery, but corticosteroid and/​or etoposide therapy mour, a red eye with signs of uveitis, unilateral glaucoma, spon-
    may be required to control the HLH until regular chemotherapy can taneous hyphaema, or heterochromia iridis. Less commonly, other
    be restarted. areas of the eye may be involved. Early diagnosis and treatment de-
    termine the final visual outcome. Ophthalmological surveillance is
    recommended for high-​risk patients less than two years of age, who
    Uncommon histiocytic disorders should undergo screening every three to six months until the age
    of three.
    Uncommon histiocytic (UCH) disorders are a diverse group of dis-
    orders that do not meet diagnostic criteria for LCH or HLH. The com- Therapy
    monest UCH disorders in childhood are juvenile xanthogranuloma The management of JXG depends upon the site(s) of involvement.
    (JXG) and sinus histiocytosis with massive lymphadenopathy Patients with JXG isolated to skin should be treated only for pain or
    (Rosai–​Dorfman disease). ulceration, and not for cosmetic appearance. Single and accessible
    Rosai–Dorfman disease 167

    lesions may be amenable to surgical excision. For systemic JXG, there RDD is uncommon but carries a significantly worse prognosis.
    is currently no established standard chemotherapeutic regimen. While skin involvement occurs commonly, disease limited to skin
    A variety of regimens have been tried with variable results. LCH-​like is uncommon.
    therapy, with corticosteroids alone or with vinblastine or vincristine, The clinical course of SHML is unpredictable, with episodes of
    appear to have the highest response rate, but chemotherapy response exacerbation and remission which may extend over many years. The
    is less predictable than in LCH. Patients responding to cladribine outcome is usually good and disease is often self-​limiting; nonethe-
    and cytosine arabinoside, alone and in combination, have been re- less, about 5% of patients die from disease.
    ported, and clofarabine has been reported to have activity in a small
    number of heavily pretreated, refractory patients. Other published Therapy
    therapies include methotrexate, etoposide, cyclosporine, interferon, Patients with RDD isolated to nodes should be observed without
    and radiation therapy. A trial of different agents may be needed in therapy. Surgery may be indicated when localized nodal obstruc-
    individual patients. tion of a vital organ is present. First-​line systemic therapy, utilized
    For ocular JXG, therapy includes topical, intralesional, and for patients with vital organ compression or dysfunction, is usually
    subconjunctival corticosteroids; surgery may be required to treat corticosteroids, which are often effective, but off-​therapy relapses
    complications such as hyphaema and glaucoma; and systemic cor- are common. Chemotherapy, including vinblastine (alone or in
    ticosteroids, chemotherapy, or low-​dose ‘non-​cataractogenic’ radi- combination with corticosteroids), cytosine arabinoside, metho-
    ation therapy (300–​400 cGy) may be required for non-​responders. trexate, and other agents, produces variable to poor responses.
    CNS involvement has been successfully treated with cladribine, but Cladribine, alone and in combination, has been successful in some
    the disease does not always respond. Thalidomide has been tried patients, as has clofarabine. Imatinib, targeting platelet-​derived
    successfully in a few cases. growth factor receptor and KIT-​positive tumours, has resulted in
    As in LCH, when BRAF mutations are found in JXG, inhibitor good responses in some patients but not others. Rituximab therapy
    therapy may be a useful therapy option. induced remission in a child with long-​standing refractory disease.
    Other innovative therapies (including interferon-​ alfa, thalido-
    mide, and retinoids) have been reported, and therapies that target
    Rosai–​Dorfman disease cytokines such as TNF-​alfa and IL6 have been used successfully
    in some patients. Recurrence of SHML after cessation of therapy
    Sinus histiocytosis with massive lymphadenopathy (SHML), or (including after rituximab, interferon, and thalidomide) may
    Rosai-​Dorfman disease (RDD), is a non-​ neoplastic, polyclonal, occur, but retreatment with the same agents may induce a second
    usually self-​limited disease due to accumulation of S100+, CD1a−, response.
    CD68++, CD163++ histiocytes. Pathologically, lymph nodes show Treatment options for CNS and orbital RDD include surgery for
    massive sinus infiltration of large histiocytes admixed with lympho- diagnosis, relief of compression or excision if possible, stereotactic
    cytes and plasma cells. Intact erythrocytes, lymphocytes, and fractionated radiotherapy or radiosurgery, or chemotherapy (as out-
    plasma cells may be engulfed by histiocytes—​a process known as lined earlier). As with JXG, different therapies may have to be tried
    emperipolesis. Emperipolesis in histiocytes expressing S100 in the with an individual patient until a response is obtained.
    appropriate clinico-​pathological setting is considered diagnostic of
    SHML. The mean age of onset is 20.6 years, with a wide age distribu-
    tion. A familial form of RDD with mutations in SLC29A3 has recently FURTHER READING
    been described. The aetiology of SHML is unknown. An increased Abla O, Janka G (eds) (2018) Histiocytic Disorders. Switzerland: Springer
    incidence of autoimmune disorders, such as autoimmune haemolytic International Publishing.
    anaemia, and various rheumatological disorders have been reported. Abla O, Weitzman S (2015) Treatment of Langerhans cell histiocytosis:
    Evaluation of pathology specimens for BRAF-​V600E mutations in role of BRAF/​MAPK inhibition. Hematology Am Soc Hematol Educ
    SHML have shown no mutations except in a few adult patients pre- Program 565–​70. doi: 10.1182/​asheducation-​2015.1.565
    Allen CE, et al. (2015) How I treat Langerhans cell histiocytosis. Blood
    senting with simultaneous SHML and Erdheim–​Chester disease.
    126, 26–​35.
    Clinical presentation Aricò M, et al. (2001) Pathogenesis of haemophagocytic lympho­
    histiocytosis. Br J Haematol 114, 761–​9.
    The commonest presentation is with bilateral painless cervical Aricò M (2016) Langerhans cell histiocytosis in children: from the
    adenopathy associated with fever, night sweats, malaise, weight loss, bench to bedside for an updated therapy. Br J Haematol 173, 663–​70.
    and polyclonal hypergammaglobulinaemia. Other nodal groups Badalian-​Very G, et al. (2013) Pathogenesis of Langerhans cell
    may be involved, and involvement of extranodal tissue such as skin, histiocytosis. Ann Rev Pathol 8, 1–​20.
    soft tissue, upper respiratory tract, bone, eye, retro-​orbital tissue, and Berres ML, et al. (2014) BRAF-​V600E expression in precursor versus
    other sites occurs in 43% of cases as the sole initial manifestation or differentiated dendritic cells defines clinically distinct LCH risk
    with lymphadenopathy. Involvement of the CNS, seen in 4% (8 out groups. J Exp Med 211, 669–​83.
    of 200) of RDD cases in an early series, poses a significant diagnostic Brisse E, et al. (2016) Advances in the pathogenesis of primary and
    and therapeutic challenge, usually occurring without extracranial secondary hemophagocytic lymphohistiocytosis: differences and
    lymphadenopathy and clinically and radiologically resembling similarities. Br J Haematol 174, 203–​17.
    Brown NA, et al. (2014) High prevalence of somatic MAP2K1 muta-
    meningioma due to the propensity of all of the intracranial UCH
    tions in BRAF V600E-​negative Langerhans cell histiocytosis. Blood
    disorders to be dural-​based. Patients presenting with isolated intra-
    124, 1655–​8.
    cranial disease tend to be older (mean age 37.5 years). Pulmonary
    168 CHAPTER 20 Childhood Histiocytoses

    Bryceson YT, et al. (2012) A prospective evaluation of degranula- Maghnie M, et al. (2000) Central diabetes insipidus in children and
    tion assays in the rapid diagnosis of familial hemophagocytic syn- young adults. N Engl J Med 343, 998–​1007.
    dromes. Blood 119, 2754–​63. Marcenaro S, et al. (2006) Analysis of natural killer-​cell function in
    Cetica V, et al. (2016) Genetic predisposition to hemophagocytic familial hemophagocytic lymphohistiocytosis (FHL): defective
    lymphohistiocytosis: report on 500 patients from the Italian CD107a surface expression heralds Munc13-​4 defect and dis-
    registry. J Allergy Clin Immunol 137, 188–​96. criminates between genetic subtypes of the disease. Blood 108,
    Delprat C, Aricò M (2014) Blood spotlight on Langerhans cell 2316–​23.
    histiocytosis. Blood 124, 867–​72. Meazza R, et al. (2016) Diagnosing XLP1 in patients with
    Donadieu J, et al. (2015) Cladribine and cytarabine in refractory hemophagocytic lymphohistiocytosis. J Allergy Clin Immunol 134,
    multisystem Langerhans cell histiocytosis: results of an inter- 1381–​7.
    national phase 2 study. Blood 126, 1415–​23. Picarsic J, et al. (2019) BRAF V600E mutation in Juvenile
    Donadieu J, et al. (2019) Vemurafenib for refractory multisystem Xanthogranuloma family neoplasms of the central nervous system
    Langerhans Cell Histiocytosis in children: An international obser- (CNS-JXG): a revised diagnostic algorithm to include pediatric
    vational study. J Clin Oncol 37, 2857–65. Erdheim-Chester disease. Acta Neuropathol Commun 7(1), 168.
    Eckstein OS, et al. (2019) Clinical responses and persistent BRAF doi:10.1186/s40478-019-0811-6.
    V600E+ blood cells in children with LCH treated with MAPK Ravelli A, et al. (2016) Classification criteria for macrophage activa-
    pathway inhibition. Blood 133, 1691–94. tion syndrome complicating systemic juvenile idiopathic arth-
    Gadner H, et al. (2013) Therapy prolongation improves outcome in ritis: a European League Against Rheumatism/​American College
    multisystem Langerhans cell histiocytosis. Blood 121, 5006–​14. of Rheumatology/​ Paediatric Rheumatology International Trials
    Hackmann Y, et al. (2013) Syntaxin binding mechanism and disease-​ Organisation Collaborative Initiative. Ann Rheum Dis 75(3), 481–​9.
    causing mutations in Munc18-​2. Proc Natl Acad Sci USA 110, Rollins BJ (2015) Genomic alterations in Langerhans cell histiocytosis.
    E4482–​91. Hematol Oncol Clin North Am 29, 839–​51.
    Haroche J, et al. (2012) High prevalence of BRAF V600E mutations Sieni E, et al. (2015) Early diagnosis and monitoring of
    in Erdheim–​Chester disease but not in other non-​Langerhans cell neurodegenerative Langerhans cell histiocytosis. PLoS One 10(7),
    histiocytoses. Blood 120(13), 2700–​3. e0131635.
    Haroche J, et al. (2015) Reproducible and sustained efficacy of targeted Sieni E, et al. (2014) Familial hemophagocytic lymphohistiocytosis:
    therapy with vemurafenib in patients with BRAFV600E-​mutated when rare diseases shed light on immune system functioning. Front
    Erdheim–​Chester disease. J Clin Oncol 33, 411–​18. Immunol 5, 167.
    Haroche J, et al. (2017) Histiocytoses: emerging neoplasia behind in- Simko SJ, et al. (2014) Clofarabine salvage therapy in refractory multi-
    flammation. Lancet Oncol 18, e113–​25. focal histiocytic disorders, including Langerhans cell histiocytosis,
    Henter JI, et al. (2007) HLH-​2004: diagnostic and therapeutic guide- juvenile xanthogranuloma and Rosai-​Dorfman disease. Pediatr
    lines for hemophagocytic lymphohistiocytosis. Pediatr Blood Blood Cancer 61, 479–​8.
    Cancer 48, 124–​1. Tran TH, et al. Cutaneous histiocytoses. In: C. Griffiths, et al.
    Hutter C, Minkov M. (2016) Insight into the pathogenesis of (eds) Rooks Textbook of Dermatology (9th edn), Chapter 136.
    Langerhans cell histiocytosis: the development of targeted ther- Oxford: Wiley-​Blackwell.
    apies. Immunotargets Ther 5, 81–​91. Trottestam H, et al. (2011) Chemoimmunotherapy for hemophagocytic
    Jordan MA, et al. (2011) How I treat hemophagocytic lympho­ lymphohistiocytosis: long-​term results of the HLH-​94 treatment
    histiocytosis. Blood 118, 4041–​52. protocol. Blood 118(17), 4577–​84.
    Lehmberg K, et al. (2015) Malignancy-​associated haemophagocytic Veys PA, et al. (2015) Haematopoietic stem cell transplantation for re-
    lymphohistiocytosis in children and adolescents. Br J Haematol fractory Langerhans cell histiocytosis: outcome by intensity of con-
    170, 539–​49. ditioning. Br J Haematol 169(5), 711–​18.
    21
    Symptoms and Emergencies in Children
    with Brain and Spinal Tumours
    Michel Zerah, David Walker, and Shaarna Whitton

    Introduction outcome in survivors. It has also been associated with a reduced


    likelihood of achieving complete tumour resection, which is an im-
    Childhood central nervous system (CNS) tumours are the most portant prognostic indicator in some tumour types.
    common solid tumour in childhood, accounting for a quarter of all The reasons for prolonged TDIs are multi-​factorial. Firstly, the
    childhood cancers. Despite being the commonest cancer cause of rarity of brain tumours means health professionals do not routinely
    death, the time from symptom onset to diagnosis, known as the total consider this as a diagnosis when presented with non-​specific symp-
    diagnostic interval (TDI), is one of the longest for all childhood can- toms. The presenting symptoms of brain tumours include head-
    cers. The international median TDI for brain tumours is reported as ache, vomiting, and lethargy, which are commonly misdiagnosed
    8.5 weeks (range) and 5.6 months (range) for spinal tumours. Five-​ as gastroenteritis, migraine, or behavioural problems. Furthermore,
    year survival rates for brain tumours have improved to over 80% the majority of children who present with a brain tumour have a
    across Europe; however, two-​thirds of survivors are left with a life-​ completely normal neurological examination, contrary to the beliefs
    changing disability. of many health professionals. Reluctance to consider brain imaging
    As oncologists, we receive children who have been given a new in children also contributes to the delay; it is an alarming investiga-
    diagnosis of a brain or spinal cord tumour. The start of symptoms tion for the child and family, often requires general anaesthesia or
    indicates the start of brain/​spinal injury and accelerating diagnosis sedation, and is mistakenly considered costly. All too frequently, the
    is increasingly recognized as a crucial step in order to reduce long-​ diagnosis is suggested by the parents who have noticed an abnormal
    term brain injury for all. Childhood CNS tumours pose a particular behaviour or insidious onset of symptoms and have asked, for many
    diagnostic challenge for all clinicians, but especially for family and weeks, for access to imaging for their child.
    hospital clinicians who meet these children at the start of their A very specific situation is represented by the antenatal diagnosis
    journey to diagnosis; the signs and symptoms that precede diagnosis of a brain tumour. The diagnosis is made predominantly during the
    are non-​specific, fluctuate in severity, and can mimic other common third trimester due to observation of macrocrania, disorganization
    illnesses. The aim of highlighting this issue is to understand these of the brain, and hydramnios. In most of the cases, they are very ma-
    difficulties and raise awareness amongst colleagues to seek ways to lignant tumours, that lead to a termination of pregnancy or to a pre-
    accelerate diagnosis so as to minimize neurological injury. natal or neonatal death. Occasionally, prenatally diagnosed choroid
    plexus papillomas or craniopharyngiomas can be treated postnatally
    with good results.
    Delays in diagnosis In order to diagnose brain tumours effectively, we need to have
    knowledge of the risk of the disease in children and young people,
    Along this pathway of initial symptom onset to diagnosis, symptom and be aware of the signs and symptoms of brain tumours at all ages.
    progression linked to tumour growth inevitably occurs. In the early
    stages, due to the rarity of the disease, conventional approaches in
    primary care, using positive predictive values of red-​flag symptoms, Predisposing factors
    are recognized as insufficiently sensitive. In the later stages, the ac-
    cumulation of signs and symptoms becomes increasingly helpful in There are a number of predisposing factors which are associated
    localizing the problem within the nervous system, at which point with an increased risk of childhood CNS tumours. These include
    the neurological outcome is poorer. A prolonged TDI in childhood important familial genetic syndromes such as neurofibromatosis
    brain tumours is associated with an increased risk of life-​threatening types 1 and 2 (see Table 21.1), as well as a personal or family history
    neurological complications at presentation and a worse cognitive of a brain tumour, leukaemia, sarcoma or early-​onset breast cancer,
    170 CHAPTER 21 Brain and Spinal Tumours

    Table 21.1 Central nervous system tumours and their genetic and prior therapeutic CNS irradiation. Recognition of predisposing
    association factors would not only lower the threshold for investigation but also
    affect treatment options and prognosis. Table 21.1 outlines a com-
    Syndrome (gene) Associated CNS tumours
    prehensive list of the genetic associations. The more common pres-
    Ataxia-​telangiectasia (ATM) • Medulloblastoma entations are discussed below.
    Constitutional mismatch repair • Astrocytoma
    deficiency syndrome (MSH2, • Glioblastoma Neurofibromatosis type 1
    MSH6 MLH1, PMS2) • CNS embryonic tumour
    • Medulloblastoma Neurofibromatosis type 1 (NF1) is associated with visual pathway
    gliomas and other tumours of the CNS. Approximately 20% of NF1
    Familial adenomatous • Astrocytoma
    polyposis (APC) • Medulloblastoma cases are associated with low-​grade gliomas which are almost always
    • Ependymoma pilocytic astrocytomas. They can occur anywhere in the CNS, with
    • Pineoblastoma the commonest locations being the visual pathways, cerebellum, and
    Carney complex (PRKAR1A) • Schwannoma (psammomatous brainstem. They classically affect the visual pathways in early child-
    melanotic) hood, which justifies visual screening during early childhood.
    Cowden syndrome (PTEN) • Dysplastic gangliocytoma of cerebellum
    • Meningioma Neurofibromatosis type 2
    DICER1 syndrome (DICER1) • CNS embryonic tumour (specifically Neurofibromatosis type 2 (NF2) is characterized by the presence of
    medulloepithelioma) acoustic neuromas or auditory nerve sheath schwannomas, causing
    • Pineoblastoma hearing difficulties. Other clinical features include skin nodules or
    • Pituitary blastoma
    plaque-​like lesions, café au lait patches, mono-​or polyneuropathies,
    Fanconi anaemia (FANCD1/​ • CNS embryonic tumour and visual loss due to cataracts or post-​lenticular opacities. In add-
    BRCA2, FANC-​N, or PALB2) • Medulloblastoma
    ition, there is a high risk of intracranial meningioma affecting 18–​
    Hereditary retinoblastoma • Pineoblastoma 58% of patients, and astrocytomas and ependymomas in about 3% of
    (RB1)
    patients. When present in the adolescent age group, they may repre-
    Li–​Fraumeni syndrome (TP53) • Astrocytoma sent the more aggressive version of the disease called the ‘Wiseheart’
    • High-​grade glioma (diffuse astrocytoma,
    anaplastic astrocytoma, glioblastoma) type, which is progressive and has a reduced life expectancy.
    • Choroid plexus tumour
    • CNS embryonic tumour Tuberous sclerosis complex
    • Medulloblastoma
    Tuberous sclerosis complex (TSC) phenomena can affect any
    Neurofibromatosis type 1 • Pilocytic astrocytoma/​optic pathway/​ organ; however, they affect the brain and skin in up to 80%. Other
    (NF1) brainstem, neuraxial glioma
    • Diffuse astrocytoma organ phenomena manifest at different times in life. Thus, cardiac
    • High-​grade glioma rhabdomyomas occur in infancy, grow, and then involute within the
    • Malignant peripheral nerve first year. Subependymal giant-​cell astrocytomas (SEGAs) present
    sheath tumour
    • Neurofibroma symptomatically or on brain imaging as part of diagnostic screening
    during childhood, and grow during adolescence. Recently pub-
    Neurofibromatosis type 2 • Astrocytoma
    (NF2) • Ependymoma lished guidelines recommend surveillance imaging to diagnose the
    • Meningioma development of SEGAs and other TSC phenomena earlier during
    • Neurofibroma childhood. Worsening epilepsy and raised intracranial pressure
    • Schwannoma
    due to obstruction of the lateral/​third ventricles are common pre-
    • Vestibular schwannoma
    senting symptoms; occasionally, massive spontaneous haemorrhage
    Nevoid basal-​cell carcinoma • Astrocytoma
    can occur.
    syndrome (PTCH/​SUFU) • Craniopharyngioma
    • Medulloblastoma (desmoplastic subtype)
    • Meningioma
    • Oligodendroglioma Brain tumours
    Rhabdoid tumour • Atypical teratoid/​rhabdoid tumour
    predisposition syndrome • Malignant peripheral nerve sheath Symptomatology
    (SMARCB1) tumour
    Symptoms of CNS tumours can be non-​specific and can occur singu-
    Rubinstein–​Taybi syndrome • Medulloblastoma
    (CREBBP) larly or in combination. Symptoms frequently fluctuate in severity;
    Schwannomatosis (SMARCB1) • Schwannoma resolution and then recurrence does not exclude a CNS tumour. It is
    important to note that presentation depends on the age of the child,
    Tuberous sclerosis complex • Subependymal giant-​cell astrocytoma as well as tumour location and biology.
    (TSC)
    Wilne and colleagues provided data on the most common pre-
    Von Hippel–​Lindau (VHL) • Haemangioblastoma (intracranial, spinal)
    senting signs and symptoms. Ranking of symptoms based on lo-
    Source: data from de Kock L, et al. (2014); Dubuc AM, et al. (2010); Foulkes W, et al. cation is outlined in Figure 21.1 and highlights the variability of
    (2011); Hasselblatt M, et al. (2014); Hottinger AF, Khakoo Y (2007); Kimmelman A, Liang
    BC (2001); Kirschner L, et al. (2000); Louis D, et al. (2007); Ohgaki H, et al. (2010); Stefanaki presentation depending on location. For example, brainstem tu-
    K, et al. (2012); Tabori U, et al. (2015); Taylor M, et al. (2001); and Villani A, et al. (2012) mours present most commonly with abnormal gait and cranial
    Brain tumours 171

    Brainstem tumours
    Supratentorial tumours Abnormal gait and coordination difficulties 78%
    Unspecified symptoms of raised ICP* 47% Cranial nerve palsies (unspecified) 52%
    Seizures 38% Pyramidal signs (unspecified) 33%
    Papilloedema* 21% Headache* 23%
    Focal neurological signs 17% Squint 19%
    Headache* 11% Focal motor weakness 19%
    Hemiplegia 10% Facial palsy 15%
    Nausea and vomiting* 8% Papilloedema* 13%
    Macrocephaly* 6% Unspecified symptoms of raised ICP* 10%
    Abnormal eye movements 6%
    Behavioural change or school difficulties 5%

    Posterior fossa tumours


    Nausea and vomiting* 75% Central tumours
    Headache* 67% Headache* 49%
    Abnormal gait and coordination difficulties 60% Abnormal eye movements and squint 21%
    Papilloedema* 34% Nausea and vomiting* 19%
    Abnormal eye movements 20% Papilloedema* 18%
    Lethargy 13% Reduced visual acuity 16%
    Nausea without vomiting* 10% Unspecified symptoms and signs of raised ICP* 13%
    Unspecified symptoms and signs of raised ICP* 9% Diabetes insipidus 12%
    Weight loss 9% Abnormal gait and coordination difficulties 10%
    Focal motor weakness 9% Optic atrophy 9%
    Macrocephaly* 7% Behavioural change or school difficulties 9%
    Impaired consciousness 7% Altered level of consciousness 9%
    Vertigo or auditory symptoms 7% Reduced visual fields 8%
    Squint 6% Seizures 7%
    Stiff neck 6% Hemiplegia 7%
    Head tilt Focal motor deficit 7%
    Accidental head injury 5% Developmental delay 7%
    Short stature 7%
    Weight loss 5%
    Spinal-cord tumours Vertigo or auditory symptoms 5%
    Back pain 67% Visual or eye abnormalities (unspecified) 5%
    Abnormal gait or coordination diffuculties 42%
    Spinal deformity 39%
    Foral motor weakness 21%
    Sphincter disturbance 20%
    Decreased upper-limp movement 17%
    Developmental delay 8%
    Head tilt 7%
    Headache* 7%

    Figure 21.1 Presenting symptoms of central nervous system tumours ranked according to location, based on a systematic meta-​analysis.
    Reproduced with permission from Wilne S, et al. 'Presentation of childhood CNS tumours: a systematic review and meta-​analysis', Lancet Oncology, Volume 8, Issue 8, pp. 685–​
    95. Copyright © 2007 Elsevier Ltd. All rights reserved. DOI: https://​doi.org/​10.1016/​S1470-​2045(07)70207-​3

    nerve palsies, whereas posterior fossa tumours present with nausea signs and symptoms amongst paediatricians nationally. About 20%
    and/​or vomiting and headaches. of the UK public also became aware of the campaign.
    A UK-​based cohort study highlighted that the symptom profile Further analysis of the TDI data showed that the tumour sub-
    also changes over time from symptom onset to diagnosis. Visual types with the greatest proportion of prolonged TDI were mid-​line
    and motor symptoms increase the most, by 53% and 45% respect- supratentoral tumours including craniopharyngiomas, intracranial
    ively, from symptom onset to diagnosis. This is important to con- germ-​cell tumours, low-​grade gliomas, and optic pathway gliomas,
    sider when reassessing patient symptoms over time; looking at the with median TDIs of 15.1, 5.6, 11.9, and 10.4 weeks respectively.
    progression of visual, motor, or behaviour symptoms over multiple Moreover, the 12–​18 age group were found to have the longest delay,
    consultations should alert clinicians to a possible diagnosis of brain with a median of 12.1 weeks, compared to six weeks for the 0–​5 age
    tumour. group and eight weeks for the 5-​11 age group.
    It is also recognized that the non-​specificity of symptoms can re- The next phase of the campaign involves strategies targeting the
    sult in numerous consultations with different subspecialties in order adolescent age group, as well as those tumour subtypes with longest
    to find a cause. Subspecialty clinicians need to be aware of a brain delays; the website and media campaign has been designed to appeal
    tumour as a potential diagnosis in order to accelerate diagnosis, to the adolescent age group, with symbols to represent the symp-
    thereby limiting brain injury (Figure 21.2). toms, a website designed for mobile devices, and the use of musical
    content to attract attention. In the UK, the campaign was relaunched
    ‘HeadSmart: Early Diagnosis of Brain Tumour’ campaign by the Royal College of General Practitioners in conjunction with
    In the UK, a national public and professional awareness campaign the Royal College of Paediatrics and Child Health, to emphasize the
    named ‘HeadSmart: Early Diagnosis of Brain Tumour’ was launched priority across the primary–​secondary care interface. The overall
    to amplify the national evidence-​based clinical professional guid- success of this programme, using public-​awareness techniques, is
    ance in order to promote earlier diagnosis (www.headsmart.org.uk). worthy of consideration.
    As a quality-​improvement project, the driver for change was chosen
    as the TDI, which was collected nationally across 18 UK Children’s Emergency presentations
    CNS tumour treatment centres. Following the launch of the guide-
    Sudden coma
    line and two years after the subsequent awareness campaign, the
    median TDI had reduced from 14 weeks (mean 35.4 weeks) to 6.7 Sudden coma due to brain tumour is diagnosed in the emergency
    weeks (mean 22.3 weeks). The campaign did enhance awareness of room, necessitating admission to an intensive-​care unit (ICU),
    PAEDIATRIC SUBSPECIALISTS: ARE YOU HEADSMART?

    A young child with


    hydrocephalus caused by a brain
    tumour will have an increasing
    head circumference and
    developmental delay or
    regression
    Supratentorial tumours
    can cause change in
    personality, mood or A supratentorial
    disinhibition. They can Psychiatry cortical tumour will
    Community Neurology present with focal
    also cause symptoms of Seizures
    anorexia. A brain tumour neurological signs
    Anorexia Developmental delay Motor weakness such as weakness.
    needs to be considered
    Behavioural change Developmental regression CN palsies
    as part of the differential
    Depression Increasing head circumference Ataxia/cerebellar
    diagnosis.
    Psychosis Focal neurological deficits

    Ophthalmology Endocrinology

    Papilloedema Growth problem


    Central tumours
    Decreased visual acuity Hypo-pit/pituitary Central tumours such as
    such as an optic
    Nystagmus/Parinauds dysfunction a craniopharyngioma are
    pathway glioma are
    Diplopia Diabetes insipidus slow growing and will
    slow growing and
    Squint Precocious or delayed present with abnormal
    will present with
    Visual fileld defect puberty growth or precoious or
    progressive visual
    Blindness Menstrual irregularities delayed puberty. These
    symptoms that may
    Ptosis Galactorrhoea children may also have
    present to an
    Proptosis Gynaecomastia visual symptoms.
    ophthalmologist.
    Ocular palsies Cushing’s
    Ophthalmoplegia Obesity/weight gain

    Head tilt or
    Ear, nose & throat Gastroenterology
    torticollis can be Respiratory
    caused by a Dizziness Recurrent respiratory
    Vertigo Nuusea & vomiting
    posterior fossa infections can occur
    Torticollis Abdominal pain Recurrent chest infections
    tumour. These secondary to
    Reflux Apnoeas
    symptoms may Head tilt aspiration caused by a
    present to Failure to thrive
    Hearing loss bulbar palsy. This MRI
    ENT specialists as Dysphagia
    Tinnitus shows a brainstem
    head tilt and tumour which causes
    totticollis have cranial nerve palsies.
    other common ENT A child with hydrocephalus caused
    causes. by a brain tumour will have
    persistent vomiting. In infants
    where the sutures are not yet fused
    there will be no other signs of
    hydrocephalus aside from
    macrocephaly.

    Figure 21.2 Subspecialty involvement in central nervous system tumour diagnosis.


    Spinal-cord tumours 173

    intubation, and emergency computed tomography (CT) (mag- Head ache Voice change (1%)
    netic resonance imaging (MRI) is often difficult to achieve in (7%)
    Head tilt (7%)
    such emergencies). The sudden deterioration is frequently due to Dysphagia (2%)
    intra-​
    tumoural haemorrhage in hemispheric tumours, or acute Development delay
    hydrocephalus in pineal or posterior fossa tumours. Mannitol or (8%)
    hyperosmotic saline must be given, and emergency surgery is man- Quadriplegia (4%)
    Vomiting (4%)
    datory to debulk the tumour, control the bleeding, and/​or treat the
    hydrocephalus. Upper-limp weakness
    Back pain (17%)
    Hydrocephalus (67%) Sensory abnormality
    Spinal deformity (7%)
    If recognized and diagnosed, hydrocephalus is not life-​threatening, (38%)
    but it does need urgent treatment. If a pineal tumour is the cause, Sphincter disturbance
    then endoscopic third ventriculostomy (ETV) with sampling of Paraplegia (2%)
    (20%)
    cerebrospinal fluid (CSF) markers and, if it is safely possible, tumour Focal motor weakness
    biopsy, is recommended. In posterior fossa tumours, the approach (21%)
    is more controversial as surgical approaches include four options:
    • direct approach of the tumour,
    • external ventricular drainage, Abnormal Gait or Co-ordination (42%)
    • ventriculo-​peritoneal shunt, or Figure 21.3 Presenting symptoms and signs of spinal tumours in
    • ETV. childhood and young people.
    Reproduced with permission from Wilne S, Walker D 'Spine and spinal cord tumours
    ETV remains the best option as it allows the surgeon to operate on in children: a diagnostic and therapeutic challenge to healthcare systems’, British
    the tumour when the patient is in a better clinical condition, and also Medical Journal, Volume 95, Issue 2, pp. 47–​54. Copyright © BMJ Publishing Group
    Ltd. DOI: http://​dx.doi.org/​10.1136/​adc.2008.143214
    allows the surgeon time to discuss alternative therapeutic strategies,
    especially in very young children, children with metastatic disease,
    and those with the possibility of desmoplastic histology.
    spinal nerve roots through to tumours of the spinal cord itself or
    Vision or visual-​field loss tumours of spinal coverings, vertebrae, or tissues of the paraspinal
    Acute visual impairment can be the first noticed symptom. The most regions (Figure 21.3).
    frequent situation is a cystic craniopharyngioma with neglected
    endocrinological symptoms or ignored diabetes insipidus. There is Risks of delays in diagnosis
    a high risk of irreversible blindness associated with immediate tu- Initial signs are often relatively subtle and their initial symptoms
    mour surgery, and most neurosurgical teams favour stereotactic or may easily be disregarded as trivial complaints. Delays are common
    endoscopic tapping of cysts with the positioning of an intracystic due to multiple consultations where the significance of symptoms of
    reservoir for further tapping or local treatment. Similarly, optic spinal origin are not appreciated at first. A review of the diagnostic
    pathway gliomas, whether associated with NF1 or not, can pre- and therapeutic challenges that these initial symptoms pose is pro-
    sent with visual impairment. Sustained raised intracranial pres- vided by Wilne and Walker (see ‘Further reading’).
    sure causing papilloedema and subsequent optic atrophy can cause A study of 134 cases from Italy provides information on delay
    lifelong visual impairment. Together, about 6% of children are regis- between clinical presentation and diagnosis, age incidence, clin-
    tered blind within two years of diagnosis, and historical follow-​up ical presentation, and outcome. This large, single-​centre ex-
    studies identify visual impairment in up to 20% of survivors. perience identified that delay in diagnosis was frequent, with a
    median symptom interval before diagnosis of 5.3 months (range
    A diagnostic emergency one day to two years). There was a bimodal age distribution, with
    In some situations, the emergency is not to treat but to have access to a peak in the first 12–​18 months followed by a second peak in
    a diagnosis to initiate non-​surgical approaches to treatment. Many adolescence. Symptoms reported at presentation were, in order
    such situations have been described here: CSF and CSF markers or of frequency, pain, palpable mass, paraparesis, torticollis, spinal
    endoscopic biopsy for pineal tumours, for example. More and more deformity, sphincteric problems, dyspnoea, weight loss, head-
    often, as stereotactic biopsy is the first diagnostic step for diffuse ache, hemiparesis, and vomiting. From this analysis, the most
    intrinsic pontine gliomas (DIPGs), it is also a requirement in new progressive problems at the time of diagnosis were pain, para-
    therapeutic trial protocols. paresis, spinal deformity, sphincter problems, and hemiparesis.
    The tumours predominantly involved the upper spinal cord
    in cervical and thoracic regions, but all regions were repre-
    Spinal-​cord tumours sented. Intramedullary tumours accounted for 34.3%, intradural
    extramedullary tumours, 18.7%, and extradural tumours, 39.5%,
    Spinal-​cord tumours in childhood account for about 5–​10% of all of which 60% involved the vertebrae, 39% did not, and 7.5% were
    CNS tumours. There is a wide range of tumour types which can classified as paravertebral. The range of histologies extended to
    present in this way, ranging from benign slow-​growing tumours of 29 subtypes of benign and malignant tumours.
    174 CHAPTER 21 Brain and Spinal Tumours

    A second population-​ b ased study of children aged less Box 21.1 Classification of spinal-​cord tumours
    than three years is reported, where gait disorder affected over
    70%, followed by developmental delay, abnormal eye move- Grade one
    ments, seizures, head tilt, vomiting, and failure to thrive. The • Is neurologically normal.
    range of tumour types was much more restricted, involving • The mild focal deficit does not significantly affect the function of the
    low-​g rade astrocytoma (70%), CNS embryonic tumour (15%), involved limb.
    and rhabdoid tumour (15%). Outcomes were significantly dif- • There is mild spasticity or reflex abnormality.
    ferent between benign and malignant tumours, the malignant • Patient’s gait is normal.
    tumours doing significantly worse. Grade two
    A third institutional report focused upon presentation, imaging, • Sensorimotor deficit affects the function of the involved limb.
    and pathological characteristics in addition to clinical features and • Severe pain or dysaesthetic syndrome present which impairs the
    outcomes. Mean symptom delay was 7.8 months (0.25–​60 months). patient’s quality of life.
    A notable presenting feature was early handedness in under three • Patient has mild to moderate gait difficulty.
    year olds. Radiographically, homogenous gadolinium enhancement • Patient can still function and ambulate independently.
    on MRI correlated with low-​grade histology. The degree of resection Grade three
    was strongly predictive of survival. • The neurological deficit is more severe.
    From these reports, we can conclude that tumours can arise any- • Patient requires a cane or brace for ambulation, or there is significant
    where along the spinal cord, and that delays in diagnosis are frequent bilateral upper-​extremity impairment.
    and longer for low-​grade tumours. The distribution between extradural • Patient may or may not be able to function independently.
    and intradural tumours is similar. The range of histological subtypes Grade four
    is extremely wide, especially when the whole childhood age range is • The neurological deficit is severe.
    studied. Children less than three years of age present with a narrower • Patient requires a wheelchair or a cane/​brace, and there is bilateral
    range of tumour types. Surgical management is important for obtaining upper-​extremity impairment.
    histology in this extremely diverse range of tumour types but, more im- • Patient is usually not independent.
    portantly, complete resection, where it is achieved, contributes to sig- Source: data from McCormick, P. et al. ‘Intramedullary ependymoma of the spinal
    nificant functional and survival advantage. cord’. Journal of Neurosurgery. Volume 72, Issue 4. pp. 523–​32. Copyright © 1990,
    American Association of Neurological Surgeons

    Clinical description
    The clinical features of spinal-​
    cord compression include three
    groups of symptoms and signs: Diagnosis
    • back pain and rigidity, The diagnosis of a spinal-​cord tumour should be quickly confirmed
    • signs of segmental spinal involvement, and by MRI since the risk of sudden aggravation, with complete para-
    plegia due to compromised circulation in the anterior cerebral ar-
    • signs resulting from interruption of long tracts within the
    tery and its branches, although very rare, may occur at any time.
    spinal cord.
    One should be exceedingly suspicious of recurrent back pain and, in
    The importance of back pain and, especially, of spinal rigidity as an early some patients, of recurrent abdominal pain. Atypical presentations
    symptom of tumour of the cord has long been known. Pain may have with paroxysmal attacks of arm pain or isolated abdominal discom-
    a radicular distribution but is more often diffuse and is usually pre- fort simulating an irritable bowel syndrome may delay recognition
    dominantly nocturnal. Rigidity is frequently associated with scoliosis. of a tumour.
    It must be re-​emphasized that a painful or rigid scoliosis is a completely Plain X-​rays of the spine may be diagnostic by showing bone
    abnormal situation and must lead in any case to MRI. Segmental weak- destruction, erosion of the spinal pedicles or of the posterior ver-
    ness, hyporeflexia, sensory disturbances, and amyotrophy can result tebral bodies, and widening of both the anteroposterior and trans-
    from involvement of the central grey matter or nerve roots. Segmental verse diameters of the spinal canal. MRI is the imaging technique
    myoclonus is a rare presentation. Spasticity evolving into paraplegia, of choice. CT scanning involves relatively high doses of radiation
    sphincter disturbances, and sensory deficits results from long-​tract in- for young children; it does show bony structures but is less precise
    volvement. Disturbances of walking, with claudication and stiffness, at showing neural tissue. MRI clearly shows both intra-​and extra-​
    may be present long before other symptoms. The presence of a defin- medullary lesions and associated cystic formation, thus permitting
    able sensory level, which may be difficult to demonstrate in young chil- separation of intramedullary tumours with associated cysts from
    dren, indicates the upper level of compression. Bilateral pyramidal tract syringomyelia. However, distinguishing a tumour from an area of
    signs are present with Babinski and Rossolimo signs. Initial bladder increased signal due to oedema, haemorrhage, or inflammation
    symptoms are increased urgency followed by retention or incontinence. can be difficult. Gadolinium enhancement, when homogenous,
    Sphincter disturbances, especially urinary retention, indicate an urgent selects low-​grade lesions, and diffusion-​weighted MRI may be
    need for decompression. useful in some cases. CSF protein, where obtained, is usually in-
    The main clinical problem with these tumours is functional dis- creased below the level of the block. However, there is a serious
    turbance (gait, sensory, sphincters, pain, spinal deformity, and pul- risk of damage to the cord by lumbar puncture at the site of com-
    monary restriction). McCormick’s widely used classification divides pression, possibly due to slight downward displacement of the cord
    these spinal-​cord tumour patients into four groups (see Box 21.1). (Figure 21.4).
    Intramedullary spinal-cord tumours 175

    and haemangio‑blastomas. They are frequently associated with


    cystic cavitation of the central cord, which may be difficult to
    differentiate from purely cystic intramedullary lesions. Their
    presentations are often subacute or diagnosed in patients under
    investigation for spinal surgery or associated with inherited pre-
    disposition states (NF2, NF1, and von Hippel–​L indau). They
    can produce symmetrical weakness of the limbs and are often
    unassociated with pain. They often do not produce a clear-​c ut
    sensory level and some of them may be surprisingly well toler-
    ated for long periods, despite considerable anatomical extension.
    Some spinal astrocytomas may present with the clinical mani-
    festations of raised intracranial pressure (ICP) without local signs.
    Cinalli and colleagues reported eight cases in which hydrocephalus
    was the initial manifestation and reviewed 38 cases from the litera-
    ture. Intracranial hypertension may result from metastatic disease,
    meningeal fibrosis due to haemorrhage from the tumour, or de-
    creased resorption of CSF because of its high protein content ‘clog-
    ging’ the arachnoid granulations.

    Diagnosis and management


    The diagnosis of intramedullary tumours is dependent upon MRI
    clearly delineating the extent, the gross tumour morphology, and
    neuronal involvement. Astrocytomas may extend over consider-
    able lengths of cord. Yet their removal, specifically in the paediatric
    population, has been made possible in most (more than 90% of
    Figure 21.4 Paraspinal neuroblastoma entering the spinal canal the ependymomas and 80% of the astrocytomas) cases by modern
    through the intervertebral foramen, displacing and compressing the
    lower spinal cord roots. neurosurgical techniques. Obtaining a tissue biopsy is the prime
    Reproduced with permission from Kennedy C, et al. ‘Tumours of the central nervous reason for operating and surgical resectability is discovered at op-
    system, other space-​occupying lesions and pseudotumour cerebri’. In: Aicardi’s eration. This, however, is not always possible, and so selected cases
    Diseases of the Nervous System in Childhood, 4th Edition, edited by A. Arzimanoglou.
    London, UK: Mac Keith Press. Copyright © 2018, Mac Keith Press, 9781909962804.
    may be offered a trial of chemotherapy to see if tumour shrinkage
    and symptom relief can be achieved after histological diagnosis.
    Spinal ependymomas that complicate NF2 are often multiple and
    Emergency presentation are progressive during adolescence and adulthood. When not asso-
    The emergency context of clinical presentation is worthy of specific ciated with NF2, they are uncommon and not infrequently present
    comment. If a spinal tumour is diagnosed, then the rate of progres- in conus medullaris where they often demonstrate myxo-​papillary
    sion of neurological signs must be kept under regular review. The histology and are reported to metastasize throughout the CNS. The
    nature of many children’s tumours is that they can progress rap- role of chemotherapy versus radiotherapy has not been studied in
    idly. Extramedullary tumours in particular (e.g. neuroblastoma, low-​grade astrocytomas. Primary chemotherapy, particularly in
    lymphoma, Ewing’s tumour) can grow across the spinal canal in a younger children, with vincristine and carboplatin (or other drug
    few days to a week or two. Clinical priority for referral, diagnostic combinations), is being investigated in clinical trials and in selected
    biopsy, and possible medical or surgical decompression needs to be spinal-​cord cases. Non-​progression rates of 70% and response rates
    classified as urgent, imaging and referral actions being completed in of 50% can be expected with all drug regimens. As in all CNS tu-
    emergency. Where spinal-​cord compression is identified, the use of mours, treatment selection by the neuro-​oncology multidisciplinary
    immediate steroid medication such as dexamethasone in high dose team is the recommended approach.
    (up to 10 mg/​m2/​day), in preparation for surgical decompression, Other rare spinal tumour types, such as gangliogliomas and
    can be critical to saving neurological function. In parallel, achieving a haemangioblastomas, may be a part of the von Hippel–​Lindau com-
    histological diagnosis remains a strong priority. Lymphomas can melt plex. These tumours are often cystic and commonly extend over con-
    away under the influence of steroids and the possibility of a tissue siderable lengths in the cervico-​dorsal part of the cord. Some of them
    diagnosis can be lost as a consequence. Close collaboration with the also involve the lower brainstem. Surgery is the only option in symp-
    neuro-​oncology team can help minimize these risks. Individualized tomatic haemangioblastomas. It is very complex and dangerous sur-
    management is critical as the use of steroids in a lymphoma can also gery which can be helped by pre-​surgical embolization (Figure 21.5).
    precipitate tumour lysis syndrome with the risk of acute renal failure. Treatment of intramedullary tumours raises special problems.
    If removal proves impossible, decompression of surrounding bony
    structures, chemotherapy, or irradiation may afford fairly long
    Intramedullary spinal-​cord tumours symptomatic relief in 30–​50% of cases. Astrocytomas, even when
    incompletely resected and treated with adjuvant chemotherapy, can
    Intramedullary tumours are mainly represented by remain static for sustained periods. One study found that the overall
    astrocytomas, ependymomas and gangliogliomas/​cavernomas, long-​term survival was only of the order of 50%, but such figures
    176 CHAPTER 21 Brain and Spinal Tumours

    (a) (b) (a) (b)

    Figure 21.5 Intramedullary astrocytoma of the spinal cord: (a) T1-​


    weighted MRI showing enlarged cervical cord with multiple low-​signal Figure 21.6 Spinal schwannoma (neurinoma) of cauda equine: (a) T1-​
    areas; and (b) T2-​weighted MRI showing high signal in the same areas. weighted MRI; (b) T2-​weighted high-​signal image.
    Reproduced with permission from Kennedy C, et al. ‘Tumours of the central nervous Reproduced with permission from Kennedy C, et al. ‘Tumours of the central nervous
    system, other space-​occupying lesions and pseudotumour cerebri’. In: Aicardi’s system, other space-​occupying lesions and pseudotumour cerebri’. In: Aicardi’s
    Diseases of the Nervous System in Childhood, 4th Edition, edited by A. Arzimanoglou. Diseases of the Nervous System in Childhood, 4th Edition, edited by A. Arzimanoglou.
    London, UK: Mac Keith Press. Copyright © 2018, Mac Keith Press, 9781909962804. London, UK: Mac Keith Press. Copyright © 2018, Mac Keith Press, 9781909962804.

    require interpretation with specific histologies. Patients with malig- patients and in tumours of cervical or junctional regions). On the
    nant tumours such as glioblastomas do poorly. other hand, if the neurological injury is long-​standing (more than
    Intradural tumours account for approximately 25% of cases, and 24–​48 hours), decompression may not be effective in saving spinal-​
    meningiomas are the most common histological type, followed by cord function, although remarkable late recoveries can occur in chil-
    schwannomas. Metastases from intracranial tumours, especially dren, when not expected in adults.
    ependymomas and medulloblastomas, are usually multiple. Selection of the tissue site for biopsy is critical, especially if it can
    offer decompression to prevent further progression of spinal injury
    associated with delayed tumour response to chemotherapy. The sig-
    Extramedullary tumours nificant differences between the treatment programmes justifies
    the need for tissue for histology and biology, as well as for research
    Approximately two thirds of extramedullary tumours are extradural purposes. In acute presentations, where spinal-​nerve injury is pro-
    in origin, spreading from nearby bone or through intervertebral for- gressive, chemotherapy may be used as empirical treatment pending
    amina. About a quarter are intradural. Neuroblastoma is the most histological classification, if steroids alone are not holding neuro-
    frequent cause but other tumours are of osseous or soft-​tissue origin logical deterioration or a full histological/​molecular diagnosis will
    such as rhabdomyosarcoma, Ewing sarcoma, aneurysmal bone be delayed.
    cysts, metastases, lymphomas, or bony deposits of Langerhans cell Spinal-​cord tumours can arise throughout childhood, are fre-
    histiocytosis (LCH). quently delayed in diagnosis, and, when presenting acutely, are
    Extramedullary tumours tend to manifest initially with unilat- associated with risk of serious spinal injury justifying timely inter-
    eral pain in segmental distribution, often in association with par- vention. Histological diagnosis is crucial for treatment planning and
    aesthesiae and weakness. Brown–​Séquard syndrome is observed prognostication. Successful surgical intervention is central to a suc-
    in a small proportion of children with extramedullary tumours. cessful outcome in most cases.
    More commonly, there is only predominant weakness, spasticity, There are two completely different situations for emergency pres-
    and deep sensory loss on the side of the tumour, with more marked entations of spinal-​cord tumours:
    contralateral loss of pain and temperature sensation. Schwannomas
    1. Intramedullary tumours
    (neurinomas) are more common than meningiomas (Figure 21.6).
    True emergency situations are exceptional, requiring mostly
    Acute spinal-​cord compression speedy surgery in the case of intractable pain or rapidly progres-
    There is some controversy regarding the role of surgical decom- sive symptoms.
    pression in tumour-​related presentations. On the one hand, acute 2. Extramedullary tumours
    neurological deterioration is reversible with early decompressive On the other hand, emergency presentation is frequent in
    surgery; it offers the opportunity to obtain tissue for histological extramedullary tumours. The decision making depends on the
    diagnosis without attempting to access retroperitoneal tissues in the age, the topography, the supposed diagnosis, the clinical condi-
    thorax or abdomen. The risk of subsequent spinal deformity linked tion, the evolution, and the MR images. However, acute surgery
    to delamination is thought to be limited by the use of osteoplastic is very often mandatory. Laminectomy, which was once the only
    laminotomy and replacement of lamina post surgery, subsequent option, has been progressively in competition with different or
    surgical instability affecting only a minority (mainly the youngest combined approaches and the use of instrumentation, even in
    Conclusion 177

    emergency surgery. These types of approaches need well-​trained Di Rocco C, et al. (1994) Spinal epidural meningiomas in childhood: a
    neurosurgical and orthopaedic paediatric teams with significant case report. J Neurosurg Sci 38, 251–​4.
    experience of these complex spinal-​surgery techniques. Two Dixon-​Woods M, et al. (2001) Parents’ accounts of obtaining a diag-
    specific situations where this surgery must be discussed are: nosis of childhood cancer. Lancet 357, 670e4.
    Dubuc AM, et al. (2010) The genetics of pediatric brain tumors. Curr
    • Neuroblastoma: There is no surgical indication in case of
    Neurol Neurosci Rep 10(3), 215–​23.
    antenatal or neonatal neuroblastoma. The possibility of re-
    Epstein F, Wisoff J (1987) Intra-​axial tumors of the cervicomedullary
    covery is nil and the effect on the spine growing is cata- junction. J Neurosurg 67, 483–​7.
    strophic. Surgery is not indicated as a first option except to Flores LE, et al. (1986) Delay in the diagnosis of paediatric brain tu-
    do a biopsy, if impossible by another approach (exceptional), mors. Am J Dis Child 140, 684e6.
    or in case of clinical impairment after one or two courses of Foulkes W, et al. (2011) Extending the phenotypes associated with
    chemotherapy. DICER1 mutations. Hum Mutat 32(12), 1381–​4.
    • Ewing sarcoma: Spinal Ewing sarcomas are very often diag- Gelabert M, et al. (1990) Hydrocephalus and intraspinal tumor in
    nosed in emergency, in the face of very severe clinical deteri- childhood. Child Nerv Syst 6, 110–​12.
    oration. The role of the surgery is to do a biopsy and to protect Hasselblatt M, et al. (2014) SMARCA4-​mutated atypical teratoid/​
    the spinal cord by a reasonable decompression. Sparing of the rhabdoid tumors are associated with inherited germline alterations
    spinal capital and the planning process for the next surgery and poor prognosis. Acta Neuropathol 128(3), 453–​6.
    must always be taken into account, and a frozen section can HeadSmart website: www.headsmart.org.uk
    Hottinger AF, Khakoo Y (2007) Update on the management of familial
    be helpful to challenge the diagnosis and stop the surgery.
    central nervous system tumor syndromes. Curr Neurol Neurosci
    Rep 7(3), 200–​7.
    Houten J, Weiner H (2000) Pediatric intramedullary spinal cord tu-
    Conclusion mors: special considerations. J Neuro-​oncol 47, 225–​30.
    Jallo GI, et al. (2004) Spinal cord gangliogliomas: a review of 56 pa-
    Clinical presentations of brain and spinal tumours include slow, grad- tients. J Neuro-​Oncol 68, 71–​7.
    ually progressive symptomatology which can mimic a wide variety of Kimmelman A, Liang BC (2001) Familial neurogenic tumor syn-
    clinical conditions, leading to avoidable delays. Acute presentations, dromes. Hematol Oncol Clin North Am 15(6), 1073–​84.
    on the other hand, can constitute emergencies threatening life due Kirschner L, et al. (2000) Mutations of the gene encoding the protein
    to raised ICP or sudden onset of paralysis due to spinal cord com- kinase A type I-​α regulatory subunit in patients with the Carney
    pression, the risk of acute intra-tumoural haemorrhage can compli- complex. Nat Genet 26(1), 89–​92.
    cate avoidable delays in diagnosis. Any neurological symptoms of the Louis D, et al. (2007) The 2007 WHO classification of tumours of the
    central nervous system. Acta Neuropathol 114(5), 97–​109.
    brain can constitute the beginning of brain injury, which is likely to
    McCormick P, et al. (1990) Intramedullary ependymoma of the spinal
    be lifelong. Accelerating diagnosis is therefore emerging as a top pri-
    cord. J Neurosurg 72, 523–​32.
    ority for reducing the late neurological consequences for survivors.
    Mora J, et al. (2007) Successful treatment of childhood intramedullary
    Emergencies associated with brain and spinal tumours need to be spinal cord astrocytomas with irinotecan and cisplatin. Neuro
    recognized as such by health systems to offer the optimal response. Oncol 9(1), 39–​46.
    Nagib M, O’Fallon M (1997) Myxopapillary ependymoma of the
    Conus medullaris and filum terminale in the pediatric age group.
    FURTHER READING Pediatr Neurosurg 26, 2–​7.
    Albanese V, Platania N (2002) Spinal intradural extramedullary tu- Ohgaki H, et al. (2010) Nervous system tumors associated with fa-
    mors. Personal experience. J Neurosurg Sci 46, 18–​24. milial tumor syndromes. Curr Opin Neurol 23(6), 583–​91.
    Allen J, et al. (1998) Treatment of high-​grade spinal cord astrocytoma Orbach D, et al. (2013) Neonatal cancer. Lancet Oncol 14, e609–​20.
    of childhood with ‘8-​in-​1’ chemotherapy and radiotherapy: a pilot Pascual-​Castraviejo I (ed.) (1990) Spinal Tumors in Children and
    study of CCG-​945. J Neurosurg 88, 215–​20. Adolescents. New York: Raven Press.
    Celli P, et al. (2005) Spinal extradural schwannorna. J Neurosurg 2, Phan T, et al. (2000) Recurrent lumbar ependymoma presenting as
    447–​56. headache and communicating hydrocephalus. Mayo Clin Proc
    Cinalli G, et al. (1995) Hydrocephalus associated with intramedullary 75, 850–​2.
    low-​grade glioma—​illustrative cases and review of the literature. J Pui M, et al. (2005) Diffusion-​weighted magnetic resonance imaging
    Neurosurg 83, 480–​5. of spinal infection and malignancy. J Neuroimaging 15, 164–​70.
    Costello F, et al. (2002) Papilledema as the presenting manifestation of Renault F, et al. (1995) Segmental myoclonus in a child with spinal
    spinal schwannoma. J Neuro-​ophthalmol 22, 199–​203. cord tumour. Dev Med Child Neurol 37, 354–​61.
    Crawford J, et al. (2009) Primary spinal tumors of childhood: effects Rifkinson-​Mann S, et al. (1990) The association of hydrocephalus
    of clinical presentation, radiographic features, and pathology on with intramedullary spinal cord tumors: a series of 25 patients.
    survival. J Neuro-​oncology 95(2), 259–​69. Neurosurgery 27, 749–​54.
    Daley M, et al. (2003) Primary epidural Burkitt lymphoma in a child: case Robertson P, et al. (1994) Cervicomedullary tumors in children: a dis-
    presentation and literature review. Pediatr Hematol Oncol 20, 333–​8. tinct subset of brainstem gliomas. Neurology 44, 1798–​803.
    de Kock L, et al. (2014) Germ-​line and somatic DICER1 mutations in Roig M, et al. (1988) Congenital spinal cord haemangioblastoma: an-
    pineoblastoma. Acta Neuropathol 128(4), 583–​95. other cause of spinal cord section syndrome in the newborn. J
    de Kock L, et al. (2014) Pituitary blastoma: a pathognomonic fea- Neurol Neurosurg Psychiatry 51, 1091–​103.
    ture of germ-​line CICER1 mutations. Acta Neuropathol 128(1), Royal College of Paediatric and Child Health (RCPCH) (2008)
    111–​22. The Brain Pathways Guideline: A Guideline to Assist Healthcare
    178 CHAPTER 21 Brain and Spinal Tumours

    Professionals in the Assessment of Children Who May Have a Brain Walker D, et al. (2013) A multi-​disciplinary consensus statement con-
    Tumour. London: RCPCH. cerning surgical approaches to low-​grade, high-​grade astrocytomas
    Spacca B, et al. (2015) Spinal tumors in children: long term retro- and diffuse intrinsic pontine gliomas in childhood (CPN Paris
    spective evaluation of a series of 134 cases treated in a single unit of 2011) using the Delphi method. Neuro Oncol 15(4), 462–​8.
    pediatric neurosurgery. Spinal J 15, 1949–​55. Williams A, et al. (1987) Differentiation of intramedullary neoplasms
    Stefanaki K, et al. (2012) Tumors of central and peripheral nervous and cysts by MR. Am J Roentgenol 149, 159–​64.
    system associated with inherited genetic syndromes. Pediatr Wilne S, et al. (2012) Progression from first symptom to diagnosis in
    Neurosurg 48, 271–​85. childhood brain tumours. Eur J Pediatr 171(1), 87–​93.
    Tabori U, et al. (2015) Cancer predisposition in children with brain Wilne S, et al. (2007) Presentation of childhood CNS tumours: a sys-
    tumors. In: K Scheinemann, E Bouffet E (eds) Pediatric Neuro-​ tematic review and meta-​analysis. Lancet Oncol 8(8), 685–​95.
    Oncology, pp. 69–​89. New York: Springer. Wilne S, et al. (2010) The diagnosis of brain tumors in children: a
    Taylor M, et al. (2001) Medulloblastoma in a child with Rubenstein–​ guideline to assist healthcare professionals in the assessment of
    Taybi syndrome: case report and review of the literature. Pediatr children who may have a brain tumour. Arch Dis Child 95, 534e9.
    Neurosurg 35(5), 235–​8. Wilne S, Walker D (2010) Spine and spinal cord tumours in children: a
    Uesaka T, et al. (2003) Intradural, extramedullary spinal Ewing’s sar- diagnostic and therapeutic challenge to healthcare systems. Arch
    coma in childhood. J Clin Neurosci 10, 112–​25. Dis Child Educ Pract Ed 95, 47–​54.
    Villani A, et al. (2012) Syndromes predisposing to pediatric central Zelcer S, et al. (2011) Spinal cord tumors in children under the age
    nervous system tumors: lessons learned and new promises. Curr of 3 years: a retrospective Canadian review. Child Nerv Syst 27,
    Neurol Neurosci Rep 12(2), 153–​64. 1089–​94.
    22
    Glial Central Nervous System Tumours
    of Childhood
    Dannis van Vuurden, Darren Hargrave, Dominik Sturm, and David T.W. Jones

    Low-​grade gliomas myxoid background, is termed pilomyxoid astrocytoma. By far the


    most frequent (more than 70%) genetic MAPK pathway alteration is
    a tandem duplication involving the BRAF gene, resulting in an onco-
    Classification and pathology genic KIAA1549-​BRAF fusion.
    Low-​grade gliomas (LGGs) are the most common primary central Subependymal giant-​ cell astrocytomas (WHO grade I) typ-
    nervous system (CNS) tumours in children and adolescents (more ically present as solid, partially calcified masses in the wall of the
    than 30%) and consist of a variety of histopathological entities clas- lateral ventricles adjacent to the foramen of Monro. They are com-
    sified as grade I or grade II (relatively benign, slow-​growing lesions) posed of large ganglionic astroyctes arranged in sweeping fascicles,
    by the World Health Organization (WHO) classification of tumours sheets, and nests. Up to 15% of patients with tuberous sclerosis syn-
    of the CNS. The largest proportion of these tumours presents as drome (TSC), an autosomal dominant genetic disease (caused by
    WHO grade I pilocytic astrocytoma (PA), accounting for approxi- inactivating germline mutations in TSC1/​2) predisposing to benign
    mately 20% of brain tumours in patients under the age of 20 years, tumour growth, develop SEGAs during childhood or adolescence. It
    with an average age-​adjusted annual incidence rate of 0.84 cases is uncertain whether SEGAs occur outside a TSC setting.
    per 100,000. Other frequent histopathological entities classified PXAs (WHO grade II) account for about 0.3 cases per 100,000
    as paediatric LGG include subependymal giant-​cell astrocytoma population, typically developing in children and adolescents. They
    (SEGA), pleomorphic xanthoastrocytoma (PXA), paediatric-​ arise almost exclusively in supratentorial regions, often involving
    type diffuse astrocytoma, and mixed low-​ neuronal the leptomeninges and cerebrum. Neuroimaging shows a tumour
    grade glial-​
    tumours (LGGNTs) such as dysembryoplastic neuroepithelial tu- mass with frequently cystic portions, moderate contrast enhance-
    mours (DNTs), ganglioglioma (GG), and desmoplastic infantile ment, and no pronounced oedema. The histological appearance
    astrocytoma/​ganglioglioma (DIA/​DIG), as well as other, less fre- of the tumour is very variable, containing large pleomorphic and
    quent diagnoses. The majority of paediatric LGGs harbour a variety frequently multinucleated cells, spindle and lipidized cells, a dense
    of genetic abnormalities frequently causing aberrant intracellular pericellular reticulin network, and numerous eosinophilic granular
    signalling via the Ras mitogen-​activated protein kinase (MAPK) bodies. An often present diffusely infiltrating component may
    pathway (see the section ‘Biology’). render paediatric-​type diffuse astrocytoma (WHO grade II) a differ-
    Paediatric PAs (WHO grade I) can occur throughout the neuraxis, ential diagnosis, especially in the absence of a typical pleomorphic
    but arise predominantly in infratentorial regions. Commonly af- pattern. The majority of PXAs harbour point mutations in the BRAF
    fected sites include (but are not restricted to) the cerebellum, optic gene (V600E; 50–​80%), and concomitant homozygous 9p21.3 de-
    nerve/​chiasm, hypothalamus, thalamus, and cerebral hemispheres. letions targeting the CDKN2A/​CDKN2B loci are common. If five
    PAs appear circumscribed upon imaging, with little (or no) oedema, or more mitoses per ten high-​power fields are present, tumours are
    and display variable proportions of cyst-​like areas intermixed with diagnosed as anaplastic PXAs (WHO grade III), which are associ-
    contrast-​enhancing solid components. Optic-​pathway tumours can ated with comparably worse overall survival.
    be sporadic but are a hallmark of neurofibromatosis type 1 (NF1), Among the group of mixed glial-​neuronal tumours are DNTs
    an autosomal dominant tumour predisposition syndrome caused (WHO grade I), which typically occur in children (or young adults)
    by germline mutations affecting the neurofibromin 1 gene; about with early-​onset epilepsy and account for more than 20% of long-​
    20% of NF1 patients develop a PA. Upon microscopy, they are char- term epilepsy-​associated tumours. They predominantly arise in cor-
    acterized by a biphasic pattern with compacted bipolar cells, with tical regions of the temporal lobe and, in contrast to diffuse gliomas,
    Rosenthal fibres alongside loose-​ textured multipolar cells with show little mass effect and no significant perifocal oedema upon
    microcysts. A histological variant, with tumour cells in a prominent imaging. Specific glioneuronal elements, characterized by columns
    180 CHAPTER 22 Glial Central Nervous System Tumours

    made up of bundles of axons orientated perpendicularly to the cor- Outside of the MAPK pathway, and enriched in paediatric-​type
    tical surface, are a histological hallmark of DNTs. diffuse astrocytomas and angiocentric gliomas (AGs—​ epilepsy-​
    Gangliogliomas (WHO grade I) most frequently occur in the tem- associated, slow-​growing cerebral tumours histologically character-
    poral lobe (more than 70%) of children and young adults. Imaging ized by an angiocentric growth pattern), are alterations in the MYB
    usually demonstrates a circumscribed solid or mixed solid and cystic and MYBL1 oncogenes. In contrast to the typical kinase fusions,
    mass with variable contrast enhancement and occasional tumour MYB and MYBL1 rearrangements result in loss of their Cʹ-​terminal
    calcification. Gangliogliomas are well-​differentiated glioneuronal portion (encoding a negative regulatory domain) but retention of
    neoplasms composed of dysplastic ganglion cells in combination the transactivating Nʹ-​terminus. The most commonly identified
    with neoplastic glial cells. BRAF V600E point mutations are present variant is a MYB:QKI fusion gene, acting through a tripartite mech-
    in up to 50% of gangliogliomas . The presence of an anaplastic glial anism of MYB activation by truncation, enhancer translocation
    component with elevated mitotic activity corresponds to a diagnosis driving aberrant MYB-​QKI expression, and hemizygous loss of the
    of anaplastic ganglioglioma (WHO grade III). tumour suppressive functions of QKI at the same time.
    DIAs/​DIGs (WHO grade I) are rare neoplasms of early child- Low-​grade tumours arising in children with NF1 represent a dis-
    hood, accounting for about 15% of infant tumours. Tumour location tinct molecular subset of LGG, characterized by biallelic inactiva-
    is always supratentorial, frequently involving more than one lobe. tion of the NF1 gene accompanied by increased MAPK signalling.
    Neuroimaging usually demonstrates a solid, superficial, enhancing While certain histologies are enriched for recurrent molecular al-
    portion extending to the leptomeninges, and a cystic portion terations (e.g. BRAF in PA/​ganglioglioma, FGFR1 in DNT, MYB in
    reaching into the cerebrum. Tumours are characterized by a prom- AG), it is becoming increasingly evident that classification schemes
    inent desmoplastic stroma with a neoplastic glial component (DIAs) based purely on histological criteria fail to recapitulate the substan-
    or neoplastic glioneuronal component (DIGs). tial heterogeneity of molecular LGG/​LGGNT subgroups. Some of
    these represent very rare but distinct biological entities, reflecting
    Biology the complex interplay among cellular origins, anatomical tumour
    Genetic aberrations targeting the MAPK signalling pathway are location, and susceptibility to particular oncogenic insults. Future
    a hallmark molecular feature of paediatric LGGs as well as some molecularly guided classification approaches may therefore distin-
    LGGNTs. As the most prominent example, a variety of single-​hit guish between BRAF fusion-​positive, BRAF V600E-​positive, and
    MAPK abnormalities, including gene fusions and point mutations, BRAF wildtype LGG/​LGGNT, and paediatric-​type diffuse gliomas
    can be found in essentially all PAs, which therefore has been hy- driven by MYB/​MYBL1, FGFR, and NRTK abnormalities, while
    pothesized to represent a prototypical single-​pathway disease. still acknowledging histopathological variants in each of these
    The most frequent genetic event are KIAA1549:BRAF fusion genes categories.
    detectable in 70–​80% of PAs, with a number of less frequent BRAF However, the diagnostic, prognostic, or predictive value of the
    fusion gene partners representing variants of a common molecular aforementioned molecular genetic markers remains to be deter-
    mechanism of MAPK activation (i.e. loss of the autoinhibitory Nʹ-​ mined. Their integration into a new standard of an integrated, multi-​
    terminal region of BRAF while retaining its active kinase domain). layered, histopathologic-​molecular diagnosis is therefore crucial
    Similar rearrangements involving upstream receptor tyrosine kinase for their prospective evaluation in not only improving diagnostic
    genes NRTK2 or NTRK3 (previously observed in infant high-​grade accuracy, but also identifying patients that may potentially benefit
    gliomas) have recently been identified across various histological from targeted therapeutic strategies such as BRAF V600E or MEK
    subtypes of LGG, as well as fusion events involving FGFR1 or FGFR3, inhibitors.
    previously observed in adult high-​grade gliomas and resulting in
    constitutive FGFR signalling. Duplications of the whole kinase do- Current treatment strategies
    main (internal tandem duplication, ITD) have been described as a Paediatric low-​grade gliomas (pLGGs) are the most common brain
    frequent alternative mechanism of FGFR1 activation in DNT and tumours in children, and the majority of patients will survive, but
    some PAs. The occurrence of NTRK and FGFR gene-​family re- often with significant long-​term deficits either from the tumour it-
    arrangements in both low-​and high-​grade gliomas suggests some self or secondary to treatment. These tumours can return multiple
    overlap in their molecular genetic make-​up in some instances. times over many years and follow a chronic disease course, with a
    The most common point mutations in LGG and LGGNT also cumulative treatment burden. This makes the choice and sequencing
    target the BRAF gene, typically resulting in a hotspot V600E al- of therapy complex, with a need to balance durable tumour control
    teration, and occur with high incidence in GG and PXA as well with protecting neurological function.
    as supratentorial PA. Hotspot mutations were also shown to occur Surgery is the treatment of choice for some hemispheric and cere-
    in FGFR1 (at p.N546 or p.K656) in some tumours, highlighting bellar LGGs when total resection can be accomplished with little
    the growing evidence for FGFR1 alterations as a molecular driver morbidity. The six-​year progression-​free survival (PFS) of a gross
    in LGG. total resection alone is 70% to 100%. However, owing to their prox-
    In line with location-​specific differences in DNA methylation and imity to vital brain structures (e.g. the hypothalamus, pituitary, optic
    gene-​expression patterns of LGG and LGGNT, the aforementioned pathways), many pLGGs can only undergo partial resection to facili-
    aberrations in the canonical MAPK pathway differ in their anatom- tate mechanical decompression or biopsy, and have a significantly
    ical distribution, with BRAF fusions predominating in cerebellar higher rate of progression post surgery. Asymptomatic pLGGs (e.g.
    tumours, BRAF V600E and NTRK gene-​family fusions more com- optic-​pathway gliomas in NF1 patients, tectal-​plate tumours post
    monly found in hemispheric supratentorial tumours, and FGFR1 al- CSF diversion) may not require immediate or even any treatment,
    terations typically occurring in midline tumours. whereas those that progress or present with symptoms often require
    High-grade gliomas 181

    non-​surgical treatment. Poor prognostic factors include young age downstream targets with a MEK inhibitor is thought to overcome
    (under one year old), grade II histology, and dissemination. this problem, and several of these inhibitors are being studied in pro-
    The selection of adjuvant therapy varies according to age, loca- gressive/​relapsed pLGG with preliminary results reported in a phase
    tion, and the presence of NF1. Children with NF1 and optic-​pathway I study of the MEK inhibitor selumetinib. In this study, selumetinib
    gliomas often have an indolent course and, in many patients, careful demonstrated encouraging preliminary activity, with 20% sus-
    observation may safely be undertaken, as up to 50% of tumours will taining partial responses and a two-​year PFS for those treated at the
    not progress. Radiotherapy has been considered a definitive treat- recommended phase II dose of 69%. The phase 2 study has been
    ment of pLGGs, with overall survival (OS) in excess of 90% and published and confirms this preliminary activity: NF1-​associated
    PFS of 60–​85%. However, it is associated with late effects which pLGG had a 40% objective response rate and 96% two-​year PFS, and
    depend on location, dose, volume, and age of the child. The most sporadic pLGG (either BRAF fusion-​or BRAF V600E-​positive) had
    frequent reported deficits are in neurocognition and neuroendo- a 36% objective response rate and 70% two-​year PFS. The toxicity
    crine function, but vasculopathy and second malignancy have also profile was similar to that seen in adults, with rash and mucositis,
    been documented, especially for NF1 patients, where irradiation is asymptomatic CPK elevation, and no retinal toxicity, all of which
    usually avoided. Advances in radiotherapy have led to increasingly will need monitoring in subsequent late phase trials.
    conformal focal approaches to reduce treatment volumes, and to try Confirmatory data, including response rates according to
    and reduce side effects. Proton beam radiotherapy is one such tech- underlying biology (e.g. NF1, BRAF fusion, BRAF V600E) and
    nique which can decrease the dose to normal brain by the possible duration of tumour control off treatment, are eagerly awaited from
    lower entrance dose and elimination of the exit dose, compared with studies of selumetinb, trametinib, cobimetinib, and MEK-​162 in
    conventional photon beams. Reports from pLGG show good disease pLGG. For BRAF V600E-​mutated pLGG, selective BRAF inhibitors
    control with 83% PFS and 100% OS rates, respectively, at eight years. represent an attractive option and the first results using dabrafenib
    Even with proton beams, significant neurocognitive decline has have been reported in abstract form. The drug was well tolerated,
    been seen in younger patients (under seven years), but for older pa- with the most common side effects related to skin rash being similar
    tients and with sparing of the left temporal lobe, hippocampus, and to those observed in adults. Promising response rates were reported
    hypothalamic-​pituitary good clinical outcomes have been reported. for all but three out of the 32 enrolled, demonstrating tumour con-
    To avoid or delay radiotherapy, those young patients (typic- trol with a high objective response rate. Such responses have also
    ally under seven to eight years of age) and patients with NF1 re- been reported with other BRAF inhibitors in pLGG.
    quiring treatment usually receive chemotherapy. The combination
    of vincristine and carboplatin is currently regarded as the standard Future outlook
    regimen, and several studies report five-​year PFS rates of 35–​45% It is likely that future clinical trials and treatment strategies will con-
    for sporadic pLGG and higher in those with NF1 (60–​65%). Disease centrate on the dual aim of durable tumour control and improving
    control tends to plateau in many NF1 patients over the first five the quality of survival. This will be achieved by focusing on protecting
    years, but in sporadic patients this tends to fall over the first five to or improving neurological function and decreasing the overall treat-
    ten years. This means that patients will often require several rounds ment burden, with an emphasis on using biology to optimize patient
    of treatment and may have to discontinue initial therapy due to selection and risk stratification. The use of molecular analysis as a
    treatment-​related effects (e.g. carboplatin allergy). This has resulted prognostic and predictive tool will lead to more patients being biop-
    in many alternative chemotherapy regimens being explored, either sied and considered for targeted therapies. Over the next few years,
    to try and improve PFS by intensification (e.g. multi-​agent regimens studies will assess the impact of MAPK inhibitors as single agents
    such as TPCV (thioguanine, procarbazine, CCNU, vincristine) or and in combinations, and explore agents such as bevacizumab which
    to reduce toxicity and maintain PFS (e.g. single-​agent carboplatin, may play a role not just in tumour control but also in protecting
    vinblastine, vinorelbine). against or reversing neurological deficits.
    With increasing knowledge of the underlying biology of pLGG
    and the desire to improve function whilst decreasing treatment
    burden, targeted therapy is now being explored. One example has
    been SEGA in TSC, where mTOR inhibition has been used when
    High-​grade gliomas
    surgical resection is difficult. Whether mTOR inhibitors have a role
    in PA remains unclear. The anti-​angiogenic agent bevacizumab has Classification and pathology
    been combined with chemotherapy and shown good short-​term re- High-​grade gliomas (HGGs) in children and adolescents comprise
    sponse rates with functional improvement (e.g. vision recovery) in a group of diffusely infiltrating, malignant CNS tumours, that are
    some patients, and optimization of this agent with standard chemo- situated supratentorially, as well as in midline structures of the
    therapy is now being investigated. brain and (more rarely) the spinal cord. These tumours occur in
    Most recently, the main focus has been on agents targeting the any age group, ranging from neonates to young adults, with a peak
    MAPK pathway, which has been shown to be activated in the ma- incidence in the second decade of life, and display a relatively equal
    jority of pLGGs. The first attempt at targeting this pathway in pLGG sex distribution.
    used the agent sorafenib as a BRAF inhibitor, but this resulted in rapid Apart from specific genetic tumour-​predisposition syndromes
    disease progression in patients with tumours with KIAA1549:BRAF such as DNA mismatch repair, Li–​Fraumeni syndrome (LFS), and
    fusions. This occurred as the KIAA1549:BRAF protein functions as neurofibromatosis type I, aetiology is largely unknown.
    a homodimer, which results in resistance to first-​generation BRAF On magnetic resonance imaging (MRI), supratentorial HGGs
    inhibitors due to paradoxical MAPK activation. Targeting more usually present as poorly defined tumour masses. Irregular-​shaped
    182 CHAPTER 22 Glial Central Nervous System Tumours

    contrast enhancement around a dark central core of necrosis, which (three lobes or more) to both cerebral hemispheres with add-
    may have a haemorrhagic component, is a typical radiological fea- itional involvement of the brainstem, cerebellum, and spinal cord.
    ture of glioblastomas. Surrounding oedema is frequent and often Since gliomatosis cerebri is lacking a unique underlying molecular
    contains infiltration by neoplastic cells. Where possible, tumour biology, it is considered a more invasive phenotype of HGG (rather
    resection or biopsy, followed by histopathological and molecular than a distinct nosological entity). The mechanisms by which these
    work-​up, establish an integrated diagnosis for clinical management. tumours display this extensively diffuse phenotype are yet unknown.
    On pathology, paediatric HGGs have a varying histological ap-
    pearance. Most display morphological and immunohistochemical Biology
    features of anaplastic astrocytoma (i.e. atypia, anaplasia, and mi- Recent collaborative efforts directed at the underlying molecular
    totic activity) or glioblastoma (with additional microvascular pro- biology of paediatric HGG have unravelled fundamental differ-
    liferation and/​ or pseudopalisading necrosis). The vast majority ences compared to other CNS neoplasms arising in both adults and
    of supratentorial paediatric HGGs are classified as ‘anaplastic children, highlighted by the discovery of somatic histone muta-
    astrocytoma, isocitrate dehydrogenase (IDH)-​ wildtype’ (WHO tions in up to 50% of these tumours, affecting key residues for post-​
    grade III) or ‘glioblastoma, IDH-​wildtype’ (WHO grade IV) ac- translational histone tail modification.
    cording to the revised 2016 fourth edition of the WHO classification The majority of diffuse midline gliomas in the brainstem (more
    of tumours of the CNS. In contrast to adult HGG, IDH1/​2 mutations than 85%), thalamus (about 50%), and spinal cord (about 60%) har-
    in this age group are very rare (see the section ‘Biology’). bour mutations at position 27 (K27M; lysine replaced by methio-
    Of specific interest, the 2016 WHO classification has introduced a nine) in genes coding for histone variants H3.3 (H3F3A; about 75%)
    new category of HGGs situated in the midline of the brain, called ‘dif- and H3.1 (HIST1H3B/​C; about 25%), or rare variations. Independent
    fuse midline glioma, H3 K27M-​mutant’ (WHO grade IV), referring of which histone 3 variant is affected, the K27M-​mutant protein in-
    to infiltrative midline HGGs with predominantly astrocytic differen- hibits polycomb repressive complex 2 (PRC2) activity—​via seques-
    tiation that primarily affect children, harbouring highly specific mu- tration of its catalytic subunit EZH2—​resulting in a global decrease
    tations in histone genes (see the section ‘Biology’). On morphology/​ of H3 K27 trimethylation (which can serve as a molecular surrogate
    immunohistochemistry, these tumours can show low-​to high-​grade marker). The exact downstream processes driving the evolution of
    histology with equally poor outcomes, clearly discerning them from H3 K27M-​mutated gliomas remain to be elucidated, but likely in-
    non-​histone-​mutated LGGs. This new pathological entity not only volve other non-​recurrent mutations targeting the epigenetic ma-
    relates to tumours that have been termed ‘diffuse intrinsic pontine chinery such as members of the lysine-​specific methyltransferase
    glioma’ (DIPG), but also to diffuse gliomas in non-​pontine midline (KMT), lysine-​specific demethylase (KDM), and chromodomain
    structures such as the thalamus, medulla oblongata, or spinal cord. helicase DNA-​binding protein (CHD) families. Additional mu-
    Initially thought to only occur in paediatric high-​grade astrocytic tations in canonical cancer pathways target the receptor tyrosine
    tumours, the histological spectrum of CNS tumours with H3 K27M kinase/​RAS/​PI3K pathway (e.g. mutations in PDGFRA, PIK3CA,
    mutations has recently been expanded to lower-​grade lesions and PIK3R1, or PTEN; occurring in about 50% of tumours), the p53
    rare instances of posterior fossa ependymomas. pathway (e.g. mutations in TP53, PPM1D, CHEK2, or ATM;
    Although paediatric HGGs are rare compared to their adult coun- occurring in about 70% of tumours), and the retinoblastoma pro-
    terparts, they altogether add up to be one of the most common ma- tein pathway. Structural variations across diffuse midline gliomas
    lignant childhood CNS tumours, with an age-​adjusted incidence of include focal gene amplifications of PDGFRA, MYC/​MYCN, CDK4/​
    0.26 per 100,000. This might also be an underestimation, as tumours 6, CCND1–​3, ID2, or MET.
    classified as H3 K27-​mutant diffuse midline gliomas might have pre- Location-​specific patterns of genetic alterations and clinical char-
    viously been classified as low-​grade malignant/​diffuse astrocytomas acteristics suggest biological diversity within diffuse midline K27M-​
    (or not have been biopsied at all), and another fraction might pre- mutated gliomas. Tumours harbouring H3.3 mutations develop
    viously have been classified as CNS-​primitive neuroectodermal tu- across midline structures (co-​occurring with FGFR1 mutations
    mours, a term describing a heterogeneous group of CNS neoplasms in some thalamic gliomas), typically affect children aged seven to
    and no longer used in the WHO classification. ten years, and are associated with very poor outcome. In contrast,
    Due to their critical anatomical location precluding surgical re- H3.1 mutations are largely restricted to diffuse pontine gliomas with
    section, the diagnosis of DIPG has historically been based on the earlier onset (four to six years), have been associated with distinct
    combination of characteristic clinical symptoms (signs of brain- clinicopathological and radiological features as well as a slightly
    stem dysfunction and/​or CSF obstruction, developing over a short better prognosis, and frequently co-​occur with mutations in ACVR1.
    period of time) and radiological parameters (large, expansile, often Up to one third of non-​midline paediatric HGGs carry muta-
    asymmetrical brainstem mass occupying more than two thirds of tions at position 34 (G34R/​V; glycine replaced by arginine/​valine)
    the pons). Reintroducing safe stereotactic biopsies in these pontine in the H3F3A gene (encoding histone 3.3), the exact molecular
    tumours has led to the identification of pathognomonic DNA mu- consequences of which are not yet fully understood. However, as-
    tations (H3 K27M; see the section ‘Biology’) that also occur across sociations with mutations in ATRX and/​or DAXX and subtelomeric
    other midline locations, and that now serve as robust molecular DNA hypomethylation point towards a role for telomerase-​
    markers for pathological diagnosis with additional prognostic value. independent telomere maintenance mechanisms (i.e. alternative
    Both anaplastic astrocytomas and glioblastomas, as well as diffuse lengthening of telomeres; ALT) in this subset of tumours, in which
    midline gliomas, may manifest at initial clinical presentation with a mutations in TP53 and MGMT promoter methylation are also com-
    ‘gliomatosis cerebri pattern’—​an extremely diffuse phenotype, with monly observed. Restricted to hemispheric locations, H3.3 G34R/​V-​
    the affected area ranging from most of one cerebral hemisphere mutated gliomas manifest during adolescence or young adulthood
    High-grade gliomas 183

    fh a
    bd b
    g e
    h 6 7 f
    1 h
    d 5
    h
    4 c
    b 2
    h a
    3
    e b
    a h
    b
    Hemispheric/cortical
    h
    Supratentorial midline a f
    b a
    7
    d h
    c
    5
    b
    h 3
    b 1
    a
    h f b h
    a h
    c 7
    b f h 3
    C a
    b
    Histology Genetics
    1 PA a BRAF Fusion
    1
    2 DNET b BRAF V600E
    3 GG c FGFR Infratentorial
    4 PXA d MYB/MYBL1
    5 DA e NTRK
    6 SEGA f NF1
    7 OTHER g TSC1/2
    h OTHER/NA

    Figure 22.1 Distribution of paediatric low-​grade glioma histologies and molecular genetic alterations by anatomical tumour location. Inner pie
    charts represent relative frequencies of the most common paediatric low-​grade glioma histological entities, indicated by different shading. Outer rings
    represent the most common molecular genetic alterations associated with each histological entity in a given location.
    DA, diffuse astrocytoma; DNET, dysembryoplastic neuroepithelial tumour; GG, ganglioglioma; PA, pilocytic astrocytoma; PXA, pleomorphic xanthoastrocytoma; SEGA,
    subependymal giant-​cell astrocytoma.
    Adapted with permission from Sturm D, et al. ‘Pediatric Gliomas: Current Concepts on Diagnosis, Biology, and Clinical Management’. Journal of Clinical Oncology, Volume 35,
    Issue 21, pp. 2370–​2377. Copyright © 2017 American Society of Clinical Oncology. All rights reserved. DOI: 10.1200/​JCO.2017.73.0242

    (10–​25 years) and are associated with prolonged overall survival gliomas, resection is not an option. Radiotherapy is considered to
    compared with other HGGs. be the standard, albeit palliative care modality in these patients,
    A minor fraction (less than 5%) of hemispheric HGGs in older improving PFS and OS by some months and positively influencing
    adolescents display hotspot mutations in IDH1/​2 genes—​a quasi-​ quality of life. Although not investigated in randomized studies and
    defining hallmark of secondary glioblastomas and lower-​grade dif- distilled from retrospective data, chemotherapy in addition to radio-
    fuse gliomas from which they develop—​and therefore represent the therapy seems to convey a minor benefit in PFS and OS. Very rarely,
    lower-​age spectrum of gliomas in young adults (40–​50 years) that long-​term survival is reported in these diseases.
    are enriched in the frontal lobe and share the overall best prognosis. For supratentorial HGG, standard therapy consists of multi-
    The associations of hotspot mutations in histone (H3.3/​H3.1) modal treatment combining maximal resection, post-​ operative
    and IDH1/​2 genes with anatomical location, patient age, and out- conformal (chemo-​) radiotherapy, and adjuvant chemotherapy.
    come have led to our understanding of distinct biological variants Neurosurgical resection of supratentorial and spinal paediatric
    of (paediatric) HGG (Figure 22.2). Each of these variants displays HGG is an important part of therapy, as the extent of tumour re-
    a unique DNA methylation pattern thought to be an imprint of dif- section is positively correlated with long-​term survival. As HGGs
    ferent cell-​of-​origin populations during neurodevelopment, and invade brain tissue beyond the tumour margins visible on MRI,
    shaped by mutations rewiring their epigenome. maximal surgery alone is not considered curative, necessitating
    The remaining proportion of non-​H3/​non-​IDH-​mutated HGGs adjuvant radio-​and chemotherapy. Since the 1960s, radiation
    (less than 50%) is likely to include further biological subtypes, therapy (RT) adjuvant to surgery has been the mainstay of therapy,
    some of which might be driven by MYC(N) amplifications, receptor improving survival, although not proven in a randomized fashion.
    tyrosine kinase amplifications, BRAF mutations, or fusion events Focal radiation of 54 Gy is usually employed in fractions of 1.8–​2
    involving, for example, NTRK or FGFR genes. However, there are Gy, with boosts up to 59.4 Gy to residual tumour. Intensification
    a substantial number of paediatric malignant gliomas with rela- of radiotherapy by hyperfractionation to a total dose of 72 Gy has
    tively stable genomic profiles for which the molecular mechanisms failed to improve outcome. New radiotherapy techniques, such as
    underlying gliomagenesis are yet to be elucidated. intensity-​modulated radiation therapy (IMRT) and proton-​beam
    radiotherapy, have been developed to allow for more precise de-
    Current treatment strategies livery to the tumour bed while sparing normal brain tissue, but the
    The treatment of paediatric HGG patients is challenging. For pa- efficacy in the context of diffusely infiltrating gliomas needs to be
    tients suffering from DIPG and other WHO grade IV diffuse midline ascertained.
    184 CHAPTER 22 Glial Central Nervous System Tumours

    (a) Pons (b)


    H3.3 K27M
    H3.1 K27M
    H3.2 K27M
    H3.1 K27I Hemispheric/cortical
    ACVR1
    TP53/PPM1D
    PDGFRA
    MYC/MYCN amp.
    CCND/CDK amp.
    PIK3CA/PIK3R1
    H3 K27me3 loss
    Supratentorial midline

    Supratentorial midline
    H3.3 K27M
    H3.1 K27M
    TP53/PPM1D
    Pons
    PDGFRA
    CDKN2A/B del.
    MYC/MYCN amp.
    CCND/CDK amp.
    PIK3CA/PIK3R1
    FGFR1 & NF1 (c)
    H3 K27me3 loss

    30
    Hemispheric/cortical
    H3.3 G34R/V
    BRAF V600E
    IDH1/2 20

    Age (Years)
    NTRK
    TP53
    PDGFRA
    MYC/MYCN amp. 10
    CCND/CDK amp.
    PIK3CA/PIK3R1
    CDKN2A/B del.
    EGFR/MET 0
    ATRX Loss NTRK H3.1 K27 H3.3 K27 H3 G34 IDH1/2 Other

    Figure 22.2 Molecular patterns and clinical features of paediatric high-​grade glioma subgroups. (A) Oncoprints schematically depicting selected
    common molecular alterations in paediatric high-​grade gliomas by anatomical location. Shaded boxes indicate alterations being present. (B) Pie charts
    representing relative frequencies of common paediatric high-​grade glioma molecular alterations represented by shading corresponding to (A) and (C).
    (C) Age distribution of common paediatric high-​grade glioma molecular alterations in children and young adults (age younger than 30 years).
    amp, amplification; del, deletion
    Adapted with permission from Sturm D, et al. ‘Pediatric Gliomas: Current Concepts on Diagnosis, Biology, and Clinical Management’. Journal of Clinical Oncology, Volume 35,
    Issue 21, pp. 2370–​2377. Copyright © 2017 American Society of Clinical Oncology. All rights reserved. DOI: 10.1200/​JCO.2017.73.0242

    Because of its detrimental neurological long-​term sequelae, es- With significant improvement in survival reported in adults with
    pecially in tumours situated supratentorially, radiotherapy is usu- GBM treated with temozolomide (TMZ) added to radiation and
    ally either omitted or postponed in children under four years of maintenance therapy, the non-​randomized single-​arm ACNS0126
    age, and replaced by high-​dose chemotherapy regimens. Although study investigated the use of TMZ in paediatric HGG, showing equal
    profound evidence of efficacy is lacking, most children suffering survival rates to the more toxic CCG-​943 and CCG-​945 trials. Based
    from HGGs are nowadays treated with chemotherapy concomi- on this study, radiochemotherapy with TMZ has been more or less
    tant and adjuvant to radiotherapy, based on experiences in adult considered a backbone for HGG therapy. More recently, results from
    HGG. In the late 1980s, the Children’s Cancer Group (CCG) the non-​randomized, single-​arm ACNS0423 study indicate that the
    showed improved survival in a first randomized study (CCG-​943) reintroduction of CCNU to the backbone of TMZ-​RT has a survival
    for paediatric patients with non-​brainstem anaplastic astrocytoma benefit over TMZ-​RT, most notably in GBM, although survival is
    and glioblastoma treated with radiotherapy with concomitant and still poor, with a three-​year PFS of 22%. Anti-​angiogenesis strategies
    maintenance prednisone, CCNU (procarbazine, 6-​thioguanine, to the TMZ-​RT backbone, in analogy to adult GBM studies, are cur-
    dibromodulcitol, and lomustine) and vincristine (pCV), com- rently being investigated.
    pared to radiotherapy alone. Glioblastoma multiforme (GBM) In younger children in whom radiotherapy is postponed or
    patients seemed to particularly benefit from the addition of omitted, high-​dose chemotherapy with stem-​cell support regimens
    chemotherapy. A later study of central neuropathological review has yielded promising results, albeit these are studies with small
    of cases treated in the CCG-​943 study indicated that a large pro- numbers of patients with potentially different tumour biology and/​
    portion of patients randomized could actually be reclassified to or blood–​brain barrier dynamics.
    LGGs, possibly influencing the reported survival. The subsequent
    randomized CCG-​945 study—​two cycles of post-​operative pre-​ Future outlook
    radiotherapy multi-​drug chemotherapy (vincristine, hydroxyurea, Future studies need to be focused on analysis of retrospective and
    procarbazine, CCNU, cisplatin, cytarabine, methylprednisolone, prospective paediatric HGG patient data from multiple trials, cor-
    and cyclophosphamide or dacarbazine)—​failed to show a benefit relating clinical, radiological, pathological, and molecular biology
    over the CCG-​943 pCV scheme. data to treatment and outcome data. As it is more and more evident
    High-grade gliomas 185

    that these diffusely infiltrative diseases show large intratumoural Cohen KJ, et al. (2011) Temozolomide in the treatment of high-​grade
    heterogeneity, both of the tumour cells and the micro-​environment, gliomas in children: a report from the Children’s Oncology Group.
    including the blood–​brain barrier, research needs to be addressed at Neuro Oncol 13, 317–​23.
    combinational (targeted) treatment strategies. Novel drug targeting Dodgshun AJ, et al. (2016) Single agent carboplatin for pediatric low-​
    and delivery systems, such as convection-​ enhanced, interstitial grade glioma: a retrospective analysis shows equivalent efficacy to
    multiagent chemotherapy. Int J Cancer 138, 481–​8.
    delivery of drugs and oncolytic viruses, and focused ultrasound-​
    Dohrmann GJ, et al. (1976) Glioblastoma multiforme in children. J
    assisted (nanocarrier-​mediated) drug delivery, as well as immuno-
    Neurosurg 44, 442–​8.
    therapies, are of particular interest in this respect.
    Dougherty MJ, et al. (2010) Activating mutations in BRAF charac-
    terize a spectrum of pediatric low-​grade gliomas. Neuro Oncol 12,
    FURTHER READING 621–​30.
    Fangusaro J (2012) Pediatric high grade glioma: a review and update
    Adamski J, et al. (2014) Advances in the management of paediatric on tumor clinical characteristics and biology. Front Oncol 2, 105.
    high-​grade glioma. Curr Oncol Rep 16, 414. Fangusaro J, et al. (2019) Selumetinib in paediatric patients with BRAF-​
    Ater JL, et al. (2016) Nonrandomized comparison of neurofibromatosis aberrant or neurofibromatosis type 1-​associated recurrent, refrac-
    type 1 and non-​neurofibromatosis type 1 children who received tory, or progressive low-​grade glioma: a multicentre, phase 2 trial.
    carboplatin and vincristine for progressive low-​grade glioma: a re- Lancet Oncol 20(7), 1011–​22. doi: 10.1016/​S1470-​2045(19)30277-​3.
    port from the Children’s Oncology Group. Cancer 122, 1928–​36. Fisher PG, et al. (2008) Outcome analysis of childhood low-​grade
    Ater JL, et al. (2012) Randomized study of two chemotherapy regimens astrocytomas. Pediatr Blood Cancer 51, 245–​50.
    for treatment of low-​grade glioma in young children: a report from Fouladi M, et al. (2003) Outcome of children with centrally reviewed
    the Children’s Oncology Group. J Clin Oncol 30, 2641–​7. low-​grade gliomas treated with chemotherapy with or without
    Bandopadhayay P, et al. (2014) Long-​term outcome of 4,040 chil- radiotherapy on Children’s Cancer Group high-​grade glioma study
    dren diagnosed with pediatric low-​grade gliomas: an analysis of CCG-​945. Cancer 98, 1243–​52.
    the Surveillance Epidemiology and End Results (SEER) database. Franz DN, et al. (2013) Efficacy and safety of everolimus for
    Pediatr Blood Cancer 61, 1173–​9. subependymal giant cell astrocytomas associated with tuberous
    Bandopadhayay P, et al. (2016) MYB-​ QKI rearrangements in sclerosis complex (EXIST-​1): a multicentre, randomised, placebo-​
    angiocentric glioma drive tumorigenicity through a tripartite controlled phase 3 trial. Lancet 381, 125–​32.
    mechanism. Nat Genet 48, 273–​82. Fulton DS, et al. (1992) Increasing radiation dose intensity using
    Banerjee A, et al. (2017) A phase I trial of the MEK inhibitor selumetinib hyperfractionation in patients with malignant glioma. Final re-
    (AZD6244) in pediatric patients with recurrent or refractory low-​ port of a prospective phase I-​II dose response study. J Neurooncol
    grade glioma: a Pediatric Brain Tumor Consortium (PBTC) study. 14, 63–​72.
    Neuro Oncol 19(8), 1135–​44. Gessi M, et al. (2016) Evidence of H3 K27M mutations in posterior
    Blumcke I, et al. (2014) A neuropathology-​based approach to epilepsy fossa ependymomas. Acta Neuropathol 132, 635–​7.
    surgery in brain tumors and proposal for a new terminology use Gnekow AK, et al. (2012) Long-​term follow-​up of the multicenter,
    for long-​term epilepsy-​associated brain tumors. Acta Neuropathol multidisciplinary treatment study HIT-​LGG-​1996 for low-​grade
    128, 39–​54. glioma in children and adolescents of the German Speaking
    Brada M, Haylock B (2014) Is current technology improving outcomes Society of Pediatric Oncology and Hematology. Neuro-​Oncology,
    with radiation therapy for gliomas? Am Soc Clin Oncol Educ Book 14, 1265–​84.
    e89–​94. doi: 10.14694/​EdBook_​AM.2014.34.e89 Gnekow AK, et al. (2004) Low grade chiasmatic-​hypothalamic glioma-​
    Broniscer A, Gajjar A (2004) Supratentorial high-​grade astrocytoma carboplatin and vincristin chemotherapy effectively defers radio-
    and diffuse brainstem glioma: two challenges for the pediatric on- therapy within a comprehensive treatment strategy—​report from
    cologist. Oncologist 9, 197–​206. the multicenter treatment study for children and adolescents with
    Broniscer A, et al. (2016) Gliomatosis cerebri in children shares a low grade glioma—​HIT-​LGG 1996—​of the Society of Pediatric
    molecular characteristics with other pediatric gliomas. Acta Oncology and Hematology (GPOH). Klin Padiatr 216, 331–​42.
    Neuropathol 131, 299–​307. Greenberger BA, et al. (2014) Clinical outcomes and late endocrine,
    Cappellano AM, et al. (2015) Single agent vinorelbine in pediatric pa- neurocognitive, and visual profiles of proton radiation for pediatric
    tients with progressive optic pathway glioma. J Neurooncol 121, low-​grade gliomas. Int J Radiat Oncol Biol Phys 89, 1060–​8.
    405–​12. Grill J, et al. (1999) Radiation-​induced cerebral vasculopathy in chil-
    Castel D, et al. (2015) Histone H3F3A and HIST1H3B K27M muta- dren with neurofibromatosis and optic pathway glioma. Ann Neurol
    tions define two subgroups of diffuse intrinsic pontine gliomas with 45, 393–​6.
    different prognosis and phenotypes. Acta Neuropathol 130, 815–​27. Gururangan S, et al. (2014) Efficacy of bevacizumab plus irinotecan
    Chamdine O, et al. (2016) Metastatic low-​grade gliomas in chil- in children with recurrent low-​grade gliomas—​a Pediatric Brain
    dren: 20 years’ experience at St Jude Children’s Research Hospital. Tumor Consortium study. Neuro Oncol 16, 310–​17.
    Pediatr Blood Cancer 63, 62–​70. Gururangan S, et al. (2007) Temozolomide in children with progres-
    Chappe C, et al. (2013) Dysembryoplastic neuroepithelial tumors sive low-​grade glioma. Neuro Oncol 9, 161–​8.
    share with pleomorphic xanthoastrocytomas and gangliogliomas Hargrave D, et al. (2006) Diffuse brainstem glioma in children: critical
    BRAF(V600E) mutation and expression. Brain Pathol 23, 574–​83. review of clinical trials. Lancet Oncol 7, 241–​8.
    Chintagumpala M, et al. (2015) A pilot study using carboplatin, vin- Hersh DS, et al. (2016) Emerging applications of therapeutic ultra-
    cristine, and temozolomide in children with progressive/​symptom- sound in neuro-​ oncology: moving beyond tumor ablation.
    atic low-​grade glioma: a Children’s Oncology Group study dagger. Neurosurgery 79, 643–​54.
    Neuro Oncol 17, 1132–​8.
    186 CHAPTER 22 Glial Central Nervous System Tumours

    Hutt-​Cabezas M, et al. (2013) Activation of mTORC1/​ mTORC2 Louis DN, et al. (2016) WHO Classification of Tumours of the Central
    signaling in pediatric low-​grade glioma and pilocytic astrocytoma Nervous System. Switzerland: WHO Press.
    reveals mTOR as a therapeutic target. Neuro Oncol 15, 1604–​14. Massimino M, et al. (2010) Is there a role for myeloablative chemo-
    Jakacki RI, et al. (2011) A phase 1 study of vinblastine in combination therapy with autologous hematopoietic cell rescue in the manage-
    with carboplatin for children with low-​grade gliomas: a Children’s ment of childhood high-​grade astrocytomas? Pediatr Blood Cancer
    Oncology Group phase 1 consortium study. Neuro Oncol 13, 54, 641–​3.
    910–​15. Merchant TE, et al. (2009) Late effects of conformal radiation therapy
    Jakacki RI, et al. (2016) Phase 2 study of concurrent radiotherapy for pediatric patients with low-​grade glioma: prospective evalu-
    and temozolomide followed by temozolomide and lomustine in ation of cognitive, endocrine, and hearing deficits. J Clin Oncol 27,
    the treatment of children with high-​grade glioma: a report of the 3691–​7.
    Children’s Oncology Group ACNS0423 study. Neuro Oncol 18, Miller C, et al. (2017) Report of effective trametinib therapy in 2
    1442–​50. children with progressive hypothalamic optic pathway pilocytic
    Jansen MH, et al. (2012) Diffuse intrinsic pontine gliomas: a systematic astrocytoma: documentation of volumetric response. J Neurosurg
    update on clinical trials and biology. Cancer Treat Rev 38, 27–​35. Pediatr 19, 319–​24.
    Jansen MH, et al. (2015) Survival prediction model of children with Northrup H, et al. (2013) Tuberous sclerosis complex diagnostic cri-
    diffuse intrinsic pontine glioma based on clinical and radiological teria update: recommendations of the 2012 International Tuberous
    criteria. Neuro Oncol 17, 160–​6. Sclerosis Complex Consensus Conference. Pediatr Neurol 49, 243–​54.
    Jaspan T, et al. (2016) Response assessment in pediatric neuro-​ Olow A, et al. (2016) BRAF status in personalizing treatment ap-
    oncology: implementation and expansion of the RANO criteria in proaches for pediatric gliomas. Clin Cancer Res 22, 5312–​21.
    a randomized phase II trial of pediatric patients with newly diag- Ostrom QT, et al. (2016) CBTRUS statistical report: primary brain and
    nosed high-​grade gliomas. Am J Neuroradiol 37, 1581–​7. other central nervous system tumors diagnosed in the United States
    Johansson G, et al. (2016) Recent developments in brain tumor in 2009–​2013. Neuro-​Oncology, 18, v1–​75.
    predisposing syndromes. Acta Oncol 55, 401–​11. Packer RJ, et al. (1997) Carboplatin and vincristine chemotherapy for
    Jones C, Baker SJ (2014) Unique genetic and epigenetic mechanisms children with newly diagnosed progressive low-​grade gliomas. J
    driving paediatric diffuse high-​ grade glioma. Nat Rev Cancer Neurosurg 86, 747–​54.
    14(10). doi: 10.1038/​nrc3811 Packer RJ, et al. (2009) Objective response of multiple recurrent low-​
    Jones DT, et al. (2013) Recurrent somatic alterations of FGFR1 and grade gliomas to bevacizumab and irinotecan. Pediatr Blood Cancer
    NTRK2 in pilocytic astrocytoma. Nat Genet 45, 927–​32. 52, 791–​5.
    Jones DT, et al. (2008) Tandem duplication producing a novel Qaddoumi I, et al. (2016) Genetic alterations in uncommon low-​grade
    oncogenic BRAF fusion gene defines the majority of pilocytic neuroepithelial tumors: BRAF, FGFR1, and MYB mutations occur
    astrocytomas. Cancer Res 68, 8673–​7. at high frequency and align with morphology. Acta Neuropathol
    Kalra M, et al. (2015) Confirmation of bevacizumab activity, and 131, 833–​45.
    maintenance of efficacy in retreatment after subsequent relapse, in Ramkissoon LA, et al. (2013) Genomic analysis of diffuse pediatric
    pediatric low-​grade glioma. J Pediatr Hematol Oncol 37, e341–​46. low-​grade gliomas identifies recurrent oncogenic truncating re-
    Karajannis MA, et al. (2014) Phase II study of sorafenib in children arrangements in the transcription factor MYBL1. Proc Natl Acad
    with recurrent or progressive low-​grade astrocytomas. Neuro Oncol Sci USA 110, 8188–​93.
    16, 1408–​16. Sanders RP, et al. (2007) High-​grade astrocytoma in very young chil-
    Karim R, et al. (2016) Nanocarriers for the treatment of glioblastoma dren. Pediatr Blood Cancer 49, 888–​93.
    multiforme: current state-​of-​the-​art. J Control Release 227, 23–​37. Schindler G, et al. (2011) Analysis of BRAF V600E mutation in 1,320
    Kieran MW, et al. (2016) The first study of dabrafenib in pediatric pa- nervous system tumors reveals high mutation frequencies in pleo-
    tients with BRAF V600—​mutant relapsed or refractory low-​grade morphic xanthoastrocytoma, ganglioglioma and extra-​cerebellar
    gliomas. Ann Oncol 27(Suppl 6), LBA19_​PR. of Oncology, Volume pilocytic astrocytoma. Acta Neuropathol 121, 397–​405.
    27, Issue suppl_​6, 1 October 2016, LBA19_​PR. doi.org/​10.1093/​ Schwartzentruber J, et al. (2012) Driver mutations in histone H3.3 and
    annonc/​mdw435.09 chromatin remodelling genes in paediatric glioblastoma. Nature
    Krueger DA, et al. (2010) Everolimus for subependymal giant-​cell 482, 226–​31.
    astrocytomas in tuberous sclerosis. N Engl J Med 363, 1801–​11. Sievert AJ, et al. (2013) Paradoxical activation and RAF inhibitor re-
    Lassaletta A, et al. (2016) Profound clinical and radiological response sistance of BRAF protein kinase fusions characterizing pediatric
    to BRAF inhibition in a 2-​month-​old diencephalic child with hypo- astrocytomas. Proc Natl Acad Sci USA 110, 5957–​62.
    thalamic/​chiasmatic glioma. Pediatr Blood Cancer 63, 2038–​41. Skrypek M, et al. (2014) Pilomyxoid astrocytoma treated successfully
    Lassaletta A, et al. (2016) Phase II weekly vinblastine for chemotherapy-​ with vemurafenib. Pediatr Blood Cancer 61, 2099–​100.
    naive children with progressive low-​grade glioma: a Canadian Stokland T, et al. (2010) A multivariate analysis of factors determining
    Pediatric Brain Tumor Consortium Study. J Clin Oncol 34(29), tumor progression in childhood low-​grade glioma: a population-​
    3537–​43. based cohort study (CCLG CNS9702). Neuro Oncol 12, 1257–​68.
    Lassaletta A, et al. (2016) An integrative molecular and genomic ana- Stupp R, et al. (2005) Radiotherapy plus concomitant and adjuvant
    lysis of pediatric hemispheric low-​grade gliomas: an update. Childs temozolomide for glioblastoma. N Engl J Med 352, 987–​96.
    Nerv Syst 32, 1789–​97. Sturm D, et al. (2014) Paediatric and adult glioblastoma: multiform
    Lewis PW, et al. (2013) Inhibition of PRC2 activity by a gain-​of-​ (epi)genomic culprits emerge. Nat Rev Cancer 14, 92–​107.
    function H3 mutation found in pediatric glioblastoma. Science 340, Sturm D, et al. (2016) New brain tumor entities emerge from mo-
    857–​61. lecular classification of CNS-​PNETs. Cell 164, 1060–​72.
    Listernick R, et al. (1999) Intracranial gliomas in neurofibromatosis Sturm D, et al. (2017) Pediatric gliomas: current concepts on diagnosis,
    type 1. Am J Med Genet 89, 38–​44. biology, and clinical management. J Clin Oncol 35(21), 2370–​7.
    High-grade gliomas 187

    Sturm D, et al. (2012) Hotspot mutations in H3F3A and IDH1 de- low-​ grade glioma treated with conformal radiation therapy. J
    fine distinct epigenetic and biological subgroups of glioblastoma. Neuro-​oncol 122, 161–​8.
    Cancer Cell 22, 425–​37. Wisoff JH, et al. (2011) Primary neurosurgery for pediatric low-​grade
    Tatevossian RG, et al. (2010) MYB upregulation and genetic aberra- gliomas: a prospective multi-​institutional study from the Children’s
    tions in a subset of pediatric low-​grade gliomas. Acta Neuropathol Oncology Group. Neurosurgery 68, 1548–​54.
    120, 731–​43. Wu G, et al. (2012) Somatic histone H3 alterations in pediatric diffuse
    Vanan MI, Eisenstat DD (2014) Management of high-​grade gliomas in intrinsic pontine gliomas and non-​brainstem glioblastomas. Nat
    the pediatric patient: Past, present, and future. Neuro-​Oncol Pract Genet 44, 251–​3.
    1, 145–​57. Wu G, et al. (2014) The genomic landscape of diffuse intrinsic pontine
    Vogelbaum MA, Aghi MK (2015) Convection-​enhanced delivery for glioma and pediatric non-​brainstem high-​grade glioma. Nat Genet
    the treatment of glioblastoma. Neuro Oncol 17(Suppl 2), ii3–​ii8. 46, 444–​50.
    Weber RG, et al. (2007) Frequent loss of chromosome 9, homozygous Zhang J, et al. (2013) Whole-​genome sequencing identifies genetic al-
    CDKN2A/​ p14(ARF)/​ CDKN2B deletion and low TSC1 mRNA terations in pediatric low-​grade gliomas. Nat Genet 45, 602–​12.
    expression in pleomorphic xanthoastrocytomas. Oncogene 26, Zhang X, et al. (2016) Targeting immune checkpoints in malignant
    1088–​97. glioma. Oncotarget 8(4), 7157–​74.
    Willard VW, et al. (2015) Prospective longitudinal evaluation of Zuccaro G, et al. (1986) Intracranial neoplasms during the first year of
    emotional and behavioral functioning in pediatric patients with life. Surg Neurol 26, 29–​36.
    23
    Embryonal Tumours
    Maura Massimino, Eric Bouffet, and Vijay Ramaswamy

    Introduction and epidemiology paediatric brain tumours, but is predominantly diagnosed in chil-
    dren under the age of three. An additional difficulty in discerning
    Among the various tumours of the central nervous system (CNS) the incidence of ATRTs is that prior to the late 1990s, they were com-
    occurring in childhood, 20% are medulloblastomas (MB). The dis- monly diagnosed as MBs or CNS-​PNETs, as SMARCB1 loss as its
    ease has an annual incidence of 5–​6.8 per million children (aged defining feature had not been well established. ATRTs can be familial
    0–​14 years) in the USA and Europe. Higher rates are observed in in up to 25% due to germline inactivating lesions in SMARCB1.
    Southern and Central Europe. The disease has a higher incidence Until recently, there have been no dedicated data pertaining to ei-
    in boys. Embryonal tumours represented 23.2% of all cases seen at ther demographics or survival of ATRTs. The most recent issue of
    the Children’s Cancer Hospital in Egypt between 2007 and 2013, the Central Brain Tumour Registry of the United States (CBTRUS)
    similar to Europe and North America. Some genetic disorders that estimates five-​year survival to be 12.1% in those under one year of
    predispose patients to cancer, including Turcot (APC mutation), age and 33.9% in those over one year of age.
    Li–​Fraumeni (TP53 mutations), and Rubinstein–​Taybi (CREB/​BP
    mutation) syndromes, mismatch repair deficiency (PMS2, MLH1,
    MSH2, or MSH6 mutations), and abnormalities in the ‘sonic Diagnosis
    hedgehog’ (SHH) pathway (Gorlin syndrome with PTCH1 muta-
    tion, SUFU mutations), are associated with an increased risk of MB. The differential diagnoses of posterior fossa tumours include MB,
    The one-​, three-​, and five-​year survival rates among European cerebellar astrocytoma, ependymoma, brainstem glioma, and
    children with MB diagnosed from 2000 to 2007 were 81%, 63%, and ATRT. Computed tomography (CT) is often the first choice for
    56%, respectively. Survival rates remained much the same during neuroimaging because of its availability in an emergency setting.
    the period between 1999 and 2007, having improved before the end A typical picture of MB on CT is a midline, homogeneous, contrast-​
    of the 1990s. Very recent data from the American National Cancer enhancing cerebellar vermian mass. Follow-​on magnetic resonance
    Data Base indicate that five-​year overall survival (OS) is around 82% imaging (MRI) is needed before any surgery, and typically shows a
    for children with no metastases who are treated with post-​operative heterogeneous hypointense mass on T1-​weighted images. Contrast
    radiotherapy. enhancement is usually heterogeneous. Up to 40% of patients have
    The incidence and epidemiology of non-​MB and non-​atypical spinal metastases, best seen on post-​contrast images, making spinal
    teratoid rhabdoid (ATR) CNS embryonal tumours, known as MRI essential before any adjuvant treatment.
    ‘other CNS embryonal tumours’ (and previously as primitive It has been suggested recently that MRI features can distinguish
    neuroectodermal tumours (PNETs)) is extraordinarily difficult to between different histological subtypes and molecular subgroups of
    discern. Best estimates suggest they represent approximately 2–​3% MB: WNT tumours occur frequently in the cerebellar peduncle and
    of all paediatric brain tumours. Due to their heterogeneity and diag- cerebellar-​pontine angle, while SHH tumours arise in the cerebellar
    nostic uncertainty, their incidence is unclear, varying significantly hemispheres; and Group 3 and 4 tumours predominate within the
    by study. This has led to the removal of ‘other CNS embryonal tu- midline fourth ventricle. Midline Group 4 tumours also show little
    mours’ from the 2016 revised World Health Organization (WHO) or no enhancement.
    classification of CNS tumours. Moreover, entities comprising CNS On neuroimaging, the appearance of CNS-​PNET is highly vari-
    embryonal tumours vary tremendously in outcome, with ETMRs able, although usually displaying restricted diffusion (Figure 23.1).
    (embryonal tumours with multilayered rosettes) having a dismal Controversy has always existed as to whether CNS-​PNET truly
    prognosis, while survival in other CNS embryonal tumours ranges represents one entity, and the key to diagnosis is to exclude other
    from 17–​60%. morphological mimics. A study by the Children’s Oncology Group
    The incidence and epidemiology of atypical teratoid rhabdoid (COG) suggests that up to 50% of institutionally diagnosed CNS-​
    tumours (ATRTs) is also difficult to discern. In the USA, the best PNETs are high-​grade gliomas on centralized review, and an inte-
    estimate is 0.12 per 100,000 children, accounting for 2–​3% of all grated genomic study suggests that by using a combined molecular
    Diagnosis 189

    (a) ≥ 3–5 years (b) < 3–5 years

    Standard-risk patients: High-risk patients: Standard/low-risk patients: High-risk patients:


    Negative CSF + Positive CSF ± DMB/MBEN, even M+ Classic histology LC-A
    No macrometastases + Macrometastases + MYC amplified, even M0
    < 1.5 cm2 residual tumour > 1.5 cm2 residual tumour M1-M3
    LCA histology, MYC status
    Standard CT ± i.t CT

    Standard CT ± HD CT
    23.4*/54 Gy CSI + CT Standard/Higher doses/altered RT focal/CSI
    fractionation RT
    * for WNT subgroup, a ‘lower’ risk plus standard CT/HDCT
    can be recognized giving the
    possibility of de-escalating therapy if
    in controlled clinical trials

    Figure 23.1 Current clinical-​risk stratification employed by both European and North American Cooperative Groups. (A) Patients over the age
    of three to five are classified as standard risk if they have no evidence of residual disease either in the tumour bed or along the leptomeninges, and
    classified as high risk if they have residual disease. In some studies, large-​cell/​anaplastic histology and MYC amplification are also considered higher-​
    risk features. (B) Proposed risk stratification for children under the age of three to five in the upcoming International Society of Paediatric Oncology
    infant medulloblastoma study. Children with desmoplastic histology and/​or SHH-​driven tumours are considered standard risk, and receive lower doses
    of chemotherapy. Children with non-​desmoplastic histology and/​or non-​SHH-​driven tumours are classified as higher risk, and proceed to either high-​
    dose chemotherapy with autologous stem-​cell support and/​or external beam irradiation.
    Reproduced courtesy of Maura Massimino, IRCCS National Cancer Institute Foundation, Milan, Italy

    approach, up to 70% of CNS-​PNETs can be reclassified as a multi- and TP53-​mutated SHH. The prognostic of LC/​A in WNT and
    tude of different entities. Group 4 is unclear.
    ATRTs can occur anywhere within the nervous system, and on The revised 2016 WHO classification has removed CNS-​PNET
    neuroimaging can mimic other embryonal tumours and high-​grade as a distinct entity and replaced it with CNS neuroblastoma, CNS
    gliomas. Thus, as such, INI1 immunoreactivity should be deter- ganglioneuroblastoma, and CNS embryonal tumour not otherwise
    mined in all malignant infant brain tumours. The diagnosis of ATRT specified (NOS). ETMR is now accepted as a distinct entity defined
    should include renal ultrasounds to exclude extracranial rhabdoid by an amplification of a microRNA cluster on chromosome 19. ATRT
    tumours, and genetic counselling to determine germline SMARCB1 is described as mixed small blue-​cell tumour and rhabdoid cells,
    status. In rare instances, INI1 immunoreactivity is retained with ei- characterized by loss of INI1/​BAF47 immunoreactivity. A small
    ther functional loss of INI1 or mutations in SMARCA4. proportion of ATRTs have retained INI1/​BAF47 immunoreactivity,

    Pathology
    The recent WHO classification of CNS tumours acknowledges both
    the histological classification and the biological subtypes of MB.
    There are four histological variants: classic, desmoplastic/​nodular,
    MB with extensive nodularity (MBEN), and large-​cell/​anaplastic
    (LC/​A) (Figure 23.2).
    The combined LC/​A form accounts for 10–​22%, while nodular/​
    desmoplastic MB and MBEN represent approximately 7% and 3% of
    cases, respectively. Classic tumours make up the remainder. Clinical
    data firmly support a favourable prognosis for nodular/​desmoplastic
    MB in some age and risk groups, and especially in young children.
    The prognosis of the LC/​A variant is significantly worse. However,
    the high interobserver variability of these histological subtypes sig-
    nificantly impairs global interpretation and clinical applicability.
    Currently, the biological subtypes are divided into WNT-​activated,
    SHH-​activated TP53 wild-​type, SHH-​activated TP53 mutant, and
    non-​WNT/​SHH with Group 3 and Group 4 listed as provisional
    Figure 23.2 MRI of CNS embryonal tumour: axial T1-​enhanced image
    entities. There is overlap between the histological and biological
    of the entity (formerly termed CNS-​PNET) showing a heterogeneously
    classifications, where desmoplastic and MBEN tumours are almost enhancing mass in the left frontal lobe with associated cyst.
    exclusively SHH-​activated MBs, while the LC/​A variant is enriched Reproduced courtesy of Maura Massimino, IRCCS National Cancer Institute
    for very high-​risk biological groups, specifically high-​risk Group 3 Foundation, Milan, Italy
    190 CHAPTER 23 Embryonal Tumours

    with functional loss of the SMARCB1 gene or mutations in the and integrated clustering has suggested two distinct infant SHH
    SMARCA4 gene. groups, SHH1 and SHH2, with significant differences in outcome.
    PTCH1 and SUFU mutations can be germline, and all infants should
    Biology undergo genetic counselling for potential Gorlin syndrome. Adults
    with SHH medulloblastomas frequently harbour somatic mutations
    Medulloblastoma
    in PTCH1, TERT, and SMO, with an intermediate prognosis.
    Over the past ten years there have been considerable advances in SHH tumours most frequently recur locally, regardless of therapy.
    our knowledge of the biology of MB. Historically considered one Efforts are underway to incorporate specific SHH-​pathway antag-
    disease, it is now accepted that MB actually comprises four distinct onists into therapy. TP53-​mutated SHH MBs harbour lesions down-
    molecular subgroups termed WNT, SHH, Group 3, and Group 4. stream of SMO and are unlikely to respond to the current generation
    These subgroups have distinct demographics, cytogenetics, somatic of SHH inhibitors.
    nucleotide variants, and outcomes, and are now included in the 2016
    revised WHO classification of CNS tumours. The clinical utility of Group 3 medulloblastoma
    molecular subgrouping is apparent when applying risk-​stratification Group 3 tumours have the worst prognosis, are frequently metastatic,
    schemes, which are significantly more robust than standard-​risk and occur mainly in infants and young children. Approximately
    stratification when incorporating subgrouping, and will form the 20% of Group 3 tumours harbour high-​level amplifications of MYC,
    basis of future clinical trials. are frequently metastatic, and comprise a very high-​risk group.
    Histologically, Group 3 are frequently LCA/​anaplastic and are likely
    WNT medulloblastoma
    associated with MYC amplifications. Other aberrations in Group 3
    WNT tumours are unique in that they have an excellent prognosis include OTX2 amplifications, isochromosome 17q, gains of ACVR1/​
    across many studies, where survival consistently approaches 95% in 2, and activation of GFI1/​GFI1b through enhancer hijacking. Infant
    patients under the age of 16. Biologically, they are characterized by Group 3 patients pose a particularly difficult challenge. However, re-
    activation of the WNT pathway, resulting in nuclear translocation cent studies have suggested that survival for non-​metastatic Group 3
    of beta-​catenin. WNT tumours have a relatively quiet genome with can approach 40–​ 50% using high-​ dose chemotherapy (HDCT)
    mutations in exon 3 of the gene coding for b-​catenin (CTNNB1) with autologous stem-​ cell transplantation (ASCT), such as the
    being the most frequent aberration, followed by monosomy 6. Less CCG99703 protocol. Metastatic Group 3 have a dismal prognosis,
    frequently, TP53 and DDX3X mutations are observed. WNT tu- particularly in infants.
    mours can be diagnosed using a variety of methods including direct Recent drug screening has suggested pemetrexed and gemcitabine
    sequencing of exon 3 of CTNNB1, methylation profiling, or gene-​ as rationale therapies for Group 3 patients—​a combination currently
    expression profiling. Immunohistochemistry showing nuclear beta-​ being evaluated in the SJMB12 study. Panobinostat has also been
    catenin immunoreactivity is a hallmark of WNT tumours. However, proposed as a Group 3 specific therapy. Irradiated Group 3 patients
    it should be used with caution and in isolation, due to false positive almost exclusively recur with metastatic dissemination, usually with
    results. Similarly, identification of monosomy 6, while highly en- a tumour bed devoid of disease, suggesting that future therapies
    riched in WNT tumours, should be used with caution. need to be tailored to the metastatic compartment.
    Demographically, WNT tumours present in older children and a
    minority of adults. An international consensus suggests that childhood Group 4 medulloblastoma
    WNT tumours are now considered very low-​risk tumours. They seem Group 4 tumours comprise over 40% of all patients, and are pri-
    to have a worse prognosis in patients over the age of 16. Histologically, marily observed between the ages of three and seventeen. They are
    most WNT tumours are classic histology, although a small proportion usually of classic histology, although it is unclear what the signifi-
    can have LC/​A histology, which is of dubious clinical significance. cance of anaplasia is in this group. Group 4 are frequently metastatic,
    portending to a worse prognosis, although non-​metastatic Group 4
    SHH medulloblastoma patients with loss of chromosome 11 have an excellent prognosis and
    SHH tumours are characterized by lesions within the SHH pathway, are now considered very low-​risk patients. Nearly 80% of Group 4
    and are thought to arise from the external granular layer, hence their tumours harbour isochromosome 17q, with less frequent aberra-
    cerebellar location. They are observed across all ages, and are the tions being amplifications of CDK6, MYCN, chromosome 7q gain,
    predominant subgroup in infants and adults. The most common and chromosome 8p loss. KDM6A mutations are the most common
    aberrations are inactivating mutations of PTCH1, SMO, and SUFU recurrent SNV, occurring in 10% of patients, and tandem duplica-
    and amplifications of GLI2 and MYCN. TP53 mutations (usually tions of SNCAIP are found in 20% of Group 4 patients. Unlike SHH,
    germline mutations) are found in children aged 3–​16, and these pa- the presence of MYCN amplifications in Group 4 confers no prog-
    tients have a near universally fatal prognosis, succumbing to local re- nostic significance. Infant and adult Group 4, while rare, appear to
    currence or, less frequently, secondary malignancies. TP53-​mutated have a poor prognosis, and similar to Group 3, irradiated Group 4
    SHH usually occurs with significant genomic instability and high-​ patients recur almost exclusively with metastatic dissemination.
    level amplifications of GLI2 and MYCN.
    All SHH MBs in non-​infants warrant urgent testing for germline Other embryonal tumours
    TP53 mutations in the context of Li–​Fraumeni syndrome. Non-​ The first suggestion that PNET was not a single entity was a gene-​
    TP53-​mutated SHH MBs in older children confer an excellent expression study of medulloblastoma, ATRT, and PNET, in which
    prognosis. Infant SHH MBs frequently harbour mutations in PNETs did not form a clear cluster on their own. Further work sug-
    PTCH1 and SUFU, and display desmoplastic or MBEN histology, gested that a proportion of PNETs harbour high-​level amplifications
    Prognosis 191

    of a microRNA cluster on chromosome 19, and should be con- Table 23.1 Chang’s classification for metastases
    sidered a distinct entity known as ETMR/​ETANTR (embryonal tu-
    Stage Description
    mour with abundant neuropil and true rosettes). ETMRs arise in
    very young children and have a dismal prognosis. An integrated gen- M0 No gross nodular or laminar subarachnoid or haematogenous
    metastasis
    omic study of PNET showed that it comprises three groups, with one
    group harbouring the C19MC amplicon. Further work has shown M1 Microscopic tumour cells found in CSF
    that up to 70% of institutionally diagnosed PNET can be reclassi- M2 Gross nodular or laminar seeding in the cerebellum, cerebral
    subarachnoid space, or third or fourth ventricles
    fied as another entity after careful molecular analysis. As such, the
    diagnosis of PNET should include evaluation for RELA fusion for M3 Gross nodular or laminar seeding in the spinal subarachnoid
    space
    ependymoma, SMARCB1 alterations for ATRT, G34V and K27M
    mutations for high-​grade glioma, and C19MC amplicon for ETMR. M4 Extra-​neuraxial metastasis
    Genome-​wide methylation profiling is emerging as a powerful tool Adapted with permission from Chang C, et al. ‘An operative staging system and a
    in the evaluation of CNS embryonal tumours. Pineoblastoma repre- megavoltage radiotherapeutic technic for cerebellar medulloblastomas’. Radiology,
    Volume 93, Issue 6, pp.1351–​9. Copyright © 1969 Radiological Society of North
    sents a distinct entity, although little is known regarding its genomic America. DOI: https://​doi.org/​10.1148/​93.6.1351
    landscape. Germline DICER1 mutations are found in a small pro-
    portion of patients and warrant genetic testing for all patients.
    The small proportion of PNETs/​other embryonal tumours, which MB patients are stratified as standard-​or high-​risk for treatment
    could not be reclassified, harbour unique rearrangements com- purposes, depending on the clinical presentation, metastatic status
    prising four distinct groups. Alterations of FOXR2, CIC, MN1, (M1–​M4), and/​or residual disease of more than 1.5 cm2 on early post-​
    and BCOR comprise these new molecular entities provisionally operative MRI (within 24–​72 hours). The risk grouping of MB patients is
    designated ‘CNS neuroblastoma with FOXR2 activation (CNS NB-​ based on Chang’s classification for metastases (Table 23.1). Historically,
    FOXR2)’, ‘CNS Ewing sarcoma family tumour with CIC alteration patients with other embryonal tumours and ATRTs have been treated
    (CNS EFT-​CIC)’, ‘CNS high-​grade neuroepithelial tumour with on MB protocols and, due to their poorer survival, as high-​risk patients,
    MN1 alteration (CNS HGNET-​MN1)’, and ‘CNS high-​grade neuro- although dedicated ATRT studies have become more prevalent.
    epithelial tumour with BCOR alteration (CNS HGNET-​BCOR)’. Patients have generally been risk stratified by age, extent of residual
    disease, spread, LC/​A-​MB in some protocols, and (more recently) by
    The clinical significance of these alterations is still unknown and re-
    quires large multicentre collaboration. biological features such as MYC and WNT status (Figure 23.3).
    The standard-​risk group applies to 60–​70% of patients over three
    Atypical teratoid rhabdoid tumour years old. High-​risk patients are those with disseminated disease
    ATRTs are almost universally defined by a genetic lesion in and/​or less than a gross or near-​total resection (arbitrarily de-
    SMARCB1, either broad events resulting in deletions or point mu- fined as 1.5 cm2 or more residue). Histologically, LC/​A MB is con-
    tations, with a small proportion harbouring lesions in SMARCA4. sidered high-​risk. The value of staging in the context of molecular
    Recent integrated genomic studies have shown that ATRT com- subgrouping is unclear, particularly for WNT and Group 4 patients.
    prises three distinct molecular variants. These groups are termed In forthcoming trials, staging will be done by combining histological
    Group 1/​ATRT-​SHH, Group 2A/​ATRT-​T YR, and Group 2B/​ and biological variables. CNS embryonal tumour patients over the
    ATRT-​ MYC, defined by distinct gene expression and DNA-​ age of three receive similar staging to medulloblastoma patients;
    methylation patterns. Group 1/​ATRT-​SHH are more frequently however, the value of biological markers is currently unclear.
    supratentorial with focal alterations in SMARCB1, with a trend to-
    wards a more favourable prognosis. This group have a more closed
    chromatin pattern and activation of NOTCH and SHH path- Prognosis
    ways. Group 2A/​ATRT-​T YR are more frequently infratentorial,
    with a mix of broad and focal alterations in SMARCB1, and a With modern treatment protocols combining surgery with
    trend towards a poorer prognosis. This group have activation of craniospinal irradiation (CSI) and CT, five-​year OS for MB is now
    PDGFRB and overactivation of multiple receptor tyrosine kin- higher than 80% for standard-​risk patients. Until a few years ago,
    ases. Spinal tumours are enriched in group 2B/​ATRT-​MYC, with the outcome for metastatic MB was poor, with five-​year OS around
    broad alterations in SMARCB1 and a globally hypomethylated 30–​ 40%. Intensified CT regimens and high-​ dose RT schedules
    DNA-​methylation landscape. The clinical implications of these (including hyperfractionated accelerated radiotherapy—​ HART)
    subgroups are still being discerned, although clear trends are have considerably improved outcomes, achieving five-​year survival
    emerging from retrospective data. rates around 70%.
    Younger children (up to five years of age, depending on group pol-
    icies) have traditionally been treated with prolonged CT schedules
    Staging in an effort to contain late sequelae, especially of RT. More recently,
    using systemic CT combined with intraventricular methotrexate or
    MB, other embryonal tumour, and ATRT staging and risk stratifi- HDCT, five-​year progression-​free survival (PFS) has risen to more
    cation are crucial for management. Staging is based on MRI of the than 70%, showing that a tailored use of CT can make RT unneces-
    brain and spine, with and without gadolinium, and on cytology of sary, in some patient subgroups at least.
    cerebrospinal fluid (CSF) drawn from the lumbar region two weeks Other embryonal tumours have a variable prognosis. ETMR and
    after surgery (to avoid false positive findings). CNS embryonal tumours harbouring C19MC amplicons are near
    192 CHAPTER 23 Embryonal Tumours

    Classic Desmoplastic/Nodular

    Large cell/Anaplastic MBEN

    Figure 23.3 Histological classification of medulloblastoma: microscopic haematoxylin and eosin-​stained sections of medulloblastoma depicting the
    four histological variants in the WHO classification.
    Reproduced with permission from Maura Massimino, et al. ‘Childhood medulloblastoma’, Critical Reviews in Oncology/​Hematology, Volume 105, pp. 35–​51. © 2016 Elsevier
    Ireland Ltd. All rights reserved. DOI: https://​doi.org/​10.1016/​j.critrevonc.2016.05.012

    universally fatal. A major factor in the interpretation of studies of 3. Introducing any available targeted therapies in suitable
    other embryonal tumours is the inclusion of small-​cell high-​grade constellations;
    gliomas which have a worse outcome. However, a pineal location 4. Considering other radiation techniques, such as proton therapy,
    portends to a better outcome, with three-​to five-​year event-​free with a view to limiting late sequelae. This is particularly of
    survival (EFS) between 38% and 69%. Currently, there is a paucity interest in younger children who are at high risk to suffer long-​
    of studies enrolling other embryonal tumours, making it difficult to term side effects of cranial radiation.
    provide accurate risk stratification and prognosis.
    Surgery

    Treatment Surgical resection is fundamental to the treatment of MB. Though


    there is a substantial risk of morbidity, expert paediatric neuro-
    surgeons can usually achieve gross total resection without causing
    The aim of ongoing and future clinical trials is to improve the prog-
    major damage. Recent studies have shown that the value of extent
    nosis for MBs and CNS embryonal tumours, and to contain the
    of surgery in MB is less pronounced when accounting for molecular
    sequelae of treatments by:
    subgroups. A permanent CSF diversion will be necessary for about
    1. Testing the feasibility of reducing the dose of CSI and the volume 20–​30% of patients due to scarring of the CSF pathways. In addition
    of the posterior fossa boost in standard-​risk patients in the light to the risk of infections and mechanical complications (fluid leak,
    of the recently-​identified biological stratification factors; pseudomeningocele), surgery can lead to posterior fossa mutism
    2. Establishing whether intensifying CT or RT can improve out- syndrome developing 48–​72 hours after the surgical procedure. In
    come in high-​risk MB patients; around one third of cases, this condition is associated with severe
    Treatment 193

    cerebellar impairments, including dysmetria, hypotonia, paresis, None of the studies conducted to date have demonstrated that
    and depressed mood, which may persist for months. combinations of high-​dose CSI (be it hyperfractionated/​accelerated
    Surgical resection is also fundamental in the treatment of other or not) with HDCT can improve disease control. Wider, phase 3
    embryonal tumours. Although historically, a gross total resection is trials may soon begin to seek more evidence, but any effort to im-
    thought to confer a better prognosis, it fails to reach significance, prove the prognosis by intensifying treatments should meet accurate
    likely due to small patient numbers. The prognostic role of surgical quality assurance and feasibility criteria to avoid undue toxicities.
    resection in ATRTs is not clear. A pooled retrospective data analysis
    Infant medulloblastoma
    failed to show a clear benefit to gross total resection when correcting
    for HDCT and radiation; however, this has not been confirmed in a Infants have either SHH tumours with a better prognosis, or Group 3
    prospective study. tumours with a higher incidence of metastatic disease and a worse
    Early post-​operative brain MRI (within 48 hours after the pro- prognosis.
    cedure) should be obtained to assess the extent of the resection. Some studies have tried intensifying systemic CT, while others
    In the event of residue, patients whose tumour can be reduced by have added intraventricular CT or HDCT in an effort to avoid the
    means of a second surgical resection to no more than 1.5 cm2 are unacceptably severe sequelae seen in long-​term survivors treated
    eligible for standard-​ risk protocols—​ providing this approach is with CSI at a young age. High response rates have been reported after
    undertaken in good time (generally within two weeks after the first sequential HDCT, particularly in the CCG99703 protocol with three
    surgical procedure). tandem cycles of myeloablative carboplatin and thiotepa, which is
    being investigated in second-​generation studies. A French group has
    Non-​surgical treatment also reported that smaller volumes of RT after HDCT contributed
    to long-​term survival. Further research should ascertain whether
    Standard-​risk medulloblastoma
    these regimens can also increase the proportion of patients curable
    After surgical resection, the mainstay for patients aged three to five without RT in the M0/​T0 group, particularly when accounting for
    years or more is ‘reduced-​dose’ CSI, with a total dose of 23.4 Gy, plus a molecular subgrouping.
    boost to the tumour bed up to a total dose of 54–​55.8 Gy. Historically, Whether the subset of infants cured in each study had particular
    whole posterior fossa boosts were considered the standard; however, biological features that favoured survival remains to be seen. The
    reduced-​volume tumour-​bed boost spares a significant proportion HIT (Hirntumoren Säuglinge und Kleinkinder—​the German group
    of the occipital lobes from exposure to radiation without a reduction for brain tumour study in children) SKK 1992 study reported a
    in survival. This is often combined with weekly vincristine, and is higher frequency of desmoplastic MBs (40%), which carry a signifi-
    then followed by a multi-​drug platinum-​based regimen. With this cantly better prognosis than the classic variant. A meta-​analysis of
    approach, the five-​year EFS rate is over 80%. infant studies conducted in Europe and North America, between
    The clinical target volume (CTV) for CSI includes the whole brain 1987 and 2004, included 270 children under five years old and con-
    and spinal cord with the meninges, down to the thecal sac on level L5 firmed a significantly better EFS and OS for infants with nodular/​
    or S3. Modern protocols include frontline central review of RT plan- desmoplastic MB and patients with completely resected localized
    ning to assure the best possible quality control. An 18 Gy CSI dose disease (M0R0). This histological subtype is mainly associated with
    has been investigated in the randomized COG trial ACNS0331, and SHH activation and emerges as a major positive prognostic factor
    showed a higher rate of relapse in the reduced-​dose CSI arm. Still, in young children, whatever therapy they receive, whereas HDCT
    reduced-​dose CSI is being tried in a subset of standard-​risk patients seems to be needed for higher-​risk histotypes and clinical presen-
    under 16 years old with WNT-​MB in the SIOP-​PNET5, ACNS1422, tations. Every attempt should be made to avoid RT for infants with
    and SJMB12 trials. associated Gorlin syndrome or at risk of developing secondary tu-
    Another solution that might reduce the late sequelae of RT lies mours later in life. In conclusion, the last 10-​15 years have seen im-
    in proton therapy, but the follow-​up is very short in the experi- provements in the treatment for infant MB, tending towards less RT
    ences published to date, and only carefully selected cases have been and more CT. Recent evidence suggests that age per se is no longer an
    treated so far. Obviously, proton therapy may only spare normal adverse prognostic factor. The SIOPE is considering an international
    brain during the boost phase of the treatment, not during the CSI phase III trial for young children with MB, to compare survival rates
    phase. However, during the CSI phase, proton therapy may spare and neurocognitive outcomes achievable with different treatment
    other normal tissues like the oesophagus and intestine, which often strategies based on recognized prognostic factors (Figure 23.2).
    receive a higher dose with photon therapy.
    Treatment of other embryonal tumours
    High-​risk patients
    Post-​surgical treatment for other embryonal tumours has historic-
    Intensified cyclophosphamide-​based regimens, including HDCT ally followed that of MB. Results of these protocols have consistently
    with ASCT and high-​dose RT schedules, seem to have improved the shown inferior survival rates.
    prognosis for high-​risk patients, whose five-​year survival rates have
    risen from under 40% to 70%. The COG is currently assessing a ran- Results in infants
    domized treatment with concurrent carboplatin and vincristine plus Infants have classically been treated with a CT-​only approach.
    RT (versus vincristine and RT), followed by maintenance CT with or However, the results of early trials should be interpreted with cau-
    without 13-​cis retinoic acid, for newly diagnosed high-​risk MB (un- tion, as they certainly included a number of patients with ATRTs and
    published data). The 13-​cis retinoic-​acid arm was closed early due to high-​grade gliomas. Duffner reported the result of the first large,
    lack of efficacy in an interim analysis. infant, brain-​tumour collaborative study that included 17 patients
    194 CHAPTER 23 Embryonal Tumours

    with supraPNETs/​supratentorial CNS embryonal tumours and 11 using multidisciplinary management, with an approach similar to
    with pineoblastomas. Patients were treated with a four-​drug com- that used to treat rhabdomyosarcomas. The possibility to success-
    bination (vincristine–​cyclophosphamide/​cisplatin–​etoposide) until fully manage ATRT patients was further confirmed by Hilden and
    the age of three, and then received CSI. The five-​year EFS rate was colleagues in a report on 42 patients enrolled into the ATRT registry.
    27% for patients with sPNETs and 0% for those with pineoblastomas. Although 14 patients were alive in this report, the heterogeneity of
    Similar results were reported by the French and UK groups and the treatments administered precluded any conclusions regarding the
    Children’s Cancer Group (CCG). A subsequent study conducted by respective role of surgery, CT, and radiation. However, a simultan-
    the CCG used a more intensive CT approach without adjuvant radi- eous publication from the St Jude group described a high survival
    ation, and reported 8% five-​year PFS. rate in older children treated with CSI and chemotherapy, and em-
    The best results were reported by Fangusaro and colleagues, using phasized the role of early radiation. This approach was used by Chi
    consolidation with HDCT and ASCT. In this experience, EFS and and colleagues, who reported two-​year PFS and OS rates of 53%
    OS were 39% and 49%, with a better survival rate for non-​pineal and 70%, respectively, in a prospective multi-​institutional trial of 20
    sPNETs (57% versus 23%). Twelve patients did not receive any radi- evaluable patients treated with a chemotherapy backbone of the sar-
    ation. However, these results should be taken with caution, as there coma IRS-​III regimen.
    was no central pathology review in this trial. The HIT-​SKK group The results of the recently closed COG ACNS0333 protocol are
    analysed its experience of CT only followed by delayed radiation in awaited. This protocol used an intensive induction, followed by focal
    29 young children. They reported a dismal EFS and OS of 14.9% irradiation and HDCT with ASCT. Although RT is advocated by
    and 17.2%, respectively. Based on their findings, they recommended many, its role continues to be controversial, and some series report
    limiting any delay of RT to a maximum of six months. successful outcomes with a radiation-​sparing approach. There are
    Overall, the management of young children with CNS embryonal similar controversies regarding the respective role of anthracyclines,
    tumours remains a challenge. A benefit of HDCT has been sug- high-​dose methotrexate, HDCT, and intrathecal therapy. Whether
    gested in some series, and this approach has become standard in these questions should be addressed in future protocols, or whether
    many places. In CT-​only protocols, there is a lack of agreement re- the focus should be on tailored management of ATRTs based on their
    garding the role of focal irradiation following completion of therapy molecular subgroups, is a challenge for the years to come. The effi-
    in patients with no evidence of disease. cacy of molecularly targeted agents, such as alisertib, suggests that
    future protocols should not be limited to addressing chemotherapy
    Results in older children questions but, rather, should look for new drug opportunities.
    Children over the age of three to five are treated with CSI, with a
    boost to the primary tumour bed. Due to their relatively poor out-
    come, sPNET/​supratentorial CNS embryonal tumour patients have Follow-​up
    traditionally been treated as high-​risk patients with 36 Gy of CSI,
    even in the absence of metastatic dissemination. Currently, accrual Patients enrolled in study protocols undergo scheduled imaging, but
    for sPNET patients in the COG ACNS0332 was discontinued in after they have completed their treatment, the timing of their MRI
    2013 and there are no open North American or European studies. scans becomes more arbitrary. Imaging of the brain and spine gen-
    However, the CCG99701 study suggested PFS of almost 50% for erally occurs every three months for the first two years; then brain
    sPNETs treated with 36 Gy of CSI with a focal boost and concurrent MRI is every four months in the third year, six-​monthly up until the
    carboplatin as a radiosensitizer, followed by six cycles of cisplatin, fifth year, and yearly thereafter. Further imaging of the spine is usu-
    cyclophosphamide, and vincristine. A limited series has shown ally only warranted for suspect clinical signs. Late medulloblastoma
    encouraging survival rates using CSI followed by four cycles of HDCT relapses in previously irradiated patients should be approached with
    and ASCT. Interestingly, in this protocol, children with average-​risk caution; a significant proportion are radiation-​induced secondary
    features were treated with reduced-​dose CSI. Massimino reported malignancies, particularly in locally relapsing Group 4 patients.
    the result of a pilot study using HDCT and HART in 15 sPNET pa- Guidelines for patients with other embryonal tumours and ATRTs
    tients. Three-​year EFS and OS were 51% and 61%, respectively. follow the same rules. However, overall, patients with previously ir-
    Currently, there is no clear consensus, particularly in the context radiated embryonal tumours are extremely challenging, with dismal
    of recent advances in biology. However, the standard of care in most survival rates.
    centres remains CSI with a tumour-​bed boost, followed by cisplatin-​ The follow-​up should include all the appropriate clinical, radio-
    based maintenance CT. By contrast with the results observed in logical, and biochemical assessments, and tests tailored to a patient’s
    younger children, outcomes of pineoblastoma patients treated with neurofunctional capabilities (see the chapter ‘Late Effects of Therapy
    CSI are constantly better than those observed in other sPNETs, with and Survivorship Issues’).
    five-​year survival rates in the range of 60–​70% in patients without
    metastatic disease.
    Treatment at relapse
    Atypical teratoid rhabdoid tumour
    There are still a number of uncertainties concerning optimal man- MB relapses occur within two years of diagnosis in nearly 75% of
    agement of ATRT patients. Early reports stated that ATRTs rarely paediatric patients, and more than half of them have disseminated
    respond to CT, with an overall response rate of less than 20% and disease. Several factors influence how patients with relapsing MB are
    median survival of six months in an early report. In 1995, Olson treated. The initial treatment, and specifically whether the patient
    and colleagues described patients who showed sustained response was treated with radiation, is crucial when considering treatment for
    Future directions 195

    relapse. Infants who were treated with a radiation-​sparing approach consider therapy de-​escalation for low-​risk patients such as those
    comprise a distinct group where salvage is possible with CSI. This with WNT MB and low-​risk Group 4 tumours with chromosome 11
    principle guided the initial studies of deferred radiation; however, loss. As preclinical models of MB mature, new therapies are likely
    more recent studies attempted to avoid radiation altogether. CSI at to emerge for high-​risk patients which can be rapidly incorporated
    relapse, in this context, can result in survival approaching 60%, al- into clinical trials.
    though the optimal dose of CSI remains unknown. Although still in the conceptual stages, the next generation of
    Older children who relapse after CSI cannot be cured by salvage ATRT studies will also likely be molecularly informed. More im-
    therapy, with some very rare exceptions (less than 5%). Further sur- portantly, the discovery of discrete biological subsets of ATRT may
    gery, followed by focal (re-​)irradiation with or without CT, might help inform novel therapies. Unfortunately, other embryonal tu-
    be an option for solitary recurrences, but should be considered case mours are a group of rare tumours that have no open clinical trials.
    by case, and only after complete staging to ascertain the extent of As such, there is a paucity of existing and ongoing prospective clin-
    dissemination. ical data, which will make the design of any future clinical trial ex-
    Many phase II trials have reported a poor response in nearly all ceedingly difficult. International collaborations and pooled analyses
    recurrent MB patients. Several groups are investigating low-​dose will be required to determine the optimal therapy for the subsets of
    CT regimens called ‘metronomic’ therapy. It has proven feasible other embryonal tumours, to plan any future molecularly informed
    for treating paediatric brain tumours in case series, though it raises studies.
    concerns about haematological toxicities and the risk of secondary
    malignancies. Furthermore, Group 4 patients relapse later and have
    significantly longer survival post recurrence, and, as such, should be FURTHER READING
    taken into account when interpreting Phase II trials which suggest Bakst RL, et al. (2011) Reirradiation for recurrent medulloblastoma.
    activity based on a small proportion of short-​term survivors. Cancer 117, 4977–​82.
    There are ‘druggable’ mutations and copy number variations Bergthold G, et al. (2014) High-​dose busulfan-​thiotepa with autolo-
    in the genomic landscape of MB, but the only targeted therapy to gous stem cell transplantation followed by posterior fossa irradi-
    ation in young children with classical or incompletely resected
    have been examined so far in the clinical setting involved the use
    medulloblastoma. Pediatr Blood Cancer 61, 907–​12.
    of smoothened (SMO) inhibitors. These inhibitors are only effective
    Biegel JA, et al. (1999) Germ-​line and acquired mutations of INI1 in
    in the case of PTCH or SMO mutations (more common in adult
    atypical teratoid and rhabdoid tumours. Cancer Res 59, 74–​9.
    MB), while they are useless when mutations downstream, including Broniscer A, et al. (2004) High-​dose chemotherapy with autologous
    SUFU, GLI2, or MYCN, are responsible for SHH-​pathway activation stem-​cell rescue in the treatment of patients with recurrent non-​
    (as in childhood tumours), and any response to SMO inhibitors is cerebellar primitive neuroectodermal tumours. Pediatr Blood
    transient due to downstream pathway activation. In addition to their Cancer 42, 261–​7.
    limited efficacy, their use in children has to be balanced against the Chi SN, et al. (2009) Intensive multimodality treatment for children
    risk of their long-​term toxicity. with newly diagnosed CNS atypical teratoid rhabdoid tumour. J
    The pattern of relapse in MB is highly subgroup-​dependent, with Clin Oncol 27, 385–​9.
    SHH patients recurring most frequently locally, and Group 3 and 4 Chintagumpala M, et al. (2009) A pilot study of risk-​adapted radio-
    patients (particularly those who were irradiated) recurring almost therapy and chemotherapy in patients with supratentorial PNET.
    exclusively with metastatic dissemination. As such, future therapies Neuro Oncol 11, 33–​40.
    for Group 3 and 4 MB will undoubtedly need to include new ther- Clifford SC, et al. (2015) Biomarker-​driven stratification of disease-​
    risk in non-​metastatic medulloblastoma: results from the multi-​
    apies directed at the metastatic compartment. Moreover, it has been
    center HIT-​SIOP-​PNET4 clinical trial. Oncotarget 6(36), 38827–​39.
    recently shown that the relapse sample is highly divergent from the
    Cohen BH, et al. (2015) Pilot study of intensive chemotherapy with
    diagnostic sample, and biopsy at the time of relapse should be con-
    peripheral hematopoietic cell support for children less than 3 years
    sidered if targeted therapies are to be employed. of age with malignant brain tumours, the CCG-​99703 phase I/​
    There is no consensus on the treatment of other embryonal II study. A report from the Children’s Oncology Group. Pediatr
    tumours or ATRTs at relapse, except for infants who were not Neurol 53, 31–​46.
    irradiated, who may benefit from salvage radiation. The recent iden- Cohen BH, et al. (1995) Prognostic factors and treatment results for
    tification of novel groups of embryonal brain tumours and ATRTs supratentorial primitive neuroectodermal tumours in children
    with distinct molecular alterations may soon provide potential sal- using radiation and chemotherapy: a Children’s Cancer Group ran-
    vage options for these patients. domized trial. J Clin Oncol 13, 1687–​96.
    de Kock L, et al. (2014) Germ-​line and somatic DICER1 mutations in
    pineoblastoma. Acta Neuropathol 128, 583–​95.
    Future directions Duffner PK, et al. (1993) Postoperative chemotherapy and delayed ra-
    diation in children less than three years of age with malignant brain
    tumours. N Engl J Med 328, 1725–​31.
    The advent of integrated genomics has significantly changed our ap-
    Dufour C, et al. (2014) Tandem high-​dose chemotherapy and autolo-
    proach to MB, and has the potential to influence future therapies
    gous stem cell rescue in children with newly diagnosed high-​risk
    for ATRTs and other embryonal tumours. The current generation of
    medulloblastoma or supratentorial primitive neuro-​ectodermic tu-
    MB studies are more molecularly informed, and the next generation mours. Pediatr Blood Cancer 61, 1398–​402.
    of trials in MB will incorporate more robust molecular risk stratifi- Eberhart CG, Burger PC (2003) Anaplasia and grading in
    cation and introduce novel agents for very high-​risk subsets of pa- medulloblastomas. Brain Pathol 13, 376–​85.
    tients. Moreover, the current and next generation of MB studies will
    196 CHAPTER 23 Embryonal Tumours

    Ezzat S, et al. (2016) Pediatric brain tumours in a low/​middle income Kool M, et al. (2014) Genome sequencing of SHH medulloblastoma
    country: does it differ from that in developed world? J Neurooncol predicts genotype-​ related response to smoothened inhibition.
    126, 371–​6. Cancer Cell 25, 393–​405.
    Fangusaro J, et al. (2008) Intensive chemotherapy followed by Korshunov A, et al. (2014) Embryonal tumour with abundant
    consolidative myeloablative chemotherapy with autologous hem- neuropil and true rosettes (ETANTR), ependymoblastoma, and
    atopoietic cell rescue (AuHCR) in young children with newly medulloepithelioma share molecular similarity and comprise a
    diagnosed supratentorial primitive neuroectodermal tumours single clinicopathological entity. Acta Neuropathol 128, 279–​89.
    (sPNETs): report of the Head Start I and II experience. Pediatr Lafay-​Cousin L, et al. (2016) Clinical, pathological, and molecular
    Blood Cancer 50, 312–​18. characterization of infant medulloblastomas treated with sequen-
    Friedrich C, et al. (2013) Treatment of young children with CNS-​ tial high-​dose chemotherapy. Pediatr Blood Cancer 63(9), 1527–​34.
    primitive neuroectodermal tumours/​pineoblastomas in the pro- Louis DN, et al. (2016) The 2016 World Health Organization classifi-
    spective multicenter trial HIT 2000 using different chemotherapy cation of tumours of the central nervous system: a summary. Acta
    regimens and radiotherapy. Neuro Oncol 15, 224–​34. Neuropathol 131, 803–​20.
    Gajjar A, et al. (2006) Risk-​adapted craniospinal radiotherapy followed Marec-​Berard P, et al. (2002) Supratentorial embryonal tumours in
    by high-​dose chemotherapy and stem-​cell rescue in children with children under 5 years of age: an SFOP study of treatment with
    newly diagnosed medulloblastoma (St Jude Medulloblastoma-​ postoperative chemotherapy alone. Med Pediatr Oncol 38, 83–​90.
    96): long-​term results from a prospective, multicentre trial. Lancet Massimino M, et al. (2009) No salvage using high-​dose chemotherapy
    Oncol 7, 813–​20. plus/​minus reirradiation for relapsing previously irradiated
    Gandola L, et al. (2008) Hyperfractionated accelerated radiotherapy medulloblastoma. Radiat Oncol Biol 73, 1358–​63.
    in the Milan strategy for metastatic medulloblastoma. J Clin Oncol Massimino M, et al. (2006) Supratentorial primitive neuroectodermal
    27, 566–​71. tumours (S-​PNET) in children: a prospective experience with ad-
    Garre ML, et al. (2009) Medulloblastoma variants: age-​dependent oc- juvant intensive chemotherapy and hyperfractionated accelerated
    currence and relation to Gorlin syndrome—​a new clinical perspec- radiotherapy. Int J Radiat Oncol Biol Phys 64, 1031–​7.
    tive. Clin Cancer Res 15, 2463–​71. Morrissy AS, et al. (2016) Divergent clonal selection dominates
    Gatta G, et al. (2014) Childhood cancer survival in Europe 1999–​ medulloblastoma at recurrence. Nature 529, 351–​7.
    2007: results of EUROCARE-​5—​a population-​based study. Lancet Muller K, et al. (2014) Postponed is not canceled: role of craniospinal
    Oncol 15, 35–​47. radiation therapy in the management of recurrent infant
    Gerber NU, et al. (2014) Treatment of children with central nervous medulloblastoma—​an experience from the HIT-​REZ 1997 & 2005
    system primitive neuroectodermal tumours/​pinealoblastomas in studies. Int J Radiat Oncol Biol Phys 88, 1019–​24.
    the prospective multicentric trial HIT 2000 using hyperfractionated Northcott PA, et al. (2012) Medulloblastomics: the end of the begin-
    radiation therapy followed by maintenance chemotherapy. Int J ning. Nat Rev Cancer 12, 818–​34.
    Radiat Oncol Biol Phys 89, 863–​71. Northcott PA, et al. (2014) Enhancer hijacking activates GFI1 family
    Geyer JR, et al. (2005) Multiagent chemotherapy and deferred radio- oncogenes in medulloblastoma. Nature 511, 428–​34.
    therapy in infants with malignant brain tumours: a report from the Olson TA, et al. (1995) Successful treatment of disseminated central
    Children’s Cancer Group. J Clin Oncol 23, 7621–​31. nervous system malignant rhabdoid tumour. J Pediatr Hematol
    Grill J, et al. (2005) Treatment of medulloblastoma with postoperative Oncol 17, 71–​5.
    chemotherapy alone: an SFOP prospective trial in young children. Ostrom QT, et al. (2015) Alex’s Lemonade Stand Foundation Infant
    Lancet Oncol 6, 573–​80. and Childhood Primary Brain and Central Nervous System
    Grundy RG, et al. (2010) Primary postoperative chemotherapy Tumours Diagnosed in the United States in 2007–​2011. Neuro
    without radiotherapy for treatment of brain tumours other Oncol 16 (Suppl 10), x1–​x36.
    than ependymoma in children under 3 years: results of the first Packer RJ, et al. (2006) Phase III study of craniospinal radiation therapy
    UKCCSG/​SIOP CNS 9204 trial. Eur J Cancer 46, 120–​33. followed by adjuvant chemotherapy for newly diagnosed average-​
    Hasselblatt M, et al. (2014) SMARCA4-​mutated atypical teratoid/​ risk medulloblastoma. J Clin Oncol 24, 4202–​8.
    rhabdoid tumours are associated with inherited germline alter- Pei Y, et al. (2016) HDAC and PI3K antagonists cooperate to inhibit
    ations and poor prognosis. Acta Neuropathol 128, 453–​6. growth of MYC-​driven medulloblastoma. Cancer Cell 29, 311–​23.
    Hilden JM, et al. (2004) Central nervous system atypical teratoid/​ Peris-​
    Bonet R, et al. (2006) Childhood central nervous system
    rhabdoid tumour: results of therapy in children enrolled in a tumours—​incidence and survival in Europe (1978–​1997): report
    registry. J Clin Oncol 22, 2877–​84. from Automated Childhood Cancer Information System project.
    Jakacki RI, et al. (2015) Outcome and prognostic factors for children Eur J Cancer 42, 2064–​80.
    with supratentorial primitive neuroectodermal tumours treated Perreault S, et al. (2014) MRI surrogates for molecular subgroups of
    with carboplatin during radiotherapy: a report from the Children’s medulloblastoma. Am J Neuroradiol 35, 1263–​9.
    Oncology Group. Pediatr Blood Cancer 62, 776–​83. Picard D, et al. (2012) Markers of survival and metastatic potential in
    Jakacki RI, et al. (2012) Outcome of children with metastatic childhood CNS primitive neuro-​ectodermal brain tumours: an in-
    medulloblastoma treated with carboplatin during craniospinal tegrative genomic analysis. Lancet Oncol 13, 838–​48.
    radiotherapy: a Children’s Oncology Group phase I/​II study. J Clin Pietsch T, et al. (2014) Prognostic significance of clinical, histopatho-
    Oncol 30, 2648–​53. logical, and molecular characteristics of medulloblastomas in
    Johann PD, et al. (2016) Atypical teratoid/​rhabdoid tumours are com- the prospective HIT2000 multicenter clinical trial cohort. Acta
    prised of three epigenetic subgroups with distinct enhancer land- Neuropathol 128, 137–​49.
    scapes. Cancer Cell 29, 379–​93. Pizer B, et al. (2011) Treatment of recurrent central nervous system
    Kann BH, et al. (2016) Postoperative radiotherapy patterns of care primitive neuroectodermal tumours in children and adoles-
    and survival implications for medulloblastoma in young children. cents: results of a Children’s Cancer and Leukaemia Group study.
    JAMA Oncol 2, 1574–​81. Eur J Cancer 47, 1389–​97.
    Future directions 197

    Pomeroy SL, et al. (2002) Prediction of central nervous system embry- Smith MJ, et al. (2014) Germline mutations in SUFU cause Gorlin
    onal tumour outcome based on gene expression. Nature 415, 436–​42. syndrome-​associated childhood medulloblastoma and redefine the
    Ramaswamy V, et al. (2011) FISH and chips: the recipe for improved risk associated with PTCH1 mutations. J Clin Oncol 32, 4155–​61.
    prognostication and outcomes for children with medulloblastoma. Spence T, et al. (2014) A novel C19MC amplified cell line links Lin28/​
    Cancer Genet 204, 577–​88. let-​7 to mTOR signaling in embryonal tumour with multilayered
    Ramaswamy V, et al. (2016) Medulloblastoma subgroup-​specific out- rosettes. Neuro Oncol 16, 62–​71.
    comes in irradiated children: who are the true high-​risk patients? Sturm D, et al. (2016) New brain tumour entities emerge from mo-
    Neuro Oncol 18, 291–​7. lecular classification of CNS-​PNETs. Cell 164, 1060–​72.
    Ramaswamy V, et al. (2016) Risk stratification of childhood Taylor MD, et al. (2012) Molecular subgroups of medulloblastoma: the
    medulloblastoma in the molecular era: the current consensus. Acta current consensus. Acta Neuropathol 123, 465–​72.
    Neuropathol 131, 821–​31. Tekautz TM, et al. (2005) Atypical teratoid/​ rhabdoid tumours
    Ramaswamy V, et al. (2013) Recurrence patterns across (ATRT): improved survival in children 3 years of age and older with
    medulloblastoma subgroups: an integrated clinical and molecular radiation therapy and high-​dose alkylator-​based chemotherapy. J
    analysis. Lancet Oncol 14, 1200–​207. Clin Oncol 23, 1491–​9.
    Raybaud C, et al. (2015) Posterior fossa tumours in children: devel- Thompson EM, et al. (2016) Prognostic value of medulloblastoma ex-
    opmental anatomy and diagnostic imaging. Childs Nerv Syst 31, tent of resection after accounting for molecular subgroup: a retro-
    1661–​76. spective integrated clinical and molecular analysis. Lancet Oncol
    Remke M, et al. (2011) Adult medulloblastoma comprises three major 17(4), 484–​95.
    molecular variants. J Clin Oncol 29, 2717–​23. Timmermann B, et al. (2002) Role of radiotherapy in the treatment
    Remke M, et al. (2013) TERT promoter mutations are highly recurrent of supratentorial primitive neuroectodermal tumours in child-
    in SHH subgroup medulloblastoma. Acta Neuropathol 126, 917–​29. hood: results of the prospective German brain tumour trials HIT
    Rorke LB, et al. (1996) Central nervous system atypical teratoid/​ 88/​89 and 91. J Clin Oncol 20, 842–​9.
    rhabdoid tumours of infancy and childhood: definition of an entity. Torchia J, et al. (2016) Integrated (epi)-​genomic analyses identify
    J Neurosurg 85, 56–​65. subgroup-​specific therapeutic targets in CNS rhabdoid tumours.
    Rudin CM, et al. (2009) Treatment of medulloblastoma with hedgehog Cancer Cell 30, 891–​908.
    pathway inhibitor GDC-​0449. N Engl J Med 361, 1173–​8. Torchia J, et al. (2015) Molecular subgroups of atypical teratoid
    Rutkowski S, et al. (2005) Treatment of early childhood rhabdoid tumours in children: an integrated genomic and
    medulloblastoma by postoperative chemotherapy alone. N Engl J clinicopathological analysis. Lancet Oncol 16, 569–​82.
    Med 352, 978–​86. Villani A, et al. (2012) Syndromes predisposing to pediatric central
    Rutkowski S, et al. (2010) Medulloblastoma in young children. Pediatr nervous system tumours: lessons learned and new promises. Curr
    Blood Cancer 54, 635–​7. Neurol Neurosci Rep 12, 153–​64.
    Rutkowski S, et al. (2010) Survival and prognostic factors of early Wetmore C, et al. (2015) Alisertib is active as single agent in recur-
    childhood medulloblastoma: an international meta-​analysis. J Clin rent atypical teratoid rhabdoid tumours in 4 children. Neuro Oncol
    Oncol 28, 4961–​68. 17, 882–​8.
    Schrey D, et al. (2016) Multimodal therapy in children and adolescents Wetmore C, et al. (2014) Reirradiation of recurrent medullo-
    with newly diagnosed atypical teratoid/​rhabdoid tumour: indi- blastoma: does clinical benefit outweigh risk for toxicity? Cancer
    vidual pooled data analysis and review of the literature. J Neuro-​ 120, 3731–​7.
    Oncol 126, 81–​90. Yock TI, et al. (2016) Long-​term toxic effects of proton radiotherapy
    Shih DJ, et al. (2014) Cytogenetic prognostication within for paediatric medulloblastoma: a phase 2 single-​arm study. Lancet
    medulloblastoma subgroups. J Clin Oncol 32, 886–​96. Oncol 17(3), 287–​98.
    24
    Other Central Nervous System Tumours
    of Childhood
    Cecile Faure Conter, Didier Frappaz, Kristian W. Pajtler, and Stefan M. Pfister

    Germ-​cell tumours positivity on IHC. Teratomas are either mature, with the presence of
    mature structures derived from all three germ layers (e.g. hair, squa-
    mous and mucous epithelium, teeth, cartilage, bone, glial or thyroid
    Diagnosis: radiology, (molecular) pathology tissue), or immature, with varying components of immature tissue
    All germ-​cell tumours (GCT), whether intra-​or extracranial, derive that mostly resemble immature neural tissue with neurotubular
    from the primordial germ cell. During early embryogenesis, prim- structures. GCT are either pure or mixed. In the latter case, each com-
    ordial germ cells migrate along the dorsal mesentery to populate ponent has its specific sensitivity to treatment, germinomas being
    the gonads. Aberrant migration may occur axially (i.e. suprasellar, exquisitely sensitive, teratomas resistant to radio-​ chemotherapy
    pineal regions with 10% being bifocal). Intracranial GCT are pure or (indicating the need for surgical removal), and other components
    mixed tumours including undifferentiated (germinoma, embryonal being somewhere in between.
    carcinoma) and/​or extraembryonic (choriocarcinoma, yolk sac) Intracranial GCT have a distinctive age distribution: teratomas
    and/​or somatic (teratomas) differentiating components. and pure YSTs develop during infancy and childhood and show im-
    In the pineal location, Parinaud’s syndrome and/​or raised intra- balances at chromosome 1 (loss at 1p, gain at 1q), whereas all others
    cranial pressure is recognized promptly. In the opposing suprasellar have a peak incidence during adolescence and early adulthood and
    location, symptoms—​ diabetes insipidus, impairment of visual are characterized by isochromosome 12p or 12p amplification.
    fields, or variation in onset of puberty—​are often overlooked for There is a male predominance, especially for germinomas.
    long periods in adolescents due to the natural changes occurring Pineal tumours may bulge in the third and/​or fourth ventricle
    within the body and mood. Basal ganglia germinomas usually de- and invade the thalamic region. Suprasellar masses involve the pi-
    velop in Asiatic patients following a longstanding history of focal tuitary stalk and third ventricle, and may compress or invade the
    deficit and/​or mental deterioration. optic chiasm. Primary tumours tend to propagate along the ven-
    Pathology is not always available. It may include one or several of tricular walls often causing endocrinological disorders even when
    the following components that may not all appear in a limited bi- they are not radiologically detectable. Disappearance of the spon-
    opsy. Germinomas, which represent the most frequent histology, are taneously bright spot of the posterior pituitary on T1 is correlated to
    composed of monotonous undifferentiated cells with some isolated diabetes insipidus. Germinomas are iso-​to hyperdense, sometimes
    syncitiotrophoblastic cells (explaining why germinomas may secrete calcified on computer tomography (CT) scan, iso-​or hypointense
    ß-​human chorionic gonadotropin (HCG)) intermixed with an in- on T1-​and T2-​weighted magnetic resonance imaging (MRI), with
    flammatory infiltrate. Immunohistochemistry (IHC) stains positive moderate to marked gadolinium enhancement without necrosis,
    for placenta-​like alkaline phosphatase (PLAP). The other compo- a drop in apparent diffusion coefficient (ADC) values is classically
    nents may show aberrant differentiation toward extraembryonic observed. The diffusion-​weighted imaging (DWI) is similar to that
    and/​or somatic features: yolk sac tumours (YST) show a lace-​like, of Langerhans cell histiocytosis and lymphocytic hypophysitis, but
    papillary, or cord-​ like pattern of cuboidal/​ elongated cells; 50% differs from craniopharyngiomas and gliomas. Bifocal tumours (if
    of tumours have Schiller–​Duval bodies; IHC is positive for alpha pineal and suprasellar) are a specific entity, not a metastatic disease
    fetoprotein (AFP) and glypican 3; choriocarcinomas (CC) are com- (Figure 24.1), and are typical of GCT, though not pathognomonic.
    posed of mononucleated (cytotrophoblastic) and multinucleated Teratomas may be suggested when the three compartments (paren-
    (syncitiotrophoblastic) cells that stain on IHC for HCG; and em- chymal, fat, and calcifications) are present. Basal ganglia germinoma
    bryonal carcinomas (EC) are composed of primitive epithelial-​type MRI shows atrophy of the cerebral peduncle and/​or basal ganglia
    cells with minimal features of differentiation arranged as solid, and/​or cerebral hemisphere. The contrast enhancement is incon-
    pseudo glandular, alveolar, tubular, or papillary patterns with CD30 stant and patchy. Spinal MRI without and with contrast will look
    Germ-cell tumours 199

    As all Japanese patients have histological documentation, the


    first Japanese study promoted a three-​ category strategy. Pure
    germinomas reached ten-​year overall survival (OS) of 97% when
    treated with three courses of etoposide and carboplatin concomitant
    with whole ventricular irradiation (WVI) of 24 Gy. The intermediate
    group (HCG-​secreting germinomas and mixed tumours composed
    of germinomas and immature teratomas) received similar treat-
    ment, with an increased dose to the ventricles (30 Gy) and a local
    boost (20 Gy); their ten-​year OS was 89%. The high-​risk group
    (non germinomatous tumours) received three courses of platinum–​
    etoposide–​ifosfamide concomitant with whole-​brain irradiation (30
    Gy) with a local boost (30 Gy) and spinal (24 Gy) radiotherapy; their
    ten-​year OS was 58%. In the ongoing study, strategy for germinomas
    (including HCG-​secreting tumours) remains the same. Those in the
    intermediate group receive similar chemotherapy with a decreased
    dose to the ventricles (23.4 Gy) and an increased boost to the tu-
    mour bed (27 Gy). Those in the high risk group receive a 30.6 Gy
    CSI plus 30.6 Gy additional boost to the tumour with concomitant
    Figure 24.1 Bifocal germinoma on T1-​weighted MRI showing and maintenance courses of platinum–​etoposide–​ifosfamide. Basal
    gadolinium enhancement. ganglia primaries with good or intermediate prognosis have a poor
    Reproduced courtesy of Cécile Faure-Conter, Institute of Hematology and Pediatric outcome and receive chemotherapy followed by a 27 Gy whole-​
    Oncology (iHOPe), Lyon, France
    brain irradiation. Those in the intermediate group receive similar
    chemotherapy but with a decreased dose to the ventricles (23 Gy)
    for nodules and/​or linear enhancement. It should be performed pre- and an increased boost (30 Gy). Those in the high risk group re-
    operatively to avoid bloody artefacts. If not, and if doubtful, it should ceived eight courses of platinum–​etoposide–​ifosfamide and a 30 Gy
    be repeated with both sagittal and axial sequences. CSI with 30 Gy additional boost. Basal ganglia primaries with good
    or intermediate prognosis have a poor outcome and receive a 27 Gy
    Biology whole-​brain irradiation.
    Certain markers are measurable in serum and cerebrospinal fluid American and European groups do not advocate front-​line bi-
    (CSF) that can be used in diagnosis and monitoring: AFP is mainly opsy in secreting tumours (i.e. non germinoma) and thus use a two-​
    produced by YST, though low levels may be detected from EC and category strategy: germinomas and non-​germinomas.
    immature teratomas. Its half-​life is five to seven days. The high level Localized germinomas were treated within the SIOP (International
    at birth falls to adult limits at 12 months of age. During this period, Society of Paediatric Oncology) GCT 96 protocol with either CSI
    an isolated measure should prompt further sampling at one-​week of 24 Gy with a local boost of 16 Gy, or combined treatment with
    intervals. Total HCG is produced by the placenta and is secreted by etoposide–​carboplatin–​ifosfamide (CarboPEI) followed by a 40-​Gy
    CC and at lower levels by germinomas and EC. Its half-​life is two focal boost. Due to potential neurotoxicity in the former, and excess
    days. Total HCG measures both HCG and subunit β and is expressed of ventricular relapses in the latter, the current CGTII SIOP protocol
    as mUI/​ml, while free-​chain HCG only measures free subunit β. explores the same induction chemotherapy followed by WVI of 24
    Positive cytology of CSF is found in 10% of cases. Gy with a 16-​Gy boost, delivered exclusively in patients who do not
    reach complete remission (CR) prior to radiation.The Children’s
    Current treatment strategies Oncology Group (COG) launched ACNS 1123 that explores com-
    The initial strategy requires a multidisciplinary approach. The initial bined treatment by etoposide–​carboplatin, followed by 18 Gy of
    workflow includes craniospinal MRI and determination of markers WVI with a 12-​Gy boost in patients reaching CR, or 24 Gy and 12
    (AFP and HCG) both in blood and CSF, and cytology of CSF. The in- Gy respectively otherwise.
    cidence of relapses increased from 4% in fully staged to 24% in par- Localized non-​germinomas were treated in the two successive
    tially staged germinoma patients. Apart from the Japanese strategy SIOP trials (GCT96 and SIOPGCTII) with ifosfamide–​etoposide–​
    (explained later), if markers (serum and CSF) are positive, histo- cisplatin (PEI) followed by a 54-​Gy boost to the primary. The
    logical confirmation is not required. Moreover, for non-​secreting GCT96 cohort reported a poorer prognosis for patients with AFP
    bifocal tumours (most likely germinomas) biopsy may be skipped. more than 1,000 and/​or with persisting viable tumour after chemo-
    In the remaining cases, open surgery, stereotactic biopsy, or endo- therapy. Hence, GCTII explores the role of high-​dose therapy fol-
    scopic third ventriculostomy may be used to obtain histological lowed by stem-​cell rescue, prior to radiation, in high-​risk patients.
    proof. The latter is often used to relieve hydrocephalus and allows ACNS 0122 delivered etoposide–​carboplatin–​ifosfamide with 36
    inspection of the ventricles. In ongoing protocols, detection of ven- Gy of CSI and a 18-​Gy boost for patients in CR, and consolidation
    tricular nodules during this procedure does not qualify the patient by high-​dose etoposide thiotepa with stem-​cell rescue followed by
    as metastatic, provided they are not detectable on MRI, as it is con- CSI otherwise. Five-​year OS was 93%. The current ACNS 1123 for
    sidered that ventriculoscopy is a recent procedure that deserves pro- non-​germinomas explores the feasibility of reducing the field to
    spective evaluation. WVI. The overall prognosis for patients with recurrent germinomas
    200 CHAPTER 24 Other Central Nervous System Tumours

    remains excellent. Non-​germinomatous GCT that recur may be sal- aggressive with higher rate of recurrences than papillary tumours
    vaged, although the long-​term outcome is worse. that occur primarily in adults, and have a well-​differentiated squa-
    mous epithelium with picket-​like palisades and wet keratin.
    Future outlook The diagnosis is often overlooked for years, possibly due both to
    Biology of intracranial (non-​)germinomas is lagging behind due to public and medical ignorance of the non-​specific signs of intracra-
    the lack of availability of tumour tissue in most chemo-​naive pa- nial pressure. Compression of optic nerves and/​or chiasm induces
    tients. Intracranial and extracranial tumours have similar mutations loss of visual field and/​or visual acuity in at least half of the cases.
    in KIT (26%: mostly exon 17), KRAS (15%), NRAS (5%; exclusive Most present with at least one hormonal deficit of growth hormone
    from KIT), mTor (8%), CBL (11%), with copy gain of AKT1 (19%). (GH: 75%), gonadotropins (40%), adrenocorticotropic hormone
    Potentially effective drugs in the future include sunitinib and mTor (ACTH: 25%), thyroid-​stimulating hormone (TSH: 25%) secre-
    inhibitors. EC is CD30-​positive and may benefit from brentuximab tions, or diabetes insipidus (30%). Significant weight gain, predictive
    vedotin. of hypothalamic obesity, occurs shortly before diagnosis.
    Liquid biopsy, especially in CSF, is an attractive surrogate for The majority of craniopharyngiomas (95%) have a suprasellar
    tissue sampling; miRNAs are potential candidates for the follow-​up component (pure or with an intrasellar part), 5% are purely
    of tumour-​mass response to therapy. intrasellar. They combine solid, cystic, and calcified components. CT
    scan shows calcifications. On T1-​weighted MRI, cyst-​content inten-
    sity varies according to the protein content of fluid. They may extend
    Craniopharyngiomas intracranially and have an ectopic origin. The differential diagnoses
    include all tumours of the opto-​chiasmato-​hypothalamic region.
    Diagnosis: radiology, (molecular) pathology
    Current treatment strategies
    Craniopharyngiomas are World Health Organization (WHO) grade
    I tumours of the intra/​suprasellar region. Due to the proximity of the In emergency, the management of raised intracranial pressure, visual
    opto-​chiamatic and hypothalamo-​pituitary axis, the quality of life is compression, and rapid restoration or substitution of hormone
    often compromised in these children whose life expectancy is long deficits are crucial. According to the risk–​benefit balance, first-​line
    (up to 95% at 20 years after diagnosis). With less than two cases per treatment—​radical surgery and/​or radiation therapy and/​or cyst
    million persons per year, this orphan disease has a bimodal distribu- drainage—​should be discussed in a multidisciplinary meeting.
    tion: half before the end of adolescence (rarely prenatal or neonatal Complete resection, preserving visual and hypothalamic func-
    period), half during adulthood. No sex preference or predisposing tion, is widely advised. No consensus exists when optic nerve and/​or
    factors are reported. the hypothalamus are involved: complete resection versus planned
    Adamantinomatous craniopharyngiomas (Figure 24.2) are com- limited resection followed by radiation therapy. In a report from
    posed of a solid portion formed by nests and trabeculae of epithelial Elowe Gruau, the rate of severe obesity decreased when extensive
    tumour cells with calcifications, and a cystic component filled with resective surgery was compared with hypothalamus-​sparing surgical
    a dark, oily fluid. They are most frequent in children, more locally strategies tailored to an algorithm that took into account presurgical
    hypothalamic tumour involvement: 54% versus 28%. Merchant re-
    ported a 9.8 points IQ loss after complete resection versus 13.1 points
    after repeated surgery at relapse, versus 1.25 points after limited re-
    section followed by radiation therapy. Treatment in large centres was
    less radical, and the rates of complete resection and hypothalamic
    surgical lesions were lower. A multinational prospective study is
    under way to adopt strategies tailored to risk factors for morbidity
    and quality of life.
    Various surgical approaches have been described: pterional or
    subfrontal transcranial (for suprasellar tumours extending into
    the infundibulum); transventricular (for large tumours within
    the ventricle); and transsphenoidal (for infradiaphragmatic tu-
    mours with sellar enlargement). Tumour cysts can be opened for
    removal and decompression or insertion of catheters. Endoscopic
    procedures may treat occlusive hydrocephalus, and allow limited
    biopsies and catheter placement under optic control. Radiotherapy
    is very effective. The patient’s age, site, and rate of progression dic-
    tate whether re-​ operation and/​ or radiotherapy should be per-
    formed at relapse. A prospective paediatric multinational trial
    Figure 24.2 Adamantinomatous craniopharyngioma (hematoxylin-​ (KRANIOPHARYNGEOM2007) randomizes between immediate
    phloxine-​saffron staining, original magnification × 100). Solid epithelial radiotherapy versus radiotherapy at time of progression for patients
    nests, that resemble the developing tooth, are bordered by basal cells older than five years of age with incomplete resection.
    (single arrow) and include loose areas called ‘stellate reticulum’ (black
    Craniopharyngiomas are not infiltrative and have sharp borders.
    round) and a ‘ghost’ keratin called ‘wet keratin’ (double arrow).
    Reproduced courtesy of Cécile Faure-Conter, Institute of Hematology and Pediatric
    They are ideally suited for using high-​precision, three-​dimensional
    Oncology (IHOPE), Lyon, France conformational or intensity-​modulated technology with a maximum
    Ependymomas 201

    5-​mm safety margin, delivering a conventionally fractionated target of specific markers and important embryonic signalling pathways
    (total) volume dose of 54 Gy. The high-​precision technique is some- in adamantinomatous craniopharyngiomas: apart from signifi-
    times also called fractionated stereotactic radiotherapy. Stereotactic cant upregulation of several direct targets of the WNT/​β-​catenin
    radiosurgery requires sophisticated immobilization devices and is signalling pathway, important components of the hedgehog
    indicated in small, mainly solid tumours with more than 3 mm space signalling pathway (e.g. GLI2, PTCH1, as well as SHH) were also
    between the tumour and critical structures. No treatment-​related over-​expressed specifically. Gene-​expression data showed increased
    mortality has been reported. expression of MAP2, tenascin C (TNC), and the stem-​cell marker
    Proton therapy is frequently considered to be the best option, CD133 (PROM1). CD44 and claudin 1 (CLDN1) are significantly
    when accessible, although no comparative study between proton-​ downregulated. All these alterations may be targeted by new com-
    beam therapy and modern photon therapy is available. As reported pounds in the future and take an active part in the treatment of
    by Fitzek, the combination of proton with photon irradiation for craniopharyngiomas.
    residual or recurrent disease provides a ten-​year local control rate
    of 85% and OS of 72%, without treatment-​related neurocognitive
    deficits, although newly diagnosed panhypopituitarism, cerebrovas- Ependymomas
    cular accident, and cyst enlargement have been described.
    Multivariate analysis of a prospective programme reported by Ependymomas are the third most common paediatric tumour of
    Muller showed that the risk for relapses after complete resection was the central nervous system (CNS), accounting for 10% of brain tu-
    80% lower when compared to incomplete resection. The risk for pro- mours in children. Ependymal tumours can arise along the entire
    gression was 88% lower in irradiated patients when compared with neuraxis. In paediatric patients, more than 90% of ependymomas
    patients without or before irradiation. The tumour control rate was are located intracranially, with two thirds found in the infratentorial
    identical if radiation was offered after primary surgery or at the time region. More than half of paediatric ependymomas occur in chil-
    of recurrence. dren younger than five years of age, with an overall male:female
    Local intracystic instillations—​bleomycin or interferon-​α—​may ratio of about 2:1. The aetiology of ependymal tumours is not yet
    be used in predominantly monocystic craniopharyngiomas. Emitting known, and there are no reliable associations with environmental or
    isotopes may be used for relapsing monocystic craniopharyngiomas infection-​related factors. Ependymomas rarely occur in the context
    after failure of surgery and external-​beam irradiation. of cancer predisposition syndromes. Patients with neurofibroma-
    Treatment-​related morbidities are a major concern. Transient tosis type 2 may develop spinal ependymomas, while in both Turcot
    post-​surgical diabetes insipidus occurs in almost all cases and per- and Gorlin syndrome intracranial ependymomas may occur.
    sists in around 80%. Inappropriate antidiuretic hormone secretion Progression-​free survival (PFS) and OS rates of children with
    and cerebral salt-​wasting syndrome are usually transient. GH defi- intracranial ependymomas vary significantly between different
    ciency is found in most patients. Its therapy induces excellent linear series, while all prior studies that evaluated the therapeutic value of
    growth, but has only a slight beneficial effect on body mass index interventions in these tumours have not accounted for molecular
    (BMI) gain. A third of patients have neurological sequelae including subgroup affiliation, and might therefore be confounded by clinical
    both visual and neurological complications (e.g. epilepsy, motor or differences in response to therapy between these subgroups. Thus,
    cranial nerve deficits), especially when the initial tumour volume clinical management of intracranial ependymomas (WHO Grade II/​
    was large. A series by Liu et al. described that one third of patients III) is challenging, and the optimal treatment strategy is contentious.
    treated with irradiation develop some type of vasculopathy, half of Intracranial ependymomas, particularly before administration of
    them being symptomatic. However, treatment techniques dated back any therapy, demonstrate predominantly locally invasive growth
    to the nineties and some patients received intracystic bleomycin. patterns and have only very low metastatic potential. However, late
    Hypothalamic dysfunction is present in one third of patients recurrence is frequent, especially in aggressive molecular subgroups
    preoperatively and manifests in behavioural changes, disturbed (see Table 24.1).
    attention and impaired processing speed, obesity, perturbation Classic MRI appearance of ependymomas demonstrates a low T1-​
    of circadian rhythm and sleep, daytime sleepiness, imbalances in signal intensity, enhanced with gadolinium, and a heterogeneous T2
    regulation of body temperature, thirst, heart rate, and/​or blood hyperintensity. Cystic components may be seen in supratentorial tu-
    pressure. Obesity and eating disorders result in increased risks of mours. Posterior fossa ependymomas typically show an extension
    metabolic syndrome and cardiovascular disease, and increased pre- through the foramina of Luschka into the cerebellopontine angle
    mature mortality. Pharmacological interventions have little effect. and/​or through the foramen of Magendie. This highly suggestive fea-
    Its incidence increases after radical surgery, especially when surgery ture may help to distinguish ependymomas from medulloblastomas.
    reaches the mammillary bodies. A classification of presurgical in-
    volvement and post-​surgical lesions of hypothalamic structures is Histopathology, molecular classification, and risk
    now proposed to guide surgical decisions. stratification of ependymal tumours
    Ependymomas from throughout the CNS are currently subdiv-
    Future outlook ided by three histology-​based grades used to predict the natural
    CTNNB1 mutations were exclusively detected in adamantinomatous course of the disease. WHO Grade I tumours include myxopapillary
    craniopharyngiomas, whether adult or paediatric, whereas BRAF ependymomas, which typically occur in the spine, as well as
    V600E mutations were solely found in the papillary subgroup. The subependymomas, which are usually intracranial and only rarely
    methylome fingerprints perfectly matched with histology-​based found in the spine. Grade I ependymomas are relatively easier to dis-
    diagnosis. Gene-​
    expression analysis confirmed over-​ activation tinguish, occur predominantly in adults, and are usually associated
    202 CHAPTER 24 Other Central Nervous System Tumours

    Table 24.1 Molecular subgroups of ependymal tumours

    Molecular subgroup Age group Gender distribution Prognosis Genetics Histopathology


    SE Adults m>f Good Balanced Subependymomo (WHO I)
    EPN-​YAP1 Children m<f Good Focal aberrations Chrosome 11; (Anaplastic) ependymoma
    YAP1-​fusion (WHO I–​III)
    EPN-​RELA Children and adults m>f Poor Aberrations (Anaplastic) ependymoma
    Chrosome 11; (WHO I–​III)
    chromothripsis; RELA-​fusion
    SE Adults m>f Good Balanced Subependymoma (WHO I)
    EPN-​A Children m>f Poor Balanced (Anaplastic) ependymoma
    (WHO I–​II)
    EPN-​B Children and adults m<f Good CIN (Anaplastic) ependymoma
    (WHO I–​II)
    SE Adults m=f Good 6q deletion Subependymoma (WHO I)
    MPE Predominantly adults m=f Good CIN Myxopapillary ependymoma
    (WHO I)
    EPN Predominantly adults m>f Good CIN; NF2 mutation (Anaplastic) ependymoma
    (WHO I–​II)

    Adapted from Kristian W. Pajtler, et al. ‘Molecular dissection of ependymomas,’ Oncoscience, Volume 2, Number 10, pp. 827–​828, 2015. Reproduced under the Attribution 3.0
    Unported (CC BY 3.0), https://​creativecommons.org/​licenses/​by/​3.0/​

    with favourable patient outcomes. The majority of ependymomas of supratentorial ependymomas are characterized by recurrent fu-
    constitute WHO Grade II (classic) and WHO Grade III (anaplastic) sions to the oncogene YAP1, and are mainly diagnosed in childhood.
    tumours. While classic WHO Grade II ependymomas show a pap- Multivariate survival analyses demonstrated that molecular
    illary, clear-​cell, or tanycytic phenotype, anaplastic ependymomas subgrouping is superior to histopathological grading of epen-
    exhibit additional features of elevated mitotic activity, microvascular dymal tumors, and the only other variables that endured as inde-
    proliferation, and tumour necrosis. Studies demonstrated a high pendent prognostic factors were extent of surgical resection and 1q
    variance in the utility of the Grade II versus Grade III distinction as gain. Molecular subclassification is expected to significantly sim-
    a robust marker for clinical outcome, and conventional histological plify risk stratification processes in the near future. An intercon-
    grading may be further confounded by anatomical compartment. nection of tissue-​based information with more precise diagnostic
    Thus, the utility of histological grading of these tumours for risk aids gained from novel, unbiased, and observer-​independent mo-
    stratification is controversial. lecular analyses, as well as data from imaging methods including
    Recent advances in the biological characterization of epen- central pathology and radiological review, might finally converge
    dymal tumours using genome-​wide DNA methylation profiling at a refined diagnostic setting and improve treatment strategies for
    have demonstrated the existence of nine clinically, demographic- ependymoma. The updated fourth edition of the WHO classification
    ally, and molecularly distinct entities, with three occurring in each of CNS tumours already recognizes the supratentorial molecular
    anatomical compartment of the CNS—​spine (SP), posterior fossa variant, ST-​EPN-​RELA, as a distinct pathological disease entity.
    (PF), and supratentorial (ST) region. One of the molecular sub-
    groups within each compartment was enriched with WHO Grade Surgery
    I subependymomas (SE)—​ named ST-​ SE, PF-​SE, and SP-​ SE—​ Surgery is the mainstay for local tumour control and the extent of
    occurring in adults only. The two other molecular subgroups within tumour resection has been demonstrated to be the most consistent
    the spine predominantly matched the histopathology-​based diag- independent prognostic factor in intracranial ependymomas.
    noses myxopapillary ependymoma (SP-​MPE) and WHO Grade II/​ Treatment strategies increasingly comprise second-​look surgeries,
    III ependymoma (SP-​EPN). The remaining two molecular types of since there are marked outcome differences between patients with
    ependymoma occurred in the posterior fossa, termed PF-​EPN-​A complete versus incomplete tumour resection. However, these
    and PF-​EPN-​B (or, alternatively, posterior fossa Group A and B), concepts require experienced and comprehensive radiological as-
    and were independently identified in retrospective studies. PF-​ sessment of the initial and residual disease status on which to base
    EPN-​A tumours are predominantly diagnosed in infants and young potential secondary neurosurgical intervention decisions.
    children, and are associated with high recurrence rates and poor In the context of molecular subgroups of ependymal tumours, it
    clinical outcome. In contrast, PF-​EPN-​B tumours are mainly found was recently demonstrated, in a huge retrospective study comprising
    in adolescents and young adults, and are associated with a better four independent non-​overlapping cohorts of PF ependymomas
    prognosis. The majority of supratentorial ependymomas (more than (n=820 cases), that patients with either PF-​EPN-​A or PF-​EPN-​B
    70%) are characterized by fusions between a gene with unknown tumours benefit from gross total resection. However, survival rates
    function, C11orf95, and the NFkB effector, RELA, and are desig- are particularly poor for incomplete resected PF-​EPN-​A tumours,
    nated as ST-​EPN-​RELA tumours. ST-​EPN-​RELA tumours occur in even after completion of radiation therapy. Notably, a subset of pa-
    both children and adults, while the remaining molecular subgroup tients with gross totally resected PF-​EPN-​B tumours did not recur,
    Choroid plexus tumours 203

    even in the absence of radiotherapy. For ST ependymomas, series local radiotherapy. There is no evidence to recommend routine use
    and collections of single cases markedly differ regarding age distri- of chemotherapy outside of clinical trials. Randomized trial-​based
    bution, therapy modalities, and availability of molecular data. Thus, therapy de-​escalation strategies should be explored in gross total re-
    variations in outcome cannot be reliably linked to specific treatment sected PF-​EPN-​B ependymoma patients in order to test the possi-
    approaches or molecular subgroups. Routine molecular analyses bility of avoiding radiotherapy.
    in the setting of prospective trials for ST ependymomas, as well as There is an urgent need to clarify questions about the clinical out-
    investigation of clinically well-​annotated, large international retro- come of the molecular variants of ST ependymoma, eventually al-
    spective cohorts with long-​term follow-​up, are expected to further lowing for explicit therapy recommendations of these subgroups.
    clarify questions about the clinical outcome and adequate therapy Further in-​depth and comprehensive analyses of the molecular
    regimens. heterogeneity of ependymal tumours are expected to create new
    opportunities to identify promising therapeutic targets in these
    Radiotherapy and chemotherapy treatment-​resistant tumours and facilitate risk stratification.
    In addition to surgery, post-​operative involved-​field radiotherapy
    dosed at 54–​59.4 Gy is considered the standard of care for patients
    with non-​disseminated ependymomas, to lower the risk of local re- Choroid plexus tumours
    currence. Safety margins around the target volume have decreased
    from 2.0 cm to 1.0 cm, or even 0.5 cm, without increased frequency Choroid plexus tumours (CPT) are rare intraventricular neoplasms
    of tumour recurrence. It is of note that extension of immediate post-​ derived from the choroid plexus epithelium and represent 0.2% of
    operative high-​dose conformal radiotherapy, even to children under all CNS neoplasms. Within the first year of life, choroid plexus tu-
    the age of three years, led to seven-​year PFS rates of 77%, albeit long-​ mours even account for 10–​20% of brain tumours. Paediatric CPT
    term follow-​up for toxic effects on development is still pending. mainly occur in the lateral (80%) or third (15%) ventricles, whereas
    Since laterally located hindbrain ependymomas occur especially in infratentorial tumours in the fourth ventricle or cerebellopontine
    young children, proton-​beam therapy is under exploration to better angle are rare.
    spare eloquent neurological structures. There is also some evidence WHO classification of CNS tumours defines three histopathology-​
    for reasonable treatment effects through re-​irradiation in case of tu- based entities. Choroid plexus papilloma (CPP, WHO Grade I)
    mour recurrence, although this strategy was associated with a de- closely resembles normal choroid plexus tissue. Although malig-
    cline in intellectual function of affected patients. nant transformation of CPP to choroid plexus carcinoma (CPC)
    In contrast to surgery and radiotherapy, the role of chemotherapy has been described, and even some cases with metastatic disease,
    in the management of ependymoma remains unclear. Retrospective CPP are considered as being of relatively low aggressiveness. Gross
    analyses of cohorts from paediatric patients, that investigated the total resection of CPP is frequently achieved and is regarded as the
    role of chemotherapy, either failed to demonstrate a survival advan- therapeutic mainstay of these tumours. The five-​year OS rates reach
    tage or showed substantial variation between individual patients. more than 90% with surgery alone. The defining feature of atypical
    Two ongoing international randomized trials in children—​SIOP choroid plexus papilloma (aCPP, WHO Grade II) is the presence of
    Ependymoma II (Europe) and ACNS0831 (USA)—​are currently increased mitotic activity (two or more mitoses per ten high-​power
    comparing post-​irradiation chemotherapy to an observation-​only fields).
    strategy. In order to delay radiotherapy in infants, driven by con- In infants, CPP and aCPP share molecular signatures and mostly
    cerns about long-​term treatment toxicity, post-​operative chemo- show excellent outcome. However, in older children (above three
    therapy approaches in children under three years old were applied, years old) and adults, aCPPs are also associated with recurrent dis-
    with 42% being the highest rate of five-​year PFS reached to date. ease and intermediate outcome, and might also be difficult to dis-
    Radiotherapy deferral strategies using chemotherapy in patients tinguish from CPC by histology alone. It remains unclear which
    older than one year are, therefore, not frequently applied. Ongoing molecular mechanisms are driving these differences in aCPP be-
    trials are currently exploring the role of neoadjuvant chemotherapy, tween different age groups. CPC (WHO Grade III) is characterized
    since there is evidence for responses to chemotherapy after subtotal by clear signs of malignancy including high mitotic activity, nuclear
    resection. To date, there is no chemotherapeutic regimen that can pleomorphism, high cellularity, blurring of the papillary growth pat-
    routinely be recommended, outside the context of a clinical trial, in tern, and necrosis. CPCs exhibit a dismal prognosis, with long-​term
    paediatric patients with intracranial ependymoma. OS of about 30%. Despite aggressive treatment strategies, including
    surgical resection and combined radio-​chemotherapy, the clinical
    Conclusion and outlook behaviour of CPCs is variable and most of the few survivors exhibit
    Ependymal tumours from different compartments of the CNS are long-​term sequelae in cognitive functions and development.
    biologically distinct, and future therapy optimization trials, as well To date, the optimal treatment regimen has not been defined, and
    as testing of novel therapeutics, should accompany molecular clas- only a few standardized trials have been performed. In an attempt to
    sification in order to be useful to ependymoma patients. The dif- delay radiotherapy in very young children, a UKCCSG-​SIOP trial
    ferentiation between histologically defined WHO Grade II versus applied post-​operative chemotherapy only, but this resulted in dra-
    Grade III ependymomas is of limited utility for clinical decision matically low survival rates, with the primary tumour bed being the
    making, and therefore should be used with great caution outside the most frequent site of relapse. The previous study, CPT-​SIOP-​2000,
    setting of a clinical trial. For patients over the age of one year with established CPTs to be chemotherapy-​responsive to combinations of
    PF-​EPN-​A ependymomas, the currently recommended standard etoposide, vincristine, and either carboplatin or cyclophosphamide,
    of care is maximal safe neurosurgical resection complemented by but which agent was superior could not be determined. The current
    204 CHAPTER 24 Other Central Nervous System Tumours

    CPT-​SIOP-​2009 protocol randomizes CPCs, aCPP with residual dis- Elowe-​ Gruau E, et al. (2013) Childhood craniopharyngioma:
    ease or at relapse, and CPP in cases of metastases or unresectability hypothalamus-​sparing surgery decreases the risk of obesity. J Clin
    to four different chemotherapy schedules. The majority of patients Endocrinol Metab 98(6), 2376–​82.
    will not receive irradiation. Fitzek MM, et al. (2006) Combined proton and photon irradiation for
    The indication for radiation, the treatment volume, and the dose craniopharyngioma: long-​term results of the early cohort of pa-
    tients treated at Harvard Cyclotron Laboratory and Massachusetts
    is specified based upon the patient’s age, tumour histology, staging,
    General Hospital. Int J Radiat Oncol Biol Phys 64(5), 1348–​54.
    and response to the two first cycles of chemotherapy. Maximal radi-
    Flitsch J, et al. (2011) Surgical strategies in childhood craniopharyn­
    ation will be given to patients over three years of age with metastatic
    gioma. Front Endocrinol 2, 96.
    CPC not responsive to the first two cycles. These patients will also Geffner M, et al. (2004) Changes in height, weight, and body mass
    receive CSI. It is of note that over 50% of CPC harbour somatic TP53 index in children with craniopharyngioma after three years of
    mutations, while germline TP53 mutations are also frequently iden- growth hormone therapy: analysis of KIGS (Pfizer International
    tified, since CPC is one of the hallmark cancers of the Li–​Fraumeni Growth Database). J Clin Endocrinol Metab 89(11), 5435–​40.
    syndrome. This cancer predisposition syndrome should be con- Goldman S, et al. (2015) Phase II trial assessing the ability of
    sidered in the planning process for radiation therapy. neoadjuvant chemotherapy with or without second-​look surgery
    Recent classification approaches of CPT using DNA methylation to eliminate measurable disease for nongerminomatous germ cell
    profiling in a cohort across all age groups and CNS locations iden- tumors: a Children’s Oncology Group study. J Clin Oncol 33(22),
    tified three distinct molecular clusters. These clusters are clinically 2464–​71.
    relevant. Notably, some paediatric aCPP with favourable prognosis González Briceño L, et al. (2014) Water and electrolyte disorders at
    cluster together with CPP, thereby confirming previous studies that long-​term post-​treatment follow-​up in paediatric patients with
    suprasellar tumours include unexpected persistent cerebral salt-​
    identified molecular similarities of CPP and aCPP in young chil-
    wasting syndrome. Horm Res Pædiatr 82(6), 364–​71.
    dren. However, a significant proportion of histologically diagnosed
    Hardenbergh PH, et al. (1997) Intracranial germinoma: the case for
    paediatric CPP and aCPP cluster together with CPC, and indeed lower dose radiation therapy. Int J Radiat Oncol Biol Phys 39(2),
    have been associated with a higher risk of tumour progression. 419–​26.
    Further validation of these data in future prospective trials and in-​ Hölsken A, et al. (2016) Adamantinomatous and papillary craniopharyn­
    depth molecular characterization of different CPT subgroups will giomas are characterized by distinct epigenomic as well as mutational
    inform risk stratification of affected patients and set the basis for and transcriptomic profiles. Acta Neuropathol Comm 4(1), article 20.
    future treatment intensification in high-​risk patients and potentially Khatua S, et al. (2016) Current therapy and the evolving molecular
    de-​escalation strategies in low-​risk patients. landscape of paediatric ependymoma. Eur J Cancer 70, 34–​41.
    doi: 10.1016/​j.ejca.2016.10.013
    Laffond C, et al. (2012) Quality-​of-​life, mood and executive func-
    FURTHER READING tioning after childhood craniopharyngioma treated with surgery
    Amendola BE, et al. (2003) Role of radiosurgery in craniopharyngiomas: and proton beam therapy. Brain Inj 26(3), 270–​81.
    a preliminary report. Med Pediatr Oncol 41(2), 123–​7. Liu AK, et al. (2009) Vascular abnormalities in pediatric craniopharyn­
    Aoyama H (2009) Radiation therapy for intracranial germ cell tumors. gioma patients treated with radiation therapy. Pediatr Blood Cancer
    Prog Neurolog Surg 23, 96–​105. 52(2), 227–​30.
    Bahar M, et al. (2015) Radiation therapy for choroid plexus carcinoma Mason PW, et al. (2002) The use of dextroamphetamine to treat obesity
    patients with Li–​ Fraumeni syndrome: advantageous or detri- and hyperphagia in children treated for craniopharyngioma. Arch
    mental? Anticancer Res 35(5), 3013–​17. Pediatr Adol Med 156(9), 887–​92.
    Beltran C, et al. (2012) On the benefits and risks of proton therapy Merchant TE (2017) Current clinical challenges in childhood
    in pediatric craniopharyngioma. Int J RadiatOncol Biol Phys 82(2), ependymoma: a focused review. J Clin Oncol 35(21), 2364–​9.
    e281–​7. doi: 10.1200/​JCO.2017.73.1265
    Calaminus G, et al. (2013) SIOP CNS GCT 96: final report of outcome Merchant TE, et al. (2002) Craniopharyngioma: the St Jude Children’s
    of a prospective, multinational nonrandomized trial for children Research Hospital experience 1984–​2001. Int J Radiat Oncol Biol
    and adults with intracranial germinoma, comparing craniospinal Phys 53(3), 533–​42.Merchant TE, et al. (2013) Disease control after
    irradiation alone with chemotherapy followed by focal primary site reduced volume conformal and intensity modulated radiation
    irradiation for pat. Neuro Oncol 15, 788–​96. therapy for childhood craniopharyngioma. Int J Radiat Oncol Biol
    Cavalheiro S, et al. (2010) Craniopharyngiomas: intratumoral chemo- Phys 85(4), e187–​92.
    therapy with interferon-​alpha: a multicenter preliminary study Merino DM, et al. (2015) Molecular characterization of choroid plexus
    with 60 cases. Neurosurg Foc 28(4), E12. tumors reveals novel clinically relevant subgroups. Clin Cancer Res
    Chakrabarti I, et al. (2005) Long-​ term neurological, visual, 21(1), 184–​92. doi: 10.1158/​1078-​0432.CCR-​14-​1324
    and endocrine outcomes following transnasal resection of Modak S, et al. (2004) Thiotepa-​based high-​dose chemotherapy with
    craniopharyngioma. J Neurosurg 102(4), 650–​7. autologous stem-​cell rescue in patients with recurrent or progres-
    Derrey S, et al. (2008) Management of cystic craniopharyngiomas with sive CNS germ cell tumors. J Clin Oncol 22(10), 1934–​43.
    stereotactic endocavitary irradiation using colloidal 186Re: a retro- Mortini P, et al. (2011) Neurosurgical treatment of craniopharyngioma
    spective study of 48 consecutive patients. Neurosurgery 63(6), 1045–​52. in adults and children: early and long-​term results in a large case
    Dudley RW, et al. (2015) Pediatric choroid plexus tumors: epidemi- series. J Neurosurg 114(5), 1350–​9.
    ology, treatments, and outcome analysis on 202 children from Mottolese C, et al. (2001) Intracystic chemotherapy with bleomycin
    the SEER database. J Neurooncol 121(1), 201–​7. doi: 10.1007/​ in the treatment of craniopharyngiomas. Childs Nerv Syst 17(12),
    s11060-​014-​1628-​6 724–​30.
    Choroid plexus tumours 205

    Müller HL (2011) Diagnostics, treatment, and follow-​ up in Pajtler KW, et al. (2015) Molecular classification of ependymal tu-
    craniopharyngioma. Front Endocrinol 2, 70. mors across all CNS compartments, histopathological grades,
    Müller HL (2014) Craniopharyngioma. Endocr Rev 35(3), 513–​43. and age groups. Cancer Cell 27(5), 728–​ 43. doi: 10.1016/​
    Müller HL, et al. (2004) Longitudinal study on growth and body mass j.ccell.2015.04.002
    index before and after diagnosis of childhood craniopharyngioma. Phi JH, et al. (2010) Germinomas in the basal ganglia: magnetic reson-
    J Clin Endocrinol Metab 89(7), 3298–​305. ance imaging classification and the prognosis. J Neuro-​Oncol 99(2),
    Müller HL, et al. (2011) Post-​operative hypothalamic lesions and 227–​36.
    obesity in childhood craniopharyngioma: results of the multi- Ramaswamy V, et al. (2016) Therapeutic impact of cytoreductive
    national prospective trial KRANIOPHARYNGEOM 2000 after 3-​ surgery and irradiation of posterior fossa ependymoma in the
    year follow-​up. Eur J Endocrinol 165(1), 17–​24. molecular era: a retrospective multicohort analysis. J Clin Oncol
    Murray MJ, et al. (2016) A pipeline to quantify serum and cerebro- 34(21), 2468–​77. doi: 10.1200/​JCO.2015.65.7825
    spinal fluid microRNAs for diagnosis and detection of relapse Thomas C, et al. (2016) Methylation profiling of choroid plexus tumors
    in paediatric malignant germ-​cell tumours. Br J Cancer 114(2), reveals 3 clinically distinct subgroups. Neuro Oncol 18(6), 790–​6.
    151–​62. doi: 10.1093/​neuonc/​nov322
    Özyurt J, et al. (2014) Neuropsychological outcome in patients with Wang L, et al. (2014) Novel somatic and germline mutations in intra-
    childhood craniopharyngioma and hypothalamic involvement. J cranial germ cell tumours. Nature 511(7508), 241–​5.
    Pediatr 164(4), 876–​81.Pajtler KW, et al. (2016) The current con- Winkfield KM, et al. (2009) Surveillance of craniopharyngioma cyst
    sensus on the clinical management of intracranial ependymoma growth in children treated with proton radiotherapy. Int J Radiat
    and its distinct molecular variants. Acta Neuropathol 133(1), 5–​12. Oncol Biol Phys 73(3), 716–​21.
    25
    Soft-​Tissue Sarcomas
    Gianni Bisogno and Hans Merks

    Introduction Pathology and biology


    RMS is characterized by a variable degree of myogenic differenti-
    Soft-​tissue sarcomas (STS) are a heterogeneous group of tumours ation, a feature that results from the biological impact of aberrant
    derived from mesenchymal cells. As these cells normally mature transcription signals and the resultant production of myogenic pro-
    into, for example, muscle, fibrous structures, and fat, the different teins. Classically, RMS is histologically distinguished into two main
    histotypes of STS are designated according to the line of differenti- subtypes: embryonal RMS (E-​RMS), which accounts for approxi-
    ation that may be recognized in the tumour. mately 70% of all RMS, and alveolar RMS (A-​RMS) (20%). A third
    STS comprise approximately 8% of all paediatric malignan- subtype, pleomorphic RMS, is described in adults but very rarely
    cies. They are conventionally divided into rhabdomyosarcoma encountered in children; this entity is probably more closely related
    (RMS), which account for 50–​60% of the whole group, and non-​ to adult pleomorphic STS than to other RMS subtypes.
    rhabdomyosarcoma soft-​tissue sarcoma (NRSTS) (Figure 25.1). E-​RMS is characterized by a spindle and/​or round cell tumour in
    a loose myxoid or dense collagenous stroma. Two variants have been
    described:
    Rhabdomyosarcoma
    1. The botryoid subtype, found typically at vaginal or nasopharyn-
    geal sites, where tumour grows into organ cavities. Histologically,
    Rhabdomyosarcoma (RMS) is an aggressive tumour that typically
    it shows a characteristic condensed layer of tumour cells (cam-
    affects young children and can develop in almost any part of the
    bium layer) under the overlying mucosa.
    body where mesenchymal tissue is present. Historically, until the
    2. The spindle-​cell variant, with either collagen-​poor or collagen-​
    1960s, less than one third of children survived. In the 1970s, large co-
    rich leiomyomatous forms, with a storiform pattern, arising in
    operative national and international study groups started to adopt a
    paratesticular locations. Its distinction from other forms of STS
    systematic multidisciplinary approach, including multidrug chemo-
    relies on the presence of well-​differentiated rhabdomyoblasts in
    therapy coordinated with surgery and radiotherapy. This led to a
    the spindle-​cell population.
    progressive increase of survival (now above 70%) and to the identifi-
    cation of a number of prognostic factors that can be utilized to tailor Classical A-​RMS displays well-​ defined alveolar-​ like spaces sep-
    the treatment. More recently, clinical protocols have been linked to arated by thick collagenous bands and lined with round tumour
    pathology and biological studies that have added important insights cells showing variable myogenic features. A solid alveolar variant
    into the nature of RMS and may provide new therapeutic opportun- has been described and tumours exist with a mixed embryonal-​
    ities in the near future. alveolar appearance. Previously, it was generally accepted that the
    percentage of cells showing an alveolar pattern was unimportant
    Epidemiology and aetiology
    and that even a focal presence was sufficient to justify the diagnosis
    RMS is a rare tumour with an annual incidence of 4.5 per mil- of A-​RMS. More recently, this concept has been challenged and the
    lion children under the age of 20. A male predominance is gen- current histological definition of A-​RMS requires that the majority
    erally reported. In two thirds of cases, RMS arises before six years of a tumour shows alveolar histology, or solid sheets of cells with
    of age, but a second, smaller peak is evident in adolescence and monotonous round nuclei characteristic of A-​RMS. Other variants
    the number of adults with RMS should not be neglected. The aeti- have been proposed, such as pseudovascular sclerosing, rhabdoid,
    ology is not known. Genetic factors may play an important role, or anaplastic RMS, but the clinical significance of these entities is yet
    as demonstrated by an association between RMS and a familial to be confirmed.
    cancer syndrome (Li–​Fraumeni), congenital anomalies (involving Desmin and muscle-​ specific actin are relatively sensitive
    the genitourinary and central nervous system), and other genetic immunohistochemical markers of RMS, but the most sensitive
    conditions. markers are the muscle transcription factors MyoD and myogenin.
    Rhabdomyosarcoma 207

    (a) 4500

    4000

    3500

    No of cases (SEER 1973 to 2007)


    3000

    2500

    2000

    1500 Rhabdomyosarcoma
    Synovial Sarcoma
    1000
    MPNST
    500 Fibrosarcoma
    Leiomyosarcoma
    0
    0 to 18 19 to 29 30 to 39 40 to 49 50 to 59 60 to 69 70 + Liposarcoma
    Age (years)

    (b) 600
    No of cases (SEER 1973 to 2007)

    500

    400

    300
    Rhabdomyosarcoma

    200
    Synovial sarcoma

    100 Fibrosarcoma

    MPNST
    0
    <1 1 to 4 5 to 9 10 to 14 15 to 18
    Age (years)

    Figure 25.1 Frequency of soft-​tissue sarcoma subtypes: (a) at all ages; and (b) for patients aged 0–​18 years.
    Source: data from the Survival Epidemiology and End result (SEER) public-​access database (1973–​2007) (www.seer.cancer.gov).

    A high degree of nuclear staining of myogenin seems correlated future trials, fusion-​gene status will be used in risk stratification of
    with the alveolar subtype. Recently, gene-​expression data have iden- RMS irrespective of histology. No specific genetic alterations have
    tified AP2Beta and P-​cadherin as specific markers for A-​RMS, and been found in E-​RMS but many of them exhibit a loss of heterozy-
    epidermal growth factor (EGF) receptor and fibrillin-​2 as markers gosity at chromosome 11p15.5, although the biological implications
    for E-​RMS. The search for specific genetic alterations has identified of this are not yet understood.
    two chromosomal translocations specifically associated with A-​ The International Classification of Rhabdomyosarcoma identified
    RMS, and the t(2;13)(q35;q14) and t(1;13)(p36;q14) translocations three risk groups:
    are found in more than 80% of cases. These abnormalities fuse the
    1. superior prognosis: botryoid and spindle-​cell RMS
    FKHR locus on chromosome 13 to either the PAX3 gene on chromo-
    2. intermediate prognosis: E-​RMS
    some 2 (60%) or the PAX7 gene on chromosome 1 (20%), produ-
    cing the PAX3–​FKHR and PAX7–​FKHR chimeric proteins that are 3. poor prognosis: A-​RMS
    powerful transcription factors influencing cell growth, differen- This classification was shown to be strongly predictive of survival,
    tiation, and apoptosis. A minority of A-​RMS do not show known but it does not take into account the recent advances of molecular
    translocations: this may be explained by the presence of rare alterna- biology.
    tive translocations such as t(2;2) and t(2;8). It seems, however, that
    most translocation-​negative A-​RMS are clinically and molecularly Clinical presentation and diagnosis
    indistinguishable from embryonal cases and significantly different RMS is encountered at almost all anatomical sites, although head
    from fusion-​positive A-​RMS. As a consequence, in most current and and neck (40%), genitourinary (20%), and extremities (15%) are the
    208 CHAPTER 25 Soft-Tissue Sarcomas

    (a)

    (b)

    Figure 25.2 Orbital tumour: (a) clinical presentation with proptosis and deviation of globe; (b) CT scan shows large antero-​medial soft-​tissue mass
    without bone erosion.

    most common locations. It tends to invade nearby organs and re- RMS occurs in other sites; for example, systematic retroperitoneal
    gional lymph nodes. Metastases at diagnosis are present in approxi- lymph-​node sampling is recommended in older children with
    mately 20% of cases, more frequently in the lungs, bone marrow, paratesticular RMS in North American protocols, while radio-
    bones, and distant lymph nodes. Clinical presentation is strongly logical evaluation and selective biopsy is considered sufficient in
    influenced by site; for example, tumours within the orbit tend to Europe.
    present early with obvious displacement of the ocular globe and are Diagnosis must be confirmed histologically and, although needle
    rarely associated with regional lymph node or distant metastatic biopsy may be the simplest approach favoured by some clinicians,
    spread (Figure 25.2), whilst tumours in the nasopharynx may result it has the disadvantage of limiting the tissue available for conven-
    in a history of nasal discharge and obstruction and frequently involve tional histological examination, including immunohistochemistry,
    local extension to the skull base, with the potential for associated and may impede the need for fresh frozen tissue for cytogenetic and
    cranial nerve palsies or visual loss. The definition of certain head–​ molecular genetic investigation. Open biopsy is therefore frequently
    neck sites as ‘parameningeal’ relates to the risk of direct tumour preferred and should, when possible, be undertaken at the oncology
    extension into the meninges and beyond intracranially. Tumours centre where the initial surgical approach can be determined by
    within the genitourinary tract may present with urinary obstruction the multidisciplinary team responsible for the child’s subsequent
    (in bladder and prostate sites), as a scrotal mass (paratesticular), or treatment.
    as a vaginal polyp or discharge (vaginal and uterine tumours).
    Treatment
    Diagnostic and staging investigations The multidisciplinary risk-​adapted management of RMS promotes
    These must include adequate imaging of the primary site and re- chemotherapy as systemic treatment for both local and metastatic
    gional lymph nodes by magnetic resonance imaging (MRI) scans, lesions, coordinated with local treatment that includes surgery and/​
    and accurate assessment of sites of potential metastatic spread or radiotherapy. Since the early 1970s, patients have been enrolled in
    by chest computed tomography (CT) scan, fluorodeoxyglucose large multi-​institutional trials sequentially addressing crucial ques-
    (FDG)-​ positron emission tomography (PET)/​ CT scan, bone tions. Examples of successful models of clinical trials are the studies
    marrow biopsy, and aspirates. Cerebrospinal fluid should be sam- coordinated by the Intergroup Rhabdomyosarcoma Study (IRS)
    pled in the case of parameningeal/​paraspinal tumours. The risk of Group (now the Soft-​Tissue Sarcoma Committee of the Children’s
    involvement of regional lymph nodes depends on the primary site, Oncology Group (COG)) in the USA and those coordinated by the
    and there is a general consensus to biopsy nodes when the RMS European paediatric Soft-​Tissue Sarcoma Study Group (EpSSG)
    arises in the extremities (the use of sentinel lymph-​node mapping and the German Soft-​Tissue Sarcoma Cooperative Group (CWS) in
    techniques may be of help). Different strategies are adopted when Europe.
    Rhabdomyosarcoma 209

    Prognostic factors and staging a better prognosis. Tumour dimension of more than 5 cm maximum
    The presence of metastasis at diagnosis is the most powerful factor diameter is the usually adopted point for identifying tumour size as
    in determining an adverse outcome and groups of patients with lo- an adverse factor within treatment-​stratification systems, although
    calized or metastatic disease have often been investigated separately recent evidence suggests that tumour dimension should be cor-
    in the search for prognostic factors. related with patient size, and the use of a fixed-​size cut-​off may be
    questioned in the future. Tumour invasiveness is difficult to assess
    Localized disease in numerous cases, so many investigators feel it is not practical to be
    For non-​metastatic RMS patients, several prognostic factors have included in a stratification system.
    emerged with the evolution of treatment. Table 25.1 describes the Local lymph-​node involvement (N1)
    tumour and patient characteristics currently used to assign risk
    group and treatment strategy. Local lymph-​node involvement has always been regarded as a sign
    of tumour aggressiveness. In particular, patients N1 with alveolar
    Histology histology and positive fusion status seem to present a prognosis
    The IRS IV Study showed 88% relapse-​free survival in E-​RMS versus similar to metastatic RMS.
    66% in A-​RMS. This prognostic distinction has been confirmed in Post-​surgical status
    most other studies and becomes more striking if only fusion-​positive
    A-​RMS is compared with E-​RMS. The IRS grouping system is used:
    • Group I: completely excised tumours with negative microscopic
    Fusion status
    margins
    There is recent but increasing evidence that biological factors like the • Group II: grossly resected tumours with microscopic residual
    presence of the PAX3-​FKHR or PAX7-​FKHR gene fusions resulting disease.
    from t(2;13) or t(1;13) translocations, typical of alveolar RMS, may be • Group III: macroscopic residual disease after incomplete resec-
    useful to better stratify patients; their presence identifies more precisely tion or initial biopsy
    those patients at higher risk of treatment failure. Unfortunately, these • Group IV: metastatic disease
    data derive mainly from retrospective studies; therefore fusion status as
    a prognostic factor will be studied in ongoing and future trials.
    Age
    Tumour site, size, and invasiveness Prognosis is worse in patients over ten years of age. Young children
    Site is a major determinant of treatment strategy and outcome: orbit, (below one year old) generally have a worse prognosis, which may be
    non-​bladder prostate genitourinary (i.e. paratesticular and vagina/​ associated with more restricted use of radiotherapy due to the higher
    uterus), and non-​parameningeal head–​neck sites are associated with risk of sequelae in this age group.
    It is important to note that factors predictive of prognosis are often
    interdependent: limb tumours are generally of the alveolar type and
    Table 25.1 Prognostic factors in localized rhabdomyosarcoma tend to spread more frequently, both to regional nodes and to distant
    metastatic sites. The surgical status is also strongly determined by
    Favourable Unfavourable
    the site; for example, complete tumour resection is usually impos-
    Histology/​Biology Embryonal Alveolar,
    Spindle cell translocation-​positive
    sible for parameningeal RMS but easily performed for paratesticular
    Botryoid RMS tumour.
    Alveolar fusion-​negative
    Tumour site Orbit Parameningeal
    Metastatic disease
    Genitourinary non-​bladder Extremities A pooled analysis of 788 patients treated in nine studies performed by
    prostate (i.e. paratesticular and Genitourinary
    vagina/​uterus) bladder-​prostate European and American cooperative groups identified factors with
    Head and neck ‘Other’ (i.e. trunk, an independent and significant negative impact on survival: age,
    non-​parameningeal pelvis) unfavourable primary site (extremity and ‘other’ sites), presence of
    Tumour size Maximum diameter <5 cm >5 cm three or more sites of metastatic disease, and the specific presence
    Tumour T1: tumour confined to the T2: tumour involving of bone or bone-​marrow involvement. Patients without any of these
    invasiveness* organ or tissue of origin one or more unfavourable characteristics showed 50% event-​free survival (EFS)
    contiguous organs or in comparison with 42%, 12%, and 5% in patients with one, two, or
    tissues
    more than two adverse factors, respectively.
    Lymph node N0 = no clinical or N1 = clinical or
    involvement* pathological node pathological nodal Treatment stratification
    involvement involvement
    Initial surgery Complete (IRS Group I) Incomplete (IRS
    To take into account all the known prognostic factors, risk-​adapted
    Group II or III) treatment strategies are currently based on complex staging systems.
    Patient age 1–​9 years <1 year and ≥10 years As an example, the EpSSG stratification takes into consideration six
    different prognostic factors leading to the identification of eight sub-
    * TNM system; although T status is a prognostic factor in many analyses, tumour groups that receive different treatments according to the following
    invasiveness is difficult to assess in many cases and therefore is no longer part of
    stratification systems four derived risk groups:
    210 CHAPTER 25 Soft-Tissue Sarcomas

    1. Low-​risk group: includes 6–​8% of the whole population of local- cosmetic disadvantage: this may be particularly useful in limb and
    ized RMS. These children present with a small (less than 5 cm) paratesticular tumours.
    tumour, completely resected at diagnosis (IRS group I), with fa- RMS usually shrinks significantly with initial chemotherapy, and
    vourable histology and site. Most of these patients are children secondary conservative surgery to achieve local control may become
    with paratesticular RMS. Five-​year EFS is approximately 90%, more feasible and remains an important aspect of treatment. Its im-
    so the goal for this selected group of patients is to avoid aggres- plementation depends, however, on the site of disease; for example,
    sive treatment. In the current EpSSG protocol, these children are surgery has little or no role in the primary management of orbital
    treated with a relatively short course of non-​intensive chemo- or parameningeal RMS. At other sites, surgery should aim to com-
    therapy (vincristine and actinomycin-​D). pletely resect the residual tumour, although marginal resections may
    2. Intermediate-​risk group: includes patients with favourable hist- be acceptable provided that they are followed by radiotherapy. The
    ology and site, with survival around ​80%. The goal for this group efficacy and the potential morbidity of surgery and radiotherapy
    includes to reduce treatment cautiously without compromising must be carefully assessed: in some cases, the morbidity of radio-
    survival. Intensive alkylating agent-​based chemotherapy is util- therapy may be more acceptable than radical operations which re-
    ized, but there is an attempt to reduce cumulative dose exposure sult in important functional (e.g. total cystectomy) and/​or cosmetic
    and to limit the use of radiotherapy. (e.g. amputation) consequences; in other circumstances, the mor-
    3. High-​risk group: this group includes patients with large E-​RMS bidity of radical surgery to achieve local control may be preferred,
    localized in unfavourable sites or with nodal involvement, and for example to avoid pelvic or extremity irradiation in very young
    most A-​RMS. These patients still have an unsatisfactory prog- children.
    nosis (survival around 70%) and are therefore the major focus of
    Radiotherapy
    randomized trials to identify more effective strategies.
    4. Very high-​risk group: this includes mainly metastatic RMS. The role of radiotherapy in the local control of RMS is unquestion-
    In this group survival is between 20% and 30%, and has not able but awareness of the possible late effects, especially in young
    improved significantly in the past 40 years. Experimental ap- children, has led to attempts to identify groups of patients where
    proaches are therefore needed for these patients. irradiation can be withheld or modulated in its dose and target
    volume.
    Strategies determining the use of radiotherapy require consider-
    Therapeutic strategy ation of diverse factors including histological subtype, patient age,
    Surgery, radiotherapy, and chemotherapy have been combined tumour size and site, response to initial chemotherapy, and presence
    differently over the years by international cooperative groups, in of residual tumour after surgery. The only patients where radio-
    an attempt to improve outcome whilst finding the correct balance therapy can be safely withheld are those with E-​RMS completely
    between the intensity of treatment and its possible cost in terms of resected at diagnosis. It is also known that a certain proportion of
    late effects. In general, there is concordance on the use of intensive patients with E-​RMS can be cured without radiotherapy when the
    chemotherapy and aggressive non-​mutilating surgery. Controversies tumour is completely resected at secondary surgery but, as most
    relate more to the method and timing of local treatment, and, more studies suggest a survival advantage for radiotherapy, this strategy is
    specifically, to the place of radiotherapy in guaranteeing local con- often limited to very young children.
    trol for patients who appear to achieve complete remission with Studies from the European groups have attempted to relate the
    chemotherapy, with or without surgery. This represented an im- use of radiotherapy to the response to initial chemotherapy. The
    portant philosophical difference between the Société Internationale most radical approach has been used in the SIOP protocols, where
    d’Oncologie Pediatrique (SIOP) malignant mesenchymal tumours radiotherapy was avoided if complete remission was achieved after
    (MMT) studies and those of the IRS group and, to some extent, initial chemotherapy, with or without second surgery (except for
    those of the German and Italian cooperative groups. Local relapse parameningeal RMS). This approach has proven feasible, especially
    rates are higher when radiotherapy is not systematically used, al- in the subsets of patients with orbital embryonal RMS where up to
    though the SIOP experience has also made it clear that, at least for 40% were cured without irradiation. However, the higher rate of
    RMS localized in certain sites like orbit, a significant number of pa- relapse observed in this population requires consideration of the
    tients who relapse may be cured with salvage treatment. burden of second-​line therapy, and the psychological impact of re-
    lapse, in the total cost of cure. There is good evidence from all groups
    Surgery that survival is improved when radiotherapy is given to all patients
    Tumour resection at diagnosis should be attempted only if the tu- with A-​RMS and to those with E-​RMS with microscopic or macro-
    mour can be realistically excised with clear margins without danger, scopic residual disease. The timing of radiotherapy in treatment
    functional impairment, or mutilation. If this is not possible, a diag- may be important and, at least in some subgroups, such as those
    nostic biopsy is required. An attempt at surgical resection which with parameningeal RMS, there is evidence that timely irradiation
    leaves microscopic residual disease makes treatment decisions (around week 12) may improve local control.
    more complicated because the patient is unassessable for the effi- Current guidelines for therapy within COG and EpSSG recom-
    cacy of chemotherapy and may still require further local treatment. mend doses from 40 to 55 Gy, given by conventional fractionation
    Primary re-​excision (i.e. a second surgical resection before chemo- (single daily doses of 1.8 Gy) depending on the various clinical vari-
    therapy) may be worthwhile in cases where there is confidence ables. A reduction of dose to 32–​36 Gy is under exploration in chil-
    that clear excision margins can be achieved without functional or dren with favourable characteristics. Hyperfractionated schedules
    have been explored but did not show any advantage in terms of local
    Rhabdomyosarcoma 211

    control or survival. In adult sarcoma, preoperative radiotherapy is COG has compared VAC versus alternating VAC/​VI (vincristine
    becoming the new standard in the case of a resectable tumour; in the and irinotecan). The two arms produced similar results in term of
    next EpSSG study, pre-​versus post-​operative radiotherapy will be progression-​free survival, but the VAC/​VI regimen proved to be less
    studied in a randomized trial. toxic and therefore preferable to standard VAC.
    The reduction of side effects can be pursued not only by avoiding or The addition of other drugs (originally, cisplatin and etoposide;
    reducing the doses of radiotherapy but also through better treatment later, epirubicin, carboplatin, and topotecan) has failed to demon-
    planning and the use of the most modern techniques. Radiotherapy strate any convincing evidence for superiority over VAC or IVA,
    should be performed only in a limited number of experienced centres although some or all may have value in second-​line therapies. The
    and CT-​based 3D conformal planning is mandatory to reduce ir- addition of temozolomide to the VI regimen is has proven beneficial
    radiation of critical structures near the target volume. The target for relapsed patients.
    volume should be determined as the initial tumour volume plus ad-
    equate margins (1–​2 cm), but after good response to chemotherapy, Approaches with high-​dose and
    the residual tumour can be considered as target volume (plus mar- dose-​intensified chemotherapy
    gins). In large tumours, a boost can be delivered to a more limited The very poor prognosis of some subgroups of patients, particularly
    volume with high risk of relapse. Interstitial radiotherapy (brachy- those with metastases detectable at diagnosis, has justified the search
    therapy) using intracavitary moulds or implanted wires may be of for novel strategies. The place of high-​dose chemotherapy (HDCT)
    particular relevance for small tumours at selected sites, notably in followed by haematopoietic stem-​cell rescue (HSCR) has been ex-
    the vagina and perineum. Occasionally, this technique is utilized at plored in different trials, although there has been no direct random-
    other genitourinary sites, including tumours of the bladder base and ized comparison with conventional-​dose chemotherapy. The closest
    prostate, and there is emerging experience of its application to head study was a non-​randomized European protocol that compared 12
    and neck sites. Although clinical studies to better understand the in- cycles of a six-​drug regimen (CEVAIE: carboplatin, epirubicin, vin-
    dications and potential advantages and disadvantages of intensity-​ cristine, actinomycin D, ifosfamide, and etoposide) to six courses of
    modulated radiotherapy (IMRT) and proton therapy are still CEVAIE followed by high-​dose melphalan as consolidation therapy
    required, both are already increasingly utilized; in some countries, for patients who achieved complete remission. No obvious differ-
    proton therapy is standard therapy for head and neck localizations. ence in survival was identified. Similarly, unsatisfactory results were
    achieved with other high-​dose regimens (including combinations
    Chemotherapy of melphalan, etoposide, carboplatin, thiotepa, and cyclophospha-
    Chemotherapy is an essential component of treatment for all chil- mide) and with attempts to use rapid sequential HDCT with HSRC
    dren with RMS as the majority of patients with localized disease have earlier in the treatment pathway. In contrast, recent data from the
    micrometastases at diagnosis. Another important goal of chemo- German group suggested a survival advantage for metastatic pa-
    therapy is to reduce tumour volume and extension to allow for tients who received low-​dose ‘maintenance’ chemotherapy with
    more conservative subsequent surgery or radiation therapy. Finally, oral trofosfamide and idarubicin when compared (albeit in a non-​
    it is the only available tool against metastatic lesions not otherwise random manner) with patients undergoing HDCT.
    amenable to local therapy (e.g. bone-​marrow involvement). Nevertheless, the strategy to add a ‘metronomic’ (regular and fre-
    Experience since the 1970s has defined the efficacy of a var- quent low-​dosage) maintenance treatment after completing conven-
    iety of individually active chemotherapeutic agents, subsequently tional chemotherapy is a promising alternative approach, which aims
    combined in different multidrug regimens to increase their effect. to control presumed minimal residual disease resistant to standard
    Vincristine, actinomycin D, cyclophosphamide (or its analogue chemotherapy. The EpSSG RMS 2005 protocol has investigated the role
    ifosfamide), and doxorubicin have been the most frequently utilized of a six-​month schedule of maintenance treatment using vinorelbine
    agents in the treatment of RMS. Early experience from the IRS group and low-​dose oral cyclophosphamide in high-​risk patients and showed
    failed to show a convincing difference in survival when patients were a survival benefit; it is the new standard of care for this subgroup.
    treated with VAC (vincristine, actinomycin, and cyclophospha- An alternative way to increase dose intensity is by compressing the
    mide), with or without doxorubicin, and concern about potential interval between cycles of chemotherapy with granulocyte colony-​
    cardiotoxicity justifies caution over its incorporation into primary stimulating growth factor (G-​CSF) support. The recently completed
    treatment. A randomized trial recently concluded by EpSSG con- COG trial studying this approach showed improved survival (when
    firmed that the omission of doxorubicin did not impair patients’ compared to historical controls) for patients with lower-​risk meta-
    survival in high-​risk localized RMS. static E-​RMS, while prognosis for those with higher-​risk stage IV
    The substitution of cyclophosphamide by ifosfamide, in combin- disease remained dismal.
    ation with vincristine and actinomycin D (with or without doxo-
    rubicin), has been the hallmark of all recent European studies. Experimental approaches
    There is no evidence that an ifosfamide-​based combination with A number of potentially new approaches are being investigated
    vincristine and actinomycin D (IVA) is superior to VAC, although in preclinical and clinical studies, including the use of novel anti-​
    ifosfamide appears to convey some advantages over its analogue, neoplastic agents, immunotherapy, and molecules directed against
    showing a lack of cross resistance, a lower myelotoxicity profile, specific biological targets. Various new anti-​neoplastic agents are in
    and (probably) a lower risk of gonadal toxicity. However, it carries current phase I and II trials (as single agents or in combination).
    a risk of renal toxicity. Currently, VAC remains the combination of Preclinical studies have shown that the insulin signalling pathway
    choice for North American studies, while the IVA regimen is the is frequently altered in RMS, leading to development and clinical
    gold standard for European groups. A recent trial coordinated by testing of several anti-​IGF-​1-​R antibodies. Despite responses, this
    212 CHAPTER 25 Soft-Tissue Sarcomas

    approach has not been taken forward. ALK aberrations on the gen- acceptably low with cumulative doses below 60 g/​m2 as in all cur-
    omic and protein level are frequently found in RMS, and are the basis rent European RMS protocols. The continuing use of high doses of
    for current inclusion of RMS patients in early clinical trials testing alkylating agents (and etoposide) has been linked to an increased
    efficacy of small-​molecule inhibitors of ALK. As RAS isoforms are risk of second malignancy, and the long-​term follow-​up and pro-
    frequently mutated in fusion-​negative RMS, intervening in the spective evaluation of survivors of current protocols is required in
    RAS–​MEK–​ERK pathway and PIK3CA–​MTOR–​AKt pathway is a order to document the frequency and functional significance of all
    policy currently being tested preclinically and clinically. Strategies possible late effects of therapy.
    to directly target the PAX3-​FOXO1 fusion gene are under develop-
    ment but still in a preclinical phase. The observation that RMS cell
    lines express vascular endothelial growth factor (VEGF) and that Non-​rhabdomyosarcoma soft-​tissue sarcomas
    VEGF-​receptor-​1 antibody inhibits VEGF signalling and delays
    RMS proliferation has made anti-​angiogenic treatment an attractive The term non-​rhabdomyosarcoma soft-​tissue sarcoma (NRSTS)
    option. A phase II randomized study on the use of bevacizumab describes a very heterogeneous group of mesenchymal extra-​
    (a humanized monoclonal antibody which blocks the binding of skeletal malignant tumours, with different clinical and biological
    human VEGF to its receptors) in association with chemotherapy characteristics, classified on the basis of their differentiation ac-
    has recently been concluded in Europe and, unfortunately, failed to cording to the adult tissue they resemble. The term NRSTS reflects
    demonstrate any survival advantage. the fact that these tumours have often been treated according to
    Antibodies targeting the programmed cell-​ death protein 1 the principles adopted for RMS, despite the fact they represent dif-
    pathway (PD-​1/​PD-​L1) function by removing the brakes on T cells, ferent entities.
    which can then perform active anti-​tumour immune surveillance; Most of the NRSTS histotypes are more common in adults than in
    several antibodies are currently being investigated preclinically in children, but the frequency of the different subtypes differs signifi-
    RMS and in early clinical basket trials. cantly according to age: infantile fibrosarcoma is a peculiar entity
    The demonstration of a graft effect against some solid tumour of young children, synovial sarcoma and malignant peripheral
    cells has prompted the consideration of allogeneic bone-​marrow nerve sheath tumour (MPNST) are typical of adolescents and young
    transplantation in paediatric STS. Recent studies were however adults, and liposarcoma and leiomyosarcoma occur in adults and are
    negative and, also taking into account its high toxicity, this approach very rare in children (see Figure 25.1).
    is difficult to recommend. The pathogenesis of NRSTS is still unknown and there are no
    well-​established risk factors. Ionizing radiation (angiosarcoma) and
    Outcome oncogenic viruses (leiomyosarcoma in immunocompromised pa-
    The advent of multidisciplinary protocols dramatically increased tients) have been associated with the development of some NRSTS,
    the overall survival (OS) of patients with RMS from around 20% but the aetiological relation is yet unclear. A well-​known associ-
    in the 1970s to over 70% in the early 1990s, since which time fur- ation exists between MPNST and neurofibromatosis type 1 (NF1).
    ther improvement has been slower. Treatment failures usually occur The lifetime risk of developing MPNST in NF1 patients has been
    within three years from diagnosis, and are represented mostly by estimated at 8–​13% compared to 0.001% in the general population.
    local or regional relapse (accounting for approximately 60% of all Both NF1-​associated and sporadic MPNST show loss or mutation of
    relapses). Survival after relapse is generally poor (below 30% at five the NF-​1 gene (chromosome 17q11.2).
    years), but patients with favourable tumour characteristics (embry- Tumour aggressiveness, and its propensity to local relapse and dis-
    onal histology, favourable primary site) who did not receive radio- tant metastases, correlates with histological diagnosis and tumour
    therapy during first-​line treatment have a realistic chance of cure grade and size. In general, low-​grade tumours are locally aggressive,
    with further therapy. especially if they are small, but they rarely metastasize, whilst high-​
    grade tumours are more aggressive and have a high risk of metastasis
    Late effects of treatment (Table 25.2).
    The cost of the cure is the difficult, yet essential, issue to be ad-
    dressed when reviewing the outcome of survivors of all forms of Diagnosis
    cancer in childhood, particularly when survival is achieved using Similarly to RMS, NRSTS can arise anywhere in the soft tissues of
    different philosophies and treatment modalities. The importance the body, although the most common clinical presentation is a pain-
    of accurate prognostic assessment at diagnosis is as much to ensure less, growing mass localized in lower extremities; less frequent sites
    that patients with a good prognosis are not overtreated as to iden- are the trunk or the head and neck region.
    tify those with a poorer prognosis who require a more aggressive The diagnostic work-​up is similar to the one used for RMS, with
    approach. MRI to investigate the primary tumour and further investigations
    Much concern has been focused on the late sequelae of local performed according to the histological subtype (i.e. lung CT is ne-
    treatment for RMS, particularly in patients with head–​neck and cessary in most cases, but brain MRI is necessary only in cases of
    genitourinary primaries. Chemotherapy is also associated with rhabdoid tumour or alveolar soft-​part sarcoma).
    significant sequelae in some patients, and the concept that more The importance of initial surgery in NRSTS is demonstrated by
    intensive chemotherapy may help to reduce the use of local treat- the fact that they are often staged according to the IRS post-​surgical
    ment must be balanced against the additional toxicity that the grouping system as used for RMS, while the TNM (tumour–​node–​
    chemotherapy may bring. The more recent use of ifosfamide has metastasis) system seems less important, at least because nodal in-
    raised concern about long-​term renal damage, although risks seem volvement is, in general, rare for NRSTS.
    Non-rhabdomyosarcoma soft-tissue sarcomas 213

    Table 25.2 Peak age of incidence, molecular findings, and clinical characteristics of the most frequent non-​
    rhabdomyosarcoma soft-​tissue sarcoma subtypes

    Histotype Peak age of incidence Molecular findings Clinical characteristics


    Synovial sarcoma Second to third decade t(X;18)(p11;q11) Extremity (more frequent) or axial site
    SYT-​SSX1, RR up to 60%
    SYT-​SSX2, Overall satisfactory prognosis
    SYT-​SSX4
    Malignant peripheral nerve Third to fourth decade Loss or rearrangement of 10p, 30% associated with NF-​1
    sheath tumour (MPNST) 11q, 17q, and 22q Frequently located in the trunk
    RR up to 50–​60% in non-​NF1, <10% in NF1
    patients
    Poor prognosis
    Adult-​type fibrosarcoma Fourth to fifth decade t(2,5) and t(7,22) Diagnosis of exclusion. Tendency to metastatic
    spread according to tumour grade
    Epithelioid sarcoma Third decade -​ Superficial, distant sites (fingers)
    Indolent course
    Lymph nodal spread
    RR <30%
    Intermediate prognosis
    Liposarcoma Sixth to seventh decade Myxoid liposarcoma: Different biology and clinical behaviour according
    (earlier for the myxoid t(12;16)(q13;p11) to the subtype (i.e. well-​differentiated, de-​
    type) t(12;22)(q13;q12) differentiated, or myxoid/​round-​cell subtype)
    FUS-​CHOP Retroperitoneal location
    Leiomyosarcoma Sixth decade -​ Retroperitoneum
    Immunocompromised patients
    Alveolar soft-​part sarcoma Third decade t(X;17)(p11.2;q25) Head and neck and other unusual locations
    ASPSCR-​TFE3 RR <20%
    Poor prognosis
    (better than adults)
    Clear-​cell sarcoma Third to fourth decade t(12;22)(q13;q12) Extremity site, deep-​seated
    t(9;22)(q22;q12) Poor response to chemotherapy
    ATF1-​EWS Poor prognosis
    Angiosarcoma Fourth to sixth decade -​ Associated with lymphoedema, may occur after
    radiotherapy
    Superficial or deep
    Poor prognosis
    Extraskeletal myxoid Fifth to sixth decade t(9;22)(q22;q12) Slow-​growing tumour of extremities
    chondrosarcoma t(9;17)(q22;q11.2)
    EWS-​CHN
    Extraskeletal mesenchymal Second to third decade Complex cytogenetic Head–​neck region (orbit)
    chondrosarcoma alteration Highly aggressive tumour
    t(11;22) (q24;q12) (as Ewing
    family tumours)
    Infantile fibrosarcoma First year of life (rare after t(12;15;)(p13;q25) Rapid growth
    2 years of age) ETVG(TEL)-​NTRK3 Relatively high chemosensitivity
    Good prognosis
    Desmoplastic small round-​ Second to third decade t(11;22) (p13;q12) Abdominal mass widely disseminated at onset
    cell tumour EWS-​WT1 Poor prognosis
    Dermatofibrosarcoma Third to fifth decade t(17;22) Subcutaneous
    protuberans t(2;17)(p23;q23) Indolent growth
    ALK-​CLTC Good response to ALK inhibitors
    PDGFβ-​COL1A1
    Rhabdoid tumour Infants and young children hSNF5/​INI 1 gene mutations Poor prognosis

    RR: response rate (to chemotherapy; in most cases this also includes minor responses)
    N.B. fusion transcripts are in italic

    After the radiological assessment, a biopsy should be taken to de- diagnosis of histotype requires the use of immunohistochemistry
    fine the histological diagnosis and to grade the tumour appropriately. and cytogenetic/​molecular investigation. A progressively increasing
    The World Health Organization (WHO) classification of STS, last number of specific translocations have been identified in NRSTS
    updated in 2013, classifies tumours (on the basis of their differen- (Table 25.2).
    tiation) as adipocytic, fibroblastic/​myofibroblastic, fibrohistiocytic, Once the histotype is defined, the grade of malignancy should be
    smooth muscle, pericytic/​ perivascular, skeletal muscle, nerve estimated. Although this usually indicates how aggressively a tu-
    sheath, vascular, and of uncertain differentiation. In most cases, the mour will behave, some diagnoses (e.g. synovial sarcoma, alveolar
    214 CHAPTER 25 Soft-Tissue Sarcomas

    sarcoma, angiosarcoma) should be considered high grade regardless 5. tumour size: five-​year OS is 95% for tumours 5 cm or smaller,
    of their morphological parameters, whereas in others (e.g. clear-​cell 55% for those larger than 5 cm.
    sarcoma, extra-​skeletal myxoid chondrosarcoma), biological behav-
    Tumour invasiveness (T stage) and superficial/​deep location are
    iour is less predictable. Tumour grade is usually established from
    often linked to tumour size and site, and are not commonly used in
    a combined assessment of histological features including degree
    risk stratification of NRSTS in children. Many of these prognostic
    of cellularity, cellular pleomorphism or anaplasia, mitotic activity,
    variables are interrelated; for example, MPNST are often large, axial
    and degree of necrosis. Over the years, two different grading sys-
    tumours and unresectable at diagnosis.
    tems have emerged: the Paediatric Oncology Group (POG) system
    and the French system (FNCLCC) (developed for adult STS but also Synovial sarcoma
    used in Europe for paediatric NRSTS).
    Synovial sarcoma (SS) is a malignant undifferentiated mesenchymal
    Treatment strategy tumour, resembling normal synovial tissue. It represents 5–​10% of
    all soft-​tissue tumours in adults and is one of the most common
    The rarity and heterogeneity of these tumours suggest that children
    NRSTS in the paediatric age range (15–​25%), occurring primarily
    and adolescents with NRSTS should be referred to institutions with
    in adolescents. Histologically, two major subtypes are described: the
    adequate experience and committed to enrolling patients in clin-
    monophasic form, characterized by spindle cells, and the more fre-
    ical trials. Treatment is usually based on histotype, grading, and tu-
    quent biphasic form, characterized by spindle and epithelial cells or-
    mour resectability. In consideration of the limited chemosensitivity,
    ganized in glandular structures. Cytogenetic studies have reported
    a more aggressive approach to local treatment is generally required,
    the presence of a characteristic translocation t(X;18)(p11;q11) in
    and surgery still represents the mainstay of treatment in adult as
    approximately 90% of cases. Two fusion transcripts, SYT-​SSX1 and
    in paediatric NRSTS. A complete tumour resection gives the best
    SYT-​SSX2, are derived from fusion of the SYT and SSX genes local-
    chance of cure and may be the only treatment needed for most
    ized on chromosome 18q11 and Xp11 respectively. SS usually arises
    histotypes.
    in the extremities, and the lungs are by far the most common meta-
    Surgery aims to be more conservative in children, preferring
    static site. Prognosis is related to the presence of metastatic lesions,
    limb-​ sparing procedures and non-​ compartmental resections.
    tumour size (larger than 5 cm) and site (extremity locations fare
    Therefore the concept of radical surgery in children allows closer
    better than axial locations), and complete resection.
    surgical margins than are generally acceptable in adults. If resection
    Comparative studies have highlighted that survival is lower in
    is considered unlikely at diagnosis, an incisional biopsy is required
    adults than in children, even after correction for the presence of un-
    for diagnostic purposes (fine-​needle biopsy does not usually provide
    favourable risk factors. This might reflect differences in the biology
    sufficient accuracy) and other treatment modalities need to be con-
    of the tumour at different ages and/​or the different use of chemo-
    sidered before referring the child back to the surgeon for a defini-
    therapy. In paediatric oncology, SS is considered as a relatively
    tive procedure. Radiotherapy is adopted when post-​surgical residual
    chemosensitive tumour, with a reported response rate of up to 60%.
    disease is present (IRS III) or suspected (IRS II). Depending on the
    Chemotherapy has therefore been systematically used in all patients.
    age of a child and likeliness of a sufficient margin after surgery, pre-
    In 2005, EpSSG promoted a multinational study with the aims to
    operative irradiation of the tumour may be considered, increasingly
    homogenize SS treatment in Europe and collect prospective data
    resembling the approach in adults.
    (see Figure 25.3). Patients were stratified according to surgical stage,
    NRSTS have demonstrated different chemosensitivity: from
    tumour size and site (extremity versus axial), and nodal involve-
    80% in undifferentiated soft-​tissue sarcoma, to 60% in synovial
    ment: low risk (IRS 1, size less than 5 cm), intermediate risk (IRS 1,
    sarcoma, and less than 30% in other histotypes. Recent reports
    size more than 5 cm, and all IRS II), and high risk (all IRS III or any
    have suggested that the use of chemotherapy seems justified
    N1 tumour). The study recently concluded and reported five-​year
    in patients with resected but large and high-​ grade tumours
    EFS of 80.7% and OS of 90.7%—​superior to the results reported in
    (to reduce the risk of distant relapse) and also in patients with
    previous series. This study reported a 55.2% response rate to chemo-
    unresectable tumours (to attempt to reduce the lesion and make
    therapy (including also minor responses) and demonstrated that
    it operable). As in adult NRSTS, the ifosfamide–​doxorubicin
    chemotherapy can be safely omitted in low-​risk patients.
    combination represents the preferred regimen for paediatric
    NRSTS, but interesting new, more histotype-​directed, drugs are Adult-​type non-​rhabdomyosarcoma
    under investigation. soft-​tissue sarcomas
    The overall cure rate for NRSTS patients is around 70%, but it cor-
    A series of heterogeneous tumours typical of adults are collected
    relates with a series of prognostic variables including:
    under this definition (see Table 25.2). Whilst their histological ap-
    1. extent of disease at diagnosis: survival is very poor in children pearance and biology are very different to paediatric-​type NRSTS,
    with disseminated disease (OS less than 20%); they share the same low chemosensitivity and critical role of com-
    2. completeness of initial surgery: five-​year OS is around 90% in plete surgery to achieve cure.
    IRS group I, 80% in those who had marginal resection (group For adult-​type NRSTS, the possibility to resect the tumour, using
    II), and 50% in initially unresected cases (group III); a conservative approach but with adequate margins, should be
    3. tumour grade: five-​year OS is 90% for Grade 1, 80% for Grade 2, considered at diagnosis. Unfortunately, this is only the case for a
    and 65% for Grade 3; minority of usually large and invasive lesions. Neoadjuvant chemo-
    4. tumour site: five-​year OS is 80% for extremity tumours and 60% therapy is therefore recommended because experience suggests that
    for axial location; up to 40% of these patients may achieve some degree of tumour
    Non-rhabdomyosarcoma soft-tissue sarcomas 215

    Synovial sarcoma

    IRS group I, ≤5 cm Surgery alone Ifosfamide-adriamycin chemotherapy


    IRS group I, ≤5 cm Surgery Ifosfamide chemotherapy

    IRS group II, ≤5 cm Surgery Surgery S


    RT
    IRS group II, >5 cm Surgery Radiotherapy RT
    RT
    IRS group III/N1/axial Biopsy
    S RT

    ‘Adult-type’ NRSTS

    IRS group I-II, G3, >5 cm Surgery RT


    IRS group III/N1 Biopsy
    S RT

    Figure 25.3 Risk-​adapted treatment strategy for synovial sarcoma and ‘adult-​type’ soft-​tissue sarcoma in the European Paediatric Soft Tissue Sarcoma
    Study Group (EpSSG) NRSTS 2005 protocol.
    IRS, Intergroup Rhabdomyosarcoma Study; G, tumour grade.
    Adapted with permission from Ferrari A., et al. ‘Synovial sarcoma in children and adolescents: the European Pediatric Soft Tissue Sarcoma Study Group prospective trial (EpSSG
    NRSTS 2005)’. Annals of Oncology, Volume 26, Issue 3, pp. 567–​572. Copyright © 2014, Oxford University Press. DOI: https://​doi.org/​10.1093/​annonc/​mdu562

    shrinkage, possibly facilitating tumour resection and giving some Finally, eribulin has recently been approved on the basis of im-
    survival advantage. proved overall survival for patients with adipocytic sarcomas, and
    The role of chemotherapy after complete tumour resection, to trabectedin is now approved for patients with leiomyosarcomas and
    reduce the risk of distant recurrence, is controversial. There are no liposarcomas. Whilst these agents have been tested mainly in adult
    randomized trials in paediatric NRSTS (and probably never will be patients and have produced only temporary responses, they also
    considering the limited number of patients) and the results of adult represent a promising approach for paediatric NRSTS.
    trials are clearly not in favour of chemotherapy. Survival is around
    90% in patients with small, low-​grade tumours after initial tumour Other histotypes
    resection alone. The scenario is completely different for patients with Other rare lesions arise from soft tissue in children, with peculiar
    high-​grade, large and invasive tumours: even after initial complete clinical biological characteristics and different levels of aggres-
    resection, they have a high risk of developing metastases if the treat- siveness. A brief description of the most encountered histotypes is
    ment is limited to local therapy alone. This would suggest, in prin- given here.
    ciple, the use of systemic chemotherapy to try to improve survival,
    despite the lack of clear evidence.
    Radiotherapy plays a well-​ defined role in local control for Table 25.3 Histology-​directed therapeutic options in adult-​type
    adult STS but its indications are more strictly controlled in chil- non-​rhabdomyosarcoma soft-​tissue sarcoma
    dren, given the higher risk of late effects. In general, irradiation is Hitotype Agent
    planned when microscopic or macroscopic residuals remain after
    Synovial sacoma Ifosfamide
    surgery or in conjunction with chemotherapy to make the tumour
    Taxanes
    resectable.
    In the EpSSG NRSTS 2005 protocol, ifosfamide–​doxorubicin Gemcitabine
    adjuvant chemotherapy is recommended for patients with large, Vascular endothelial growth factor (VEGF)
    high-​grade tumours in IRS I or II group, or as neoadjuvant treat- inhibitors
    ment in unresectable lesions (Figure 25.3). COG uses a very Gemcitabine
    similar approach, but with radiotherapy concomitant to chemo- Gemcitabine and docetaxel
    therapy, in initially unresectable tumours to maximize tumour Trabectedine
    shrinkage.
    Vascular endothelial growth factor (VEFG)
    In recent years, various drugs other than the ifosfamide–​ inhibitors
    doxorubicin combination have shown some activity against par- Mammalian targets of rapamycin (mTOR)
    ticular STS histotypes (Table 25.3). The gemcitabine–​docetaxel inhibitors
    combination represents a standard second-​line treatment in adult Liposarcoma (myxoid/ Trabectedine
    STS. Taxanes and biological agents like bevacizumab have pro- round cell)
    cured responses in angiosarcoma. The peculiar vascular pattern Dermatofibrosarcoma Imatinib
    of alveolar soft-​part sarcoma (ASPS) can be affected by the use protuberans
    of sunitinib and cedirabinib. The extremely low response rate to Alveolar soft part sarcoma Sunitinib, cedarinibid
    chemotherapy makes the use of these agents attractive in first-​line Clear cell sarcoma Sunitinib
    treatment.
    216 CHAPTER 25 Soft-Tissue Sarcomas

    Infantile fibrosarcoma increased the ability to recognize MRT and probably is increasing
    Infantile fibrosarcoma (IFS) represents a distinct entity with pecu- the number of patients being diagnosed with this entity.
    liar characteristics and a better prognosis in comparison with adult-​ Despite progress in the understanding of the biological mechan-
    type fibrosarcoma. IFS is the most common sarcoma under one year isms underlying tumour development, only little progress has been
    of age and is rarely diagnosed after the age of two. made in the treatment of MRT. The outcome remains unsatisfac-
    It is characterized by the presence of a characteristic t(12;15) tory, with three-​year OS of 38% in the most recent study. Treating
    translocation coding for the ETV6-​NTRK3 gene fusion. Recently, these patients is extremely difficult because a substantial propor-
    transcript variants like LMNA-​NTRK1 have been described, ex- tion progress early during treatment and the use of radiotherapy is
    plaining, at least partially, the minor part of IFS where the ETV6-​ limited by their young age. An age of under one year and the pres-
    NTRK3 gene fusion is not detected. It must be kept in mind, however, ence of metastatic disease are the strongest predictors of outcome.
    that the same genetic alteration can be found in other tumours like Improved understanding of the biology and role of SMARCB1 has
    nephroblastic nephroma. enabled identification of new targets for small molecules, like his-
    IFS usually arises in the distal extremities, and presents as a pain- tone deacyetylase inhibitors, that are currently under investigation.
    less, rapidly enlarging mass. Despite initial rapid growth, evolution Fybroblastic-​myofibroblastic lesions
    may be indolent. It is considered an intermediate malignancy ac-
    cording to WHO classification because metastatic spread is rare. In This is a homogenous group of tumours characterized by invasive
    addition, few cases of spontaneous regression have been described. local growth but no propensity to distant dissemination. They are
    The overall prognosis is very good (survival rates are above 90% in therefore considered intermediate malignancy lesions and need to
    recent studies). Surgery is the mainstay of treatment, but chemo- be differentiated from more aggressive tumours. Their management
    therapy may be needed to reduce large invasive masses in very is often problematic: for a long time, surgical resection has consti-
    young children. The recommended first-​line chemotherapy is the tuted the mainstay of treatment, but increased knowledge of their
    combination of vincristine and actinomycin D, with the addition of biological behaviour and the availability of active agents has pro-
    alkylating agents or anthracylines in case of no response. moted a multidisciplinary approach.
    Despite good prognosis, mutilating surgery could not be avoided Aggressive fibromatosis (AF), also known as desmoid tumour,
    in 6% of patients in a recent EpSSG study, and toxicity due to chemo- is a rare, deep-​seated, fibroblastic tumour that arises in musculo-​
    therapy was not negligible. For this reason, new agents targeting aponeurotic structures, mainly in extremities and deep structures
    the NTRK fusion gene seem a promising alternative approach to in the abdomen and pelvis. Tumours usually grow fairly slowly
    standard chemotherapy. but diffusely along muscle bundles and fascial planes, lacking a
    pseudocapsule, which contributes to the difficulty of obtaining clear
    Desmoplastic small round-​cell tumour margins when the tumour is resected. The pathogenesis of AF is most
    likely multifactorial: genetic predisposition, endocrine factors, and
    DSRCT is a very aggressive neoplasm, identified by the specific trauma all seem to play an important part. The incidence is higher
    translocation t(11;22)(p13;q12), with the chimeric transcript EWS-​ in families with familial adenomatous polyposis (FAP) within which
    WT1. It usually arises in the abdominal cavity and is often dissem- there is a particular association with intra-​abdominal fibromatosis
    inated at diagnosis, with peritoneal seeding (sometimes not clearly (Gardner’s syndrome).
    visible on imaging investigations) and liver metastases. OS rates at ten years are above 90%, but many patients experience
    Intensive multimodality treatments (with various drug com- long-​term morbidity due to the tumour growth or, in some cases, to
    binations and including high-​dose chemotherapy with stem-​cell the treatment (i.e. repeated aggressive resections).
    rescue) and hyperthermic intraperitoneal chemotherapy (HIPEC) Surgery retains an important role in the treatment, but micro-
    have been investigated, but prognosis remains poor. Adjuvant radio- scopically or macrospically incomplete resections are of little use be-
    therapy as part of multimodal treatment also seems beneficial. New cause AF usually rapidly regrows, probably stimulated by the tissue
    tailored therapeutic approaches are urgently needed for this tumour. trauma caused by surgery. The possibility that AF can stabilize or
    even reduce its dimensions, spontaneously or after medical treat-
    Malignant rhabdoid tumour ment, has largely diminished the use of aggressive measures like sur-
    Extracranial malignant rhabdoid tumours (MRT) are rare and often gery or radiotherapy. However, radiotherapy can be of use in case of
    affect very young children. They may occur sporadically or as part relapse and after second surgery, especially in older children.
    of a cancer syndrome known as rhabdoid tumour predisposition Various pharmacological treatments have proved to be relatively
    syndrome. Although the kidney is the most common site, MRT effective: the most-​used strategy is the administration of prolonged
    can arise anywhere in the body. It is characterized by alterations of low-​dose chemotherapy comprising combinations of weekly low-​
    SMARCB1 (also known as INI1), a tumour suppressor gene located dose methotrexate plus a vinca alkaloid (vinblastine or vinorelbine).
    on chromosome 22. Germline mutations or deletions in one allele Other possible options include hydroxyurea, anti-​ inflammatory
    are the basis of the predisposition syndrome, and somatic loss or drugs, tamoxifene (although this is rarely used in children because
    mutation of the other allele causes MRT or other tumours. of the side effects), or, more recently, imatinib and tyrosine kinase
    Histological diagnosis was mainly based on the recognition inhibitors like sorafenib. Studies, especially in children, are lacking.
    of rhabdoid cells, but this aspect is not always present. The diag- The observation that desmoid tumours may remain stable for a
    nostic landscape changed after the discovery that the inactivation long time, with or without primary treatment, has prompted the
    of SMARCB1 causes a loss of protein expression in the nucleus of suggestion that a ‘wait-​and-​see’ strategy (i.e. clinical–​radiological
    the cells, which can be detected by immunohistochemistry. This has monitoring alone) may be suitable in cases where tumour growth
    Non-rhabdomyosarcoma soft-tissue sarcomas 217

    is slow or static. All this demonstrates the difficulty of treating AF Chisholm JC, et al. (2017) Open-​ label, multicentre, randomised,
    that may become a chronic disease that often requires a personal- phase II study of the EpSSG and the ITCC evaluating the addition
    ized approach. of bevacizumab to chemotherapy in childhood and adolescent pa-
    Inflammatory myofibroblastic tumour (IMT) generally affects tients with metastatic soft tissue sarcoma (the BERNIE study). Eur
    the first two decades of life, can occur at any body site (with vis- J Cancer 83, 177–​84.
    Crist WM, et al. (2001) Intergroup rhabdomyosarcoma study-​IV: results
    ceral tumours more frequent in children), and is often accompanied
    for patients with nonmetastatic disease. J Clin Oncol 19, 3091–​102.
    by an inflammatory syndrome characterized by fever, malaise, and
    Davicioni E, et al. (2010) Gene expression profiling for survival predic-
    raised inflammatory markers. It was considered a reactive lesion but
    tion in pediatric rhabdomyosarcomas: a report from the Children’s
    monoclonality and the presence of a specific genetic alteration sup- Oncology Group. J Clin Oncol 28, 1240–​6.
    port its neoplastic nature. The finding that a substantial part of IMT Ferrari A, et al. (2002) Paratesticular rhabdomyosarcoma: report
    presents ALK rearrangements has supported the use of ALK inhibi- from the Italian and German Cooperative Group. J Clin Oncol 20,
    tors like crizotinib, with good results. In the case of ALK-​negative 449–​55.
    IMT, sarcoma-​like regimens including intensive chemotherapy with Ferrari A, et al.(2011) Non-​metastatic unresected paediatric non-​
    vincristine, actinomycin D, and alkylating agents, or prolonged ad- rhabdomyosarcoma soft tissue sarcomas: results of a pooled ana-
    ministration of low-​dose chemotherapy with vincristine–​etoposide lysis from United States and European groups. Eur J Cancer 47,
    or vinorelbine and methotrexate can be used. It is to be noted that 724–​31.
    cases of remission have been reported after anti-​inflammatory or Ferrari A, Casanova M (2005) New concepts for the treatment of pedi-
    steroid therapy. atric non-​rhabdomyosarcoma soft tissue sarcomas. Expert Rev
    Dermatofibrosarcoma protuberans (DFP) is a rare entity (0.10 per Anticancer Ther 5, 307–​18.
    Ferrari A, et al. (2013) Fibroblastic tumors of intermediate malignancy
    100,000 children aged 0–​19 years) that occurs especially in the trunk
    in childhood. Expert Rev Anticancer Ther 13(2), 225–​36.
    and extremities. Surgery is the treatment of choice and survival is
    Ferrari A, et al. (2015) Synovial sarcoma in children and adoles-
    above 95%. Unfortunately, a complete conservative resection may be cents: the European Pediatric Soft Tissue Sarcoma Study Group
    difficult due to tumour dimensions and localization in difficult sites. prospective trial (EpSSG NRSTS 2005). Ann Oncol 26, 567–​72.
    DFP is characterized by the specific translocation t(7;22) that ultim- Hawkins DS, et al. (2014) What is new in the biology and treatment of
    ately causes tumour cell growth, activating the PDGFRbeta protein pediatric rhabdomyosarcoma? Curr Opin Pediatr 26(1), 50–​6.
    tyrosine kinase. This mechanism can be interrupted by the adminis- Klingebiel T, et al. (2008) Treatment of children with metastatic soft
    tration of imatinib that has demonstrated interesting activity both in tissue sarcoma with oral maintenance compared to high dose
    adults and children with DFP chemotherapy: report of the HD CWS-​ 96 trial. Pediatr Blood
    Cancer 50, 739–​45.
    Martelli H, et al. (1999) Conservative treatment for girls with
    FURTHER READING nonmetastatic rhabdomyosarcoma of the genital tract: a report
    Admiraal R, et al. (2010) High-​dose chemotherapy for children and from the Study Committee of the International Society of Pediatric
    young adults with stage IV rhabdomyosarcoma. Cochrane Database Oncology. J Clin Oncol 17, 2117–​22.
    Syst Rev (12) CD006669. doi: 10.1002/​14651858.CD006669.pub2 Meza JL, et al. (2006) Analysis of prognostic factors in patients with
    Alaggio R, Coffin CM (2015) The evolution of pediatric soft tissue sar- nonmetastatic rhabdomyosarcoma treated on intergroup rhabdo-
    coma classification in the last 50 years. Pediatr Dev Pathol 18(6), myosarcoma studies III and IV: the Children’s Oncology Group. J
    481–​94. doi: 10.2350/​15-​07-​1666-​MISC.1 Clin Oncol 24, 3844–​51.
    Arndt CA, et al. (2009) Vincristine, actinomycin, and cyclophospha- Oberlin O, et al. (2008) Prognostic factors in metastatic rhabdo-
    mide compared with vincristine, actinomycin, and cyclophospha- myosarcomas: results of a pooled analysis from United States and
    mide alternating with vincristine, topotecan, and cyclophosphamide European cooperative groups. J Clin Oncol 26, 2384–​9.
    for intermediate-​ risk rhabdomyosarcoma: Children’s Oncology Okcu MF, et al. (2003) Synovial sarcoma of childhood and adoles-
    Group study D9803. J Clin Oncol 27, 5182–​8. cence: a multicenter, multivariate analysis of outcome. J Clin Oncol
    Bisogno G, et al. (2009) Sequential high-​dose chemotherapy for chil- 21, 1602–​11.
    dren with metastatic rhabdomyosarcoma. Eur J Cancer 45, 3035–​41. Orbach D, et al. (2016) Conservative strategy in infantile fibrosarcoma
    Bisogno G, et al. (2018) Addition of dose-intensified doxorubicin to is possible: the European Paediatric Soft Tissue Sarcoma Study
    standard chemotherapy for rhabdomyosarcoma (EpSSG RMS Group experience. Eur J Cancer 57,1–​9.
    2005): a multicentre, open-label, randomised controlled, phase 3 Orbach D, et al. (2017) The EpSSG NRSTS 2005 treatment protocol
    trial. Lancet Oncol 19(8), 1061–71. for desmoid-​type fibromatosis in children: an international pro-
    Bisogno G, et al. (2019) Vinorelbine and continuous low-dose cyclo- spective case series. Lancet Child Adolesc Health 1(4), 284–​92.
    phosphamide as maintenance chemotherapy in patients with high- Pappo AS, et al. (2005) Phase II trial of neoadjuvant vincristine,
    risk rhabdomyosarcoma (RMS 2005): a multicentre, open-label, ifosfamide, and doxorubicin with granulocyte colony-​stimulating
    randomised, phase 3 trial. Lancet Oncol 20(11), 1566–75. factor support in children and adolescents with advanced-​stage
    Breitfeld PP, Meyer WH (2005) Rhabdomyosarcoma: new windows of nonrhabdomyosarcomatous soft tissue sarcomas: a Pediatric
    opportunity. Oncologist 10, 518–​27. Oncology Group Study. J Clin Oncol 23, 4031–​8.
    Carli M, et al. (2005) Malignant peripheral nerve sheath tumours in Rosenberg AR, et al. (2014) Early response as assessed by anatomic
    pediatric age: a report from the Italian and German Soft Tissue imaging does not predict failure-​free survival among patients with
    Sarcoma Cooperative Group. J Clin Oncol 23, 8422–​30. Group III rhabdomyosarcoma: a report from the Children’s Oncology
    Casanova M, et al. (2004) Vinorelbine and low-​dose cyclophosphamide Group. Eur J Cancer 50(4), 816–​23. doi: 10.1016/​j.ejca.2013.11.031
    in the treatment of pediatric sarcomas: pilot study for the upcoming Shern JF, et al. (2015) Pediatric rhabdomyosarcoma. Crit Rev Oncog
    European Rhabdomyosarcoma Protocol. Cancer 101, 1664–​71. 20(3–​4), 227–​43.
    218 CHAPTER 25 Soft-Tissue Sarcomas

    Skapek SX, et al. (2013) PAX-​FOXO1 fusion status drives unfavour- van der Graaf WT, et al. (2017) Soft tissue sarcomas in adolescents and
    able outcome for children with rhabdomyosarcoma: a children’s young adults: a comparison with their paediatric and adult counter-
    oncology group report. Pediatr Blood Cancer 60(9), 1411–​17. parts. Lancet Oncol 18(3), e166–​75.
    doi: 10.1002/​pbc.24532 Van Rijn RR, et al. (2008) Imaging findings in noncraniofacial child-
    Spalding AC, et al. (2013) The effect of radiation timing on patients hood rhabdomyosarcoma. Pediatr Radiol 38(6), 617–​34.
    with high-​risk features of parameningeal rhabdomyosarcoma: an Wachtel M, Schäfer BW (2010) Targets for cancer therapy in childhood
    analysis of IRS-​IV and D9803. Int J Radiat Oncol Biol Phys 87(3), sarcomas. Cancer Treat Rev 36, 318–​27.
    512–​16. doi: 10.1016/​j.ijrobp.2013.07.003 Weigel BJ, et al. (2016) Intensive multiagent therapy, including dose-​
    Stevens MC (2005) Treatment for childhood rhabdomyosarcoma: the compressed cycles of ifosfamide/​etoposide and vincristine/​doxo-
    cost of cure. Lancet Oncol 6, 77–​84. rubicin/​cyclophosphamide, irinotecan, and radiation, in patients
    Stevens MC, et al. (2005) Treatment of nonmetastatic rhabdomyosar- with high-​risk rhabdomyosarcoma: a report From the Children’s
    coma in childhood and adolescence: third study of the International Oncology Group. J Clin Oncol 34(2), 117–​ 22. doi:10.1200/​
    Society of Paediatric Oncology—​SIOP Malignant Mesenchymal JCO.2015.63.4048
    Tumor 89. J Clin Oncol 23, 2618–​28. Williamson D, et al. (2010) Fusion gene-​negative alveolar rhabdomyo-
    Sultan I, Ferrari A (2010) Selecting multimodal therapy for rhabdo- sarcoma is clinically and molecularly indistinguishable from em-
    myosarcoma. Expert Rev Anticancer Ther 10, 1285–​301. bryonal rhabdomyosarcoma. J Clin Oncol 28, 2151–​8.
    26
    Bone Tumours
    Stefan Bielack, Michael Paulussen, and Lee Helman

    Epidemiology administered. Exposure to alkylating agents may also contribute.


    The use of radiation and alkylating-​agent chemotherapy, together
    Osteosarcoma and Ewing sarcoma (ES), including peripheral with genetic tumour predisposition, make osteosarcoma one of the
    neuroectodermal tumour (PNET) and Askin tumours of the chest more frequent secondary solid malignancies following therapy for
    wall, account for the majority of malignant bone tumours in chil- childhood cancer.
    dren and adolescents. In both, males are affected approximately 1.5
    times as often as females. The median age at presentation is around
    age 15 years and at least half of all diagnoses are made in the second Clinical presentation
    decade of life.
    Primary osteosarcoma is usually located in the metaphyses of
    long extremity bones, particularly the distal femur or the prox-
    Incidence imal tibia, followed by the proximal humerus. In total, two thirds
    of all paediatric osteosarcomas arise around the knee (Figure 26.1).
    Osteosarcoma is the most frequent primary bone cancer. Its ap- Osteosarcomas of the axial skeleton or craniofacial bones generally
    proximate annual incidence is 2–​3 per million in the general popu- occur in older patients. Sites most frequently affected by ES of bone
    lation, peaking at 8–​11 per million at age 15–​19. A second, smaller are the pelvis, long bones of the extremities, ribs, scapulae, and ver-
    peak in older patients is due to osteosarcoma arising in abnormal tebrae. Compared to osteosarcoma, involvement of flat bones of the
    bones, such as those affected by Paget disease or prior irradiation. In trunk is much more common (Figure 26.1). If ES arises in long ex-
    paediatric and adolescent populations of Caucasian descent, Ewing tremity bones, it usually does so in the diaphysis, in distinction to
    sarcoma is the second most common primary malignancy of bone. the typical metaphyseal presentation of osteosarcoma.
    Annual incidence rates are approximately 2–​3 per million in chil- At diagnosis, accurate staging will reveal metastases in 10–​30%
    dren, adolescents, and young adults below the age of 25. It is exceed- of the patients. Without systemic therapy, however, most patients
    ingly rare in African and Chinese populations. with seemingly localized bone sarcomas will develop metachronous
    metastases within one to two years. In osteosarcoma, primary and
    secondary metastases are limited to the lungs in 80% of affected in-
    Predisposition and genetic links dividuals. The remainder have bone metastases with or without add-
    itional pulmonary involvement. Skip metastases, defined as isolated
    In most paediatric and adolescent patients, the aetiology remains tumour foci within the same bone as the primary tumour, occur in
    obscure. Trauma has been implicated but little evidence exists to a minority of patients. In ES, lungs and bones contain metastatic
    support this relationship. There are no known predisposing fac- deposits at diagnosis in approximately equal proportions (around
    tors for ES. The predilection of osteosarcoma for the age of the 10% of patients each), and there may be bone-​marrow involvement
    pubertal growth spurt and sites of maximum growth suggest a cor- (around 5%). In both diseases, lymph-​node metastases are rare and
    relation with bone growth. The incidence of osteosarcoma is in- other sites of metastatic disease are almost never involved at initial
    creased in several disorders associated with germline alterations of diagnosis.
    tumour suppressor genes and these may account for 10% or more Patients with bone tumours usually do not feel ill until late in the
    of all osteosarcomas. Survivors of hereditary retinoblastoma with course of their disease. They typically seek medical attention because
    germline mutations of the retinoblastoma gene RB1 carry a risk that of, at first, intermittent and then continuous pain, often erroneously
    is 500–​1,000 times greater than in the general population. The Li–​ attributed to recent trauma. Tumour-​related swelling and loss of
    Fraumeni cancer family syndrome, caused by germline mutations in function of adjacent joints usually develop later. In approximately
    the p53 gene, is associated with a 15-​fold increase. Radiation is an- 10%, the first sign of disease is a pathological fracture. Pain at bony
    other risk factor for osteosarcoma, the risk being related to the dose sites other than the primary may represent metastatic involvement.
    220 CHAPTER 26 Bone Tumours

    Osteosarcoma Ewing
    Craniofacial 0.9% 2%
    Clavicle 0.1% 2%
    Scapula 0.2% 4%
    Rib 0.4% 10%
    Sternum – 0.5%
    Proximal 10%

    Humerus 10% 6%

    Spine 0.4% 6%
    Ulna 0.4% 1%
    Radius 1% 1%
    Pelvis 3% 26%

    Hand 0.1% 1%

    Femur 49% 20%

    Distal 44%

    Proximal 24%

    Tibia 28% 10%

    Foot 0.3% 3%
    Fibula 5% 8%

    Figure 26.1 Skeletal distribution of osteosarcoma and Ewing sarcoma.


    Source: data from 1,791 primary high-​grade central osteosarcomas and 1,426 ES from Cooperative Osteosarcoma Study Group (COSS) and European Intergroup Cooperative
    Ewing’s Sarcoma Study ((EI)CESS), respectively. The percentages shown for proximal, diaphyseal, and distal-​extremity locations are for osteosarcoma.

    Metastases in the lungs produce respiratory symptoms only with ex-


    tensive involvement. Systemic symptoms, particularly fever, may be
    present in ES.
    (a) (b)

    Diagnosis

    Most patients present with a history of pain of the involved region,


    and physical examination is typically remarkable only for a mass at
    the primary site. Loss of motion of neighbouring joints, infiltration
    of the skin, and neurological deficits may occur, depending on the
    location and extent of the tumour. There are no laboratory tests that
    can identify patients with malignant bone tumours. Blood tests may
    show moderately elevated erythrocyte sedimentation rates and may
    reveal some degree of anaemia and leucocytosis. Elevated serum
    levels of lactate dehydrogenase or alkaline phosphatase may cor-
    relate with tumour burden or bone destruction, or, in osteosarcoma,
    with tumour phenotype, and can thus be of prognostic significance.

    Imaging of the primary tumour


    Plain radiographs and cross-​sectional imaging, preferably by mag-
    netic resonance imaging (MRI), are required to assess tumour size
    and its loco-​regional extension.
    Plain radiographs are used to describe the bony compartment.
    Osteosarcoma can present with lytic or sclerotic changes, or both.
    Figure 26.2 Conventional radiograph (left) and T1, gadolinium-​
    Ossification in the soft tissues in a radial or ‘sunburst’ pattern is typ-
    enhanced, fat-​suppressed MRI (right) of an osteosarcoma of the right
    ical, but neither sensitive nor specific. Periosteal bone formation proximal tibia. X-​ray shows cortical destruction and Codman’s triangle on
    with lifting of the cortex may lead to the appearance of Codman’s tri- the lateral aspect of the tumour.
    angle (Figure 26.2). The typical radiograph of ES shows a destructive Reproduced courtesy of T. Von Kalle, Klinikum Stuttgart, Germany
    Histopathology 221

    osteolytic lesion of the diaphysis with destruction of the osseous


    Differential diagnosis
    cortex, elevation of the periosteum, and infiltration of surrounding
    soft tissue. The layered periosteal reaction may give rise to a classic
    The differential diagnosis of osteosarcoma includes traumatic le-
    ‘onion-​skin’ appearance. MRI is the most useful tool to define the
    sions, osteomyelitis, benign tumours such as exostosis, fibroma,
    intramedullary tumour extent, its soft-​tissue component, and the re-
    osteoid osteoma, chondroma, giant-​cell tumour of bone, bone cysts,
    lation to vessels and nerves (Figure 26.2).
    as well as other primary malignant lesions of bone such as ES or
    Serial MRIs can also be used to monitor response to induction
    lymphoma, and metastases from malignancies such as neuro-
    chemotherapy. In ES, a marked shrinkage of the soft-​tissue compo-
    blastoma or soft-​tissue sarcoma. The most important clinical differ-
    nent can be used as a surrogate marker for response. In osteosar-
    ential diagnosis of ES is osteomyelitis. The radiological appearance
    coma, where the osteoid matrix often prevents tumour shrinkage,
    may be very similar and, on occasion, ES may be secondarily in-
    response can be predicted based upon the results of serial dynamic
    fected. On histological examination, ES must be differentiated from
    contrast-​enhanced MRIs. Other methods used to predict response
    other small round-​cell tumours, in particular embryonal rhabdo-
    preoperatively include serial evaluation by angiography, quantitative
    myosarcoma, neuroblastoma, small-​cell osteosarcoma, and non-​
    bone scans, or positron emission tomography (PET).
    Hodgkin lymphoma.

    Systemic staging Histopathology


    The search for metastases focuses on those sites where the majority
    By definition, osteosarcoma is a mesenchymal malignancy in which
    occur: lungs and bones. The bone marrow is evaluated in ES. Plain
    the malignant cell population produces osteoid, but production
    radiographs and a computed tomography (CT) scan of the thorax
    can vary considerably. Both abundant production, leading to hard
    must be performed to rule out pulmonary metastases. Bone metas-
    sclerotic tumours, and very scanty production of osteoid are con-
    tases are searched for by a 99mTc-​MDP bone scan. Skip metastases
    sistent with the diagnosis. Conventional osteosarcoma, a high-​grade
    may also be visualized on the bone scan, but MRI of the whole bone
    central malignancy of bone, accounts for 80–​90% of all osteosar-
    is more sensitive. Whole-​body MRI may lead to a higher detection
    comas. Its most frequent subtypes are osteoblastic, chondroblastic,
    rate of bone metastases, but its exact role remains to be defined.
    and fibroblastic, but various less frequent subtypes are also con-
    PET/​CT can detect bone metastases with a higher sensitivity than
    sidered conventional osteosarcomas. Conventional, telangiectatic,
    bone scans, but is also less specific.
    high-​grade surface, and small-​cell osteosarcomas all have a similar
    Ewing sarcoma cells may be inhomogeneously distributed within
    clinical course and are treated by multimodal regimens that include
    the bone marrow. It is therefore recommended that marrow samples
    chemotherapy. Low-​grade central, parosteal, and perisosteal osteo-
    be obtained from several sites distant from the primary. In addition
    sarcomas are treated by surgery only. Craniofacial osteosarcomas,
    to standard pathology assessment, marrow micro-​metastases can be
    apart from those of the skull, metastasize less frequently than con-
    detected by reverse transcriptase polymerase chain reaction (RT-​
    ventional osteosarcomas. Never the less, the trend is to treat them
    PCR) for tumour-​specific ES gene (EWS) fusion transcripts. While
    by surgery plus chemotherapy. Extraskeletal osteosarcomas, usually
    involvement detectable by light microscopy carries adverse prog-
    high-​grade malignancies, are exceedingly rare in young people.
    nostic implications, assessment of micro-​metastatic dissemination
    ES of bone are composed of firm, grey-​white soft tissue with a
    by RT-​PCR remains investigational.
    glistening, moist appearance on sectioning. Macroscopically, the
    intraosseous component of the tumour is usually firm, while the
    extraosseous component tends to be less firm with areas of haemor-
    Biopsy
    rhage and cystic degeneration secondary to tumour necrosis. All ES
    are high-​grade malignancies. When stained with haematoxylin and
    Imaging will often result in a high index of suspicion, but diagnosis
    eosin, they appear as a monomorphic small, blue, round, primitive-​
    must always be verified histologically. In order to ensure appro-
    cell population, with round nuclei and scanty cytoplasm. Variable
    priate biopsy techniques and optimal evaluation of the obtained
    amounts of glycogen deposition stain periodic-​acid-​Schiff (PAS)
    material, it is strongly recommended that biopsies should be per-
    positive. Peripheral primitive neuroectodermal tumours (pPNET)—​
    formed in specialized paediatric bone-​tumour centres. Recent
    differing from other ES by showing features of neural differentiation
    years have witnessed a shift from open to tru-​cut needle biop-
    with prominent neurite-​like cell processes containing neurosecretory
    sies. Whichever technique is chosen, sufficient material must be
    granules and neurofilaments on electron microscopy, or rosettes and
    obtained for both histological evaluation and ancillary studies. The
    Homer–​ Wright pseudorosettes and/​ or expressing S-​ 100 protein,
    biopsy specimen should be forwarded to the pathologist without
    neuron-​specific enolase (NSE), or synaptophysin—​were historically
    prior fixation. Tissue is required for standard histology and
    distinguished from ES, but share the same molecular biology and are
    immunohistochemistry, and fresh/​frozen samples are required
    now regarded as part of the ES family of tumours.
    for molecular biology studies. The biopsy track and adjacent tis-
    ES cannot be differentiated from other PAS-​positive small, blue,
    sues must be regarded as tumour-​contaminated and the approach
    round-​cell tumours on morphological features alone, and the diagnosis
    should be planned so that these can subsequently be included into
    must be assisted by immunocytochemistry and molecular biology (see
    definitive local treatment. If metastases are suspected, but there is
    the section ‘Biology’). CD99 (Mic-​2 antigen) expression is positive in
    no unequivocal evidence from imaging, histological confirmation
    more than 95% of cases but is not unique for ES. Other useful markers
    should be attempted.
    222 CHAPTER 26 Bone Tumours

    include vimentin, desmin, smooth-​muscle actin, and CD45 (leukocyte


    Binds to RNA
    common antigen). These help to differentiate between small round-​cell
    tumours of myogenic, fibrogenic, and haematopoetic origin.
    EWS COOH chr22
    NH2
    Transactivation domain

    Biology
    EWS-FLI1
    Biology of osteosarcoma chimera NH2
    COOH
    abn22
    Although germline mutations of p53 and RB are rare, these genes are
    altered in many osteosarcoma tumour samples. Recent data from
    next-​generation sequencing analysis of osteosarcoma tumour speci- DNA binding domain
    FLI1 COOH
    mens demonstrates that approximately 95% of these tumours have NH2

    p53 pathway alterations, with the majority of these alterations con-


    sisting of either sequence alterations or structural variants within Binds to DNA
    the p53 gene, with a much smaller number of cases having MDM2
    amplifications. This data strongly suggests that p53 alterations are chr11

    the major oncogenic driver in these tumours. Intriguingly, these Figure 26.3 Chromosomal translocation t(11;22)(q24;q12) in Ewing
    tumours also have the highest known rate of chromothripsis, best sarcoma. This leads to the generation of a DNA-​binding fusion protein.
    described as a single catastrophic event resulting in massive chromo- The gene for the RNA-​binding protein EWS is located on chromosome
    22 (chr22) in the region q12, the gene for the DNA-​binding transcription
    somal rearrangements. Also common is a phenomenon known as
    factor FLI1 on chromosome 11 (chr11) at q24. The EWS-​FLI1 chimeric
    kataegis, that is localized hypermutated chromosomal regions. protein is expressed from the abnormal chromosome 22 (abn22), while the
    These events are no doubt the reason that the hallmarks of osteo- reciprocal translocation product on chromosome 11 has no gene product.
    sarcoma are a chaotic-​appearing karyotype and aneuploidy. Beyond Reproduced courtesy of Heinrich Kovar, St. Anna Kinderkrebsforschung Children’s
    p53, alterations are seen in RB in approximately 50% of tumours and Cancer Research Institute, Vienna, Austria
    there are also, frequently, alterations in ATRX. Consequently, loss of
    function of the p53 and RB tumour-​suppressor genes, which regu- in other solid tumours and in acute leukaemia, the specific combin-
    late cell-​cycle progression, are believed to have important roles in ation of EWS with an ETS gene is restricted to, and therefore diag-
    osteosarcoma tumourigenesis. Rothmund–​Thomson syndrome and nostic of, ES. EWS–​ETS fusion is essential for ES pathogenesis.
    Bloom syndrome, rare conditions caused by mutations in tumour-​ The classical cytogenetic demonstration of t(11;22)(q24;q12) has
    suppressor genes coding for helicases, are also associated with an been largely replaced by fluorescence in situ hybridization (FISH)
    increase in osteosarcoma, as is Werner syndrome (adult progeria). using probes flanking the EWS breakpoint region on chromosome
    Numerous oncogenes are also altered in osteosarcoma tumour 22. Using this method, even complex chromosomal rearrangements
    cells. These include amplifications of the product of the murine double and rare EWS translocations with other ETS partner genes can be
    minute 2 gene, amplification of cyclin-​dependent kinase 4, and over-​ revealed. However, the most frequently used method to detect the
    expression of human epidermal growth factor receptor 2. Although fusion gene is RT-​PCR. This requires immediate snap freezing of
    it is clear that alterations in tumour-​suppressor genes and oncogenes the biopsy since it relies on extraction of good-​quality RNA. The
    are necessary to produce osteosarcomas, it is not clear which of these strength of the method lies in its high sensitivity, enabling identifica-
    events occurs first and why or how it occurs. Moreover, it is not clear tion of single tumour cells among 105–​106 normal cells.
    which, if any, of the alterations are essential to tumour development Recently, it has become clear that at least two distinct Ewing-​like
    and might therefore represent therapeutic targets. tumours exist that histologically appear very similar to Ewing tu-
    mours but lack the EWS/​FUS-​ETS fusion genes. The most common
    Biology of Ewing sarcoma
    entity is a tumour harbouring CIC-​DUX4/​DUX4L fusion. These are
    The molecular key to ES is the rearrangement of the ES gene EWS most frequently soft-​tissue tumours and, clinically, appear to be-
    on chromosome 22q12 with the FLI-​1 gene on 11q24. EWS encodes have quite aggressively. Less common are the tumours associated
    a ubiquitously expressed RNA-​binding protein which appears to be with BCOR-​CCNB3 fusions, that appear to occur predominantly in
    involved in transcription and processing of messenger RNA. FLI-​1 males. To date, most of these tumours are treated with Ewing-​type
    codes for a protein with a carboxy-​terminal DNA-​binding domain
    that is characteristic for members of the ETS transcription factor
    Table 26.1 Gene fusions in Ewing sarcoma
    family. As a consequence of gene fusion, the EWS RNA binding do-
    main is replaced by the FLI-​1-​derived ETS DNA binding domain, Gene fusion Cytogenetic equivalent Observed frequency
    resulting in a novel transcription factor with strong in vitro transacti- EWS–​FLI1 t(11;22)(q24;q12) 85%
    vation potential (Figure 26.3). The most frequent gene fusions in ES
    EWS–​ERG t(21;22)(q22;q12) 10%
    result in the type 1 (EWS exon 7 to FLI1 exon 6: 50% of cases) or type
    2 (EWS 7/​FLI1 5: 27%) fusions. In 10–​15%, alternative EWS gene fu- EWS–​ETV1 t(7;22)(p22;q12) <1%
    sions to other ETS gene family members (predominantly ERG) can EWS–​ETV4 t(17;22)(q12;q12) <1%
    be observed (Table 26.1). While EWS, as well as its close relatives TLS EWS–​FEV t(2;22)(q35;q12) 1%
    (FUS) and RPB56 (hTAFII68) (FET gene family for FUS, EWS, TAF), FUS-​ERG t(16;26)(p11;q22) <1%
    FUS-​FEV t(2;16)(q35;p11) <1%
    and ERG genes are involved in several chromosomal translocations
    Treatment 223

    therapy, although given their distinct molecular features, it is likely Osteosarcoma


    they should be treated and analysed separately. 1.0 1.0
    0.9 0.9
    0.8 0.8
    0.7 0.7
    Risk classification 0.6 0.6

    Survival

    Survival
    0.5 0.5
    0.4 0.4
    Staging 0.3 0.3
    The tumour–​node–​metastasis (TNM) classification for malignant 0.2 0.2
    0.1 0.1
    bone tumours is an expansion of a staging system originally devel- 0.0 0.0
    oped by the Musculoskeletal Tumor Society (MSTS). In extremity, 0 5 10 15 20 0 5 10 15 20
    Years Years
    trunk, and craniofacial locations, T1 describes a tumour 8 cm or
    less in dimension, T2 has the greatest dimension of more than
    8 cm, while T3 is used for discontinuous tumours in the primary Ewing sarcoma
    bone site, skip metastases. (For the first time, the eighth edition of
    1.00 No pMet : N = 477
    the TNM classification provides distinct T definitions for the rarer Lung pMet : N = 68
    pelvic and vertebral tumour sites, which are not elaborated upon in Other pMet : N = 93

    Estimated probability
    this chapter.) Tumour stages are categorized by grade and anatom- 0.75
    0.60
    ical extent, with stage I reserved for localized low-​grade sarcomas 0.53
    (IA T1; I B T2,3); stage II describing localized high-​grade tumours 0.50 0.42
    (II A T1; II B T2); and stage III, high-​grade sarcomas with skip me- 0.32
    tastases (T3). Stage IV A describes bone sarcomas with lung metas- 0.25
    tases (M1a); stage IV B tumours are those with lymph node (N1) or 0.26 0.20
    extrapulmonary distant metastases (M1b), regardless of grade.
    0.00
    0 1 2 3 4 5 6 7 8 9 10
    Prognostic factors Time in years p = 0.0001
    Several prognostic factors have been identified in osteosarcoma.
    Figure 26.4 Osteosarcoma (top): left—​overall survival with multimodal
    Primary metastases, axial or proximal extremity location, and large tu- therapy; right—​comparison of good responders (solid line) and poor
    mour size are of independent negative prognostic value. Factors asso- responders (dashed line); good response = <10% viable tumour. Results
    ciated with an adverse outcome in some series also include very young are from 1,152 COSS patients under 20 years of age with localized
    or older age, high levels of alkaline phosphatase or lactic dehydro- extremity osteosarcoma. Ewing sarcoma (bottom): event-​free survival
    genase, and the immunohistochemical detection of p-​glycoprotein by disease stage at initial presentation. Without primary metastases
    (no pMet), with lung or pleural metastases only (lung pMet), and
    (still debatable: HER2/​erbB2). The relative risk associated with any with metastases to bones and/​or bone marrow (other pMet). (CESS/​
    presenting factor, however, is lower than that of two treatment-​related EICESS data)
    variables: incomplete surgery is the most important negative prog-
    nostic indicator, and response to induction chemotherapy is the most
    important prognostic factor for resectable osteosarcoma. Response, suboptimal initial therapy can irrevocably compromise a patient’s
    however, is not all or nothing. Even a very moderate response is asso- chances, and patients should be treated in specialized, experienced,
    ciated with better outcomes than none at all. paediatric bone-​tumour centres able to provide access to the full
    The classical factors related to poor prognosis in ES include pres- diagnostic and therapeutic spectrum. Treatment within prospective
    ence of metastases, older age, large tumours, and primary site in the clinical trials is considered standard clinical practice in many coun-
    trunk and pelvis. Specific prognostic factors that currently influence tries. Multinational trials such as the European and American
    the choice of treatment regimens include initial tumour size, the pres- Osteosarcoma Study EURAMOS-​1 or the studies of the Children’s
    ence of pulmonary and/​or extrapulmonary metastases, the presence Oncology Group (COG) and EURO-​EWING groups are recent
    of tumour cells at the surgical margins, and poor histological response examples.
    to induction chemotherapy. The previously suggested prognostic im- Local treatment (surgery in osteosarcoma, surgery and/​or radio-
    portance of the specific EWS–​FLI1 gene fusion type has recently been therapy in ES) is an essential part of curative therapy. Most patients,
    refuted. As in osteosarcoma, the most clear-​cut treatment-​related however, will already have micro-​metastatic disease by the time
    prognostic indicator is tumour response to induction chemotherapy. their sarcoma is detected. Prior to the 1970s, when treatment was ex-
    Several large analyses have shown that early relapse (within two years clusively targeted towards the primary, the outcome was extremely
    from the time of initial diagnosis) is another predictor of death. poor. This dismal outlook was dramatically improved when multi-​
    agent chemotherapy was added to local treatment. Many trials have
    since reported disease-​free survival rates in the range of 50–​70%.
    A multimodal approach has therefore emerged as the undisputed
    Treatment standard of care.
    Currently, most institutions use an approach consisting of preopera-
    Treatment principles and strategy tive chemotherapy (also called neoadjuvant or induction chemo-
    Many children and adolescents can be cured with appropriate therapy), followed by definitive local therapy, and postoperative,
    therapy (Figure 26.4). Inappropriate use of diagnostic tools and adjuvant chemotherapy both for osteosarcoma and ES. Theoretical
    224 CHAPTER 26 Bone Tumours

    advantages of this approach include early treatment of micro-​ Table 26.2 Surgical margins in musculoskeletal oncology
    metastatic disease and facilitation of subsequent local therapy because
    Type Resection
    of shrinkage and decreased vascularity of the tumour. Furthermore,
    preoperative treatment permits tumour response to induction chemo- Intralesional Within the lesion
    therapy to be evaluated. A theoretical concern is that delayed removal Marginal Through the pseudocapsule or reactive tissue
    of the bulk tumour could lead to the emergence of chemotherapy Wide Lesion (including biopsy scar), pseudocapsule and/​
    resistance. Only one relatively small, randomized osteosarcoma or reactive zone, and an intact cuff of normal tissue
    completely surrounding the mass removed as a single
    trial has prospectively compared patients treated by both pre-​and block
    postoperative chemotherapy with patients treated by primary sur-
    Radical Entire anatomical compartment containing the tumour
    gery followed by the same chemotherapy. Treatment results did not removed as one block
    differ between the two arms. Similarly, the Cooperative Osteosarcoma
    Study Group could not detect a survival difference between both ap- Source: data from Enneking WF, Spanier SS, Goodman MA, (1980). ‘A system for the
    surgical staging of musculoskeletal sarcoma.’ Clinical Orthopaedics and Related Research.
    proaches in a retrospective comparison of 157 patients undergoing Volume 153, Nov–​Dec. pp. 106–​20. Wolters Kluwer Health Inc.
    primary surgery and 1,451 who received initial chemotherapy. As the
    delivery of initial chemotherapy also facilitates preparation for limb-​
    salvage surgery, as well as offering the value of measuring early re- soft tissues, and skin which will remain after wide resection; and the
    sponse, induction chemotherapy has become standard. patient’s growth expectancy must all be evaluated before deciding
    Local therapy upon the type of surgery. En bloc resection with limb salvage is
    possible in many cases. Amputation, however, may still be the most
    Choice of modality appropriate type of surgery for selected patients with unfavourable
    In osteosarcoma, complete surgical removal of the primary tu- tumour characteristics. Advantages of amputation include onco-
    mour and, if present, all metastases is the local treatment of choice. logical safety, low complication rates, and a low rate of revision sur-
    Indications for radiotherapy are limited and basically restricted to gery. Disadvantages include mutilation and phantom sensations.
    inoperable tumours. The functional outcome is often rather poor after proximal amputa-
    The radiosensitivity of ES was first recognized by James Ewing tions, but better with below the knee amputations.
    in1921 and has long played a major role. Currently, however, radio- Rotation-​plasty, used for sarcomas of the distal femur and, oc-
    therapy is no longer the local treatment of first choice for most ES. casionally, the proximal tibia, is the classic example of a resection-​
    Increasing awareness of the risk of local recurrence has encouraged reimplantation procedure. The knee is removed en bloc together
    the use of surgery, and its use was associated with a survival advan- with the tumour, the only structure left in situ being the popliteal
    tage in several series. Situations in which surgery is preferable to nerve bundle. The distal part of the tibia, together with the foot,
    irradiation include lesions in expendable bones, pathological frac- is then rotated by 180°, the tibia is fused with the femoral stump,
    tures, distal extremity sites, bulky primary tumours, and in patients and anastomoses between the femoral and tibial vessels are created.
    with poor response to initial chemotherapy. Preoperative radio- The result of this reimplantation is a shortened extremity, in which
    therapy may be indicated when there is tumour progression during the rotated foot substitutes for the knee and carries the prosthesis.
    chemotherapy, in emergencies such as spinal-​cord compression, or Advantages of rotation-​plasty include oncological safety, even in
    when incomplete surgery is anticipated. Postoperative radiotherapy very large tumours, lack of phantom sensations, and an infrequent
    is indicated after incomplete resection or, in the opinion of some, need for revision surgery, as well as an extremity function rivalling
    in patients with poor histological response to induction chemo- that of limb salvage. The highly unusual cosmetic appearance is its
    therapy. In one recent publication of the EURO-​EWING consor- main disadvantage, and the procedure has therefore fallen out of
    tium, postoperative radiotherapy reduced local relapse rates even in favour.
    a subset of patients with good chemotherapy response. Retaining the extremity requires reconstruction of bony and
    soft-​tissue defects. Only a minority of tumours are situated in ex-
    Principles of bone-​sarcoma surgery pendable bones such as the proximal fibula. Far more often, limb
    Bone-​sarcoma surgery has three aims. First and foremost, the tumour salvage implies replacement of a major joint. Allografts are used by
    must be completely removed. Second, the patient should be left with some surgeons, but modular endoprosthetic systems, assembled to
    good function. Third, it should, if possible, result in a cosmetically fit in the operating room, are employed more commonly. Autologous
    acceptable appearance. Complete tumour removal is of paramount bones, such as vascularized fibulae bridging diaphyseal defects, can
    importance; functional and cosmetic aspects are secondary goals. be used in selected situations. Special expandable endoprostheses
    The aim should be that the tumour, including the biopsy scar and are available for growing extremities. Innovative models can be
    biopsy track, should be removed together with an intact cuff of the lengthened non-​invasively. Unfortunately, such devices still carry
    surrounding normal tissue, corresponding to the definition of ‘wide’ a high complication risk with the associated need for frequent re-
    margins within the MSTS definition (Table 26.2). Wide and radical visions. Disadvantages of limb salvage in general already include
    margins are considered adequate, whilst marginal or intralesional a significant complication risk, to which infections, fractures, and
    margins are associated with an increased risk of local recurrence, prosthetic wear contribute. Several publications report local recur-
    which in turn carry a grave prognosis. rence rates approximately three times higher than after ablative pro-
    The location, size, and regional anatomy of the tumour; its rela- cedures, suggesting that resection margins may sometimes not be as
    tion to nerves and vessels as well as neighbouring joints, the bones, wide as expected.
    Treatment 225

    Principles of radiotherapy such as dexrazoxane and the reduction of anthracycline peak levels
    In osteosarcoma, radiotherapy is limited to few indications. by continuous doxorubicin infusions featuring most prominently.
    Postoperative irradiation is indicated when definitive surgery did Sequential studies by American and European groups suggest that
    not result in safe margins and in which revision surgery, even when there is no major loss of efficacy with these approaches, but no con-
    mutilating, could not do so. In these cases, local control may be at- trolled studies have been reported.
    tempted by irradiating with very high doses, often 60 Gy and above. Methotrexate, a folate antagonist, blocks the action of dihydrofolate
    Both proton and heavy-​ion radiotherapy are of interest and worthy reductase, the enzyme responsible for reducing folate to its active
    of investigation in inoperable osteosarcoma. form, tetrahydrofolic acid. In osteosarcoma, methotrexate is given
    In ES, whole-​bone irradiation had traditionally been advised at very high doses in the range of 8–​12 g/​m2. Toxicity must be an-
    because the tumour was thought to arise in the bone marrow, put- tagonized by the use of leucovorin, an activated folate. The treat-
    ting the whole marrow cavity at risk. The advent of MRI, which ment concept of high-​dose methotrexate is based on the assumption
    accurately demonstrated the extent of marrow involvement, and that normal cells can be rescued more effectively than tumour cells,
    of effective chemotherapy, questioned this approach. A random- which may lack active folate transporters. Observations published
    ized study undertaken by the North American Pediatric Oncology in 1972 that high-​dose methotrexate monotherapy was able to in-
    Group (POG) has demonstrated that radiotherapy to the initial tu- duce responses of pulmonary metastases led to its use as an adju-
    mour volume, plus a 2-​cm safety margin, produced results that were vant, first as monotherapy and, soon later, also as part of multidrug
    equivalent to that achieved by whole-​bone irradiation. Scars after regimens. A significant survival advantage conferred by high-​dose
    biopsy or tumour resection should be included in the radiation field. over moderate-​dose methotrexate-​containing combination chemo-
    To avoid constrictive fibrosis, an adequate strip of skin and sub- therapy was demonstrated in a randomized Italian study. However,
    cutaneous tissue should be left when irradiating limb sites, and epi- the assumption that osteosarcoma polychemotherapy regimens
    physeal plates should be spared if possible, particularly in children. which include methotrexate must necessarily be superior to other
    In inoperable ES, a compartment dose of 45 Gy is recommended, regimens has been repeatedly challenged, both in randomized set-
    with a tumour boost to at least 54–​55.8 Gy: whether this can be tings and in uncontrolled studies.
    achieved depends on the site of the tumour and the age of the pa- High-​dose methotrexate therapy requires meticulous attention
    tient. For preoperative radiotherapy, the standard target volume to detail and extensive supportive measures, including hydration,
    dose is 45 Gy. For postoperative radiotherapy, dosage recommenda- alkalinization of the urine, and leucovorin administration adapted
    tions depend on histological assessment of the resection margins to methotrexate serum levels. Inadequate supportive care will result
    and the tumour response to induction chemotherapy, and vary be- in severely delayed methotrexate clearance and excessive toxicity,
    tween 45 and 55.8 Gy. including myelosuppression, renal toxicity, and mucositis. Some
    patients will experience such toxicity despite adequate supportive
    care, the risk apparently increasing with increasing patient age. The
    Systemic therapy for osteosarcoma enzyme glucarpidase, which cleaves methotrexate, may be of benefit
    The majority of current treatment protocols are based upon varying in selected patients with renal failure and severely delayed metho-
    combinations of only four active agents: doxorubicin, cisplatin, trexate clearance, but most cases of delayed methotrexate clearance
    high-​dose methotrexate, and ifosfamide. Even after several decades can be handled by maintaining hydration and increasing leucovorin
    and numerous clinical trials, the exact individual role of each of dose and frequency.
    these agents and the optimal way in which they are to be combined Given that, under most circumstances, patients do not experience
    and delivered are still being debated, as is the potential benefit of significant myelotoxicity and given the resulting ability to schedule
    additional drugs. methotrexate at times when other agents cannot be administered,
    Doxorubicin was introduced into osteosarcoma treatment in most groups incorporate high-​dose methotrexate into their treat-
    the 1970s and has remained an integral part ever since. A meta-​ ment protocols. Protocols devoid of high-​dose methotrexate, but
    analysis of trials published until 1991 concluded that doxorubicin containing several of the other three ‘standard’ agents, can be used
    dose intensity is an important determinant of favourable outcome in healthcare settings which cannot guarantee the life-​saving sup-
    and that the dose intensities of other agents do not contribute as sig- portive measures without which the drug should not be prescribed.
    nificantly to outcome. However, the agent’s cardiotoxicity remains The activity of cisplatin against osteosarcoma was proven in early
    a major issue and multiple efforts aiming to reduce the incidence phase II trials. Cisplatin therapy requires supportive hyperhydration.
    and severity of this complication have been pursued. Prospective, Ototoxicity and nephrotoxicity are dose-​limiting, but both can be
    randomized studies comparing doxorubicin-​free with doxorubicin-​ reduced by administering the drug as a continuous infusion. Intra-​
    containing regimens have produced conflicting results. Most re- arterial administration of cisplatin into the vessel supplying the tu-
    cently, French investigators from the SFOP (Societe Francaise mour was previously tried but has largely been abandoned since
    d’Oncologie Pediatrique) OS94 trial reported that treatment could comparative trials failed to demonstrate enhanced anti-​tumour ef-
    be initiated with an anthracycline-​(and cisplatin-​) free preopera- fects compared to intravenous administration.
    tive regimen of high-​ dose ifosfamide and methotrexate, which Following positive phase II trials, ifosfamide has been part of
    would be continued postoperatively in the case of good response many osteosarcoma protocols since the mid-​ 1980s. Its efficacy
    and changed to doxorubicin-​based polychemotherapy only in the may be related to the dose administered. Supportive measures ne-
    case of poor histological response to anthracycline-​free induction. cessary to prevent haemorrhagic uropathy include hyperhydration
    Some doxorubicin-​based protocols include measures aimed to- and mesna (uromitexan). Ifosfamide may also lead to chronic renal
    wards reducing doxorubicin cardiotoxicity, with cardioprotectants tubular toxicity and to sterility.
    226 CHAPTER 26 Bone Tumours

    Data from two European trials (by the Istituto Rizzoli and the has been argued that further trials are needed before the agent might
    Cooperative Osteosarcoma Study Group (COSS)) supported the be considered for routine use.
    use of ifosfamide, but a randomized trial by the North American Adding the bisphosphonate zoledronic acid to chemotherapy was
    POG and Children’s Cancer Group (CCG) could not demonstrate investigated in a recent randomized French trial. The study was ter-
    that adding standard-​dose ifosfamide to high-​dose methotrexate, minated early for futility.
    doxorubicin, and cisplatin improved outcomes. The results of the
    Systemic therapy for Ewing sarcoma
    North American trial INT0133 should, however, be interpreted with
    caution as a second randomization of liposomal muramyl tripep- ES is responsive to alkylating agents such as ifosfamide and cyclo-
    tide (MTP-​PE) may have interfered with the ifosfamide question. phosphamide, to doxorubicin, and to other agents such as vincris-
    Furthermore, the ifosfamide arm contained no preoperative cis- tine, actinomycin D, and the topoisomerase II inhibitor etoposide.
    platin. Results of a French trial, SFOP-​91, which did not add, but Combinations of these agents with complementary mechanisms of
    rather substituted ifosfamide, suggest that favourable results may action have improved disease-​free survival rates. Intergroup Ewing
    be obtained by regimens that rely on ifosfamide rather than on Sarcoma (IESS) studies demonstrated the superiority of a four-​drug
    doxorubicin. regimen using vincristine (V), actinomycin D (A), cyclophospha-
    No other cytostatic agents have come close to replacing the four mide (C), and doxorubicin (D) over three-​drug VAC in terms of
    standard substances previously described. A combination of bleo- both disease-​free survival (60% versus 24%) and local control (96%
    mycin, cyclophosphamide, and actinomycin D (BCD) was used in versus 86%). Rosen, from the Memorial Sloan-​Kettering Cancer
    the early days of chemotherapy but was largely abandoned due to Center (MSKCC), reported an advantage of using these drugs in
    questionable efficacy. Carboplatin has some activity against osteo- combination rather than sequentially as early as 1978, but it is im-
    sarcoma, but less than cisplatin. Etoposide is almost inactive when portant to realize that a considerable number of patients in earlier
    given as a single agent, but may enhance the effect of carboplatin or series received cumulative doxorubicin doses of over 700 mg/​m2 be-
    ifosfamide. Amongst others, negative phase II studies have been re- fore its use was restricted to a maximum of 400–​500 mg/​m2 because
    ported for paclitaxel, docetaxel, and topotecan. Gemcitabine seems of the risk of doxorubicin-​related cardiomyopathy.
    to be marginally active, with stable disease achieved in some patients. The IESS and MSKCC experiences led to the widespread use
    Overall, progress in the treatment of osteosarcoma has been slow of similar four-​drug regimens, and treatment was later extended
    and survival rates have improved little over more than two decades. to incorporate ifosfamide (I) and etoposide (E). In the European
    Attempts to improve prognosis through treatment intensification by EICESS-​92 study, patients with localized high-​risk disease (tumour
    interval compression achieved with granulocyte colony-​stimulating volume more than 200 ml) receiving VIDA benefitted from the add-
    growth factor (G-​CSF) support or by combining maximally toler- ition of etoposide, and in the randomized POG-​CCG Ewing trial
    ated doses of high-​dose ifosfamide, high-​dose methotrexate, cis- (VACD versus EVADI), patients treated with EVADI appeared to
    platin, and doxorubicin have failed to show benefit. High-​dose have a more favourable outcome.
    chemotherapy with autologous peripheral blood stem-​cell trans- The incorporation of G-​ CSF into treatment regimens allows
    plantation was also unsuccessful in the reported series. dose intensification by increasing the dose per cycle or by short-
    Improving the prognosis for patients with poor response to in- ening the interval between treatments. The IESS-​II study compared
    duction chemotherapy remains a particular challenge. The pro- high-​dose intermittent chemotherapy with moderate-​ dose con-
    spective, randomized European and American Osteosarcoma tinuous chemotherapy, resulting in a significant benefit from the
    Study EURAMOS-​1 investigated whether the postoperative add- more intensive regimen (68% versus 48% disease-​free survival at
    ition of high-​dose ifosfamide and etoposide to a therapy backbone five years). Another POG-​CCG randomized study explored dose
    of methotrexate, doxorubicin, and cisplatin (MAP) was superior to intensification (maintaining a dosing interval of 21 days) by con-
    MAP alone in patients whose tumours had shown a poor response densing treatment duration to 30 weeks by increasing the individual
    to preoperative MAP. Unfortunately, this was not the case: the doses of chemotherapy in comparison to the delivery of the same
    five-​drug regimen was associated with additional morbidity and a cumulative doses over 48 weeks: so far, results between the standard
    higher risk to develop secondary leukaemias, but did not decrease and the dose-​intensified arms do not differ. More recently, the COG
    the recurrence risk, and thus its use is no longer recommended. The AEWS0031 trial has explored dose intensification with the use of
    EURAMOS-​1 trial also investigated whether outcomes for good G-​CSF by decreasing the intervals between cycles (interval com-
    responders to induction MAP chemotherapy might be improved pression) from three to two weeks while maintaining the same dose
    by adding maintenance therapy with pegylated interferonα after per cycle. Published results confirm that relapse rates are lower and
    completing postoperative MAP. Again, no significant benefit was toxicity manageable. The role of high-​dose therapy with autologous
    observed. haematopoietic stem-​cell rescue in certain risk groups of patients
    Whether the addition of the immunomodulator L-​ MTP-​ PE was investigated in the EURO-​EWING trial. The study’s results
    (liposomal muramyl tripeptide phenol ethanolamine) to chemo- suggest that a subgroup of patients with high-​risk localized disease
    therapy improves prognosis, as evaluated in the randomized North would benefit from such a procedure.
    American POG-​CCG INT0133 trial, is still a matter of debate. Treatment of metastatic and recurrent disease
    A first analysis of INT0133 concluded that there was an interaction
    between L-​MTP-​PE and ifosfamide, which, within a rather complex Primary metastases
    factorial trial design, was evaluated in the same study. A later publi- In osteosarcoma, all detectable metastases must be removed by
    cation reported a survival advantage for L-​MTP-​PE, but could still surgery if therapy is to be curative. As most metastases develop in
    not prove that adding L-​MTP-​PE improved event-​free survival. It the lung, this usually implies thoracotomy. There is evidence that
    Follow-up 227

    some patients with apparently unilateral lung metastases may in- cyclophosphamide, irinotecan and temozolomide, or gemcitabine
    deed have bilateral disease. Complete resection of osteosarcoma and docetaxel. High-​dose chemotherapy with autologous haemato-
    pulmonary metastases requires palpation of the lung. Acceptable poietic stem-​cell rescue can be considered in patients who achieved
    surgical approaches include thoracotomy or median sternotomy. a complete remission with conventional multimodal treatment.
    Approximately one quarter of all osteosarcoma patients with proven Unfortunately, outcomes after bone sarcoma recurrence remain
    metastatic disease at diagnosis and 40% of those who achieve a com- poor and have not improved much over the years. Participation in
    plete surgical remission of both the primary and all metastases in studies investigating tumour biology and innovative treatment ap-
    the context of an intensive polychemotherapy regimen will go on to proaches should be encouraged in order to help break this stalemate.
    become long-​term survivors. Solitary primary metastases may have
    a prognosis similar to that of localized disease.
    Metastatic disease at diagnosis is the most prominent adverse Follow-​up
    prognostic factor in ES. A recently published analysis of the EURO-​
    EWING consortium described survival rates of 10–​40% at three Suggestions for follow-​up are given in Table 26.3. Intervals between
    years, with patient age, tumour size and site, and number of metas- visits mirror the declining risk of relapse with time. As there is little
    tases defining distinct prognostic groups. Patients with isolated lung prospectively gathered evidence, suggestions must necessarily re-
    metastases have a better prognosis than those with extrapulmonary main somewhat arbitrary. Follow-​ up should, however, include
    metastases (Figure 26.4). If pulmonary disease is present, patients regular assessments of remission as well as tests for possible late ef-
    also receive bilateral pulmonary irradiation at doses of 14–​18 Gy. fects of treatment.
    Patients with solitary or circumscribed bony lesions can receive ir-
    radiation to those sites at doses of 40–​50 Gy. However, survival rates Remission status
    for patients with multiple bony and/​or bone-​marrow metastases are Tumour-​directed follow-​up focuses on the few organ systems where
    below 10% at two years. These discouraging results have led to more relapses are likely to occur. Pulmonary metastases contribute to over
    aggressive approaches utilizing high-​dose chemotherapy with au- 80% of recurrences of osteosarcoma and approximately half of all
    tologous stem-​cell reinfusion. A variety of agents such as busulfan, recurrences of ES. They will not usually cause symptoms until they
    treosulfan, melphalan, cyclophosphamide, thiotepa, etoposide, and have reached a very large size or penetrated the pleura. In order to
    carboplatin have been employed but without definitive evidence of
    improved outcome. Total body irradiation does not seem to offer
    benefit, but significantly contributes to toxicity. Local therapy aimed Table 26.3 Suggestions for follow-​up after multimodal therapy
    at identified metastases may improve survival. for bone sarcoma

    Recurrent disease Time Tumour-​directed Late-​effects monitoring


    investigations
    Osteosarcoma recurrences usually involve the lung. Bone metas-
    Baseline Chest X-​ray & CT Echocardiogram,
    tases and local recurrences are much less common, and other sites audiogram*, liver & kidney
    are rarely affected. Unfortunately, long-​term survival rates remain function
    below 20%. A short latency period and the presence of more than one X-​ray & MRI of primary site
    or two metastases are associated with particularly poor outcomes.
    Years 1 & 2 Chest X-​ray q 6–​12 wk Echocardiogram q 1–​2
    Complete surgery of all sites of recurrence is the only therapy with yr, audiogram*, liver** &
    unequivocally proven impact on survival. Selected patients may be kidney** function
    cured after multiple recurrences, provided that surgery results in a X-​ray of primary site q 4 mo
    renewed complete remission. It may be prudent to irradiate suitable Years 3 & 4 Chest X-​ray q 2–​4 mo Echocardiogram q 1–​2 yr
    inoperable lesions in order to slow the progression of disease, but
    X-​ray of primary site q 4 mo
    this is unlikely to lead to cure. Patients with inoperable recurrent
    osteosarcoma who receive chemotherapy may survive longer than Years 5–​10 Chest X-​ray q 6 mo*** Echocardiogram q (1–​)2–​4
    yr****
    those who do not. Recently, sorafenib (with and without everolimus)
    Thereafter A few relapses reported as Echocardiogram q (1–​)2–​4
    has also shown some activity in recurrent osteosarcoma. The exact late as two decades after yr
    role of adjuvant chemotherapy in recurrent operable osteosarcoma treatment: discuss with patient
    remains to be defined. Success has been limited, and there is no uni- whether to continue chest X-​ray
    versally accepted standard regimen, although many would consider monitoring
    the use of a combination of ifosfamide and etoposide in patients who Every clinic visit should include detailed history and physical examination. Many
    had not previously received those drugs. institutions will add complete blood counts. Evaluate any site with unexplained pain or
    swelling. Chest CT scan is optional, but should always be performed if chest X-​ray shows
    The survival of ES patients who develop local or metastatic recur- metastasis or is inconclusive. Add consultation with orthopaedic surgery and physical
    rence is poor and second remissions are usually short-​lived. Patients therapy as indicated. Offer fertility testing for males. Additional investigations may be
    indicated.
    with a longer disease-​free interval represent the subset of patients * If treatment included cisplatin.
    most likely to survive. The role of local therapy is less clear than in ** Need not be repeated if normal at one year.
    osteosarcoma. Chemotherapy and radiotherapy are commonly ap- *** Some groups recommend annual radiographs of the primary site until year 10.
    plied in an attempt to prolong survival. Chemotherapy regimens **** Longer interval if normal function and post-​pubertal at diagnosis.
    currently used include combinations of previously used drugs, or of Adapted with permission from Ritter J, Bielack SS, ‘Osteosarcoma’. Annals of Oncology,
    Volume 21, Suppl. 7, pp. vii320–​vii325. Copyright © 2010, Oxford University Press.
    platinum compounds with ifosfamide and etoposide, topotecan and DOI: https://​doi.org/​10.1093/​annonc/​mdq276
    228 CHAPTER 26 Bone Tumours

    detect them at an earlier stage, they must be searched for by appro- has been approved for the treatment of neuroblastoma. Since GD2
    priate imaging, which includes serial X-​rays and/​or CT scans of the is expressed in osteosarcomas as well as neurblastomas, the COG is
    thorax. In ES, where bone metastases occur more frequently, tech- currently testing this antibody in recurrent osteosarcoma.
    netium bone scans may be included, with other imaging methods Given the recent better characterization of osteosarcoma as a p53-​
    like FDG (fluorodeoxyglucose)-​PET scans or whole-​body MRI used driven cancer with extensive chromosomal aberrations in which
    less commonly. Local recurrences are often first detected because of multiple oncogenic pathways drive chromosomal instability during
    symptoms, most noticeably pain. Most centres still include sequen- tumour evolution, clearly new approaches attempting to exploit
    tial imaging of the primary site into their follow-​up programme. these alterations as a potential vulnerability are needed.
    In patients with ES, the addition of multi-​agent chemotherapy to
    Late effects local control with radiation and/​or surgery has improved disease-​
    Fortunately, many former bone sarcoma patients can lead relatively free survival from 10% to 50–​70%. High-​throughput microarray-​
    normal and productive lives. Nevertheless, they are more likely to suffer based gene-​expression profiling studies comparing different types
    from chronic health conditions than survivors of most other paediatric of small, blue, round-​cell tumours have identified a unique gene-​
    cancers. Late complications may be caused by the tumour itself or by expression pattern in ES, and gene-​expression studies on large co-
    surgery, radiotherapy, or chemotherapy. Functional and cosmetic con- horts may lead to the discrimination of prognostically distinct
    sequences for the musculoskeletal system depend on the location and groups, allowing better treatment stratification.
    extent of the tumour as well as the type of local treatment employed. The combinations of topotecan and cyclophosphamide, and
    Amputations and rotation-​plasties carry the stigma of mutilation, irinotecan andtemozolomide, have shown activity in patients with
    but may lead to functional results that are similar to those after limb-​ recurrent ES. Topotecan and cyclophosphamide combined with
    salvage procedures. Revision surgery is more frequently needed after vincristine (VTC) has been added to the compressed vincristine,
    limb-​salvage procedures, where periprosthetic infection, loosening, doxorubicin, cyclophosphamide, ifosfamide, etoposide (VDC-​IE)
    fractures, and prosthetic wear can occur. The use of expandable pros- schedule to study the feasibility of interval compression with that
    theses in skeletally immature patients is predictably associated with additional combination. Once the tolerable interval has been estab-
    the need for multiple revision procedures. Radiotherapy may lead to lished, it is likely that a randomized study of compressed VDC-​IE
    growth reduction, functional limitations, and secondary cancers. (The compared with VDC-​IE-​VTC will follow.
    late consequences of chemotherapy are more completely described in There was early excitement surrounding the use of IGF1R anti-
    Chapter 14, ‘Late Effects of Therapy and Survivorship Issues’.) bodies in the treatment of Ewing sarcoma, based upon early reports
    of clinical activity. The largest study, evaluating Roche IGF1R anti-
    body R1507, assessed 115 patients with recurrent Ewing sarcoma
    Future and remaining challenges and demonstrated a response rate of only 10%, with a median dur-
    ation of response of 29 weeks. Although the single-​agent activity is
    Osteosarcoma remains a challenging disease. It is sensitive to a small low, the potential to improve survival in high-​risk Ewing sarcoma
    number of medications, all having significant short-​and long-​term patients with the addition of an IGF1R antibody to standard chemo-
    toxicities. Radiation is of limited value and even sensitive tumours therapy is currently being tested by the COG.
    are incompletely eliminated by chemotherapy, mandating sur- Another area of interest and potential vulnerability of Ewing
    gical resection and reconstruction, with their attendant morbidity. sarcomas is PARP inhibition. Several groups independently re-
    Patients with recurrent disease are difficult to cure, with a some- ported that Ewing sarcomas were highly sensitive to the inhibition
    what better outlook for those with resectable metastases. The role of this enzyme. However, the first clinical trial with the PARP in-
    of MTP-​PE (mifurmatide) therapy in osteosarcoma requires better hibitor olaparib showed zero out of twelve responses. Ongoing
    definition. studies are evaluating combinations of DNA-​damaging agents such
    New initiatives are required. A way to affect the osteoclastic ac- as temozolamide with PARP inhibitors.
    tivity important for the progression of bone lesions is to interfere with Finally, since the known oncogenic driver of Ewing sarcoma is the
    the receptor activator for nuclear factor kappa-​B (RANK) pathway. mutant transcription factor fusing EWS (or other members of the
    RANK, RANK-​Ligand, and osteoprotegrin maintain normal bone FET family) to ETS genes, a first-​in-​class inhibitor of ETS activity is
    metabolism. Osteosarcoma cells express RANK-​ Ligand. Those currently being evaluated in a clinical trial targeting relapsed Ewing
    resistant to chemotherapy show increased expression of genes in- sarcoma patients.
    volved in osteoclast activation. Denosumab, a humanized antibody
    that binds RANK-​Ligand, inhibiting osteoclasts and therefore bone
    resorption, has been successfully used in patients with giant-​cell tu- Conclusions
    mour of bone, and has decreased skeletal-​related events in patients
    with some adult cancers. A COG study is currently enrolling recur- The prognosis for bone-​sarcoma patients is determined primarily by
    rent osteosarcoma patients. local treatment options, tumour dissemination, and tumour burden,
    The antibody drug conjugate, glembatumumab vedoitin, as well as by the biology underlying the tumour and its response to
    recognizes the cell-​surface glycoprotein GNMB and is conjugated treatment. Tumour burden can only be limited by early diagnosis, but
    to the microtubule inhibitor MMAE. GNMB is highly expressed on response is subject to the impact of better therapies including the de-
    the surface of more than 90% of primary osteosarcoma primary tu- sign of more effective chemotherapy combinations, increased drug
    mours and is currently being tested in recurrent osteosarcoma by intensity with improved supportive care, and surgical removal of all
    the COG. Similarly, the chimeric anti-​GD2 antibody, dinutuximab, the tumour. In addition, more precise determination of prognostic
    Acknowledgements 229

    factors should result in opportunities to better stratify treatment in- Goorin AM, et al. (2003) Presurgical chemotherapy compared with
    tensity, offering the opportunity to limit the risk of late effects from immediate surgery and adjuvant chemotherapy for nonmetastatic
    treatment in patients with favourable disease. Because of the rarity osteosarcoma: Pediatric Oncology Group Study POG-​8651. J Clin
    of these tumours, treatment should always be performed in centres Oncol 21(8), 1574–​80.
    experienced in bone-​sarcoma treatment, and within the framework Gorlick R, et al. (2013) Children’s Oncology Group’s 2013 blueprint for
    research: bone tumors. Pediatr Blood Cancer 60, 1009–​15.
    of controlled clinical trials where available. The evaluation of experi-
    Grier HE, et al. (2003) Addition of ifosfamide and etoposide to standard
    mental approaches to treatment should be restricted to clinical trial
    chemotherapy for Ewing’s sarcoma and primitive neuroectodermal
    settings.
    tumor of bone. N Engl J Med 34, 694–​701.
    Isakoff MS, et al. (2015) Osteosarcoma: current treatment and a collab-
    orative pathway to success. J Clin Oncol 33, 3029–​35.
    Acknowledgements Jaffe N (1972) Recent advances in the chemotherapy of metastatic
    osteogenic sarcoma. Cancer 30, 1627–​31.
    The authors acknowledge the important contributions of Heinrich Jaffe N, et al. (eds) (2009) Pediatric and Adolescent Osteosarcoma (Cancer
    Kovar, Heribert Jürgens, and Mark Bernstein to the chapter ‘Bone Treatment and Research series, Vol. 152). New York: Springer.
    Tumours’ in the sixth edition of Cancer in Children, upon which this Jiang Y, et al. (2015) Targeted therapies for advanced Ewing sarcoma
    chapter is based. family of tumors. Cancer Treat Rev 41, 391–​400.
    Kager L, et al. (2016) The ENCCA-​WP7/​EuroSarc/​EEC/​PROVABES/​
    EURAMOS Third European Bone Sarcoma Networking Meeting/​
    FURTHER READING Joint Workshop of EU Bone Sarcoma Translational Research
    Abed R, Grimer R (2010) Surgical modalities in the treatment of bone Networks; Vienna, Austria, September 24–​25, 2015. Workshop re-
    sarcoma in children. Cancer Treat Rev 36, 342–​7. port. Clin Sarcoma Res 6, 3. doi: 10.1186/​s13569-​016-​0043-​5
    Bacci G, et al. (1990) Primary chemotherapy and delayed surgery Kansara M, et al. (2014) Translational biology of osteosarcoma. Nat
    (neoadjuvant chemotherapy) for osteosarcoma of the extremities. Rev Cancer 14, 722–​35.
    The Istituto Rizzoli Experience in 127 patients treated preopera- Kempf-​Bielack B, et al. (2005) Osteosarcoma relapse after combined mo-
    tively with intravenous methotrexate (high versus moderate doses) dality therapy: an analysis of unselected patients in the Cooperative
    and intraarterial cisplatin. Cancer 65, 2539–​3. Osteosarcoma Study Group (COSS). J Clin Oncol 23, 559–​68.
    Bielack S, et al. (2002) Prognostic factors in high-​grade osteosarcoma Khanna C, et al. (2014) Toward a drug development path that tar-
    of the extremities or trunk. An analysis of 1,702 patients treated on gets metastatic progression in osteosarcoma. Clin Cancer Res 20,
    neoadjuvant Cooperative Osteosarcoma Study Group protocols. J 4200–​9.
    Clin Oncol 20, 776–​90. Kovac M, et al. (2015) Exome sequencing of osteosarcoma reveals mu-
    Bielack S, et al. (2013) Controversies in childhood osteosarcoma. tation signatures reminiscent of BRCA deficiency. Nat Commun 6,
    Minerva Pediatr 65, 125–​48. 8940. doi: 10.1038/​ncomms9940
    Bielack SS, et al. (2015) Methotrexate, doxorubicin, and cisplatin Kovar H, et al. (2016) The second European interdisciplinary Ewing
    (MAP) plus maintenance pegylated interferon alfa-​2b versus MAP sarcoma research summit—​a joint effort to deconstructing the
    alone in patients with resectable high-​grade osteosarcoma and multiple layers of a complex disease. Oncotarget 7(8), 8613–​24.
    good histologic response to preoperative MAP: first results of the doi: 10.18632/​oncotarget.6937
    EURAMOS-​1 good response randomized controlled trial. J Clin Ladenstein R, et al. (2010) Primary disseminated multifocal Ewing sar-
    Oncol 33, 2279–​87. coma: results of the Euro-​EWING 99 trial. J Clin Oncol 28, 3284–​91.
    Bishop MW, et al. (2016) Future directions in the treatment of osteo- Le Deley MC, et al. (2007) SFOP OS94: a randomised trial comparing
    sarcoma. Curr Opin Pediatr 28, 26–​33. preoperative high-​dose methotrexate plus doxorubicin to high-​
    Brierley JD, et al. (eds.) (2016) TNM Classification of Malignant dose methotrexate plus etoposide and ifosfamide in osteosarcoma
    Tumours (8th edn). Chichester, West Sussex: Wiley-​Blackwell. patients. Eur J Cancer 43, 752–​61.
    Casali PG, et al. (2018) Bone sarcomas: ESMO-​PaedCan-​EURACAN Machado I, et al. (2016) Ewing sarcoma and the new emerging Ewing-​
    clinical practice guidelines for diagnosis, treatment and follow-​up. like sarcomas: CIC and BCOR-​rearranged-​sarcomas. A systematic
    Ann Oncol 29(Suppl_​4), iv79–​95. review. Histol Histopathol 31,1169–​81.
    Ciccarese F, et al. (2015) The many faces of pulmonary metastases of Marina NM, et al. (2016) Comparison of MAPIE versus MAP in patients
    osteosarcoma: retrospective study on 283 lesions submitted to sur- with a poor response to preoperative chemotherapy for newly diag-
    gery. Eur J Radiol 84, 2679–​85. nosed high-​grade osteosarcoma (EURAMOS-​1): an open-​label, inter-
    Chen X, et al. (2014) Recurrent somatic structural variations contribute national, randomised controlled trial. Lancet Oncol 17, 1396–​408.
    to tumorigenesis in pediatric osteosarcoma. Cell Rep 7,104–​12. Meyers PA, et al. (2008) Osteosarcoma: the addition of muramyl tri-
    DuBois SG, et al. (2010) Angiogenesis and vascular targeting in Ewing peptide to chemotherapy improves overall survival. A report from
    sarcoma: a review of preclinical and clinical data. Cancer 116, the Children’s Oncology Group. J Clin Oncol 26, 633–​8.
    749–​57. Mirabello L, et al. (2015) Germline TP53 variants and susceptibility
    Ferrari S, et al. (2003) Postrelapse survival in osteosarcoma of the ex- to osteosarcoma. J Natl Cancer Inst 107, pii: djv101. doi:10.1093/​
    tremities: prognostic factors for long-​term survival. J Clin Oncol 21, jnci/​djv101
    710–​15. Oeffinger KC, et al. (2006) Chronic health conditions in adult sur-
    Fletcher CDM, et al. (eds) (2013) WHO Classification of Tumours of vivors of childhood cancer. N Engl J Med 355, 1572–​82.
    Soft Tissue and Bone (4th edn), Vol. 5. Lyon: IARC. Omer N, et al. (2017) Phase-​II trials in osteosarcoma recurrences: a
    Gaspar N, et al. (2015) Ewing sarcoma: current management and fu- systematic review of past experience. Eur J Cancer 75, 98–​108.
    ture approaches through collaboration. J Clin Oncol 33, 3036–​46. doi: 10.1016/​j.ejca.2017.01.005
    230 CHAPTER 26 Bone Tumours

    Pappo AS, et al. (2014) A phase 2 trial of R1507, a monoclonal anti- Smith MA, et al. (1991) Influence of doxorubicin dose intensity on
    body to the insulin-​like growth factor-​1 receptor (IGF-​1R), in response and outcome for patients with osteogenic sarcoma and
    patients with recurrent or refractory rhabdomyosarcoma, osteo- Ewing’s sarcoma. J Natl Cancer Inst 83, 1460–​70.
    sarcoma, synovial sarcoma, and other soft tissue sarcomas: results Sobin LH, et al. (2009) UICC TNM Classification of Malignant Tumors
    of a Sarcoma Alliance for Research Through Collaboration study. (7th edn). New York: Wiley.
    Cancer 120(16), 2448–​56. doi: 10.1002/​cncr.28728 Subbiah V, et al. (2009) Ewing’s sarcoma: standard and experimental
    Paulussen M, et al. (2008) Results of the EICESS-​92 Study: two ran- treatment options. Cur Treat Options Oncol 10, 126–​40.
    domized trials of Ewing’s sarcoma treatment—​cyclophosphamide Whelan JS, et al. (2012) Survival from high-​grade localised extremity
    compared with ifosfamide in standard-​risk patients and assessment osteosarcoma: combined results and prognostic factors from three
    of benefit of etoposide added to standard treatment in high-​risk pa- European Osteosarcoma Intergroup randomised controlled trials.
    tients. J Clin Oncol 26, 4385–​93. Ann Oncol 23, 1607–​16.
    Petrilli AS, et al. (2006) Results of the Brazilian Osteosarcoma Whelan J, et al. (2018) High-​dose chemotherapy and blood autolo-
    Treatment Group Studies III and IV: prognostic factors and impact gous stem-​cell rescue compared with standard chemotherapy in
    on survival. J Clin Oncol 24, 1161–​8. localized high-​risk Ewing sarcoma: results of Euro-​E.W.I.N.G.99
    Picci P, et al. (1994) The relationship of chemotherapy-​induced ne- and Ewing-​2008. J Clin Oncol. JCO2018782516. doi: 10.1200/​
    crosis and surgical margins to local recurrence in osteosarcoma. J JCO.2018.78.2516 [Epub ahead of print]
    Clin Oncol 12, 2699–​705. Winkler K, et al. (1988) Neoadjuvant chemotherapy of osteosar-
    Piperno-​Neumann S, et al. (2016) Zoledronate in combination with coma: results of a randomized cooperative trial (COSS-​82) with
    chemotherapy and surgery to treat osteosarcoma (OS2006): a random- salvage chemotherapy based on histological tumor response. J Clin
    ised, multicentre, open-​label, phase 3 trial. Lancet Oncol 17, 1070–​80. Oncol 6, 329–​37.
    Roth M, et al. (2016) Targeting glycoprotein NMB with antibody-​drug Womer RB, et al.(2012) Randomized controlled trial of interval-​
    conjugate, glembatumumab vedotin, for the treatment of osteosar- compressed chemotherapy for the treatment of localized Ewing sar-
    coma. Pediatr Blood Cancer 63, 32–​8. coma: a report from the Children’s Oncology Group. J Clin Oncol
    Schuck A, et al. (2003) Local therapy in localized Ewing tumors: results 30, 4148–​54.
    of 1058 patients treated in the CESS 81, CESS 86, and EICESS 92
    trials. Int J Radiat Oncol Biol Phys 55, 168–​77.
    27
    Wilms and Other Renal Tumours
    Norbert Graf and Christophe Bergeron

    Wilms tumour (nephroblastoma) Pathogenesis


    Molecular biology and genetics
    Introduction Nephroblastoma is a genetically heterogeneous tumour. Besides
    In 1814, Rance classified Wilms Tumour (WT) or nephroblastoma gene mutations, further mechanisms of tumour development are
    for the first time as a renal neoplasia. In 1899, the surgeon Max found as loss of heterozygosity (LOH) and loss of imprinting (LOI).
    Wilms described this tumour entity in a 90-​page monograph Die Nephroblastomas mainly show an euploid set of chromosomes and
    Mischgeschwülste der Niere in much more detail. Today, this tumour the frequency of LOH is very low, in the range of less than 5%.
    is regarded as a prime example of a curable malignant disease. The Several genes are known which play a crucial role in the devel-
    improvements in the treatment of nephroblastoma are particularly opment of nephroblastoma. WT candidate genes are on chromo-
    based on the progress in surgery and radiotherapy, and in the devel- somes 11p13 (WT1), 11p15.5 (WT2), 16q (WT3), 17q12-​q21 (WT4;
    opment of effective chemotherapies. Interdisciplinary cooperation FWT1), 7p15-​p11.2 (WT5). Deletions of wt1 are found in 10–​30% of
    and prospective randomized trials have made considerable contri- WT. Numeric and structural changes concern particularly chromo-
    butions to this success. somes 1, 6, 7, 8, 11, 12, 16, 17, and 18. LOH is mainly found on 11p
    (40%) and, in a lower frequency, on 1p, 7p, and 16q. Epigenetic al-
    Epidemiology terations primarily affect 11p15.5 and 11p13. Further genes with
    The incidence of WT is seven per one million children less documented mutations include CTNNB1, WTX (AMER1), TP53,
    than 15 years of age. With 6–​7% of all childhood malignancies, FBXW7, MYCN, SIX1/​2, DICER1, DROSHA, and DGCR8. A gain
    nephroblastoma represents the most frequent malignant renal tu- of 1q is associated with poorer event-​free survival (EFS) in tumour
    mour. Worldwide, WT are more common in girls than boys. The age subsets and will be validated for better stratification of postoperative
    distribution shows a peak between the second and third year of life. treatment.
    Children with a bilateral tumour are younger. The genetics of the overgrowth syndromes are complex. As with
    BWS, they arise from abnormalities of imprinting in 11p15.5 with
    Aetiology and risk factors differences between BWS and hemi-​hypertrophy. WAGR syndrome
    In 1964, Miller and colleagues reported, for the first time, an asso- is caused by a complete deletion of one copy of the WTgene, WT1,
    ciation between WT and aniridia. Today, different malformations and the adjacent aniridia gene, PAX6 on chromosome 11p13. In pa-
    linked to WT are known and include, particularly, hemihypertrophy tients with aniridia, this information helps to identify those at risk
    and urogenital malformations (cryptorchidism, hypospadias, for developing WT, by screening them for the combined deletion of
    pseudohermaphroditism, and gonadal dysgenesis). Aniridia and WT1 and PAX6. WT1 is also involved in DDS, caused by a germline
    hemihypertrophy are very rare syndromes justifying regular check-​ point mutation. In familial WT, accounting for approximately 1%
    ups in children with such anomalies, for an early diagnosis. A high of patients, there is usually no associated congenital abnormality
    incidence of WT is seen in children with hemihypertrophy, WAGR or predisposition to other tumour types. They show an autosomal
    syndrome (WT, aniridia, genital deformity, retardation), Beckwith–​ dominant inheritance with a variable penetrance. Genetic linkage
    Wiedemann syndrome (BWS; hemihypertrophy, omphalocele, studies in different families have shown no association to 11p13 or
    macrosomy, macroglossia, hypoglycaemia, and deep-​ rooted 11p15.5. One gene for familial WT, FWT1, could be localized on
    ears), Denys–​ Drash syndrome (DDS; pseudohermaphroditism, chromosome 17q and another, FWT2, on 19q.
    glomerulopathy, and WT), Perlman syndrome (macrocephaly, Nephrogenic rests and nephroblastomatosis
    deep-​rooted eyes and ears, macrosomia, and organomegalia), and
    von Recklinghausen’s disease or neurofibromatosis type 1 (Table Nephrogenic rests can appear as preliminary stages of a
    27.1). Environmental factors do not play a major role in the aeti- nephroblastoma. Nephroblastomatosis is defined as the diffuse
    ology of nephroblastoma. or multifocal occurrence of nephrogenic rests. Different forms of
    232 CHAPTER 27 Wilms and Other Renal Tumours

    Table 27.1 Syndromes and congenital anomalies with a predisposition for nephroblastoma,
    and its frequency

    SIOP [%]‌ NWTS [%]‌


    Syndromes/​anomalies with increased growth
    Beckwith–​Wiedemann syndrome (BWS) Not known ?Not known
    • Risk for tumours 5–​10%
    • Loss of imprinting of LIT1 1–​5%
    • Loss of imprinting of H19 35–​45%
    • Loss of imprinting of LIT1 and H19 (UPD 11 p15) 25–​30%
    • Normal methylation sample 10–​15%
    Isolated hemihypertrophy 3.13 2.47
    Perlman syndrome <1.00 –​
    Sotos syndrome <1.00 –​
    Simpson–​Golabi–​Behmel syndrome <1.00 –​
    Syndromes/​anomalies without disturbances of growth
    WAGR syndrome <1.00 –​
    Eye deformities 2.17 1.21
    Isolated aniridia –​ 0.84
    Denys–​Drash syndrome (DDS) <1.00 –​
    Urogenital deformities 4.41 4.61
    CNS deformities 1.66 0.47
    Bloom syndrome <1.00 –​
    Trisomy 18 <1.00 –​
    M. Recklinghausen <1.00 –​
    Familial Wilms tumour 1.00 –​

    nephrogenic rests are distinguished according to their localization Pathology


    in the renal lobe: Histopathology
    • perilobar nephroblastomatosis The diagnosis of nephroblastoma is still based on the histo-
    • intralobar nephroblastomatosis pathological classification. In accordance with Wilms’ concept,
    • mixed perilobar/​intralobar nephroblastomatosis nephroblastoma is a tumour of mesodermal origin that develops in
    • panlobar nephroblastomatosis the embryonic kidney. Classic nephroblastomas (mixed type) are
    renal tumours with a blastemic, epithelial (tubules), and mesen-
    Despite the histological similarity to WT, nephroblastomatosis be-
    chymal component (stroma). The distribution of these three com-
    haves like residua of primitive embryonic tissue and does not seem
    ponents can vary, explaining epithelial, stromal, and blastemal types
    to have any invasive or metastatic behaviour. The malignant poten-
    of nephroblastoma. Multi-​centric growth of the tumour in one or
    tial of nephrogenic rests is justified in their high mitotic activity and
    both kidneys is possible. Approximately 5% are bilateral.
    the narrow association with WT. They occur in approximately 40%
    The histopathological classification of childhood renal tumours
    of all children with WT. A transformation of nephroblastomatosis to
    is dependent on the initial treatment. Patients receiving preopera-
    a true nephroblastoma is possible.
    tive chemotherapy are classified according to the International
    The histogenetic origin of other renal tumours of childhood is
    Society of Paediatric Oncology (SIOP) Stockholm Working
    unclear. Clear-​cell sarcoma (CCSK) possibly represents the ma-
    Classification of 2002, and patients undergoing primary sur-
    lignant variant of mesoblastic nephroma. Malignant rhabdoid
    gery are classified according to the classification of the Children’s
    tumour of the kidney (RTK) is not exclusively located in the
    Oncology Group (COG, formerly the National Wilms Tumour
    kidney but in many other sites including the central nervous
    Study Group). According to the Stockholm classification, three
    system (CNS) (AT/​RT: atypical teratoid rhabdoid tumour) and
    risk groups can be distinguished (low, intermediate, and high),
    soft tissue (MRT: malignant rhabdoid tumour). In paediatric
    whereas in the COG classification, only ‘favourable’ and ‘un-
    renal-​cell carcinoma (RCC), the characterization of the trans-
    favourable’ risk groups are defined. The tumours with a low and
    location RCC is most important. Molecular biology studies
    intermediate risk in the SIOP classification correspond largely to
    have shown that the cellular subtype of congenital mesoblastic
    tumours of the group with a favourable histology according to
    nephroma is identical with the infantile fibrosarcoma of soft
    COG. The revised SIOP classification of childhood renal tumours
    tissue, both sharing histopathological features and the t(12;15)
    is given in Box 27.1.
    ETV6-​NTRK3 gene fusion.
    Wilms tumour (nephroblastoma) 233

    Box 27.1 The International Society of Paediatric Oncology metastasis, survive in more than 90% of cases. Patients with a tu-
    (SIOP) classification of childhood renal tumours for patients mour with diffuse anaplasia have a considerably poorer outcome. In
    with a preoperative chemotherapy stage IV, the prognosis of patients depends on treatment response
    and histology. Patients with a complete remission after preoperative
    A. For pre-​treated cases chemotherapy and tumour nephrectomy have an overall survival
    I. Low-​risk tumours (OS) of 80% if the histology is not high risk. The prognosis of CCSK
    Mesoblastic nephroma*
    Cystic, partially differentiated nephroblastoma
    Completely necrotic nephroblastoma
    II. Intermediate-​risk tumours
    Nephroblastoma—​epithelial type
    Nephroblastoma—​stromal type Box 27.2 International Society of Paediatric Oncology (SIOP)
    Nephroblastoma—​mixed type staging criteria for renal tumours of childhood
    Nephroblastoma—​regressive type
    Nephroblastoma—​focal anaplasia Stage I
    III. High-​risk tumours a. The tumour limited to kidney or surrounded with fibrous
    Nephroblastoma—​blastemal type pseudocapsule if outside of the normal contours of the kidney, the
    Nephroblastoma—​diffuse anaplasia renal capsule, or pseudocapsule may be infiltrated with the tumour
    Clear-​cell sarcoma of the kidney* but it does not reach the outer surface, and is completely resected
    Rhabdoid tumour of the kidney* (resection margins ‘clear’)
    B. For primary nephrectomy cases b. The tumour may be protruding (‘bulging’) into the pelvic system and
    ‘dipping’ into the ureter (but it is not infiltrating their walls)
    I. Low-​risk tumours
    Mesoblastic nephroma* c. The vessels of the renal sinus are not involved
    Cystic, partially differentiated nephroblastoma d. Intrarenal vessel involvement may be present
    II. Intermediate-​risk tumours Fine-​
    needle aspiration or percutaneous core-​ needle biopsy (‘tru-​
    Non-​anaplastic nephroblastoma and its variants cut’) does not upstage the tumour. The presence of necrotic tumour or
    Nephroblastoma—​focal anaplasia chemotherapy-​induced changes in the renal sinus/​hilus fat and/​or out-
    III. High-​risk tumours side of the kidney should not be regarded as a reason for upstaging a
    Nephroblastoma—​diffuse anaplasia tumour.
    Clear-​cell sarcoma of the kidney* Stage II
    Rhabdoid tumour of the kidney* a. The tumour extends beyond kidney or penetrates through the renal
    * Non-​Wilms tumour capsule and/​or fibrous pseudocapsule into perirenal fat but is com-
    Reproduced with permission from Vujanic GM, et al. ‘Revised International Society pletely resected (resection margins ‘clear’)
    of Paediatric Oncology (SIOP) Working Classification of Renal Tumors of Childhood’.
    Med Pediatr Oncol, Volume 38, Issue 2, pp.79–​82. Copyright © 2002 Wiley–​Liss Inc. b. The tumour infiltrates the renal sinus and/​or invades blood and
    DOI: https://​doi.org/​10.1002/​mpo.1276 lymphatic vessels outside the renal parenchyma but is completely
    resected
    c. The tumour infiltrates adjacent organs or vena cava but is com-
    pletely resected
    Prognostic factors
    Stage III
    As a result of preoperative chemotherapy, the distribution of the dif-
    a. Incomplete excision of the tumour which extends beyond resection
    ferent histological subtypes changes. In particular, the percentage of margins (gross or microscopical tumour remains postoperatively)
    the blastemal type goes down from 40% to 10%. As the remaining b. Any abdominal lymph nodes are involved
    blastema after preoperative chemotherapy is chemoresistant, these c. Tumour rupture before or intra-​operatively (irrespective of other
    tumours are treated as high risk in SIOP protocols. After preoperative criteria for staging)
    chemotherapy, some tumours become completely necrotic or regres- d. The tumour has penetrated through the peritoneal surface
    sive in the form of necrosis, fibrosis, and a collection of xanthomatous e. Tumour implants are found on the peritoneal surface
    cells. The detection of anaplasia is of prognostic importance. f. The tumour thrombi present at resection margins of vessels or ur-
    eter, transsected or removed piecemeal by surgeon
    The extension of the tumour (staging) must be described exactly
    g. The tumour has been surgically biopsied (wedge biopsy) prior to
    and is essential for correct treatment stratification after surgery. preoperative chemotherapy or surgery
    Growth of the tumour beyond the tumour capsule represents the
    The presence of necrotic tumour or chemotherapy-​induced changes
    decisive feature between stage I and II. Tumour in vessels corres- in a lymph node or at the resection margins should be regarded as
    ponds to stage II. Every incomplete operative removal, every tumour stage III.
    rupture or open tumour biopsy and lymph-​node infiltration should
    Stage IV
    be assigned to stage III. In stage IV, distant metastases are diagnosed.
    Haematogenous metastases (lung, liver, bone, brain, etc.) or lymph-​
    Bilateral tumours indicate stage V (Box 27.2). node metastases outside the abdomino-​pelvic region.
    In the European approach, the prognosis of nephroblastoma pri-
    marily depends on tumour stage, histological type, and response to Stage V
    Bilateral renal tumours at diagnosis. Each side should be substaged ac-
    preoperative chemotherapy. In addition, some molecular markers
    cording to the above criteria.
    correlate with a poor prognosis (gain of 1q; LOH 1p, 11q, 16q, 22q; Reproduced with permission from Vujanic GM, et al. ‘Revised International Society
    and p53 mutations). of Paediatric Oncology (SIOP) Working Classification of Renal Tumors of Childhood’.
    Med Pediatr Oncol, Volume 38, Issue 2, pp. 79–​82. Copyright © 2002 Wiley–​Liss Inc.
    Of all children with nephroblastoma, 90% can be cured today. DOI: https://​doi.org/​10.1002/​mpo.1276
    Patients with a low-​and intermediate-​ risk tumour, without
    234 CHAPTER 27 Wilms and Other Renal Tumours

    has improved considerably, with OS rates of 80%, whereas the prog- (a)
    nosis of patients with an RTK remains dismally below 20%.

    Prevention and early diagnosis


    Most WT appear sporadically. A general screening is not possible
    due to the lack of tumour markers. On the other hand, in some chil-
    dren having syndromes associated with nephroblastoma, the indi-
    vidual risk to develop a nephroblastoma can be defined by molecular
    genetic examinations. Ultrasound scans are recommended at three-​
    month intervals in such children. In Germany, 10% of children with
    nephroblastoma are diagnosed in the context of preventive medical
    check-​ups by paediatricians.

    Clinical symptoms
    The typical first sign of a WT is an asymptomatic palpable or visible
    abdominal mass. Pain or haematuria are rare. In addition, WT is
    known to present with hypertension, that is caused by an increase in (b)
    renin activity and is sometimes difficult to control, or coagulopathy,
    caused by acquired von Willebrand syndrome. In most cases, the
    acquired coagulation disorder and hypertension will resolve after
    tumour nephrectomy. Intratumoural haemorrhage may occur, re-
    sulting in rapid abdominal enlargement that requires emergency
    surgery. Prenatal diagnoses by ultrasound are described (Table 27.2).

    Diagnostic procedures
    Besides the clinical examination, the diagnosis of nephroblastoma
    is primarily based on imaging studies (abdominal ultrasound, com-
    puted tomography, and/​or magnetic resonance imaging) (Figure
    27.1). As patients in the SIOP trials and studies are treated without
    histological diagnosis, all imaging studies need to be of high quality
    to reduce the risk of a false positive diagnosis and not indicated
    chemotherapy. By using abdominal ultrasound, this risk was as-
    sessed below 2% for SIOP 6 and SIOP 9. The most important criteria Figure 27.1 Imaging study (CT scan) of unilateral nephroblastoma: (a)
    with contrast enhancement; (b) without contrast enhancement.
    allowing a safe diagnosis are listed in Box 27.3.

    Differential diagnoses ganglioneuroma, cystic nephroma, hamartoma, renal cysts, hema-


    The following diseases need to be ruled out: neuroblastoma (but toma, renal abscess, xanthogranulomatous pyelonephritis,
    urine catecholamines are negative in nephroblastoma), lymphoma angiomyolipoma, and adenoma.
    of the kidney, RCC, nephroblastomatosis, RTK, teratoma, The diagnosis of nephroblastomatosis can only be proven by hist-
    ology. It is of utmost importance that the pathologist is aware of the
    patient’s history and of the imaging studies.
    Table 27.2 First symptoms at diagnosis in children with a
    nephroblastoma Treatment
    Symptom Frequency [%]‌ The treatment of this tumour should always be carried out in pro-
    Asymptomatic tumour swelling 61.6 spective trials in which surgery, chemotherapy, and radiotherapy
    Haematuria 15.1 are combined according to the individual risk to a patient. Today,
    two major study groups are treating these patients. The fundamental
    Preventive medical check-​up 9.2
    difference between the two groups is in their initial treatment ap-
    Constipation 4.3 proach: primary surgery (COG: North America) or preoperative
    Weight loss 3.8 chemotherapy (SIOP: European countries). Arguments for pre-
    Urinary tract infection 3.2 operative chemotherapy are the reduction of the tumour volume
    Diarrhoea 3.2 (downstaging) with the possibility to measure in vivo response. The
    Diagnosis at trauma 2.7
    number of intraoperative tumour ruptures is decreased and the
    number of patients with a postoperative local stage I is raised (Table
    Nausea, vomiting, pain, hernia, pleural effusion, high Rare
    blood pressure 27.3 and Figure 27.2). Moreover, the chemotherapy response gives
    some additional prognostic factors. For more than 40 years both
    Adapted with permisson from Graf N, Gutjahr P. ‘Wilms-​Tumoren (Nephroblastome).’
    In: P. Gutjahr (Ed.) Krebs bei Kindern und Jugendlichen, 5. Überarbeitete und erweiterte
    study groups have gained a lot of success in treating these patients.
    Auflage. Copyright © 2004 Deutscher Ärzteverlag.
    Wilms tumour (nephroblastoma) 235

    Box 27.3 Criteria for a safe diagnosis of nephroblastoma using restricted to patients with higher risk of local relapse and, in some
    imaging studies cases, for the treatment of pulmonary metastases or after incomplete
    resection of liver metastases. Radiotherapy can also provide excel-
    • The intrarenal tumour is solid and inhomogeneous with partly lent palliation for symptomatic brain or bone metastases.
    cystic areas
    Radiotherapy is given in SIOP and COG studies in local stage III
    • The tumour has a sharp border
    • The surrounding tissue is replaced rather than infiltrated
    in non-​high-​risk or favourable histology, and in stage II in high-​risk
    • The destruction, inclination, and replacement of calices and the renal or unfavourable histology. The total dose is 14.4 Gy in intermediate-​
    pelvis is characteristic risk and 25.2 Gy in high-​risk histology in the SIOP studies. For
    • The inhomogeneity of the tumour increases after contrast macroscopic residual disease, a local boost of 10–​15 Gy is given. In
    enhancement COG, the irradiation dose is 10.2 Gy. Whole abdominal irradiation
    • Intratumoural bleedings are relatively frequent (27%) (15–​19.5 Gy) is needed in cases of diffuse peritoneal seeding due to
    • Calcifications in the tumour are rare (8%) tumour rupture. The dose to the liver and, particularly, to the un-
    affected kidney should be reduced to 12 Gy. Irradiation to the lungs
    can be avoided if complete remission of the metastases is achieved
    The most important results from their trials form the basis of today’s after chemotherapy. Radiotherapy treatment planning should
    treatment and are listed in Box 27.4. homogeneously include the whole width of the vertebral column,
    particularly at doses of more than 15 Gy, to avoid disturbances of
    Surgery growth and subsequent development of scoliosis. The dose to the
    Surgery for a WT is almost always an elective intervention and rarely genital organs should be below 2 Gy on at least one ovary and under
    due to emergency situations after traumatic or spontaneous tumour 1 Gy for both testicles, if possible.
    rupture. Even if the tumour nephrectomy seems technically simple,
    Chemotherapy
    the most experienced team of surgeons and anaesthesiologists should
    carry out the intervention. The tumour has to be removed radically, International Society of Paediatric Oncology studies
    and tumour spread has always to be documented. Simultaneously, Preoperative chemotherapy is given to all patients aged over six
    an adequate lymph-​node dissection should be carried out. It has to months and under 18 years at diagnosis of a renal tumour by
    be taken into account that the quality of the imaging studies has im- imaging studies. Vincristine and actinomycin D are given for
    proved fundamentally and can rule out bilateral disease with great four weeks in localized stages. Anthracyclines are added in case
    certainty. If the contralateral kidney looks normal on good-​quality of metastatic disease and treatment is prolonged to six weeks.
    imaging studies, exposure of this normal kidney is not needed during Patients with bilateral disease are treated individually with vincris-
    surgery. If there are doubts about the diagnosis of a nephroblastoma tine and actinomycin as long as there is tumour regression, but
    (aspect and age) before preoperative treatment, a tru-​cut biopsy of for no longer than 12 weeks. Babies less than six months of age
    the tumour, with a posterior approach only, may be considered. On and teenagers over 18 years are primarily operated in the context
    the other hand, primary complete tumour removal should always be of SIOP studies, as other renal tumours (CMN, RCC) are more
    preferred in doubt of malignancy. Renal-​sparing surgery in unilat- common in these patients.
    eral tumours should only be done in special cases and by experienced Postoperative treatment is always carried out on the basis of the
    surgeons. Surgery for thromboses of the vena cava often requires the local stage and the histological type. The following guidelines apply
    use of the heart–​lung machine. Prognosis of these patients remains to the individual sub-​types:
    good after adequate surgery. Experienced surgeons should remove
    primary liver metastases after preoperative chemotherapy. • Low-​risk tumours: in stage I, no further treatment is given. In
    Congenital mesoblastic nephromas (CMN) show an unusual stage II and III, vincristine and actinomycin-​D (AV) are given for
    growth pattern, with fine finger-​shaped extensions into the adjacent 26 weeks.
    kidney fat capsule. Because of these extensions, the fat capsule needs • Intermediate-​risk tumours: relapse-​free survival (RFS) of pa-
    to be removed with wide excision margins. tients with stage 1 is 90% with only four weeks of postoperative
    treatment. In stage II and III, patients receive only two drugs (AV)
    Radiotherapy for 26 weeks. Doxorubicin can be avoided in these patients.
    Nephroblastoma is a radiosensitive tumour. With the development • High-​risk tumours: with RFS below 50%, treatment is intensi-
    of effective chemotherapeutic combinations, its place has been fied with carboplatin and etoposide. In the high-​risk scheme, both

    Table 27.3 Number of tumour ruptures at primary surgery and after preoperative treatment in SIOP
    (International Society of Paediatric Oncology) 1 and SIOP 2

    SIOP 1 Rupture (%) SIOP 2 Rupture (%)


    Preoperative irradiation 4 p = 0.001 Preoperative chemotherapy 5 p = 0.0025
    Primary surgery 32 Primary surgery including small tumours 20

    Reproduced with permission from Graf N, et al. ‘The Role of Preoperative Chemotherapy in the Management of Wilms Tumor—​The SIOP Studies’.
    Urologic Clinics of North America, Volume 27, Issue 3, pp. 443–​454. Copyright © 2000 W.B. Saunders Company. Published by Elsevier Inc. All rights
    reserved. DOI: https://​doi.org/​10.1016/​S0094-​0143(05)70092-​6.
    236 CHAPTER 27 Wilms and Other Renal Tumours

    stage I stage II NO stage II NO & III % tumour ruptures

    Primary surgery, siop 1,2 [n = 110]

    Preoperative irradiation, siop 1, [n = 73]

    Preoperative irradiation & amd, siop 2, 5, [n = 162]

    Preoperative chemotherapy siop 5, 6, 9, [n = 1077]

    0 20 40 60 80 100
    %

    Figure 27.2 Postoperative stage distribution and rupture rate in different SIOP (International Society of Paediatric Oncology) studies.

    drugs are alternately given with a combination of cyclophospha- radiation treatment has to be very restrictive. During the course of
    mide and doxorubicin. Patients with stage II and III need local the treatment, regular abdominal ultrasounds are needed, at short
    irradiation. intervals, to recognize the development of a possible anaplastic WT
    • Stage IV disease: postoperative treatment depends on the re- as early as possible (see Figure 27.3).
    sponse to preoperative chemotherapy. In case of complete re-
    mission (CR) after preoperative treatment (good responder),
    postoperative treatment is given according to local stage and hist-
    ology as in non-​metastasized tumours, but at least according to Box 27.4 Major findings of NWTSG (COG), SIOP, and UK trials
    stage II. Patients with non-​CR do need a more intensified treat-
    ment (VP16, cyclophosphamide, doxorubicine, carboplatin). • Treatment needs to be stratified according to stage and histology
    Radiotherapy to the lungs is necessary in patients with non-​CR or • The combination of vincristine and actinomycin D is more effective
    than the use of a single drug (with the exception of single VCR in
    with high-​risk tumours. stage I after primary surgery)
    • In stage I, no irradiation is needed
    Children’s Oncology Group studies • In stage II and intermediate risk or favourable histology, no irradiation
    is needed
    Preoperative treatment is not given. Postoperative treatment is al- • With the addition of doxorubicin in stage III, the dose of radiation
    ways carried out on the basis of age, local stage, histological type, can be reduced
    LOH 1p and 16q, and response to chemotherapy in case of stage • Cyclophosphamide does not increase prognosis in stage IV in case of
    IV. The most important drugs in the treatment are vincristine, anaplasia
    actinomycin-​D, and doxorubicine. The NWTS 4 study evaluated • Anthracyclines are decisive for the prognosis of clear-​cell sarcoma
    an early intensification of the treatment. It turned out that this so-​ • A single bolus of actinomycin-​D is as effective as a split dose
    called ‘pulse intensive treatment’ is equal to the standard treatment • The duration of treatment is eight weeks in stage I and up to six
    months in stage II and III in intermediate risk or favourable histology
    and more cost-​effective by minimizing hospital visits. The duration
    • A tumour volume of more than 500 ml after preoperative chemo-
    of the treatment is shortened to six months at the longest. In the therapy is of prognostic significance
    ongoing COG trial, LOH of 1p and 16q are used for stratification. • The number of tumour ruptures decreases as the number of stage
    Patients with LOH 1p and LOH 16q receive a more intensified treat- I increases significantly after preoperative treatment
    ment to improve their survival rates. There is a very low-​risk arm • Preoperative chemotherapy with vincristine and actinomycin D is as
    within COG, where children below two years of age with a unilat- effective as preoperative irradiation with 20 Gy
    eral nephroblastoma, favourable histology, stage I, and a tumour • After four and eight weeks of preoperative chemotherapy with vin-
    weight of less than 550 g receive surgery only, without adjuvant cristine and actinomycin D, the same stage distribution is achieved
    chemotherapy. • Patients with pulmonary metastases and complete remission after
    preoperative chemotherapy do not need lung irradiation
    Treatment of nephroblastomatosis • Some very low-​risk patients can be cured by surgery alone
    • Gain 1q, LOH of 1 p and 16q are of prognostic significance
    The treatment of nephroblastomatosis is still controversial. Initial • Remaining blastema after preoperative chemotherapy is a bad prog-
    nephrectomy should be avoided. Treatment is based on WT treat- nostic feature
    ment. In cases of a simultaneous WT, treatment is given according NWTSG: National Wilms Tumour Study Group; COG: Children’s Oncology Group;
    to the histological results and the stage of the tumour. In cases of SIOP: International Society of Paediatric Oncology; VCR: vincristine
    diffuse perilobar nephroblastomatosis (without WT), prolonged Adapted with permission from Graf N, Gutjahr P. ‘Wilms-​Tumoren (Nephroblastome).’
    In: P. Gutjahr (Ed.) Krebs bei Kindern und Jugendlichen, 5. Überarbeitete und erweiterte
    chemotherapy with vincristine and actinomycin-​ D is recom- Auflage. Copyright © 2004 Deutscher Ärzteverlag.
    mended. Impressive regressions are described. The indication for
    Clear-cell sarcoma of the kidney 237

    (a) (b) prognostic factors of the primary tumour. Children having only re-
    ceived actinomycin and vincristine during their first treatment have
    a high chance of survival with conventional therapy. The chance of
    survival is less than 40% in all other cases. For these patients, sev-
    eral highly effective chemotherapy regimens, including ifosfamide,
    carboplatin, and etoposide (ICE), are considered as a first treatment
    choice. To what extent the prognosis can be improved in such pa-
    tients by giving high-​dose chemotherapy followed by stem-​cell
    rescue is still unclear. International cooperation is needed to achieve
    better outcomes for these patients.

    Follow-​up
    Figure 27.3 Imaging study (MRI): (a) right diffuse hyperplastic perilobar
    nephrogenic rests; (b) regression after two months of vincristin and The most important acute toxicity is venous occlusive disease (VOD),
    actinomycine D. currently often called sinusoid obstruction syndrome (SOS) of the
    liver. This is attributed to relative over-​dosage of actinomycin-​D in
    dystrophic children. Further risk factors are right-​sided tumours,
    Treatment of bilateral nephroblastomas including the liver, treated with postoperative radiotherapy. Other
    The therapeutic goal is to preserve as much healthy kidney tissue as acute side effects are rare and concern mainly infections, neuropathy
    possible. Preoperative chemotherapy is always indicated. Treatment (vincristine), and side effects on the intestine caused by the tumour.
    should be individualized and given as long as it enables renal-​sparing Surgical complications are lower after preoperative treatment and lie
    surgery, but should not exceed 12 weeks. Renal-​sparing surgery on within a range of about 5%. In the long term, one needs to pay atten-
    both sides is possible in more than 50% of patients. Postoperative tion to skeletal growth and soft tissue which may have been affected
    therapy is based on the histology and the local stage, in which the by radiotherapy. Delayed cardiotoxicity (anthracycline) and renal
    highest risk group and the highest local stage of both sides are de- failure are possible and need to be taken seriously. Second malig-
    cisive for the treatment. Postoperative treatment will be given as for nant neoplasms occur in about 1%. A follow-​up schema for patients
    unilateral cases. Essential drugs in the treatment are vincristine and with nephroblastoma is given in Table 27.4.
    actinomycin-​D, which might be supplemented with doxorubicine.
    In cases of no shrinkage of the tumour during preoperative chemo-
    therapy, surgery is of utmost importance. The treatment of high-​risk Clear-​cell sarcoma of the kidney
    tumours is the same as for unilateral cases. The prognosis of bilateral
    tumours is good, with survival rates of about 70% after ten years. Clear-​cell sarcoma of the kidney (CCSK) represents 3% of malig-
    As a late effect, kidney malfunctions are observed that can lead to nant tumours of the kidney. It occurs with a peak incidence at one
    renal failure and dialysis, necessitating renal transplantation. In to four years of age, predominantly affecting boys (in a ratio of 2:1).
    every case, the treatment of such a patient should be carried out in a Four percent of CCSK have metastases at the time of diagnosis. In
    paediatric oncology centre. addition to bone and pulmonary metastases, CCSK may also spread
    to brain and soft tissue. The prognosis for CCSK improved after
    Treatment of adults with a nephroblastoma the introduction of anthracyclines to modern treatment regimens,
    Nephroblastoma is rarely diagnosed in adults. The treatment can be with survival rates approaching 90% for non-​metastatic tumours.
    carried out according to the treatment for children. In adults, the Imaging studies do not allow this tumour to be distinguished from
    tumour is primarily operated. The tumour is frequently more ad- WT. Bone scintigraphy and cranial magnetic resonance imaging
    vanced than in children, with a high rate of patients with metastases (MRI) have to be done in the work-​up of this tumour.
    at initial presentation The reasons for a poorer outcome can partially A large series of 351 cases from the National Wilms Tumour Study
    be attributed to non-​standardized treatments in adulthood. Adult Group (NWTSG, now part of COG) Pathology Centre has provided
    patients who are treated according to the guidelines for children detailed insight into the pathology of CCSK and its microscopic and
    have a better outcome. However, the increased neurotoxicity of vin- immunohistochemical features. According to the SIOP experience
    cristine needs to be considered in this age group. with chemotherapy first, 42% of patients had stage I disease, 23%
    had stage II, 28% had stage III, and 7% had stage IV disease. These
    Treatment of relapses tumours are characteristically composed of a mixture of cord cells
    Approximately 15% of patients with favourable-​histology WT and and very clear septal cells with an extensive capillary network. CCSK
    50% of patients with anaplastic WT develop recurrence. Most re- is considered as a high-​risk renal tumour. Gene-​expression profiling
    currences occur within two years after diagnosis. The general pro- studies have reported upregulation of neural markers and apparent
    file of relapse site shows that the lungs and pleura alone account for expression of members of the sonic hedgehog signalling pathway
    50–​60%, abdominal recurrences make up to 30% (isolated abdomen and the AKT cell-​proliferation pathway (including strong activa-
    or combined to other sites), while other sites (brain or bone) are in- tion of EGFR) in CCSK. Cytogenetic studies of CCSK have reported
    volved in less than 10% of cases. balanced translocations t(10;17)(q22;p13) and t(10;17)(q11;p12)
    Treatment regimens for recurrent WT include drugs that have not and del(14)(q24.1q31.1). Very recently, recurrent internal tandem
    been used during primary chemotherapy. Therapy of recurrent dis- duplications (ITD) of the X-​linked BCL-​6 co-​repressor (BCOR)
    ease depends on the nature of initial treatment and of recognized gene have been described in CCSK.
    238 CHAPTER 27 Wilms and Other Renal Tumours

    Table 27.4 Follow-​up schema for children with nephroblastoma

    End of therapy 1st year 2nd year 3rd–​5th year As of 6th year
    Psychosocial history + Every 12 m Every 12 m Every 12 m Every 12 m
    Clinical examination, RR + Every 3 m Every 3 m Every 6 m Every 12 m
    Imaging studies
    X-​ray of the lung + Every 3 m Every 3 m Every 6 m Every 12 m
    Sonography, abdomen + Every 3 m Every 3 m Every 3 m Every 12 m
    Laboratory
    Blood picture + Every 4 w Every 3 m Every 6 m Every 12 m
    Urinalysis + Every 4 w Every 3 m Every 6 m Every 12 m
    Kidney values + Every 2 m Every 3 m Every 6 m Every 12 m
    Fanconi syndrome* + Every 3 m Every 3 m Every 6 m Every 12 m
    Vaccination status, HBV, HCV, HIV + End of 1st year After vaccination End of 5th year
    T3, T4, TSH** + Every 6 m Every 12 m Every 12 m
    Audiogram*** + Once Once, if path. Once, if path. Once, if path.
    ECG/​Echocardiography**** + Every 6 m Every 12 m Every 12 m Every 24 m
    Bone scintigraphy***** + Every 6 m Only at relapse Only at relapse Only at relapse

    * after ifosfamide, ** after lung irradiation, ***after carboplatin, ****after anthracyclines, ***** in case of clear-​cell sarcoma
    Adapted with permission from Vujanić GM, et al. ‘The UMBRELLA SIOP-​RTSG 2016 Wilms tumour pathology and molecular biology protocol’. Nature Reviews Urology, Volume 15,
    Issue 11, pp. 693–​701. Copyright © 2018 Springer Nature Publishing. DOI: 10.1038/​s41585-​018-​0100-​3

    With current intensive treatment schedules, including radio- and/​or molecular genetic techniques showing the loss of INI1 pro-
    therapy and multi-​ agent chemotherapy regimens, outcome has tein expression resulting from SMARCB1 mutations.
    significantly improved (five-​year EFS ranging from 65% to 85%, The prognosis of patients with RTK remains dismal despite
    five-​year OS ranging from 75% to 90%). Today, NWTS/​COG as well aggressive treatments. RTK is a very chemoresistant tumour.
    as SIOP use four drugs to treat this tumour (cyclophosphamide, Common therapeutic regimens use intensive anthracycline-​based
    VP16, vincristine, and doxorubicine). All patients with this tumour polychemotherapeutic regimens and aggressive local therapy. The
    have to be referred to a centre for paediatric oncology. 142 children recruited into all NWTS studies showed an OS of
    23.2% after four years. Additionally, the outcome of infants under
    six months of age at diagnosis was only 8% compared to 41% in pa-
    Rhabdoid tumour of the kidney tients aged two years or older (p<0·0001). In the UK experience of 11
    patients, OS was 36% at four years. The only statistically significant
    Rhabdoid tumours of the kidney (RTK) are rare (2% of all renal prognostic factor is metastatic disease, as shown in a retrospective
    malignancies) and extremely aggressive. Eighty percent of pa- analysis of 70 patients in Gesellschaft fur Pädiatrische Onkologie
    tients are younger than two years of age and 60% are under the und Hämatologie (GPOH).
    age of one year. There is an overall male predominance (ratio In most reports, treatment is based on regimens with vincris-
    1.5:1). The most common location is the kidney (RTK), but tine, dactinomycin, and doxorubicin (with or without cyclophos-
    brain and soft tissue can be involved. Fever and haematuria in a phamide). In today’s strategies, radical nephrectomy is followed by
    young patient with a renal malignancy with lung metastasis and carboplatin and etoposide alternating with cyclophosphamide and
    hypercalcaemia should suggest the diagnosis of RTK. Among 639 doxorubicin for half a year, including radiotherapy. The use of high-​
    cases of kidney tumours in the first seven months of life with spe- dose chemotherapy followed by autologous stem-​cell transplant-
    cified histology and stage, nine out of eleven stage IV tumours ation is still controversial. Treatment of these patients should be
    were RTK as reported by van den Heuvel-​Eibrink and colleagues centralized and given according to prospective trials. A registry for
    in 2008. Up to 15% of patients with RTK have brain lesions. all rhabdoid tumours, independent of their primary site, is available
    Because of the coincidence with brain metastasis, cerebral MRI in Europe—​the so-​called EURHAB protocol.
    is always indicated.
    Common to rhabdoid tumours are deletions of the tumour-​
    suppressor gene SMARCB1 (hSNF5/​ INI1) on chromosome 22. Renal-​cell carcinoma
    Today, mutations can be detected in at least about 80% of cases on
    chromosome 22q11.2. Germline mutations of SMARCB1 should be Renal-​cell carcinoma (RCC) is a rare malignancy in the kidney
    searched, especially in very young children and in those with syn- of children and accounts for only 1.9—​5% of paediatric renal tu-
    chronous rhabdoid tumours of the CNS and the kidney. mours. The average age of children diagnosed with RCC is about ten
    The diagnosis of RTK can only be made by histology. Today, the years (range 9 to 15 years), with a male predominance. Metastases
    diagnosis of RTK needs to be confirmed by immunohistochemical to the lungs, bone, liver, or brain are identified in about 20% of
    Congenital mesoblastic nephroma 239

    cases. Imaging is not characteristic and shows a non-​specific solid


    intrarenal lesion with little contrast enhancement, and is often less FURTHER READING
    huge than WT. Calcifications are more frequent in RCC than in Suggested reading for nephroblastoma
    Wilms tumour (25% versus 9%, respectively). Chagtai T, et al. (2016) Gain of 1q as a prognostic biomarker in Wilms
    The signs and symptoms of RCC in children are most frequently tumours treated with pre-​operative chemotherapy in the SIOP WT
    abdominal pain and haematuria, sometimes accompanied by fever. 2001 trial: a SIOP Renal Tumours Biology Consortium study. J Clin
    Recent data suggest that paediatric RCC may be different from adult Oncol 34(26), 3195–​203.
    RCC, with distinct morphological characteristics and unique gen- de Kraker J, et al. (2004) Reduction of postoperative chemotherapy
    etic abnormalities. Some studies have shown that most RCC can be in children with stage I intermediate risk and anaplasia Wilms’
    classified according to specific genetic translocations. Such genetic tumour. The SIOP 93–​ 01 randomised trial. Lancet 364(9441),
    translocations involve chromosome Xp11.2 resulting in TFE3 fu- 1229–​35.
    sions. Childhood RCC are more frequently of the papillary subtype Dome JS, et al. (2014) Risk stratification for Wilms tumor: current
    and the Xp11 translocation type or the related t(6;11) translocation approach and future directions. Am Soc Clin Oncol Educ Book 34,
    type. An association of RCC to neuroblastoma is found in children, 215–​23.
    Dome JS, et al. (2015) Advances in Wilms tumor treatment and
    as well as to other syndromes (tuberous sclerosis, urogenital malfor-
    biology: progress through international collaboration. J Clin Oncol
    mations, Saethre–​Chotzen, XYY syndrome). Clear-​cell RCC with
    33, 2999–​3007.
    3p25 abnormalities is rare in children. Furtwängler R, et al. (2014) Pretreatment for bilateral
    Survival of children with RCC depends on the stage of disease at nephroblastomatosis is an independent risk factor for progressive
    presentation and the possibility of complete resection, with an OS disease in patients with stage V nephroblastoma—​the GPOH ex-
    rate of 60–​90%. Lymph-​node (LN) involvement (N+) was found perience. Klin Pädiatrie 226, 175–​81.
    in 45% for which LNs were sampled. The five-​year survival rate is Godzinski J, et al. (2014) Current concepts in surgery for Wilms
    90–​100% for stage I, 50–​80% for stage II and III, and 10% for stage tumour—​the risk-​and function-​adapted strategy. Eur J Pediatr
    IV. Surgical resection of tumour and metastases remains the crucial Surg 24, 457–​60.
    mainstay of successful treatment in paediatric RCC, even in stage IV. Graf N, et al. (2000) The role of preoperative chemotherapy in the
    RCC in children, as in adults, is not chemosensitive. Most localized management of Wilms tumor—​the SIOP studies. Urol Clin North
    RCC can be cured by surgery alone. In patients with distant metas- Am 27, 443–​54.
    tasis, a combination of surgery, radiotherapy, and immunotherapy Graf N, et al. (2012) Characteristics and prognosis of non-​anaplastic
    Wilms’ tumor stage II and III treated according to the SIOP trial
    or targeted therapy (as sunitinib or sorafenib) needs to be discussed
    and study 93–​01. Eur J Cancer 48, 3240–​8.
    in every single case, within an adult multidisciplinary meeting.
    Green DM, et al. (2001) Treatment with nephrectomy only for small,
    stage I/​ favorable histology Wilms’ tumor: a report from the
    National Wilms’ Tumor Study Group. J Clin Oncol 19, 3719–​24.
    Congenital mesoblastic nephroma Grundy PE, et al. (2005) Loss of heterozygosity for chromosomes
    1p and 16q is an adverse prognostic factor in favorable-​histology
    Congenital mesoblastic nephroma (CMN) is a rare paediatric tu- Wilms tumor: a report from the National Wilms Tumor Study
    mour with a favourable clinical outcome. Although the most Group. J Clin Oncol 23, 7312–​21.
    common renal tumour under the age of seven months is a WT, a Ha TC, et al. (2013) An international strategy to determine the role
    relatively high percentage of children suffer from CMN, especially of high dose therapy in recurrent Wilms’ tumour. Eur J Cancer 49,
    in the first two months of life. 194–​210.
    The clinical presentation of CMN is similar to other paediatric Lehnhardt A, et al. (2015) Clinical and molecular characterization of
    renal neoplasms. CMN usually presents as an asymptomatic abdom- patients with heterozygous mutations in Wilms tumour suppressor
    inal mass in an infant. CMN is a spindle-​cell neoplasm with low ma- Gene 1 (WT1). Report of the WT1-​Registry of the GPN. Clin J Am
    Soc Nephrol 10, 825–​31.
    lignant potential. Several subtypes including classical, cellular, and
    Malogolowkin M, et al. (2013) Incidence and outcomes of patients
    mixed were identified on the basis of molecular characterizations
    with late recurrence of Wilms tumor. The international experience.
    and morphological features. The cellular variant, which represents
    Ped Blood Cancer 60, 1612–​15.
    40–​60% of CMN, shows a t(12;15)(p13;q25) translocation with a re- Mitchell C, et al. (2000) The treatment of Wilms’ tumour: results of
    sulting ETV6-​NTRK3 fusion. An identical gene-​fusion transcript the United Kingdom Children’s Cancer Study Group (UKCCSG)
    has also been reported in congenital or infantile fibrosarcoma (IFS), second Wilms’ tumour study. Br J Cancer 83, 602–​8.
    which also affects the same age group and shows similar morph- Neville HL, Ritchey ML (2000) Wilms’ tumor. Overview of National
    ology, suggesting that cellular CMN represents intrarenal IFS. Wilms’ Tumor Study Group results. Urol Clin North Am 27, 435–​42.
    As CMN patients usually have a good outcome, immediate neph- Pastore G, et al. (2006) Malignant renal tumours incidence and survival
    rectomy is advocated and no further treatment needed. Overall EFS in European children (1978–​1997): report from the ACCIS Project.
    is more than 95%, but the occasional stage III cellular CMN in older Eur J Cancer 42(13), 2103–​14.
    patients has a worse prognosis. The main factors of importance as Pein F, et al. (1998) High-​dose melphalan, etoposide, and carboplatin
    risk factors for local recurrence are cellular subtype and incom- followed by autologous stem-​cell rescue pediatric high-​risk recur-
    plete local excision at primary surgery. They rarely relapse within rent Wilms’ tumor: a French Society of Pediatric Oncology study. J
    Clin Oncol 16, 3295–​301.
    12 months after the diagnosis, so during this period close follow-​up
    PerlmanEJ,etal.(2006)Hyperplasticperilobarnephroblastomatosis:long-​
    is recommended.
    term survival of 52 patients. Pediatr Blood Cancer 46, 203–​21.
    240 CHAPTER 27 Wilms and Other Renal Tumours

    Pritchard-​Jones K, et al. (2015) Doxorubicin omission from the report from the National Wilms’ Tumor Study Group. J Clin Oncol
    treatment of stage II/​III, intermediate risk histology Wilms tu- 22, 468–​73.
    mour: results of the SIOP WT 2001 randomised trial. Lancet 386,
    1156–​64. Suggested reading for rhabdoid tumour
    Pritchard-​Jones K, et al. (2016) Evidence for a delay in diagnosis of Biegel JA, et al. (1999) Germ-​line and acquired mutations of INI1 in
    Wilms’ tumour in the UK compared to Germany: implications for atypical teratoid and rhabdoid tumors. Cancer Res 59, 74–​9.
    primary care for children. Arch Dis Child 101, 417–​20. Chi SN, et al. (2008) Intensive multimodality treatment for children
    Reinhard H, et al. (2004) Wilms tumor in adults—​results of the SIOP with newly diagnosed CNS atypical teratoid rhabdoid tumor. J Clin
    93-​01/​GPOH study. J Clin Oncol 22, 4500–​6. Oncol 8, 8–​14.
    Ritchey M, et al. (1992) Surgical complications after nephrectomy for Furtwängler R, et al. (2014) Malignant rhabdoid tumor of the
    Wilms’ tumor. Surg Gynecol Obstetr 175, 507–​14. kidney: significantly improved response to preoperative treatment
    Spreafico F, et al. (2009) Treatment of relapsed Wilms tumors: lessons intensified with doxorubicin. Cancer Genetics 207(9), 434–​6.
    learned. Expert Rev Anticancer Ther 12, 1807–​15. Seeringer A, et al. (2014) Feasibilty of intensive multimodal therapy in
    van den Heuvel-​Eibrink MM, et al. (2015) Outcome of blastemal type infants affected by rhabdoid tumors—​experience of the EU-​RHAB
    Wilms tumour patients treated according to intensified treatment registry. Klin Pädiatr 226, 143–​8.
    in the SIOP WT 2001 protocol—​a report of the SIOP Renal Tumor Weeks DA, et al. (1989) Rhabdoid tumour of kidney: a report of 111
    Study Group (SIOP-​RTSG). Eur J Cancer 51, 498–​506. cases from the NWTS pathology center. Am J Surg Pathol 13,
    Verschuur AC, et al. (2012) Treatment of pulmonary metastases in 439–​58.
    children with stage IV nephroblastoma with risk-​based use of pul-
    monary radiotherapy. J Clin Oncol 30, 3533–​9. Suggested reading for renal-​cell carcinoma
    Vujanic GM, Sandstedt (2010) The pathology of Wilms’ tumour Argani P, Ldanyi M (2006) The evolving story of renal translocation
    (nephroblastoma): the International Society of Paediatric Oncology carcinomas. Am J Clin Pathol 126, 332–​4.
    approach. J Clin Pathol 63, 102–​9. Chowdhury T, et al. (2013) Persistent complete response after single-​
    Wegert J, et al. (2015) Mutations in the SIX1/​2 pathway and the agent sunitinib treatment in a case of TFE translocation positive re-
    DROSHA/​ DGCR8 miRNA microprocessor complex underlie lapsed metastatic pediatric renal cell carcinoma. J Pediatr Hematol
    high-​risk blastemal type Wilms tumors. Cancer Cell 27, 298–​311. Oncol 35(1), e1–​3.
    Weirich A, et al. (2001) Clinical impact of histologic subtypes in local- Geller JI, et al. (2015) Characterization of adolescent and pediatric
    ized non-​anaplastic nephroblastoma treated according to the trial renal cell carcinoma: a report from the Children’s Oncology Group
    and study SIOP-​9/​GPOH. Ann Oncol 12, 311–​19. study AREN03B2. Cancer 121(14), 2457–​64.
    Wilde JCH, et al. (2014) Nephron sparing surgery (NSS) for unilateral Moch H, et al. (2016) World Health Organization Classification
    Wilms tumor (UWT): the SIOP 2001 experience. Ped Blood Cancer of Tumours of the Urinary System and Male Genital Organs.
    61, 2175–​9. Lyon: IARC Press.
    Molina AM, et al. (2014) Sunitinib objective response in metastatic
    Suggested reading for clear-​cell sarcoma
    renal cell carcinoma: analysis of 1059 patients treated on clinical
    Argani P, et al. (2000) Clear cell sarcoma of the kidney: a review of 351 trials. Eur J Cancer 50(2), 351–​8.
    cases from the National Wilms Tumour Study Pathology Center. Mourad WF, et al. (2014) State-​of-​the-​art management of renal cell
    Am J Surg 24, 4–​18. carcinoma. Am J Clin Oncol 37(5), 498–​505.
    Furtwängler R, et al. (2013) Clear cell sarcomas of the kidney regis- Selle B, et al. (2006) Population-​based study of renal cell carcinoma
    tered on International Society of Pediatric Oncology (SIOP) 93–​01 in children in Germany, 1980–​2005. More frequently localized tu-
    and SIOP 2001 protocols: a report of the SIOP Renal Tumour Study mors and underlying disorders compared with adult counterparts.
    Group. Eur J Cancer 49(16), 3497–​506. Cancer 107, 2906–​14.
    Gooskens S, et al. (2012) Clear cell sarcoma of the kidney: a review. Eur
    J Cancer 48, 2219–​26. Suggested reading for congenital mesoblastic nephroma
    Gooskens SL, et al. (2014) Outcome of patients with relapsed clear Furtwaengler R, et al. (2006) Mesoblastic nephroma—​a report from
    cell sarcoma of the kidney (CCSK) registered on recent SIOP, UK-​ the Gesellschaft fur Padiatrische Onkologie und Hamatologie
    CCLG and AIEOP study. Br J Cancer 111, 227–​33. (GPOH). Cancer 106(10), 2275–​83.
    Gooskens SL, et al. (2016) The clinical phenotype of the YWHAE-​ Sebire NJ, Vujanic GM (2009) Paediatric renal tumours: recent devel-
    NUTM2B/​E fusion transcript in paediatric clear cell sarcoma of the opments, new entities and pathological features. Histopathology 54,
    kidney. Genes Chrom Cancer 55(2), 143–​7. 516–​28.
    Kenny C, et al. (2016) Mutually exclusive BCOR internal tandem du- van den Heuvel-​Eibrink MM, et al. (2008) Characteristics and survival
    plications and YWHAE-​NUTM2 fusions in clear cell sarcoma of of 750 children diagnosed with a renal tumor in the first seven
    kidney: not the full story. J Pathol 238(5), 617–​20. months of life: a collaborative study by the SIOP/​GPOH/​SFOP,
    Seibel NL, et al. (2004) Effect of duration of treatment on treatment NWTSG, and UKCCSG Wilms tumor study groups. Pediatr Blood
    outcome for patients with clear-​cell sarcoma of the kidney: a Cancer 50, 1130–​4.
    28
    Neuroblastoma
    Angelika Eggert, Garrett M. Brodeur, and Gudrun Schleiermacher

    Introduction spontaneous regression of neuroblastoma without clinical detection


    is at least as prevalent as clinically detected neuroblastoma.
    Neuroblastoma is one of the most challenging malignancies for The aetiology of neuroblastoma is unknown, and to date there
    paediatric oncologists and one of the most fascinating tumours for are no data supporting a major role for environmental exposures.
    scientists. In 1910, Wright introduced the term ‘neuroblastoma’ Nonetheless, the predisposing effects of prenatal exposure to noxa
    after demonstrating its origin from embryonal neuroblasts of the or maternal food patterns during pregnancy warrant further inves-
    sympathetic peripheral nervous system. The tumour itself was tigation. However, the consistent incidence rates of neuroblastoma
    first described by Virchow, in 1864, in a child with an abdominal in children support the hypothesis that genetic factors play the larger
    mass. Neuroblastoma is characterized by an extreme clinical and role in aetiology. Recent advances in identifying neuroblastoma pre-
    biological heterogeneity, ranging from low-​stage, biologically fa- disposition genes may lead to the identification of individuals who
    vourable tumours that can spontaneously regress or differentiate to are at increased risk and would benefit from future screening efforts.
    highly aggressive, metastasized tumours that progress early or re-
    lapse and cause the death of the patient.
    To date, neuroblastoma serves as a paradigm for the prognostic Tumour predisposition, genetics, and
    benefit of combining molecular data with clinically defined risk molecular pathogenesis
    factors, as well as for the potential to tailor therapy for patient sub-
    groups according to their individual risk for recurrence. However, Familial neuroblastoma
    despite tremendous progress in understanding the pathogenesis A familial history of neuroblastoma is observed in 1–​2% of pa-
    and molecular mechanisms of neuroblastoma, prognosis of patients tients. Neuroblastoma pedigrees show an autosomal dominant pat-
    with advanced disease remains dismal, with long-​term event-​free tern of inheritance with incomplete penetrance. Two main neural
    survival (EFS) of less than 50%. crest-​derived developmental disorders are associated with an in-
    creased risk to develop neuroblastoma: Hirschsprung’s disease and
    Ondine’s curse. Both diseases are linked to mutation of the PHOX2B
    Epidemiology gene, which was subsequently identified as the first predisposition
    mutation in neuroblastoma. PHOX2B encodes a transcription
    Neuroblastoma accounts for 8–​10% of all paediatric cancers. Its age factor that promotes cell-​cycle exit and neuronal differentiation.
    distribution is characterized by a peak of incidence in the first year Perturbations of the PHOX2B-​regulated differentiation pathway
    of life, followed by a rapid decline in the following years. It is rare in the sympathoadrenal lineage may contribute to neuroblastoma
    after the age of six, and is exceptional among adults. Neuroblastoma tumourigenesis. Although the involvement of PHOX2B in familial
    occurs at slightly higher rates in males than in females (male:female cases of neuroblastoma is compelling, its contribution to the de-
    ratio 1.1:1.2). Its mean annual incidence is 7–​12 cases per million velopment of sporadic neuroblastoma is limited, with mutations
    children in western countries, and its prevalence is one in 7,000 live detected in only 5% of cases. Associations between neuroblastic
    births. However, incidence rates are rather heterogeneous, except for tumours and other cancer predisposition syndromes including
    the extremely low rates observed in some African regions. Screening neurofibromin 1 (NF1), defined by the constitutive activation of the
    programmes were implemented in regions of several countries RAS–​MAPK pathway, and Costello syndrome (HRAS) have been
    including Japan, Germany, France, and Canada. An increase in the also reported.
    prevalence of neuroblastoma has been reported in these regions, The discovery of a second neuroblastoma predisposition gene has
    but without a significant decrease in the prevalence or mortality represented a breakthrough in the understanding of neuroblastoma
    of neuroblastoma in patients older than one year. The increased pathogenesis. Activating mutations in the anaplastic lymphoma re-
    prevalence of neuroblastoma in screened populations suggests that ceptor tyrosine kinase (ALK) gene were identified in the majority of
    242 CHAPTER 28 Neuroblastoma

    familial neuroblastoma cases and 10% of sporadic neuroblastomas. survival in multivariate analyses, and is significantly associated with
    ALK encodes a transmembrane receptor tyrosine kinase preferen- aggressive disease markers. Thus, this genomic region might contain
    tially expressed in the developing sympathoadrenal lineage of the one or more tumour-​suppressor genes.
    neural crest, where it may regulate the balance between proliferation Deletions in a consensus region at 11q23 occur in about 40% of
    and differentiation through multiple cellular pathways. Three types cases. Chromosome 11q loss is inversely correlated with MYCN
    of ALK germline mutations with incomplete penetrance have been amplification and, therefore, identifies a molecularly distinct high-​
    described in neuroblastoma families, with the R1275Q substitution risk patient subset. This patient subset is characterized by advanced
    being the most frequent. stage, older age, and unfavourable pathology. Allelic loss at 11q is also
    Germline mutations of ALK and PHOX2B do not account for an independent marker of reduced EFS in low-​and intermediate-​
    all familial neuroblastoma cases. Interestingly, various types of ab- risk subgroups and might be a useful biomarker to improve further
    normal constitutional karyotypes have been observed in patients risk stratification of medium-​risk patients.
    with neuroblastoma, including constitutional copy number anom- The most common aberration in neuroblastoma is the unbalanced
    alies, balanced and unbalanced translocations, as well as specific chromosome 17q21-​qter gain occurring in around 70% of tumours.
    chromosome deletions. Further whole-​genome sequencing (WGS) Numerous studies have reported that 17q gain is significantly associ-
    efforts are likely to identify additional germline mutations in pa- ated with advanced disease stage, increased patient age, 1p LOH, 11q
    tients with a family history of neuroblastoma. LOH, and MYCN amplification. However, the independence of 17q
    gain as a prognostic factor is controversial. Whether an independent
    Sporadic neuroblastoma prognostic factor, or merely a modifying factor, identifying the gene
    Although a few neuroblastomas occur in familial contexts, almost aberrations caused by 17q imbalance will be crucial to fully under-
    98% of cases occur sporadically. In light of the developmental origins, stand neuroblastoma progression.
    it is likely that underlying genetic factors are at least partly respon- The overall genomic profile has been shown to be of prognostic
    sible for the development of sporadic neuroblastomas. Genome-​ impact in neuroblastoma. Whereas numerical chromosome al-
    wide association (GWAS) studies of constitutional DNA samples terations, consisting of gains or losses of whole chromosomes, are
    have identified numerous single-​nucleotide polymorphisms (SNPs) associated with a favourable outcome, segmental chromosome al-
    associated with different clinical features of neuroblastoma and with terations are associated with a higher risk of relapse.
    potential roles in the disease pathogenesis. SNPs have been identi- ALK amplifications and protein overexpression have been de-
    fied in the CASC15, NBAT1, BARD1, LMO1, DUSP12, HSD17B12, scribed in sporadic neuroblastomas in addition to the activating
    DDX4, IL31RA, HACE1,and LIN28B genes. Neuroblastoma can point mutations in ALK. Although the most aggressive somatic ALK
    demonstrate varying levels of genomic instability and harbour a mutation, the F1174L substitution, contributes to neuroblastoma
    wide variety of numerical and structural genetic abnormalities, re- tumourigenesis in mice, this mutation has not been observed in the
    flecting the heterogenous clinical and biological behaviour of the germline in cases of familial neuroblastoma. It appears to behave as
    disease. a superactivator that is not tolerated in the germline. Co-​expression
    Neuroblastoma cells frequently harbour gross cytogenetic ab- of ALKF1174L and MYCN produced a synergistic effect on tumour
    errations such as double-​minute (DM) chromosomes or homoge- development in mouse and zebrafish models, and ALK mutations
    neously staining regions (HSRs) of chromosomes, which contain are consistently correlated with aggressive neuroblastoma subtypes
    ectopic amplifications of the MYCN oncogene. Amplification of the in patients. However, the mechanistic link between ALK mutations
    MYCN oncogene, mapping to 2p24.1, is present in about 25% of all and tumour biology has not yet been fully elucidated. Nonetheless,
    neuroblastomas and 40% of high-​risk tumours. It remains one of ALK is a promising target for molecular therapy in clinical trials re-
    the most important genetic abnormalities associated with advanced cruiting neuroblastoma patients.
    disease stages, a highly malignant phenotype, and poor outcome, Several neuroblastoma deep-​sequencing studies in recent years
    since its discovery by M. Schwab and others in 1983. MYCN status is have consistently reported a low frequency of recurrent mutations.
    assessed in the diagnostic tumour biopsy or a sample of the resected More recent studies investigating relapsed neuroblastomas have
    tumour, and is routinely used in clinical practice to assign thera- suggested that clonal evolution is common and results in the acqui-
    peutic intensity. sition of targetable somatic aberrations in known oncogenic path-
    Several lines of evidence indicate that overexpression of MYC ways. Early evidence suggests that activation of the MAPK pathway
    proteins has a major role in oncogenesis through enhancing the and other signalling pathways inducing epithelial-​mesenchymal
    expression of numerous genes involved in cell proliferation, and transition (EMT) processes are contributing to treatment failure
    repressing expression of genes related to differentiation and apop- and might be promising targets for molecular targeted treatment
    tosis. Ectopic MYCN expression in the neural crest is sufficient to approaches.
    drive neuroblastoma tumourigenesis in zebrafish and mice. As a Inactivating mutations of the ATRX chromatin-​remodelling gene
    major oncogenic driver and one of the few tumour-​specific targets were also recently identified in a small subset of high-​risk neuroblas-
    in paediatric cancer, MYCN deserves to be a priority of therapeutic tomas, and were mutually exclusive to MYCN amplifications. ATRX
    targeting. mutations are associated with activation of a telomere maintenance
    Other structural chromosomal alterations commonly recurring mechanism termed alternate lengthening of telomeres (ALT), which
    in tumours and associated with advanced disease stage and poor may be a marker of an indolent neuroblastoma course with pri-
    outcome in neuroblastoma patients include deletions on 1p, 3p, 4p, mary chemotherapy resistance. Fischer et al. discovered recurrent
    and 11q chromosome arms and gains of a consensus region from genomic rearrangements of the telomerase reverse transcriptase
    2p and 17q. Loss of 1p36 is observed in 20–​35% of cases, predicts (TERT) gene on chromosome 5p15.33 in 13% of neuroblastoma
    Biological aspects 243

    Spontaneous regression/differentiation stromal septa with Schwannian proliferation ranging from none
    to limited. This category consists of three subtypes: undifferenti-
    Low TERT No telomere
    3n LR
    expression elongation ated, poorly differentiated, and differentiating.
    2. Ganglioneuroblastoma, intermixed (Schwannian stroma-​
    Progression/aggressive tumour growth rich): A tumour containing well-​defined microscopic nests of
    Alternative
    HR Low TERT neuroblastic cells intermixed or randomly distributed in the
    lengthening
    2n expression
    of telomeres ganglioneuromatous stroma. The nests are composed of a mix-
    RX
    AT ture of neuroblastic cells in various stages of differentiation.
    TERT High TERT
    2n 2n/4n expression 3. Ganglioneuroblastoma, nodular (composite Schwannian
    Telomerase stroma-​rich/​stroma-​dominant and stroma-​poor): This le-
    MNA High TERT activation
    2n sion is characterized by the presence of grossly visible, usually
    expression
    haemorrhagic neuroblastic nodules (stroma-​poor component,
    Figure 28.1 Current model of neuroblastoma evolution. representing an aggressive clone) coexisting with intermixed
    Adapted with permission from Peifer M, et al. ‘Telomerase activation by genomic ganglioneuroblastoma (stroma-​ rich component) or with
    rearrangements in high-​risk neuroblastoma’, Nature, Volume 526, pp. 700–​704. ganglioneuroma (stroma-​ dominant component), both repre-
    Copyright © 2015, Springer Nature. DOI: https://​doi.org/​10.1038/​nature14980
    senting a non-​aggressive clone. The term ‘composite’ implies that
    the tumour is composed of biologically different clones.
    cases. TERT rearrangements define a subgroup of high-​risk disease 4. Ganglioneuroma (Schwannian stroma-​ dominant): This
    and occur in mutually exclusive fashion with MYCN amplification variant is divided into maturing and mature subtypes. The pre-
    and ATRX mutations. dominant ganglioneuromatous stroma of the maturing subtype
    These data fundamentally advance our understanding of the mo- contains scattered collections of differentiating neuroblasts and/​
    lecular pathogenesis of neuroblastoma (Figure 28.1). Now we know or at least some maturing ganglion cells. Some fully mature
    that the majority of high-​risk neuroblastomas are affected by genetic ganglion cells can also be present in the maturing subtype. The
    alterations of MYCN, TERT, or ATRX, all of which converge on ac- mature subtype contains only mature Schwannian stroma and
    tivation of telomere lengthening mechanisms. Telomeres get short- ganglion cells.
    ened with each cell division in somatic cells, eventually leading to Morphological features used in INPC staging significantly correlate
    replicative senescence or cell death. Cancer cells, in contrast, acquire with biological properties of peripheral neuroblastic tumours, such
    the capability of maintaining their telomeres to provide themselves as MYCN amplification or NTRK1 expression.
    with infinite proliferative capacity. The most aggressive neuro-
    blastoma subtypes appear to be defined by activation of telomerase,
    which is induced either by TERT rearrangements or by MYCN amp- Biological aspects
    lification. In the vast majority of the remaining high-​risk tumours,
    the ALT pathway is activated, which is associated with ATRX muta- A number of biological pathways regulating major hallmarks of
    tions in a fraction of this subtype. Low-​risk tumours appear to lack cancer appear to be disrupted or at least affected in neuroblastoma.
    any telomere maintenance mechanisms, which may explain their Important biological characteristics include abnormal patterns of
    inability to gain immortal proliferative capacity and consequently gene or protein expression that are associated with regions of tu-
    their potential for spontaneous regression. Advances in the develop- mour differentiation, apoptosis, drug resistance, angiogenesis, inva-
    ment of telomerase inhibitors and their evaluation in clinical trials sion, or metastasis. Insight into the molecular regulation of these
    open up a new opportunity to establish effective treatment strategies biological features will ultimately lead to the identification of novel
    for high-​risk neuroblastoma patients. drug targets.
    As neurotrophin signalling has a central role in normal neur-
    onal development and may be involved in both differentiation and
    Pathology regression of neuroblastoma, there has been a particular interest
    in alterations of these pathways. The clinical and biological roles
    Peripheral neuroblastic tumours, including neuroblastoma, belong of neurotrophin receptor tyrosine kinases 1 and 2 (NTRK1 and
    to the small, round, blue-​cell neoplasms of childhood. They are de- NTRK2) and their corresponding ligands have been extensively
    rived from sympathogonia, the progenitor cells of the sympathetic investigated. Both NTRK1 and NTRK2 tumour expression are im-
    nervous system in the sympathoadrenal lineage. After migrating portant prognostic factors influencing the clinical and biological be-
    from the neural crest, these pluripotent sympathogonia form the haviour of neuroblastoma.
    sympathetic ganglia, the chromaffin cells of the adrenal medulla and Delayed activation or disruption of normal apoptotic pathways
    the paraganglia, reflecting the typical localizations of neuroblastic may be an important phenomenon involved in spontaneous regres-
    tumours. sion of neuroblastoma as well as therapy resistance. Abnormal expres-
    The International Neuroblastoma Pathology Classification sion or activation patterns occur in major proteins in the apoptotic
    (INPC) developed by Shimada assigns peripheral neuroblastic tu- signalling cascade, including the BCL2 family, BIRC5, and caspase
    mours to one of four basic morphological categories: 8. Caspase 8 inactivation mainly occurs via epigenetic silencing.
    1. Neuroblastoma (Schwannian stroma-​poor): A tumour com- Alterations in DNA methylation represent one of the most common
    posed of neuroblastic cells forming groups or nests separated by molecular events in neoplasia, and CpG-​island hypermethylation of
    244 CHAPTER 28 Neuroblastoma

    Table 28.1 Tumour site and clinical characteristics

    Tumour site Main symptoms and signs Incidence (%) Metastases (%) Outcome
    Neck Mass mimicking lymphadenopathy, Horner syndrome 4 2 Good
    Thorax: Upper Respiratory symptoms 6 5 Good
    Thorax: Lower Incidental X-​ray detection 6 2 Good
    Abdomen Central or lateral mass, symptoms related to metastatic 70 75 Poor
    disease (fever, pallor, anorexia, bone pain, irritability)
    Pelvis Lower abdominal mass, dysuria, constipation 5 5 Good
    Not identified Fever, pallor, anorexia, bone pain, irritability 1 100 Very Poor

    gene promoters is a frequent mechanism for functional inactivation Altogether, neuroblastoma infiltration with suppressive mye-
    of relevant tumour-​associated genes. loid cells can decrease anti-​tumour immune responses and favour
    Acquired resistance to chemotherapeutic agents may also be tumour growth in high-​risk disease. Both active and passive im-
    conferred by enhanced drug efflux via overexpression of classical munotherapies, including antibody-​based therapies, are now being
    multidrug resistance proteins, including the ATP-​binding cassette explored for neuroblastoma patients.
    subfamily B member 1 (ABCB1), and ATP-​binding cassette sub-
    family C member 1 (ABCC1) proteins. The ATP-​binding cassette
    subfamily G member 2 (ABCG2) is another important contributor Clinical presentation, diagnosis, and staging
    to a drug-​resistance phenotype in neuroblastomas, especially imma-
    ture and stem-​like cell populations. Their potential clinical signifi- Neuroblastoma can arise at any site along the sympathetic nervous
    cance in neuroblastoma has been addressed in several studies. system chain, but the distribution varies across the sites with age.
    Enhanced tumour angiogenesis and elevated expression of About 70% of primary tumours occur within the abdomen, and
    proangiogenic factors, such as vascular endothelial growth factor more than half of these arise in the adrenal gland. Infants more
    A (VEGFA), both correlate with an aggressive neuroblastoma frequently present with adrenal, thoracic, and cervical primary tu-
    phenotype. Inhibitors of angiogenesis are currently being evaluated mours. The clinical presentation of neuroblastoma reflects the site of
    in clinical trials as a potential treatment option. Deep-​sequencing tumour origin, the extent of regional and metastatic disease, and the
    approaches for comprehensive molecular characterization of neuro- presence of paraneoplastic syndromes (Table 28.1).
    blastoma are currently doing a great deal to advance our under- Cervical neuroblastomas (4% of cases) mimic adenopathies and
    standing of neuroblastoma biology. A detailed molecular profile often include Horner syndrome (ptosis, miosis, anhydrosis) and
    from the tumour, metastases, and/​or circulating tumour cells or heterochromia. Thoracic neuroblastomas represent 15% of all cases.
    tumour-​associated DNA in the peripheral bloodstream could allow Tumours in the upper mediastinum cause respiratory distress as well
    patient-​tailored precise prognostication, while also identifying key as Horner syndrome, while tumours arising in the middle and lower
    targets that could be therapeutically exploited. mediastinum are usually asymptomatic and discovered on routine
    It is likely that immune factors and interactions with the tumour chest X-​ray. The typical neuroblastoma patient is a toddler, failing
    microenvironment contribute to the variation in neuroblastoma to grow, with abdominal distension and a palpable lateral or cen-
    biology. These can also strongly influence the efficacy and application tral hard, fixed mass. Pelvic tumours are uncommon and present as
    of immunotherapies, which currently hold high potential for high-​ suprapubic masses causing constipation and urinary disturbances.
    risk patients. Since age at diagnosis is of strong prognostic impact in Approximately half of all patients have disseminated disease at the
    patients with neuroblastoma, it could be suggested that stage-​and age-​ time of diagnosis. The most commonly involved organs are the bone
    related differences in the host reaction to a developing tumour might marrow, skeleton, liver, and lymph nodes, with less common involve-
    influence outcome. Further evidence for the involvement of the im- ment of the lung and central nervous system. Disseminated disease
    mune system in neuroblastoma control derives from observations of is usually associated with non-​specific symptoms, including fever,
    the neuroblastoma-​associated paraneoplastic syndrome, opsoclonus-​ pallor, anorexia, and bone pain, with consequent refusal to walk and
    myoclonus syndrome (OMS). Lymphoid infiltrates are frequently ob- mood changes. Retro-​orbital and orbital metastases are rather fre-
    served in the tumours in OMS patients, and these patients typically quent, and produce a typical appearance of proptosis and periorbital
    have an excellent oncological prognosis. Higher levels of CD163+ M2 ecchymoses (Figure 28.2). In infants, stage 4S or MS is characterized
    macrophages have also been detected in tumour tissue from patients by massive liver involvement with or without the presence of bluish
    with INSS (International Neuroblastoma Staging System) stage 3 or nodules in the subcutaneous tissue and bone-​marrow infiltration.
    4S neuroblastoma compared with tumour tissue from patients with Infants with stage 4S can present with significant respiratory distress
    INSS stage 4 metastatic disease, a finding associated with outcome in due to a massively enlarged liver. By definition, stage 4S patients do
    several other cancers. Age-​dependent differential expression of in- not have bone lesions. Neuroblastoma in adolescents and adults has
    flammatory genes associated with tumour-​associated macrophages a slow, indolent clinical course with a worse overall outcome com-
    has been demonstrated in neuroblastomas. Cancer-​associated fibro- pared to that in children, although the primary tumour and site dis-
    blasts in the neuroblastoma microenvironment have been suggested tributions and metastatic patterns are similar.
    to contribute to immunosuppression, which could consequentially The anatomical connection between the sympathetic ner-
    favour tumour development. vous system and the spinal cord accounts for the propensity of
    Clinical presentation, diagnosis, and staging 245

    Figure 28.2 Retro-​orbital and orbital metastases with a typical appearance of proptosis and periorbital ecchymoses in a child with stage 4
    neuroblastoma.

    neuroblastoma to infiltrate the intervertebral foramina. Paraspinal concurrent or retrospective mortality. As a result, routine screening
    tumours tend to grow through the neural foramina of the vertebral was ceased because of the low clinical impact for treating neuro-
    bodies, causing spinal-​cord compression as a presenting sign of the blastoma. However, a great deal of biological information on favour-
    tumour. Motor deficits are the most frequent symptoms experienced able neuroblastoma subtypes has been derived from these studies.
    by patients with paraspinal tumours, followed by radicular or back Lactate dehydrogenase (LDH), ferritin, and enolase 2 (formerly
    pain, bladder or bowel dysfunction, and, rarely, sensorial manifest- neuron-​specific enolase (NSE)) have been identified as markers for
    ations. Only 5% of patients have clinical evidence of epidural com- neuroblastoma in patient serum, and correlate with tumour stage,
    pression. Early detection of this phenomenon is crucial because extension, and prognosis. New serum tumour markers and useful
    symptoms may rapidly worsen, leading to irreversible paraplegia. molecular markers to monitor minimal residual disease might be
    Ganglioneuroblastomas are well-​differentiated locoregional tu- identified by ongoing liquid-​biopsy studies.
    mours, and may occasionally cause symptoms either directly related
    to effects of the tumour mass or related to paracrine secretion of Diagnostic criteria
    various substances. Ganglioneuromas can entrap and impinge on The following examinations are required for accurate diagnosis and
    vessels, nerves, vessels, gastrointestinal (GI) track, or other tissues. staging according to the INSS:
    Vasoactive intestinal peptide (VIP) causes watery diarrhoea, which
    • Magnetic resonance imaging (MRI) or computed tomography
    may require treatment with somatostatin. Antineuronal antibodies
    (CT) of the primary tumour site.
    can trigger an autoimmune reaction that may occur as a cerebellar
    • Biopsy of the primary tumour or metastatic site for histological
    syndrome or OMS. OMS occurs in 2% of neuroblastoma patients
    examination and determination of biological prognostic factors
    and is characterized by multidirectional rapid eye movement
    (MYCN amplification).
    (opsoclonus), myoclonus, and brainstem ataxia. These symptoms
    may precede the detection of a tumoural mass by several months • Bone-​marrow aspirate and trephine biopsy at two different sites.
    and improve if the primary tumour is removed, but in many cases • 123I Metaiodobenzylguanidine (MIBG) scintigraphy before tu-
    patients require specific immunosuppressive treatment. Symptoms mour excision to evaluate the primary tumour site and detect
    may reappear during infectious episodes and the majority of these metastatic sites (Figure 28.3). Detection of a single equivocal le-
    children appear to have long-​term neurological deficits. sion requires confirmation by another imaging modality.
    Increased catecholamine levels may cause high blood pressure, in • Bone scintigraphy or 18-​ fluorodeoxyglucose (FDG)-​ positron
    particular when elevated epinephrine or norepinephrine levels are emission tomography (PET)/​CT is required only in exceptional
    present, although hypertensive crises are very rare in patients with cases when MIBG cannot confirm a primary tumour. An isolated
    neuroblastoma. Hypertension may also be caused by stimulation of bone uptake should be confirmed by another imaging modality
    the renin-​angiotensin system from pressure by the retroperitoneal and/​or a biopsy.
    mass on the renal artery. • Assay of urinary catecholamine levels (HVA and VMA).
    Additional assays should include serum ferritin, LDH, and NSE
    Tumour markers levels.
    Urinary metabolites produced by catecholamine degradation,
    namely vanillylmandelic acid (VMA) and homovanillic acid (HVA), Staging
    are the most sensitive (90% positive cases) and specific disease Disease staging is necessary to define prognosis and select appro-
    markers. Plasma-​free and total normetanephrine, metanephrine, priate treatment intensity. The INSS definitions are based on local
    and methoxytyramine represent a convenient alternative to urine and distant extension as well as tumour resectability (Table 28.2).
    markers. Overall levels and the ratio of VMA to HVA in the urine The INSS stage of locoregional tumours is dependent on the de-
    have prognostic value, and are related to age, histology, stage, and gree of surgical resection, which is subjectively assessed and prone
    MYCN amplification. Screening for neuroblastoma using a urinary to variation between surgeons. The International Neuroblastoma
    assay for catecholamines at six months of age was pioneered in Japan Risk Group (INRG) classification system was developed to facilitate
    in the early 1970s. Two large population-​based studies in Quebec comparison of risk-​based clinical trials conducted in different world
    and Germany showed that screening at 12 months of age detected regions by defining homogeneous pretreatment patient cohorts. The
    many neuroblastomas that would never have been diagnosed clin- premise is that a staging system based on preoperative, diagnostic
    ically, but observed no reduction in the incidence of advanced-​ images will be more robust and reproducible than one based on op-
    stage neuroblastoma. Cumulative mortality was not different from erative findings and approaches. Since the surgical risk factors are
    246 CHAPTER 28 Neuroblastoma

    (a) (b) Table 28.2 The International Neuroblastoma Staging System (INSS)

    Stage Description
    Stage 1 Localized tumour with complete gross excision, with or without
    microscopic residual disease. Representative ipsilateral lymph
    node microscopically negative for tumour.
    Stage 2A Localized tumour with incomplete gross excision. Representative
    ipsilateral non-​adherent lymph nodes microscopically negative
    for tumour.
    Stage 2B Localized tumour with or without complete gross excision,
    with ipsilateral non-​adherent lymph nodes positive for tumour.
    Enlarged contralateral lymph nodes must be microscopically
    negative for tumour.
    Stage 3 Unresectable unilateral tumour, infiltrating across the midline,
    with or without regional lymph-​node involvement; or localized
    unilateral tumour with contralateral lymph-​node involvement; or
    midline tumour with bilateral extension by infiltration or lymph-​
    node involvement.
    Stage 4 Any primary tumour with dissemination to distant lymph nodes,
    bone, bone marrow, liver, and/​or other organs (except as defined
    for Stage 4S).
    Stage 4S Localized primary tumour (as defined for Stage 1, 2A, or 2B), with
    dissemination limited to liver, skin, and/​or bone marrow*. Limited
    to infants <18 months of age.

    ** Marrow involvement in stage 4S should be minimal (<10% nucleated cells in bone-​


    marrow biopsy). More extensive marrow involvement should be considered stage 4. The
    MIBG scan should be negative in the marrow for stage 4S.
    Figure 28.3 MIBG scan: (a) uptake by the primary tumour (right adrenal Reproduced with permission from Brodeur GM, et al. ‘Revisions of the international
    criteria for neuroblastoma diagnosis, staging, and response to treatment’, Journal of
    gland) as well as bone and bone-​marrow metastases are visualized; Clinical Oncology, Volume 11, Issue 8, pp. 1466–​7. Copyright © 1993 American Society
    (b) normal uptake after chemotherapy. of Clinical Oncology. DOI: 10.1200/​JCO.1993.11.8.1466

    based on radiographic images, the term ‘image-​defined risk factors’ The assessment of more complex molecular patterns and the re-
    (IDRFs) was chosen, and a consensus was reached for the IDRFs cent establishment of mRNA-​based molecular risk classifiers will
    listed in Table 28.3. allow more refined and precise risk-​stratification approaches in the
    These criteria, based on the relation of the tumour with the adja- near future, and will consequently enable physicians to tailor appro-
    cent structures and vasculature (Figure 28.4) predict severe surgical priate treatment strategies to individual patients.
    complications and are the basis for the new INRG Staging System
    (INRGSS; Table 28.4).
    Several international groups have developed models of clinical-​ Treatment
    risk stratification to facilitate the delivery of risk-​adapted treatments.
    Various biological features were added to the different clinical char- The treatment of children with neuroblastoma primarily depends
    acteristics, but the most important prognostic factors remain pa- on patient age, disease extension, and some features of tumour
    tient age, tumour stage, and MYCN amplification. Two extremely histology and biology. The currently employed treatment mo-
    different clinical courses of disease can easily be discriminated. The dalities include surgery, chemotherapy, radiotherapy, immuno-
    favourable subtype of neuroblastoma arises in the first months of therapy, and biological response modifiers. Major clinical phase
    life, with many patients showing spontaneous regression of the dis- 3 randomized trials from the European International Society of
    ease and most having excellent survival with minimal or no treat- Paediatric Oncology European Neuroblastoma Network (SIOPEN),
    ment, provided the tumour is not MYCN amplified. The aggressive, the Children’s Oncology Group (COG), the German Society for
    unfavourable neuroblastoma subtype occurs in children older than Paediatric Oncology and Haematology (GPOH), and the Japanese
    18 months and is associated with a poor outcome. It presents with Neuroblastoma Study Group (JNBSG) are currently being per-
    metastatic disease at the time of diagnosis, and tumours often har- formed, have completed inclusion, or are in preparation.
    bour a MYCN amplification. Between these two extremes there are
    Surgery
    less well-​defined subgroups with intermediate characteristics.
    The INRG Task Force has developed the INRG Consensus Surgery plays a key role for both diagnosis and treatment. The
    Pretreatment Classification Schema to establish a consensus ap- goals of primary surgery are to confirm the diagnosis, acquire
    proach for current pretreatment risk stratification. The prognostic tissue samples for classification as well as for histological and bio-
    effect of 13 variables in an 8,800-​patient cohort was analysed, and a logical studies, and to resect the tumour with minimal morbidity.
    schema with four main prognostic groups (Table 28.5) and 16 pre- When the anatomical characteristics (site, size, and relationship
    treatment designations was developed (Table 28.6). This approach with the surrounding structures) indicate that surgical resection is
    greatly facilitates the comparison of risk-​based clinical trials con- feasible, surgery is the treatment of choice for patients presenting
    ducted in different regions of the world. with localized disease. An international panel of surgeons and
    Treatment 247

    Table 28.3 Image-​defined risk factors (IDRFs) in neuroblastic tumours

    Ipsilateral tumour extension within two body compartments


    Neck
    • Tumour encasing carotid and/​or vertebral artery and/​or internal jugular vein
    • Tumour extending to base of the skull
    • Tumour compressing the trachea
    • Cervico-​thoracic junction
    • Tumour encasing brachial plexus roots
    • Tumour encasing subclavian vessels and/​or vertebral and/​or carotid artery
    • Tumour compressing the trachea
    Thorax
    • Tumour encasing the aorta and/​or major branches
    Figure 28.4 Angio-​CT, coronal view, showing a large abdominal
    • Tumour compressing the trachea and/​or principal bronchi
    neuroblastoma encasing the major retroperitoneal vessels.
    • Lower mediastinal tumour, infiltrating the costo-​vertebral junction between
    T9 and T12
    Thoraco-​abdominal of the locoregional lymph nodes, especially in abdominal and
    pelvic sites. In paravertebral sites with infiltration through the
    • Tumour encasing the aorta and/​or the vena cava
    intervertebral foramina, laminectomy is indicated only in selected
    • Abdomen/​pelvis cases with acute, severe neurological symptoms. In fact, chemo-
    • Tumour infiltrating the porta hepatis and/​or the hepatoduodenal ligament therapy can rapidly reduce both tumour size and the impact of
    • Tumour encasing branches of the superior mesenteric artery and the spinal-​cord compression in these cases.
    mesenteric root In contrast to its pivotal role in treating localized disease, the
    • Tumour encasing the origin of celiac axis and/​or of the superior mesenteric function of aggressive surgery remains controversial for metastatic
    artery disease. Meta-​analyses in large patient cohorts are hampered by
    • Tumour invading one or both renal pedicles fragmented and inconsistent documentation of surgical approaches
    and a lack of immediate postoperative imaging results. Both limi-
    • Tumour encasing the aorta and/​or the vena cava
    tations should be addressed in upcoming trial protocols. However,
    • Tumour encasing the iliac vessels considering the high incidence of local relapse, the current strategy
    • Pelvic tumour across the sciatic notch in most treatment protocols is to resect the primary tumour after
    Intraspinal tumour inducing remission of the metastases. In delayed primary or second-​
    look surgery, the surgeon evaluates therapy response and removes
    • Extension, regardless of location, provided that more than one third of
    the spinal canal in the axial plane is invaded and/​or the perimedullary the residual tumour to a maximum extent, whenever possible.
    leptomeningeal spaces are not visible and/​or the spinal cord signal is
    abnormal Radiotherapy
    Infiltration of adjacent organs/​structures Neuroblastoma is a radiosensitive tumour, and tumouricidal doses
    • Pericardium, diaphragm, kidney, liver, duodeno-​pancreatic block and range from 15 to 36 Gy (with fraction doses ranging from 150 to 400
    mesentery cGy) depending on tumour site, volume, and patient age. Definition
    Conditions to be recorded, but not considered as IDRFs of the role of external-​beam radiotherapy (EBRT) and the optimal
    • Multifocal primary tumours mode of its application are under continuous refinement. The lack of
    randomized trials addressing the contribution of radiotherapy ham-
    • Pleural effusion, with or without malignant cells
    pers proper evaluation of the impact of this treatment modality on
    • Ascites, with or without malignant cells clinical outcome to date. EBRT is also successfully used as palliative
    Adapted with permission from Monclair T, et al. ‘The International Neuroblastoma Risk treatment on painful sites.
    Group (INRG) Staging System: An INRG Task Force Report’, Journal of Clinical Oncology, Radiometabolic salvage therapy is used for high-​risk neuroblas-
    Volume 27, Issue 2, pp. 298–​303. Copyright © 2009 American Society of Clinical
    Oncology. DOI: 10.1200/​JCO.2008.16.6876 tomas with active residual tumour prior to myeloablative chemo-
    therapy. It utilizes 131I carried by 131I-​MIBG, a noradrenalin analogue
    radiologists identified surgical risk factors detectable by imaging that is actively taken up by tumour cells via the norepinephrine
    and established radiological criteria for a safe resection. Surgery is transporter and incorporated into the neurosecretory granules of
    the only treatment necessary to cure the majority of patients with neuroblastoma cells. Dosimetry problems, together with the non-​
    localized disease, even if macroscopic remnants of the tumour homogeneous uptake of 131I-​MIBG throughout the tumour and
    are left behind. If surgical risk factors are detected, presurgical toxicity have limited the use of this therapeutic approach to selected
    chemotherapy is necessary to shrink the tumour prior to resec- treatment centres. Some groups have used radiometabolic therapy as
    tion. However, frequently IDRFs persist after neoadjuvant chemo- first-​line treatment, but the results from long-​term patient follow-​up
    therapy. Since neuroblastoma has an elevated tropism towards were not favourable in these cases. Other approaches have included
    lymphatic vessels, it is important to perform a surgical exploration general radiometabolic therapy using 131I-​MIBG in the conditioning
    248 CHAPTER 28 Neuroblastoma

    Table 28.4 International Neuroblastoma Risk Group Staging survival rates. The predicted survival is almost 100%, both for pa-
    System (INRGSS) tients who are treated with a moderate dose of chemotherapy (COG
    and SIOPEN groups) and patients who are treated without cyto-
    Stage Description
    toxic agents (Memorial Sloan–​Kettering Cancer Center (MSKCC)
    Stage L1 Radiological risk factors absent. Localized tumour not involving
    vital structures as defined by the list of image-​defined risk
    and German groups). German studies have shown that infants with
    factors and confined to one body compartment. radiologically diagnosed neuroblastoma may safely be observed over
    Stage L2 Locoregional tumour with presence of one or more image-​ time without obtaining a definitive surgical histological diagnosis,
    defined risk factors. thus avoiding the potential complications of surgery. Chemotherapy
    Stage M Distant metastatic disease (in contrast to stage MS). is mandatory if severe symptoms (especially neurological symp-
    toms) are present or if tumour progression occurs. Neuroblastomas
    Stage MS Metastatic disease in children <18 months with metastases
    confined to the skin, liver, and/​or bone marrow (bone-​marrow lacking MYCN amplification or in the stage 4S/​MS group undergo
    involvement should be limited to <10% of total nucleated cells spontaneous regression in the majority of cases. Chemotherapy or
    on smears or biopsy). low-​dose radiotherapy should be reserved for patients with large tu-
    Reproduced with permission from Monclair T, et al. ‘The International Neuroblastoma mours or massive hepatomegaly causing mechanical obstruction,
    Risk Group (INRG) Staging System: An INRG Task Force Report’, Journal of Clinical respiratory insufficiency, or liver dysfunction.
    Oncology, Volume 27, Issue 2, pp. 298–​303. Copyright © 2009 American Society of
    Clinical Oncology. DOI: 10.1200/​JCO.2008.16.6876
    Intermediate-​risk group
    The intermediate-​risk group encompasses a wide spectrum of con-
    phase prior to haematopoietic stem-​cell transplantation. Its use as ditions, including patients with stage 3 (L2) disease with tumours
    salvage treatment in relapsed disease is under investigation. lacking MYCN amplification who are more than 18 months of age,
    Chemotherapy and patients with stage 4 (M) disease who are under 18 months
    old. Stage 1 (L1) MYCN-​amplified tumours have also recently been
    Chemotherapy has an important role in treating neuroblastoma, included in this group by SIOPEN. Survival is nearly 90% in this
    since the majority of patients present with metastatic or locally group, but may be poorer in some subgroups, and so the challenge
    advanced disease at diagnosis and require systemic treatment. is to identify patients for whom therapy may be further reduced as
    Alkylating agents (cyclophosphamide, ifosfamide, busulfan, opposed to those who require more treatment. Surgical resection
    melphalan), platinum analogues (cisplatin and carboplatin), vinca and moderate-​dose, multi-​agent chemotherapy are the backbone
    alkaloids (vincristine), epipodophyllotoxins (VP16, VM26), and of therapy. High-​dose chemotherapy with PBSC rescue is not pro-
    anthracyclines (doxorubicin) have well-​established activity and ef- posed, and the role of radiotherapy is less well-​defined.
    ficacy against neuroblastoma, and are considered standard options.
    Only a few new drugs have been introduced into first-​line therapy in High-​risk group
    recent years. Topotecan, irinotecan, and temozolomide are expected This group includes all stage 4 (M) patients more than 18 months of
    to slightly improve response in high-​risk neuroblastoma patients. age and stage 2, 3, and 4S patients with MYCN-​amplified tumours.
    High-​dose chemotherapy with various combinations of busulfan, These patients are generally treated with dose-​intensive, multi-​agent
    melphalan, carboplatin, and etoposide, followed by autologous per- chemotherapy in the induction phase that consists of high-​dose cis-
    ipheral blood stem-​cell (PBSC) rescue, is currently used as consoli- platin with etoposide, vincristine, doxorubicin, and cyclophospha-
    dation treatment in high-​risk patients. mide or ifosfamide, and, more recently, topotecan. Current treatment
    Risk-​adapted treatment regimens for induction chemotherapy employed by the SIOPEN,
    GPOH, and COG trial groups will be compared in a randomized
    Low-​and very low-​risk groups fashion in the near future to identify the most beneficial approach.
    This group includes children with INSS stage 1 and 2, infants with Induction chemotherapy is followed, where possible, by primary
    stage 3 (L2) and 4S neuroblastomas without unfavourable bio- tumour resection, then myeloablative chemotherapy with PBSC
    logical markers. Since these patients have an excellent prognosis, rescue in the consolidation phase. The idea of eliminating resistant
    the goal is to reduce the risk of side effects while maintaining high tumour clones with supralethal chemotherapy has been studied
    in neuroblastoma since the early 1980s, and the results of several
    randomized phase III cooperative trials (COG 4 and GPOH NB
    1997) have shown improved EFS. Current international evidence
    Table 28.5 The four main prognostic neuroblastoma subgroups
    points to a superiority of busulphan/​melphalan in comparison to
    Pretreatment risk 5-​year EFS (%) Proportion of patients (%) carboplatin, etoposide, and melphalan as the conditioning regimen
    group prior to PBSC. Recent data also suggest that tandem intensification
    Very low >85 28.2 is feasibile and may be of benefit for certain patients.
    Low >75 to ≤85 26.8 Irradiation of the tumour bed or residual primary tumour is per-
    formed after myeloablative therapy. Most protocols usually em-
    Intermediate ≥50 to ≤75 9.0
    ploy a dose of 21 Gy to the primary site as consolidation therapy
    High <50 36.1
    (SIOPEN) or 36 Gy to MIBG-​active residual tumour at the primary
    Reproduced with permission from Susan L. Cohn, et al. ‘The International site (GPOH). The indication for irradiating metastatic lesions is less
    Neuroblastoma Risk Group (INRG) Classification System: An INRG Task Force Report.’ clear, other than for palliation. Randomized trials and implemen-
    Journal of Clinical Oncology, Volume 27, Issue 2, pp. 289–​97. Copyright © 2009 American
    Society of Clinical Oncology. DOI: 10.1200/​JCO.2008.16.6785. tation of quality-​assured radiotherapy are now being set up since
    Treatment 249

    Table 28.6 International Neuroblastoma Risk Group (INRG) Treatment Classification System

    INRG stage Age (months) Histological Grade of tumour MYCN 11q aberration Ploidy Pretreatment risk
    classification differentiation group
    L1/​L2 Any GN maturing A Very low
    GNB intermixed
    L1 Any Any except NA B Very low
    GN maturing or Amp K High
    GNB intermixed
    L2 <18 Any except No D Low
    GN maturing or NA Yes G Intermediate
    GNB intermixed
    ≥18 GNB nodular; Differentiating NA No E Low
    neuroblastoma Poorly differentiating or NA Yes H Intermediate
    undifferentiating
    L2 Any NA High risk
    M <18 NA Hyperdiploid F Low
    <12 NA Diploid I Intermediate
    12 to <18 NA Diploid J Intermediate
    <18 Amp O High
    ≥18 P High
    MS <18 NA No C Very low
    Yes Q High
    Amp R High

    Amp = amplified; GN = ganglioneuroma; GNB = ganglioneuroblastoma; NA = not aftablemplified


    Reproduced with permission from Susan L. Cohn, et al. ‘The International Neuroblastoma Risk Group (INRG) Classification System: An INRG Task Force Report.’ Journal of Clinical
    Oncology, Volume 27, Issue 2, pp. 289–​97. Copyright © 2009 American Society of Clinical Oncology. DOI: 10.1200/​JCO.2008.16.6785.

    current irradiation protocols, including fractionation protocols, are is still under investigation. Future trials will mostly likely assess
    rather heterogeneous. the effect of ch14.18 chimeric antibodies in parallel to induction
    Most high-​risk neuroblastomas initially respond to therapy but chemotherapy.
    subsequently relapse. Therefore, much attention has been given to
    biological therapies including differentiation-​inducing agents such Treatment of spinal-​cord compression
    as retinoid derivatives or immunotherapy with monoclonal anti- The routine use of MRI has increased the number of cases with docu-
    bodies (MoAbs) targeting the disialoganglioside GD2. Retinoids mented infiltration of neural foramina or the spinal canal. In the
    are a class of compounds that can induce terminal differentiation majority of cases, no neurological symptoms arise and no specific
    of neuroblastoma cells in vitro. Current international evidence for a treatment is required. There is very little evidence that an asymp-
    clinical benefit of retinoic-​acid therapy in high-​risk neuroblastoma tomatic intraspinal tumour will continue to grow after resection
    patients is very limited and does not necessarily warrant further im- of the extraspinal component. Some cases suggest that intraspinal
    plementation of this approach in first-​line treatment. neuroblastoma in patients with no neurological symptoms might re-
    Although the profound immunosuppression produced by main stable or even regress without specific treatment. If more than
    high-​dose chemotherapy regimens provides unfavourable con- 50% of the spinal canal is taken by the tumour or in case of clinical
    ditions for the application of active immunotherapy, the use of symptoms of spinal-​cord compression, dexamethasone should be
    passive antibody-​ based immunotherapy is feasible. GD2 is a administered in addition to chemotherapy.
    surface glycolipid antigen present on neuroblastoma cells. It is Patients with neuroblastoma who present with signs of spinal-​cord
    expressed only weakly in the normal tissues of the central and compression require urgent, specific treatment. Rapid therapeutic
    peripheral nervous systems. Therapies using various MoAbs have decisions must be made within a few hours if neurological deficits
    been assessed in phase I and II settings, and their safety profile and clinical progression are detected. The decision whether to ad-
    has been established. Anti-​GD2 therapies have been studied, minister emergency chemotherapy should be made immediately
    with or without cytokines, in phase III trials for neuroblastoma after detailed conference between the oncologist and neurosurgeon,
    patients in first response. A COG randomized phase III trial of and the start of treatment should never be delayed. Laminectomy
    the anti-​GD2 ch14.18 MoAb plus cytokines after PBSC rescue should be considered only in infants showing very rapid neuro-
    demonstrated a clear survival advantage in patients receiving im- logical deterioration due to the long-​term consequences of this
    munotherapy compared to untreated controls. Application of the procedure. If chemotherapy has been started, the tumour should be
    anti-​GD2 ch14.18 MoAb without immune-​modulating cytokines biopsied when the patient is stable a few days after chemotherapy.
    250 CHAPTER 28 Neuroblastoma

    Treatment of relapse Targeting MYCN


    Survival of children with recurrent or refractory neuroblastoma re- In light of the frequency and importance of MYCN amplification
    mains very poor, and very few can be salvaged. A small cohort of in the pathogenesis of high-​risk neuroblastoma, blocking MYCN
    patients, mainly represented by children with stages 1 and 2 who signalling represents an important approach for developing new
    developed local recurrence or late relapse, may still benefit from therapeutics. Efforts are limited by difficulties associated with se-
    further conventional treatment including aggressive surgery, ex- lective pharmacological inhibition of nuclear transcription factors,
    ternal or metabolic radiotherapy, and chemotherapy. However, ex- but various strategies aimed at counteracting MYCN activity more
    perimental therapies are proposed for the majority of patients. An indirectly are actively being pursued:
    umbrella trial incorporating all promising current options for re-
    Targeting MYCN binding to DNA. Development of small mol-
    lapsed neuroblastoma treatment is currently being developed by the
    ecules directly targeting MYC family members has focused on
    SIOPEN trial group.
    blocking the interaction between MYCN and MAX, which is ne-
    New drugs cessary for DNA binding.
    Targeting MYCN transcription. Bromodomain and extra-​
    To best confront the heterogenous nature of neuroblastoma,
    terminal (BET) family adaptor proteins localize to MYC/​MYCN
    comprehensive molecular profiling must be utilized to tailor
    promoters. The prototypic BET inhibitor, JQ1, disrupts BET re-
    treatment and identify novel therapeutic targets. The major hope
    cruitment to chromatin and has been shown to downregulate
    is now focused on new drugs directed against these molecular
    the MYC/​MYCN transcriptional programme and suppress tran-
    targets. Here, we concentrate on the most promising ones close
    scription of MYCN itself. The orally bioavailable OTX015 is cur-
    to clinical exploitation. Current established collaborative net-
    rently the only BET inhibitor in early-​phase clinical trials.
    works, such as Innovative Therapies for Children with Cancer
    (ITCC) in Europe, are testing an increasing number of targeted Targeting synthetic lethal interactions of MYCN. Genes that are
    agents in paediatric cell lines and animal models, providing synthetically lethal with MYCN expression include AURKA,
    strong evidence supporting its biology-​driven use, and testing CDK1, CDK2, and CHEK1. Several CDK inhibitors with excel-
    feasible and effective combinations with chemotherapy or other lent selectivity and potency are under development.
    targeted agents. The inclusion of robust predictive biomarkers in Targeting oncogenic stabilization of MYCN. MYC family pro-
    phase I/​II clinical trials will enhance the success in drug develop- teins bind to a proteasomal degradation complex that includes
    ment and reduce exposures and toxicity for populations unlikely AURKA and E3 ubiquitin ligases. An attractive strategy to target
    to benefit from a particular therapy. These approaches will also oncogenic stabilization of MYCN is to promote dissociation
    help determine the indication for inclusion of novel therapeutic of the AURKA:MYCN complex to trigger rapid proteasomal
    strategies in first-​line treatment, in particular of high-​risk neuro- degradation of MYCN. AURKA inhibitors, such as alisertinib
    blastoma patients. (MLN8237), induce a conformational change in the kinase that
    actively reinitiates MYCN degradation, independent of any re-
    Targeting ALK quirement for enzymatic inhibition of the kinase itself.
    Targeting mutated and activated ALK in neuroblastoma is likely to Targeting MYCN expression. A novel mechanism regulating
    produce clinical benefit since: MYCN expression involves the modulation of the let-​7 family
    of miRNAs by LIN28B, which negatively regulate MYCN
    1. activating point mutations or amplification occurs in 10% of pri- expression.
    mary neuroblastomas and up to 30% of relapse tumours;
    Targeting RTKs and PI3K. Aberrant PI3K/​ mTOR activity in
    2. normal ALK expression is confined to developing neural tissues; neuroblastoma correlates with poor outcome, drives oncogenic
    3. cells expressing mutated ALK exhibit oncogene addiction with stabilization of MYCN, and can be targeted using clinical PI3K/​
    ALK inhibition leading to cell death; mTOR inhibitors. Early-​phase trials of compounds active against
    4. ALK mutations can be inhibited by small molecules; and PI3K, mTOR, or AKT are underway.
    5. ALK can be activated by mechanisms other than mutation and
    amplification, suggesting that ALK inhibition might be useful Targeting histone deacetylases
    for a larger patient subset.
    Histone deacetylase (HDAC) inhibitors are an emerging class of
    Crizotinib (PF-​02341066) was the first ALK inhibitor evaluated promising novel anti-​cancer drugs. Neuroblastoma is the first tu-
    for neuroblastoma. While excellent clinical responses have been mour entity in which expression of all classical HDAC family mem-
    achieved in patients with non-​small-​cell lung cancer and adult bers have been systematically investigated. All 11 classical HDAC
    anaplastic large-​cell lymphoma (ALCL), only very limited responses family members are expressed, but HDAC8 is the only isozyme that
    were achieved in neuroblastoma patients. The ALK F1174L muta- significantly correlates with advanced disease stage, age, unfavour-
    tion confers primary as well as acquired crizotinib resistance, since able tumour histology, 11q aberration, and poor survival. Since
    this mutant protein does not directly interact with the crizotinib-​ HDAC8-​selective inhibitors are available, HDAC8 may be a prom-
    binding site within the ATP-​binding pocket. Several structurally ising target for neuroblastoma differentiation therapy.
    distinct, third-​generation ALK inhibitors with enhanced potency
    and specificity are currently in early-​phase clinical trials, including Targeting MDM2/​TP53
    ceritinib (LDK378) and lorlatinib (PF-​06463922), a dual-​specificity Another appealing approach for molecular therapy of neuroblastoma
    ALK/​ROS inhibitor. is provided by a crucial protein–​protein interaction that regulates
    Conclusion 251

    the cellular decision between life and death: the physical association Brodeur GM (2003) Neuroblastoma: biological insights into a clinical
    between the TP53 tumour-​suppressor protein and its major negative enigma. Nat Rev Cancer 3, 203–​16.
    regulator, MDM2. Escape from the TP53-​orchestrated failsafe pro- Cohn SL, et al. (2004) The International Neuroblastoma Risk Group
    gramme is a necessary event for tumourigenesis, and is achieved in (INRG) Classification System: an INRG Task Force. J Clin Oncol
    neuroblastoma cells by inappropriately increased activity of MDM2 27, 289–​97.
    De Grandis E, et al. (2009) Long-​term follow-​up neuroblastoma-​
    to keep TP53 in check. Consequently, targeted disruption of the
    associated opsoclonus-​myoclonus-​ataxia syndrome. Neuro­
    TP53-​MDM2 interaction may be a powerful means to unleash the
    pediatrics 40, 103–​11.
    potent anti-​tumour capacities of wild-​type TP53 in neuroblastoma
    DuBois SG, Matthay KK (2008) Radiolabeled metaiodobenzylguanidine
    cells without inflicting the collateral genotoxic damage associated for the treatment of neuroblastoma Nucl Med Biol 35(S1), S35–​48.
    with conventional chemotherapeutic drugs. DuBois SG, et al. (2016) Phase I study of the aurora A kinase inhibitor
    alisertib in combination with irinotecan and temozolomide for pa-
    tients with relapsed or refractory neuroblastoma: a NANT (New
    Conclusion Approaches to Neuroblastoma Therapy) trial. J Clin Oncol 34,
    1368–​75.
    Neuroblastoma has remained a challenging disease for both clin- Eleveld TF, et al. (2015) Relapsed neuroblastomas show frequent RAS-​
    icians and researchers for many decades. Advances in our under- MAPK pathway mutations. Nat Genet 47, 864–​71.
    standing of the genetics genomics, epigenetics, and biology of Fulda S (2009) The PI3K/​Akt/​mTOR pathway as therapeutic target in
    neuroblastoma have provided molecular markers that predict tu- neuroblastoma. Curr Cancer Drug Targets 9, 729–​37.
    mour behaviour and patient outcome with very high accuracy. Gripp KW (2005) Tumor predisposition in Costello syndrome. Am J
    Med Genet C Semin Med Genet 137C(1), 72–​7.
    Together with recent information on altered signalling pathways in
    Haas-​Kogan DA, et al. (2003) Impact of radiotherapy for high-​risk
    aggressively growing tumours, this knowledge will help to establish
    neuroblastoma: a Children’s Cancer Group study. Int J Radiat Oncol
    therapeutic strategies specifically targeting key molecular factors of
    Biol Phys 56, 28–​39.
    neuroblastoma tumour progression. Hart LS, et al. (2016) Preclinical therapeutic synergy of MEK1/​2
    Risk classification continues to be optimized, and future ap- and CDK4/​6 inhibition in neuroblastoma. Clin Cancer Res 23(7),
    proaches will need to integrate profiling of mRNA and microRNA 1785–​96.
    expression, epigenetic modifications, and whole-​ genome copy Henssen A, et al. (2016) Targeting MYCN-​driven transcription by
    number variations with the current INRG system. The remaining BET-​bromodomain inhibition. Clin Cancer Res 22, 2470–​81.
    challenges will be to bridge the gap between characterizing the mo- Hirschmann-​Jax C, et al. (2004) A distinct ‘side population’ of cells
    lecular properties of neuroblastoma and understanding the pre- with high drug efflux capacity in human tumor cells. Proc Natl Acad
    cursor cells giving rise to the disease, focusing on those features Sci USA 101(39), 14228–​33.
    that make the cells susceptible to malignant transformation and on Janoueix-​Lerosey I, et al. (2008) Somatic and germline activating mu-
    determining how to combine targeted agents with standard cytotoxic tations of the ALK kinase receptor in neuroblastoma. Nature 455,
    agents, and when to introduce them during therapy. Furthermore, 967–​70.
    Maris JM (2010) Recent advances in neuroblastoma. N Engl J Med 362,
    interactions between neuroblastoma and the host will need to be
    2202–​11.
    more clearly defined, in terms of the role of the immune response
    Matthay KK, et al. (2016) Neuroblastoma. Nat Rev Dis Primers
    and other host-​dependent factors. 2, 16078.
    It is imperative that new agents or therapeutic strategies be moved Matthay K, et al. (2009) Long-​term results for children with high-​
    into frontline therapy to enhance induction response or maximize risk neuroblastoma treated on a randomized trial of myeloablative
    treatment of minimal residual disease. The focus must now shift to- therapy followed by 13-​cis-​retinoic acid: a Children’s Oncology
    wards multidisciplinary translational research teams using validated Group study. J Clin Oncol 7, 1007–​13.
    preclinical models to select the most promising combinations of mo- Monclair T, et al. (2009) The International Neuroblastoma Risk
    lecularly targeted therapies, broad-​spectrum chemotherapy, and im- Group Staging System: an INRG task force report. J Clin Oncol
    munotherapy to cure patients who currently have a poor prognosis 27, 298–​303.
    and to reduce therapy for those with more favourable outcomes. Oberthuer A, et al. (2015) Revised risk estimation and treatment strati-
    fication of low-​and intermediate-​risk neuroblastoma patients by
    integrating clinical and molecular prognostic markers. Clin Cancer
    Res 21, 1904–​15.
    FURTHER READING Ohira M, Nakagawara A (2010) Global genomic and RNA profiles
    Acosta S, et al. (2009) Comprehensive characterization of neuro- for novel risk stratification of neuroblastoma. Cancer Sci 101,
    blastoma cell line subtypes reveals bilineage potential similar to 2295–​301.
    neural crest stem cells. BMC Dev Biol 9, 12. Parodi S, et al. (2007) The prognostic role of urinary catecholamines
    Barone G, et al (2013) New strategies in neuroblastoma: therapeutic in infants with disseminated neuroblastoma may be mediated by
    targeting of MYCN and ALK. Clin Cancer Res 19, 5814–​21. MYCN amplification. Ped Blood Cancer 48, 593.
    Bourdeaut F, et al. (2005) Germline mutations of the paired-​like Peifer M, et al. (2015) Telomerase activation by genomic rearrange-
    homeobox 2B (PHOX2B) gene in neuroblastoma. Cancer Lett ments in high-​risk neuroblastoma. Nature 526, 700–​4.
    228, 51–​8. Pinto NR, et al. (2015) Advances in risk classification and treatment
    Brockmann M, et al. (2016) Small molecule inhibitors of aurora-​A strategies for neuroblastoma. J Clin Oncol 33, 3008–​17.
    induce proteasomal degradation of N-​Myc in childhood neuro- Pugh TJ, et al. (2013) The genetic landscape of high-​risk neuro-
    blastoma. Cancer Cell 30, 357–​8. blastoma. Nat Genet 45, 279–​84.
    252 CHAPTER 28 Neuroblastoma

    Schleiermacher G, et al. (2014) Recent insights into the biology of Villani A, et al. (2017) Recommendations for cancer surveillance in in-
    neuroblastoma. Int J Cancer 135, 2249–​61. dividuals with RASopathies and other rare genetic conditions with
    Schramm A, et al. (2015) Mutational dynamics between primary and increased cancer risk. Clin Cancer Res 23(12), e83–​90.
    relapse neuroblastomas. Nat Genet 47, 872–​7. Weiss WA, et al. (1997) Targeted expression of MYCN causes neuro-
    Schramm A, et al. (2005) Biological effects of TrkA and TrkB receptor blastoma in transgenic mice. EMBO J 16, 2985–​95.
    signaling in neuroblastoma. Cancer Lett 228, 143–​53. Witt O, et al. (2009) Targeting histone deacetylases in neuroblastoma.
    Schulte JH, Eggert A (2015) Neuroblastoma. Crit Rev Oncog 20, Curr Pharm Des 15, 436–​47.
    245–​70. Yalçin B, et al. (2010) High-​dose chemotherapy and autologous haem-
    Schwab M, et al. (1983) Amplified DNA with limited homology to myc atopoietic stem cell rescue for children with high-​risk neuro-
    cellular oncogene is shared by human neuroblastoma cell lines and blastoma. Cochrane Database Syst Rev 5, cd 006301.
    a neuroblastoma tumour. Nature 305, 245–​8. Yu AL, et al. (2010) Anti-​GD2 antibody with GM-​CSF, interleukin-​
    Shimada H, et al. (1999) Terminology and morphologic criteria 2, and isotretinoin for neuroblastoma. N Engl J Med 363,
    of neuroblastic tumors: recommendations by the International 1324–​34.
    Neuroblastoma Pathology Committee. Cancer 86, 349–​63.
    29
    Germ-​Cell Tumours
    Gabriele Calaminus and James C. Nicholson

    Introduction mediastinum, retroperitoneum, and urogenital system (10%). In


    adolescence and young adulthood, testicular GCT are the most fre-
    Germ-​cell tumours (GCT) constitute a highly heterogeneous group quent cancer in males and females.
    of tumours characterized by a wide range of histologies, benign and Sacrococcygeal, vaginal, and head and neck GCT develop
    malignant, and clinical presentation at many different anatomical during the neonatal period and infancy/​ early childhood.
    sites. In adolescents and adults, over 90% of cases develop within Histologically, they present as teratomas and YST (Table 29.1).
    the gonads, mostly in males, whilst around half of childhood GCT Testicular GCT may also develop during infancy and child-
    occur at extra-​gonadal midline sites such as the sacrococcygeal re- hood; however, they show a bimodal age distribution with a
    gion, central nervous system (CNS), and mediastinum. The diverse second incidence peak during adolescence and adulthood. In
    nature of GCT necessitates cooperation between paediatric oncolo- this age group, GCT present as seminomas and malignant non-​
    gists and surgeons (according to tumour site) and radiation oncolo- seminomatous GCT, which may include embryonal carcinoma
    gists, as well as related disciplines of endocrinology and neurology. (EC), YST, and choriocarcinoma, often with mixed histology. In
    Coordination of this multidisciplinary care is critical in the success contrast, ovarian and CNS GCT predominantly develop after the
    of treatment. onset of puberty and during the second (and third) decade of life.
    During the past three decades, a dramatic improvement in prog- The remainder of this chapter will focus on the management of
    nosis of malignant GCT (MGCT) has been achieved. This progress extracranial GCT as management of intracranial (CNS) GCT is
    can mainly be attributed to national and international cooperative addressed elsewhere.
    therapeutic protocols using platinum-​based combination chemo-
    Conditions predisposing to germ-​cell tumours
    therapy as part of a multimodal therapeutic approach. Identification
    of risk groups relating to age, histology, primary site, and stage have In adolescents, there is an association between testicular GCT
    led to stratification of chemotherapy. and testicular dysgenesis, resulting from inadequately treated un-
    descended testes. GCT are also associated with constitutional dis-
    turbances of sex chromosomes such as Klinefelter syndrome. The
    Epidemiology and biology risk of developing GCT is increased in phenotypic females with
    testicular feminization and streak gonads (frequently associated
    GCT can occur in all age groups, ranging from the fetal period to with ambiguous sexual differentiation, Swyer syndrome) and in
    adulthood. The annual incidence is approximately 0.5 per 100,000 Turner’s syndrome; in these patients, Y-​chromosome fragments are
    in children under 15. However, the overall incidence of teratomas is often detected and dysgerminoma can occur in association with
    likely to be underestimated because of under-​reporting to registries. gonadoblastoma. Maternal exposure to oestrogen may influence the
    Registry and trial data show a distinct distribution pattern with appearance of testicular GCT in young men, an association not seen
    regard to site and tumour histology of GCT. The histological sub- for GCT in infants and childhood.
    types of GCT in neonates are almost exclusively mature and imma-
    Embryology and histological classification
    ture teratomas (0.9 per 100,000 in girls; 2.6 per 100,000 in boys),
    sometimes admixed with subtle foci of yolk-​sac tumours (YST). YST GCT are believed to arise from primordial germ cells (PGC), which
    predominate later in infancy and during early childhood. Whilst the are totipotent and capable of differentiating into all GCT sub-
    incidence of GCT drops steadily during childhood to below 0.1 per types. During normal development, PGC migrate from the extra-​
    100,000, a significant increase is observed after the onset of puberty embryonic yolk sac to the gonads; the occurrence of GCT at midline
    (Figure 29.1). extra-​gonadal sites is thought to result from aberrant migration.
    In children, the most frequent primary sites of GCT are the Despite their common origin, GCT are characterized by marked
    testes (25%), ovaries (25%), sacrococcygeal region (20%), and the heterogeneity, impacting on their behaviour and management.
    CNS (20%); they may also arise in the head–​neck region, anterior Their complex classification reflects their degree of differentiation,
    254 CHAPTER 29 Germ-Cell Tumours

    < 18 years1) > 18 years2) Table 29.2 Histological summary of paediatric germ-​cell tumours
    Brain 20 % 3% I Germinoma1 A. Intratubular
    Mediastinum 4% 10% ( male) B. Invasive

    Abdomen 4% 2% II Teratoma A. Mature


    B. Immature2,4 Grade 1. Immature tissue <10% of the
    Coccyx 18 % sample surface
    Ovary 27 % 10% Grade 2. Immature tissue 10–​50% of the
    sample surface
    Testis 25 % 85% Grade 3. Immature tissue 50% of the
    sample surface
    Other 2% ?
    C. Malignant teratoma (teratoma with non-​germ-​cell
    malignant component)
    III Embryonal carcinoma
    Figure 29.1 Distribution of germ-​cell tumours. IV Yolk sac tumour3
    Source: data from: (1) Kaatsch P, et al. (2015) Pediatric germ cell tumours from
    1997–2011. Incidence rates, time trends and survival. Pediatrics. 135(1), 136. V Choriocarcinoma
    doi: 10.1542/peds.2014-1989 and (2) Siegel R, et al. (2014) Cancer Statistics 2014. VI Gonadoblastoma
    CA Cancer J Clin. 64(1), 9. doi: 10.3322/caac.21208.
    VII Mixed malignant germ-​cell tumour (each component to be listed)
    and is based on the 2006 World Health Organization (WHO) classi- 1. Synonyms: dysgerminoma (ovary), seminoma (testis), germinoma (extragonadal).
    fication of testicular cancer (Table 29.1). 2. Degree of immaturity is sometimes expressed according to Gonzalez–​Crussi in terms
    Germinomatous tumours (also known as seminomas (testis), of % of sampled surface (grade 1, <10%; grade 2, 10–​50%; grade 3, >50%).
    3. Yolk sac tumour is also known as endodermal sinus tumour.
    dysgerminomas (ovary), and germinomas (extragonadal)) arise 4. Common to each of these: +/​–​microfoci of probable yolk sac tumour.
    from cells that remain in the pluripotent undifferentiated state with
    morphological features of undifferentiated germ-​cell epithelium.
    YST and choriocarcinoma (CHC) follow an extra-​embryonic dif-
    ferentiation pattern and are characterized by secretion of alpha-​ only in the presence of the Y chromosome. Left untreated it may be-
    1-​fetoprotein (AFP) or human chorionic gonadotropin (HCG or come malignant, typically as a germinoma. In contrast to testicular
    ß-​HCG), respectively. EC consists of immature totipotent cells re- GCT of adult patients, there is no evidence to suggest that paediatric
    sulting from embryonic differentiation. Somatic differentiation GCT develop from carcinoma in situ.
    results in the formation of teratomas, mature or immature, and In most patients, response to different treatment modalities can
    mimicking structures of all germ layers. The histological grade of be predicted from histology and tumour-​marker profile. Paediatric
    immaturity is defined by the extent of immature (predominantly GCT often present with more than one histological subtype; in this
    neuroepithelial) elements (Table 29.2). This correlates with the situation selection of therapy is based on the strategy employed for
    risk of loco-​regional relapse, particularly after incomplete resec- the most aggressive malignant component.
    tion, and is therefore essential in risk assessment of pure teratomas.
    Gonadoblastoma is a benign but premalignant condition, with Tumour markers
    stromal components associated with dysgenetic gonads, occurring The secretion of AFP and HCG by some subtypes of GCT is well rec-
    ognized and can be exploited in diagnosis and monitoring, on and
    Table 29.1 Revised World Health Organization (WHO) off treatment. AFP, an α-​globulin, has very high levels at birth as it is
    classification for germ-​cell tumours the main fetal serum-​binding protein, but falls to adult levels of less
    Group Histology Epigenetics Genetics
    than 12 ng/​ml (less than 10 kU/​l; 1 ng/​ml ≡ 0.84 kU/​l) between one
    and two years of age. Therefore, isolated AFP levels should be inter-
    Prepubertal Germ-​cell Teratoma Premeiotic Normal
    tumours of Loss of preted cautiously in children under two. AFP is produced by YST
    Yolk sac –​1p,+1q, and, at low levels, by some EC and immature teratomas.
    infancy and imprinting
    tumour –​6q,+20
    childhood HCG is a glycoprotein produced by the placenta and made up of
    Postpubertal Germ-​cell Teratoma Meiosis I +12p α and β subunits. The α subunit is structurally similar to other hor-
    tumours of Loss of mones including luteinizing hormone (LH) and follicle-​stimulating
    Seminoma
    adolescence imprinting
    and adulthood Malignant hormone (FSH). So the more specific β subunit is used in assays
    non-​ to avoid cross reactivity with other hormones. βHCG secretion by
    seminoma tumour is seen typically in the presence of choriocarcinoma, with
    Spermatocytic seminoma (testis) Meiosis II +9 lower levels produced by some germinomas, indicating the presence
    Gametal of syncytiotrophoblastic giant cells. Since relative levels of ‘free β’ and
    imprinting
    intact dimer (α + β) vary between different tumour subtypes, meas-
    Cystic teratoma (ovary) Meiosis II (23,X)x2 urement of ‘total HCG’ (both free β and dimer) is recommended.
    Gametal
    imprinting Lactate dehydrogenase (LDH) and placental alkaline phos-
    phatase (PLAP) are sometimes measured in GCT. LDH is a non-​
    Reproduced with permission from Stark DP, et al. ‘Tumors in Adolescents and Young
    Adults’. Prog Tumor Res, Volume 43, pp. 115–​127. Copyright © 2016 Basel, Karger.
    specific marker of malignancy of limited value in the assessment
    DOI: https://​doi.org/​10.1159/​000447081 of paediatric tumours, and PLAP is raised in only around 20% of
    Diagnosis 255

    germinomas and may therefore be used as an additional diagnostic suggestion that they differ biologically, segregating by both histology
    tool if elevated. and age. In contrast, and in keeping with their presumed primordial
    germ-​cell origin, it has also been shown that MGCT all overexpress
    Molecular biology two microRNA clusters (miR 371-​373 and miR 302-​367), regardless
    The rarity of GCT and their subtypes has hampered progress in of patient age, histological subtype, and anatomical site of the tu-
    understanding their biology, although patterns are now emerging mour. These are not only detectable in tumours but also in the serum
    in terms of genomic changes and gene expression. The group and in cerebrospinal fluid (CSF) in intracranial sites. These circu-
    of Ichimura et al. described similar mutations of the MTOR lating microRNAs have been found to be both sensitive and specific
    pathway in intracranial and testicular GCT, which underlines the for the diagnosis of MGCT, with potential use in diagnosis, disease
    common origin of the disease and the possibility to develop diag- monitoring, and prognostication (Figure 29.2).
    nostic and therapeutic targets for the disease regardless of the
    location.
    Gain of the short arm of chromosome 12 (12p) is a well-​recognized Diagnosis
    abnormality found almost universally in adult testicular GCT, usu-
    ally arising by isochromosome formation. It has also been observed Clinical presentation
    in ovarian and mediastinal GCT but is relatively rare in children. The GCT tend to occur as relatively indolent masses, with clinical symp-
    incidence of 12p gain increases through childhood, and the pattern toms determined by their site of origin. Gonadal GCT comprise
    of imbalances seen on chromosomes 1, 6, 13, and 20 differs between around half of childhood GCT. Ovarian tumours typically occur
    pre-​and postpubertal patients with mediastinal GCT. No con- in girls over the age of five years; they are usually teratomas or
    sistent correlation between cytogenetic imbalance and primary site dysgerminomas, more complex mixed tumours being seen during
    has been observed, and teratomas are associated with complete ab- adolescence. They present as palpable abdominal masses, often with
    sence of chromosome imbalance. Pediatric GCT commonly arise at visible distension, and there may be pain associated with torsion or
    extragonadal sites. It has been proposed that non-​gonadal GCT arise rupture. Symptoms associated with hormonal disturbances such as
    from ectopic primordial germ cells that have aberrantly migrated virilization and menstrual disturbance may be seen in postpubertal
    during embryogenesis. During a time between their migration and patients with ovarian primaries. Testicular GCT present as pain-
    development to mature gametes, primordial germ cells are char- less scrotal swellings, either in the first years of life when YST and
    acterized by their lack of imprinting, which can be assessed by the teratomas predominate, or in adolescence, when seminomas and
    evaluation of allelic gene expression and DNA methylation in differ- mixed MGCT are seen.
    entially methylated control regions. Gonadal and non-​gonadal GCT The most common sites for extragonadal GCT are the CNS (dis-
    are derived from primordial germ cells that have consistently lost the cussed in Chapter 24, ‘Other Central Nervous System Tumours of
    imprinting of SNRPN and partly lost imprinting of H19 and IGF-​2. Childhood’), sacrococcygeal region, mediastinum, retroperitoneum,
    More recent work has also suggested an association between global-​ and head and neck, typically in the midline. Most sacrococcygeal
    methylation patterns and resistance to treatment.In addition, paedi- tumours are teratomas (mature or immature) and occur more com-
    atric and adult MGCT differ significantly in their protein-​coding monly in girls than boys. They may be diagnosed antenatally, by rou-
    gene-​expression profiles (messenger RNA, mRNA), supporting the tine scanning, or as a visible mass on neonatal examination. Cases

    Tumour Markers
    Precursor Cell Differentation Histology AFP ß-HCG

    Normal Normal testis and ovary - -


    intermediate

    Somatic Teratoma ± -
    Benigne/

    - Mature
    - Immature
    Embryonic
    Embryonal ± ±
    Totipotent carcinoma (EC)
    Primordial
    Germ Cell
    Yolk sac tumour (YST) ++ -
    Malignant

    Extra-Embryonic
    - ++
    Choriocarcinoma (CHC)

    Seminoma (testis)
    None: Dysgerminoma (ovary) - ±
    Germinal state
    Germinoma (other)

    Figure 29.2 Biological characteristics of germ-​cell tumours.


    256 CHAPTER 29 Germ-Cell Tumours

    that present later, with partly or exclusively internal tumours, are tumour tissue for genetic and biological studies. Where a biopsy is
    more likely to have malignant components, typically YST, and to performed in a mixed tumour, the GCT components in the tissue
    present with functional disturbances resulting from the effects of tu- diagnosis may differ from the subtype suggested by marker estima-
    mour bulk or nerve damage. Mediastinal GCT are more common in tion, or may under-​report the range of histologies present as a result
    boys than girls and the majority occur from adolescence onwards, of sampling error. This is unlikely to influence treatment, provided
    almost exclusively in the anterior mediastinum. Symptoms may re- that the tumour is fully staged.
    sult from airway compression, superior vena cava obstruction, and, Thresholds above which marker levels can be considered diag-
    occasionally, aberrant hormone production. nostic are controversial. For AFP, moderately elevated levels may be
    Approximately 20% of MGCT have metastases at the time of diag- consistent with the presence of microfoci of YST or immature hep-
    nosis, although they are rarely associated with presenting symp- atic/intestinal elements in immature teratomas. HCG may be mod-
    toms, and mostly identified as a result of staging investigations. GCT erately elevated in germinomas, and levels up to 50 IU/​l are usually
    most commonly spread to the lungs, but occasionally to liver, bone, considered consistent with this diagnosis, although caution should
    or bone marrow. Metastasis to the CNS is associated with CHC, and be exercised in labelling levels any higher than this as CHC, given
    may be seen in approximately 4% of young males with malignant that CHC are rare in young people and typically associated with
    testicular GCT, but in less than 1% of cases in younger children. This HCG levels over 1,000 IU/​l.
    possibility should therefore be borne in mind in patients with very In female patients with ovarian GCT and in male patients with
    high levels of HCG. mediastinal GCT, constitutional karyotyping is also recommended.
    In addition to routine blood tests before chemotherapy, special
    Investigations consideration should be given to assessment of renal function as
    Radiology and staging most regimens used for treatment of GCT incorporate nephrotoxic
    In many patients with extracranial GCT, initial assessment will be agents including platinum compounds and ifosfamide.
    made with ultrasound of the abdomen and pelvis, and should in-
    clude assessment of tumour position and size, as well as the presence
    of lymphatic spread or peritoneal seeding. This is followed by de- Treatment
    tailed three-​dimensional imaging with computed tomography (CT)
    or magnetic resonance imaging (MRI) scan. CT scan is required for Surgery
    assessment of lung metastases and extent of abdominal disease. MRI Most extracranial GCT are amenable to primary resection, and this
    is the investigation of choice for sacrococcygeal GCT, in order to is the treatment of first choice for localized tumours. If the diag-
    define extension into the sacral and spinal canal. Plain X-​rays have nosis of an MGCT has been made histologically or with raised
    limited value in initial assessment, except for mediastinal tumours, markers, and the initial radiological assessment shows infiltration
    but are valuable in follow-​up, particularly chest X-​ray. Skeletal me- or metastatic spread, preoperative chemotherapy will increase the
    tastases and bone-​marrow involvement are seen in less than 10% of likelihood of microscopically complete resection. In general, there
    patients overall, and are rare in the absence of pulmonary metas- is no role for debulking surgery in pediatric GCT. In patients with
    tases, but are particularly associated with malignant sacrococcygeal tumour residual after initial resection, second-​look surgery is essen-
    GCT. Technetium bone scan should therefore be considered in the tial. Surgical removal of metastases is not indicated unless they show
    presence of organ metastases. CNS imaging only needs to be con- insufficient response to chemotherapy. Close attention to surgical
    sidered in patients (mainly adolescent males) in which CHC is iden- recommendations for GCT is essential to optimize outcomes whilst
    tified histologically or by markers. The use of positron emission minimizing morbidity.
    tomography (PET) scans is controversial but might add important
    information for patients with bulky disease after chemotherapy and Testis
    with negative markers under treatment, to evaluate response other The resection of testicular tumours is performed as unilateral orchid-
    than by reduction of tumour volume. ectomy after high inguinal incision, to avoid scrotal contamination;
    transscrotal surgery will lead to upstaging of tumours which might
    Laboratory investigations otherwise be stage 1, with the result that patients may receive chemo-
    AFP and total HCG should be measured in all cases of suspected therapy that could have been avoided. There is no evidence that retro-
    GCT, as raised tumour markers may be of value in diagnosis and peritoneal lymph-​node dissection improves prognosis in children, and
    monitoring of GCT, both on and off treatment: caution should be it leads to increased morbidity including risk to future sexual function.
    exercised in interpretation of AFP levels in infants. GCT may be
    treated without tissue diagnosis on the basis of radiological findings Ovary
    and raised tumour markers, although histological confirmation is Primary surgical resection (salpingo-​oopherectomy) is standard
    usually available at diagnosis in gonadal tumours, for which primary management in ovarian GCT. Ascitic fluid or peritoneal washings
    excision is part of treatment. Neoadjuvant chemotherapy with de- should be collected for cytological examination, and biopsies of any
    layed surgery may be the treatment of choice for some extragonadal suspicious area on peritoneal surfaces, lymph nodes, omentum, or
    GCT, in which case raised markers may obviate the need for poten- contralateral ovary performed. Unresectable or bilateral tumours
    tially hazardous biopsy. However, in equivocal cases (non-​diagnostic should be biopsied with a view to fertility-​preserving surgery where
    markers, hepatic or upper retroperitoneal tumours), a diagnostic possible, following neoadjuvant chemotherapy, except in the pres-
    biopsy should be performed, and should include storage of fresh ence of gonadoblastoma or XY gonadal dysgenesis, when bilateral
    Treatment 257

    gonadectomy is recommended. Whilst oophorectomy remains particularly in older patients and the very young, and in the presence
    standard practice for mature teratomas of the ovary, there is a trend of renal impairment. These concerns have led to the development
    towards more conservative fertility preservation surgery. This kind of regimens omitting bleomycin or using it at a reduced dose. The
    of surgery should be done in experienced centres to avoid tumour-​ cisplatin, etoposide, and ifosfamide (PEI) combination is associated
    cell spillage and to deliver the balance between maximal resection of with a higher degree of myelosuppression and also carries a risk of
    the tumour and preservation of ovarian tissue. tubular nephropathy, particularly in small children. Etoposide, used
    in the majority of GCT strategies in both adults and children, carries
    Coccyx the risk of secondary leukaemia, estimated to be in the region of 1%
    A dorsal approach to sacrococcygeal tumours is appropriate for most over ten years of follow-​up.
    cases in view of their predominantly external components. Some
    have substantial internal components deep in the pelvis for which Development of cooperative protocols
    an additional anterior approach via laparotomy is required. The Prospective trials for gonadal and non-​gonadal GCT in children
    coccyx should be resected en bloc to avoid tumour rupture, and in- and adolescents were initiated by national groups in the 1980s.
    filtrated skin areas should also be removed en bloc with the tumour. Early experience from North American (Children’s Cancer Group;
    Separation of tumour and rectum is mostly possible with digital CCG) studies showed that patients with ovarian GCT had a better
    control, and bowel and bladder function recover postoperatively in prognosis than children with non-​gonadal GCT due to the higher
    the majority of patients. rate of incomplete tumour resections in non-​gonadal tumours.
    A subsequent American Intergroup protocol adopted the ‘watch
    Other sites and wait’ policy for stage I testicular GCT and achieved good re-
    MGCT at other extragonadal sites such as the mediastinum, sults for intermediate-​risk patients (testicular stage II, ovarian and
    retroperitoneum, and genitourinary tract are less likely to be amen- non-​gonadal stages I and II) with four cycles of cisplatin, etoposide,
    able to complete resection, and the surgical approach should be and bleomycin. In high-​risk patients (stages III and IV), the thera-
    individualized. peutic impact of cisplatin-​dose intensification (200 mg/​m2/​cycle)
    was evaluated, indicating that higher doses of cisplatin may result in
    Evolution of chemotherapeutic strategies—​balancing higher response and complete remission rates, but without improve-
    efficacy and toxicity ment in overall survival (OS)OS. Furthermore, high-​dose cisplatin
    Until 1980, the prognosis of children with MGCT was poor. The was also associated with higher rates of renal and auditory toxicity.
    modern era of GCT chemotherapy began in the mid-​1970s with the A subsequent report from this study group failed to demonstrate a
    identification of the efficacy of cisplatin in testicular GCT, given in protective benefit of amifostine against hearing loss during cisplatin
    combination with vinblastine and bleomycin (PVB) following tu- therapy at escalated doses.
    mour resection. With this regimen, the overall good response rate UK experience with the UKCCSG GC I and GC II protocols also
    was also translated into durable remissions. Nevertheless, the chal- demonstrated the high therapeutic efficacy of platinum-​based re-
    lenge of relapsed or refractory tumours established the need for gimens and promoted the use of carboplatin-​based chemotherapy
    second-​line therapies. Etoposide soon emerged as an active drug (JEB (carboplatin–​etoposide–​bleomycin) utilizing carboplatin at
    with a single-​agent efficacy superior to vinblastine. The resulting 600 mg/​m2/​cycle) in preference to cisplatin (BEP). Results of 184 pa-
    combination of cisplatin, etoposide, and bleomycin (BEP) is now tients treated between 1989 and 1997 confirmed the high efficacy of
    standard chemotherapy for GCT in adult patients. The efficacy of the JEB regimen, with five-​year event-​free survival (EFS) of 88% and
    ifosfamide in cisplatin-​refractory GCT has since been demonstrated, OS of 93%, comparable to outcomes with cisplatin-​based strategies
    and combinations of cisplatin with ifosfamide and either etoposide from other national groups and with a favourable toxicity profile.
    or vinblastine for recurrent testicular GCT are considered standard The French cooperative protocol TGM 85 used a similar
    relapse treatment. These observations led to studies incorporating chemotherapeutic approach, and in the consecutive TGM 90
    ifosfamide into the first-​line treatment of GCT, and although the protocol, cisplatin was replaced by carboplatin (400 mg/​m2/​cycle).
    inclusion of ifosfamide with cisplatin and etoposide (PEI) did not The results were less favourable than with the UKCCSG JEB regimen,
    result in significantly improved outcome, this demonstrated that the difference attributed mainly to the lower dose of carboplatin. In
    MGCT could be treated successfully with combination chemo- both the French TGM 90 and the UK GC II studies, the analysis
    therapy from which bleomycin is excluded. of prognostic factors revealed the prognostic impact of high serum
    Selection of suitable chemotherapeutic regimes is based on a AFP levels at diagnosis, a finding that was not reproduced by other
    balance between efficacy and adverse effects. Cisplatin has been re- studies that used a cisplatin-​based regimen.The German protocols
    tained in the majority of first-​line treatment approaches for GCT but for testicular (MAHO) and non-​testicular (MAKEI) GCT included
    it is associated with a risk of toxicity to kidneys and hearing. For this cisplatin and etoposide-​based chemotherapy regimens, with recom-
    reason, strategies employing carboplatin instead of cisplatin have mendations for delayed tumour resection after neoadjuvant chemo-
    been explored. However, although the auditory and renal toxicity of therapy. Even from the beginning of the 1980s, EFS rates above 80%
    carboplatin-​based regimens are less than for cisplatin, carboplatin is were achieved with the first MAKEI and MAHO protocols, resulting
    more myelotoxic at therapeutically equivalent doses. Whilst this has in a step-​wise reduction in duration of chemotherapy from a max-
    not been found to be a dose-​limiting effect in children, adult patients imum of eight cycles to six and, currently, four or five. As with strat-
    may tolerate this less well and there has been some reluctance to egies used by other national groups, an expectant ‘watch and wait’
    pursue a carboplatin-​based approach in adult protocols. Pulmonary strategy was introduced in the MAKEI 96 protocol for patients with
    toxicity is the main concern associated with the use of bleomycin, completely resected low-​stage tumours. This resulted in sparing
    258 CHAPTER 29 Germ-Cell Tumours

    of chemotherapy for almost 75% of patients with stage Ia MGCT. Gonadal tumours are considered lower risk than extragonadal, but
    A neoadjuvant approach to locally invasive and/​or metastatic tu- this is largely because they are more likely to be completely resected.
    mours was shown to facilitate complete tumour resection. Increasing age is generally associated with increased risk, and often
    associated in older teenagers with higher-​risk adult-​type histology.
    Current therapeutic approaches The impact of metastases on risk and the need for treatment inten-
    In general, current paediatric GCT strategies stratify therapy ac- sity varies according to the type of tumour, and the risk associated
    cording to initial diagnostic parameters, and only in cases of insuffi- with high levels of AFP is of uncertain significance. Complete con-
    cient response to treatment is therapy further intensified. Although trol of disease at primary and metastatic sites is essential in all cases,
    there is some variation between protocols in the application of risk and residual disease following treatment confers a relatively adverse
    factors, there are common themes, summarized in Table 29.3. prognosis. A collaboration between the UK’s Children’s Cancer

    Table 29.3 Standard chemotherapy regimens for paediatric extracranial germ-​


    cell tumours

    PEI MAKEI 96, MAHO 98


    Cisplatin1 20 mg/​m2 Over 1 h Day 1,2,3,4,5
    Etoposide 100 mg/​m2 Over 3 h Day 1,2,3
    Ifosfamide2 1500 mg/​m2 Over 3 or 20 h Day 1,2,3,4,5
    Two to four cycles
    PVB MAHO 98
    Cisplatin1 20 mg/​m2 Over 1h Day 4,5,6,7,8
    Vinblastin 3 mg/​m2 or 0.15 mg/​kg IV bolus Day 1,2
    Bleomycin3 15 mg/​m2 Over 24 h Day 1,2,3
    Three cycles
    BEP MAHO 98
    Bleomycin3 15 mg/​m2 Over 24 h Day 1,2,3
    Etoposide 80 mg/​m2 Over 3 h Day 1,2,3
    Cisplatin1 20 mg/​m2 Over 1 h Day 4,5,6,7,8
    Three cycles
    BEP USA: Children’s Oncology Group
    Bleomycin 15 mg/​m2 Over 24 h Day 1
    Etoposide 100 mg/​m2 Over 3 h Day 1,2,3,4,5
    Cisplatin1 20 mg/​m2 Over 1 h Day 1,2,3,4,5
    Four cycles
    High-​dose BEP USA: Children’s Oncology Group
    Bleomycin 15 mg/​m2 Over 24 h Day 1
    Etoposide 100 mg/​m2 Over 3 h Day 1,2,3,4,5
    Cisplatin1 40 mg/​m2 Over 1 h Day 1,2,3,4,5
    Four cycles
    JEB UKCCSG/​CCLG GC3
    Carboplatin 600 mg/​m2 Over 1 h Day 2
    Etoposide 120 mg/​m2 Over 1 h Day 1,2,3
    Bleomycin3 15 mg/​m2 Over 15 min Day 3
    Four or six cycles
    PE Brazilian GCT group
    Cisplatin 20/​30 mg/​m2 Over 1 h Day 1,2,3,
    Etoposide 100/​120 mg/​m2 Over 1 h Day 1,2,3,4,5
    Five cycles

    1. Given with mannitol-​forced diuresis.


    2. Given with mesna uroprotection.
    3. Omitted in children <1 year, and at reduced dose (7.5 mg/​m2) in children <2 years.
    Future perspectives 259

    and Leukaemia Group (CCLG) and the USA’s Children’s Oncology chemotherapy, further surgery is indicated, even in a secreting tu-
    Group (COG), with pooling of trial data, helped to refine risk factors mour with normalized markers, since mature teratomas may re-
    further, with age and stage being the most important predictors; the main after successful treatment of the malignant component of a
    greater risk often attributed to extragonadal tumours is more likely mixed tumour, imposing a risk of tumour progression. PET scans
    to be a consequence of age and stage than site in itself, as shown have not been proven useful in this situation, as they cannot readily
    also in publications by other groups.The approach to treatment of distinguish between mature teratomas and residual necrosis or
    MGCT can be summarized as follows: fibrosis.
    Monitoring for late effects of treatment needs to be tailored to the
    • Surgical resection at diagnosis, where possible, with inguinal ap-
    treatment received. Audiometry should be performed during and at
    proach for testicular tumours, and sampling of ascites and nodules
    the end of treatment in patients who have received cisplatin therapy.
    for ovarian tumours.
    Patients who have received bleomycin require lung-​function testing.
    • Biopsy or marker diagnosis in unresectable tumours.
    Although routine monitoring of blood counts is not indicated, the
    • ‘Watch and wait’ for completely resected stage 1 gonadal primaries risk of treatment-​related secondary leukaemia should be considered
    avoids chemotherapy in over 70% of cases, and effective salvage is as part of routine follow-​up, in patients who have received etoposide.
    available for the remainder.
    • Platinum-​based chemotherapy for all stage II–​IV patients; ranging Relapse treatment
    from two cycles of a two-​drug regimen for completely resected There is no international consensus on strategies for treatment of re-
    stage II (MAKEI) to four to six cycles of a three-​drug regimen for current GCT. In patients previously treated with a carboplatin-​based
    highest risk (stage IV extragonadal). or non-​platinum-​based therapy, cisplatin-​based regimens such as
    • Surgery following chemotherapy for residual tumour. PEI or VIP (vinblastine, ifosfamide, cisplatin) have been used suc-
    • There is no place for radiotherapy in the first-​line treatment of cessfully. Failure following primary treatment with cisplatin may be
    extracranial GCT. treated with carboplatin and etoposide. Newer relapse strategies are
    increasingly exploiting the efficacy of gemcitabine, oxaliplatin, and
    Approach to treatment of teratomas taxol in combination.
    Mature teratomas are considered benign and are cured by sur- In Germany, a standardized strategy is used for patients with local
    gical excision which should be complete in order to avoid recur- recurrent extracranial MGCT in the abdomen. This is a combination
    rence. Sacroccoccygeal teratomas may recur as malignant tumours of local deep hyperthermia with platinorientated chemotherapy
    in around 10% of cases, underlining the need to aim for complete (HyperPEI). Results are promising and reveal a salvage rate of about
    excision and also to monitor AFP closely for recurrence. Surgical 70%. Aggressive local therapy is central to the strategy because most
    excision is also the mainstay of treatment for immature teratomas relapse occurs at the primary site. Local radiotherapy, at doses above
    and although they may show clinical features of malignancy, a role 45 Gy, has shown some benefit after incomplete resection of the re-
    for chemotherapy has not been established. The risk of recurrence current tumour. Radiotherapy may also be useful in the treatment of
    can be estimated from the primary site of the tumour, histological brain metastases.
    grade of immaturity, and completeness of the tumour resection. The therapeutic value of high-​dose chemotherapy with autologous
    Pure immature teratomas of the ovary may be associated with sec- stem-​cell transplantation has been investigated in a small number of
    ondary peritoneal deposits for which surgery remains the treatment patients with relapsed or refractory disease, but further clinical trial
    of choice. The significance of small foci of YST and/​or mildly raised evidence is needed to establish the role of high-​dose therapy more
    AFP in immature teratomas is uncertain and management subject clearly. Long-​term survival is likely to be achieved only in those pa-
    to controversy. If small YST foci are present and an incomplete re- tients in whom clinical complete remission can be achieved by other
    section has been obtained, a close ‘watch and wait’ procedure with treatment prior to high-​dose chemotherapy.
    tumour-​marker follow-​up (AFP) is mandatory.

    Follow-​up Future perspectives


    Most relapses occur within the first two years after diagnosis, al-
    though much later recurrences have been observed. Most groups In view of the high cure rates achieved by current protocols for
    would therefore recommend close monitoring of markers for se- standard-​risk GCT, research should now focus on the better iden-
    creting GCT and imaging, at least for marker-​negative GCT, for tification of patients more likely to fail first-​line treatment, the in-
    three to five years from the end of treatment. AFP and ß-​HCG are tensification of treatment in high-​risk patients, and the reduction of
    usually monitored weekly until they reach the normal range, then at therapy with its associated toxicities in low-​risk cases.
    reduced frequency (every one to three months) for the remaining As insufficient local tumour control at the primary site represents
    period of follow-​up. In ‘watch and wait’ patients, the decline of the main problem in most relapsing patients, further advances in
    AFP level should be evaluated with respect to its serum half-​life outcomes for relapsed GCT may depend on improvements in local
    of approximately six to seven days. The interpretation of AFP may therapy. Novel approaches need to be explored for high-​risk GCT.
    be difficult in infants but a slowing of the rate of decline or a sec- Locoregional hyperthermia combined with platinum-​based chemo-
    ondary rise in AFP level strongly suggests the presence of residual therapy offers a therapeutic concept in locally recurrent extracranial
    or recurrent YST. tumours. The augmentation of standard treatment with the develop-
    Imaging of the primary site of tumour should be undertaken ment of targeted therapies to oncogenes, growth factors, and recep-
    at the end of treatment. In the case of residual abnormalities after tors should also be a goal for the future.
    260 CHAPTER 29 Germ-Cell Tumours

    Timely progress in these areas for such a rare tumour group re- Kaatsch P, et al. (2015) Pediatric germ cell tumours from 1997–2011.
    quires multinational cooperation in both biological studies and clin- Incidence rates, time trends and survival. Pediatrics 135(1), 136.
    ical trials. A strong focus is also needed on the rehabilitation and Lopes LF, et al. (2009) Cisplatin and etoposide in childhood germ cell
    quality of life of the survivors of GCT, and such outcomes should be tumor: Brazilian Pediatric Oncology Society protocol GCT-​91. J
    taken into consideration when planning treatment. Clin Oncol 27(8), 1297–​303.
    Mann JR, et al. (1998) UKCCSG’s germ cell tumour (GCT)
    studies: improving outcome for children with malignant extra-
    cranial non-​gonadal tumours—​carboplatin, etoposide, and bleo-
    FURTHER READING mycin are effective and less toxic than previous regimens. United
    Baranzelli MC, et al. (1999) Prognostic factors in children with local- Kingdom Children’s Cancer Study Group. Med Pediatr Oncol
    ized malignant nonseminomatous germ cell tumors. J Clin Oncol 30(4), 217–​27.
    17(4), 1212. Mann JR, et al. (2000) The United Kingdom Children’s Cancer Study
    Billmire D, et al. (2004) Outcome and staging evaluation in malig- Group’s second germ cell tumor study: carboplatin, etoposide,
    nant germ cell tumors of the ovary in children and adolescents: an and bleomycin are effective treatment for children with malignant
    intergroup study. J Pediatr Surg 39(3), 424–​9. extracranial germ cell tumors, with acceptable toxicity. J Clin Oncol
    Blohm ME, et al. (1998) Alpha 1-​fetoprotein (AFP) reference values 18(22), 3809–​18.
    in infants up to 2 years of age. Pediatr Hematol Oncol 15(2), Marina N, et al. (2005) Amifostine does not protect against the oto-
    135–​42. toxicity of high-​dose cisplatin combined with etoposide and bleo-
    Bokemeyer C (2002) Extragonadal germ cell tumors of the medias- mycin in pediatric germ-​cell tumors: a Children’s Oncology Group
    tinum and retroperitoneum. Results from an international analysis. study. Cancer 104(4), 841–​7.
    J Clin Oncol 20(7), 1864–​73. Murray MJ, et al. (2016) A pipeline to quantify serum and cerebro-
    Calaminus G, et al. (2003) Prognostic value of tumor size, metastases, spinal fluid microRNAs for diagnosis and detection of relapse
    extension into bone, and increased tumor marker in children with in paediatric malignant germ-​cell tumours. Br J Cancer 114(2),
    malignant sacrococcygeal germ cell tumors. A prospective evalu- 151–​62.
    ation of 71 patients treated in the German cooperative protocols Murray MJ, et al. (2016) The present and future of serum diagnostic
    Maligne Keimzelltumoren (MAKEI) 83/​86 and MAKEI 89. J Clin tests for testicular germ cell tumours. Nat Rev Urol 13(12), 715–​25.
    Oncol 21(5), 781–​6. Palmer RD, et al. (2007) Malignant germ cell tumours of child-
    Cushing B, et al. (2004) Randomized comparison of combination hood: new associations of genomic imbalance. Br J Cancer 96(4),
    chemotherapy with etoposide, bleomycin, and either high-​dose 667–​76.
    or standard-​dose cisplatin in children and adolescents with high-​ Palmer RD, et al. (2010) Malignant germ cell tumors display common
    risk malignant germ cell tumors. A pediatric intergroup study—​ microRNA profiles resulting in global changes in expression of
    Pediatric Oncology Group 9049 and Children’s Cancer Group 8882. messenger RNA targets. Cancer research 70(7), 2911–​23.
    J Clin Oncol 22(13), 2691–​700. Pashankar F, et al. (2016) Is adjuvant chemotherapy indicated in
    De Giorgi U, et al. (2005) Salvage high-​dose chemotherapy for children ovarian immature teratomas? A combined data analysis from the
    with extragonadal germ-​cell tumours. Br J Cancer 93(4), 412–​17. Malignant Germ Cell Tumor International Collaborative. Cancer
    Eble JN, et al. (eds) (2004) World Health Organization Classification of 122(2), 230–​7.
    Tumours. Pathology and Genetics of the Urinary System and Male Perlman EJ, et al. (2000) Genetic analysis of childhood endodermal
    Genital Organs. Lyon: IARC. sinus tumors by comparative genomic hybridization. J Pediatr
    Einhorn LH, Donohue JP (1977) Improved chemotherapy in dissem- Hematol Oncol 22(2), 100.
    inated testicular cancer. J Urol 117(1), 65–​9. Poynter JN, et al. (2010) Trends in incidence and survival of pediatric
    Frazier AL, et al. (2015) Revised risk classification for pediatric extra- and adolescent patients with germ cell tumors in the United States,
    cranial germ cell tumors based on 25 years of clinical trial data 1975 to 2006. Cancer 116, 4882–​91.
    from the United Kingdom and United States. J Clin Oncol 33(2), Schmoll H-​J, et al. (2003) Long-​term results of first-​line sequential
    195–​201. high-​dose etoposide, ifosfamide, and cisplatin chemotherapy plus
    Fukushima S, et al. (2014) Mutually exclusive mutations of KIT and autologous stem cell support for patients with advanced meta-
    RAS are associated with KIT mRNA expression and chromo- static germ cell cancer: an extended phase I/​II study of the German
    somal instability in primary intracranial pure germinomas. Acta Testicular Cancer Study Group. J Clin Oncol 21(22), 4083–​91.
    Neuropathol 127(6), 911–​25. Schneider DT, et al. (1999) Acute myelogenous leukemia after treat-
    Göbel U, et al. (2001) Multimodal treatment of malignant ment for malignant germ cell tumors in children. J Clin Oncol
    sacrococcygeal germ cell tumors. A prospective analysis of 66 pa- 17(10), 3226–​33.
    tients of the German cooperative protocols MAKEI 83/​86 and 89. J Schneider DT, et al. (2000) Primary mediastinal germ cell tumors in
    Clin Oncol 19(7), 1943–​50. children and adolescents: results of the German cooperative proto-
    Göbel U, et al. (2010) Brain metastases in children and adolescents cols MAKEI 83/​86, 89, and 96. J Clin Oncol 18(4), 832–​9.
    with extracranial germ cell tumor—​data of the MAHO/​MAKEI-​ Schneider DT, et al. (2001) Genetic analysis of childhood germ cell
    registry. Klin Pädiatrie 222(3), 140–​4. tumors with comparative genomic hybridization. Klin Padiatrie
    Göbel U, et al. (2014) Testicular germ cell tumors in adolescents—​ 213(4), 204–​11.
    results of the protocol MAHO 98 and the identification of good risk Schneider DT, et al. (2001) Multipoint imprinting analysis indicates a
    patients. Klin Padiatrie 226(6–​7), 316–​22. common precursor cell for gonadal and nongonadal pediatric germ
    Heifetz SA, et al. (1998) Immature teratomas in children: patho- cell tumors. Cancer Res 61(19), 7268–​76.
    logic considerations. A report from the combined Pediatric Schneider DT, et al. (2002) Genetic analysis of mediastinal
    Oncology Group/​Children’s Cancer Group. Am J Surg Pathol nonseminomatous germ cell tumors in children and adolescents.
    22(9), 1115–​24. Genes Chromosomes Cancer 34(1), 115–​25.
    Future perspectives 261

    Schneider DT, et al. (2004) Epidemiologic analysis of 1,442 children Wermann H, et al. (2010) Global DNA methylation in fetal human
    and adolescents registered in the German germ cell tumor proto- germ cells and germ cell tumours: association with differentiation
    cols. Pediatr Blood Cancer 42(2), 169–​75. and cisplatin resistance. J Pathol 221(4), 433–​42.
    Siegel R, et al. (2014) Cancer Statistics 2014. CA Cancer J Clin. 64(1), 9. Wessalowski R, et al. (2013) Regional deep hyperthermia for salvage
    Veltman I, et al. (2005) Identification of recurrent chromosomal treatment of children and adolescents with refractory or recurrent
    aberrations in germ cell tumors of neonates and infants using non-​testicular malignant germ-​cell tumours: an open-​label, non-​
    genomewide array-​ based comparative genomic hybridization. randomised, single-​institution, phase 2 study. Lancet Oncol 14(9),
    Genes Chromosomes Cancer 43(4), 367–​76. 843–​52.
    30
    Childhood Liver Tumours
    Giorgio Perilongo and József Zsiros

    Introduction the APC and AXIN2, have also been reported. Thus, overall, the
    Wnt pathway is involved in about 90% of the genetic abnormalities
    Childhood liver tumours are a heterogeneous group of neo- documented in hepatoblastoma. Due to mutations in β-catenin it
    plasms encompassing individually rare histological varieties. bypasses the proteasomal degradation pathway and is translocated
    Hepatoblastoma and hepatocellular carcinoma are the two most into the nucleus where it acts as a transcriptional activator and
    frequent subtypes; undifferentiated sarcoma of the liver, atyp- can be detected by immunohistochemistry. Potential prognostic
    ical teratoid/​rhabdoid tumour, primary hepatic germ-​cell tumour, relevance of this finding has been proposed. It has been recently
    lymphoma, angiosarcoma are examples of other, very rare pri- shown that the Hippo pathway (YAP) most likely cooperates with
    mary liver tumours in childhood. This chapter will focus only on β-​catenin in the pathogenesis of hepatoblastoma. The mechanism
    hepatoblastoma and hepatocellular carcinoma. of YAP activation and its cooperation with β-​catenin remain to be
    discovered.
    Microarray analysis shows that the glypican 3 (GPC3) gene, mu-
    Hepatoblastoma tated in Simpson–​Golabi–​Behmel tissue overgrowth syndrome, in
    which an association with an increased risk of hepatoblastoma de-
    Biology velopment has been hypothesized, is one of the most overexpressed
    Hepatoblastoma is an embryonal tumour of the liver. The term genes in hepatoblastoma.
    ‘embryonal’ implies that its origin is related to a derangement Growth regulation of hepatoblastoma involves the insulin-​like
    in the mechanisms regulating normal hepatic organogenesis. growth factor II (IGF2) signalling pathway. IGF2 is a maternally im-
    Various hypotheses have already linked the different histo- printed gene and encodes a fetal peptide hormone that regulates cell
    logical subtypes of hepatoblastoma to specific stages of the ar- proliferation, differentiation, and cell migration. IGF2 is expressed
    rest of normal hepatic organogenesis. It is hypothesized that in hepatoblastoma, inversely correlated with the degree of differen-
    hepatoblastoma arises from undifferentiated hepatic stem cells or tiation: it is lacking in fetal-​type cells and high in embryonal-​type
    human fetal liver multipotent progenitor cells (hFLMPC) capable cells. High expression of IGF2 is the hallmark of the Beckwith–​
    of differentiation into a variety of tissues. This phenomenon could Wiedemann syndrome—​an overgrowth syndrome characterized
    explain why there are histological differences between the various by hemihypertrophy and organomegaly—​and this finding could ex-
    hepatoblastomas and why almost 40% of them have mixed histo- plain the higher incidence of hepatoblastoma in children affected by
    logical components including both epithelial and mesenchymal this condition.
    elements. Further investigation is required to clarify which de-
    Epidemiology
    velopmental signalling, and transcriptional pathways are af-
    fected and at what level, and for what reason such changes occur. The estimated incidence of hepatoblastoma is 1.2 per million, with
    Nevertheless, different genetic abnormalities have been described a male prevalence, but no geographical variations have been re-
    in association with hepatoblastoma. ported. Little is know about the risk factors for hepatoblastoma de-
    Hepatoblastoma, as most of paediatric tumours do, presents velopment. A consistent association appears to exist between risk
    with a limited number of cytogenetic alterations and genetic mu- of hepatoblastoma and very low birthweight; duration of oxygen
    tations (average 2.9 mutations per case). Mutation of the β-​catenin therapy has been identified as a significant factor associated with
    (CTNNB1) gene, the key molecule of the Wnt signalling pathway, that risk. The International Agency for Research on Cancer has re-
    represents the only recurrent alteration found in two thirds of cently classified tobacco smoke (via parental smoking) as a human
    hepatoblastoma. Hepatoblastoma is the tumour with the most carcinogen for childhood hepatoblastoma.
    interstitial deletions or missense mutations in the CTNNB1 gene Most cases of hepatoblastoma are sporadic but at least three
    (reported in 60–​70% of the cases). Mutations in other components congenital conditions have been linked to an excess risk: trisomy
    of the β-​catenin degradation complex, including those affecting 18, Beckwith–​Wiedemann syndrome, and familial adenomatous
    Hepatoblastoma 263

    polyposis (FAP). For patients with Beckwith–​Wiedemann syn- Laboratory investigations should include a full blood count and
    drome, the risk of developing hepatoblastoma in those children serum AFP and β-​HCG. In most children with hepatoblastoma,
    presenting uniparental disomy on chromosome 15.5 is increased liver-​function tests are within normal limits.
    to 4.7%. For this reason, routine screening with ultrasound and
    serum AFP is recommended at least during the first years of life. Histological classification
    Children with a positive family history of FAP have a 750-​to 7,500-​ In order to reach a consensus on the histopathology and classi-
    fold increased risk of developing hepatoblastoma. An accurate fication of paediatric liver tumours, and hepatoblastoma in par-
    family history aiming to identify the presence of a positive his- ticular, an International Pathology Symposium was organized in
    tory of FAP is part of the diagnostic work-​up of all children with 2011. Pathologists expert in paediatric liver tumors, including those
    hepatoblastoma. serving as central reviewers of the major cooperative clinical trials,
    The suggestion that hepatoblastoma is a developmental aberra- convened together. The histological classification of malignant
    tion is supported by the fact that a wide spectrum of congenital mal- paediatric liver tumours that they elaborated is reported in Box 30.1.
    formation syndromes and isolated anomalies have been reported in
    cases of hepatoblastoma. The overall frequency of anomalies among
    children with hepatoblastoma appears to be higher than for any
    Box 30.1 SIOPEL classification of primary malignant paediatric
    paediatric tumour other than Wilms. liver tumours
    Clinical presentation Epithelial tumours
    Hepatoblastoma is a tumour of early childhood, with more than Hepatocellular
    80% of the patients being diagnosed under the age of two years. Hepatoblastoma
    An asymptomatic abdominal mass in an otherwise healthy young • Epithelial variants
    child is the classical presenting feature of hepatoblastoma, and • Pure fetal with low mitotic activity
    • Fetal, mitotically active
    systemic symptoms are rarely present. The presence of clinical • Pleomorphic, poorly differentiated
    signs of associated syndromes, such as Beckwith–​ Wiedemann • Embryonal
    syndrome, should always be sought and may dominate the clin- • Small-​cell undifferentiated
    ical picture. The hallmark of hepatoblastoma is an elevated serum • INI1-​negative
    level of alpha fetoprotein (AFP), although tumours associated with • INI1-​positive
    • Epithelial mixed (any/​all of above)
    normal or low (less than 100 ng/​ml) levels of AFP do occur, but
    • Cholangioblastic
    are very rare (less than 2%). Levels may be more difficult to inter- • Epithelial macrotrabecular pattern
    pret in very young infants because of the high physiological pro- Mixed epithelial and mesenchymal
    duction of AFP, and should always be interpreted in relation to • Without teratoid features
    the normal range for age. Extremely high levels of AFP in some • With teratoid features
    patients can generate erroneously low results on standard labora- Hepatocellular carcinoma (HCC)
    • Classic HCC
    tory testing and require appropriate serum dilution for a reliable as-
    • Fibrolamellar HCC
    sessment. Paraneoplastic abnormalities of calcium metabolism can • Hepatocellular neoplasm not otherwise specified (NOS)
    occur with a picture of pseudo-​osteoporosis and spontaneous frac-
    Biliary
    tures. Rarely, hepatoblastoma can also secrete β-​human chorionic
    Cholangiocarcinoma
    gonadotropin (β-​HCG) which may result in pseudo-​precocious Combined (hepatocellular cholangioca)
    puberty in males. Thrombocytosis, related to the hyperproduction
    of thrombopoietin by tumour cells, is frequently documented at Mesenchymal tumours
    presentation. Infrequently, a child with hepatoblastoma will present Malignant
    with an acute abdomen due to tumour rupture or intra-​tumoural Embryonal sarcoma
    Rhabdomyosarcoma
    haemorrhage.
    Vascular tumours
    Diagnosis Epithelioid haemangioendothelioma
    Whilst the presence of a hepatic mass in a young child, associ- Angiosarcoma
    ated with elevated AFP, is extremely suggestive of the diagnosis of Other malignancies/​tumours of uncertain origin
    hepatoblastoma, surgical biopsy is required to establish histological Malignant rhabdoid tumour INI1-negative (documented INI1 mutation or
    diagnosis. INI1-positive)
    Radiological investigation is required to define the extent of Nested epithelial stromal tumour
    the primary disease and confirm or exclude the possibility of me-
    Other
    tastases. Abdominal ultrasound and computed tomography (CT)
    Germ-​cell tumour
    and/​or magnetic resonance imaging (MRI) scans are the conven-
    Teratoma
    tional tools to document the presence, extension, and charac- Yolk sac tumour
    teristics of a hepatic mass. As slightly more than 20% of patients
    Reproduced with permission from López-​Terrada et al. ‘Towards an international
    present with lung metastases, contrast-​enhanced CT of the lung pediatric liver tumor consensus classification: proceedings of the Los Angeles COG
    is part of the diagnostic work-​up for any child presenting with Liver Tumors Symposium’. Modern Pathology, Volume 27, pp. 472–​491. Copyright ©
    2014 Springer Nature Publishing. DOI: https://​doi.org/​10.1038/​modpathol.2013.80
    hepatoblastoma.
    264 CHAPTER 30 Childhood Liver Tumours

    About 55% of hepatoblastomas are epithelial (30% fetal, 20% fetal-​ (for age) or low level (less than 100 ng/​ml) of AFP is a hallmark of
    embryonal, 3% macrotrabecular, 2% small-​cell undifferentiated), a poorer prognosis: in the SIOPEL group experience, three-​year
    and 45% are mixed epithelial and mesenchymal. However, when all overall survival (OS) of children presenting with low or normal AFP
    types are considered, around 85% contain both fetal and embryonal was around 15%. An association between small-​cell undifferentiated
    components in variable proportions. Evidence is accumulating that hepatoblastoma and low or normal for age AFP has been proposed.
    small-​cell undifferentiated hepatoblastoma is associated with un- Finally, the age of the patients also emerged as a potential prognostic
    favourable outcome, whilst pure fetal hepatoblastoma with a low factor, with a different cut-​off within each PRETEXT category.
    mitotic index (less than 2 mitoses per 10 HPF (high-​power fields)), The biological meaning of this risk stratification is still to be inves-
    if completely resected, has an excellent outcome even without adju- tigated. Similarly, the impact of the different histological subtypes
    vant treatment. of hepatoblastoma is a matter of current analysis. However, solid
    evidence supports the concept that hepatoblastoma with pure fetal
    Staging and risk stratification histology and low mitotic activity is associated with a very good out-
    Two different systems to describe tumour extension or stage are come whilst, alternatively, small-​cell undifferentiated tumours carry
    presently in use: one used by the North American Cooperative a dismal prognosis.
    Group on Childhood Hepatic Tumors, and another one used by the
    International Childhood Liver Tumour Strategy Group (SIOPEL Treatment strategies
    group), consisting mainly of European investigators. The American The goal of any therapeutic strategy directed at the cure of children
    system is based on the results of the initial surgical procedure with hepatoblastoma is to achieve complete tumour resection, al-
    (postsurgical system). It encompasses four stages: though surgery alone is not sufficient and the best chance of cure can
    • Stage I—​ complete tumour resection without microscopical only be realized with a multidisciplinary treatment strategy that also
    residual includes systemic chemotherapy. The two major cooperative groups
    • Stage II—​tumour grossly resected with evidence of microscopical offer different approaches to primary treatment. The North American
    residual Children’s Oncology Group (COG) utilizes a strategy based on an ini-
    • Stage III—​tumour unresectable or resected with gross residual tial attempt at surgical resection (primary surgery) regardless of the
    tumour or positive lymph nodes tumour extension within and/​or outside the liver. In contrast, the
    SIOPEL group takes an approach based on the use of primary chemo-
    • Stage IV—​tumours presenting with metastases.
    therapy before definitive surgery (neoadjuvant chemotherapy) regard-
    A progressive decline in the survival rate with evolution of stage in less of the initial tumour extension. The rationale for using primary
    this system has been firmly documented. chemotherapy is to reduce the tumour volume and to make the tu-
    The SIOPEL group developed an alternative system that describes mour more solid, less prone to bleed, and better demarcated from the
    tumour extension before any surgery (presurgical system); this is surrounding healthy liver parenchyma and, ultimately, more easily re-
    known as the PRETEXT system (pretreatment extension of the dis- sected. The results of these two strategies have been comparable, both
    ease). It has been developed in accordance with the surgical anatomy of achieving five-​year OS and event-​free survival (EFS) rates of over 70%.
    the liver, based on the vascular structures. This identifies four sectors, Historically, the most significant step in improving the prognosis
    two in each lobe. The PRETEXT category is defined by the number of of hepatoblastoma was the introduction of cisplatin-​based chemo-
    sectors involved by the tumour (see Table 30.3). The presence of dis- therapy, after which resection rate and three-​year OS improved from
    tant metastases is identified by the addition of the letter ‘M’; extension historical rates of 44% and 30% to more than 80% and 70%, respect-
    of tumour into portal vein is indicated by the letter ‘P’; extension into ively. COG conducted a trial that compared cisplatin (90 mg/​m2
    the vena cava and/​or hepatic veins by the letter ‘V’; and the presence of as an infusion over six hours) followed by doxorubicin (80 mg/​m2
    intra-​abdominal extrahepatic disease by the letter ‘E’. The prognostic as a continuous infusion over 96 hours) with the combination of
    relevance of the PRETEXT system has also been well documented. cisplatin (delivered as per above) on day one, and vincristine (1.5
    In 2011, a global coalition, the Children’s Hepatic Tumours mg/​m2) and 5-​fluorouracil (600 mg/​m2) on day two (C5V). They
    International Collaboration (CHIC), was constituted with the main subsequently chose C5V as their standard chemotherapy, not be-
    aim of creating a common approach to staging and risk stratification cause there was a significant difference between the two regimens
    in hepatoblastoma. The data derived from the prospective random- but because of a difference in toxicity profiles. It should be noted,
    ized trials conducted by the North American, European, German, however, that while tumour progression accounted for 86% of all
    and Japanese cooperative trials on hepatobalstoma were merged in reported events for patients treated with C5V, it represented only
    a single large database. A total of 1,650 children were entered into 56% of all events observed in those patients treated with cisplatin/​
    that database. The patient risk-​stratification system they elaborated doxorubicin. The SIOPEL group has adopted the combination cis-
    is reported in Figure 30.1 platin (90 mg/​m2 as an infusion over 24 hours) followed by doxo-
    In summary, within the backbone of the PRETEXT system it clearly rubicin (60 mg/​m2 infusion over 48 hours) as its therapeutic gold
    emerged that the completeness of surgical resection still remains the standard. The COG recently conducted an upfront window vincris-
    most important single prognostic factor for hepatoblastoma, and tine /​irinotecan (VI) treatment of high-​risk hepatoblastoma. In the
    only those children in whom complete tumour resection is achieved 32 children they enrolled, three-​year EFS and OS rates were 49%
    can hope for cure. The second most important, negative prognostic (95% CI, 30–​65%) and 62% (95% CI, 42–​77%), respectively. These
    factor is the presence of metastases at diagnosis. Furthermore, the data allowed them to conclude that the VI combination appears to
    analysis of the data so far collected confirms the fact that a normal have substantial activity against high-​risk hepatoblastoma.
    Hepatoblastoma 265

    PRETEXT I
    Any age Resectable at diagnosis Yes Very low

    VPEFR- No Low

    PRETEXT I M- Any AFP VPEFR+ <8 years Intermediate

    ≥8 years
    High
    M+

    PRETEXT II
    Resectable at diagnosis Yes Very low

    VPEFR- No Low

    AFP >100 ng/mL VPEFR+ Intermediate

    <8 years AFP ≤100 ng/mL High

    PRETEXT II M- ≥8 years High

    M+ High

    PRETEXT III
    VPEFR- Low

    AFP >1000 ng/mL VPEFR+ Intermediate

    AFP 101–1000 ng/mL Intermediate

    <8 years AFP ≤100 ng/mL High

    PRETEXT III M- ≥8 years High

    M+ High

    PRETEXT IV
    AFP >100 ng/mL VPEFR+ or VPEFR- Intermediate

    <3 years AFP ≤100 ng/mL High

    PRETEXT IV M- ≥3 years High

    M+ High

    Figure 30.1 Risk-​stratification trees from the Children’s Hepatic Tumors International Collaboration
    AFP = α-​fetoprotein; M = metastases; PRETEXT = PRETreatment EXTent of disease group; VPEFR = macro-​vascular involvement of hepatic veins (V) or portal bifurcation (P),
    contiguous extrahepatic tumour (E), multifocal tumour (F), and spontaneous rupture (R).
    Reproduced with permission from Meyers RL, et al. ‘Risk-​stratified staging in paediatric hepatoblastoma: a unified analysis from the Children’s Hepatic Tumors International
    Collaboration’. Lancet Oncology, Volume 18, Issue 1, pp. 122–​131. Copyright © 2017 Elsevier Ltd. DOI: 10.1016/​S1470-​2045(16)30598-​8
    266 CHAPTER 30 Childhood Liver Tumours

    Table 30.1 Results of the North American trials for childhood hepatoblastoma

    Trial Type of trial/​inclusion Regimen Outcome


    criteria
    CCG 862 Single arm Vincristine + cyclophosphamide/​doxorubicin alternating 3-​year OS:
    All patients with vincristine, cyclophosphamide, 5-​fluorouracil Stage I = 94%
    Stage II = 57%
    Stage III = 20%
    Stage IV = 14%
    CCG 823F Single arm Cisplatin/​doxorubicin 2-​year OS:
    Stage II = 86%
    Stage III = 58%
    Stage IV = 32%
    POG 8697 Single arm Cisplatin/​vincristine/​5-​fluorouracil 4-​year OS:
    Stage I/​PFH = 100%
    Stage I/​II = 90%
    Stage III = 67%
    Stage IV = 12%
    INT 0089 Prospective randomized Cisplatin/​vincristine/​5-​fluorouracil vs cisplatin/​doxorubicn 4-​year OS:
    trial Stage I/​II 100% vs 96%
    Stage III 66% vs 71%
    Stage IV 33% vs 42%
    POG 9345 Single arm Carboplatin followed by carboplatin/​vincristine/​ 4-​year OS:
    Stage III/​IV 5-​fluorouracil Stage III =73%
    Stage IV = 27%
    COG P9645 Prospective randomized Cisplatin/​vincristine/​5-​fluorouracil vs cisplatin/​carboplatin 3-​year OS:
    Stage III/​IV All patients = 75% vs 56%

    Stage I = microscopic complete resection; Stage II = microscopical residuals; Stage III = macroscopical residual; Stage IV = presence of metastases; OS = overall survival.

    Risk-​adapted treatment administration. Three-​year OS and EFS of the whole cohort of high-​
    In the course of subsequent clinical trials, these two principal re- risk hepatoblastoma were 79% (95% CI, 66–​87%) and 83% (73–​93%)
    gimens have been further refined. COG achieved 100% three-​year respectively. Particularly significant were the results achieved in the
    survival in children with pure fetal hepatoblastoma after gross com- group of children presenting with metastases, whose corresponding
    plete resection (Stage I and II, favourable histology) with adjuvant figures were 77% (95% CI, 63–​90%) and 79% (95% CI, 66–​92%)
    chemotherapy using four cycles of C5V. These excellent results have respectively
    prompted the further reduction of chemotherapy to only two cycles Concerns about the ototoxicity and nephrotoxicity of cisplatin
    for most of these patients and, for children with completely resected have stimulated both the COG and the SIOPEL group to inves-
    (Stage I) pure fetal hepatoblastoma and low mitotic index (less than tigate the possibility of integrating the use of toxicity-​protective
    2 mitoses per 10 HPF), omission of all adjuvant chemotherapy. On agents within their standard regimens. The SIOPEL group has
    the other hand, children with unresectable and metastatic tumour recently closed a prospective randomized controlled trial in
    still do not enjoy satisfactory long-​term outcome and have OS of low-​risk hepatoblastoma that tested the role of sodium thiosul-
    approximately 50% and 30% respectively. The current COG strategy fate (STS), an otoprotective agent, in preventing hearing loss in
    is to explore whether the addition of doxorubicin will benefit pa- children treated with cisplatin alone. The preliminary data indi-
    tients with unresectable localized disease, whilst the very poor re- cate that STS does not jeopardize the prognosis of these children
    sults achieved in patients with metastases at diagnosis has prompted and significantly reduces the hearing loss that these children may
    a search for new agents using a therapeutic ‘window approach’. suffer.
    The SIOPEL group have been able to show that cisplatin alone is The results of the series of trials so far run by the North American
    as effective as combination cisplatin/​doxorubicin in children with Study Group and by SIOPEL are reported in Tables 30.1 and 30.2,
    standard-​risk hepatoblastoma (data from SIOPEL 3 SR study—​ while their present therapeutic recommendations are summarized
    see Table 30.2). For children with high-​risk hepatoblastoma, the in Table 30.3.
    SIOPEL group undertook a trial (SIOPEL 3) exploring the use of Following the earlier cooperation between the SIOPEL group and
    alternating courses of cisplatin and the combination of carboplatin/​ COG, large cooperative trials on childhood hepatoblastoma will be
    doxorubicin. The results were encouraging, with improved EFS launched in the near future. The therapeutic recommendations of
    and OS estimates at three years of 65% (95% CI, 57–​73%) and 69% these trials are reported in Table 30.3.
    (95% CI, 62–​77%) for the whole group, 75% and 77% for patients
    with PRETEXT-​IV tumours, and 57% and 63% for those with lung Orthotopic liver transplantation
    metastases. In the past two decades, the role of orthotopic liver transplantation
    In the subsequent generation of clinical trials directed at children for children with hepatoblastoma has been quite clearly delineated.
    with advanced hepatoblastoma, the SIOPEL group looked into a In summary, children with hepatoblastoma eligible for transplant
    further intensification in the use of cisplatin, based on the weekly should meet the following criteria:
    Hepatocellular carcinoma 267

    Table 30.2 Results of the International Childhood Liver Tumor Strategy Group (SIOPEL) trials for childhood hepatoblastoma

    Trial Type of trial/​inclusion criteria Regimen Outcomes


    SIOPEL 1 Single arm/​all patients Cisplatin/​doxorubicin 5-​year OS 66%
    3-​year EFS 75%
    SIOPEL 2 SR Single arm Cisplatin alone 3-​year OS 91% (±7%)
    SR-​HB 3-​year PFS 89% (±7%)
    SIOPEL 2 HR Single arm Cisplatin and carboplatin/​doxorubicin 3-​year OS 53% (±13%)
    HR-​HB 3-​year PFS 48% (±13%)
    SIOPEL 3 SR Prospective randomized trial Cisplatin/​doxorubicin vs cisplatin alone Cisplatin/​doxorubicin:
    SR-​HB 3-​year OS 93% (±5%)
    3-​year EFS 85% (±7%)
    Cisplatin alone:
    3-​year OS 95% (±4%)
    3-​year EFS 83% (±7%)
    SIOPEL 3 HR Single arm Cisplatin and carboplatin/​doxorubicn 3-​year OS 69(±7)%
    HR-​HB 3-​year EFS 65(±8)%
    SIOPEL 4 HR Single arm Weekly cisplatin with doxorubicin and 3-​year OS 83% (±10%)
    HR-​HB carboplatin 3-​year EFS 79% (±8%)

    SR = standard risk; HR = high risk; SR-​HB = standard-​risk hepatoblastoma: tumour confined to the liver, involving at the most three hepatic sectors associated with alpha fetoprotein
    >100 ng/​ml; HR-​HB = high-​risk hepatoblastoma: tumour involving the entire liver, and/​or presenting with metastases, and/​or with vascular invasion, and/​or with extrahepatic
    abdominal disease, and/​or with alpha fetoprotein <100 ng/​ml; EFS = event-​free survival; OS= overall survival; PFS = progression-​free survival.

    • They have had a response to chemotherapy—​a criterion which candidate for transplantation, it is important to ensure early assess-
    implies that any child with hepatoblastoma should be treated ment by a transplant team, so that the treatment strategy can be
    with chemotherapy regardless of the type of resection which is carefully timed.
    envisaged.
    • They have a tumour completely confined to the liver but deemed
    unresectable. Usually these are those clearly involving all four sec- Hepatocellular carcinoma
    tors of the liver as judged by magnetic resonance imaging (MRI)
    scan and/​or angiography, or which lie so close to the main ves- Hepatocellular carcinoma is an even rarer tumour than
    sels at the hilum of the liver and/​or to the hepatic veins to make a hepatoblastoma. Among the 213 liver tumour cases diagnosed in
    tumour-​free excision margin highly unlikely. the UK National Registry of Childhood Tumours between 1995
    and 2006, there were 154 (72.3%) cases of hepatoblastoma and 22
    The presence of metastases at diagnosis is not an absolute contra- of hepatocellular carcinoma, including five fibrolamellar tumours
    indication to transplant but requires complete clearing of the lungs (10.3%). All other diagnoses represented a heterogeneous group of
    with chemotherapy. In managing a child who may be a potential even rarer primary liver tumours. Hepatoblastoma dominated the
    under one year and one to four years age groups, with low rates for
    other tumours. In 5–​9 and 10–​14 year olds there was a more even
    Table 30.3 Hepatoblastoma: summary of the therapeutic distribution of rates between the tumour groups, but hepatocellular
    recommendations of the Children’s Oncology Group and of the
    SIOPEL group, according to the PRETEXT system carcinoma had a higher incidence than other tumours in 10–​14 year
    olds, albeit still very low in absolute terms.
    PRETEXT Therapeutic recommendation The majority of hepatocellular carcinoma cases across all ages
    Small I/​M–​ Surgery at diagnosis (primary surgery) seen in developing countries are due to hepatitis B virus infection
    Completely resected UH–​CDDP x 2 (HBV) and to aflatoxin exposure, while in developed countries, the
    If complete resection and pure fetal histology, no further
    more common aetiological links are to alcohol abuse and smoking.
    therapy
    Hepatitis C virus (HCV) is also a factor, especially in Japan. The in-
    II/​M–​ Upfront chemotherapy CDDP x 2
    Surgery CDDP x 2 vs CDDP x 4
    cidence of hepatocellular carcinoma in HBV hyperendemic areas
    in China and Thailand appears to have fallen. Hepatocellular car-
    III and IV/​M–​ Three-​arm randomization:
    Historical COG regimen—​CV5D vs SIOPEL 3 cinoma may also arise in children with various inherited metabolic
    High-​risk regime—​CDDP alternating with carbo/​DOXO vs disorders.
    SIOPEL 3 Similar to hepatoblastoma, a hard abdominal mass is the classical
    Standard-​risk regimen—​CDDP alone
    presenting symptom of hepatocellular carcinoma, but is more com-
    M+ Upfront chemotherapy as per SIOPEL 4 regimen—​weekly monly associated with systemic symptoms such as fatigue, anorexia,
    CDDP/​DOXO and, based on clearing of the metastases,
    CDDP/​DOXO vs I/​V or CDDP/​DOXO vs CDDP/​DOXO and abdominal pain. Jaundice is rare but has been reported. In cases
    alternating I/​V that arise in the context of pre-​existing liver conditions, the clinical
    features of the underlying disease may dominate the clinical picture.
    C5VD = cisplatin, vincristine, 5-​fluorouracil, doxorubicin; CDDP = cisplatin;
    DOXO = doxorubicin; I = irinotecan; V = vincristine; M–​or M+ = absence or presence Seventy per cent of patients with hepatocellular carcinoma pre-
    of metastases; S = surgery; UH = unfavourable histology, meaning no pure fetal sented with elevated AFP. In patients with a fibrolamellar carcinoma,
    hepatoblastoma with low mitotic index.
    268 CHAPTER 30 Childhood Liver Tumours

    AFP level is normal but high-​serum vitamin B12-​binding capacity Eichenmüller M, et al. (2014) The genomic landscape of hepatoblastoma
    (transcobalamin) has been shown to be elevated and may be used and their progenies with HCC-​like features. J Hepatol 61, 1312–​20.
    as a tumour marker. Finally, the serological evidence of pre-​existing Katzenstein HM, et al. (2002) Treatment of unresectable and meta-
    infection or of metabolic conditions may be present in affected static hepatobalstoma: a Pediatric Oncology Group Phase II Study.
    patients. J Clin Oncol 20, 3438–​44.
    Katzenstein HM, et al. (2002) Hepatocellular carcinoma in children
    The diagnostic strategy is similar to that described for
    and adolescents: results from the Pediatric Oncology Group and
    hepatoblastoma. Abdominal ultrasound and contrast-​enhanced CT
    the Children’s Cancer Group Intergroup Study. J Clin Oncol 20,
    or MRI scans serve to delineate tumour extension within the liver
    2789–​97.
    and abdomen, chest CT is recommended to rule out lung metas- Kitanovski L, et al. (2009) Multifocal hepatoblastoma in a 6-​month-​old
    tases, and the definitive diagnosis relies on the histological assess- girl with trisomy 18: a case report. J Med Case Reports 3, 8319.
    ment of a biopsy. Li J, et al. (2008) Cancer incidence among children and adolescents in
    The prognosis of children with hepatocellular carcinoma is un- the United States, 2001–​2003. Pediatrics 121, e1470–​77.
    satisfactory. Results from the use of cisplatin and doxorubicin, with López-​Terrada D, et al. (2014) Towards an international pediatric liver
    surgery, in the first SIOPEL study gave five-​year EFS and OS of 17% tumor consensus classification: proceedings of the Los Angeles
    and 28%, respectively. No improvement in the prognosis was ob- COG Liver Tumors Symposium. Mod Pathol 27(3), 472–​91.
    served in the subsequent study, where an alternating regimen with Malogolowkin MH, et al. (2006) Intensified platinum therapy is an in-
    cisplatin, carboplatin, and doxorubicin was used, and long-​term effective strategy for improving outcome in pediatric patients with
    survival is reported only among the group of patients who achieve advanced hepatoblastoma. J Clin Oncol 24, 2879–​84.
    complete resection. The experience of the North American group is Malogolowkin MH, et al. (2008) Redefining the role of doxorubicin
    for the treatment of children with hepatoblastoma. J Clin Oncol 26,
    comparable: 46 patients were enrolled in the INT-​0098 trial which
    2379–​83.
    compared treatment with cisplatin/​ doxorubicin with the C5V
    Meyers RL, et al. (2017) Risk-​ stratified staging in paediatric
    regimen. EFS at five years was 19% for the whole cohort, but ranged hepatoblastoma: a unified analysis from the Children’s Hepatic
    from 88% for children with completely resected (Stage I) tumour to Tumors International Collaboration. Lancet Oncol 18(1), 122–​31.
    0% for those with metastatic disease. O’Neill AF, et al. (2017) Characterization of pulmonary metastases
    There is still some uncertainty about the role of chemotherapy in in children with hepatoblastoma treated on Children’s Oncology
    hepatocellular carcinoma and the place of liver transplantation is Group Protocol AHEP0731 (the treatment of children with all
    less well established than for hepatoblastoma. However, the intro- stages of hepatoblastoma): a report from the Children’s Oncology
    duction of biological agents, specifically sorafenib, a small-​molecule Group. J Clin Oncol 35(30), 3465–​73.
    multikinase inhibitor, is opening new approaches to the treatment of Ortega JA, et al. (1991) Effective treatment of unresectable or meta-
    this challenging tumour. static hepatoblastoma with cisplatin and continuous infusion doxo-
    rubicin chemotherapy: a report from the Children’s Cancer Study
    Group. J Clin Oncol 9, 2167–​76.
    FURTHER READING Ortega JA, et al. (2000) Randomized comparison of cisplatin/​vincris-
    Alkhouri N, et al. (2010) Familial adenomatous polyposis in children tine/​fluorouracil and cisplatin/​continuous infusion doxorubicin for
    and adolescents. J Pediatr Gastroenterol Nutr 51, 727–​32. treatment of pediatric hepatoblastoma: a report from the Children’s
    Aronson DC, et al. (2005) Predictive value of the pretreatment extent Cancer Group and the Pediatric Oncology Group. J Clin Oncol 18,
    of disease system in hepatoblastoma: results from the International 2665–​75.
    Society of Paediatric Oncology Liver Tumour Study Group Otte JB, et al. (2004) Liver transplantation for hepatoblastoma: results
    SIOPEL-​1 study. J Clin Oncol 23, 1245–​52. from the International Society of Pediatric Oncology (SIOP) Study
    Brown J, et al. (2000) Pretreatment prognostic factors for children with SIOPEL-​1 and review of the world experience. Pediatr Blood Cancer
    hepatoblastoma—​results from the International Society of Paediatric 42, 74–​83.
    Oncology (SIOP) study SIOPEL 1. Eur J Cancer 36, 1418–​25. Perilongo G, et al. (2000) Hepatoblastoma presenting with lung metas-
    Bell A, et al. (2017) Novel advances in understanding of molecular tases: treatment results of the first cooperative, prospective study of
    pathogenesis of hepatoblastoma: a Wnt/​βcatenin prospective. Gene the International Society of Paediatric Oncology on childhood liver
    Expr 17(2), 141–​54. tumors. Cancer 89, 1845–​53.
    Bulterys M, et al. (1999) Hepatic tumors. In: LAG Ries, et al. (eds) Cancer Perilongo G, et al. (2004) Risk-​ adapted treatment for child-
    Incidence and Survival among Children and Adolescents: United hood hepatoblastoma. Final report of the second study of the
    States SEER Program 1975–​1995, pp. 91–​7. NIH Pub. No. 99-​4649. International Society of Paediatric Oncology—​SIOPEL 2. Eur J
    Bethesda, MD: National Cancer Institute, SEER Program. Cancer 40, 411–​21.
    Czauderna P, et al. (2002) Hepatocellular carcinoma in children; re- Perilongo G, et al. (2009) Cisplatin versus cisplatin plus doxorubicin
    sults of the first prospective study of the International Society of for standard risk hepatoblastoma. N Engl J Med 361, 1662–​70.
    Pediatric Oncology. J Clin Oncol 20, 2798–​804. Pritchard J, et al. (2000) Cisplatin, doxorubicin, and delayed surgery
    Czauderna P, et al. (2016) The Children’s Hepatic Tumors International for childhood hepatoblastoma: a successful approach—​results of
    Collaboration (CHIC): novel global rare tumor database yields the first prospective study of the International Society of Pediatric
    new prognostic factors in hepatoblastoma and becomes a research Oncology. J Clin Oncol 18, 3819–​28.
    model. Eur J Cancer 52, 92–​101. Roebuck DJ, et al. (2007) 2005 PRETEXT: a revised staging system
    De Ioris M, et al. (2008) Hepatoblastoma with a low serum alpha-​ for primary malignant liver tumours of childhood developed by
    fetoprotein level at diagnosis: the SIOPEL group experience. Eur J the SIOPEL Group International Childhood Liver Tumor Strategy
    Cancer 44, 545–​50. Group. Pediatr Radiol 37, 123–​32.
    Hepatocellular carcinoma 269

    Roebuck DJ, et al. (2007) Assessment of extrahepatic abdominal ex- Zsíros J, et al. (2010) Successful treatment of childhood high risk
    tension in primary malignant liver tumours of childhood. Pediatr hepatoblastoma with dose intensive multiagent chemotherapy and
    Radiol 37, 1096–​100. surgery—​final results of the SIOPEL-​3HR study of the Childhood
    Trobaugh-​Lotrario AD, et al. (2009) Small cell undifferentiated variant Liver Tumor Strategy Group. J Clin Oncol 28, 2584–​90.
    of hepatoblastoma: adverse clinical and molecular features similar Zsiros J, et al. (2013) Dose-​dense cisplatin-​based chemotherapy and
    to rhabdoid tumors. Pediatr Blood Cancer 52(3), 328–​34. surgery for children with high-​ risk hepatoblastoma (SIOPEL-​
    Zimmermann A (2005) The emerging family of hepatoblastoma tu- 4): a prospective, single-​arm, feasibility study. Lancet Oncol 14(9),
    mours: from ontogenesis to oncogenesis. Eur J Cancer 41, 1503–​14. 834–​42.
    31
    Retinoblastoma
    Guillermo Chantada and Carlos Rodríguez-​Galindo

    Introduction the family history, more than 90% of neonatal cases have either bi-
    lateral disease at presentation or will develop asynchronous bilateral
    Retinoblastoma is a malignant tumour arising from the embryonic retinoblastoma.
    neural retina and occurs in about one in 14,000–​18,000 live births. The presenting features of retinoblastoma vary according to the
    Thus, an estimated 8,000 children develop retinoblastoma each year level of socioeconomic development. In developing countries, late
    worldwide. In some less developed countries, an increased incidence diagnosis is common and patients usually present with proptosis
    of retinoblastoma has been reported; however, in others, a similar and an orbital mass, often with pre-​auricular adenopathy (Figure
    pattern to developed nations was found. The reason for this is not 31.2). In developed countries, the most common presenting signs
    known; environmental or genetic factors may play a role. When ret- are the presence of a white reflex in the child’s eye (leukocoria;
    inoblastoma is diagnosed in the intraocular stage (as it usually oc- Figure 31.2), which occurs in two thirds of cases, or, less frequently,
    curs in the most affluent countries), disease-​free survival has been strabismus. Less common presenting signs include a painful red
    over 90% for the past decades. However, in less developed countries, eye, glaucoma, poor vision, orbital cellulitis, hyphema, or unilateral
    retinoblastoma is diagnosed later, when the disease has dissemin- mydriasis. Even though leukocoria is frequently present, it may be
    ated to extraocular sites. Consequently, survival in these countries is overlooked since it may be only seen when the child looks sideways.
    lower than 50% for most affected children. It may also be noticeable in a flash photograph. Whilst leukocoria
    is a relatively specific sign with few differential diagnoses, stra-
    bismus is very common in children and often due to a benign cause.
    Presenting signs and symptoms However, the occurrence of strabismus in a young child calls for a
    dilated examination of the retina, with special attention given to
    Retinoblastoma presents in two distinct clinical forms: the macula.
    1. The bilateral or multifocal, hereditary form (30–​35% of cases) A syndrome associated with deletion of the long arm of chromo-
    is characterized by the presence of germline mutations of the some 13 (the 13q deletion syndrome) has been reported, with mental
    RB1 gene. These may be the result of a new germline mutation retardation and dysmorphic features. Identification of these anom-
    (in 75% of cases) or may be inherited from an affected survivor alies may precede the recognition of concomitant retinoblastoma.
    (accounting for the remaining 25% of cases). In less developed Such children require karyotype analysis and retinal examination.
    countries, familial cases are less common. Trilateral retinoblastoma refers to the association of bilateral ret-
    inoblastoma with an asynchronous intracranial embryonal tumour.
    2. The unilateral form (65–​70% of cases) is mainly sporadic (in
    This association can occur in about 3% of patients with the gen-
    about 90% of cases); the remaining 10% have germline muta-
    etic form, and appears to be more common in familial cases. The
    tions. However, in the absence of a positive family history, it is
    prognosis has until recently been almost uniformly fatal, but with
    not possible, without performing a mutational analysis, to deter-
    current therapies results may be more encouraging. Even though it
    mine which unilateral cases are inheritable (Figure 31.1).
    can be detected concomitantly with intraocular retinoblastoma, the
    Patients with bilateral retinoblastoma tend to present before one interval between the diagnosis of bilateral retinoblastoma and the
    year of age, and those with unilateral disease often present in the diagnosis of the brain tumour is usually around 20 months.
    second or third year of life. However, there are wide variations in
    the age of diagnosis in different countries, according to the de-
    gree of socioeconomical development; children in less developed Biology
    countries are diagnosed at a later age. It is uncommon for retino-
    blastoma to be diagnosed during the first month of life, except in Retinoblastoma was the first cancer to be described as a genetic
    familial cases where screening is performed. However, regardless of disease. In 1971, Knudson proposed the ‘two-​hit hypothesis’, in
    Biology 271

    If parent was ...


    Bilateral Unilateral Unaffected
    Chance of 45% 55% 7–5% 85–93% <<1% 99%
    offspring having affected unaffected affected unaffected affected unaffected
    retinoblastoma

    85% 15% 85% 15% 33% 67%


    Laterality bilateral unilateral 0% bilateral unilateral 0% bilateral unilateral 0%

    Focality 100% 96% 4% 100% 96% 4% 100% 15% 85%


    multifocal multifocal unifocal 0% multifocal multifocal unifocal 0% multifocal multifocal unifocal 0%

    Chance of next
    sibling having
    retinoblastoma 45% 45% 45% 45% 45% 45% 45% 7–15% 5%* <1%* <1%* <1%

    *If parent is a carrier, then 45%

    Figure 31.1 Card for genetic counselling of retinoblastoma.


    Reproduced courtesy of David H. Abramson, Chief, Ophthalmic Oncology Service, Memorial Sloan-Kettering Cancer Center, New York, USA

    which two mutational events in a developing retinal cell lead to its lack or inactivation removes the constraint on cell-​c ycle con-
    the development of retinoblastoma. This hypothesis was subse- trol. The RB1 is a large gene, and mutations have been described
    quently extended to suggest that the two events could be mutations in almost every exon. There are no mutational hot spots, although
    of both alleles of the RB1 gene. RB1 is located within chromosome new germline mutations have an overwhelming preference for the
    13q14. Its product, pRb, is a key substrate for G1 cyclin-​cdk com- paternal allele. This second hit occurs at a much higher frequency
    plexes, which phosphorylate target gene products required for than the first hit, and it is more sensitive to environmental factors
    the transition of the cell through the G1 phase of the cell cycle. such as ionizing irradiation, thus explaining the increased risk of
    pRb is the major gatekeeper controlling growth regulation, and irradiation-​induced malignancies in survivors of retinoblastoma.

    (a)

    (b)

    Figure 31.2 Clinical presentation of retinoblastoma: (top) massive orbital dissemination; (bottom) leukocoria.
    Reproduced courtesy of Kahaki Kimani, Senior Lecturer, Department of Ophthalmology, University of Nairobi, Kenya.
    272 CHAPTER 31 Retinoblastoma

    Recently, cases of retinoblastoma with no RB1 mutation but Box 31.1 International Grouping System for intraocular disease
    MYCN amplification in small children with unilateral involve-
    ment were reported. Group A
    Identification of the cell of origin of retinoblastoma remains Small tumours away from foveola and disc
    elusive. Recent studies suggested that it derives from cone pre- • Tumours <3 mm in greatest dimension confined to the retina, and
    cursors, but other reports suggest that more than one cell can be • Located at least 3 mm from the foveola and 1.5 mm from the
    implicated. optic disc
    Group B
    All remaining tumours confined to the retina
    Histology • All other tumours confined to the retina not in Group A
    • Subretinal fluid (without subretinal seeding) <3 mm from the base of
    Retinoblastoma is a tumour of neuro-​epithelial origin con- the tumour
    sisting of small, undifferentiated, anaplastic cells with scanty Group C
    cytoplasm and large nuclei that stain deeply with haematoxylin, Local subretinal fluid or seeding
    arising from the nucleated layer of the retina. Calcification usu- • Local subretinal fluid alone >3 mm to <6 mm from the tumour
    ally occurs in necrotic areas. Retinoblastoma cells often express • Vitreous seeding or subretinal seeding <3 mm from the tumour
    photoreceptor-​differentiation antigens, neuron-​specific enolase Group D
    and the ganglioside GD2 (but not CD99), glial fibrillary acidic Diffuse subretinal fluid or seeding
    protein, or S-​100. Retinoblastoma cells may show features of • Subretinal fluid alone >6 mm from the tumour
    photoreceptor differentiation and Flexner–​Wintersteiner ros- • Vitreous seeding or subretinal seeding >3 mm from the tumour
    ettes, which are characteristic of retinoblastoma but may be
    Group E
    present in other ophthalmic tumours (medulloepithelioma). Presence of one or more of these poor prognosis features
    Alternatively, it may present with highly undifferentiated • More than two thirds of globe filled with the tumour
    retinoblasts or a mixed pattern. • Tumour in anterior segment
    The tumour may grow either in an endophytic pattern seeding the • Tumour in or on the ciliary body
    vitreous or in an exophytic form into the subretinal space. The active • Iris neovascularization
    seeds of retinoblastoma can remain viable for long periods following • Neovascular glaucoma
    chemotherapy or radiation. When the tumour grows from the retina • Opaque media from haemorrhage
    outwards into the subretinal space (exophytic pattern), it produces • Tumour necrosis with aseptic orbital cellulitis
    retinal detachment, sometimes with no clear view of the mass, and • Phthisis bulbi
    can resemble Coats disease or other forms of exudative retinal de- Reproduced from Aerts et al., ‘Retinoblastoma’, Orphanet Journal of Rare Diseases,
    Volume 1, Issue 31. Copyright © 2006 Aerts et al; licensee BioMed Central Ltd. doi.
    tachment. Both vitreous and subretinal seeding are the major obs- org/​10.1186/​1750-​1172-​1-​31, under the terms of the Creative Commons Attribution
    tacles to curing intraocular tumours with chemotherapy, because of License CC-​BY-​2.0 (http://​creativecommons.org/​licenses/​by/​2.0)
    the difficulty of chemotherapy to reach these areas with low vascular
    supply.
    Retinoblastoma can disseminate outside the eye, through the
    optic nerve and/​or the subarachnoid space to the chiasm, the brain, consensus, a staging system that articulates with the intraocular
    and the meninges. It can also escape from the eyeball, through the grouping system proposed for the eye-​conserving therapies with
    sclera, and invade the orbit and beyond it to the surrounding struc- chemoreduction (Table 31.1). More recently, a new version of the
    tures. The tumour cells can also reach the choroid and, from there, TNM system was developed which also incorporates imaging and
    may gain access to the systemic circulation, giving rise to haema- heredity data.
    togenous metastasis involving the bone, bone marrow, or any other
    Imaging studies
    organ. Metastatic retinoblastoma usually involves the central ner-
    vous system (CNS), either as a solitary mass or multiple lesions, All children with retinoblastoma should undergo gadolinium-​
    or with leptomeningeal dissemination. It can also invade facial enhanced magnetic resonance imaging (MRI) of the orbits and
    structures such as the pre-​auricular lymph nodes and the bones of brain. MRI is preferred to computed tomography (CT) scan be-
    the skull. cause it allows for a better estimation of the invasion to the optic
    nerve, it evaluates accurately the pineal area. and it avoids the ex-
    posure to radiation in these susceptible patients. MRI is also helpful
    in the differential diagnosis with Coats disease and other inflam-
    Diagnostics and extent of disease evaluation matory conditions, and with persistent hyperplastic primary vit-
    reous (PHPV). Given the short interval between the diagnosis of
    Grouping and staging retinoblastoma and the occurrence of trilateral retinoblastoma,
    A grouping system for intraocular disease, generated by inter- routine screening with MRI every six months until five years of
    national experts, has been in use for the past 15 years (Box 31.1), age might be recommended, yet it is not clear whether earlier diag-
    and recently data were incorporated into the tumour–​ node–​ nosis can impact survival. Benign pineal cysts may be encountered
    metastasis (TNM) system. For staging of extraocular disease, in these children, causing occasional diagnostic and management
    a group of international retinoblastoma experts developed, by dilemmas.
    Treatment 273

    Table 31.1 International Staging for Retinoblastoma less developed countries. In some countries, up to 50% of affected
    families opt out of enucleation as initial treatment; these children
    Stage Description
    die of a potentially curable disease. Therefore, in these problematic
    Stage 0 Patients treated conservatively (subject to presurgical settings, comprehensive strategies, including psychosocial support,
    ophthalmological classifications)
    should be implemented to improve survival. Some surgical proced-
    Stage I Eye enucleated, completely resected histologically ures, such as pars plana vitrectomy, are contraindicated in children
    Stage II Eye enucleated, microscopic residual tumour with retinoblastoma, and other aggressive surgeries, like orbital ex-
    Stage III Regional extension enteration, are currently seldom performed.
    a) Overt orbital disease
    b) Preauricular or cervical lymph-​node extension Focal therapies
    Stage IV Metastatic disease Focal treatments are used for small tumours (less than 3–​6 mm),
    a) Haematogenous metastasis—​1. Single lesion; 2. Multiple lesions
    b) CNS extension—​1. Prechiasmatic lesion; 2. CNS mass; usually in patients with bilateral disease, and in combination with
    3. Leptomeningeal disease chemotherapy. These treatments include photocoagulation with an
    argon laser (for the treatment of tumours posterior to the equator),
    Reproduced from ‘Retinoblastoma Treatment (PDQ®)—​Health Professional Version’
    © 2018 National Cancer Institute (https://​www.cancer.gov/​types/​retinoblastoma/​hp/​ transpupillary thermotherapy (which applies focused heat, usually
    retinoblastoma-​treatment-​pdq). with a diode laser, at subphotocoagulation levels), and cryotherapy
    (for the treatment of small anterior lesions).

    Ancillary studies Radiotherapy


    Additional evaluation of metastatic disease needs to be considered Radiotherapy, in combination with focal treatments, can provide
    only in patients with International Retinoblastoma Staging System excellent tumour control. However, radiation therapy increases the
    (IRSS) stage I and high-​risk features, and all patients with IRSS stage risk of second malignancies in heritable retinoblastoma survivors,
    II onwards. These investigations include bilateral bone-​marrow as- and so contemporary management of intraocular retinoblastoma
    pirates and biopsies, and lumbar puncture with examination of the is designed to avoid or delay its use; the role of external-​beam ra-
    cytospin. Since retinoblastoma cells may adhere to the tube, the diation (EBRT) is currently very limited in developed countries.
    examination of the cytocentrifugate should be done to improve the Radiation therapy continues to have a major role in the treatment of
    yield of this procedure. Bone scintigraphy is only recommended in patients with extraocular disease.
    children with confirmed metastatic disease or those with bone pain. Several techniques can be used, usually through lateral or anterior
    fields. Recommended total doses are 4000–​4500 cGy, although doses
    of 3600 cGy, in conjunction with other techniques, are under inves-
    Treatment tigation. The use of modern radiotherapy techniques like intensity-​
    modulated radiotherapy (IMRT) and proton therapy may lower
    Retinoblastoma is special among paediatric tumours in that two the risk of long-​term toxicity, especially by reducing the dose to the
    different treatments goals are sought by treating physicians: patient brain and other non-​ocular structures.
    survival, usually led by a paediatric oncologist, and eye survival, Radioactive plaque technique is useful when treating localized
    usually led by a paediatric ophthalmologist. Each need different tumours, both because the procedure time is short and also be-
    parameters to be considered for treatment in the context of a multi- cause a high dose of irradiation is delivered to the areas of interest
    disciplinary team. So, the treatment of retinoblastoma includes sur- while minimizing radiation effects to the extraocular structures. The
    gical procedures, local therapies, radiotherapy, and chemotherapy radioactive implant is placed on the sclera over the base of the tu-
    that can be given systemically or locally. Conservative therapies mour, and is kept there for two to four days—​the time needed to
    should be used with caution in less developed countries where pa- deliver approximately 4000 cGy to the apex of the tumour. Different
    tient compliance is poor or uncertain. Deaths after refusal of enucle- radioactive episcleral plaques can be used, although 125I and 106Ru
    ation or toxicity have been published. are the most widely used. Control rates of 85–​90% can be achieved.
    This treatment has not been associated with an increased risk of sec-
    Surgery ondary tumours when used in children with germline mutations.
    Enucleation is the simplest and safest therapy for retinoblastoma.
    A prosthetic implant is fitted after the procedure to minimize cos- Chemotherapy
    metic deficits. Enucleation should be performed by an experienced The uses of chemotherapy for the treatment of retinoblastoma in-
    paediatric ophthalmologist in order to obtain a long optic-​nerve clude three different scenarios.
    stump. Initial enucleation is indicated in all eyes with secondary glau-
    coma, invasion of anterior segment (anterior chamber, iris), rubeosis Tumour chemoreduction for conservative treatment
    iridis, impossibility of close follow-​up or limitations for using local Chemoreduction using systemic chemotherapy has been used as
    therapies, and group E eyes in most unilateral non-​hereditary pa- primary treatment for intraocular disease not amenable for local
    tients. Secondary enucleation is indicated after failure of conserva- therapy, in order to decrease tumour size and make the tumours
    tive approaches and after neoadjuvant chemotherapy in extraocular suitable for local therapy, for the past two decades. Carboplatin in
    disease. Even though enucleation is a curative treatment for retino- combination with vincristine or etoposide, given intravenously,
    blastoma and may be performed in most settings, regardless of their have been the most frequently used agents. A recent randomized
    complexity, it is not always accepted by the families, especially in study showed that the combination of carboplatin and etoposide
    274 CHAPTER 31 Retinoblastoma

    yielded marginally better results than carboplatin and vincristine. Even though chemoreduction followed by local treatment has
    Topotecan in combination with periocular carboplatin, alternating become the standard therapy for intraocular retinoblastoma,
    with the carboplatin–​etoposide–​vincristine regimen, proved to be there are several limitations to this approach. Systemic chemo-
    effective in advanced intraocular retinoblastoma. therapy may increase the risk of secondary malignancies, certainly
    The effectiveness and short-​ term toxicity of systemic as far as their potential induction is concerned, and especially
    chemoreduction were reported by many groups. With this treat- treatment-​induced leukaemia. Finally, this treatment is tedious
    ment, most intraocular tumours usually show dramatic shrinkage. and needs meticulous management; resources are only available
    However, consolidation with local treatment is needed in most cases in specialized centres.
    to prevent relapse. Tumour location and size correlate with respon-
    siveness to chemotherapy. Systemic chemoreduction followed by Adjuvant chemotherapy for high-​risk stage I and II disease
    local therapy is still the standard therapy for patients with less ad- After enucleation, the eye should be comprehensively evaluated by
    vanced disease, such as groups A to C (see Box 31.1), in whom up an experienced ocular pathologist: histopathological microstaging
    to 90% of the eyes can be salvaged without using EBRT. For patients using standardized criteria for risk assessment is essential to define
    with advanced intraocular tumours, belonging to group D (see Box groups with a different risk of extraocular relapse. Invasion to the
    31.1), and especially those with vitreous seeds, ocular salvage rates choroid, sclera, and optic nerve are considered as significant risk
    are not so encouraging with this therapy. Most of these patients factors for extraocular relapse. The role of adjuvant chemotherapy
    have been traditionally considered for initial enucleation. However, for some of these patients, to reduce the relapse rate, is a matter
    single-​institution experiences using superselective intra-​ arterial of controversy. Because these factors are infrequent in developed
    chemotherapy administration in the ophthalmic artery have shown countries, there have been no randomized studies published in the
    improved salvage rates. Melphalan has been the most commonly used literature as for other malignancies. There is almost universal agree-
    agent for intra-​arterial chemotherapy, but carboplatin and topotecan ment that there is no need of adjuvant chemotherapy for patients
    have also been employed. This treatment was reported to be highly with intraretinal disease and in those with prelaminar optic-​nerve
    effective in eyes with retinal detachment and subretinal seeds, which invasion. The role of chemotherapy in isolated choroidal invasion is
    were considered for enucleation in the past. Tandem administration controversial because the relapse rate, when there is no concomitant
    of intra-​arterial chemotherapy to both eyes in the same procedure postlaminar optic-​nerve involvement, is only about 3%. Different
    is feasible. This strategy is usually recommended in children older degrees of choroidal invasion may have different prognostic impli-
    than six months, weighing more than 6 kg, and it has also been ef- cations and some subgroups, such as those with massive invasion
    fective as salvage therapy after failure of systemic chemoreduction (defined as greater than 3 mm), might benefit from adjuvant chemo-
    and EBRT, with a salvage rate of 58% in results of single-​referral in- therapy. However, since the overall survival of these children is in
    stitutions. However, intra-​arterial chemotherapy is being used in the range of 95% with no other therapy than enucleation, it is un-
    middle-​income countries as well, with encouraging results. likely that a benefit from adjuvant chemotherapy will be unequivo-
    Intra-​arterial chemotherapy has a favourable systemic toxicity cally proven.
    profile compared to systemic chemotherapy. However, ocular tox- Invasion to the optic nerve beyond the lamina cribrosa is a
    icity, including severe choroidal damage leading to reduced vision major risk factor for relapse, and adjuvant therapy is recom-
    has been reported. Long-​term effects of intra-​arterial chemotherapy mended by most groups. When the resection margin is free of tu-
    are less known, but reports from Japan failed to find evidence of in- mour, most treatment groups recommend adjuvant chemotherapy
    creased risk of secondary malignancies. Intra-​arterial chemotherapy alone. With the use of current high-​resolution MRI, postlaminar
    as a frontline eye-​salvage procedure for unilateral group D and optic-​nerve invasion may be presumed preoperatively by detecting
    selected E patients is increasingly used by some treatment groups, contrast enhancement in the distal portion of the nerve. Some of
    with ocular survival higher than 80%. Even though no toxic mor- these patients have received neoadjuvant chemotherapy in an at-
    tality associated to this treatment was reported, occasional events tempt to avoid tumour residue at the resection margin. However,
    of metastatic relapse were reported in this population. So, this mo- identifying postlaminar optic-​nerve invasion through imaging
    dality should be used with caution for unilateral group D and E studies is a challenge and the sensitivity of MRI when the optic
    disease in areas where high-​risk pathology is more prevalent and nerve is not enlarged is about 60% in centres with extensive experi-
    a higher risk of extraocular relapse is seen, since enucleation yields ence. Patients with invasion of the resection margin of the optic
    excellent survival outcome and is usually well tolerated. nerve are classified as IRSS stage II because microscopic tumour
    More recently, intravitreous injection of melphalan became an ef- residue is left behind after enucleation, and they are uniformly
    fective therapy for children with vitreous seeds. Though contraindi- considered for adjuvant therapy with chemotherapy and orbital
    cated for decades because of the fear of accidental orbital seeding radiotherapy. In recent series, about 70% of these patients sur-
    of the tumour, an improved technique, reported recently, has made vive, albeit with important cosmetic and neuroendocrinological
    it a safe and relatively inexpensive treatment. However, it needs to sequelae because of radiation.
    be administered by experienced ophthalmologists. The combined The importance of microscopic scleral invasion as a risk factor for
    sequential use of intra-​ arterial chemotherapy and intravitreous relapse is less clear because it usually occurs with other concomitant
    chemotherapy appears to be a promising strategy for eyes with factors. When there is evidence of microscopic transscleral invasion,
    vitreous seeds. The optimal dose and schedule of intravitreous these children should be considered as stage II, but patients with
    melphalan are still not known. According to the current classifica- only intrascleral invasion (stage I) need adjuvant chemotherapy as
    tion of vitreous seeds, those presenting as clouds would need more well. The chemotherapy regimens used for the adjuvant therapy
    prolonged and intensive therapy for tumour control. of children with stage I disease and pathology risk factors include
    Second malignancies in retinoblastoma survivors 275

    the combination of carboplatin and etoposide, or more intensive


    regimens including the same agents at higher doses together with FURTHER READING
    alkylating agents such as cyclophosphamide or ifosfamide and an Abramson DH, et al. (2008) A phase I/​II study of direct intraarterial
    anthracycline, as it is used in other centres. The use of more intensive (ophthalmic artery) chemotherapy with melphalan for intraocular
    regimens is warranted for children with stage II disease such as those retinoblastoma initial results. Ophthalmology 115, 1398–​404.
    with postlaminar optic-​nerve and tumor at the resection margin or Aerts I, et al. (2013) Results of a multicenter prospective study on the
    microscopical scleral invasion. However, there is no consensus on postoperative treatment of unilateral retinoblastoma after primary
    enucleation. J Clin Oncol 31, 1458–​63.
    the adequate treatment intensity for these children.
    Antoneli CB, et al. (2003) Extraocular retinoblastoma: a 13-​year ex-
    Treatment of overt extraocular retinoblastoma (stages III perience. Cancer 98, 1292–​8.
    and IV) Canturk S, et al. (2010) Survival of retinoblastoma in less-​developed
    countries: impact of socioeconomic and health-​related indicators.
    In all these cases, neoadjuvant chemotherapy is warranted. Extensive Br J Ophthalmol 94, 1432–​6.
    and mutilating surgeries such as orbital exenteration are of no use. Chantada GL, et al. (2010) Results of a prospective study for the treat-
    About 70% of patients with stage III disease are curable with an ag- ment of unilateral retinoblastoma. Pediatr Blood Cancer 55, 60–​6.
    gressive approach using neoadjuvant chemotherapy, enucleation Corson TW, Gallie BL (2007) One hit, two hits, three hits, more?
    of the residual mass, and adjuvant chemo-​and radiotherapy. The Genomic changes in the development of retinoblastoma. Genes
    agents used are the same as those used for high-​risk stage I and II Chromosomes Cancer 46, 617–​34.
    disease. Orbital relapse after enucleation is also curable with the De Camargo B, et al. (2010) Cancer incidence among children and
    same approach. adolescents in Brazil: first report of 14 population-​based cancer
    Although a complete remission is usually achieved with con- registries. Int J Cancer 126, 715–​20.
    ventional chemotherapy in patients with stage IV disease, it is De Graaf P, et al. (2012) Guidelines for imaging retinoblastoma: imaging
    principles and MRI standardization. Pediatr Radiol 42, 2–​14.
    usually short-​lived and ultimate survival is infrequent. However,
    De Jong MC, et al. (2014) Trilateral retinoblastoma: a systematic re-
    recent series report that a high proportion of metastatic retino-
    view and meta-​analysis. Lancet Oncol 15, 1157–​67.
    blastoma outside the CNS (stage IVa) may be cured using high-​ Dunkel IJ, et al. (2010) Intensive multimodality therapy for patients
    dose chemotherapy and autologous stem-​cell rescue. Patients with stage 4a metastatic retinoblastoma. Pediatr Blood Cancer
    usually receive four to six cycles of conventional induction 55, 55–​9.
    chemotherapy, followed by high-​dose chemotherapy and autolo- Francis JH, et al. (2015) The classification of vitreous seeds in retino-
    gous stem-​cell rescue. Different preparative regimens have been blastoma and response to intravitreal melphalan. Ophthalmology
    used, but thiotepa-​containing regimens appear to be effective be- 122, 1173–​9.
    cause of its high penetration to the CNS. This modality has also Kleinerman RA, et al. (2005) Risk of new cancers after radiotherapy
    been explored for children with CNS metastasis (stage IVb), but in long-​term survivors of retinoblastoma: an extended follow-​up. J
    very few patients survive, even with this treatment. Recent re- Clin Oncol 23, 2272–​9.
    ports suggest that this strategy may also be effective for trilateral Laurie NA, et al. (2006) Inactivation of the p53 pathway in retino-
    retinoblastoma. blastoma. Nature 444, 61–​6.
    Leal-​Leal CA, et al. (2006) Survival in extra-​orbital metastatic retino-
    blastoma: treatment results. Clin Transl Oncol 8, 39–​44.
    Lumbroso-​ LeRouic L, et al. (2016) Conservative treatment of
    Second malignancies in retinoblastoma survivors
    retinoblastoma: a prospective phase II randomized trial of
    neoadjuvant chemotherapy followed by local treatments and
    The cumulative incidence of second cancers in patients with chemothermotherapy. Eye (Lond) 30, 46–​52.
    germline mutations of the RB1 gene is greatly increased with the use Munier FL, et al. (2008) New developments in external beam radio-
    and dose of radiation therapy and, to a lesser extent, chemotherapy. therapy for retinoblastoma: from lens to normal tissue-​sparing
    This incidence is reported to increase steadily with age, to up to 40–​ techniques. Clin Exp Ophthalmol 36, 78–​89.
    60% at 40–​50 years of age, although more recent studies estimate a Qaddoumi I, et al. (2012) Topotecan and vincristine combination is
    considerably lower risk for patients irradiated with more modern effective against advanced bilateral intraocular retinoblastoma and
    techniques. Patients with non-​hereditary retinoblastoma are not in- has manageable toxicity. Cancer 118, 5663–​70.
    herently at an increased risk. Rushlow DE, et al. (2013) Characterisation of retinoblastomas without
    Almost any neoplasm has been described in survivors of retino- RB1 mutations: genomic, gene expression, and clinical studies.
    blastoma and 60–​70% of the tumours occur in the head and neck Lancet Oncol 14, 327–​34.
    areas. The most common second tumours are osteogenic sarcoma Sastre X, et al. (2009) Proceedings of the consensus meetings from the
    International Retinoblastoma Staging Working Group on the path-
    (arising both inside and outside the irradiation field), soft-​tissue
    ology guidelines for the examination of enucleated eyes and evalu-
    sarcomas, and melanomas. Uterine leiomyosarcomas are more fre-
    ation of prognostic risk factors in retinoblastoma. Arch Pathol Lab
    quent in women surviving hereditary retinoblastoma. Survivors Med 133, 1199–​202.
    of bilateral retinoblastoma are also at risk of developing epithe- Shields CL, et al. (2014) Intra-​ arterial chemotherapy for retino-
    lial cancers late in adulthood, particularly in the lung. There is an blastoma in 70 eyes: outcomes based on the international classifica-
    ongoing debate about the use of whole-​body MRI for screening of tion of retinoblastoma. Ophthalmology 121, 1453–​60.
    second malignancies, but no formal recommendation besides fre- Xu XL, et al. (2009) Retinoblastoma has properties of a cone precursor
    quent medical follow-​up, avoidance of sun exposure, and healthy life tumor and depends upon cone-​specific MDM2 signaling. Cell 137,
    habits is currently issued. 1018–​31.
    Index

    Notes
    Abbreviations used in subentries can be found on pages (ix) to (xiii)
    Tables, figures and boxes are indicated by an italic t, f and b following the page number
    vs. indicates a comparison or differential diagnosis

    AAPCOB (American Academy of acute lymphoblastic leukaemia adamantinomatous AFP see alpha-​fetoprotein (AFP)
    Pediatrics Committee on treatment, 21, 118–​120, 121t craniopharyngiomas, The African Pediatric Fellowship
    Bioethics), 55 allogeneic HSCT, 57–​59 200, 200f Program, 93t
    abdominal surgery, late adverse chemotherapy, 27 adapted-​treatment regimens, age, myelodysplastic syndrome, 134
    effects, 114 reduced-​intensity conditioning LMIC, 94 age-​standardized rate (ASR), 1
    ABL genes, 27 vs. myeloablative adaptive clinical trials, 42 aggressive fibromatosis (AF), 216
    tyrosine kinase inhibitors, 27–​28 conditioning, 56–​57 adaptive radiotherapy, 33–​34 agitation, 87
    ABL tyrosine kinase, 122 total body irradiation, 56 ADCC (antibody-​dependent cellular AIEOP (Associazione Italiana
    abnormal gait, spinal cord acute lymphoid leukaemia, 7f cytotoxicity), 60 Ematologia Oncologia
    tumours, 173f acute megakaryoblastic leukaemia ADC (apparent diffusion coefficient), Pediatrica), 101
    abstracts, structured, 48 (AMKL), 138 germinomas, 198 ALCL see anaplastic large-​cell
    ABVD (doxorubicin, bleomycin, acute monoblastic leukaemia, 126 adjuvant analgesics, 86 lymphoma (ALCL)
    vinblastine, dacarbazine), acute myeloblastic leukaemia, 126 adjuvant chemotherapy, 274–​275 ALCL99 chemotherapy, 152
    158, 160 acute myeloid leukaemia (AML), administration routes, analgesia, 85 ALiCCS (Nordic Adult Life After
    aciclovir, 72 124–​133 adolescents and young adults Childhood Cancer in
    ACP (American College of Physicians) aetiology, 124–​125 (AYA), 97–​105 Scandinavia), 106
    Journals, 48 AYA, 3t, 97 access to care, 98 ALK
    aCPP (atypical choroid plexus challenges, 132 age-​specific psychological anaplastic large-​cell
    papillomas), 203 clinical presentation, 125–​126 aspects, 98–​99 lymphoma, 145
    ACTH see adrenocorticotrophic diagnosis, 126–​128 ALL, 120 familial neuroblastoma, 241–​242
    hormone (ACTH) epidemiology, 2t, 124–​125 biology, 98 sporadic neuroblastoma, 242
    actinomycin B, bilateral minimal residual disease, 129–​130 clinical trials, 101 alkylating agents, 22t
    nephroblastoma, 237 mutations, 124, 125f diagnostic delay, 98 ALL see acute lymphoblastic
    actinomycin D myelodysplastic syndrome vs., diffuse large B-​cell lymphoma, 147 leukaemia (ALL)
    cyclophosphamide, bleomycin 136–​137, 137t epidemiology, 97–​98 ALL-​HSCT-​BFM 2003 Study, 55
    (BCD), 226 pathophysiology, 124 future work, 102 ALL-​I-​BFM HSCT trial
    Ewing sarcoma, 226 perspectives, 132 incidence, 1, 3t, 97 (FORUM), 55
    rhabdomyosarcomas, 212 prognosis, 128, 130–​131, 130f late effects, 101–​102 allogeneic haematopoietic stem-​cell
    acute lymphoblastic leukaemia risk classification, 128, 129 models of care, 99–​101 transplantation, 52–​63, 64–​66
    (ALL), 117–​123 secondary malignant neoplasm outcome, 97–​98 adoptive T-​cell therapy, 65–​66
    adolescents, 120 as, 107 adoptive T-​cell therapy, allogeneic ALL, 57–​59, 118, 120
    aetiology, 117 survival, 4, 5, 5t, 6t, 7f HSCT, 65–​66 AML, 59, 131
    allogeneic HSCT, 64 acute myeloid leukaemia treatment, adrenocorticotrophic clinical studies, 65t
    AYA, 3t, 97, 98 130–​131 hormone (ACTH) conditioning regimens, 55–​57
    BCR/​ABL-​positive disease, 122 allogeneic HSCT, 59, 64 deficiency, 109 cord-​blood matching, 53
    clinical presentation, 117 chemotherapy, 27 radiotherapy side effects, 36 donor registries, 52–​53
    clinical trials, 40, 121t gemtuzumab ozogamicin, 29 Adriamycin, prednisolone, donor selection, 53
    diagnosis, 117–​118 reduced-​intensity conditioning and vincristine (APO) donor type, 52–​53
    epidemiology, 117 vs. myeloablative regimen, 152 graft-​versus-​tumour vs.
    growth hormone deficiency, 109 conditioning, 56–​57 adult-​type fibrosarcoma, 213t graft-​versus-​host
    incidence, 1, 4 total body irradiation, 56 adult-​type non-​rhabdomyosarcoma disease, 64–​65
    infants, 122 acute myeloid leukaemia-​ soft-​tissue sarcomas, 214–​215 haploidentical transplants, 54
    late effects, 122 type chemotherapy, advanced-​care planning, palliative history of, 52
    LMIC, 90, 94 myelodysplasia, 137 care, 84 HLA, 52, 53
    neuropsychological deficits, 80 acute myelomonocytic adverse effects, late see late adverse indications for, 57
    perspectives, 122–​123 leukaemia, 126 effects (LAEs) juvenile myelomonocytic
    prognostic factors, 118, 118t acute promyelocytic leukaemia advocacy, LMIC, 95 leukaemia, 140
    relapse, 122 (APL), 131 aetiology of cancer, 10 long-​term follow-​up, 61–​62
    risk classification, 118 clinical presentation, 126 AF see aggressive fibromatosis (AF) patient registries, 52–​53
    survival, 5, 6 differentiation-​inducing agents, 30 AF (aggressive fibromatosis), 216 post-​transplant
    toxicity, 122 incidence, 124 afatinib, NSCLC, 28 immunotherapy, 60–​61
    278 Index

    allogeneic haematopoietic stem-​cell antibody-​dependent cellular atypical choroid plexus papillomas craniopharyngiomas, 201
    transplantation (cont.) cytotoxicity (ADCC), 60 (aCPP), 203 cyclophosphamide, actinomycin D
    radiotherapy reduction, 55–​56 anti-​CD3 antibodies, 54 atypical teratoid rhabdoid tumours (BCD), 226
    solid tumours, 59–​60 anticonvulsants, 86 (ATRTs), 188, 189, 191 doxorubicin, vinblastine,
    stem-​cell source, 54–​55 anti-​CTLA4 antibodies, 67 imaging, 19 dacarbazine combination
    all-​trans retinoic acid (ATRA), 30 antidepressants, 86 treatment, 194 (ABVD), 158, 160
    AML, 127 antiemetics, 75t, 86t autoimmune disease, Hodgkin’s etoposide, carboplatin (JEB), 257
    alopecia anti-​ganglioside antibodies, 66 lymphoma, 158 etoposide, cisplatin (BEP), 257
    chemotherapy, 23 anti-​GD2 monoclonal autologous stem-​cell transplantation etoposide, doxorubicin,
    late adverse effect as, 113 antibodies, 29–​30 (ASCT), 190 cyclophosphamide,
    alpha-​fetoprotein (AFP) anti-​growth factor antibodies, 67 Avastin see bevacizumab (Avastin) vincristine, prednisolone,
    germ-​cell tumours, 16 anti-​growth factor receptor axitinib, 29 procarbazine (BEACOPP),
    hepatoblastoma, 263 antibodies, 67 AYA see adolescents and young 158, 159f, 160
    liver tumours, 15 anti-​metabolite drugs, 22t adults (AYA) vinblastine, cisplatin (PVT), 257
    alveolar soft-​part sarcoma, 213t anti-​microtubule compounds, 22t azacitidine, 29 Blinatumomab, 29, 67
    American Academy of Pediatrics anti-​myeloid antigen ALL treatment, 120
    Committee on Bioethics antibodies, 66–​67 balanced translocations, AML blinatumumab, 150
    (AAPCOB), 55 anti-​PD-​1 (programmed cell death risk, 129 blindness, radiotherapy side effects, 36
    American College of Physicians (ACP) protein 1) antibodies, 67 basal cell carcinoma (BCC), 36–​37 blood-​cell counts, ALL, 117
    Journals, 48 anti-​PD-​L1 (programmed cell death Bayesian design of clinical Bloom syndrome, 11
    American National Cancer Institute, ligand) antibodies, 67 trials, 42–​43 BMD (bone mineral density), 113
    Surveillance, Epidemiology, anti-​thymocyte globulins, 166 Bayesian optimal interval (BOIN), 40 BMI see body mass index (BMI)
    and End Results (SEER), 142 anti-​tumour-​specific mutation BCCSS (British Childhood Cancer BMJ (British Medical Journal) Clinical
    American Society of Haematology, 91 antibodies, 67 Survivor Study), 106 Evidence, 48
    American Society of Paediatric anxiety, 87 BCD (bleomycin, cyclophosphamide, body mass index (BMI)
    Haematology/​Oncology, 91 AYA, 102 actinomycin D), 226 craniopharyngiomas, 201
    amitriptyline, 86 APL see acute promyelocytic B-​cell non-​Hodgkin’s lymphoma obesity, 25
    AML see acute myeloid leukaemia (APL) clinical presentation/​treatment, body surface area (BSA), 23, 24t
    leukaemia (AML) aplasia, chemotherapy, 23 146–​147, 150 body weight (BW), 25
    AML-​BFM group, 131 APO (Adriamycin, prednisolone, and pathology, 143–​144 BOIN (Bayesian optimal interval), 40
    amputation, osteosarcoma, 224 vincristine) regimen, 152 relapses, 147–​148 bone marrow
    AMSTAR tool, 49 apparent diffusion coefficient (ADC), studies of, 148–​149t allogeneic HSCT, 54–​55
    anaemia, 162 germinomas, 198 BCH (benign cephalic aspiration in AML diagnosis, 126
    Fanconi anaemia see Fanconi appetite loss, 88 histiocytosis), 166 chemotherapy toxicity, 23
    anaemia ARAF gene, 162 BCR/​ABL, ALL, 118, 122 failure, 87
    analgesia, adjuvant, 86 Archives of Diseases of Childhood, 48 BCRP (breast cancer resistance myelodysplastic syndrome,
    anaplastic astrocytomas, 182 arsenic trioxide (ATO), 127 protein), 23 135, 136f
    anaplastic large-​cell ascites, malignant ascites, 87 BEACOPP (bleomycin, unmanipulated bone marrow,
    lymphoma (ALCL) Asia Pacific Leukaemia Lymphoma etoposide, doxorubicin, allogeneic HSCT, 54
    brentuximab vedotin, 29 meetings, 94 cyclophosphamide, bone mineral density (BMD), 113
    chemotherapy, 28 asparaginase vincristine, prednisolone, bone sarcoma
    clinical presentation/​treatment, adverse effects, 24t procarbazine), 158, 159f, 160 incidence, 1
    150, 152–​153, 152t chemotherapy, 21, 22t Beckwith–​Wiedemann syndrome secondary malignant neoplasm
    epidemiology, 142, 143t lymphoblastic lymphoma hepatoblastoma, 262, 263 as, 107
    immunotherapy, 66 treatment, 150 Wilms tumour, 231 survival rate, 5t
    pathology, 145 ASR (age-​standardized rate), 1 benign cephalic histiocytosis bone tumours, 219–​230
    anaplastic lymphoma kinase Associazione Italiana Ematologia (BCH), 166 AYA epidemiology, 3t
    inhibition, TKIs, 28 Oncologia Pediatrica benzodiazepines, 87 biology, 222–​223
    ANBL0032 (COG phase III (AIEOP), 101 BEP (cisplatin, etoposide, biopsy, 221
    study), 30 astrocytomas bleomycin), 257 clinical presentation, 219–​220
    anger, 79 anaplastic, 182 bereavement care, 88 diagnosis, 220–​221
    angiocentric gliomas, 180 AYA epidemiology, 3t Berlin–​Frankfurt–​Munster (BFM) differential diagnosis, 221
    angiosarcoma, 213t desmoplastic infantile protocol epidemiology, 2t, 219
    Ann Arbor revised classification, astrocytoma, 179, 180 ALL treatment, 119 follow-​up, 227–​228, 227t
    Hodgkin’s lymphoma, 158 diffuse astrocytomas, 180 AML prognosis, 130 future work, 228
    antenatal diagnosis, brain epidemiology, 2t lymphoblastic lymphoma, 145 histopathologies, 221–​222
    tumours, 169 imaging, 18f, 19 beta-​lactams, 73 imaging, 16
    anthracenediones, 22t intramedullary spinal cord bevacizumab (Avastin), 29, 67 incidence, 4, 219
    anthracyclines, 22t tumours, 175–​176 CNS toxicity, 114 metastases treatment, 226–​227
    AML treatment, 130 outcome in AYA, 97 BFM see Berlin–​Frankfurt–​Munster predisposition/​genetics, 219
    heart failure as adverse survival, 5t, 6t (BFM) protocol prognosis, 223
    effect, 111 ataxia telangiectasia, 11, 170t bilateral nephroblastoma recurrent disease therapy, 227
    lymphoblastic lymphoma ATO (arsenic trioxide), 127 treatment, 237 remission status, 227–​228
    treatment, 150 ATRA see all-​trans retinoic bilateral retinoblastoma, 270 staging, 221, 223
    toxicity, 23 acid (ATRA) birth, 10 treatment, 223–​227
    anti-​angiogenesis agents, 29 ATRTs see atypical teratoid rhabdoid bisacodyl, 86 brachytherapy, 34
    anti-​B-​cell antigen antibodies, 66 tumours (ATRTs) bispecific T-​cell engaging (BITE) BRAF, 180
    antibiotics ATRX gene antibodies, 29, 67 BRAF/​MEK inhibition
    chemotherapy, 22t osteosarcoma, 222 BL see Burkitt lymphoma (BL) multisystem Langerhans cell
    infection treatment, 73 sporadic neuroblastoma, bleomycin, 24t histiocytosis, 164
    prophylaxis, 72 242–​243 adverse effects, 24t tyrosine kinase inhibitors, 28
    Index 279

    BRAF-​V600E mutations cardiomyopathy, 35 brain tumours, 173 AML diagnosis, 126


    Langerhans cell histiocytosis, 162 cardiovascular disease, AYA, 101 germ-​cell tumours, 199 CHIC (Children’s Hepatic Tumours
    Rosai–​Dorfman disease, 167 care access, AYA, 98 rhabdomyosarcomas, 208 International Collaboration),
    brain damage, 80 care environment, AYA, 99–​100 ceritinib (LDK378), 28 264, 265f
    brain necrosis, 35 Carney complex, CNS tumours, 170t neuroblastoma, 250 Childhood Cancer Survivor Study
    brainstem gliomas, 16, 18 CARs see chimeric antigen cervical neuroblastoma, 244 (CCSS) (USA), 60, 106
    brain tumours, 170–​171, 173 receptors (CARs) cetuximab, 67 Children’s Cancer Group (CCG)
    antenatal diagnosis, 169 CaseHippo, 93t colorectal cancer, 28 cooperative chemotherapy in
    diagnostic delay, 169 cavernomas, intramedullary spinal head and neck cancer, 28 germ-​cell tumours, 257
    growth hormone deficiency, 109 cord tumours, 175 CEUS (contrast-​enhanced high-​grade gliomas, 184
    hydrocephalus, 173 cavernous haemangiomas, 35 ultrasound), 12 Children’s Hepatic Tumours
    incidence, 4 CBL gene, 139 Chang’s classification of embryonal International Collaboration
    neuropsychological deficits, 80 mutations, 139 tumours, 191, 191t (CHIC), 264, 265f
    sudden coma, 171, 173 CBTRUS (Central Brain Tumour CHD (chromodomain helicase DNA-​ Children’s Oncology Group
    symptomatology, 170–​171, Registry of the United binding protein), 182 (COG), 27, 95
    171f, 172f States), 188 checkpoint inhibitors, allogeneic embryonal tumours, 188–​189
    vision loss, 173 CCC (Cochrane Childhood HSCT, 60 hepatoblastoma treatment,
    breast cancer Cancer), 49 Chédiak–​Higashi syndrome 264, 266
    AYA, 98 CCG see Children’s Cancer (CHS), 165 Hodgkin’s lymphoma
    chemotherapy, 28 Group (CCG) chemical exposures, 10 treatment, 160
    secondary malignant neoplasm CCNU (procarbazine, 6-​thioguanine, chemotherapy, 21–​26, 22t, 84b neuroblastoma treatment, 246
    as, 107 dibromodulcitol, adjuvant, 274–​275 Soft-​Tissue Sarcoma
    breast cancer resistance protein lomustine), 184 administration routes, 22 Committee, 208
    (BCRP), 23 CCS (clear-​cell sarcoma), 13, 213t adult-​type non-​ Wilm’s tumour, 232
    breathlessness, 86 CCSK (clear-​cell sarcoma of the rhabdomyosarcoma Wilm’s tumour chemotherapy, 236
    brentuximab vedotin, 29, 66 kidney), 232, 237–​238 soft-​tissue sarcomas, 215, chimeric antigen receptors
    anaplastic large-​cell CCSS (Childhood Cancer Survivor 215f, 215t (CARs), 66
    lymphoma, 152 Study), 60 adverse effects, 24t B-​cell non-​Hodgkin’s
    British Childhood Cancer Survivor CD4 T-​helper cells, 145 anaplastic large-​cell lymphoma, 150
    Study (BCCSS), 106 CD34+ positive/​negative selection, 54 lymphoma, 150 T cells, 67–​68
    BSA (body surface area), 23, 24t CD antigens classification, 22 chondrosarcoma
    buccal pain medication, 85 AML, 128 CNS toxicity, 114 extraskeletal mesenchymal, 213t
    Burkitt lymphoma (BL) diffuse large B-​cell lymphoma, 144 development of, 26–​29 extraskeletal myxoid, 213t
    chemotherapy dose, 23 lymphoblastic lymphoma, 144 difficult situations, 25 outcome in AYA, 97
    epidemiology, 2t, 142 CDKN2A/​CDKN2B genes, 179 drug resistance, 23 survival rate, 4, 5t
    incidence, 3 Center for International Blood and ependymomas, 203 choroid plexus carcinoma (CPC), 203
    LMIC, 94 Marrow Transplant Research epigenetic agents, 29 choroid plexus papilloma (CPP), 203
    outcomes, 147 (CIBMTR), 55 fibroblastic-​myofibroblastic choroid plexus tumours (CPT),
    pathology, 143 Central Brain Tumour Registry of the lesions, 216–​217 203–​204
    survival, 4, 6t United States (CBTRUS), 188 germ-​cell tumours, 257–​259, 258t epidemiology, 2t
    treatment, 147 central nervous system (CNS) high-​dose see high-​dose survival by age, 6t
    busulfan neuroimaging, 188, 189f chemotherapy (HDCT) chromodomain helicase DNA-​
    adverse effects, 24t radiotherapy late adverse infant medulloblastoma, 193 binding protein (CHD), 182
    juvenile myelomonocytic effects, 108 infants, 24–​25, 25f, 25t chromosomal translocations
    leukaemia, 140 radiotherapy side effects, 35 intra-​arterial chemotherapy, 274 Burkitt lymphoma, 143
    myelodysplastic syndrome, 137 central nervous system tumours, 169, intrathecal see intrathecal lymphoblastic lymphoma, 144–​145
    botryoid subtype of 198–​205 chemotherapy chronic kidney disease (CKD), 111
    rhabdomyosarcomas, 206 ALL treatment, 118, 120 medulloblastoma relapse, 195 chronic myeloid leukaemia (CML)
    BW (body weight), 25 AML, 125–​126 molecular aberration allogeneic HSCT, 65
    choroid plexus tumours, 203–​204 correction, 27–​28 AYA epidemiology, 3t
    calcification, CT, 12–​13 craniopharyngiomas, 200–​201 see also tyrosine kinase tyrosine kinase inhibitors, 27
    CancerPointe, 93t diagnostic delay, 169 inhibitors (TKIs) chronic myeloproliferative disease, 2t
    cancer predisposition syndrome, 114 embryonal tumours, 188 nausea and vomiting, 74, 75t CHS (Chédiak–​Higashi
    Candida infection, 74 ependymomas, 201–​203 neuroblastoma, 248, 250–​251 syndrome), 165
    CanTeen (Australia), 102 germ-​cell tumours see germ-​cell obesity, 25 CIBMTR (Center for International
    Canteen (Eire), 102 tumours (GCTs) pharmacokinetics, 26 Blood and Marrow Transplant
    carbamazepine, 86 glial see glial central nervous phase II clinical trials, 40 Research), 55
    carbapenem, 73 system tumours preoperative chemotherapy, 21–​22 CIC-​DUX4/​DIX4L gene fusion,
    carboplatin Langerhans cell histiocytosis, 162 retinoblastoma, 273–​275 222–​223
    adverse effects, 24t late adverse effects, 114 rhabdomyosarcomas, 210, 211, 212 cisplatin
    etoposide, bleomycin (JEB), 257 predisposing factors, 169–​170, 170t signal pathway targeting, 28–​29 adverse effects, 24t
    ifosfamide, etoposide (ICE), 147, secondary neoplasms, 107 skin adverse effects, 113 cytarabine, dexamethasone
    150, 160 survival, 5, 5t, 6t, 8f toxicity, 22–​23 (DHAP), 160
    intramedullary spinal cord see also brain tumours; glial central tumour microenvironment doxorubicin, methotrexate
    tumours, 175 nervous system tumours; antagonists, 29 (MAP), 226
    osteosarcoma, 226 spinal cord tumours Wilm’s tumour (nephroblastoma), etoposide, bleomycin (BEP), 257
    retinoblastoma, 273–​274 central venous catheter 235–​236 etoposide, ifosfamide (PEI),
    cardiac system infections, 74 chemotherapy dosing, 23–​26 199, 257
    Hodgkin’s lymphoma, 160–​161 cephalosporins, 73 body surface area, 23, 24t hepatoblastoma treatment,
    late adverse effects, 111 cerebellar astrocytomas, 17, 17f high-​dose, 23 264, 266
    radiotherapy side effects, 35–​36 cerebrospinal fluid (CSF) chest X-​ray (CXR), 12 osteosarcoma, 225
    280 Index

    cisplatin (cont.) contrast-​enhanced, 13 standard-​risk vinblastine, bleomycin,


    side effects, 35 craniopharyngiomas, 19 medulloblastoma, 193 doxorubicin, combination
    vinblastine, bleomycin (PVT), 257 embryonal tumours, 188 Crizotinib (PF-​02341066), 28, 67 (ABVD), 158, 160
    cixutumumab, 67 germ-​cell tumours, 16, 256 anaplastic large-​cell dactinomycin, 24t
    CKD (chronic kidney disease), 111 germinomas, 198 lymphoma, 152 Damocles syndrome, 114–​115
    cladribine Hodgkin’s lymphoma, 16 neuroblastoma, 250 daunorubicin, 24t
    AML treatment, 130 Langerhans cell histiocytosis, 163 CRT see cranial radiotherapy (CRT) DC (dyskeratosis congenita), 135
    multisystem Langerhans cell medulloblastoma, 17, 18f CSF see cerebrospinal fluid (CSF) DCOG LATER (Dutch Childhood
    histiocytosis, 164 neuroblastoma, 245 CSF1R gene, 27 Oncology Group
    Rosai–​Dorfman disease, 167 non-​Hodgkin’s lymphoma, 16, 145 CSI see craniospinal LATER), 106
    classical phase IIA clinical trials, 40 retinoblastoma, 272 irradiation (CSI) D-​dimers, 165
    clear-​cell sarcoma (CCS), 13, 213t rhabdomyosarcoma, 15, 208 CT see computed tomography (CT) DDS (Denys–​Drash syndrome), 231
    clear-​cell sarcoma of the kidney spinal cord tumours, 174 CTCs (cytotoxic T-​cells), 145 death, fear of, 81
    (CCSK), 232, 237–​238 see also positron emission CTD (Clinical Trial Directives), 44 decision making in palliative
    clinical practice guidelines tomography (PET) CTM (continual reassessment care, 83–​84
    (CPGs), 48 computerized decision-​support and method), phase I clinical decitabine, 29
    clinical target volume (CTV), care pathways, 47–​48 trials, 39 dendritic cell immunotherapy, 68–​69
    radiotherapy, 32 cobimetinib, 181 CTNNB1 gene, 262 denosumab, 228
    Clinical Trial Directives (CTD), 44 conditioning regimens, allogeneic CTR (Clinical Trial Regulation), 44 dental toxicity, 112
    Clinical Trial Regulation (CTR), 44 HSCT, 55–​57 CTV (clinical target volume), Denys–​Drash syndrome (DDS), 231
    clinical trials, 38–​45 confirming stage, clinical trials, 42 radiotherapy, 32 depression, 87
    adaptive approach, 42 congenital mesoblastic nephroma Cure4Kids, 93t, 94 AYA, 102
    background definition, 38–​39 (CMN), 235, 239 cyclophosphamide dermatofibrosarcoma protuberans
    Bayesian designs, 42–​43 consolidation, ALL treatment, 119 adverse effects, 24t (DFP), 213t, 217
    design and implementation, constipation, 75–​76, 86–​87 allogeneic HSCT, 54 desmoid tumour, 216
    38–​42, 43f constitutional mismatch repair bleomycin, actinomycin D desmoplastic infantile astrocytoma/​
    early-​development studies, 39 deficiency, 11 (BCD), 226 ganglioglioma (DIA/​DIG),
    late-​development studies, 39 constitutional mismatch repair bleomycin, etoposide, doxorubicin, 179, 180
    middle-​development studies, 39 syndrome, 170t vincristine, prednisolone, desmoplastic small round cell
    novel strategies, 42–​43 continual reassessment method procarbazine (BEACOPP), tumour, 213t, 216
    phase I, 38, 39–​40 (CTM), phase I clinical 158, 159f, 160 dexamethasone
    phase II, 38, 40 trials, 39 Ewing sarcoma, 226 adverse effects, 24t
    phase III, 38, 40–​42 continuous quality fertility effects, 161 ALL treatment, 119
    phase IV, 38–​39 improvement, 92–​93 haemorrhagic cystitis as adverse cytarabine, cisplatin (DHAP), 160
    practical considerations, 43–​44 contrast-​enhanced computed effect, 112 long-​term follow up, 80
    translational trials, 39 tomography, 13 high-​risk medulloblastoma, 193 spinal cord tumours, 175
    see also randomized controlled contrast-​enhanced ultrasound lymphoblastic lymphoma dextroamphetamine, 87
    trials (RCTs) (CEUS), 12 treatment, 150 fatigue, 87
    CLL-​1, 129–​130 conventional external beam myelodysplastic syndrome, 137 DFP (dermatofibrosarcoma
    clofarabine, 27 radiotherapy, 33 retinoblastoma, 275 protuberans), 213t, 217
    AML treatment, 130 COPDAC (vincristine, rhabdomyosarcomas, 211 DHAP (dexamethasone, cytarabine,
    lymphoblastic lymphoma cyclophosphamide, vincristine, dacarbazine, cisplatin), 160
    treatment, 150 dacarbazine, prednisone), 160 prednisone (COPDAC), 160 diabetes insipidus, Langerhans cell
    CML see chronic myeloid COPP (vincristine, vincristine, procarbazine, histiocytosis, 162, 164
    leukaemia (CML) cyclophosphamide, prednisone (COPP), 160 DIA/​DIG (desmoplastic infantile
    CMN (congenital mesoblastic procarbazine, cyclosporin, 167 astrocytoma/​ganglioglioma),
    nephroma), 235, 239 prednisone), 160 cystic craniopharyngioma, 173 179, 180
    CNS see central nervous cord blood, allogeneic HSCT, 53, 55 cytarabine, 164 diarrhoea, chemotherapy, 23
    system (CNS) corticosteroids adverse effects, 24t dibromodulcitol, 184
    coccyx, germ-​cell tumours, 257 chemotherapy, 21, 22t AML treatment, 130 DICER1 syndrome, 170t
    Cochrane, Archie, 46 haemophagocytic Burkitt lymphoma treatment, 147 diet, infection prevention, 71
    Cochrane Childhood Cancer lymphohistiocytosis, 166 dexamethasone, cisplatin differentiation-​inducing agents, 30
    (CCC), 49 lymphoblastic lymphoma (DHAP), 160 chemotherapy, 22t
    Cochrane Collaboration, 49 treatment, 150 fludarabine (FLA), 27 diffuse astrocytomas, 180
    Cochrane Library, 49 cough, radiotherapy, 35 Langerhans cell histiocytosis, 164 diffuse intrinsic pontine gliomas
    COG see Children’s Oncology Cowden syndrome, 170t lymphoblastic lymphoma (DIPGs), 173, 182
    Group (COG) CPC (choroid plexus carcinoma), 203 treatment, 150 diffuse large B-​cell lymphoma
    cognitive impairments, 108 CPGs (clinical practice cytochrome P450 isoenzymes, 26 (DLBCL), 146
    COG phase III study guidelines), 48 cytogenetics adolescents, 147
    (ANBL0032), 30 CPP (choroid plexus papilloma), 203 AML, 128 biology, 144
    colorectal cancer chemotherapy, 28 CPT see choroid plexus myelodysplastic syndrome, epidemiology, 142
    communication tumours (CPT) 135–​136 diffuse midline gliomas, 182
    dying children, 81 cranial radiotherapy (CRT), 108 myeloid leukaemia and Down diffusion-​weighted imaging
    palliative care, 83–​84 bone mineral density, 113 syndrome, 138 (DWI), 198
    comparative genomic hybridization, craniofacial toxicity, 112 cytomegalovirus infection, 74 dinutuximab, 30, 66
    AML, 128 craniopharyngiomas, 171, 200–​201 cytosine arabinoside, 167 DIPGs (diffuse intrinsic pontine
    computed tomography (CT), 12–​13 adamantinomatous, 200, 200f cytotoxic T-​cells (CTCs), 145 gliomas), 173, 182
    astrocytomas, 19 cystic craniopharyngioma, 173 direct evidence-​based
    bone tumour staging, 221 imaging, 19, 19f dacarbazine medicine, 49–​50
    brain tumours, 173 craniospinal irradiation (CSI) cyclophosphamide, vincristine, disappointment, psychosocial
    central nervous system, 16 embryonal tumours, 191 prednisone (COPDAC), 160 care, 79
    Index 281

    DLCBL see diffuse large B-​cell prognosis, 191–​192 carboplatin, bleomycin (JEB), 257 extramedullary spinal cord
    lymphoma (DLBCL) staging, 191, 192f carboplatin, ifosfamide (ICE), 147, tumours, 176
    DLI see donor lymphocyte treatment, 192–​194 150, 160 future work, 228
    infusion (DLI) treatment at relapse, 194–​195 cisplatin, bleomycin (BEP), 257 histopathology, 221–​222
    DNTs (dysembryoplastic EMT (epithelial-​mesenchymal Ewing sarcoma, 226 imaging, 16, 221
    neuroepithelial tumours), transition), 242 haemophagocytic metastases treatment, 227
    179–​180 endocrine system lymphohistiocytosis, 166 recurrent disease therapy, 227
    docusate sodium, 86 late adverse effects, 108–​109 ifosfamide, cisplatin (PEI), remission status, 227–​228
    donor lymphocyte infusion (DLI) radiotherapy side effects, 36 199, 257 staging, 221
    allogeneic HSCT, 64, 65–​66 end-​of-​life issues retinoblastoma, 273–​274 surgery, 177
    HSCT in ALL, 58–​59 care planning, 84 uncommon histiocytoses, 167 survival, 4, 5t, 6, 6t, 9f
    juvenile myelomonocytic psychosocial care, 80–​81 ETV (endoscopic third treatment, 223
    leukaemia, 140 endoscopic third ventriculostomy ventriculostomy), 173 EWOG-​MDS study, juvenile
    donors (ETV), 173 ETV6/​RUNX1 translocation, 117 myelomonocytic
    registries, 52–​53, 58 enolase 2, 245 EU Paediatric Regulation, clinical leukaemia, 140
    type in allogeneic HSCT, 52–​53 Enpr-​EMA (European Network trials, 38 EWS gene, 222
    dose intensity, chemotherapy, 23 of Paediatric research at EUROCARE database, 4 external-​beam radiotherapy (EBRT)
    dose volume histograms (DVHs), 32 the European Medicines EuroNet Paediatric Hodgkin neuroblastoma, 247
    Down syndrome, myeloid leukaemia Agency), 26–​27 Lymphoma Group, retinoblastoma, 273
    of, 129, 137–​139, 138f ENTYAC (European Network for 159–​160, 160b extramedullary spinal cord
    doxorubicin Teenagers and Young Adults EuroNet PHL-​C2 Hodgkin’s tumours, 175
    adverse effects, 24t with Cancer), 99 lymphoma trial, 160 extraosseous sarcoma, 5t
    bleomycin, etoposide, environmental factors, 10 European Intergroup for Childhood extraskeletal mesenchymal
    cyclophosphamide, infection prevention, 71 non-​Hodgkin’s Lymphoma chondrosarcoma, 213t
    vincristine, prednisolone, enzymatic staining, AML diagnosis, (EICNHL), 142 extraskeletal myxoid
    procarbazine (BEACOPP), 126, 126t European International Society chondrosarcoma, 213t
    158, 159f, 160 :EP (lung resistance-​related protein), 23 of Paediatric Oncology ex vivo T-​cell depletion, allogeneic
    bleomycin, vinblastine, ependymomas, 201–​203, 202t European Neuroblastoma HSCT, 54
    dacarbazine combination AYA epidemiology, 3t Network (SIOPEN), 30, 246 eyes, radiotherapy, 36
    (ABVD), 158, 160 epidemiology, 2t European Network for Teenagers and
    Ewing sarcoma, 226 imaging, 16, 19 Young Adults with Cancer FAB classification see French–​
    methotrexate, cisplatin (MAP), 226 intramedullary spinal cord (ENTYAC), 99 American–​British (FAB)
    osteosarcoma, 225 tumours, 175 European Network of Paediatric classification
    rhabdomyosarcomas, 211 survival by age, 5, 6t research at the European FABLMB96 trial, Burkitt lymphoma
    Dutch Childhood Oncology Group epidemiology, 1–​11 Medicines Agency treatment, 147
    LATER (DCOG LATER), 106 classification, 1 (Enpr-​EMA), 26–​27 FAM73A-​BRAF gene fusions, 162
    DVJs (dose volume histograms), 32 incidence, 1, 2t, 3–​4, 3t European paediatric Soft-​Tissue familial adenomatous
    dysembryoplastic neuroepithelial survival rates see survival rates Sarcoma Study Group polyposis (FAP)
    tumours (DNTs), 179–​180 epidermal growth factor receptors, 28 (EpSSW), 209 CNS tumours, 170t
    dysgerminomas, 254 epipodophyllotoxins, 22t event-​free survival (EFS) hepatoblastoma, 262
    dyskeratosis congenita (DC), 135 epigenetic chemotherapy, 29 ALL treatment, 118 familial Hodgkin’s lymphoma, 158
    dyspnoea, radiotherapy, 35 EpiInfo, 92 phase III clinical trials, 41 familial neuroblastoma, 241–​242
    epirubicin, 24t everolimus, 29 family-​centred model, clinical
    early-​development studies, clinical epithelial-​mesenchymal transition EvidenceAlerts, 47 care, 99
    trials, 39 (EMT), 242 evidence-​based medicine (EBM), family support, palliative care, 88
    EBM see evidence-​based epithelioid sarcoma, 213t 46–​51, 47f Fanconi anaemia, 135
    medicine (EBM) epratuzumab, 29 acquisitions, 47 cancer risk, 11
    EBRT see external-​beam EpSSG (Soft Tissue Sarcoma Study appraisals, 49 CNS tumours, 170t
    radiotherapy (EBRT) Group), 101 computerized decision-​support FAP see familial adenomatous
    EBV see Epstein–​Barr virus (EBV) EpSSW (European paediatric and care pathways, 47–​48 polyposis (FAP)
    infection Soft-​Tissue Sarcoma Study criticisms of, 50 Farber, Sidney, 38
    echocardiography, Hodgkin’s Group), 209 definition, 46 fatigue, 87
    lymphoma, 160–​161 Epstein–​Barr virus (EBV) evidence-​based journals, AYA, 102
    Effectiveness and Efficiency: Random infection, 10 structured abstracts, 48 radiotherapy side effects, 34
    Reflections on Health Services Hodgkin’s lymphoma, 157–​158 future work, 51 FDA (Federal Drug Agency), 26
    (Cochrane), 46 juvenile myelomonocytic guideline development, 50 fear of death, 81
    EFS see event-​free survival (EFS) leukaemia vs., 139–​140 history, 46 febrile neutropenia (FN), 71
    EICNHL (European Intergroup for leukodepletion, 74 journals, 48 Federal Drug Agency (FDA), 26
    Childhood non-​Hodgkin’s non-​Hodgkin’s lymphoma, 142 question asking, 47 fellow patients, psychosocial
    Lymphoma), 142 Erdheim–​Chester disease, 162 result application, 49–​50 care, 79
    embryonal carcinomas (EC), 198 erlotinib roles of, 46 fentanyl, 85
    embryonal tumours, 188–​197 NSCLC, 28 studies, 49 ferritin, 245
    AYA, 3t pancreatic cancer, 28 systematic cancer reviews, 48–​49 fertility
    biology, 190–​191 ES see Ewing sarcoma (ES) Evidence-​Based Medicine (EBM AYA, 102
    diagnosis, 188–​189 etoposide (VP-​16) Online), 48 Hodgkin’s lymphoma, 161
    epidemiology, 2t, 188 adverse effects, 24t Ewing sarcoma (ES) radiotherapy side effects, 36
    follow-​up, 194 bleomycin, doxorubicin, AYA, 3t, 97 fever without neutropenia, 74
    future work, 195 cyclophosphamide, biology, 222–​223, 222f, 222t fibroblastic-​myofibroblastic lesions,
    intracranial, 6t vincristine, prednisolone, chemotherapy, 226 216–​217
    intraspinal, 6t procarbazine (BEACOPP), clinical presentation, 219 fibromatosis, aggressive
    pathology, 189–​190, 189f 158, 159f, 160 epidemiology, 2t, 219 fibromatosis, 216
    282 Index

    fibrosarcoma GATA2 gene, 135 glial central nervous system tumours, haematological malignancies
    adult-​type, 213t GBM (glioblastoma multiforme), 184 179–​187 AYA, 97
    epidemiology, 2t GCSF see granulocyte colony-​ high-​grade see high-​grade chimeric antigen receptor clinical
    imaging, 16 stimulating factor (GCSF) gliomas (HGGs) trials, 68
    infantile fibrosarcoma, 213t, 216 G-​CSF (granulocyte colony-​ low-​grade see low-​grade monoclonal antibodies, 29–​30
    outcome in AYA, 97 stimulating factor), 131–​132 gliomas (LGGs) haematopoietic stem-​cell
    survival rate, 5t GCTs see germ-​cell tumours (GCTs) glioblastoma multiforme (GBM), 184 transplantation (HSCT)
    finances, treatment in LMIC, 95 gefitinib, 28 glioblastomas, 182 allogeneic see allogeneic
    FISH see fluorescent in situ gemcitabine, 27 gliomas haematopoietic stem-​cell
    hybridization (FISH) prednisone, ifosfamide, angiocentric gliomas, 180 transplantation
    5S pyramid, 47, 48f, 49t vinorelbine (IGEV), 160 diffuse intrinsic pontine gliomas, juvenile myelomonocytic
    FLA (fludarabine/​cytarabine), 27 gemtuzumab ozogamicin (GO), 29, 173, 182 leukaemia, 140
    FLASH gene deletion, 145 66–​67, 132 diffuse midline gliomas, 182 late adverse effects, 113
    flow cytometry gender high-​grade see high-​grade myelodysplastic syndrome, 137
    ALL, 117–​118 AVA cancer outcome, 98 gliomas (HGGs) haemophagocytic
    AML, 127–​128 cancer epidemiology, 1 low-​grade glioma see low-​grade lymphohistiocytosis (HLH),
    haemophagocytic myelodysplastic syndrome, 134 gliomas (LGGs) 164–​166
    lymphohistiocytosis, 165 gene expression profiling, 144 optic pathway see optic pathway differential diagnosis, 164, 165b
    fluconazole, 72 gene modified donor lymphocyte gliomas genetics, 165
    fludarabine infusion, 66 global networks, LMIC, 94 pathogenesis, 165
    adverse effects, 24t gene replacement therapy, 166 Global Neuroblastoma Network, 94 therapy, 166
    AML treatment, 130 genetic epidemiology, 11 glomerular filtration rate (GFR), 25 haemopoietic stem-​cell rescue
    cytarabine (FLA), 27 genito-​urinary tract tumours, 3t glomerular toxicity, 111 (HSCR), 211
    fluorescent in situ hybridization (FISH) genome-​wide association glucuronidation pathways, 26 haemorrhages, bone marrow
    AML, 128 (GWAS), 242 P-​glycoprotein, 23 failure, 87
    anaplastic large-​cell genome-​wide screening, 123 glycopyrronium, 86 haemorrhagic cystitis (HC), 112
    lymphoma, 145 German BFM studies, 147 GO (gemtuzumab ozogamicin), haploidentical transplants, allogeneic
    Ewing sarcoma, 222 German Society for Paediatric 29, 132 HSCT, 54, 55
    18fluorodeoxyglucose (FDG), 13 Oncology and Haematology gonadal dysfunction, 109–​110 HART (hyperfractionated accelerated
    fluorodeoxyglucose-​positron (GPOH), 246 gonadoblastoma, 254 radiotherapy), 191
    emission tomography German Soft-​Tissue Sarcoma gonadotrophins, 109, 110 Hashimoto’s thyroiditis, 110
    (FDG-​PET) Cooperative Group, 209 government spending, LMICs, 90, 92f HbF (foetal haemoglobin), 135
    Hodgkin’s lymphoma, 158 germ-​cell tumours (GCTs), 198–​200, GPC3 gene, 262 HC (haemorrhagic cystitis), 112
    non-​Hodgkin’s lymphoma, 145 253–​261 GPOH (German Society for HCC (hepatocellular carcinoma),
    5-​fluorouracil, 264 AYA, 3t, 97 Paediatric Oncology and 267–​268
    FN (febrile neutropenia), 71 biology, 199, 255f Haematology), 246 HCG see human chorionic
    focal therapies, retinoblastoma, 273 classification, 254t GRADE evidence to decision (EtD) gonadotropin (HCG)
    foetal haemoglobin (HbF), 135 clinical presentation, 255–​256 framework, 50 HDACx see histone deacetylases
    follow-​up care, late adverse effects, 115 current treatment strategies, Grading of Recommendations, (HDACx)
    French–​American–​British (FAB) 199–​200 Assessment, Development, HDCT see high-​dose
    classification diagnosis, 198–​199, 255–​256 and Evaluation chemotherapy (HDCT)
    ALL, 117 distribution, 254f (GRADE), 46, 50 head and neck cancer
    AML, 126, 127t embryology, 253–​254 graft-​versus-​host disease (GVHD) B-​cell lymphoma, 146
    myelodysplastic syndrome, 134 epidemiology, 2t, 253–​255 allogeneic HSCT, 64–​65 chemotherapy, 28
    French (Lymphomes Malins B) follow-​up, 259 HSCT, 113 HeadSmart: Early Diagnosis of Brain
    (LMB) studies, 147 future work, 259–​260 stem cell source, 55 Tumours campaign, 171
    FTL3 gene, 129 histology, 253–​254, 254t graft-​versus-​tumour effect (GvT), heart failure, 111
    fungal prophylaxis, 72 imaging, 15–​16 allogeneic HSCT, 60, 64–​65 hepatic system
    fusion status, incidence, 4 granulocyte colony-​stimulating late adverse effects, 112
    rhabdomyosarcomas, 209 intracranial, 171 factor (GCSF), 131–​132 radiotherapy side effects, 36
    investigations, 256 Ewing sarcoma, 226 see also liver tumours
    gabapentin, 86 molecular biology, 255 infection prevention, 72 hepatitis B, 10
    gadolinium-​enhanced magnetic predisposing factors, 253 osteosarcoma, 226 hepatocellular carcinoma, 267
    resonance imaging, 174, 175f survival, 5t, 8f grieving, 88 vaccination, 73
    gait, abnormal, 173f tumour markers, 254–​255 Griscelli syndrome type II (GSII), 165 hepatitis C, 267
    γδ T-​lymphocytes, 60 germ-​cell tumours treatment, gross tumour volume (GTV), 32 hepatoblastoma, 262–​264, 266–​267
    ganglioglioma (GG), 179, 180 256–​259 Group Franco–​Africain d’Oncologie biology, 262
    desmoplastic infantile chemotherapy, 257–​259, 258t Pédiatrique (GFAOP), 93t, 95 clinical presentation, 263
    ganglioglioma, 179, 180 relapse treatment, 259 growth hormone (GH) diagnosis, 263
    intramedullary spinal cord surgery, 256–​257 deficiency, 109 epidemiology, 2t, 262–​263
    tumours, 175 teratoma therapy, 259 GTV (gross tumour volume), 32 histology, 263–​264, 263b
    ganglioneuroblastoma, 245 germinomas, 198, 254 Guyatt, Gordon, 46 orthotopic liver transplant,
    epidemiology, 2t germinomatous tumours, 254 GVHD see graft-​versus-​host 266–​267
    survival by age, 6t germline mutations, 11 disease (GVHD) treatment strategies, 264,
    survival rate, 5t myelodysplastic syndrome, 135 GvT (graft-​versus-​tumour effect), 266–​267
    gastrointestinal system GFAOP (Group Franco–​Africain allogeneic HSCT, 60, 64–​65 hepatocellular carcinoma (HCC),
    AYA tumours, 3t d’Oncologie Pédiatrique), GWAS (genome-​wide 267–​268
    complications, 87 93t, 95 association), 242 hepatomegaly, 162
    Langerhans cell histiocytosis, 162 GH (growth hormone) hereditary retinoblastoma, 170t
    late adverse effects, 112 deficiency, 109 haemangioblastomas, 175 herpes simple virus (HSV)
    GATA1 gene mutation, 138 glembatumumab vedotin, 228 haemangiomas, cavernous, 35 infection, 74
    Index 283

    herpes simplex virus-​derived homovanillic acid (HMA), 245 IGHG (International Guideline indeterminate cell histiocytosis, 162
    thymidine kinase Hospice training, 93t harmonization Group), 108 inducible caspase 9 (iCasp9), 66
    (HSV-​tk), 66 hospital environment, infection IGHG (International Guideline induction
    herpes virus-​6 infection, 139–​140 prevention, 71 Harmonization Group), 51 ALL treatment, 119
    heterogeneity design of clinical HOXA gene, 118 IGRT (image-​guided AML treatment, 130
    trials, 42 H/​RS cells, Hodgkin’s radiotherapy), 33–​34 infantile fibrosarcoma (IFS), 213t, 216
    HGGs see high-​grade lymphoma, 157 IKZF3 gene, 118 infants
    gliomas (HGGs) HSCT see haematopoietic stem-​cell image-​defined risk factors, ALL, 122
    HH8, 10 transplantation (HSCT) neuroblastoma, 247t chemotherapy, 24–​25, 25f, 25t
    high-​dose chemotherapy (HDCT), 23 HSV (herpes simple virus) image-​guided radiotherapy embryonal tumour treatment,
    group 3 medulloblastoma, 190 infection, 74 (IGRT), 33–​34 193–​194
    high-​grade gliomas, 184 human chorionic imaging, 12–​20 HSCT in ALL, 58
    infant medulloblastoma, 193 gonadotropin (HCG) bone tumours, 220–​221 medulloblastoma nonsurgical
    myeloablative, 23 germ-​cell tumours, 16, 199, 254 Hodgkin’s lymphoma, 158 treatment, 193
    rhabdomyosarcomas, 211 hepatoblastoma, 263 tumours outside nervous system, infections, 10, 71–​74
    high-​grade gliomas (HGGs), 181–​185 human leukocyte antigen (HLA) 13, 15–​16 AML side effects, 131
    biology, 182–​183, 183f, 184f allogeneic HSCT, 52, 53 tumours within nervous central venous catheters, 74
    classification, 181–​182 compatibility, 53 system, 16–​19 fever without neutropenia, 74
    future work, 184–​185 human papilloma virus infection, 10 see also computed tomography prevention, 71–​73
    pathology, 181–​182 hydrocephalus (CT); interventional treatment, 73–​74
    treatment, 183–​184 brain tumours, 173 radiography (IR); magnetic inflammatory myofibroblastic
    high-​income countries late adverse effects, 108 resonance imaging tumours (IMT), 28, 217
    LMIC vs., 90, 91f, 91t hyoscine butylbromide, 87 (MRI); nuclear medicine; influenza vaccination, 73
    outcome in AYA, 97 hyperfractionated accelerated radiography INFORM, 97
    high mobility group box 1 protein radiotherapy (HART), 191 IMAT (intensity-​modulated arc infratentorial tumours, imaging, 17
    (HMGB1), 152 hyperhydration, 76 therapy), 108 inherited bone-​marrow failure
    high-​risk medulloblastoma, 193 hyperleukocytosis treatment, 132 imatinib, 27 disorders (IBMFDs), 134–​135
    Hippel–​Linday complex, 175, 176f hyperthyroidism, 110 123I-​MIBG (metaiodobenzyl­ Innovative Therapies for Children
    histiocytoses, 162–​168 hyperuricaemia, 76 guanidine), 13 with Cancer (ITCC), 27
    uncommon disorders, 166–​167 hypophosphataemia rickets, 111 neuroblastoma, 245, 246f inotuzumab ozogamicin, 120
    histology, rhabdomyosarcomas, 209 hypoplasia, thoracic irradiation, 35 imipenem, 73 INRG see International
    histone deacetylases (HDACx), 29 hypothalamic immunizations, infection prevention, Neuroblastoma Risk
    neuroblastoma craniopharyngiomas, 201 72–​73, 72t Group (INRG)
    chemotherapy, 250 hypothalamic–​hypophyseal axis, 36 immunodeficiency, 79, 158 INSS (International Neuroblastoma
    HIV infection, 10 hypothalamic obesity non-​Hodgkin’s lymphoma, 142 Staging System), 244
    HLA see human leukocyte syndrome, 109 immunological dysfunction, 113 Institut national du Cancer
    antigen (HLA) hypothalamic–​pituitary immunophenotyping (INCa), 100
    HLH see haemophagocytic dysfunction, 109 ALL, 117–​118 Institute of Medicine, clinical practice
    lymphohistiocytosis (HLH) AML, 126, 127–​128 guideline definition, 48
    HMA (homovanillic acid), 245 IBMFDs (inherited bone-​marrow AML minimal residual disease, 129 insulin growth factor receptors
    HMGB1 (high mobility group box 1 failure disorders), 134–​135 Burkitt lymphoma, 143 (IGFRs)
    protein), 152 ibuprofen, 85 lymphoblastic lymphoma, 144 chemotherapy and, 28–​29
    Hodgkin’s lymphoma, 157–​161 iCasp9 (inducible caspase 9), 66 myelodysplastic syndrome targeting of, 28–​29
    AYA, 3t, 97 ICCC-​3 (International Classification (MDS), 136 type II, 262
    biology, 157 of Childhood Cancer, Third non-​Hodgkin’s disorder, 143t insulin tolerance test (ITT), 109
    cardiovascular events, 160–​161 Edition), 1 immunotherapy, 29–​30, 64–​70 INT-​0098 trial, 268
    clinical management, 158 ICD-​O (International Classification of ALL treatment, 120, 121t intensity-​modulated arc therapy
    epidemiology, 2t, 157 Diseases for Oncology), 1 anti-​B-​cell antigen antibodies, 66 (IMAT), 108
    Epstein–​Barr virus infection, ICE (ifosfamide, carboplatin, anti-​ganglioside antibodies, 66 intensity-​modulated radiotherapy
    157–​158 etoposide), 147, 150, 160 bispecific T-​cell engager (IMRT), 33, 108, 183
    familial Hodgkin’s ICP (intracranial pressure), 175 antibodies, 67 retinoblastoma, 273
    lymphoma, 158 idarubicin, 24t chimeric antigen receptor T interferon-​α
    imaging, 16, 158 IDH1/​2 genes, 183 cells, 67–​68 craniopharyngiomas, 201
    incidence, 1 ifosfamide dendritic cell uncommon histiocytoses, 167
    LMIC, 94 adverse effects, 24t, 112 immunotherapy, 68–​69 Intergroup Rhabdomyosarcoma
    nodular lymphocyte predominant carboplatin, etoposide (ICE), 147, immune checkpoint blockade, 67 Study (IRS) Group, 208–​209
    see nodular lymphocyte 150, 160 monoclonal-​based International Agency for Research on
    predominant Hodgkin’s etoposide, cisplatin (PEI), antibodies, 66–​67 Cancer, 262
    lymphoma (NLPHL) 199, 257 natural-​killer-​cell International BFW Study Group, 60
    outcomes, 97, 160–​161 Ewing sarcoma, 226 immunotherapy, 68 International Childhood Liver
    pathology, 157 haemorrhagic cystitis, 112 tumour vaccines, 68–​69 Tumour Strategy Group
    predisposition to, 158 osteosarcoma, 225–​226 see also allogeneic haematopoietic (SIOPEL)
    radiotherapy side effects, 36 prednisone, gemcitabine, stem-​cell transplantation; hepatoblastoma, 264, 266t, 267t
    secondary malignancies, 160 vinorelbine (IGEV), 160 haematopoietic stem-​cell hepatocellular carcinoma, 268
    semi-​quantitative Deauville five-​ retinoblastoma, 275 transplantation (HSCT) infant medulloblastoma, 193
    point system, 159b rhabdomyosarcomas, 211, 212 IMRT see intensity-​modulated phase III clinical trials, 41
    survival, 4, 5t, 6, 8f IFS (infantile fibrosarcoma), radiotherapy (IMRT) pre-​treatment extent (PRETEXT)
    Hodgkin’s lymphoma treatment, 213t, 216 IMT (inflammatory myofibroblastic staging system, 15
    159–​160 IGEV (gemcitabine, prednisone, tumours), 28, 217 International Classification of
    brentuximab vedotin, 29 ifosfamide, vinorelbine), 160 INCa (Instituit national du Childhood Cancer, Third
    relapse treatment, 160 IGF1R antibodies, 228 Cancer), 100 Edition (ICCC-​3), 1
    284 Index

    International Classification of Diseases JAK3 gene, 140 psychosocial outcomes, 114–​115 CNS tumours, 170t
    for Oncology (ICD-​O), 1 JAMA Users Guides, 49 pulmonary system, 111 high-​grade gliomas, 181
    International Diabetes Japanese Neuroblastoma Study renal system, 111–​112 liposarcoma, 213t
    Federation, 110 Group (JNBSG), 246 secondary malignant neoplasms, liposomal muramyl tripeptide
    International Guideline JEB (carboplatin, etoposide, 105, 107–​108 phenol ethanolamine
    Harmonization Group bleomycin), 257 skeleton, 113 (L-​MTP-​PE), 226
    (IGHG), 51, 107–​108 JKZ2 gene, 118 skin, 113 liver tumours, 262–​269
    International Network for Cancer Journal of Epidemiology, 50 surgery, 114 epidemiology, 2t
    Treatment and Research, 91 juvenile myelomonocytic leukaemia testicular dysfunction, 110 hepatoblastoma see
    International Neuroblastoma Risk (JMML), 139–​140, 166 thyroid gland, 110 hepatoblastoma
    Group (INRG), 245–​246, chemotherapy, 28 visual system, 112 hepatocellular carcinoma,
    246t, 247t juvenile xanthogranuloma (JXG), 166 late-​development studies, clinical 267–​268
    Staging System, 248t trials, 39 imaging, 15, 15f
    Treatment Classification Kaposi sarcoma, LMIC, 94 late effects of therapy see late adverse incidence, 4
    System, 249t KDM (lysine-​specific effects (LAEs) LL see lymphoblastic lymphoma (LL)
    International Neuroblastoma Staging demethylase), 182 late mortality, 106–​107 LMB (Lymphomes Malins B/​French)
    System, 246t ketamine, 86 LATER, secondary neoplasms, studies, 147
    International Neuroblastoma Staging KIAA1549:BRAF fusion genes, 180 107–​108 LMICs see low-​and middle-​income
    System (INSS), 244 kidney cancer see renal tumours Lawrence formula, obesity in countries (LMICs)
    International Paediatric NHL KMT (lysine-​specific chemotherapy, 24t, 25 L-​MTP-​PE (liposomal muramyl
    Staging System Classification methyltransferase), 182 laxatives, osmotic, 76, 86 tripeptide phenol
    (IPNHLSS), 146, 146b KRAS gene, 139 LCH see Langerhans cell ethanolamine), 226
    International Paediatric Oncology histiocytosis (LCH) lomustine, 184
    guidelines in Supportive care lacrimal gland dysfunction, 112 LDH see lactate long-​distance polymerase chain
    Network (iPOG), 51 lactate dehydrogenase (LDH), dehydrogenase (LDH) reaction (LD-​PCR), Burkitt
    International Prognostic Scoring 254–​255 LDK378 see ceritinib (LDK378) lymphoma, 143–​144
    System (IPSS), 137 haemophagocytic LEA (Leucemie Enfants et long-​term follow-​up
    International Society of Paediatric lymphohistiocytosis, 164–​165 Adolescents), 60 allogeneic HSCT, 61–​62
    Oncology (SIOP), 91, 95 neuroblastoma, 245 learning stage, clinical trials, 42 late adverse effects, 106, 107f
    Asia Continental Branch, 93t lactulose, 86 leiomyosarcoma, 213t psychosocial care, 80
    germ-​cell tumours, 199 LAEs see late adverse effects (LAEs) lenvatinib, 29 lorlatinib (PF-​06463922), 250
    phase III clinical trials, 41, 41f LAIP (leukaemia-​associated Leucemie Enfants et Adolescents lorazepam, 87
    renal tumours, 233b immunophenotype), 129 (LEA), 60 loss of heterozygosity (LOS), 145
    rhabdomyosarcomas, 210 laminectomy, 176–​177 leucocyte count, ALL, 117 low-​and middle-​income countries
    Wilm’s tumour, 232 Langerhans cell histiocytosis (LCH), leucoencephalopathy, 108 (LMICs), 90–​96
    Wilm’s tumour chemotherapy, 28, 162–​164 leukaemia-​associated advocacy, 95
    235–​236 clinical classification, 163, 163t immunophenotype high-​income countries vs., 90,
    interval-​based design, phase I clinical clinical manifestations, 162 (LAIP), 129 91f, 91t
    trials, 40 evaluation, 163, 163t ALL see acute lymphoblastic paediatric cancer units, 90–​93
    interventional radiography (IR), 13 extramedullary spinal cord leukaemia (ALL) protocol-​based care, 93–​94
    intestinal obstructions, 87 tumours, 176 leukaemias treatment challenges, 94–​95
    intra-​arterial chemotherapy, 274 multisystem, 163, 164 acute lymphoid leukaemia, 7f lower urinary tract, 112
    intracranial embryonal tumours, 6t permanent consequences, 164 acute megakaryoblastic low-​grade glial-​neuronal tumours
    intracranial germ-​cell tumours, 171 treatment, 163–​164 leukaemia, 138 (LGGNTs), 179
    intracranial pressure (ICP), 175 lapatinib, 28 acute monoblastic leukaemia, 126 mutations, 180
    intradural tumours, 176 large B-​cell lymphoma with IRF4 acute myeloblastic leukaemia, 126 low-​grade gliomas (LGGs), 94, 171,
    intramedullary spinal cord rearrangements, 144 acute myelomonocytic 179–​181
    tumours, 176 late adverse effects (LAEs), 106–​116 leukaemia, 126 biology, 180
    intraspinal embryonal tumours, 6t auditory system, 112 AML see acute myeloid classification, 179–​180
    intrathecal chemotherapy, 26 cardiac system, 111 leukaemia (AML) future work, 181
    ALL treatment, 120 craniofacial toxicity, 112 APL see acute promyelocytic mutations, 180
    intravenous immunoglobulin dental toxicity, 112 leukaemia (APL) pathology, 179–​180
    (IVIG), 164 endocrine system, 108–​109 AYA epidemiology, 3, 3t treatment, 180–​181
    intravenous infusion, pain epidemiology, 106 CML see chronic myeloid lumbar puncture, AML
    medication, 85 follow-​up care, 115 leukaemia (CML) diagnosis, 126
    invasive procedures, infection future work, 115 incidence, 1, 3 lung carcinoma, 97
    prevention, 71–​72 gastrointestinal system, 112 JMML see juvenile lung resistance-​related protein
    in vivo T-​cell depletion, allogeneic gonadal dysfunction, 109–​110 myelomonocytic (LRP), 23
    HSCT, 54 haematopoietic stem-​cell leukaemia (JMML) lymph nodes,
    ionizing radiation, 10 transplantation, 113 lymphoid leukaemia see lymphoid rhabdomyosarcomas, 209
    ipilimumab, 30 hepatic system, 112 leukaemia lymphoblastic lymphoma (LL), 143t
    IPNHLSS (International Paediatric hypothalamic–​pituitary myeloid leukaemia and Down clinical presentation/​treatment,
    NHL Staging System dysfunction, 109 syndrome, 137–​139, 138f 150, 151t
    Classification), 146, 146b immunological dysfunction, 113 leukocytopenia anaemia, 162 epidemiology, 142
    IR (interventional radiography), 13 late mortality, 106–​107 levomepromazine, 75 pathology, 144–​145
    irinotecan, 24t long-​term follow-​up, 106, 107f LGGNTs see low-​grade glial-​neuronal lymphoid leukaemia
    IRS (Intergroup Rhabdomyosarcoma lower urinary tract, 112 tumours (LGGNTs) epidemiology, 2t
    Study) Group, 208–​209 metabolic syndrome, 110 Li–​Fraumeni syndrome survival by age, 6t
    ITT (insulin tolerance test), 109 neurological toxicity, 108 bone tumours, 219 survival rate, 4, 5t
    IVIG (intravenous neuropsychological toxicity, 108 cancer risk, 11 lymphomas
    immunoglobulin), 164 ovarian dysfunction, 109–​110 choroid plexus tumours, 204 adolescent epidemiology, 3–​4, 3t
    Index 285

    ALCL see anaplastic large-​cell AML treatment, 131 ALL treatment, 119–​120 see also tyrosine kinase
    lymphoma (ALCL) MAKEI, germ-​cell cancer, 257 metabolism, 26 inhibitors (TKIs)
    Burkitt lymphoma see Burkitt malignancy-​associated meropenem, 73 molecular radiotherapy, 34
    lymphoma (BL) haemophagocytic mesenchymal chondrosarcoma, monoclonal antibodies, 29–​30, 66–​67
    diffuse large B cell see diffuse large lymphohistiocytosis, 166 extraskeletal, 213t MonoMAC syndrome, 135
    B-​cell lymphoma (DLBCL) malignant ascites, 87 metabolic syndrome (MS) monomethyl auristatin-​E
    diffuse large B-​cell lymphoma see malignant fibrous histiocytomas, late adverse effects, 110 (MMAE), 160
    diffuse large B-​cell lymphoma imaging, 16 post-​allogeneic HSCT, 60 monosomy 7, juvenile myelomonocytic
    (DLBCL) malignant melanoma radiotherapy side effects, 36 leukaemia, 139
    Hodgkin’s lymphoma see Hodgkin’s AYA, 3t, 97, 98 metaiodobenzylguanidine ( morphine, 85
    lymphoma epidemiology, 2t 123I-​MIBG), 13 mortality, late, 106–​107
    imaging, 16 late adverse effect as, 113 neuroblastoma, 245, 246f, mouth care, 75
    incidence, 3–​4 survival by age, 8f 247–​248 MPNSTs see malignant peripheral
    large B-​cell lymphoma with IRF4 survival rate, 5t metastatic retinoblastoma, 272 nerve sheath tumours
    rearrangements, 144 malignant peripheral nerve sheath methadone, 85, 86 (MPNSTs)
    lymphoblastic lymphoma tumours (MPNSTs), methotrexate (MTX) MPO (myeloperoxidase), 128
    see lymphoblastic 212, 213t adverse effects, 24t MRD see minimal residual
    lymphoma (LL) imaging, 16 ALL treatment, 119–​120 disease (MRD)
    marginal-​zone lymphomas, 145 malignant rhabdoid tumours, Burkitt lymphoma treatment, 147 MRI see magnetic resonance
    nodular lymphocyte predominant 216–​217 doxorubicin, cisplatin (MAP), 226 imaging (MRI)
    Hodgkin’s lymphoma malnourishment, treatment lymphoblastic lymphoma MRPs (multidrug resistance
    see nodular lymphocyte toxicities, 94 treatment, 150 proteins), 23
    predominant Hodgkin’s mammalian target of rapamycin osteosarcoma, 225 MT110, 67
    lymphoma (NLPHL) (mTOR) inhibitors, 29 renal toxicity, 112 MTDs see maximum tolerated
    non-​Hodgkins see non-​Hodgkin’s CNS toxicity, 114 Rosai–​Dorfman disease, 167 dose (MTD)
    lymphoma (NHL) low-​grade gliomas, 181 toxicity, 23 mTOR inhibitors see mammalian
    peripheral T-​cell lymphomas, 145 MAP (methotrexate, doxorubicin, uncommon histiocytoses, 167 target of rapamycin (mTOR)
    primary mediastinal large B-​cell cisplatin), 226 methylphenidate, 87 inhibitors
    lymphoma, 144, 147 MAP2K gene, 162 metoclopramide, 75 MTS (Musculoskeletal Tumor
    Lymphomes Malins B (LMB/​French) marginal-​zone lymphomas metronomic therapy, Society), 223
    studies, 147 (MZLs), 145 medulloblastoma relapse, 195 mucositis
    lysine-​specific demethylase maximum tolerated dose (MTD) microarrays, AML, 128 chemotherapy, 23
    (KDM), 182 chemotherapy, 23 microRNA, Burkitt lymphoma, 144 radiotherapy side effects, 34
    lysine-​specific methyltransferase phase I clinical trials, 39 microscopic scleral invasion, multi-​agent chemotherapy, Ewing
    (KMT), 182 MCV (mean corpuscular retinoblastoma, 274–​275 sarcoma, 228
    volume), 135 midazolam, 87 multidrug resistance proteins
    MAC (myeloablative MDD (minimal disseminated middle-​development studies, clinical (MRPs), 23
    conditioning), 56–​57 disease), 145 trials, 39 multifocal retinoblastoma, 270
    macrogol, 86 MDM.TP53, 250–​251 middle-​income countries see low-​ multisystem Langerhans cell
    macrophage activation syndrome MDS see myelodysplastic and middle-​income countries histiocytosis, 163, 164
    (MAS), 166 syndrome (MDS) (LMICs) Musculoskeletal Tumor Society
    magnetic resonance imaging mean corpuscular volume minimal disseminated disease (MTS), 223
    (MRI), 13 (MCV), 135 (MDD), 145 mutations, AML, 124, 125f
    astrocytomas, 18f, 19 measles, mumps, rubella (MMR) minimal residual disease (MRD) MYCN oncogene
    bone tumours, 16, 220, 220f vaccination polio ALL, 58–​59, 118, 119 neuroblastoma, 245
    brainstem gliomas, 18 vaccination, 73 AML, 59, 129–​130 sporadic neuroblastoma, 242, 243
    brain tumours, 173 mechanistic evidence-​based chemotherapy, 21 therapy target as, 250
    central nervous system, 16 medicine, 50 HSCT, 58–​59 MYC pathway deregulation, Burkitt
    clear-​cell sarcoma of the MEDI-​565, 67 phase II clinical trials, 40 lymphoma, 144
    kidney, 237 Medical Research Council (MRC), prognosis, 145 myeloablative conditioning
    craniopharyngiomas, 19, 19f AML prognosis, 130 minor sibling donors, allogeneic (MAC), 56–​57
    embryonal tumours, 188, 191 medulloblastoma, 188 HSCT, 55 myelodysplasia, 107
    ependymomas, 201 biology, 190 mismatch repair (MMR) genes, 142 myelodysplastic syndrome (MDS),
    Ewing sarcoma, 221 epidemiology, 188 mitogen-​activated protein kinase 134–​137
    gadolinium-​enhanced, 174, 175f high-​risk, 193 (MEK), 28 AML vs., 127f, 136–​137, 137t
    germ-​cell tumours, 16, 256 imaging, 16, 17, 18f mitoxantrone AYA epidemiology, 3t
    germinomas, 198, 199f LMIC, 94 adverse effects, 24t bone-​marrow features, 135, 136f
    high-​grade gliomas, 181–​182 standard-​risk AML treatment, 130 classification, 134
    Hodgkin’s lymphoma, 16 medulloblastoma, 193 ML-​DS see myeloid leukaemia of clinical features, 135
    Langerhans cell histiocytosis, 163 survival rate, 5t Down syndrome (ML-​DS) cytogenetics, 135–​136
    liver tumours, 15f WNT medulloblastoma, 190 MLL rearrangements, AML, 129 diagnosis, 135b
    neuroblastoma, 245 MEK (mitogen-​activated protein MMAE (monomethyl differential diagnosis, 136
    osteosarcoma, 220f kinase kinase), 28 auristatin-​E), 160 epidemiology, 134–​135
    renal tumours, 13, 13f MEK-​162, low-​grade gliomas, 181 MMR (measles, mumps, rubella) immunophenotyping, 136
    retinoblastoma, 272 melphalan vaccination, 73 incidence, 134
    rhabdomyosarcomas (RMS), adverse effects, 24t MMR (mismatch repair) genes, 142 laboratory features, 135
    15, 208 myelodysplastic syndrome, 137 model-​based design, phase I clinical molecular genetics, 136
    spinal cord tumours, 174 renal toxicity, 111–​112 trials, 39 natural course, 137
    MAHO, testicular cancer, 257 retinoblastoma, 274 models of care, AYA, 99–​101 non-​clonal disorders vs., 136
    maintenance therapy 6-​mercaptopurine molecular aberration correction prognosis, 137
    ALL treatment, 119–​120 adverse effects, 24t chemotherapy, 27–​28 treatment, 137
    286 Index

    myeloid leukaemia of Down neuroblastoma treatment, epidemiology, 2t, 142 osteosarcoma


    syndrome (ML-​DS), 129, 246–​251, 249t imaging, 16 AYA epidemiology, 3t
    137–​139, 138f anti-​GD2 monoclonal incidence, 142 biology, 222
    myelopathy, 108 antibodies, 29–​30 pathology, 143–​145 clinical presentation, 219, 220f
    myeloperoxidase (MPO), 128 autologous stem rescue, 23 rare types, 145 epidemiology, 2t, 219
    myocardial infarction, 35 chemotherapy, 248, 250–​251 staging, 145–​146 future work, 228
    myxoid chondrosarcoma, molecular radiotherapy, 34 survival, 5, 5t, 6, 6t, 7f histopathology, 221
    extraskeletal, 213t radiotherapy, 247–​248 non-​inferiority ‘head-​to-​head’ imaging, 16, 220–​221, 220f
    MZLs (marginal-​zone risk-​adapted treatment, 248–​249 clinical trials, 42 incidence, 219
    lymphomas), 145 surgery, 177, 246–​247 non-​ionizing radiation, 10 prognosis, 223
    neurocognition, 114 non-​melanomatous skin cancer remission status, 227
    National Cancer Institute, Common neurofibroma, plexiform, 28 (NMSC), 113 survival, 4, 5t, 6, 6t
    Terminology Criteria for neurofibromatosis type 1, 166 non-​rhabdomyosarcoma soft-​ osteosarcoma treatment, 223
    Adverse Events, 23 cancer risk, 11 tissue sarcomas (NRSTSs), chimeric antigen receptor clinical
    National Institute of Clinical CNS tumours, 169–​170, 170t 212–​217, 213t trials, 68, 69f
    Excellence (NICE), high-​grade gliomas, 181 adult-​type, 214–​215 metastases treatment, 226–​227
    Improving Outcomes juvenile myelomonocytic diagnosis, 212–​214 recurrent disease therapy, 227
    Guidance for Cancer leukaemia, 139 imaging, 16 systemic therapy, 225–​226
    in Children and Young low-​grade gliomas, 179 non-​steroidal anti-​inflammatory ovaries
    People, 100 pilocytic astrocytomas (PA), 17 drugs (NSAIDs), 85 dysfunction as adverse effects,
    national paediatric oncology Wilms tumour, 231 Noonan syndrome, 139 109–​110
    programmes, LMIC, 93, 93f neurofibromatosis type 2 NOPHO (Nordic Society of tumour surgery, 256–​257
    National Paediatric Oncology Unit cancer risk, 11 Paediatric Haematology and overt extraocular retinoblastoma, 275
    (Guatemala), 90 CNS tumours, 169–​170, 170t oncology), 138 oxazaphosphorine, 112
    natural killer (NK) cells neurofibromin 1 (NF1), 241 Nordic Adult Life After Childhood oxycodone, 85
    allogeneic HSCT, 60, 66 neurological surgery, 114 Cancer in Scandinavia
    immunotherapy, 68 neurological toxicity, 108 (ALiCCS), 106 p53 gene, 222
    solid tumours, 60 neuropathic pain, 84, 86 Nordic Society of Paediatric paediatric cancer units (PCUs),
    nausea and vomiting, 74–​75, 86, 86t neuropsychological deficits, ALL, 80 Haematology and oncology 90–​93, 93t
    radiotherapy, 32 neuropsychological toxicity, 108 (NOPHO), 138 paediatric investigational plans
    radiotherapy side effects, 34 neuropsychology, 80 North American Cooperative (PIPs), 26
    NBL late adverse effects, 114 Group on Child Hepatic pain
    chimeric antigen receptor clinical neutropenia tumours, 264 neuropathic pain, 84, 86
    trials, 68 fever without, 74 nortriptyline, 86 nociceptive pain, 84
    differentiation-​inducing agents, 30 infections, 71 NOTCH1 mutations, 145 palliative care, 84–​86
    NDI (nephrogenic diabetes nevoid basal-​cell carcinoma NPM-​1 (nucleophosmin-​1), 129 psychosocial care, 78
    insipidus), 111 syndrome, 170t NPM-​ALK gene fusion, 152–​153 somatic pain, 84
    nelarabine, 27, 150 next-​generation sequencing (NGS) NRAS gene, 139 visceral pain, 84–​85
    nephroblastoma see Wilm’s tumour AML, 128 NRSTSs see non-​rhabdomyosarcoma palliative care, 83–​89
    (nephroblastoma) Burkitt lymphoma, 144 soft-​tissue sarcomas bereavement care, 88
    nephroblastomatosis, 231–​232 haemophagocytic (NRSTSs) cancer-​directed therapy, 84, 84b
    treatment, 236, 237f lymphohistiocytosis, 165 NRTK2/​3 genes, 180 communication, 83–​84
    see also Wilm’s tumour NF1 gene, 241 NSAIDs (non-​steroidal anti-​ decision making, 83–​84
    (nephroblastoma) familial neuroblastoma, 242 inflammatory drugs), 85 family support, 88
    nephrogenic diabetes insipidus juvenile myelomonocytic NTRK1/​2 genes, 243 LMICs, 95
    (NDI), 111 leukaemia, 139 nuclear medicine, 13 nutrition, 88
    nephrogenic rests, 231–​232 NHL see non-​Hodgkin’s nucleophosmin-​1 (NPM-​1), 129 spiritual distress, 88
    nephrotoxicity, 111 lymphoma (NHL) nutrition steroids, 87–​88
    neuroblastoma, 241–​252 nitrosoureas, 112 infection prevention, 71 symptoms, 84–​87, 85b
    aetiology, 241 nivolumab, 160 palliative care, 88 pancreatic cancer, 28
    biological aspects, 243–​244 NLPHL see nodular lymphocyte radiotherapy, 32 panitumumab, 28
    cervical neuroblastoma, 244 predominant Hodgkin’s supportive care, 75 panobinostat, 29
    clinical characteristics, 244t lymphoma (NLPHL) paracetamol, 85
    clinical presentation, diagnosis, NMDA antagonists, 86 obesity, 25 parallel evidence, evidence-​based
    staging, 244–​246, 245f NMSC (non-​melanomatous skin ocular juvenile xanthogranuloma, medicine, 50
    diagnostic criteria, 245 cancer), 113 166, 167 paraplegia, 174
    epidemiology, 2t, 241 nociceptive pain, 84 OMS (opsoclonus-​myoclonus parents, 79–​80
    extramedullary spinal cord nodular lymphocyte predominant syndrome), 244 dying child worries, 81
    tumours, 176 Hodgkin’s lymphoma ondansetron, 75 Parinaud’s syndrome, 198
    familial neuroblastoma, 241–​242 (NLPHL), 157 OpenOffice.org, 92 PARP inhibitors, 228
    imaging, 14f, 15 clinical management, 160 opioids, 85 parvovirus infection, 139–​140
    incidence, 4 non-​clonal disorders, myelodysplastic breathlessness, 86 PAs see pilocytic astrocytomas (PAs)
    pathology, 243 syndrome vs., 136 induced constipation, 87 patient-​centered model, clinical care
    predisposition, genetics, non-​Hodgkin’s lymphoma (NHL), opsoclonus-​myoclonus syndrome in AYA, 99
    molecular pathogenesis, 142–​156 (OMS), 244 patient registries, allogeneic
    241–​243, 243f AYA epidemiology, 3t optic neuritis, 36 HSCT, 52–​53
    prognostic groups, 248–​249, 248t B-​cell type see B-​cell non-​ optic pathway gliomas, 171, 173 PAX6 gene, 231
    relapse, 250 Hodgkin’s lymphoma imaging, 19 pazopanib, 29
    sporadic neuroblastoma, 242–​243 classification, 146b oral pain medication, 85 PCR see polymerase chain
    staging, 245–​246 clinical presentation/​treatment, orbital tumours, 208, 208f reaction (PCR)
    survival, 4, 5t, 6, 6t 146–​147, 150, 152–​153 orthotopic liver transplant, 266–​267 PCUs (paediatric cancer units),
    tumour markers, 245 definition, 142 osmotic laxatives, 76, 86 90–​93, 93t
    Index 287

    PDGFRB gene, 27 ALL diagnosis, 118 Pseudomonas aeruginosa infection, 74 dose prescription and
    Pediatric Oncology Group (POG), AML, 128 psychology, AYA, 98–​99, 102 fractionation, 32
    myeloid leukaemia and Down polymorphisms, AML, 129 psychosocial care, 78–​82 dose reduction, 109
    syndrome, 138 population mixing hypothesis, 10 acute phase, 78–​79 ependymomas, 203
    peer support, AYA, 99 positron emission end-​of-​life issues, 80–​81 Ewing sarcoma, 223, 224, 225
    PEI (ifosfamide, etoposide, cisplatin), tomography (PET) late adverse effects, 114–​115 external-​beam see external-​beam
    199, 257 germ-​cell tumours, 256 long-​term follow-​up, 80 radiotherapy (EBRT)
    PENAT score, 74 Hodgkin’s lymphoma, 16 post treatment, 79–​80 follow-​up, 33
    peripheral blood, allogeneic Langerhans cell histiocytosis, 163 during treatment, 79 germ-​cell tumours, 199
    HSCT, 55 neuroblastoma, 245 psychosocial distress, 80 haemorrhagic cystitis as adverse
    peripheral nerve sheath tumours, 2t rhabdomyosarcomas (RMS), 208 psychostimulants, 87 effect, 112
    peripheral T-​cell lymphomas see also computed PTCL (peripheral T-​cell heart failure as adverse effect, 111
    (PTCL), 145 tomography (CT) lymphomas), 145 high-​grade gliomas, 183
    Perlman syndrome, Wilms posterior fossa tumours, 188 PTPN11 gene, 139, 140 high-​risk medulloblastoma, 193
    tumour, 231 post-​remission, AML treatment, PTSD (post-​traumatic stress Hodgkin’s lymphoma treatment,
    persistent hyperplastic primary 130–​131 disorder), 80 159–​160
    vitreous (PHPV), post-​transplant immunotherapy, PTV (planning target volume), image-​guided radiotherapy, 33–​34
    retinoblastoma, 272 allogeneic HSCT, 60–​61 radiotherapy, 32 imaging, 32
    PET see positron emission post-​traumatic stress disorder puberty, precocious puberty, 109 informed consent, 31
    tomography (PET) (PTSD), 80 pulmonary system intensity-​modulated
    PF-​02341066 see Crizotinib precocious puberty, 109 Langerhans cell histiocytosis, 162 radiotherapy, 33
    (PF-​02341066) prednisolone late adverse effects, 111 ionizing radiation, 10
    PF-​06463922 (lorlatinib), 250 Adriamycin and vincristine (APO) radiotherapy side effects, 35 low-​grade gliomas, 181
    PFS (progression-​free survival), regimen, 152 PVT (cisplatin, vinblastine, molecular radiotherapy, 34
    ependymomas, 201 adverse effects, 24t bleomycin), 257 neuroblastoma, 247–​249
    pharmacokinetics of bleomycin, etoposide, doxorubicin, PXA (pleomorphic non-​ionizing radiation, 10
    chemotherapy, 26 cyclophosphamide, xanthoastrocytoma), 179 non-​rhabdomyosarcoma soft-​
    phase I clinical trials, 38, 39–​40 vincristine, procarbazine tissue tumours, 214
    phase II clinical trials, 38, 40 (BEACOPP), 158, 159f, 160 quantitative reverse transcription normal tissue injury, 34
    phase III clinical trials, 38, 40–​42 prednisone polymerase chain reaction osteosarcoma, 225
    phase IV clinical trials, 38–​39 ALL treatment, 119 (RT-​PCR), 129 planning and dosimetry, 32
    photoreceptor-​differentiation dacarbazine, cyclophosphamide, question asking, evidence-​based positioning and immobilisation, 31
    antigens, retinoblastoma, 272 vincristine (COPDAC), 160 medicine, 47 proton beam radiotherapy, 34, 181
    PHOX2B gene, 241, 242 dexamethasone, vincristine, quinolones, 72 reduction in allogeneic
    PHPV (persistent hyperplastic L-​asparaginase, 119 HSCT, 55–​56
    primary vitreous), gemcitabine, ifosfamide, radiation necrosis, 108 retinoblastoma, 273
    retinoblastoma, 272 vinorelbine (IGEV), 160 radiation pneumonitis, 35 rhabdomyosarcomas, 210–​211
    PICO, 47 procarbazine, vincristine, radiography, 12 side effects, 34–​37
    pigmentation disorders, 113 cyclophosphamide bone tumours, 220, 220f skeletal toxicity, 113
    pilocytic astrocytomas (PAs), 179 (COPP), 160 germ-​cell tumours, 256 skin adverse effects, 113
    imaging, 16, 17, 17f pregnancy, 10 hepatoblastoma, 263 spinal irradiation, 35
    neurofibromatosis type 1, 17 preoperative chemotherapy, 21–​22 Langerhans cell histiocytosis, 163 stereotactic radiotherapy, 33
    pineal tumours, 198 prespecified interim analysis, osteosarcoma, 220f supportive care, 32–​33
    PIPs (paediatric investigational Bayesian clinical trials, 43 spinal cord tumours, 174 target volume definition, 32
    plans), 26 pre-​treatment extent (PRETEXT) radiometabolic salvage therapy, thoracic irradiation, 35
    placental alkaline phosphatase staging system, 15, 264 neuroblastoma, 247–​248 total body irradiation see total
    (PLAP), germ-​cell tumours, PRF1 gene, 165 radiotherapy, 31–​37, 84b body irradiation (TBI)
    254–​255 primary mediastinal large B-​cell accuracy verification, 32 uncommon histiocytoses, 167
    planned progressive infection lymphoma (PMBL), 144, 147 adaptive, 33–​34 Wilm’s tumour
    therapy, 73–​74 procarbazine adaptive radiotherapy, 33–​34 (nephroblastoma), 235
    planning target volume (PTV), adverse effects, 24t adult-​type non-​ Rainbow’s Trust, 88
    radiotherapy, 32 bleomycin, etoposide, doxorubicin, rhabdomyosarcoma soft-​ raised intracranial pressure, 87–​88
    PLAP (placental alkaline cyclophosphamide, tissue sarcomas, 215 randomized controlled trials (RCTs)
    phosphatase), germ-​cell vincristine, prednisolone assessment, 31 evidence-​based medicine, 46
    tumours, 254–​255 (BEACOPP), 158, 159f, 160 auditory toxicity, 112 heart failure as adverse effect, 111
    platelet transfusions, 74 fertility effects, 161 brachytherapy, 34 randomized phase IIB clinical
    platinum thioguanine, CCNU, vincristine brainstem gliomas, 18 trials, 40
    auditory system toxicity, 112 (TPCV), 181 cause as, 10 RANK (receptor activator for
    chemotherapy, 22t 6-​thioguanine, dibromodulcitol, choice, 31 nuclear factor kappa-​B)
    nephrotoxicity, 111 lomustine (CCNU), 184 choroid plexus tumours, 204 pathway, 228
    platinum–​etoposide–​ifosfamide, 199 vincristine, cyclophosphamide, CNS toxicity, 114 RAS gene, 139, 140
    pleomorphic xanthoastrocytoma prednisone (COPP), 160 conventional external beam RB1 gene, 11
    (PXA), 179 progression-​free survival (PFS), radiotherapy, 33 ALL, 118
    plexiform neurofibroma, 28 ependymomas, 201 cranial see cranial bone tumours, 219
    PMBL (primary mediastinal large B-​ proteomics, 128 radiotherapy (CRT) retinoblastoma, 270
    cell lymphoma), 144, 147 protocol-​based care, LMICs, 93–​94 craniospinal see craniospinal RB gene, 222
    Pneumocystis jiroveci pneumonia proton beam radiotherapy, 34, 181 irradiation (CSI) RCC see renal cell carcinoma (RCC)
    prevention, 72 proton therapy, 108 craniospinal irradiation RCTs see randomized controlled
    POG (Pediatric Oncology Group), craniopharyngiomas, 201 see craniospinal trials (RCTs)
    myeloid leukaemia and Down standard-​risk irradiation (CSI) receptor activator for nuclear
    syndrome, 138 medulloblastoma, 193 decision making, 31 factor kappa-​B (RANK)
    polymerase chain reaction (PCR) Provenge® (Sipuleucel-​T), 68–​69 documentation, 33 pathway, 228
    288 Index

    RECIST (response evaluation criteria rhabdoid tumour of the kidney SDS (Shwachman–​Diamond social environment, psychosocial
    in solid tumours), 40 (RTK), 232, 238 syndrome), 135 care, 78–​79
    recombinant human arginase rhabdoid tumour predisposition secondary malignancies sodium picosulphate, 86
    BCT-​100, 30 syndrome, 170t ALL treatment, 120 softening agents, 86
    recruitment, paediatric cancer rhabdomyosarcomas (RMS), AYA, 101–​102 soft-​tissue sarcomas, 206–​218
    units, 91–​92 206–​212 Hodgkin’s lymphoma, 160 epidemiology, 207f
    rectal enemas, 87 aetiology, 206 late adverse effects, 105, 107–​108 incidence, 4
    red-​cell transfusions, 74 allogeneic HSCT, 60 SEER (Surveillance, Epidemiology, secondary malignant neoplasm
    reduced-​intensity AYA, 3t, 97, 98 and End Results), as, 107
    conditioning, 56–​57 biology, 206–​207 American National Cancer survival rate, 5t
    refractory cytopenia of butyroid subtype of, 206 Institute, 142 see also
    childhood, 136 clinical presentation, 207–​208 SEGA (subependymal giant-​cell rhabdomyosarcomas (RMS)
    regional networks, LMIC, 94 diagnosis, 207–​208 astrocytoma), 179 Soft Tissue Sarcoma Study Group
    renal-​cell carcinoma (RCC), 13, 232, epidemiology, 2t, 206 SEGAs (subependymal giant-​cell (EpSSG), 101
    238–​239 imaging, 15 astrocytomas), 170, 179 solid tumours
    renal-​sparing surgery, Wilm’s metastases, 209 selective decontamination of the allogeneic HSCT, 59–​60
    tumour, 235 pathology, 206–​207 digestive tract (SDD), 72 chemotherapy, 21
    renal system, late adverse effects, prognosis, 209, 209t, 212 self-​reporting, pain, 84–​85 chimeric antigen receptor clinical
    111–​112 staging, 208 selumetinib, 28, 181 trials, 68
    renal tubular acidosis, 111 survival, 4, 5t, 6t seminomas, 254 somatic pain, 84
    renal tumours, 231–​240 treatment, 208–​212 semi-​quantitative Deauville five-​ sonic hedgehog (SHH), 17
    clear-​cell sarcoma of the kidney, ridaforolimus, 29 point system, Hodgkin’s sorafenib, 28
    237–​238 ring sideroblasts, 136 lymphoma, 159b SOS see sinusoidal obstructive
    congenital mesoblastic rituximab, 29, 66 SETPB1 gene, 140 syndrome (SOS)
    nephroma, 239 B-​cell non-​Hodgkin’s sexual activity, AYA, 98–​99 spasticity, spinal cord tumours, 174
    epidemiology, 2t lymphoma, 147 SHH medulloblastoma, 190 SPECTA-​AYA, 97
    imaging, 13, 13f chemotherapy dose, 23 SHP-​2, juvenile myelomonocytic spermatogenesis, 36
    renal-​cell carcinoma, 13, 232, Rosai–​Dorfman disease, 167 leukaemia, 139 spinal cord effects
    238–​239 RMS see rhabdomyosarcomas (RMS) Shwachman–​Diamond syndrome extramedullary spinal cord
    rhabdoid tumours, 238 ROBINS-​I tool, 49 (SDS), 135 tumours, 176–​177
    see also Wilm’s tumour ROBIS tool, 49 Simpson–​Golabi–​Behmel tissue neuroblastoma treatment, 249
    (nephroblastoma) Rosai–​Dorfman disease (sinus overgrowth syndrome, 262 spinal cord tumours, 173–​175
    Resonance Oncology, 93–​94 histiocytosis with massive single nucleotide polymorphism abnormal gait, 173f
    Resonance Patient Centre, 92–​93 lymphadenopathy), 166, 167 (SNPs) array analysis, Burkitt classification, 174b
    response-​adapted strategy, Hodgkin’s rotation-​plasty, osteosarcoma, 224 lymphoma, 144 clinical descriptions, 174
    lymphoma, 158 RTK (rhabdoid tumour of the single-​system Langerhans cell diagnosis, 174, 175f
    response evaluation criteria in solid kidney), 232, 238 histiocytosis, 163 diagnostic delay, 173–​174
    tumours (RECIST), 40 RT-​PCR see reverse transcription-​ bone, 163–​164 emergency presentation, 175
    retinoblastoma, 270–​275 polymerase chain reaction skin, 163 extramedullary, 175, 176–​177
    ancillary studies, 273 (RT-​PCR) sinus histiocytosis with massive incidence, 4
    bilateral retinoblastoma, 270 Rubenstein–​Taybi syndrome, 170t lymphadenopathy (Rosai–​ intramedullary, 175–​176
    biology, 270–​272 rule-​based design, phase I clinical Dorfman disease), 166, 167 spinal irradiation, 35
    diagnosis, 272–​273 trials, 39 sinusoidal obstructive syndrome spinal surgery, 114
    epidemiology, 2t (SOS), 36 spindle-​cell
    genetic counselling, 271t Sackett, David, 46 Wilms tumour, 237 rhabdomyosarcomas, 206
    genetics, 11 salpingo-​oophorectomy, 256–​257 SIOP see International Society of spiritual distress, 88
    grouping and staging, 272, sample size Paediatric Oncology (SIOP) splenectomy, 113
    272b, 273t Bayesian clinical trials, 43 SIOPEL see International Childhood splenic radiotherapy, 113
    hereditary, 170t phase III clinical trials, 41 Liver Tumour Strategy Group spontaneous regression,
    histology, 272 sarcomas (SIOPEL) myelodysplastic
    imaging studies, 272 adult-​type non-​ SIOPEN (European International syndrome, 137
    incidence, 4 rhabdomyosarcoma Society of Paediatric sporadic neuroblastoma, 242–​243
    metastatic retinoblastoma, 272 soft-​tissue sarcomas see Oncology European SS see synovial sarcoma (SS)
    multifocal retinoblastoma, 270 non-​rhabdomyosarcoma soft-​ Neuroblastoma Network), standardized mortality ratio
    overt extraocular tissue sarcomas (NRSTSs) 30, 246 (SMR), 106
    retinoblastoma, 275 alveolar soft-​part sarcoma, 213t Sipuleucel-​T (Provenge®), 68–​69 standard-​risk medulloblastoma, 193
    secondary malignancies, 275 bone see bone sarcoma sirolimus, 29 Staphylococcus infections, 74
    signs and symptoms, 270, 271f clear-​cell sarcoma, 13, 213t skeleton, adverse effects, 113 statistics, 40–​41, 41f, 41t, 42
    survival, 4, 5t, 6t clear-​cell sarcoma of the kidney, skin stem-​cell transplantation see
    trilateral retinoblastoma, 270 232, 237–​238 Langerhans cell histiocytosis, 162 haematopoietic stem-​cell
    unilateral retinoblastoma, 270 epithelioid sarcoma, 213t late adverse effects, 113 transplantation (HSCT)
    retinoblastoma treatment, 273–​275 Ewing sarcoma see Ewing non-​melanomatous skin stereotactic radiotherapy, 33
    chemotherapy, 273–​275 sarcoma (ES) cancer, 113 sterility, 36
    focal therapies, 273 extraosseous sarcoma, 5t radiotherapy side effects, 34 steroids, 87–​88
    radiotherapy, 273 growth hormone deficiency, 109 skin cancer stimulant agents, 86
    surgery, 273 Kaposi sarcoma, 94 basal cell carcinoma, 36–​37 stool softeners, 76
    retinoic acid, Langerhans cell soft-​tissue see soft-​tissue sarcomas epidemiology, 2t stress, 79
    histiocytosis, 164 synovial see synovial sarcoma (SS) see also malignant melanoma AYA, 102
    reverse transcription-​polymerase scarring, 113 SMO (smoothened) inhibitors, 195 radiotherapy, 32–​33
    chain reaction (RT-​PCR), 145 schwannomatosis, 170t smoothened (SMO) inhibitors, 195 structured abstracts, 48
    Ewing sarcoma, 221, 222 SDD (selective decontamination of SMR (standardized mortality subcutaneous infusion, pain
    rhabdoid tumour, 213t the digestive tract), 72 ratio), 106 medication, 85
    Index 289

    subependymal giant-​cell TCRαβ depletion, 54 total diagnostic interval (TDI), 169 unmanipulated bone marrow,
    astrocytomas (SEGAs), TDI (total diagnostic interval), 169 TP53 gene, 118 allogeneic HSCT, 54
    170, 179 TDM (therapeutic drug TPCV (thioguanine, procarbazine, UpToDate, 48
    sublingual pain medication, 85 monitoring), 26 CCNU, vincristine), 181 US see ultrasound (US)
    sub-​Saharan Africa technetium99-​MDP training uterus, late therapy effects, 110
    Burkitt lymphoma, 147 (methylene-​diphosphate), 13 AYA, 100–​101
    non-​Hodgkin’s lymphoma, 142 Teenage Cancer Trust (UK), 102 paediatric cancer units, vaccinations, 73
    sudden coma, brain tumours, 171, 173 Teen Cancer America (USA), 102 91–​92, 93t tumour vaccines, 68–​69
    supportive care, 71–​77 temozolomide, 27, 228 paediatric cancer units in LMIC, vanillylmandelic acid (VMA), 245
    blood-​product transfusions, 74 adverse effects, 24t 91, 93t varicella zoster virus infection, 74
    constipation, 75–​76 glioblastoma multiforme, 184 trametinib, 181 vascular endothelial growth
    infection see infections temsirolimus, 29 transdermal opioid patches, 85 factor (VEGF)
    mouth care, 75 teniposide (VM-​26), 24t transient abnormal myelopoiesis neuroblastoma, 244
    nausea and vomiting, 74–​75 teratomas, 198 (TAM), 138 rhabdomyosarcomas, 212
    nutrition support, 75 therapy, 259 Down syndrome, 137–​138 vasculopathy, 108
    tumour lysis syndrome, 76 TERT gene, 243 translational clinical trials, 39 VEGF see vascular endothelial
    supratentorial ependymomas, 19 testicular dysfunction, 110 trastuzumab, 67 growth factor (VEGF)
    supratentorial tumours, 19 testicular tumours treatment cessation, LMIC, 94–​95 venetoclax, 30
    surgery, 84b AYA, 97 TRH (thyrotropin-​releasing venous occlusive disease
    abdominal, late adverse effects, 114 surgery, 256 hormone), 36 (VOD), 237
    adult-​type non-​ TFE3 gene fusion, 239 trilateral retinoblastoma, 270 vinblastine
    rhabdomyosarcoma soft-​ therapeutic drug monitoring trimethoprim/​sulfamethoxazole adverse effects, 24t
    tissue sarcomas, 214–​215 (TDM), 26 (TMP/​SMZ), 72 anaplastic large-​cell
    craniopharyngiomas, 200 therapy, late effects of see late adverse trisomy 8 lymphoma, 152
    embryonal tumours, 192–​193 effects (LAEs) myelodysplastic syndrome, bleomycin, doxorubicin,
    ependymomas, 202–​203 thioguanine, procarbazine, CCNU, 135–​136 dacarbazine combination
    Ewing sarcoma, 223 vincristine (TPCV), 181 myeloid leukaemia and Down (ABVD), 158, 160
    fibroblastic-​myofibroblastic 6-​thioguanine syndrome, 138 cisplatin, bleomycin (PVT), 257
    lesions, 216 adverse effects, 24t trisomy 18, 262 Rosai–​Dorfman disease, 167
    germ-​cell tumours, 256–​257 procarbazine, dibromodulcitol, trisomy 21, 135–​136 vincristine
    high-​grade gliomas, 182 lomustine (CCNU), 184 TSC (tuberous sclerosis complex), 11, Adriamycin and prednisolone
    late adverse effects, 114 thiotepa, 24t 170, 170t (APO) regimen, 152
    low-​grade gliomas, 180–​181 thoracic irradiation, 35 TSC1/​2 gene, 179 adverse effects, 24t
    neuroblastoma, 246–​247 thoracic surgery, 114 TSH (thyroid-​stimulating hormone) bilateral nephroblastoma, 237
    non-​rhabdomyosarcoma soft-​ 3 + 3 design, phase I clinical trials, 39 deficiency, 109 bleomycin, etoposide,
    tissue tumours, 212–​213, 214 three-​step analgesia ladder TTL (target toxicity level), 39 doxorubicin,
    osteosarcoma, 223, 224, 224t (WHO), 85 t( 9;22) translocation, 58 cyclophosphamide,
    retinoblastoma, 273 thrombocytopenic anaemia, 162 tuberous sclerosis complex (TSC), 11, prednisolone, procarbazine
    rhabdomyosarcomas, 210 thyroid cancer 170, 170t (BEACOPP), 158, 159f, 160
    uncommon histiocytoses, 167 AYA, 3t, 97 tumour lysis syndrome, 76 CCNU, procarbazine, thioguanine
    Wilm’s tumour epidemiology, 2t tumour markers (TPCV), 181
    (nephroblastoma), 235 molecular radiotherapy, 34 germ-​cell tumours, 254–​255 cyclophosphamide, dacarbazine,
    Surveillance, Epidemiology, and End secondary malignant neoplasm neuroblastoma, 245 prednisone (COPDAC), 160
    Results (SEER), American as, 107 tumour–​node–​metastasis cyclophosphamide, procarbazine,
    National Cancer Institute, 142 thyroid dysfunction, Hodgkin’s (TNM) system prednisone (COPP), 160
    survival rates, 4–​6, 10 lymphoma, 161 bone tumours, 223 Ewing sarcoma, 226
    age by, 5, 6t, 7–​9f thyroid gland retinoblastoma, 272 hepatoblastoma treatment, 264
    cancer type by, 4–​5 late adverse effects, 110 tumour vaccines, 68–​69 high-​grade gliomas, 184
    European geographical radiotherapy side effects, 36 two-​hit hypothesis, intramedullary spinal cord
    variation, 5, 9f thyroid-​stimulating hormone (TSH) retinoblastoma, 270 tumours, 175
    follow-​up, 10 deficiency, 109 tyrosine kinase inhibitors (TKIs) lymphoblastic lymphoma
    trends 1999-​2007, 5t, 6, 9f, 10 thyrotropin-​releasing hormone ABL gene inhibition, 27–​28 treatment, 150
    synovial sarcoma (SS), 213t, 214 (TRH), 36 AML therapy, 132 retinoblastoma, 273–​274
    adult-​type non-​ TKIs see tyrosine kinase anaplastic lymphoma kinase rhabdomyosarcomas, 211
    rhabdomyosarcoma soft-​ inhibitors (TKIs) inhibition, 28 vinorelbine, ifosfamide,
    tissue sarcomas, 215f TLX1/​HOX11 genes, 118 BRAF/​MEK inhibition, 28 prednisone, gemcitabine
    systematic cancer reviews, evidence-​ TLX3/​HOX11L2 genes, 118 (IGEV), 160
    based medicine, 48–​49 TMP/​SMZ (trimethoprim/​ Uganda training programme, 93t viral infection treatment, 74
    systemic juvenile sulfamethoxazole), 72 UK ALL 2011 study, 23 viral prophylaxis, 72
    xanthogranuloma, 166 TNM see tumour–​node–​metastasis ultrasound (US), 12 visceral pain, 84–​85
    (TNM) system AML diagnosis, 126 visual system
    TAL/​LMO gene, 118 topoisomerase I inhibitors, 22t contrast-​enhanced ultrasound, 12 brain tumour effects, 173
    TAM see transient abnormal topotecan Langerhans cell histiocytosis, 163 late adverse effects, 112
    myelopoiesis (TAM) adverse effects, 24t liver tumours, 15 VM-​26 (teniposide), 24t
    Tanner scoring, 109 retinoblastoma, 274 non-​Hodgkin’s lymphoma, 16 VMA (vanillylmandelic acid), 245
    target toxicity level (TTL), 39 total body irradiation (TBI), 23 renal tumours, 13 VOD (venous occlusive
    taxanes, 27 acute myeloid leukaemia, 56 rhabdomyosarcoma, 15, 15f disease), 237
    TBI see total body irradiation (TBI) ALL, 56 Wilm’s tumour, 234 von Hippel-​Lindau syndrome, 170t
    T-​cell depletion, allogeneic HSCT, 65 allogeneic HSCT, 55 UNC13D gene, 165 von Recklinghausen’s disease, 231
    T-​cell replete haploidentical graft, late adverse effects, 108, 113 UN Convention of the Rights of the vorinostat, 29
    allogeneic HSCT, 55 reduced-​intensity Child (UNCRC), 44 VP-​16 see etoposide (VP-​16)
    99mTc-​MDP bone scans, 221 conditioning, 56 unilateral retinoblastoma, 270 VZV vaccination, 73
    290 Index

    WBC see white blood-​cell diagnosis, 234, 235b WNT medulloblastoma, 190 obesity, 25
    (WBC) count differential diagnosis, 234 World Child Cancer, 95 palliative care, 83
    white blood-​cell (WBC) count epidemiology, 2t, 231 World Health Organization (WHO) three-​step analgesia ladder, 85
    AML prognosis, 128 follow-​up, 237, 238t AML classification, 126–​127 WT1/​2/​3/​4/​5 genes, 231
    myelodysplastic syndrome, 135 incidence, 4 choroid plexus tumours, 203 WVI (whole ventricular
    tumour lysis syndrome, 76 LMIC, 94 CNA tumour classification, 189 irradiation), 199
    WHO see World Health molecular biology, 231 craniopharyngiomas, 200
    Organization (WHO) pathogenesis, 231–​232 embryonal tumours, 188 xanthogranuloma, systemic
    whole ventricular irradiation pathology, 232, 233b germ-​cell tumours juvenile, 166
    (WVI), 199 prevention, 234 classification, 254t xeroderma pigmentosum, 11
    Wilms, Max, 231 prognostic factors, 233–​234 high-​grade gliomas, 182 X-​linked lymphoproliferative
    Wilm’s tumour (nephroblastoma), relapse treatment, 237 Hodgkin’s lymphoma syndrome (XLP), 165
    231–​237 risk factors, 231, 232t classification, 157 X-​ray exposure, 10
    adult treatment, 237 survival, 4, 5t, 6, 6t LMIC, 95
    aetiology, 231 treatment, 234–​237, 235t, 236f low-​grade gliomas, 179 yolk-​sac tumours (YSTs), 198, 253
    bilateral, 237 Wiskott–​Aldrich syndrome, myelodysplastic syndrome, 134 young adults see adolescents and
    clinical symptoms, 234, 234t 139–​140 non-​Hodgkin’s lymphoma, 142 young adults (AYA)

    You might also like