Barbara J Bain - Haemoglobinopathy Diagnosis

Haemoglobinopathy Diagnosis
Haemoglobinopathy
Diagnosis
Barbara J. Bain MBBS FRACP FRCPath
Professor of Diagnostic Haematology
St Mary’s Hospital Campus of Imperial College London, UK
Honorary Consultant Haematologist

St Mary’s Hospital, London, UK
Third Edition
This edition first published 2020
© 2020 John Wiley & Sons Ltd
Edition History
© Barbara J. Bain (2e, 2006) Published by Blackwell Publishing Ltd.
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, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice on how to obtain permission to
reuse material from this title is available at http://www.wiley.com/go/permissions.
The right of Barbara J. Bain to be identified as the author of this work has been asserted in accordance with law.
Registered Offices
John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA
John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK
Editorial Office
9600 Garsington Road, Oxford, OX4 2DQ, UK
For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com.
Wiley also publishes its books in a variety of electronic formats and by print‐on‐demand. Some content that appears in standard print
versions of this book may not be available in other formats.
Limit of Liability/Disclaimer of Warranty

The contents of this work are intended to further general scientific research, understanding, and discussion only and are not intended
and should not be relied upon as recommending or promoting scientific method, diagnosis, or treatment by physicians for any particular
patient. In view of ongoing research, equipment modifications, changes in governmental regulations, and the constant flow of information
relating to the use of medicines, equipment, and devices, the reader is urged to review and evaluate the information provided in the
package insert or instructions for each medicine, equipment, or device for, among other things, any changes in the instructions or
indication of usage and for added warnings and precautions. While the publisher and authors have used their best efforts in preparing this
work, they make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically
disclaim all warranties, including without limitation any implied warranties of merchantability or fitness for a particular purpose. No
warranty may be created or extended by sales representatives, written sales materials or promotional statements for this work. The fact that
an organization, website, or product is referred to in this work as a citation and/or potential source of further information does not mean
that the publisher and authors endorse the information or services the organization, website, or product may provide or recommendations
it may make. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and
strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate. Further, readers
should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is
read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to
special, incidental, consequential, or other damages.
Library of Congress Cataloging‐in‐Publication Data
Names: Bain, Barbara J., author.

Title: Haemoglobinopathy diagnosis / Barbara J. Bain, MBBS, FRACP, FRCPath,
Professor of Diagnostic Haematology, St. Mary’s Hospital Campus of
Imperial College, London, UK, Honorary Consultant Haematologist, St.
Mary’s Hospital, London, UK.
Description: Third edition. | Hoboken : Wiley, 2020. | Includes
bibliographical references and index.
Identifiers: LCCN 2020000661 (print) | LCCN 2020000662 (ebook) | ISBN
9781119579953 (hardback) | ISBN 9781119579984 (adobe pdf) | ISBN
9781119579991 (epub)
Subjects: LCSH: Hemoglobinopathy–Diagnosis.
Classification: LCC RC641.7.H35 B73 2020 (print) | LCC RC641.7.H35
(ebook) | DDC 616.1/51075–dc23
LC record available at https://lccn.loc.gov/2020000661
LC ebook record available at https://lccn.loc.gov/2020000662
Cover images: blue purple abstract © Pobytov/Getty Images, graph on purple background courtesy of Barbara Bain
Cover design by Wiley
Set in 9/12pt Palatino by SPi Global, Pondicherry, India
10 9 8 7 6 5 4 3 2 1
Contents
Preface vii
Abbreviations and glossary viii
1 Haemoglobin and the genetics of haemoglobin
synthesis 1
2 Laboratory techniques for the identification of
abnormalities of globin chain synthesis 30
3 α, β, δ and γ thalassaemias and related conditions 85
4 Sickle cell haemoglobin and its interactions with other
variant haemoglobins and with thalassaemias 185
5 Other significant haemoglobinopathies 261
6 Acquired abnormalities of globin chain synthesis or
haemoglobin structure 325
7 Organisation of a haemoglobinopathy diagnostic
service 348
8 Self-assessment: test cases 365
Appendix: electronic resources 407
Index 411
v
Preface
This book is dedicated to the past and present relevant diagnostic service. It is with much pleasure
scientific staff of the haematology departments of that I dedicate this new edition to these colleagues
Princess Alexandra Hospital, Brisbane, Australia and friends.
and St Mary’s Hospital, Paddington, London. It I should also like to acknowledge many other
was the former group who first awakened my inter- colleagues throughout the world who have helped
est in this field. They also suggested that there was in diverse ways, including those who have contrib-
a need for a practical book on the laboratory diag- uted images. They are individually acknowledged
nosis of haemoglobinopathies and that I might be in the figure legends. Dr Barbara Wild kindly
the person to write it. The second group, and in par- reviewed Chapter 2.
ticular Lorry Phelan, for over four decades shared
my pleasure in solving diagnostic problems and, Barbara J. Bain
at the same time, providing an accurate, clinically London, 2019
vii
Abbreviations and glossary
α the Greek letter, alpha γ thalassaemia a thalassaemic condition resulting

α chain the α globin chain, which is required for from reduced synthesis of γ globin chain
synthesis of haemoglobins A, F and A2 and also δ the Greek letter, delta
the embryonic haemoglobin, Gower 2 δ chain a β‐like globin chain, which forms part of
α gene one of a pair of genes on chromosome 16, haemoglobin A2
HBA1 and HBA2, that encode α globin δ gene a gene of the β cluster on chromosome 11,
α thalassaemia a group of thalassaemias character- HBD, that encodes δ globin
ised by absent or reduced α globin chain synthe- δ thalassaemia reduced synthesis of δ globin and
sis, usually resulting from deletion of one or more therefore of haemoglobin A2
of the α globin genes; less often it results from ε the Greek letter, epsilon
altered structure of an α gene or mutation of the ε chain the epsilon globin chain, which is synthe-
locus control gene, LCRA, or genes encoding sised during early embryonic life and forms part
trans‐acting factors of haemoglobins Gower 1 and Gower 2
α0 thalassaemia a thalassaemic condition in which ε gene a gene of the α globin cluster on chromosome
there is no α globin chain translation from one or 16, HBE1, that encodes ε globin chain
both copies of chromosome 16 ψ the Greek letter, psi, used to indicate a pseudogene
α+ thalassaemia a thalassaemic condition in which ζ the Greek letter, zeta
there is reduced translation of α chain from one or ζ chain the ζ globin chain that is synthesized in
both copies of chromosome 16 intrauterine life and that forms part of haemo-
β the Greek letter, beta globins Gower 1, Portland 1 and Portland 2
β chain the β globin chain, which forms part of ζ gene a gene of the α globin gene cluster on chro-
haemoglobin A and haemoglobin Portland 2 and mosome 16, HBZ, that encodes ζ globin chain
is the only globin chain in the abnormal haemo- 2,3‐DPG 2,3‐diphosphoglycerate; a small molecule
globin, haemoglobin H that interacts with haemoglobin, decreasing its
β gene the gene on chromosome 11, HBB, that oxygen affinity
encodes β globin 3′ the end of a gene where transcription ceases
β thalassaemia a thalassaemia characterised by 5′ the end of a gene where transcription starts
reduced β globin synthesis, usually caused by acquired a condition that is not present at birth or is
mutation of a β globin gene; less often it results not inherited
from gene deletion or from deletion or mutation affinity the avidity of haemoglobin for oxygen
of the locus control region, LCRB AIDS acquired immune deficiency syndrome
γ the Greek letter, gamma ala δ‐aminolaevulinic acid, the first compound
γ chain the γ globin chain, which forms part of fetal formed during the process of haem synthesis
haemoglobin (haemoglobin F) and the embryonic AML acute myeloid leukaemia
haemoglobin, haemoglobin Portland 1, and is the ARMS amplification refractory mutation system, a
only globin chain in the abnormal variant haemo- PCR technique used, for example, for the detec-
globin Bart’s tion of mutations causing β thalassaemia; it
γ gene one of a pair of very similar genes on chromo- employs two primer sets, one amplifying normal
some 11, HBG1 and HBG2, encoding γ globin chain sequences and one abnormal sequences
viii
Abbreviations and glossary ix
balanced polymorphism the stable persistence of CO2 carbon dioxide, the molecule composed of one
two or more alleles of a gene in a significant pro- atom of carbon combined with two atoms of
portion of a population; a potentially deleterious oxygen
allele may show balanced polymorphism if the codon a triplet of nucleotides that encodes a specific
heterozygous state conveys an advantage amino acid or serves as a termination signal;
base a ring‐shaped organic molecule containing there are 61 codons encoding 20 amino acids and
nitrogen, which is a constituent of DNA and 3 codons that act as termination or STOP codons
RNA; DNA contains four bases – adenine, gua- congenital present at birth, often but not necessar-
nine, cytosine and thymine; RNA contains four ily inherited
bases – adenine, guanine, cytosine and uracil cooperativity the interaction between the four glo-
Bohr effect the effect of pH on oxygen affinity; the bin monomers that makes possible the Bohr effect
alkaline Bohr effect is the reduction of oxygen and the sigmoid shape of the oxygen dissociation
affinity of haemoglobin as pH falls from above to curve
below the physiological pH; there is also an acid CT computed tomography
Bohr effect which is a rise of oxygen affinity as the CV coefficient of variation
pH falls further, at a pH level that is incompatible DCIP test a screening test for haemoglobin E using
with life dichlorophenolindophenol
bp base pair, the pairing of specific bases, e.g. ade- deletion loss of part of a chromosome, which may
nine with thymine, in the complementary strands include all or part of a globin gene
of the DNA double helix deoxyhaemoglobin haemoglobin that is not com-
CAP 7‐methyl guanosine cap, added to RNA mole- bined with O2
cule during processing DGGE denaturing gradient gel electrophoresis, a
carbonic anhydrase a red cell enzyme that is the molecular genetic technique for locating a muta-
second most abundant red cell protein after hae- tion prior to precise analysis
moglobin; it may be apparent on haemoglobin DNA deoxyribonucleic acid, the major constituent
electrophoretic strips if a protein rather than a of the nucleus of a cell; a polynucleotide strand
haem stain is used that is able to replicate and that codes for the
carboxyhaemoglobin haemoglobin that has been majority of proteins synthesised by the cell; the
chemically altered by combination with carbon DNA molecule is a double helix of two comple-
monoxide mentary intertwined polynucleotides
CE capillary electrophoresis EDTA ethylene diamine tetra‐acetic acid
CE‐HPLC cation‐exchange high performance liq- EKLF erythroid Krüppel‐like factor
uid chromatography electrophoresis separation of charged suspended
chromatography a method of separating proteins particles such as proteins by application to a
from each other by means of physical characteris- membrane or gel followed by exposure to a charge
tics, such as molecular weight, charge or hydro- gradient, e.g. haemoglobin electrophoresis
phobicity, or by means of differing affinity for ELISA enzyme‐linked immunosorbent assay
lectins, antibodies or other proteins; in column elution removal of an absorbed substance from a
chromatography the proteins move through an chromatography column or membrane
absorbent column and emerge after different enhancer a DNA sequence that influences the pro-
periods of time moter of a nearby gene to increase transcription;
cis on the same chromosome (see also trans) an enhancer acts on a gene in cis and may be sited
cis‐acting a DNA sequence that affects the expression upstream, downstream or within a gene
of a gene on the same chromosome but not on the exon a part of a gene that is represented in mature
homologous chromosome (see also trans‐acting) messenger RNA; most genes are composed of
CO carbon monoxide, the molecule composed of exons and non‐translated introns
one carbon atom and one oxygen atom, formed FAB classification French–American–British
by combustion of hydrocarbons classification
x Abbreviations and glossary
FBC full blood count haemoglobin Bart’s hydrops fetalis a fatal condi-
Fe iron tion of a fetus or neonate with no α genes and
Fe2+ ferrous or bivalent iron consequently no production of haemoglobins A,
Fe3+ ferric or trivalent iron A2, F or Gower 2
fetal haemoglobin see haemoglobin F haemoglobin C a variant haemoglobin with an
G6PD glucose‐6‐phosphate dehydrogenase amino acid substitution in the β chain, mainly
GAP‐PCR a PCR technique in which there is ampli- found in those of African ancestry
fication across a ‘gap’ created by a deletion haemoglobin Constant Spring a variant haemoglo-
GATA1 an erythroid‐specific transcription factor bin with a structurally abnormal α chain that is
gene the segment of DNA that is involved in pro- synthesised at a reduced rate, leading to α
ducing a polypeptide chain; it includes regions thalassaemia
preceding and following the coding region (5′ haemoglobin D the designation of a group of
and 3′ untranslated regions) as well as interven- haemoglobin variants, some α chain variants

ing sequences (introns) between individual and some β chain variants, that have the same
coding segments (exons); genes mediate inherit- mobility as haemoglobin S on electrophoresis at
ance; they are located on nuclear chromosomes alkaline pH
or, for a minority of genes, in a mitochondrion haemoglobin dissociation curve a plot of percent-
genetic code the relationship between a triplet of age saturation of haemoglobin against partial
bases, called a codon, and the amino acid that it pressure of oxygen
encodes haemoglobin E a variant haemoglobin with an
genotype the genetic constitution of an individual amino acid substitution in the β chain, mainly
(cf. phenotype) found in South‐East Asia and parts of the Indian
globin the protein part of the haemoglobin subcontinent
molecule, usually composed of two pairs of
haemoglobin F fetal haemoglobin, the major hae-
non‐identical chains, e.g. two α chains and two moglobin of the fetus and neonate, having two α
β chains chains and two γ chains; also present as a very
H+ a proton minor component in most adults and as a larger
haem a porphyrin structure that contains iron and proportion in a minority
that forms part of the haemoglobin molecule haemoglobin G the designation of a group of hae-
haemoglobin a complex molecule composed of moglobin variants, some α chain variants and
four globin chains, each one enclosing a haem some β chain variants, that have the same
group mobility as haemoglobin S on electrophoresis at
haemoglobin A the major haemoglobin component alkaline pH
present in most adults, having two α and two β haemoglobin Gower 1 an embryonic haemoglobin,
chains having two ζ chains and two ε chains
haemoglobin A1c glycosylated haemoglobin A Haemoglobin Gower 2 an embryonic haemoglo-
haemoglobin A2 a minor haemoglobin component bin, having two α chains and two ε chains
present in most adults and, as an even lower pro- Haemoglobin H a variant haemoglobin with four β
portion of total haemoglobin, in neonates and chains and no α chains, present in haemoglobin H
infants, having two α chains and two δ chains disease and, in small quantities, in α thalassaemia
haemoglobin A2′ a haemoglobin A2 variant, also trait
known as haemoglobin B2 haemoglobin H disease a haemoglobinopathy
haemoglobin Bart’s an abnormal haemoglobin caused by marked underproduction of α chains,
with four γ chains and no α chains, present as the often consequent on deletion of three of the four
major haemoglobin component in haemoglobin α genes
Bart’s hydrops fetalis and as a minor component haemoglobin I a group of variant haemoglobins
in neonates with haemoglobin H disease or α that move more rapidly than haemoglobin A on
thalassaemia trait electrophoresis at alkaline pH
Abbreviations and glossary xi
haemoglobin J a group of variant haemoglobins HIV human immunodeficiency virus

that move more rapidly than haemoglobin A but homologue an equivalent or similar structure; the
more slowly than haemoglobin I on electrophore- α1 and α2 genes are homologues as are the two
sis at alkaline pH copies of a chromosome
haemoglobin K a group of variant haemoglobins homologous being equivalent or similar to another
moving between A and J on electrophoresis at homology the presence of structural similarity,
alkaline pH implying a common remote origin; the δ and β
haemoglobin Lepore a number of variant haemo- genes show partial homology
globins resulting from the fusion of part of a δ homozygosity the state of having two identical
globin gene with part of a β globin gene, giving a alleles of a specified autosomal or X chromosome
δβ fusion gene and a fusion protein that combines gene
with α globin to form haemoglobin Lepore homozygous having two identical alleles of a speci-
haemoglobin M a variant haemoglobin that fied autosomal gene or, in a female, two identical
oxidises readily to methaemoglobin alleles of an X chromosome gene
haemoglobin N a group of variant haemoglobins HPFH hereditary persistence of fetal haemoglobin
moving between J and I on electrophoresis at HPLC high performance liquid chromatography, a
alkaline pH method of separating proteins, such as haemo-
haemoglobin O‐Arab a β chain variant haemoglo- globin variants, from each other on the basis of
bin moving near C at alkaline pH and near S at characteristics such as size, hydrophobicity and
acid pH ionic strength; a solution of proteins is eluted
haemoglobinopathy an inherited disorder result- from a specially designed column by exposure to
ing from synthesis of a structurally abnormal various buffers, different proteins emerging after
haemoglobin; the term can also be used to encom- varying periods of time
pass the thalassaemias in which there is a reduced HS1, HS2, HS3, HS4 hypersensitive sites 1, 2, 3,
rate of synthesis of one of the globin chains and 4, upstream of the β globin gene cluster
haemoglobin Portland 1 an embryonic haemoglo- HS −40 an upstream enhancer of α globin gene
bin, having two ζ chains and two γ chains transcription
haemoglobin Portland 2 an abnormal embryonic HVR hypervariable region
haemoglobin, having two ζ chains and two β ICSH International Council for Standardization in
chains, present in some severe α thalassaemia Haematology
syndromes IEF isoelectric focusing, the separation of proteins
haemoglobin S sickle cell haemoglobin, a variant in an electric field as they move through a pH
haemoglobin with a tendency to polymerise at gradient to their isoelectric points
low oxygen tension, causing erythrocytes to inherited a characteristic that is transmitted from a
deform into the shape of a sickle parent, by means of genes that form part of
Hb haemoglobin concentration nuclear or mitochondrial DNA
Hct haematocrit initiation (i) the process by which RNA transcrip-
HDW haemoglobin distribution width tion from a gene commences; (ii) the process
heteroduplex analysis a molecular genetic tech- by which protein translation from mRNA
nique for locating a mutation prior to precise commences
analysis initiation codon the three nucleotide codon (ATG)
heterozygosity the state of having two different at the 5′ end of a gene that is essential to permit
alleles of a specified autosomal gene or, in a initiation of transcription of a gene, i.e. initiation
female, two different alleles of an X chromosomal of polypeptide synthesis
gene insertion the insertion of a DNA sequence, e.g. from
heterozygous having two different alleles of a one chromosome into another
specified autosomal or X chromosome gene intervening sequence (IVS) an intron
xii Abbreviations and glossary
intron a sequence of DNA in a gene that is not rep- PAS stain periodic acid–Schiff stain
resented in processed messenger RNA or in the PCR polymerase chain reaction, a method of
protein product making multiple copies of a DNA sequence
inversion the reversal of the normal position of a PCV packed cell volume, haematocrit
DNA sequence on a chromosome phenocopy a condition that simulates an inherited
isoelectric point the pH at which a protein has no condition; a phenocopy may be acquired or may
net charge be a genetic characteristic that simulates another
kb kilobase, a unit for measuring the length of phenotype the characteristics of an individual,
DNA; one kilobase is 1000 nucleotide base pairs which may be determined by the genotype, or
kDa kilodalton, a unit for measuring molecular may be an acquired characteristic (cf. genotype)
weight; one kilodalton is 1000 daltons Po2 partial pressure of oxygen
LCR locus control region, a DNA sequence polymorphism the occurrence of a variant form of a
upstream of genes of the α or β globin cluster that gene in a significant proportion (at least 1%) of a
enhances transcription of the genes of the cluster, population
LCRA and LCRB control the α and β gene clusters, promoter a sequence of DNA at the 5′ end of a gene
respectively that is essential for initiation of transcription
LDH lactate dehydrogenase pseudogene a non‐functioning homologue of a
MCH mean cell haemoglobin gene
MCHC mean cell haemoglobin concentration purine one of the two types of nitrogenous base
MCV mean cell volume found in nucleic acids; purines have a double
MDS myelodysplastic syndrome ring structure (see also pyrimidine)
methaemoglobin oxidised haemoglobin, which pyrimidine one of the two types of nitrogenous
does not function in oxygen transport base found in nucleic acids; pyrimidines have a
MGG May–Grünwald–Giemsa (stain) single ring structure (see also purine)
mis‐sense mutation a mutation that leads to the RBC red blood cell count
encoding of a different amino acid RDW red cell distribution width, a measure of
mRNA messenger RNA, ribonucleic acid that is anisocytosis
transcribed in the nucleus, on a DNA template, restriction endonuclease an enzyme that recog-
and moves to the cytoplasm, becoming attached nises specific sequences in a DNA molecule and
to ribosomes and serving as a template for syn- cleaves the molecule in or very near the recogni-
thesis of proteins tion site
MS mass spectrometry, electrospray ionization restriction fragment a fragment of DNA produced
mass spectrometry, a method for determining the by cleavage by a restriction endonuclease
mass and the charge of a molecule RFLP restriction fragment length polymorphism,
NO nitric oxide variation between homologous chromosomes
nonsense mutation a mutation that leads to no with regard to the length of DNA fragments pro-
amino acid being encoded that therefore func- duced by application of a specific restriction
tions as a STOP or termination codon, leading to endonuclease; can be used for the demonstration
synthesis of a truncated polypeptide chain of heterozygosity or for demonstration of a
NRBC nucleated red blood cells specific gene that removes or creates a specific
O2 oxygen cleavage site
ORF open reading frame ribosome a cytoplasmic structure on which pro-
oxyhaemoglobin haemoglobin combined with O2 teins are translated from messenger RNA;
P50 Po2 at which haemoglobin is half saturated ribosomes may be free within the cytosol or form
Pao2 partial pressure of oxygen in arterial blood part of the rough endoplasmic reticulum
partial pressure of oxygen that part of the total RNA ribonucleic acid, a polynucleotide in which
blood gas pressure exerted by oxygen the nitrogenous bases are adenine, guanine,
Abbreviations and glossary xiii
c ytosine and uracil and the sugar is ribose; RNA eterozygosity or to deletion of one or two of
h
is produced in the nucleus and in mitochondria the four α genes
from DNA templates trait a term applied to heterozygosity for an inher-
rRNA ribosomal RNA, RNA that, together with ited characteristic; in the case of disorders of
protein, constitutes the ribosomes globin genes, the term would not be used if

sickle cell an erythrocyte that becomes sickle‐ or heterozygosity were associated with a significant
crescent‐shaped as a result of polymerisation of phenotypic abnormality; rather it is used when
haemoglobin S homozygosity or compound heterozygosity
sickle cell anaemia the disease resulting from produces a clinically significant abnormality but
homozygosity for haemoglobin S simple heterozygosity does not
sickle cell disease a group of diseases including trans having an influence on a DNA sequence on
sickle cell anaemia and various compound hete- another chromosome (see also cis)
rozygous states in which clinicopathological trans‐acting a DNA sequence that affects the
effects occur as a result of sickle cell formation expression of a gene on another chromosome
(preferred definition but sometimes used as a (see also cis‐acting)
synonym for sickle cell anaemia) transcript an RNA molecule, corresponding to one
sickle cell trait heterozygosity for the βS gene that gene, transcribed from nuclear DNA
encodes the β chain of haemoglobin S transcription the synthesis of RNA on a DNA
SOP standard operating procedure template
splicing the process by which RNA sequences, transcription factor a protein capable of enhancing
corresponding to introns in the gene, are removed transcription of one or more genes
during processing of RNA translation the synthesis of protein from a mRNA
SSP stage selector protein template
sulphaemoglobin haemoglobin that has been irre- tRNA transfer RNA, RNA molecules that bind to
versibly oxidised and chemically altered by drugs specific amino acids and transport them to ribo-
or chemicals with incorporation of a sulphur somes; there they bind to specific mRNA
atom into the haemoglobin molecule sequences, leading to incorporation of amino
thalassaemia a disorder, usually inherited, in which acids into peptide chains in the sequence
one or more of the globin chains incorporated specified by the mRNA
into a haemoglobin molecule or molecules is unstable a term applied to a haemoglobin that is
synthesised at a reduced rate abnormally prone to post‐translational structural
thalassaemia intermedia a genetically heterogene- alteration, which may include loss of the normal
ous thalassaemic condition that is moderately tertiary or quaternary structure
severe but nevertheless does not require regular UTR untranslated region
blood transfusions to sustain life variant a term applied to any haemoglobin other
thalassaemia major thalassaemia that is incompat- than haemoglobins A, A2, F and the normal
ible with more than a short survival in the absence embryonal haemoglobins
of blood transfusion WBC white blood cell count
thalassaemia minor an asymptomatic thalassae- yolk sac part of an embryo, the initial site of forma-
mic condition, attributable to β thalassaemia tion of blood cells
1 Haemoglobin and the genetics of
haemoglobin synthesis
Haemoglobins and their structure

is converted back to haemoglobin mainly by
and function
the action of the enzyme NADH‐cytochrome
The haemoglobin molecule contained within red b5‐methaemoglobin reductase.
blood cells is essential for human life, being the The haemoglobin molecule can also combine with
means by which oxygen is transported to the tis- CO2 and is responsible for about 10% of CO2 transport
sues. Other functions include the transport of car- from the tissues to the lungs; transport is by reversible
bon dioxide (CO2) and a buffering action (reduction carbamation of the N‐terminal groups of the α chains
of the changes in pH that would otherwise be of haemoglobin. Because carbamated haemoglobin
expected when an acid or an alkali enters or is gen- has a lower oxygen affinity than the non‐carbamated
erated in a red cell). A normal haemoglobin mole- form, binding of the CO2 produced by the metabolic
cule has a molecular weight of 64–64.5 kDa and is processes in tissues facilitates oxygen delivery to
composed of two dissimilar pairs of polypeptide tissues. In addition, non‐oxygenated haemoglobin
chains, each of which encloses an iron‐containing can carry more CO2 than oxygenated haemoglobin so
porphyrin designated haem (Fig. 1.1). Haem is that unloading of oxygen to the tissues facilities the
essential for oxygen transport while globin serves to uptake and transport of CO2. Because of its buffering
protect haem from oxidation, renders the molecule action (mopping up of protons, H+), haemoglobin also
soluble and permits variation in oxygen affinity. The contributes to keeping CO2 in the soluble bicarbonate
structure of the haemoglobin molecule produces an form and thus transportable. The reaction CO2 + H2O
internal environment of hydrophobic radicals, → HCO3− + H+ is facilitated.
which protects the iron of haem from water and thus Haemoglobin also has a role in nitric oxide (NO)
from oxidation. External radicals are hydrophilic transport and metabolism, being both a scavenger
and thus render the haemoglobin molecule soluble. of NO and an active transporter. NO is produced in
Both haem and globin are subject to modification. endothelial cells and neutrophils by the action of
The iron of haemoglobin is normally in the ferrous nitric acid synthase [2–5]. It has a very high affinity
form (Fe2+). Haem is able to combine reversibly with for oxyhaemoglobin so that blood levels are a bal-
oxygen so that haemoglobin can function as an oxy- ance between production and removal by binding
gen‐transporting protein. Oxidation of iron to the to oxyhaemoglobin. NO is a potent vasodilator, but
ferric form (Fe3+) is a less readily reversible reaction, this effect is limited by its binding to haemoglobin.
converting haem to haematin and haemoglobin to The iron atom of a haem group of oxyhaemoglobin
methaemoglobin, a form of haemoglobin that can- (preferentially the haem enclosed in the haem
not transport oxygen. Auto‐oxidation of haemoglo- pocket of an α chain), binds NO. A haemoglobin
bin to methaemoglobin is a normal process. About molecule with NO bound to two haem groups strik-
3% of haemoglobin undergoes this process each day ingly favours the deoxy conformation so oxygen is
with about 1% (0.4–1% in one study) of haemoglo- readily released. NO–haemoglobin is subsequently
bin being methaemoglobin [1, 2]. Methaemoglobin converted to methaemoglobin with release of NO
Haemoglobinopathy Diagnosis, Third Edition. Barbara J. Bain.

© 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.
1
2 Chapter 1
β2 β1
A F A F
E E H
H
B C B
G G
D C G D C
B
E
H
Fig. 1.1 Diagrammatic representation
F
A Tertiary structure of the quaternary structure of
of β chain haemoglobin and the tertiary
α2 α1 structure of a haemoglobin monomer
(a β globin chain containing a haem
Quaternary structure
of haemoglobin A molecule group). Upper case letters indicate
homologous α helixes.
and production of nitrate ions, which are excreted. As a result of the synthesis of different globin
Since deoxyhaemoglobin has a much lower affinity chains at different stages of life (Fig. 1.2) there is a
for NO, hypoxic conditions could leave more NO free difference in the type of haemoglobin present in red
and lead to vasodilation, which is of potential physi- cells between adult life and the fetal and neonatal
ological benefit. In addition, deoxyhaemoglobin can periods (Table 1.1, Fig. 1.3). In adults, 96–98% of
convert nitrite to NO, again favouring vasodilation. haemoglobin is haemoglobin A (A = adult), which
Nitric oxide also causes S‐nitrosylation of a con- has two alpha (α) chains and two beta (β) chains.
served cysteine residue (Cys93, E15) of the β globin The name ‘haemoglobin A’ was given by Linus
chain of oxyhaemoglobin to form S‐nitrosohaemo- Pauling and colleagues in 1949 when they discov-
globin. This occurs in the lungs. In this circumstance, ered that asymptomatic carriers of sickle cell dis-
the bioactivity of NO may be retained with NO being ease had two different haemoglobins, which they
delivered to low molecular weight thiol‐containing designated haemoglobin A and haemoglobin S [6].
molecules to reach target cells such as the smooth A minor haemoglobin, haemoglobin A2, has two α
muscle of blood vessels. Oxygenation of haemoglobin chains and two delta (δ) chains. Its existence was
favours S‐nitrosylation. Conversely, deoxygenation first reported in 1955 by Kunkel and Wallenius [7].
favours release of NO. This may be an important They noted that haemoglobin A2 was increased in
physiological process with NO being released in thalassaemia minor and reduced or absent in neo-
peripheral tissues where it can facilitate arteriolar dila- nates. A very minor haemoglobin in adults but the
tion. The oxy form of S‐nitrosohaemoglobin is a vaso- major haemoglobin during fetal life and the early
constrictor, whereas the deoxy form is a vasodilator. neonatal period is haemoglobin F or fetal haemo-
Lack of oxygen could thus again favour vasodilation. globin, which has two α chains and two gamma (γ)
In normal circumstances, the ability of haemoglo- chains. There are two species of haemoglobin F, des-
bin to scavenge or destroy NO is reduced by the bar- ignated Gγ and Aγ, with glycine and alanine, respec-
rier to NO diffusion that is provided by the red cell tively, at position 136 of the γ chain. In addition, the
membrane. However, in haemolytic anaemia with A
γ chain shows polymorphism at position 75, which
increased free plasma haemoglobin, binding and may be occupied by threonine rather than the more
inactivation can be almost immediate, leading to common isoleucine [8]. This polymorphism was
impaired vascular responses to NO [5]; inactivation previously referred to as haemoglobin F‐Sardinia.
of NO by haemoglobin in the plasma thus contrib- In the early embryo, haemoglobin is synthesised
utes to the pulmonary hypertension that can be a in the yolk sac and specific embryonic haemoglobins
feature of sickle cell anaemia and also to the hyper- are produced, called Gower 1, Gower 2 and Portland
tension that has been observed with some haemo- (or Portland 1). These contain globin chains that are
globin‐based blood substitutes. synthesised in significant amounts only during
Haemoglobin and the genetics of haemoglobin synthesis 3
Yolk sac Liver Spleen Bone marrow
α
100
Globin synthesis (%)

β
80
γ
60
40
20 ε, ζ δ
of the sites and rates of synthesis of 10 20 30 40 20 40
different globin chains in the embry- Weeks from conception Weeks from birth
onic and fetal periods and during
Birth
infancy.
Table 1.1 Haemoglobins normally present during adult, fetal and embryonic periods of life.
Haemoglobin Globin chains Period when normally present

species
A α2β2* Major haemoglobin in adult life
A2 α2δ2 Minor haemoglobin in adult life; even more minor in late fetal and neonatal life
F α2 γ2 or α2 γ2
G A
Minor haemoglobin in adult life, major haemoglobin in fetal life with a declining
percentage through the neonatal period
Gower 1 ζ2ε2 Significant haemoglobin during early intrauterine life
Gower 2 α2ε2 Significant haemoglobin during early intrauterine life
Portland or ζ2γ2 Significant haemoglobin during early intrauterine life

Portland 1†
* Can also be designated αA2βA2 to distinguish the globin chains of haemoglobin A from those of variant haemoglobins.
†
Haemoglobin Portland 2 (ζ2β2) has been observed in α thalassaemia syndromes but is unlikely to occur in significant amounts during
normal development.
embryonic life, specifically zeta (ζ) and epsilon (ε) blasts in the yolk sac. From the 6th week onwards
chains (see Table 1.1). Haemoglobins Gower 1 (ζ2ε2) these same cells start to synthesise α, β and γ chains.
and Gower 2 (α2ε2) were first described by Huehns Starting from about the 10th to 12th week of gesta-
and colleagues in 1961 [9], and are named for Gower tion there is haemoglobin synthesis in the liver and
Street, in London, in which University College the spleen, with production of fetal and later adult
Hospital is situated. Portland 1 (ζ2γ2) was described haemoglobin. Production of the various embryonic,
in 1967 and was so‐named because it was first iden- fetal and adult haemoglobins is synchronous in dif-
tified in the University of Oregon in Portland, ferent sites. Later in intrauterine life, the bone mar-
Oregon [10]. By 5 weeks of gestation, ζ and ε chains row takes over as the main site of haemoglobin
are already being synthesised in primitive erythro- synthesis and increasing amounts of haemoglobin
4 Chapter 1
100
90
80 Haemoglobin F
Total haemoglobin (%)
70
60
50 Haemoglobin A
40
30
20
10 Haemoglobin A2
0
5 16 28 0 3 6 9
Weeks from conception Months of age Fig. 1.3 Diagrammatic representation
Haemoglobins Birth
of the average percentages of various
Gower 1, Gower 2 and haemoglobins present in the
Portland 1 embryonic and fetal periods and
during infancy.
A are produced. In adult life, bone marrow erythro- carboxyhaemoglobin. In normal individuals car-
blasts synthesise haemoglobin A and the minor hae- boxyhaemoglobin comprises 0.2–0.8% of total hae-
moglobins. The embryonic haemoglobins have a moglobin, but in heavy smokers it may be as much
higher oxygen affinity than haemoglobin A, similar as 10–15%. Small amounts of sulphaemoglobin
to that of haemoglobin F [11]. They differ from hae- (<0.5%) [1] and methaemoglobin are also formed in
moglobins A and F in that they continue to bind normal subjects. Methaemoglobin (see earlier) is
oxygen strongly, even in acidotic conditions [11]. In usually less than 1% of total haemoglobin. Other
the case of Gower 2, impaired binding to 2,3‐diphos- post‐translational modifications of globin chains
phoglycerate (2,3‐DPG) is the basis of the increased include carbamylation, pyruvatisation, acetylation
oxygen affinity [12]. and acetaldehyde adduct formation [14].
Haemoglobin can undergo post‐translational Glutathionylation is increased in diabetes mellitus
modification (see also Chapter 6). Glycosylation [15] and by the administration of certain anti‐epi-
occurs with formation of haemoglobins A1a–e, but leptic drugs (phenobarbital and carbamazepine)
principally of haemoglobin A1c. In healthy adults, [16]. Post‐synthetic modification of a haemoglobin
haemoglobin A1c may constitute up to 4–6% of total molecule can also occur as a consequence of a
haemoglobin, but in people with diabetes mellitus mutation in a globin gene; either the abnormal
it can be much higher. It is also increased in people amino acid or an adjacent normal amino acid can
with the acquired immune deficiency syndrome undergo post‐translational conversion to another
(AIDS) [13]. In individuals with a shortened red amino acid (see later). In addition, some abnormal
cell life span it is lower. Another minor fraction, haemoglobins in which there is a mutation of
formed on ageing, is haemoglobin AIII, in which N‐terminal amino acid are particularly prone to
glutathione is bound to the cysteine at β93. acetylation, which occurs co‐translationally [17].
Unmodified haemoglobin can be distinguished by The structure of haemoglobin is highly complex
use of the designation haemoglobin A0. In the fetus and can be viewed at four levels as follows:
about 20% of haemoglobin F shows acetylation of 1 The primary structure is the sequence of the
the γ chain, but this is not a major feature of other amino acids in the polypeptide that constitutes the
normal human globin chains [8]. Exposure to car- globin chain.
bon monoxide, the product of incomplete combus- 2 The secondary structure is the arrangement of the
tion of hydrocarbons, leads to the formation of polypeptide globin chains into α helixes (stabilised by
hydrogen bonds) separated by non‐helical turns. In The interaction between the four globin chains is
the case of the β globin chain there are eight α helixes, such that oxygenation of one haem group alters the
designated A to H, whereas the α globin chain lacks shape of the molecule in such a way that oxygena-
the D helix residues; 70 to 80% of the amino acid resi- tion of other haem groups becomes more likely.
dues of haemoglobin form part of the helixes. This is known as cooperativity and is reflected in
3 The tertiary structure is the arrangement of the the shape of the oxygen dissociation curve (Fig. 1.5).
coiled globin chain into a three‐dimensional structure The cooperativity between the globin chains is
that has a surface haem‐containing pocket between shown diagrammatically in Fig. 1.6. It is consequent
the E and F helixes; binding of haem between two on the fact that in the deoxygenated state the Fe2+
specific histidine residues in the E and F helixes, atom is out of the plane of the porphyrin ring of
respectively (Fig. 1.4), is essential for maintaining the haem. Oxygenation of Fe2+ causes it to move into
secondary and the tertiary structure of haemoglobin. the plane of the porphyrin ring and because of the
4 The quaternary structure is the relationship link between haem and the histidine residues of
between the four globin chains, which is not fixed. globin there is an alteration in the tertiary structure
The strong α1β1 and α2β2 bonds (dimeric bonds) hold of that haemoglobin monomer; this in turn causes
the molecule together in a stable form while the α1β2 the oxygenated monomer to alter its position in
and α2β1 bonds (tetrameric bonds) contribute to relation to other haemoglobin monomers (i.e. the
stability, albeit to a lesser extent than the dimeric quaternary structure of the haemoglobin molecule
bonds, and permit the chains to slide on each other is altered). The oxygenated haemoglobin molecule
and rotate; alteration in the quaternary structure of is smaller than the non‐oxygenated molecule.
haemoglobin is responsible for the sigmoid oxygen Cooperativity between the globin chains is also the
dissociation curve, the Bohr effect and the variation basis of the alkaline Bohr effect (often referred to
of oxygen affinity consequent on interaction with simply as the Bohr effect), which is the reduction of
2,3‐DPG (see later). Contacts between like chains, oxygen affinity that occurs when the pH falls from
α1α2 and β1β2, are also of physiological significance. physiological levels of 7.35 to 7.45 towards 6.0.
Increasing metabolism in tissues lowers the pH
since there is increased production of CO2 and of
carbonic acid and, in addition, in anaerobic condi-
tions there is generation of lactic acid. The Bohr
effect therefore leads to enhanced delivery of oxy-
gen to tissues such as exercising muscle. Similarly,
the quaternary structure of haemoglobin makes
possible the interaction of haemoglobin with
2,3‐DPG, which enhances oxygen delivery.
Synthesis of 2,3‐DPG is increased by hypoxia.
Marked anaemia can cause respiratory alkalosis,
which enhances 2,3‐DPG synthesis, thus compen-
sating to some extent for the anaemia. There is also
increased 2,3‐DPG synthesis in renal failure, again
partly compensating for the anaemia.
Oxygen affinity is reduced not only by acidosis and
increased levels of 2,3‐DPG but also by fever. All these
Fig. 1.4 Diagrammatic representation of a haemoglobin
molecule with a haem group within the haem pocket, effects are likely to be of physiological significance.
showing the relationship of the haem to two histidine Fever increases the metabolic rate so that decreased
residues of the globin chain, designated proximal and oxygen affinity, favouring downloading of O2, is ben-
distal histidines; haem is bound to the proximal histidine eficial in this circumstance. The lower pH in tissues
while O2 is bound to haem and to the distal histidine, favours delivery of oxygen to sites of active metabo-
both histidines being important for the integrity of the lism, whereas the efflux of CO2 in the lungs raises the
haem pocket. pH and favours uptake of oxygen by haemoglobin.
Myoglobin
Haemoglobin H
Haemoglobin Bart’s
100 100
hypothermia
90 alkalosis 90
reduced
Oxygen saturation (%) 80 80
Oxygen saturation (%)

2,3-DPG
70 70
fever
60 60
acidosis
Haemoglobin A
50 increased 2,3-DPG 50
binding of CO2
40 40
to globin
30 30
20 20
10 10
0 0
10 20 30 40 50 60 70 80 90 100 mmHg Oxygen tension
1.3 2.5 5 7.5 10 12.5 13 kPa

Oxygen tension
p50 = 27 mmHg
(a) (b)
Fig. 1.5 Oxygen dissociation curve: (a) normal oxygen dissociation curve indicating the effects of alteration of pH,
body temperature and 2,3‐diphosphoglycerate (2,3‐DPG) concentration on the oxygen affinity of haemoglobin;
(b) a comparison of the hyperbolic oxygen dissociation curve characteristic of myoglobin and of abnormal haemoglobins
that do not exhibit cooperativity with the sigmoid dissociation curve characteristic of haemoglobin A; haemoglobins A2
and F have dissociation curves similar to that of haemoglobin A but further to the right.
2,3-DPG
(extruded)
2,3-DPG
β2 β1 β2 β1
α2 α1 α2 α1
Deoxygenated Oxygenated
quaternary structure quaternary structure
Fig. 1.6 Diagrammatic representation of the effect of oxygenation and deoxygenation on the quaternary structure of
haemoglobin. The haemoglobin dimers (α1β1 and α2β2) are stable, with the dimeric bonds between the α and the β chain
having 34 contacts in both the deoxygenated and oxygenated forms. There are less strong α2β1 and α1β2 tetrameric bonds,
with 17 contacts between the α chain and the β chain, in the deoxy form and a different 17 contacts in the oxy form.
There are also α1α2 bonds with four inter‐chain contacts in the deoxy form only. 2,3‐DPG binds to the β chains (3 contacts
with each chain) only in the deoxy form of the molecule. Oxygenation is associated with breaking and reforming of
tetrameric (α2β1 and α1β2) contacts, breaking of α1α2 contacts, expulsion of 2,3‐DPG and the assumption of a more
compact form of the molecule. In the deoxygenated form the α chains are closer together and there is a cleft between the
β chains whereas in the oxygenated form the α chains are further apart and the β cleft has disappeared.
The oxygen dissociation curve is often shifted to the the mitochondrion for haem synthesis or stored as
right, as a result of acidosis, in chronic renal failure; ferritin within the cytoplasm. The transferrin mole-
this ameliorates the effect of anaemia [18]. It will be cule then detaches from the transferrin receptor and
noted that the acute effect of acidosis and the chronic is released from the cell surface. There is negative
effect of respiratory alkalosis both contribute to feedback control of haem synthesis by haem, which
improved oxygen delivery to tissues. both inhibits ferrochelatase and inhibits acquisition
of iron from transferrin. Reduced cellular uptake of
iron in turn inhibits production of δ‐aminolaevulinic
Genetics of haemoglobin synthesis
acid. Uptake of iron by erythroid cells is enhanced
Haem synthesis takes place in erythroid precursors by iron deficiency and by increased levels of eryth-
from the proerythroblast stage to the reticulocyte ropoietin. Both lead to combination of iron regula-
stage. Eight enzymes, under separate genetic control, tory proteins with iron‐responsive elements in the
are known to be necessary for haem synthesis [19]. mRNA for the transferrin receptor protein. The
Different stages of haem synthesis take place either mRNA is then protected from degradation, leading
in mitochondria or within the cytosol (Fig. 1.7). The to increased expression of transferrin receptors on
first enzymatic reaction and the last three are in the erythroid cell membranes and increased iron uptake.
mitochondrion, whereas the four intermediate enzy- Haem is necessary for normal folding of globin
matic reactions occur in the cytosol. The first rate‐ chains and prevents their precipitation [20]. Variant
limiting step in haem synthesis is formation of haemoglobins with impaired haem binding are unsta-
δ‐aminolaevulinic acid by condensation of glycine ble [20]. Haem is important in the regulation of globin
and succinyl CoA. This reaction is under the control chain synthesis. In haem‐replete cells a protein known
of δ‐aminolaevulinate synthase (ala‐synthase), with as haem‐regulated inhibitor (HRI) is inactive, with the
pyridoxal 5′‐phosphate as a cofactor. The rate of for- result that guanosine diphosphate (GDP) attached to
mation of δ‐aminolaevulinic acid is controlled by an erythroid initiation factor, eIF2, is converted to
iron availability; iron deficiency causes iron regula- guanosine triphosphate (GTP), leading to initiation of
tory proteins to bind to iron‐responsive elements in globin chain synthesis. When haem is deficient, HRI is
the messenger ribonucleic acid (mRNA) for ala‐ activated by a utophosphorylation and maintains eIF2–
synthase with resultant repression of translation. GDP in an inactive form so that globin chain translation
Synthesis of δ‐aminolaevulinic acid is followed by its is not initiated [20]. HRI is likely to be of relevance in β
entry into the cytosol, where two molecules com- thalassaemia, with increased levels resulting from oxi-
bine, under the influence of δ‐aminolaevulinate dant stress lessening the excess α chain synthesis.
dehydrase (ala‐dehydrase), to form porphobilino- Synthesis of α and β globin chains takes place in
gen. Four molecules of porphobilinogen in turn com- erythroid precursors, from the proerythroblast
bine to form uroporphyrinogen III, which is then onwards, and continues to the reticulocyte stage.
modified in two further steps to form coproporphy- Synthesis of δ chains ceases before the reticulocyte
rinogen III. Coproporphyrinogen III enters the mito- stage [21]. Haemoglobin A synthesis thus contin-
chondrion, where it is converted to protoporphyrin ues in reticulocytes, whereas synthesis of haemo-
IX. The final stage is the combination of ferrous (Fe2+) globin A2 has been completed by the late
iron with protoporphyrin IX to form haem, under the erythroblast stage [22].
influence of ferrochelatase. Haem is also referred to Globin chain synthesis takes place on ribosomes in
as ferroprotoporphyrin. the cytoplasm. Genes for globin chain synthesis are
Uptake of iron by erythroid cells is from transfer- located in two clusters, on chromosomes 11 and 16
rin (see Fig. 1.7). A molecule of transferrin with its (Figs 1.8 and 1.9). The α gene cluster is close to the
attached iron first binds to a membrane transferrin telomere of chromosome 16, at 16p13.3. The distance
receptor. The whole complex is internalised in a pro- from the telomere shows polymorphic v ariation,
cess known as endocytosis. Iron is released from its from 170 to 430 kilobases (kb), a kb being 1000 nucle-
carrier within the endocytotic vesicle and, following otide bases. The β gene is at 11p15.4. In addition to
reduction to the ferrous form, is either transferred to the functional globin genes, these clusters contain
8 Chapter 1
Fig. 1.7 Diagrammatic representation of haem synthesis. Tf, transferrin; TfR, transferrin receptor.
16p13.3
11p15.4 HBA1
HBB and
HBA2
Chromosome 16
Fig. 1.8 Diagram of chromosomes 11

and 16 showing the positions of the β Chromosome 11
and α globin gene clusters.
(a)
Chromosome 11
LCRB ε Gγ Aγ ψβ δ β
5ʹ 3ʹ
(b)
Pseudogenes
Chromosome 16
Fig. 1.9 Diagrammatic representation of the LCRA ζ ψζ ψα2 ψα1 α2 α1 θ

α and β globin gene clusters on chromo- 5ʹ 3ʹ
some 11 (a) and chromosome 16 (b).
‘pseudogenes’, which are non‐functional homologues tive invertebrates have only a single globin gene,
of globin genes; they are transcribed but not trans- whereas fish and amphibians have an α and a β gene
lated. The α cluster of chromosome 16 extends over on the same chromosome. Birds have α and β genes
28 kb and contains, in the following order: a ζ gene, on different chromosomes. All the human globin
HBZ (also referred to as ζ2); a pseudo‐ζ gene (ψζ or genes have three coding sequences (exons) and two
ψζ1); two pseudo‐α genes (ψα2 and ψα1); and two α intervening non‐coding sequences (intervening
genes, HBA2 and HBA1, usually designated α2 and sequences or introns) and are flanked by 5′ and 3′
α1. The β cluster on chromosome 11 contains, in the non‐coding sequences (referred to as untranslated
following order: an ε gene, HBE1; two γ genes, HBG2 regions, UTRs) (Fig. 1.10). The two α genes differ in
and HBG1, usually designated Gγ and Aγ, respec- structure in intron 2 and the 3′ UTR but the coding
tively; a pseudo‐β gene (ψβ); a δ gene, HBD; and a β sequences are identical.
gene, HBB. There is wide variability of the α and β As for all genes, coding is by means of triplets of
globin gene clusters between individuals and nucleotides, known as codons, which code for a
groups, with duplications and triplications of ζ, ψζ specific amino acid. Sequences known as promoters
and α being quite common. The overall structure of occur 5′ to each gene. These bind ribonucleic acid
the two clusters are remarkably conserved amongst (RNA) polymerase and transcription factors and
vertebrates and this has led to the hypothesis that all are necessary for the initiation of transcription.
the globin genes, as well as the gene for the unlinked Globin gene promoters share several conserved
but related protein myoglobin, arose from a common deoxyribonucleic acid (DNA) sequences that bind
ancestor by the processes of duplication, unequal crucial transcription factors [23, 24]. These are
crossing over and sequence divergence. Many primi- summarised in Table 1.2.
10 Chapter 1
Fig. 1.10 Diagrammatic representation of ribonucleic acid (RNA) synthesis and processing and β globin chain
synthesis. Although processes are shown sequentially, capping starts soon after transcription has started and
therefore contemporaneously with transcription, whereas polyadenylation necessarily occurs after completion of
transcription.
Table 1.2 The sequences

showing CACCC, CCAAT Gene CACCC CCAAT TATA
and TATA homology homology box homology box homology box
in the promoters of globin
ζ (HBZ) CCAAT TATAAAC
genes. Identical sequences
in different genes are shown α1 and α2 (HBA1 CCAAT CATAAAC
in bold red. and HBA2)
ε (HBE1) CCAAT AATAAAG

G
γ and γ (HBG2
A
CACCC CCAAT/CCAAT AATAAAA
and HBG1)
β (HBB) CACCC CCAAT CATAAAA
δ (HBD) CCAAC CATAAAA
The process by which globin chains are synthe- including GATA1 and histone deacetylase 1
sised is shown diagrammatically in Fig. 1.10. (HDAC1), which binds to a region near the 5′ end of
Transcription is the process by which RNA is the δ‐globin gene, repressing the gene; this interac-
synthesised from a DNA template by the action of tion is critical in the fetal‐to‐adult haemoglobin
RNA polymerase. The entire globin gene, including switch and in γ‐globin gene silencing in adults [28].
the introns and the 5′ and 3′ UTRs, is transcribed. Heterozygous inactivating mutations of KLF1 can
Transcription is controlled by interaction between lead to an increased percentage of haemoglobin F
the genes and transcription factors that bind both to [29], as can haploinsufficiency or downregulation of
promoters and to upstream regulatory elements BCL11A [30]. Inactivating mutation of KLF1 can also
referred to as the β‐locus control region (LCRB) for lead to an increase of haemoglobin A2 [31]. SSP (stage
the β cluster and the α‐locus control region (LCRA) selector protein) is an enhancer of δ and γ chain syn-
for the α cluster. The LCRA has four regulatory thesis [26]. NFE4p22 is an enhancer of Gγ and Aγ
elements, DNase sites HS −48, HS −40, HS −33 and genes [32]. FKLF and FKLF2 are enhancers of the
HS −10, of which HS −40 is the major regulatory ele- embryonic ε gene and the γ genes [32, 33]. In addition
ment. The LCRB includes five erythroid‐specific to transcription factors that are relatively specific to
DNase sites designated HS1, HS2, HS3, HS4 and erythroid cells, globin gene expression is also influ-
HS5, of which HS3 is probably the most important in enced by general transcription factors including
opening the chromatin structure to permit access of AP‐1, Sp1, YY1, USF and TAL‐1/SCL [25–27].
transcription factors, and HS2 is probably the most Expression of the genes of the β cluster is also influ-
important in enhancing globin chain synthesis [25]. enced by histone acetylases, which increase expres-
There are also enhancers within introns of genes and sion [32]. The Gγ and Aγ genes are repressed by histone
downstream of the β and Aγ genes. Trans‐acting fac- deacetylases and histone deacetylase inhibitors
tors, encoded by genes on chromosomes other than such as butyrate upregulate γ gene expression [32].
11 and 16, are vital for the expression of globin genes. Methylation of genes reduces expression and thus the
Relatively erythroid‐specific trans‐activating factors, demethylating action of azacitidine may be the mech-
including GATA1, ZFPM1 (previously known as anism by which it upregulates γ gene expression [32].
FOG1) (which interacts with GATA1 in erythroid Nascent RNA molecules resulting from tran-
and megakaryocytic development), NFE2, KLF1 scription are large and unstable and are modified
(previously known as EKLF), KLF2, SSP, Nrf‐1, Nrf‐2 in the nucleus. Initially the 5′ end acquires a
and LCR‐F1, contribute to regulation of gene expres- 7‐methylguanosine cap, which is probably added
sion by interacting either with the locus control early during transcription, protects the 5′ end of
regions or with the globin gene promoters to increase the molecule from degradation and is required
gene expression [26, 27]. KLF1 (Krüppel‐like factor 1) for initiation of translation; during this ‘capping’
is an enhancer of β chain synthesis and a repressor of process, methylation of adjacent ribose residues
γ chain synthesis, by means of its activation of also occurs. Following the completion of tran-
BCL11A. BCL11A is part of a repressor complex, also scription, the majority of transcripts acquire a 3′
12 Chapter 1
polyadenosine tail with the addition of 75 to termination codons, leading to termination of glo-
several hundred adenylate residues. There is an bin chain synthesis. Transcription thus continues
AAUAA sequence near the 3′ end (within the 3′ until a termination codon, UAA, UAG or UGA, is
UTR), which serves as a signal for 3′ cleaving of encountered. The termination codon is followed
the transcript and polyadenylation. The polyade- by the 3′ UTR.
nylate tail is important for mRNA stability, pro- The rate‐limiting step of globin chain translation
vides a signal for transfer of mRNA from the is the commencement of elongation, which is the
nucleus to the cytoplasm and probably enhances next step after initiation. Transcription from the
translation. Finally the introns are excised to give two α genes is equal up to the 8th week of gesta-
a functional mRNA, which in most cases contains tion, but thereafter the α2 gene becomes dominant
a single continuous open reading frame (ORF), and, in adult life, the ratio of α2 to α1 mRNA is
encoding the sequence of the relevant protein, 2.6–2.8:1 [36]. Translational efficiency differs some-
flanked by 5′ and 3′ UTRs. what so that the α2 gene directs synthesis of about
Molecules of mRNA move from the nucleus to twice as much α chain as the α1 gene. There is more
the cytoplasm, where they bind to ribosomes and α than β mRNA, probably about two and a half
serve as templates for the assembly of the poly- times as much, but β chain synthesis is more trans-
peptide sequences of the globin chain. Each nucle- lationally efficient than α chain synthesis and α
otide triplet serves as a template for a specific chains are therefore produced only slightly in
amino acid that is covalently bound to and trans- excess of β chains [36]. Control of globin chain syn-
ported to the ribosome by transfer RNA (tRNA). thesis is probably mainly at the level of transcrip-
tRNAs are specific both for a nucleotide triplet and tion, with translational control being less
for an amino acid. Amino acids are thus assembled important. Translation is dependent on the pres-
in the correct sequence and are covalently bound ence of haem. In iron deficiency, reduced availabil-
to each other by ribosomal enzymes, forming a ity of haem leads to inactivation of the initiation
polypeptide. This process is known as translation. factor and thus reduced synthesis of globin chains.
An initiation codon, AUG, is essential for the initi- The α and β globin chains are synthesised on dif-
ation of translation; it is the first codon after the 5′ ferent polyribosomes. Combination of free α chain
UTR and encodes methionine. Initiation requires with β chain that is still attached to the polyribo-
the amino acid methionine, tRNA specific for some, to form an αβ dimer, may contribute to
methionine, GTP and an initiation factor. When release of the β chain from the ribosome.
the nascent molecule reaches 20–30 amino acid Incorporation of haem probably occurs after
residues, the methionine is removed through the release from the polyribosome.
action of methionine aminopeptidase; this process Globin mRNA is unusually stable, and transla-
is interfered with when mutation leads to the pres- tion can continue for up to three days after cessation
ence of certain residues in position 1 or even posi- of transcription. Both the α and β globin genes have
tion 2 of the globin chain [34]. When the chain structural determinants in their 3′ UTRs that are
reaches 40–50 residues, co‐translational acetyla- important for mRNA stability [26].
tion of the N‐terminal residue can occur through
the action of several acetyltransferases [35].
Normal haemoglobins
Whether this occurs to any great extent depends
on the nature of the N‐terminal residue. Thus the The normal haemoglobins beyond the neonatal
glycine of the γ chain is 10–15% acetylated, whereas period are haemoglobin A and two minor haemo-
the valine of normal α, β and δ chains is resistant to globins, haemoglobin A2 and haemoglobin F.
acetylation. Rarely, an aberrant amino acid residue
present as a result of mutation leads to increased
Haemoglobin A2
acetylation, as is the case in haemoglobin Raleigh
[34]. There are 64 possible nucleotide triplets or In adults, haemoglobin A2 comprises about 2–3.5%
codons, 61 of which encode amino acids (20 in all) of total haemoglobin. The percentage is much lower
and 3 of which do not. The latter serve as STOP or at birth, about 0.2–0.3% or even lower, with a rise to
6
Sβ0 thalassaemia
4 Sβ+ thalassaemia
HbA2 %
3 Normal
Fig. 1.11 Diagram showing the rate of 1

rise of haemoglobin A2 in
haematologically normal Jamaican 0
0 3 6 9 12 15 18 21 24 27 30 33 36
babies and in babies with sickle cell/β
thalassaemia. (Modified from Age in months
reference [37].)
adult levels during the first 2 years of life. The the percentage of haemoglobin A2 [40]; however it
steepest rise occurs in the first year but there is a does not appear to be the HBS1L gene itself that is
continuing slow rise up to 3 years of age [37]
responsible [41]. Inactivating mutations in KLF1
(Fig. 1.11). In the normal adult population, the per- can cause a borderline‐to‐moderate increase in
centage of haemoglobin A2 shows a Gaussian distri- haemoglobin A2, up to 3.5–4.7%, with red cell indi-
bution. It has functional properties that are very ces typical of β thalassaemia heterozygosity [30,
similar to those of haemoglobin A [21] (similar 31, 42, 43]. The proportion of haemoglobin A2 is
cooperativity and interaction with 2,3‐DPG) reduced by absolute or functional iron deficiency
although, in comparison with haemoglobin A, it (see Table 6.3) and by α, δ and δβ thalassaemia. In
inhibits polymerisation of haemoglobin S [38] and γδβ thalassaemia, the rate of synthesis, but not the
has a higher oxygen affinity [12]. It has a pancellu- proportion of haemoglobin A2, is reduced since
lar distribution. synthesis of γ and β chains is reduced, as well as δ
The reduced rate of synthesis of haemoglobin A2 chain synthesis. The proportion of haemoglobin A2
in comparison with haemoglobin A reflects the is increased in the great majority of patients with β
much slower rate of synthesis of the δ chain in thalassaemia trait, in the majority of patient with
comparison with the β chain. This in turn appears haemoglobin E heterozygosity [44] and in some
to be consequent in part on a reduced rate of tran- patients with an unstable haemoglobin. The per-
scription of δ mRNA caused by a difference in the centage of haemoglobin A2 is not affected by ciga-
promoter region of these two genes; the δ gene has rette smoking [45].
a CCAAC box rather than the CCAAT box of the β Many δ chain variants and δ thalassaemias
gene [21] and in addition lacks the CACCC occur, over 80 and over 30, respectively, having
sequence that is present in the β promoter (see been described; some of the δ chain variants are
Table 1.2). In addition, γ chain mRNA is unstable thalassaemic variants. Around 1–2% of individu-
and there is also an influence of sequences in IVS2 als of African ancestry have the variant haemo-
[39]. The haemoglobin A2 percentage in adults is globin designated haemoglobin A2′ (A2 prime) or
controlled by two different genetic mechanisms haemoglobin B2 (δ16Gly→Arg). It has been reported
[40]. Analysis of single nucleotide polymorphisms to be the commonest haemoglobin variant after
(SNPs) shows that alleles in the region of the haemoglobins S and C [39]. Haemoglobin A2′ is
HBS1L and MYB loci at 6q23.3 influence both the readily detected on high performance liquid
percentage of F cells and the A2 percentage, while chromatography (HPLC) (Fig 1.12) and i soelectric
alleles around the HBB locus at 11p15.4 influence focusing. The δ thalassaemias are also common
14 Chapter 1
Fig. 1.12 High performance liquid

chromatography (HPLC)
chromatogram (Bio‐Rad Variant II)
showing a split haemoglobin A2
resulting from heterozygosity for
haemoglobin A2′. The white arrow
shows haemoglobin A2 and the black
arrow haemoglobin A2′.
100
90
80
Haemoglobin F (%)
70
60
50
40
30
20
10
Fig. 1.13 The rate of fall of
0
0 1 2 3 4 5 6 7 8 9 10 11 12 haemoglobin F percentage post‐natally
Age in months in normal and premature babies. The
pale blue areas represent premature
100% range, reference [23] babies and the deep blue normal
95% range, reference [24]
babies. (Derived from references [46]
and [47].)
in some ethnic groups (e.g. present in 1% of However, it should be noted that fetal develop-
Sardinians [12]). Although some δ chain variants ment appears to be normal in the offspring of
are unstable or have increased oxygen affinity, mothers with very high levels of haemoglobin F.
the low percentage of haemoglobin A2 means Its oxygen dissociation curve is sigmoid. The
that δ thalassaemia and δ chain variants are of no increased oxygen affinity, in comparison with
clinical significance. However, their presence haemoglobin A, is attributable to its weak affin-
complicates the diagnosis of β thalassaemia ity for 2,3‐DPG [8]. In comparison with haemo-
trait (see p. 124). globin A, haemoglobin F is less efficient at
transporting CO2. A significant proportion of
haemoglobin F is acetylated.
Haemoglobin F
During the first year of life the haemoglobin F
Haemoglobin F is the major haemoglobin during percentage falls progressively to values close to
intrauterine life. Its oxygen affinity is higher adult levels (Fig. 1.13) [46–48]. A slower fall to final
than that of haemoglobin A and this facilitates adult levels may continue for several years, even
oxygen transfer from the mother to the fetus. up to puberty and beyond. The percentage of fetal
haemoglobin present at birth is quite variable, thalassaemia major and sickle cell anaemia [54]
usually being between 60 and 95%. During intrau- and haemoglobin E homozygosity [57]; MYB is a
terine life and at birth, haemoglobin F shows a Gγ transcription factor for a number of erythroid‐
to Aγ ratio of approximately 2:1 to 3:1. Within the specific genes and the intergenic sequences atten-
first few months of birth this changes to the adult uate the function of nearby enhancers, reducing
ratio of approximately 2:3. In premature infants expression of MYB and upregulating haemoglo-
there is initially a plateau phase in haemoglobin F bin F production [59];
concentration lasting 20–60 days, followed by a • a trans‐acting locus at Xp22.2 (FCP1) (HBFQTL3),
linear decrease similar to that in term babies [47]. which affects F‐cell numbers in normal individuals
At any given period after birth the spread of val- and those with sickle cell anaemia [54];
ues is greater than in term babies. Initially there • a trans‐acting locus on chromosome 8q, desig-
are more high values but after the first month of nated haemoglobin F quantitative trait locus 4
life values both higher and lower than those of (HBFQTL4), which may be the TOX gene at 8q12.1
term infants are observed [47]. Haemoglobin F lev- [60], which interacts with the common HBG2 poly-
els at birth are higher in the babies of diabetic morphism (see earlier);
mothers and low‐for‐birth weight babies whereas • a polymorphism of the BCL11A gene at 2p16.1,
in Down’s syndrome the switch to haemoglobin A designated haemoglobin F quantitative trait locus 5
is earlier [39]. (HBFQTL5), which affects the haemoglobin F per-
In normal adults, haemoglobin F is heterogene- centage in normal Europeans, in individuals with β
ously distributed, being found in a subset of eryth- thalassaemia major and sickle cell anaemia [54], in β
rocytes designated F cells. The proportion of F cells thalassaemia heterozygosity [55] and in haemoglo-
is highly variable, in one study ranging from 0.6 to bin E heterozygosity [56] and homozygosity [57]
22% [49]. The haemoglobin F percentage is deter- (BCL11A inhibits γ gene expression);
mined by age, sex (slightly higher in women) and a • SAR1A at 10q22.1, which influences haemoglo-
number of inherited characteristics both linked and bin F response to hydroxycarbamide [54];
unlinked to the β globin gene cluster. DNA sequences • KLF1 (previously known as EKLF) at 19p13.13,
controlling the proportion of F cells and the percent- (HBFQTL6), reduced expression of which leads to
age of haemoglobin F include [27, 46–53]: reduced expression of the γ gene repressor BCL11A
• a polymorphism at position −158 of the Gγ gene and, therefore, increased haemoglobin F; various
(HBG2) (C→T being associated with a higher hae- inactivating heterozygous mutations have led to a
moglobin F) (also known as XmnI Gγ polymor- haemoglobin F of 1–3%, 1–7.4% and 3.3–19.5% and
phism), which is partly responsible for the higher a dominant mutation can lead to haemoglobin F of
haemoglobin F percentage associated with the 40–50% with congenital dyserythropoietic anae-
Senegal and Arab‐Indian haplotype in sickle cell mia [29, 61]; compound heterozygosity for S270X
anaemia [54], β thalassaemia heterozygosity [55], nonsense and K332Q mis‐sense mutations led to
haemoglobin E heterozygosity [56], haemoglobin E haemoglobin F levels of 22–31%, whereas simple
homozygosity [57] and inherited bone marrow fail- heterozygosity was not associated with an
ure syndromes [58]; elevation [62];
• variation of the number of repeats of a specific • expression of KBX3 (previously CSDA) at 12p13.2
motif at −530 in the HS2 component of the β locus (cold shock protein domain A is a trans‐acting sup-
control region – (AT)xN12GT(AT)y; pressor of HBG2) [63];
• a trans‐acting locus at 6q22.3–23.1 in the inter- • ZBTB7A at 19p13.3 inhibits γ gene expression [64];
genic region between HBS1L and MYB, desig- • PPARGC1A at 4p15.2 induces γ gene expres-
nated haemoglobin F quantitative trait locus 2 sion [64];
(HBFQTL2) or HBSIL‐MYB intergenic polymor- • the HBBP1 pseudogene is a negative regulator in
phism block 2 (HMIP‐2), which affects F‐cell num- adult red cells [65];
bers and v ariation of haemoglobin F percentage in • a BGLT3 long non‐encoding RNA is a positive
normal Europeans and in individuals with β regulator in fetal red cells [66];
16 Chapter 1
• ANTXR1 at 2p13.3 is associated with reduced or a more extensive mutation, in which there is
haemoglobin F expression with several mutations deletion, insertion or other alteration of more than
being associated with a lower haemoglobin F in one nucleotide. The types of mutation that can
sickle cell anaemia associated with the Arab‐Indian occur in globin genes are summarised in Table 1.3
haplotype [67]. [69–72]. In addition, expression of globin genes
Of these various loci, the three most important can be affected by DNA sequences outside the glo-
may be BCL11A, KLF1 and HBSIL‐MYB. There is bin genes themselves, either enhancers acting in
also a γ‐globin gene silencer in a 3.5 kb region near cis or genes on other chromosomes encoding
the 5′ end of the δ‐globin gene which, when deleted, trans‐acting transcription factors (Tables 1.3 and
can give rise to hereditary persistence of fetal hae- 1.4 [30, 42, 43, 73–76]). The phenotype in such
moglobin [28]. cases can be that of thalassaemia trait or silent tha-
The percentage of haemoglobin F is also affected lassaemia [77].
by any increase in the number of γ genes. Point mutations in globin genes sometimes have
The mechanism by which the polymorphisms in no effect on the amino acid sequence. This occurs
LCRB at −530 bp to the Gγ gene influence γ chain because, as mentioned earlier, there is redundancy
synthesis appears to be that, in comparison with in the genetic code, with a number of nucleotide tri-
(AT)7T7, the (AT)9T5 sequence shows increased bind- plets encoding the same amino acid. When a ‘same‐
ing of BP‐1, a negative trans‐acting factor [68]. sense’ mutation occurs, the new codon resulting
The distribution of haemoglobin F percentages from the mutation codes for the same amino acid as
in the population is skewed. In 85–90% of indi- the original codon and there is thus no effect on the
viduals, haemoglobin F is less than 0.6–0.7% and F final gene product. Similarly, mutation of a termina-
cells are <4.5% [50, 52]. The other 10–15% of the tion codon may be to a different termination codon.
population have values above these levels. The Many spontaneous mutations in globin genes are
upper limit of normal is rather arbitrarily taken as same‐sense mutations. Point mutations can also
1%. It would probably be more accurate to take 0.6 result in a ‘mis‐sense’ mutation, when the new
or 0.7% as the upper limit of normal, excluding codon encodes a different amino acid, leading to
the 11% of males and 21% of females who have a production of a variant haemoglobin. The site of a
slight elevation of the percentage of F cells and the mutation is critical, determining whether there is an
haemoglobin F percentage as an X‐linked domi- effect on stability, oxygen affinity, solubility and
nant characteristic [50]. However, since the meas- other critical characteristics of the haemoglobin
urement of a low percentage of haemoglobin F is molecule. Because of the redundancy in the genetic
very imprecise, 1% is a practical upper limit. code, different point mutations can give rise to the
Haemoglobin F is more markedly increased in same variant haemoglobin. For example the α chain
patients with various inherited abnormalities of β variant, G‐Philadelphia, has arisen twice, from an
globin chain synthesis (see Table 3.13) and less often AAC to AAG change in an α2α1 fusion gene and
in various acquired conditions (see Table 6.2). from an AAC to AAA change in an α2 gene [78].
Mutations in γ genes can lead not only to an There are more than 1300 known variant haemo-
increased percentage of haemoglobin F but also to globins resulting from point mutations. There are
haemoglobin F variants, more than 40 of which also more than two dozen variant haemoglobins
have been described. resulting from two point mutations in the same
gene with two amino acid substitutions. This can
result either from a new mutation occurring in the
Variant haemoglobins and
gene encoding a variant globin chain (e.g. in a
abnormalities of globin chain
parental germ cell) or from crossover between two
synthesis
variant alleles. Usually the second mutation has
Nuclear DNA, including the DNA of globin genes, occurred in a gene that would otherwise encode a
is subject to spontaneous mutation. This may be a relatively common variant haemoglobin such as
point mutation (alteration of a single nucleotide) haemoglobin S, C, E or D‐Punjab.
Table 1.3 Types of mutation that can occur in globin genes and adjoining sequences.
Type of mutation Possible consequence Examples
Point mutations
Within coding sequence, i.e. within Same‐sense or neutral mutation, Many mutations are of this type; more
an exon i.e. mutant codon codes for same than a third of theoretically possible
amino acid as normal codon so point mutations would result in no
there are no consequences alteration in the amino acid encoded
Mis‐sense mutation, i.e. mutant Haemoglobin S, haemoglobin C,
codon codes for a different amino haemoglobin E; haemoglobin Marseille
acid from the normal codon; and haemoglobin South Florida (altered
includes mis‐sense mutations in amino acid near N‐terminus plus
which an abnormal amino acid persisting methionine residue at the
interferes with the normal cleavage N‐terminus of the β chain)
of the N‐terminal methionine
Nonsense mutation, i.e. the mutant Haemoglobin McKees Rocks (two amino
codon does not code for an amino acids shorter than normal); α2 CD116
acid and thus functions as a STOP GAG→TAG creating premature STOP
or TERMINATION codon, codon and causing α thalassaemia
producing a shortened globin chain
New‐sense mutation, i.e. conversion Haemoglobin Constant Spring,
of a STOP codon to a coding haemoglobin Icaria, haemoglobin Seal
sequence producing an elongated Rock, haemoglobin Koya Dora,
globin chain haemoglobin Paksé, haemoglobin Zunyi
Mutation of the initiation codon α of β thalassaemic disorder
Gene conversion* Conversion of Gγ gene to Aγ gene, giving

γ γ genotype
A A
Conversion of Aγ gene to Gγ gene, giving

γ γ genotype
G G
Conversion of ψζ1 to a gene that

resembles ζ2 but is still non‐functional
Gene conversion between the α2 and α1
genes so that the same mutation is present
in both, e.g. α2Lys→Glu α1Lys→Glu, giving
unusually high levels of haemoglobin I
Gene conversion between a β gene and a
δ gene leading to a haemoglobin A2
variant (e.g. haemoglobin A2 Flatbush)
Gene conversion plus further point Haemoglobin F Port Royal, resulting
mutation* from a further point mutation in a Gγ‐Gγ
gene complex
Within non‐coding sequence, Production of a new splice site Some β thalassaemias

i.e. in an intron leading to a structurally
abnormal mRNA
Mutation 5′ or 3′ to the gene Mutation of an enhancer Some β thalassaemias

(i.e. outside the gene)
Reduced rate of synthesis of mRNA Some β thalassaemias
due to interference with 3′‐end
formation of mRNA
(Continued on p. 18.)
Table 1.3 Continued.
Deletion or duplication of one or more genes or pseudogenes

Deletion of one or more genes Total loss of expression of relevant Most α thalassaemias, some β
gene; occasionally also loss of thalassaemias, δβ thalassaemias and γδβ
function of an adjacent structurally thalassaemias; deletion of Gγ gene (–Aγ),
normal gene homozygosity for which causes anaemia
and a reduced haemoglobin F percentage
in the neonate; deletion of ψζ1
Deletion of genes with Loss of β and δ gene function but Deletional hereditary persistence of fetal
downstream enhancer being enhanced function of remaining Gγ haemoglobin
juxtaposed to remaining gene (± Aγ) gene
Duplication of α gene Triple, quadruple or quintuple ααα†/αα, ααα/ααα, αααα‡/αα, αααα/αααα

α gene or αα/ααααα
Triplication of entire α globin gene Six α genes on a single chromosome αα:αα:αα/αα

cluster
Duplication of Gγ gene Double, triple or quadruple Gγ gene γ γ γ, GγGγGγAγ (homozygotes have

G G A
so that there are three, four or five γ been described with a total of 8 γ genes)
genes on a chromosome or GγGγGγGγAγ
Duplication of the ζ or ψζ gene Double, triple or quadruple ζ2ψζ1ψζ1/ζ2ψζ1 or ζ2ψζ1ψζ1/ζ2ψζ1ψζ1

ζ/ψζ gene or 4 ζ‐like genes per chromosome
Abnormal crossover during meiosis leading to gene fusion

α2α1 fusion Effective loss of one α gene but −α3.7 thalassaemia
structurally normal α chain is
encoded
δβ fusion – simple crossover Reduced rate of synthesis of Haemoglobin Lepore, e.g. haemoglobin
structurally abnormal globin chain Lepore‐Washington/Boston,
haemoglobin Lepore‐Baltimore and
haemoglobin Lepore Hollandia, or δ0β+
thalassaemia [69]
δβδ fusion – double crossover Reduced rate of synthesis of Haemoglobin Parchman

with δ sequences on either side of structurally abnormal globin chain
β sequences
βδ fusion (with preservation of Anti‐Lepore haemoglobins, e.g.

intact δ and β genes on either side haemoglobin Miyada, haemoglobin
of fusion gene, with or without P‐Nilotic, haemoglobin P‐Congo
additional mutation) haemoglobin Lincoln Park
A
γβ fusion Synthesis of variant haemoglobin Haemoglobin Kenya
plus increased synthesis of
haemoglobin F
βAγ fusion (with preservation of Haemoglobin anti‐Kenya

intact Gγ and Aγ genes and
duplication of the δ gene)
G
γ‐β fusion Variant was 37%, A2 2.4%, F 6.6%, Haemoglobin Gγ‐β Ulan [70]
phenotype of β thalassaemia minor
γ γ fusion (designated ‐GγAγ‐)

G A
Reduced rate of synthesis of γ thalassaemia
haemoglobin F
Deletion of DNA sequences but without a frameshift in coding sequence

Deletion of part of a coding One to five amino acids missing but Haemoglobin Gun Hill (an unstable
sequence, either three nucleotides sequence otherwise normal haemoglobin with five amino acids
or a multiple of three missing)
Deletion plus inversion

Two deletions with inversion of Deletion involving Aγ and δ plus β Indian type of deletional Aγδβ0
intervening sequence genes respectively but with thalassaemia
preservation of an intervening
region which is inverted
Deletion plus insertion

Deletion with insertion of Same functional effect as deletion One type of α0 thalassaemia, −−MED
extraneous DNA between
breakpoints
Insertion within a coding sequence but without a frameshift

Insertion of nucleotides, either Up to five extra amino acids Haemoglobin Koriyama (an unstable
three or multiples of three, e.g. by haemoglobin with insertion of five
tandem duplication codons in β gene, anti‐Gun Hill);
haemoglobin Grady (insertion of three
codons in α gene)
Insertion of a duplicated sequence including the start codon

Tandem duplication of codons Reduced β chain synthesis β+ thalassaemia; thalassaemia intermedia
flanking the start codon of the in a homozygote [71]
HBB gene
Frameshift mutations
Alteration of the reading frame Abnormal amino acid sequence Haemoglobin Wayne (α chain),
resulting from deletion, insertion, with an elongated globin chain haemoglobin Tak (β chain), haemoglobin
deletion plus insertion or deletion (when a STOP codon is out of phase Cranston (β chain), some β thalassaemias
plus duplication and translation continues until including some dominant β
another ‘in‐frame’ STOP codon is thalassaemias, some α thalassaemias
met); abnormal amino acid
sequence with a truncated globin
chain (when a premature STOP
codon is created)
Chromosomal translocation
Unbalanced translocation (there Extra α genes on a chromosome Same significance as homozygous
may be a balanced translocation in other than chromosome 16 triplication of an α gene since there are a
a parent) total of six α genes
Loss of an α gene α thalassaemia; may be part of a

contiguous gene syndrome
Deletion of a locus control region

Locus control region deleted, with Deletion of the locus control region (ε)γδβ0 thalassaemia
or without deletion of relevant of the β gene (LCRB)
genes
Deletion of the α gene enhancer (HS α0 thalassaemia
−40) 40 kb upstream of the ζ2 gene,
also known as MCS‐R2 (LCRA)
Creation of a sequence that competes for the locus control region

Gain‐of‐function mutation Downregulation by competition α thalassaemia [72]
within α gene cluster with GATA1 for locus control region
* Gene conversion is non‐reciprocal genetic exchange between allelic or non‐allelic homologous sequences so that one gene comes to
resemble another more closely or becomes identical to it. This is responsible for maintaining the similarity between pairs of identical or
similar genes.
†
Either αααanti3.7 or αααanti4.2.
‡
Either ααααanti3.7 or ααααanti4.2.
20 Chapter 1
Table 1.4 Mutations and polymorphisms occurring outside the globin gene clusters leading to abnormal globin chain
synthesis [30, 42, 43, 73–76].
Mutation Consequence
Mutation in ATRX gene at Xq21.1 which encodes a Haemoglobin H disease plus dysmorphism and severe
trans‐acting factor regulating α gene expression learning difficulties
Mutation in FCP1 at Xp22.2 Hereditary persistence of fetal haemoglobin
An HBSIL‐MYB intergenic polymorphism at 6q22.3–23.1 Hereditary persistence of fetal haemoglobin
Mutation in the ERCC2 (XPD) gene at 19q13.22, which Recessive trichothiodystrophy and β thalassaemia
encodes one component of the general transcription
factor TFIIH [73]
Mutation in the GATA1 gene at Xp11.23 [74] X‐linked thrombocytopenia and β thalassaemia (and, in
one patient, congenital erythropoietic porphyria) [75]
BCL11A haploinsufficiency (2p16.1) Increased haemoglobin F associated with cognitive

impairment and facial dysmorphism [30]
KLF1 at 19p13.13 Increased haemoglobin A2 [42, 43]
ASH1L at 1q22 β thalassaemia [76]
Some point mutations are ‘nonsense’ mutations alanine are often acetylated and this is the case with
in which the new codon is one of the three that do this variant haemoglobin [79]. The presence in one
not code for an amino acid. A nonsense mutation individual of haemoglobins with three different β
thus functions as a ‘STOP’ or ‘TERMINATION’ chains may be attributable to post‐translational
codon, leading to termination of chain synthesis. modification. For example, the replacement of leu-
If this type of mutation is near the 3′ end of the cine by hydroxyleucine that characterises haemo-
gene an abnormal but functional globin chain is globin Coventry is not encoded by genomic DNA
produced, but if it is more proximal, the chain and is found only in the presence of an unstable
produced is likely to be not only short but also haemoglobin, either haemoglobin Atlanta or hae-
very unstable, leading to a dominant thalassaemia moglobin Sydney. Some mutations affecting the
phenotype. Point mutations can also convert a haem pocket and leading to haemoglobin instabil-
STOP codon to a coding sequence so that an ity permit oxidation of leucine to isoleucine [80].
elongated mRNA and elongated globin chain are Haemoglobin Bristol also shows post‐translational
produced. modification. It is an unstable haemoglobin result-
An unusual result of a point mutation is producing from conversion of the β67 valine codon to a
tion of an abnormal amino acid that is converted to codon for methionine; however the final haemoglo-
a different amino acid by post‐translational modifi- bin has aspartic acid rather than methionine as a
cation. This can be as the result of deamidation, result of post‐translational modification [17].
acetylation or oxidation. There are at least six In a slightly different mechanism, the abnormal
reported variant haemoglobins in which the abnor- structure of a variant haemoglobin resulting from a
mal DNA sequence codes for asparagine but this is point mutation leads to post‐translational modifica-
subsequently deamidated to aspartic acid [17]; of tion of a normal amino acid, in three cases leucine
these, the most common is haemoglobin J Sardegna being modified to hydroxyleucine [17] and in one
(α50(CD8)His→Asn→Asp), which has a prevalence of case asparagine adjacent to the abnormal residue
0.25% in northern Sardinia. Post‐translational acet- being deamidated to aspartic acid [79].
ylation occurs in haemoglobin Raleigh, which has a Mutations in the codon for the N‐terminal valine
β1Val→Ala substitution; proteins with an N‐terminal may mean that a different amino acid is encoded,
with resultant retention of the initiator methionine chain. The original STOP codon is no longer in the
and full acetylation of the N‐terminal residue (e.g. reading frame and transcription continues until
the glutamate of the α chain variant haemoglobin another STOP codon is encountered.
Thionville) or normal cleavage of methionine but Small deletions and large deletions and inser-
full acetylation of the N‐terminal residue (e.g. the tions can result from non‐homologous crossover
alanine of the α chain variant haemoglobin Lyon‐ between a pair of chromosomes during meiosis.
Bron) [35]. Similarly, a histidine to proline change in These are usually in‐frame. Non‐homologous cross-
position β2 leads to retention of the initiator methio- over can involve not only a single pair of allelic
nine [79]. If methionine is retained, the globin chain genes (e.g. two α genes) but also two structurally
is extended by one residue. similar but non‐allelic genes (e.g. a β gene and a δ
Deletions and insertions can lead to a frameshift; gene); in the latter instance there may be a loss of
that is, unless the deletion or insertion involves three the two normal genes and production of a fusion
nucleotides or multiples of three the nucleotide gene that has 5′ sequences of one gene and 3′
sequences beyond the mutation will be in a different sequences of the other gene; alternatively the two
reading frame and will be ‘read’ during translation normal genes may be retained with part of both
as coding for a completely different sequence of genes being reduplicated in the fusion gene. Some
amino acids. Frameshift mutations can lead to a pre- examples of non‐homologous crossover are shown
mature STOP codon so that both the mRNA and in Fig. 1.14. Non‐homologous crossover can also
the resultant globin chain are shorter than normal. result in the reduplication of genes; for example,
If this does not occur, a frameshift mutation is likely some individuals, instead of having two α genes on
to lead to elongated mRNA and an elongated globin each chromosome 16, have three, four or even five [81]
Anti-Lepore βδ fusion gene

ψβ δ βδ β
ψβ δ β
β
ψβ δ
ψβ δβ
(a) Lepore δβ fusion gene
Anti-Kenya
βAγ fusion gene
Gγ Aγ ψβ δ βAγ ψβ δ β
Gγ Aγ ψβ δ β
ψβ δ β
Fig. 1.14 Some examples of fusion
Gγ Aγ
genes produced by non‐homologous
crossover: (a) formation of genes
encoding Lepore and anti‐Lepore Gγ Aγβ
haemoglobins; (b) formation of genes Kenya
encoding Kenya and anti‐Kenya (b) Aγβ fusion gene
haemoglobins.
22 Chapter 1
α genes on one chromosome. Duplicated α genes with a structurally abnormal haemoglobin,

occur in many populations and in some are quite whereas others use the term to include all
frequent. For example, 2% of Sri Lankans have ααα. disorders of globin chain synthesis, encompassing
Very rarely, individuals are somatic mosaics so also the thalassaemias. The second use seems
that a variant haemoglobin is present in an unusu- preferable since some variant haemoglobins, such
ally low percentage. For example, haemoglobin as haemoglobin E and haemoglobin Constant
Costa Rica, a β chain variant arising as a result of Spring, are synthesised at a reduced rate and thus
somatic mutation, constituted 5–6% of total haemo- constitute thalassaemic haemoglobinopathies,
globin in the individual in whom it was recognised and some thalassaemias lead to synthesis of a
[82]. Similarly, a patient has been reported with hae- structurally abnormal haemoglobin, such as hae-
moglobin Korle Bu as a minor fraction as a result of moglobin H or haemoglobin Bart’s, as a result of
constitutional mosaicism [83]. unbalanced chain synthesis.
Haemoglobin dimers are stable but the tetramers Haemoglobinopathies can result from muta-
that they form are able to dissociate and re‐associate. tion of a β globin gene, in which case there is a
When both a normal and a variant haemoglobin are variant form of haemoglobin A, or mutation of an
present, heterotetramers and homotetramers will α globin gene, in which case there are variant
be present in vivo (e.g. in the case of sickle cell trait forms of haemoglobins F, A and A2. Similarly,
there will be α2β2, α2βS2 and α2ββS). When haemoglob- mutations of γ or δ genes result in mutant forms
ins are studied in vitro (e.g. by electrophoresis or of haemoglobin F and haemoglobin A2, respec-
chromatography), the heterotetramers dissociate tively. Because there are two β genes, an individ-
and re‐associate as homotetramers. Some variant ual can have both a β chain variant and
haemoglobins have abnormally stable tetramers so haemoglobin A or two β chain variants. Because
that three rather than two forms are detected on hae- there are usually four α genes, an individual
moglobin electrophoresis and similar techniques. could, in theory, have up to four different α chain
variants; in practice, a number of individuals
have been described with both haemoglobin A
Thalassaemias and variant and two different α chain variants, for example
haemoglobins haemoglobin Buda, haemoglobin Pest and hae-
Mutations can lead not only to synthesis of a moglobin A in one instance and haemoglobin
structurally abnormal haemoglobin but also to a GPhiladelphia, haemoglobin JSardegna and haemoglobin
reduced rate of synthesis of a globin chain and A in several instances.
therefore of the haemoglobin species of which it
forms a part. The term ‘thalassaemia’ is used to
The proportion of variant haemoglobins
describe disorders with a significant decrease in
the rate of synthesis of one or more globin chains. The proportion of an α chain variant in the blood
α thalassaemia indicates a reduced rate of synthe- might be expected to be around 25%, since there are
sis of α globin chain. Similarly, β, δ, δβ and γδβ tha- usually four α genes. However the situation is far
lassaemias indicate a reduced rate of synthesis of more complex. The variant is likely to be more than
β, δ, δ+β and γ+δ+β chains, respectively. In some 25% if it results from a mutation of the α2 gene
disorders there is both synthesis of a structurally (since the ratio of α2 to α1 synthesis is normally
abnormal haemoglobin and a reduced rate of about 3:1) and less than 25% if it results from muta-
synthesis of the variant haemoglobin. This is the tion of an α1 gene. The percentage is raised if there
case, for different reasons, with the α chain variant is coinheritance of α thalassaemia and is lowered if
haemoglobin Constant Spring (first described in a there is coinheritance of triple α (ααα). If a gene
Chinese patient in Constant Spring, a district of encoding an α chain variant is a mutated α1 gene in
Kingston, Jamaica), and the β chain variant hae- cis with deletion of the α2 gene then it can be upreg-
moglobin E. The term ‘haemoglobinopathy’ is ulated, increasing the percentage further. The per-
sometimes used to indicate only those disorders centage is reduced if the variant α chain is
Table 1.5 Consequences of mutation of globin genes.
Type of mutation and consequence Examples
Substitution of an external amino acid which is not involved in Haemoglobin G‐Philadelphia

inter‐chain contacts; no functional abnormality
Amino acid substitution leading to reduced solubility, polymerisation Haemoglobin S (sickle cell haemoglobin)
of haemoglobin and deformation of cells into a holly leaf or sickle shape
with consequent haemolysis and vascular obstruction
Amino acid substitution leading to reduced solubility, formation of Haemoglobin C

straight‐edged crystals and haemolysis
Replacement of haem‐binding or haem‐related histidine residue by M haemoglobins

another amino acid leading to an increased tendency to oxidation, i.e.
formation of methaemoglobin. There is cyanosis at birth if the defect is
in a γ gene, cyanosis from birth if the defect is in an α chain and
cyanosis from approximately 6 months of age if the defect is in a β
chain. There may be associated haemoglobin instability
Mutation involving amino acids of the haem pocket or α1β2 (tetrameric) Haemoglobin Köln, haemoglobin Zurich
contacts or mutation interfering with the helical structure of (haem pocket mutation), haemoglobin
haemoglobin, leading to haemoglobin instability and Heinz body Kansas (mutation affecting α1β2 contacts)
haemolytic anaemia. There may also be decreased oxygen affinity and
resultant cyanosis
Mutations involving α1β2, α2β1 tetrameric haemoglobin contacts or Haemoglobin Chesapeake, haemoglobin
C‐terminal end of β chain, where there are residues involved in 2,3‐DPG Bethesda, haemoglobin Kempsey,
interaction and stability of the deoxy form of haemoglobin, leading to haemoglobin J‐Capetown, haemoglobin
increased oxygen affinity and polycythaemia Yakima
Mutation leading to decreased oxygen affinity and therefore anaemia, Haemoglobin S, haemoglobin Seattle
since normal tissue delivery of oxygen is achieved with a lower (also unstable), haemoglobin Kansas
concentration of haemoglobin. May cause cyanosis (also unstable), haemoglobin Beth Israel
Mutation in β gene leading to markedly reduced or absent β chain β thalassaemia (major, intermedia or minor)
production, reduced synthesis of haemoglobin A and possibly
ineffective erythropoiesis consequent on damage to developing
erythroblasts by excess α chains
Mutation in β gene leading to structurally abnormal and very (Dominant) β thalassaemia phenotype
unstable β chain
Mutation in one or more α genes leading to markedly reduced or absent α thalassaemia (α thalassaemia trait,
α chain synthesis and reduced synthesis of haemoglobins F, A and A2 haemoglobin H disease or haemoglobin
Bart’s hydrops fetalis)
Mutation in α gene leading to structurally abnormal α chain synthesised α thalassaemia phenotype, e.g.
at a greatly reduced rate haemoglobin Constant Spring (mRNA and
the haemoglobin are unstable)
Mutation in δ gene leading to a structural abnormality or markedly Haemoglobin A2 variant or δ thalassaemia.

reduced or absent δ chain production No clinical significance as haemoglobin A2
is a minor haemoglobin but complicates the
diagnosis of β thalassaemia trait
Mutation in γ gene leading to structural abnormality or reduced rate of Some methaemoglobins

synthesis of γ chain and therefore haemoglobin F
2,3‐DPG, 2,3‐diphosphoglycerate; mRNA, messenger ribonucleic acid.

24 Chapter 1
synthesised at a reduced rate, if it has a lower affin- Check your knowledge

ity for β chains than does the normal α chain or if
the variant α chain or the variant haemoglobin is One to five answers may be correct. Answers to
unstable. almost all questions can be found in this chapter or
Similarly, it might be expected that a β chain vari- can be deduced from information given. Correct
ant would be about 50% of total haemoglobin in answers are given on p. 29.
heterozygotes since there are two β genes. As for α
chain variants, the situation is much more complex. 1.1 The haemoglobin molecule
The percentage may be above 50% in the case of (a) requires iron for its synthesis
variants with negatively charged β chains, which (b) is composed of three pairs of globin
have a greater affinity than the normal β chain for chains
the positively charged normal α chains (e.g. haemo- (c) alters its structure when oxygen is bound
globin J Baltimore or J‐Iran); if there is coexisting α (d) is assembled in the cytosol
thalassaemia, leading to a lack of α chains, the per- (e) binds 2,3‐diphosphoglycerate
centage of the variant is even higher. The converse
is seen with positively charged β chains, such as βS, 1.2 Haemoglobin F
βC, βO‐Arab and βD‐Punjab, which have a lower affinity (a) is the major haemoglobin present in the
than normal β chains for normal α chains. The per- fetus
centage of the variant is thus somewhat less than (b) has a lower oxygen affinity than haemo-
50% and if there is coexisting α thalassaemia the globin A
percentage is even lower. The percentage is also (c) is absent in normal adults
reduced considerably below 50% if there is a (d) percentage shows a non‐Gaussian
reduced rate of synthesis of the variant β (or δβ) distribution in the population
chain (e.g. βE, δβLepore), if the β chain is unstable or if (e) is composed of two α chains and two β
the variant haemoglobin is unstable (e.g. haemoglo- chains
bin Köln).
An alteration in the amino acid sequence of the 1.3 The functions of haemoglobin include
globin chains (i.e. an alteration in the primary struc- (a) transport of glucose
ture of haemoglobin) often has no significant effect (b) transport of CO2
on the secondary, tertiary and quaternary structure (c) transport of O2
of haemoglobin; this is the case when the substi- (d) buffering
tuted amino acid is of similar size to the normal (e) transport of creatinine to the kidney
amino acid, has the same charge and the same
hydrophobic or hydrophilic properties and does 1.4 The affinity of haemoglobin for oxygen is
not have a role in the binding of haem or 2,3‐DPG decreased by
nor in interactions between chains. In this case a (a) fever
variant haemoglobin has no consequences for the (b) metabolic alkalosis
health of the individual. In other cases an alteration (c) binding of CO2
in the primary structure of haemoglobin affects the (d) binding of 2,3‐diphosphoglycerate
secondary, tertiary or quaternary structure of the (e) glycosylation
molecule, sometimes with very profound effects.
Some of the effects of mutations in globin genes are 1.5 When blood circulates through the lungs
shown in Table 1.5. haemoglobin
Over 1000 mutations of the globin genes have (a) is oxidised
been recognised. Some 690 of them were collated in (b) takes up oxygen
a single volume [84] and this database is now avail- (c) loses CO2
able electronically, in updated form (http://globin. (d) takes up water
cse.psu.edu/). (e) dissociates into haem and globin
1.6 Structurally abnormal haemoglobins may 1.11 The proportion of a variant haemoglobin is
result from usually
(a) point mutations (a) greater in the case of an α chain variant
(b) gene fusion than a β chain variant
(c) frameshift mutations (b) greater in the case of an α chain variant
(d) mutation of STOP codon to a coding if there is coexisting deletion of an
sequence α gene
(e) mutation of a coding sequence to a STOP (c) greater if the variant β chain has a higher
codon affinity for normal α chain than does the
normal β chain
1.7 Abnormal haemoglobins may (d) greater, in the case of haemoglobin S, if
(a) have increased oxygen affinity there is coexisting α thalassaemia
(b) have decreased oxygen affinity (e) greater if the variant haemoglobin is
(c) be prone to crystallise unstable
(d) be unstable
(e) be abnormally prone to oxidation
Further reading
1.8 Mutations in globin genes Bain BJ, Wild BJ, Stephens AD and Phelan LA. Variant
(a) can occur in α, β, Gγ, Aγ and δ genes Haemoglobins: a Guide to Identification. Wiley‐
(b) always result in a structural abnormality Blackwell, Oxford, 2010.
of haemoglobin The Globin Gene Server, hosted by Pennsylvania
(c) always have harmful effects State University, USA and McMaster University,
(d) can lead to a reduced rate of globin chain Canada. http://globin.cse.psu.edu/
synthesis
(e) can convert one gene to another
References
1.9 Haemoglobin F 1 Attia AM, Ibrahim FA, Abd El‐Latif NA, Aziz SW,
(a) is present, in adult life, in a subset of Abdelmottaleb Moussa SA and Elalfy MS (2018)
erythrocytes referred to as F cells Determination of human hemoglobin derivatives.
(b) is composed of two α chains and two γ Hemoglobin, 39, 371–374.
chains, encoded by two pairs of structur- 2 Umbreit J (2007) Methemoglobin – it’s not just blue:
ally similar α genes and two pairs of a concise review. Am J Hematol, 82, 134–144.
structurally similar γ genes 3 Gross GS and Lane P (1999) Physiological reactions
of nitric oxide and hemoglobin: a radical rethink.
(c) has a sigmoid dissociation curve
Proc Natl Acad Sci U S A, 96, 9967–9969.
(d) constitutes a higher proportion of total
4 Nagel RL and Jaffé ER. CO‐, NO‐, met‐, and
haemoglobin in premature than in
sulf‐hemoglobinemias: the dyshemoglobinemias. In:
full‐term babies
Steinberg MH, Forget BG, Higgs DR and Nagel RL,
(e) on average is present at a somewhat
eds. Disorders of Hemoglobin: Genetics, Pathophysiology,
higher level in women than in men
and Clinical Management. Cambridge University Press,
Cambridge, 2001, pp. 1214–1233.
1.10 Cooperativity is essential for 5 Schechter AN and Gladwin MT (2003) Hemoglobin
(a) a sigmoid oxygen dissociation curve and the paracrine and endocrine functions of nitric
(b) the higher oxygen affinity of haemoglobin oxide. N Engl J Med, 348, 1483–1485.
F in comparison with haemoglobin A 6 Pauling L, Itano HA, Singer SY and Wells IG (1949)
(c) the Bohr effect Sickle‐cell anemia, a molecular disease. Science, 110,
(d) the binding of CO2 to haemoglobin 543–548.
(e) conversion of haemoglobin to 7 Kunkel HG and Wallenius G (1955) New hemo-
methaemoglobin globin in normal adult blood. Science, 122, 288.
26 Chapter 1
8 Bunn HF and Forget BG. Hemoglobin: Molecular, 21 Steinberg MH and Adams JG (1991) Haemoglobin
Genetic and Clinical Aspects. W.B. Saunders, A2: origin, evolution, and aftermath. Blood, 78,
Philadelphia, 1986. 2165–2177.
9 Huehns ER, Flynn FV, Butler EA and Beaven GH 22 Roberts AV, Weatherall DJ and Clegg JB (1972) The
(1961) Two new hemoglobin variants in the very synthesis of human haemoglobin A2 during eryth-
young human embryo. Nature, 189, 496–497. roid maturation. Biochem Biophys Res Commun, 47,
10 Hecht F, Jones RT and Koler RD (1967) Newborn 81–87.
infants with Hb Portland 1, an indicator of α‐chain 23 Stamatoyannopoulos G and Nienhuis AW.
deficiency. Ann Hum Genet, 31, 215–218. Hemoglobin switching. In: Stamatoyannopoulos G,
11 Fantoni A, Farace MG and Gambari R (1981) Nienhuis AW, Majerus PW and Varmus H, eds. The
Embryonic hemoglobins in man and other mam- Molecular Basis of Blood Diseases, 2nd edn. W.B,
mals. Blood, 57, 623–633. Saunders, Philadelphia, 1994, pp. 107–156.
12 Nagel RL and Steinberg MH. Hemoglobins of 24 Weatherall D and Clegg JB. The Thalassaemia
the embryo and fetus and minor hemoglobins of Syndromes, 4th edn. Blackwell Science, Oxford, 2001.
adults. In: Steinberg MH, Forget BG, Higgs DR 25 Ho PJ and Thein SL (2000) Gene regulation and
and Nagel RL, eds. Disorders of Hemoglobin: deregulation: a β globin perspective. Blood Rev, 14,
Genetics, Pathophysiology, and Clinical Management. 78–93.
Cambridge University Press, Cambridge, 2001, pp. 26 Russell JE and Liebhaber SA (1996) The stability of
197–230. human β‐globin mRNA is dependent on structural
13 Kabadi UM, Gopal V, Hood L, Kabadi MU and Platt determinants positioned within its 3′ untranslated
K (1992) Elevated glycosylated hemoglobin concen- region. Blood, 87, 5314–5322.
trations in AIDS. AIDS, 6, 236–238. 27 Jane SM and Cunningham JM (1998) Understanding
14 Zurbriggen K, Schmugge M, Schmid M, Durka S, fetal globin gene expression: a step towards effec-
Kleinert P, Kuster T et al. (2002) Analysis of minor tive Hb F reactivation in haemoglobinopathies. Br J
hemoglobins by matrix‐assisted laser desorption/ Haematol, 102, 415–422.
ionization time‐of‐flight mass spectrometry. Clin 28 Sankaran VG, Xu J, Byron R, Greisman HA, Fisher
Chem, 51, 989–996. C, Weatherall DJ et al. (2011) A functional element
15 Naito C and Niwa T (2000) Analysis of glutathionyl necessary for fetal hemoglobin silencing. N Engl J
hemoglobin levels in diabetic patients by electro- Med, 365, 807–814.
spray ionization liquid chromatography‐mass 29 Liu D, Zhang X, Yu L, Cai R, Ma X, Zheng C et al.
spectrometry: effect of vitamin E administration. (2014) KLF1 mutations are relatively more common
J Chromatogr B Biomed Sci Appl, 746, 91–94. in a thalassemia endemic region and ameliorate the
16 Niketić V, Beslo D, Rajcević S, Sredić S and Stojković severity of β‐thalassemia. Blood, 124, 803–811.
M (1992) Glutathione adduct of hemoglobin (Hb 30 Funnell AP, Prontera P, Ottaviani V, Piccione M,
ASSG) in hemolysates of patients on long‐term Giambona A, Maggio A et al. (2015) 2p15‐p16.1
antiepileptic therapy. Int J Biochem, 24, 503–507. microdeletions encompassing and proximal to
17 Rees DC, Rochette J, Schofield C, Green B, Morris BCL11A are associated with elevated HbF in addi-
M, Parker NE et al. (1996) A novel silent posttransla- tion to neurologic impairment. Blood, 126, 89–93.
tional mechanism converts methionine to aspartate 31 Jiang F, Chen GL, Li J, Zhou JY, Liao C and Li DZ
in hemoglobin Bristol (β67[E11] Val‐Met→Asp). (2018) Analysis of the genotypes in a Chinese popu-
Blood, 88, 341–348. lation with increased Hb A2 and low hematological
18 Mitchell TR and Pegrum GD (1971) The oxygen indices. Hemoglobin, 42, 161–165.
affinity of haemoglobin in chronic renal failure. Br J 32 Bank A (2006) Regulation of human fetal hemo-
Haematol, 21, 463–472. globin: new players, new complexities. Blood, 107,
19 Ponka P (1997) Tissue‐specific regulation of iron 435–443.
metabolism and heme synthesis: distinct control 33 Basu P, Morris PE, Haar JL, Wani MA, Lingrel JB,
mechanisms in erythroid cells. Blood, 89, 1–25. Gaensler KM and Lloyd JA (2005) KLF2 is essential
20 Chen J‐J (2007) Regulation of protein synthesis by for primitive erythropoiesis and regulates the
the heme‐regulated eIF2α kinase: relevance to ane- human and murine embryonic beta‐like globin
mias. Blood, 109, 2693–2699. genes in vivo. Blood, 106, 2566–2571.
34 Boissel JP, Kasper TJ, Shah SC, Malone JI and Bunn The Biochemistry of Human Development. Spastics
HF (1985) Amino‐terminal processing of proteins: International Medical, London, 1971, p. 175, quoted
hemoglobin South Florida, a variant with retention by Weatherall D and Clegg JB. The Thalassaemia
of initiator methionine and N alpha‐acetylation. Syndromes, 3rd edn. Blackwell Scientific, Oxford, 1981.
Proc Natl Acad Sci U S A, 82, 8448–8452. 47 Colombo B, Kim B, Atencio RP, Molina C and
35 Lacan P, Souillet G, Aubry M, Prome D, Richelme‐ Terrenato L (1976) The pattern of foetal haemoglo-
David S, Kister J et al. (2002) New alpha 2 globin bin disappearance after birth. Br J Haematol, 32,
chain variant with low oxygen affinity affecting the 79–87.
N‐terminal residue and leading to N‐acetylation 48 Cheron G, Bachoux I, Maier M, Massonneau M,
[Hb Lyon‐Bron α 1(NA1)Val→Ac‐Ala]. Am J Peltier JY and Girot T (1989) Fetal hemoglobin in
Hematol, 69, 214–218. sudden infant death syndrome. N Engl J Med, 320,
36 Bernini LF and Harteveld CL (1998) α‐thalassaemia. 1011–1012.
Bailliere’s Clin Haematol, 11, 53–90. 49 Garner C, Tatu T, Reittie JE, Littlewood T, Darley J,
37 Serjeant BE, Mason KP and Serjeant GR (1978) The Cervino S et al. (2000) Genetic influences on F cells
development of haemoglobin A2 in normal Negro and other hematologic variables: a twin heritability
infants and in sickle cell disease. Br J Haematol, 39, study. Blood, 95, 342–346.
259–263. 50 Miyoshi K, Kaneto Y, Kawai H, Ohchi H, Niki S,
38 Steinberg MH (1998) Pathophysiology of sickle cell Hasegawa K et al. (1988) X‐linked dominant control
disease. Bailliere’s Clin Haematol, 11, 163–184. of F‐cells in normal adult life: characterization of
39 Steinberg MH and Nagel RL. Hemoglobins of the the Swiss type as hereditary persistence of fetal
embryo, fetus, and adult. In: Steinberg MH, Forget hemoglobin regulated dominantly by gene(s) on X
BG, Higgs DR and Weatherall DJ, eds. Disorders of chromosome. Blood, 72, 1854–1860.
Hemoglobin: Genetics, Pathophysiology, and Clinical 51 Merghoub T, Perichon B, Maier‐Redelsperger M,
Management, 2nd edn. Cambridge University Press, Dibenedetto SP, Samperi P, Ducrocq R et al. (1997)
Cambridge, 2009, pp. 119–135. Dissection of the association status of two polymor-
40 Menzel S, Garner C, Rooks H, Spector TD and phisms in the β‐globin gene cluster with variations
Thein SL (2013) HbA2 levels in normal adults are in F‐cell number in non‐anemic individuals. Am J
influenced by two distinct genetic mechanisms. Br J Hematol, 56, 239–243.
Haematol, 160, 101–105. 52 Craig JE, Rochette J, Sampietro M, Wilkie AOM,
41 Sankaran VG, Joshi M, Agrawal A, Schmitz‐Abe K, Barnetson R, Hatton CSR et al. (1997) Genetic het-
Towne MC, Marinakis N et al. (2013) Rare complete erogeneity in heterocellular hereditary persistence
loss of function provides insight into a pleiotropic of fetal hemoglobin. Blood, 90, 428–434.
genome‐wide association study locus. Blood, 122, 53 Garner C, Silver N, Best S, Menzel S, Martin C,
3845–3847. Spector TD and Thein SL (2004) Quantitative trait
42 Perseu L, Satta S, Moi P, Demartis FR, Manunza L, locus on chromosome 8q influences the switch from
Sollaino MC et al. (2011) KLF1 gene mutations cause fetal to adult hemoglobin. Blood, 104, 2184–2186.
borderline Hb A2. Blood, 118, 4454–4458. 54 Akinsheye I, Alsultan A, Solovieff N, Ngo D,
43 Nitta T, Kawano F, Yamashiro Y, Takagi F, Murata T, Baldwin CT, Sebastiani P et al. (2011) Fetal hemo-
Tanaka T et al. (2015) A new Krüppel‐like factor 1 globin in sickle cell anemia. Blood, 118, 19–27.
mutation (c.947G > A or p.C316Y) in humans causes 55 Pereira C, Relvas L, Bento C, Abade A, Ribeiro ML
β‐thalassemia minor. Hemoglobin, 39, 121–126. and Manco L (2015) Polymorphic variations influ-
44 Borbely N, Phelan L, Szydlo R and Bain B (2013) encing fetal hemoglobin levels: association study in
Capillary zone electrophoresis for haemoglobinop- β‐thalassemia carriers and in normal individuals of
athy diagnosis. J Clin Pathol, 66, 29–39. Portuguese origin. Blood Cells Mol Dis, 54, 315–320.
45 Tarazi IS, Sirdah MM, El Jeadi H and Al Haddad 56 Kesornsit A, Jeenduang N, Horpet D, Plyduang T
RM (2008) Does cigarette smoking affect the diag- and Nuinoon M (2018) Quantitative trait loci influ-
nostic reliability of haemoglobin α2δ2 (HbA2)? encing Hb F levels in Southern Thai Hb E (HBB:
J Clin Lab Anal, 22, 119–122. c.79G>A) heterozygotes. Hemoglobin, 42, 23–29.
46 Huehns ER and Beaven GH. Developmental changes 57 Pakdee N, Yamsri S, Fucharoen G, Sanchaisuriya K,
in human haemoglobins. In: Benson PF, ed. Pissard S and Fucharoen S (2014) Variability of
28 Chapter 1
hemoglobin F expression in hemoglobin EE dis- 68 Tadmouri GO, Yüksel L and Basak AN (1998) HbS/
ease: hematological and molecular analysis. Blood beta(del)‐thalassemia associated with high levels of
Cells Mol Dis, 53, 11–15. hemoglobin A2 and F in a Turkish family. Am J
58 Alter BP, Rosenberg PS, Day T, Menzel S, Giri N, Hematol, 59, 83–86.
Savage SA and Thein SL (2013) Genetic regulation 69 Zertal‐Zidani S, Ducrocq R, Weil‐Oliver C, Elion J
of fetal haemoglobin in inherited bone marrow and Krishnamoorthy R (2001) A novel δβ fusion
failure syndromes. Br J Haematol, 162, 542–546. gene expresses hemoglobin A (HbA) not Hb Lepore:
59 Stadhouders R, Aktuna S, Thongjuea S, Senegalese δ0β+ thalassaemia, Blood, 98, 1261–1263.
Aghajanirefah A, Pourfarzad F, van Ijcken W et al. 70 Kim SY, Lee SH, Cho SI, Song SH, Hattori Y, Park S‐K
(2014) HBS1L‐MYB intergenic variants modulate et al. (2010) Molecular identification of the novel
fetal hemoglobin via long‐range MYB enhancers. J Gγ‐β hybrid hemoglobin: Hb Gγ‐β Ulsan (Gγ through
Clin Invest, 124, 1699–1710. 13; β from 19). Int J Lab Hematol, 45, 276–279.
60 NickAria S, Haghpanah S, Ramzi M and Karimi M 71 Blacklock HA, Case J, Chan T, Raizis T, Doocey R,
(2018) Relationship of the interaction between two Fellowes A et al. (2005) Novel sequence insertion in
quantitative trait loci with γ‐globin expression in a Maori patient with transfusion‐dependent beta‐
β‐thalassemia intermedia patients. Hemoglobin, 42, thalassaemia. Br J Haematol, 131, 400–402.
108–112. 72 De Gobbi M, Viprakasit V, de Jong PJ, Yoshinaga Y,
61 Singleton BK, Lau W, Fairweather VS, Burton NM, Cheng J‐F, Hughes JR et al. (2005) Identification of a
Wilson MC, Parsons SF et al. (2011) Mutations in the gain‐of‐function SNP causing a new model of α‐
second zinc finger of human EKLF reduce promoter thalassemia. Blood, 106, 156a.
affinity but give rise to benign and disease pheno- 73 Viprakasit V, Gibbons RJ, Broughton BC, Tolmie JL,
types. Blood, 118, 3137–3145. Brown D, Lunt P et al. (2001) Mutations in the
62 Satta S, Perseu L, Moi P, Asunis I, Cabriolu A, general transcription factor TFIIH result in beta‐
Maccioni L et al. (2011) Compound heterozygosity thalassaemia in individuals with trichothiodystro-
for KLF1 mutations associated with remarkable phy, Hum Mol Genet, 10, 2797–2802.
increase of fetal hemoglobin and red cell protopor- 74 Yu C, Niakan KK, Matsushita M, Stamatoyannopoulos
phyrin. Haematologica, 96, 767–770. G, Orkin SH and Raskind WH (2002) X‐linked throm-
63 Petruzzelli R, Gaudino S, Amendola G, Sessa R, bocytopenia with thalassemia from a mutation in the
Puzone S, Di Concilio R et al. (2010) Role of the amino finger of GATA‐1 affecting DNA binding rather
cold shock domain protein A in the transcriptional than FOG‐1 interaction. Blood, 100, 2040–2045.
regulation of HBG expression. Br J Haematol, 150, 75 Phillips JD, Steensma DP, Spangrude GJ and
689–699. Kushner JP (2005) Congenital erythropoietic por-
64 Habara AH, Shaikho EM and Steinberg MH (2017) phyria, β‐thalassemia intermedia and thrombocyto-
Fetal hemoglobin in sickle cell anemia: the Arab‐ penia due to a GATA1 mutation. Blood, 106, 154a.
Indian haplotype and new therapeutic agents. Am J 76 Breton A, Theodorou A, Aktuna S, Sonzogni L,
Hematol, 92, 1233–1242. Darling D, Chan L et al. (2016) ASH1L (a histone
65 Huang P, Keller CA, Giardine B, Grevet JD, Davies methyltransferase protein) is a novel candidate glo-
JOJ, Hughes JR et al. (2017) Comparative analysis of bin gene regulator revealed by genetic study of an
three‐dimensional chromosomal architecture iden- English family with beta‐thalassaemia unlinked to
tifies a novel fetal hemoglobin regulatory element. the beta‐globin locus. Br J Haematol, 175, 525–530.
Genes Dev, 31, 1704–1713. 77 Faà V, Meloni A, Moi L, Ibba G, Travi M, Vitucci A
66 Ivaldi MS, Diaz LF, Chakalova L, Lee J, Krivega I et al. (2006) Thalassaemia‐like carriers not linked
and Dean A (2018) Fetal γ‐globin genes are regu- to the β‐globin gene cluster. Br J Haematol, 132,
lated by the BGLT3 long noncoding RNA locus. 640–650.
Blood, 132, 1963–1973. 78 Huisman THJ (1997) Gamma chain abnormal
67 Al‐Ali ZA, Fallatah RK, Aljaffer EA, Albukhari ER, human fetal hemoglobin variants. Am J Hematol, 55,
Sadek Al‐Ali N et al. (2018) ANTXR1 intronic vari- 159–163.
ants are associated with fetal hemoglobin in the 79 Steinberg MH and Nagel RL. New and recombinant
Arab‐Indian haplotype of sickle cell disease. Acta mutant hemoglobins of biological interest. In: Steinberg
Haematol, 140, 55–59. MH, Forget BG, Higgs DR and Nagel RL, eds. Disorders
of Hemoglobin: Genetics, Pathophysiology, and Clinical 82 Rodriguez Romero WE, Castillo M, Chaves MA,
Management. Cambridge University Press, Cambridge, Saenz GF, Gu LH, Wilson JB et al. (1996) Hb Costa
2001, pp. 1195–1211. Rica or alpha 2 beta 2 77(EF1)His → Arg: the first
80 Brennan SO, Shaw J, Allen J and George PM (1992) example of a somatic cell mutation in a globin gene.
β141 Leu is not deleted in the unstable haemoglo- Hum Genet, 97, 829–833.
bin Atlanta‐Coventry but is replaced by a novel 83 Wild BJ, Green BN, Lalloz MRA and Layton DM
amino acid of mass 128 daltons. Br J Haematol, 81, (2000) When is a minor haemoglobin fraction worthy
99–103. of investigation? Br J Haematol, 108 (Suppl. 1), 40.
81 Cook RJ, Hoyer JD and Highsmith WE (2006) 84 Huisman THJ, Carver MFH and Efremov GD. A
Quintuple α‐globin gene: a novel allele in a Syllabus of Human Hemoglobin Variants. The Sickle
Sudanese man. Hemoglobin, 30, 51–55. Cell Anemia Foundation, Augusta, GA, 1996.
Answers to questions
1.1 (a) T 1.3 (a) F 1.5 (a) F 1.7 (a) T 1.9 (a) T 1.11 (a) F
(b) F (b) T (b) T (b) T (b) T (b) T
(c) T (c) T (c) T (c) T (c) T (c) T
(d) T (d) T (d) F (d) T (d) T (d) T
(e) T (e) F (e) F (e) T (e) T (e) F
1.2 (a) T 1.4 (a) T 1.6 (a) T 1.8 (a) T 1.10 (a) T
(b) F (b) F (b) T (b) F (b) F
(c) F (c) T (c) T (c) F (c) T
(d) T (d) T (d) T (d) T (d) F
(e) F (e) F (e) T (e) T (e) F
2 Laboratory techniques for the identification
of abnormalities of globin chain synthesis
The diagnosis of disorders of haemoglobin chain cord. Skin prick samples from neonates can be taken
synthesis usually requires a combination of tech into a heparinised capillary tube and sent to the lab
niques. It is important to recognise that many of the oratory as anticoagulated blood or can be absorbed
laboratory tests in routine diagnostic use indicate onto a filter paper and sent as a dried blood spot,
only the physicochemical characteristics of a hae usually referred to as a ‘Guthrie spot’ from the origi
moglobin rather than permitting its precise iden nator of the technique. Anticoagulated blood is
tification. For clinical purposes an adequate more stable than dried blood spots and the bands
presumptive identification usually requires a com obtained on electrophoresis are more distinct.
bination of at least two techniques, with results Samples should be stored at 4°C and ideally
being assessed in relation to the clinical features, the should be tested within a week, as longer storage
ethnic origin of the subject and the blood count and leads to denaturation of haemoglobin and less dis
film [1]. Principles of techniques and selection of tinct bands on electrophoresis. Dried blood spots
technique will be discussed in this chapter. Some are stable for 7–10 days at room temperature.
methods found satisfactory in the laboratory with Samples for testing should be accompanied by
which the author is associated are given as an information as to the full name, gender, date of birth
appendix to this chapter (see p. 84). For precise and ethnic origin of the individual being tested.
technical details and other recommended methods Knowledge of clinical features and family history
the reader is referred to references 2–4. is sometimes essential for adequate interpretation
and information on parental consanguinity is also
useful. When blood is taken for genetic counselling
Sample collection
of potential parents, identifying details of the partner
Laboratory investigations for haemoglobinopathies should also be given so that results of both partners
are most conveniently performed on venous blood can be assessed simultaneously and guidance on
samples anticoagulated with one of the salts of eth genetic risks can be given. Those responsible for
ylene diamine tetra‐acetic acid (e.g. K2EDTA). In the requesting tests and obtaining blood samples should
case of children, anticoagulated capillary samples ensure that samples are not inadvertently obtained
obtained by skin prick (e.g. from the heel) are also after a blood transfusion has been given.
suitable. Testing of neonates can be performed on
cord blood, venous blood or skin prick samples. To
The blood count, film and reticulocyte
reduce the chances of maternal contamination, cord
count
blood samples should be obtained from an umbilical
cord vessel by means of a syringe and needle after A full blood count (FBC) and a blood film examina
wiping away any surface blood. They should not be tion are usually indicated whenever an abnormality
obtained by squeezing blood from the end of the of globin chain synthesis is suspected. The e xception

30
Laboratory techniques for the identification of abnormalities 31
is in neonatal screening for haemoglobinopathies, be added to one volume of cool distilled water,
when usually only a small sample of blood, possibly followed by vigorous mixing on a vortex mixer,
only a dried blood spot, is available for analysis. standing for 2 minutes, followed by mixing again,
The FBC is essential in the assessment of possible centrifugation (preferably at 4°C) and pipetting off
thalassaemia and in the differential diagnosis of the supernatant for use. Lysates are prone to oxida
thalassaemia and hereditary persistence of fetal tion so should be used promptly, always within one
haemoglobin. week. If there is a requirement for long‐term storage
A reticulocyte count is indicated if a blood film of a lysate then a solution of carbon tetrachloride or
shows polychromasia or if haemoglobin H disease toluene is preferred. An alternative is to freeze
or an unstable haemoglobin is suspected (Fig. 2.1). drops of washed red cells by dropping them on to a
layer of liquid nitrogen. When the drops are frozen
and the liquid nitrogen has evaporated they can be
Preparing a red cell lysate
stored at −40°C. Long‐term storage may be needed
A red cell lysate suitable for most purposes is most for control samples for routine use or to retain a ref
simply prepared by adding one drop of red cells to erence preparation of a rare haemoglobin.
20 drops of a proprietary reagent (Haemolysate The method of preparing a red cell lysate can
reagent, Helena Biosciences Europe) and standing affect the ability of a laboratory to detect significant
for 2 minutes; the lysate is then used for testing. abnormalities. For example, in one study it was
Alternatively, two volumes of packed red cells can found that when electrophoresis on agarose at an
alkaline pH was employed it was necessary to pre
pare the lysate with carbon tetrachloride if haemo
globin H were to be quantified correctly. So much
less haemoglobin H was detected with lysates pre
pared in other ways that cases of haemoglobin H
disease could have been missed [5]. The above
methods are, however, generally satisfactory for
cellulose acetate and acid gel electrophoresis and
for high performance liquid chromatography
(HPLC) and capillary electrophoresis.
Haemoglobin electrophoresis
Haemoglobin electrophoresis [6, 7] was previously
the most common technique for the initial detection
and characterisation of a variant haemoglobin. It has
now been supplanted in this role by HPLC and cap
illary electrophoresis but remains important as a
second confirmatory technique. Haemoglobin elec
trophoresis depends on the principle that when pro
teins applied to a membrane are exposed to a charge
gradient (Fig. 2.2) they separate from each other and
can then be visualised by either a protein stain or a
haem stain (Fig. 2.3). Haemoglobin electrophoresis
can be carried out on filter paper, a cellulose acetate
membrane, a starch gel, a citrate agar gel or an
Fig. 2.1 Reticulocyte preparation showing an increased agarose gel. Haemoglobin electrophoresis is best
reticulocyte count in a patient with homozygosity for performed on lysed packed red cells so that a
haemoglobin Bushwick, a mildly unstable haemoglobin. consistent amount of haemoglobin is applied and so
32 Chapter 2
AFSC
Fig. 2.3 Haemoglobin electrophoresis on cellulose

acetate at pH 8.3 showing: (a) haemoglobins A and S
(sickle cell trait); (b) haemoglobins H and A (haemoglobin
H disease); (c) haemoglobins H and A (haemoglobin H
disease); (d) haemoglobins A and S (sickle cell trait);
(e) haemoglobins A, F and S (sickle cell trait in a baby);
(f) haemoglobins A and F (normal baby); (AFSC) control
sample containing haemoglobins A, F, S and C.
a S
b AS + y globulin
Fig. 2.2 Diagram of apparatus for performing haemoglobin c AS
electrophoresis.
that there are no bands caused by the presence of

plasma proteins. If whole blood is used, the pres
ence of a paraprotein or a very high concentration of
AFSC
polyclonal immunoglobulins can lead to a promi
nent band that can be confused with a variant hae
moglobin (Fig. 2.4). If this is suspected, in a
acetate at pH 8.3 showing an abnormal band caused by
laboratory using whole blood for lysate preparation, the presence of a paraprotein: (a) patient with sickle cell
plasma should be removed and packed red cells anaemia; (b) patient with sickle cell trait and a paraprotein
should be washed before a new lysate is prepared. band; (c) patient with sickle cell trait; (AFSC) control
In the body, the presence of a variant β globin sample containing haemoglobins A, F, S and C.
chain may lead to two types of abnormal tetramer:
one with two variant chains and one with one vari variant haemoglobins have an abnormally stable
ant chain and one normal chain. Under the condi tetramer so that hybrid tetramers persist under the
tions of electrophoresis, tetramers tend to dissociate conditions of electrophoresis and two variant bands
into dimers; fast moving dimers reassemble with are seen. The situation is more complex in the case
each other as do slow moving dimers so that only a of an α chain variant, when there will also be variant
single abnormal band representing molecules with haemoglobins A2 and F, with the possibility of
two copies of the variant chain is seen. Exceptionally, hybrid tetramers.
Cellulose acetate electrophoresis HPLC) should be used. If bands are very faint or if
at alkaline pH haemoglobin A appears to be absent (e.g. in a neo
natal sample), a supplementary alternative tech
The most useful initial electrophoretic procedure is
nique should also be used since both HPLC and
electrophoresis on cellulose acetate at alkaline pH
electrophoresis on agarose gel at acid pH are more
(pH 8.2–8.6). This permits the detection and provi
sensitive techniques for the detection of a low con
sional identification of haemoglobins A, F, S/G/D/
centration of a variant or normal haemoglobin. A
Lepore, C/E/O‐Arab, H and a number of less com
supplementary alternative procedure is also indi
mon variant haemoglobins (see Fig. 2.3). Separation
cated in patients with a positive sickle solubility test
is largely but not entirely determined by the electri
and a single band with the mobility of S, in order to
cal charge of the haemoglobin molecule. At this pH,
distinguish homozygosity for haemoglobin S from
haemoglobin is a negatively charged protein and
compound heterozygosity for S and β chain D or G
will move towards the positively charged anode. A
variants, haemoglobin Korle Bu and haemoglobin
control sample containing haemoglobins A, F, S and
Lepore. Flow charts indicating the sequential appli
C should be run with each set of samples. Control
cation of appropriate tests are given in Chapter 7.
samples containing haemoglobins A, F, S, C, D‐
Haemoglobin electrophoresis on cellulose ace
Punjab, E, G‐Philadelphia and N‐Baltimore are com
tate can be used for quantification of normal or
mercially available. It should be noted that if a
variant haemoglobins, either by scanning densi
protein stain rather than a haem stain is used, car
tometry (Fig. 2.5) or by elution followed by spec
bonic anhydrase will be apparent in addition to hae
trophotometry. Scanning densitometry requires
moglobin bands, moving behind haemoglobin A2.
that the cellulose acetate membrane be rendered
With good electrophoretic techniques, haemoglobin
transparent by use of a clearing solution, a proce
F levels greater than 2% can be recognised visually.
dure that some laboratories use routinely, even
Good techniques also permit a split A2 band and a sig
when scanning is not intended. Scanning densi
nificant elevation or reduction of the percentage of
tometry is sufficiently precise to quantify haemo
haemoglobin A2 to be recognised visually. Recognition
globins that are present as a large percentage of
of a split A2 band is useful in distinguishing α chain
total haemoglobin. For example, this technique is
variants, such as haemoglobin G‐Philadelphia, from β
adequate to determine the percentage of haemo
chain variants, such as haemoglobin D‐Punjab. A split
globin S to permit a distinction between sickle cell
A2 band will be present when there is an α chain vari
trait and sickle cell/β thalassaemia or to monitor
ant but not when there is a β chain variant. Recognition
the percentage of haemoglobin S when sickle cell
of a split A2 band is also essential if β thalassaemia trait
anaemia is being treated by exchange transfusion.
is to be diagnosed in individuals who also have a δ
Quantification by densitometry can also be used to
chain variant. Recognition of a split A2 band on cellu
help distinguish haemoglobin Lepore from other
lose acetate electrophoresis requires a fairly heavy
haemoglobins with the same mobility as haemo
application of haemolysate and can be difficult; HPLC
globin S; haemoglobin Lepore comprises about
(see later) is more reliable. Recognition of increased
10% of total haemoglobin whereas haemoglobins D
haemoglobin F on an electrophoretic strip should be
and G comprise 25–50%. Quantification of the hae
followed by precise quantification whenever this is
moglobin A2 percentage by scanning densitometry
necessary for diagnosis. A visual estimation of the pro
is not sufficiently precise for the diagnosis of β tha
portion of haemoglobin A2 can be used as a supplement
lassaemia trait. When cellulose acetate electropho
to precise measurement by an appropriate technique.
resis is used for quantification of haemoglobin A2,
If a variant haemoglobin with the mobility of
elution and spectrometry are required. This is a
haemoglobin S is detected, haemoglobin electro
labour‐intensive technique and, when large num
phoresis should be followed by a sickle solubility
bers of samples require testing, HPLC or auto
test. If this is negative or if abnormal bands with
mated capillary electrophoresis (see later) is
other mobilities are present a supplementary alter
preferred. Microcolumn chromatography is now
native technique (e.g. electrophoresis at acid pH or
rarely used for this purpose.
34 Chapter 2
Fig. 2.5 Scanning densitometry of a

cellulose acetate electrophoretic strip
showing quantification of haemoglobin
S in a patient who had had an
exchange transfusion for sickle cell
anaemia; there is approximately 64%
haemoglobin A and approximately
36% haemoglobin S.
Typical mobilities of normal and variant haemo Other characteristics of haemoglobins that have
globins on cellulose acetate electrophoresis at alka the same mobility as haemoglobin S or haemoglo
line pH are shown diagrammatically in Fig. 2.6. bin C on cellulose acetate electrophoresis are shown
It should be noted that haemoglobin A2 has the in Tables 2.1 and 2.2.
same electrophoretic mobility as haemoglobins
C, E, O‐Arab and the S‐G‐Philadelphia hybrid and
Agarose gel electrophoresis at alkaline pH
it therefore cannot be quantified by cellulose acetate
electrophoresis when any of these variant haemo Agarose gel is an alternative to cellulose acetate for
globins is present. There are subtle differences in electrophoresis at alkaline pH (Fig. 2.7). It is some
the mobility of haemoglobins D‐Punjab and Lepore, what more sensitive for the detection of variant hae
in comparison with haemoglobin S; they are both moglobins present in small amounts but is more
slightly anodal (i.e. slightly faster). Nevertheless, expensive and less convenient than cellulose acetate
various D and G haemoglobins cannot be reliably electrophoresis and is little used.
distinguished from haemoglobin S by use of this
technique in isolation and haemoglobin Lepore can
Citrate agar or agarose gel electrophoresis
be easily distinguished only because it is present in
at acid pH
a much lower amount. There are also subtle
differences between the mobilities of C and E, with If a variant haemoglobin is detected by electro
haemoglobin C moving slightly more slowly than phoresis on cellulose acetate or agarose at alkaline
haemoglobin E (i.e. being more cathodal) and usu pH it is necessary to confirm its identity by an
ally constituting a higher percentage. Nevertheless alternative technique. This can be electrophoresis
a second confirmatory technique is obligatory. on a citrate agar (Fig. 2.8) or, much more commonly,
Cellulose acetate – pH 8.2–8.6 Agarose gel – pH 6.2

A F S C F A S C
Control
S
D-Iran, D-Punjab,
G-Philadelphia,
G-Ferrara, Lepore,
D-Ouled Rabah
Korle-Bu
Hasharon
D-Norfolk
Handsworth
Q-India
C
E, A2
O-Arab
Siriraj
Setif
C-Harlem
(a)
Cellulose acetate – pH 8.2–8.6 Agarose gel – pH 6.2
A F S C F A S C
Control
H
Bart’s
N-Baltimore
J-Baltimore
J-Toronto
Detroit
Tacoma
K-Ibadan
Hofu
(b)
Fig. 2.6 Diagram showing the mobility of normal and variant haemoglobins on cellulose acetate at pH 8.2–8.6 in comparison
with the mobility on agarose gel at pH 6.0–6.2: (a) haemoglobins with mobility close to A, S or C; (b) fast haemoglobins.
Table 2.1 Characteristics of haemoglobin S and some variant haemoglobins with the same mobility as haemoglobin S
on cellulose acetate electrophoresis at alkaline pH.
Haemoglobin Abnormal Usual percentage Mobility on HPLC Usual ethnic origin

globin agarose gel
chain at acid pH
S β 40–45* S S window African ancestry, Arab, Indian
D‐Punjab β 40–45* With A D window Punjabi, Northern European, Greek,

Turkish, Yugoslav, Afro‐ American,
Afro‐Caribbean, Chinese
G‐Philadelphia α 20–25† With A D window‡ African ancestry, Chinese, Italian

25–35†
35–45†
Lepore δβ fusion 7–15 With A A2 window Greek, Italian, Turkish, Cypriot,

Eastern European, English, Spanish,
Afro‐Caribbean
Korle Bu β 40–45 With A§ A2 window West African ancestry
G‐Coushatta β 40–45% With A A2 window Native American, Chinese, Korean,

Japanese, Thai, Italian, Turkish, Algerian
D‐Iran β 36–45 With A A2 window Iran, Pakistan, Italy, Jamaica
Zurich β 22–35 With A A2 window Swiss, Japanese
Hasharon α 15–20 (if Jewish) With S Between S Ashkenazi Jewish, Italians from Ferrara
or 30–35 (if Italian) and C‡ district
HPLC, high performance liquid chromatography.

* Lower if coexisting α thalassaemia trait.
† Depending on the number of normal α genes present.
‡ Variant haemoglobin A2 also present.
§ Or slightly on the S side of A; if present with S gives a broader band than S alone.
Table 2.2 Characteristics of haemoglobin C and some variant haemoglobins with the same mobility as haemoglobin C
on cellulose acetate electrophoresis at alkaline pH.
Haemoglobin Relevant Usual Mobility on HPLC Usual ethnic origin

globin chain percentage agarose gel at
acid pH
A2 δ 2–3.5* With A A2 window All (normal minor

haemoglobin)
C β 40–45† C C window West African ancestry
E β 30–35‡ With A A2 window South‐East Asian
O‐Arab β 40–45 Slightly on C Between S and C windows Eastern European, Afro‐

side of S§ but closer to C window American, Afro‐Caribbean
C‐Harlem β 40–45† With S§ Between S and C but closer West African ancestry
to C window
E‐Saskatoon β 35–40 With A S window Scottish, Turkish
HPLC, high performance liquid chromatography.

* 3.5–8% in most β thalassaemia trait.
† Lower if coexisting α thalassaemia trait.
‡ This percentage includes haemoglobin A2, which has a very similar retention time.
§ O‐Arab and C‐Harlem are more readily distinguished from each other on citrate agar than on agarose; on citrate agar at acid pH,
C‐Harlem moves with S and O‐Arab is between S and A but closer to A.
Fig. 2.7 Haemoglobin electrophoresis

on agarose gel at alkaline pH (pH
8.6): (1) haemoglobin A and
increased A2 (β thalassaemia trait);
(2) haemoglobins A and A2
(normal); (3) haemoglobin A and S
(sickle cell trait); (4) haemoglobins A
and A2 (normal); (5) haemoglobins
A, F and C (haemoglobin C trait in a
baby); (6) haemoglobin A and S
(sickle cell trait); (7) haemoglobins A
and A2 (normal); (8) haemoglobins
A and A2 (normal).
(a) (b) (c) (d) (e) (f) (g) (h) (i) (j)
Fig. 2.8 Haemoglobin electrophoresis on citrate agar at pH

6.0–6.2 showing from left to right: (a) normal (haemoglobin
A); (b) normal (haemoglobin A); (c) sickle trait
(haemoglobins A and S) with co‐inheritance of
Fig. 2.9 Haemoglobin electrophoresis on agarose gel at
G‐Philadelphia; (d) haemoglobins A and J; (e) haemoglobins
pH 6.0–6.2 (with variant haemoglobins listed from
A and C; (f) haemoglobins S and C; (g) haemoglobin S; (h)
below upwards) showing from left to right: (a) control
haemoglobins A and S. (With thanks to Dr Barbara Wild.)
sample with haemoglobins F, A, S and C; (b) F and S
(sickle cell anaemia); (c) haemoglobins F, Köln and A
(heterozygosity for haemoglobin Köln); (d) haemoglobins
agarose gel (Fig. 2.9) at acid pH (e.g. pH 6.0–6.2).
A and C (haemoglobin C trait); (e) haemoglobin S (sickle
Separated haemoglobins are stained with a haem
cell anaemia); (f) haemoglobin S (sickle cell anaemia);
stain such as o‐dianisidine or o‐tolidine. With this (g) haemoglobins A and C (haemoglobin C trait); (h)
technique, separation of haemoglobins is depend control sample with haemoglobins F, A, S and C; (i)
ent not only on their electrical charge but also on control sample with haemoglobins F, A, S and C. (j)
their interaction with various components in the haemoglobins F, Köln and A (heterozygosity for
agar or agarose. Agar contains both agarose and haemoglobin Köln).
agaropectin, a sulphated polysaccharide [8].
Agarose polymerises and is immobile but agaro
pectin is able to complex with some of the amino acid pH will distinguish haemoglobin C from E, C‐
acids of haemoglobin and the haemoglobin–agaro Harlem and O‐Arab and will help to distinguish
pectin complex then migrates towards the anode the latter three variant haemoglobins from each
while any non‐complexed haemoglobin is carried other. Acid agar and agarose gel electrophoresis do
towards the cathode by endo‐osmotic flow [8]. not resolve haemoglobin A2 from haemoglobin A.
There are some differences in the relative mobilities Electrophoresis at acid pH is indicated in the inves
of variant haemoglobins between agarose gel (see tigation of suspected high affinity haemoglobins
Fig. 2.9) and citrate agar (Fig. 2.8). Both techniques even when electrophoresis at alkaline pH is normal,
distinguish S from D/G but do not distinguish since some high affinity haemoglobins have abnor
between most types of D and G. Electrophoresis at mal mobility only at acid pH.
38 Chapter 2
Cellulose acetate – pH 8.2–8.6 Citrate agar – pH 6.2

A F S C F A S C
Control
S
D-Iran, D-Punjab,
G-Philadelphia,
G-Ferrara, Lepore,
D-Ouled Rabah
Korle-Bu
Hasharon
Q-India
C
E, A2
O-Arab
C-Harlem
Fig. 2.10 Diagram showing the mobility of normal and variant haemoglobins on cellulose acetate at pH 8.2–8.6 in
comparison with the mobility of citrate agar gel at pH 6.0–6.2; the fast haemoglobins shown in Fig. 2.6b have the same
mobility as haemoglobin A on citrate agar at acid pH.
Capillary electrophoresis A2 from haemoglobin E but not always from

haemoglobin C, so that for some samples C and A2
In this technique electrophoresis is carried out in are measured together [9, 10]. Separation of haemo
capillary tubes, permitting higher voltages to be globin A2 from haemoglobin Lepore is also possible
used and shortening the run time. It is dependent [11]. Haemoglobin F may be overestimated in
on ion migration in an alkaline buffer and electro‐ patients with diabetes mellitus because of overlap
osmotic flow to separate haemoglobins, which are ping of the haemoglobin F and glycated haemoglo
quantified as they flow past a detector. An alterna bin A peaks [12]. In samples containing haemoglobin
tive designation is ‘capillary zone electrophoresis’, D or S, estimation of haemoglobin A2 by capillary
indicating that the position of a haemoglobin in electrophoresis can be more accurate than
relation to haemoglobin A is used to determine into estimation by HPLC. The method is suitable for the
which ‘zone’ it falls. An electropherogram is pro detection of haemoglobin Constant Spring and is
duced (Figs 2.11 and 2.12). There is a drop‐down more sensitive than HPLC [13]. Haemoglobin
menu indicating which normal and variant haemo Constant Spring can be identified in the presence of
globins are found in each zone. This method has the β thalassaemia trait or haemoglobin E trait [14].
further advantage of small sample size, which Haemoglobin Bart’s and haemoglobin H (Fig. 2.14)
makes it suitable for neonatal screening pro can also be suspected [15].
grammes (Fig. 2.13). Fractions due to post‐transla
tional modification of haemoglobins A, S, C and
other haemoglobins are not separated from the
High performance liquid
unmodified haemoglobin. This simplifies interpre
chromatography
tation but does mean that an incidental diagnosis of
diabetes mellitus, something which is not infre Cation‐exchange high performance liquid chroma
quent with HPLC (see later), will not be made. tography (CE‐HPLC or HPLC) is a process in which
Capillary electrophoresis can separate haemoglobin a mixture of molecules (such as normal and variant
Fig. 2.11 Capillary electrophoresis electropherogram (Sebia Capillarys 3) showing haemoglobins A and A2.
haemoglobins) with a net positive charge is sepa and mobility on cellulose acetate electrophoresis at
rated into its components by their adsorption onto a alkaline pH. The more positively charged
negatively charged stationary phase in a chroma haemoglobins (e.g. haemoglobins S and C) have a
tography column, followed by their elution by a longer retention time, correlating with have a
mobile phase. The mobile phase is a liquid with an slower mobility on cellulose acetate at alkaline pH.
increasing concentration of cations flowing through The application of HPLC to the identification of
the column; the cations in the mobile phase com variant haemoglobins depends on the fact that for
pete with the adsorbed proteins for the anionic each normal or variant haemoglobin on a specified
binding sites. Thus the adsorbed positively charged system there is a characteristic period of time, referred
haemoglobin molecules are eluted from the column to as the retention time, before the haemoglobin
into the liquid phase at a rate related to their affinity appears in the eluate. Many variant haemoglobins
for the stationary phase. When separated in this can be separated from each other although there are
way, they can be detected optically in the eluate, some that overlap with others. Some haemoglobins
provisionally identified by their retention time and can be resolved by HPLC that are not resolved by cel
quantitated by computing the area under the cor lulose acetate electrophoresis at alkaline pH. For
responding peak in the elution profile. There is example, haemoglobins D‐Punjab/Los Angeles and
some correlation between HPLC retention times G‐Philadelphia can be resolved from haemoglobin S
40 Chapter 2
Fig. 2.12 Capillary electrophoresis electropherogram (Sebia Capillarys 3) of an artificial mixture showing the relative
positions of haemoglobins A, F, S, A2 and C.
and from each other. A slower rate of flow of buffer rapid; they use specially designed microbore col
improves the separation of different haemoglobins umns, high precision gradient‐forming liquid
with a similar retention time but increases the time pumps and optical detectors. There is computer
taken to process each sample. The temperature of control and data handling and sometimes com
operation must be carefully controlled to ensure con puter/intranet‐assisted interpretation.
sistent retention times. Haemoglobins eluted from High performance liquid chromatography can be
the column are represented graphically and automat used not only for the detection, provisional identifi
ically quantified by spectroscopy; the retention time cation and quantification of variant haemoglobins
is stated in relation to that expected for a known nor but also for the quantification of haemoglobins A,
mal or abnormal haemoglobin (usually in relation to A2 and F. Control materials for monitoring the pre
haemoglobin A, F, S, C or D). cision of measurement of haemoglobins F and A2
Automated HPLC instruments currently in use are commercially available. HPLC has the follow
are high precision instruments that are moderately ing advantages over haemoglobin electrophoresis:
A F
(a)
A
S
(b)
F S
(c)
Fig. 2.13 Annotated screen shots showing electropherograms (Sebia Capillarys 2) of neonatal screening samples: (a) normal
baby; (b) sickle cell trait; (c) sickle cell anaemia; (d) haemoglobin C trait; (e) sickle cell/haemoglobin C disease;
(f) haemoglobin D trait; (g) haemoglobin E trait. (With thanks to Sarah Brown.) (Continued on pp. 42–43.)
A F C
(d)
Denatured S C
(e)
A
D-Punjab
(f)
Fig. 2.13 Continued.

A F
E
(g)
Fig. 2.14 Capillary electrophoresis electropherogram (Sebia Capillarys 3) in a patient with haemoglobin H disease
showing haemoglobin H (‘Hb Bart suspected’), haemoglobin A, haemoglobin F and haemoglobin A2. The nature of the
haemoglobin H was confirmed by high performance liquid chromatography (HPLC) and by detection of haemoglobin H
inclusions. Note that haemoglobin A2 is reduced, as expected in haemoglobin H disease. The genotype was −α3.7/− −FIL.
44 Chapter 2
• it is less labour intensive; verestimation but without making the result unre
o
• a very small sample is adequate; alistic; (iii) if the haemoglobin A1c is expressed as a
• quantification of normal and variant haemoglob percentage of total haemoglobin, ignoring the pres
ins is available for each sample; ence of the variant (and its glycated fraction) the
• since haemoglobin A2 is quantified, β thalassae percentage will be underestimated. Alternative
mia trait can be diagnosed in a single procedure; methods such as affinity chromatography and
• a larger range of variant haemoglobins can be immunoassay can be used but they also can fail; for
provisionally identified; example, haemoglobin Raleigh undergoes post‐
• many haemoglobin A2 variants can be detected translational acetylation and the acetylated adduct
easily, thus facilitating the differentiation of α and β cannot be glycated [17]. Information relating to spe
chain variants (even those with identical retention cific instruments and common variant haemoglob
times) and making the diagnosis of β thalassaemia ins is available from the NGSP website [18].
trait when a δ chain variant is present more accurate. Haemoglobin A1c may be increased significantly
The main disadvantage is the higher capital and (e.g. by approximately 2% or 23 mmol/mol) by iron
reagent costs. However if the greater labour costs of deficiency, although conflicting results have also
electrophoresis are considered, then in developed been published [19].
countries with high wages the overall costs are Haemoglobinopathy investigations by HPLC not
comparable [16]. infrequently lead to the diagnosis of previously
Considerable skill and experience are needed in unsuspected diabetes mellitus when an increased
interpreting the results of HPLC since the data pro glycated fraction is demonstrated (Fig. 2.15) [20].
duced are quite complex. Glycosylated variant hae Glycated fractions can lead to invalid or mislead
moglobins have a different elution time from the ing results in other ways. A clinically meaningless
non‐glycosylated forms and acetylated haemoglob elevation of glycated haemoglobin can occur in a
ins from the non‐acetylated forms (haemoglobin F patient who has been transfused as the red cells will
is partially acetylated). In addition, a variant hae have been exposed to a high glucose concentration
moglobin may have the same retention time as during storage. Glycated haemoglobin S may have
either a normal haemoglobin or another variant. a retention time the same as, or very similar to, that
For example, haemoglobin E, haemoglobin Korle of haemoglobin A so that patients with sickle cell
Bu, haemoglobin Fort Worth and haemoglobin anaemia may be thought to have a small amount of
Lepore can overlap with haemoglobin A2, and hae haemoglobin A (Fig. 2.16). In patients with sickle
moglobin A2 can be falsely elevated in the presence cell trait, the haemoglobin S percentage was under
of haemoglobin S and falsely reduced in the pres estimated by 7.4% in one study as a result of the
ence of haemoglobin D‐Punjab. closeness of glycated haemoglobin S to haemoglo
Measurement of haemoglobin A1c (haemoglobin bin A0 [21]. With some instruments and pro
A in which the N‐terminal valine is irreversibly gly grammes, the haemoglobin F can merge with the
cated) is used for predicting a diagnosis of diabetes peak resulting from glycated haemoglobin A and
mellitus and for monitoring the adequacy of control may not be detected when it is 0.6% or less [22].
of this condition, the rate of glycation and therefore Conversely, an elevated percentage of glycated hae
the percentage being higher when blood glucose is moglobin can lead to a factitious elevation of hae
often elevated. HPLC is the usual method for quan moglobin F [22]. Glycated haemoglobin A is
tification but can be inaccurate in the presence of a noticeably reduced in patients with a shortened red
variant haemoglobin for the following reasons: (i) cell life span (e.g. in hereditary spherocytosis and
the variant haemoglobin may have the same reten haemoglobin H disease).
tion time as haemoglobin A1c or the two peaks may Certain artefacts need to be recognised; for
overlap, leading to serious overestimation and example, increased bilirubin in the plasma can lead
usually an unrealistic result; (ii) post‐translationally to a sharp peak in the same general area as haemo
modified variant haemoglobin may coincide globin H, haemoglobin Bart’s and acetylated hae
with or overlap with haemoglobin A1c, leading to moglobin F (Fig. 2.17). An early small artefactual
(a)
Fig. 2.15 HPLC chromatogram (Bio‐Rad

Variant II) showing increased haemoglobin
A1c (P2) in two patients with diabetes
mellitus detected during haemoglobinopathy
investigations: (a) in a patient with no
variant haemoglobin; (b) in a patient with
haemoglobin E trait. A glycated fraction of
more than 6% is usually indicative of
diabetes. Note that the level of 8% in the
patient with E trait will be an underestimate
of the total glycated fraction as glycated
haemoglobin E is not included. (b)
46 Chapter 2
Fig. 2.16 HPLC chromatogram (Bio‐Rad Variant II) in a patient with sickle cell anaemia showing glycosylated haemoglobin S
with the same retention time as haemoglobin A (black arrow) and haemoglobin F, including the acetylated form (white arrow).
Fig. 2.17 HPLC chromatogram (Bio‐Rad Variant II) in a patient with a bilirubin of 970 mmol/l showing a peak in the
same region as haemoglobin Bart’s; from left to right the peaks are bilirubin, haemoglobin F, altered haemoglobin A
(two peaks) and haemoglobins A0, A2 and S.
peak represents the effect of injection of the sample electrophoresis. Haemoglobins A, A2, F, S, C, O‐Arab,
into the column. D‐Punjab and G‐Philadelphia can be separated from
The haemoglobins that can be distinguished from each other. However with some instruments haemo
each other vary somewhat between different instru globin E overlaps with haemoglobin A2. A broad
ments and reagent systems. However all systems peak on HPLC may indicate the presence of hybrid
permit the provisional identification of many more molecules that have the normal and the variant β
variant haemoglobins than can be distinguished by chain in the same molecule [23]. The nature of any
Fig. 2.18 Artificial mixture showing the retention times of normal and important variant haemoglobins on HPLC
(Bio‐Rad Variant II). From left to right the peaks are acetylated F (complex peak that is not integrated), F0 (shaded),
P2 (glycated A), P3 (other post‐translationally modified A), A0, A2 (shaded), S and C.
variant haemoglobin detected by HPLC should be Typical elution patterns of normal and variant
confirmed by an alternative technique. haemoglobins are shown in Figs 2.18–2.21 and some
Some HPLC instruments have dual kits and pro of the variants that can have retention times over
tocols that permit both haemoglobin A1c and haemo lapping with those of haemoglobin A2 and haemo
globin A2 to be quantified accurately [24, 25]. They globin S are shown in Table 2.3 [27, 30–34]. It should
also can measure haemoglobin A2 accurately in the be noted that the ‘window’ of retention times
presence of haemoglobin S [24]. However some vari allocated to a normal or a variant haemoglobin by a
ants that separate from haemoglobin A with other manufacturer may be considerably wider than the
instruments may fail to separate with a dual kit [25]. range of retention times actually observed with that
Evaluations of a number of the automated HPLC haemoglobin. It is therefore necessary to consider
systems that are now commercially available have the actual retention time as well as the window in
been published [22, 26–29]. which it falls in making a provisional identification.
48 Chapter 2
(a)
Fig. 2.19 Typical elution patterns on HPLC with the Bio‐Rad Variant II system for normal samples and for common or
clinically important variant haemoglobins; in each specimen, peaks in the F and A2 windows are generally shaded, P2
represents glycated haemoglobin A, P3 other post‐translationally modified A and A0 unmodified A; the small early peak
(far left) is an injection artefact and is ignored in the descriptions that follow: (a) normal adult; (b) normal neonate, peaks
from left to right are acetylated haemoglobin F (a complex of peaks), F0 and A0 (19.5%); (c) premature neonate, peaks
from left to right are acetylated haemoglobin F, F0 and A0 (9.1%); (d) sickle cell trait, peaks from left to right are F, P2, P3,
A0 (with the irregularity on the upward slope being caused by glycated S), A2 (including post‐translationally modified S)
and S; (e) sickle cell anaemia, peaks from left to right are acetylated F, F0, glycated S (‘unknown’ and in A0 window), A2
and S; (f) haemoglobin C trait, peaks from left to right are F, P2, P3, A0, A2, glycated C (in S window) and C (with a
shoulder on the upwards slope representing other post‐translationally modified C); (g) haemoglobin C homozygosity,
peaks from left to right are F, a low peak in the A0 window that probably represents carry‐over from the preceding
specimen, glycated C (in S window) and C (with its shoulder of other post‐translationally modified C); (h) sickle cell/
haemoglobin C compound heterozygosity, peaks from left to right are glycosylated S (in A0 window), A2 (including
other post‐translationally modified S), S plus glycated C and C (with its shoulder of other post‐translationally modified
haemoglobin C); (i) haemoglobin E trait, peaks from left to right are F, P2, P3, A0 and E plus A2; (j) haemoglobin E
homozygosity, peaks from left to right are F, P3 (in this instance, glycated E), other post‐translationally modified E (in A0
window) and E plus A2; (k) haemoglobin Lepore trait, peaks from left to right are small irregular peaks of acetylated F,
F0, P2, P3, A0 and Lepore plus A2; (l) haemoglobin D‐Punjab trait, peaks from left to right are F, P2, P3, unidentified
(probably glycated D), A0, A2 and D‐Punjab; (m) haemoglobin G‐Philadelphia trait, peaks form left to right are F, P2, P3,
A0 (with a shoulder that is likely to represent glycated G‐Philadelphia), A2, G‐Philadelphia and G2 (an A2 variant with a
G‐Philadelphia α chain); (n) heterozygosity for both haemoglobin G‐Philadelphia and haemoglobin C, peaks from left to
right are F, P2, P3, A0, A2, G‐Philadelphia, glycated C (in S window), C and C‐G‐Philadelphia hybrid; (o) haemoglobin
O‐Arab trait, the peaks from left to right are F, P2, P3, A0, A2, glycated O‐Arab, other post‐translationally modified O‐
Arab and unmodified O‐Arab; (p) haemoglobin S/haemoglobin O‐Arab compound heterozygosity, the peaks from left
to right are acetylated F, F0, a peak in the A0 window representing glycated haemoglobin S, A2, glycated O‐Arab, S, other
post‐translationally modified O‐Arab and unmodified O‐Arab. (Continued on pp. 49–56.)
(b)
(c)
50 Chapter 2
(d)
(e)
(f)
(g)
52 Chapter 2
(h)
(i)
(j)
(k)

54 Chapter 2
(l)
(m)
(n)
(o)
56 Chapter 2
(p)
(a)
Fig. 2.20 Typical elution patterns for normal and variant haemoglobins with the Primus Ultra Plus HPLC system, showing:
(a) a normal pattern (with haemoglobins A and A2); (b) a control sample containing haemoglobins F, A, S and C; (c) a patient
sample containing haemoglobins F, A0, A2 and S; (d) a sample from a neonate with compound heterozygosity for haemoglobins
S and C showing acetylated F, F0, S and C. (With thanks to Lisa Farrell.) (Continued on pp. 57–58.)
(b)
(c)
and their N‐containing basic groups are fully

Isoelectric focusing charged (NH3+), giving a net positive charge. At
Isoelectric focusing (IEF) depends on the fact that high pH the converse occurs; the carboxylic acid
the net charge of a protein depends on the pH of the groups are negatively charged (COO−) and the basic
surrounding solution. At a low pH the carboxylic groups are uncharged, giving a net negative charge.
acid groups of proteins are generally uncharged In IEF, various haemoglobins are separated in a gel
(d)
(a)
(b)
Fig. 2.21 Typical elution patterns for normal and variant haemoglobins with the Tosoh G11 HPLC system showing:
(a) a normal pattern (with haemoglobins A and A2); (b) sickle cell trait; (c) haemoglobin C trait; (d) haemoglobin
D‐Punjab trait; (e) haemoglobin E trait (note that with this instrument there is separation of haemoglobins E and A2);
(f) β thalassaemia trait (haemoglobin A2 4.9%). Haemoglobins F and A2 are shown in red. (With thanks to Dr Sukhjinder
Marwah.)
(c)
(d)
(e)
(f)

60 Chapter 2
Table 2.3 Retention times of common normal and variant haemoglobins on the Bio‐Rad Variant II system compared
with other haemoglobins that may have overlapping retention times; common and diagnostically important haemoglobins
are in bold. Information is from reference [30] unless otherwise cited. Some of the haemoglobins that can appear in one
of two windows are listed twice.
‘Window’ Retention Window Haemoglobins that may overlap

time
F 1.10 0.98–1.22 A fraction of glycosylated haemoglobin A, J‐Iran, Marseille, South Florida,

Okayoma [27], Dagestan [27]
Between F 1.23–1.27 A fraction of glycosylated haemoglobin A, Camperdown [27]

and P2
‘P2’* 0.11 1.28–1.50 Beckman, Sherwood Forest, Raleigh, glycosylated A, K‐Woolwich, Hope, I,
Helsinki, and probably Osu‐Christiansborg [31], Pyrgos [27], J‐Abidjian [27],
I‐Texas [27], Olomouc [27], N‐Baltimore [27], Camden [27], J‐Oxford [27]
‘P3’† 1.70 1.50–1.90 N‐Baltimore, Camden, J‐Oxford [32], Grady [27], J‐Guantanamo, Andrew‐
Minneapolis, Hallamshire, Hopkins‐II, Dublin, J‐Norfolk, Le Lamentin, Zambia,
Harlow, J‐Rajapen, J‐Habana, Santa Clara, Buffalo, Austin [32], Luton,
J‐Broussais, I‐High Wycombe, Tatras, J‐Paris I, Gouda, J‐Baltimore, Chicago [27],
Hikari [27], Fukuyama [32], Fannin‐Lubbock [32], J‐Anatolia [32], J‐Mexico [32],
J‐Meerut, Old Dominion, J‐Toronto
A 2.50 1.90–3.10 J‐Toronto [32], Detroit, Ramban, J‐Calabria, J‐Bangkok, K‐Ibadan, Southwark, North
Shore, Milne, Hofu, Athens‐GA, Niigata, Seattle, San‐Diego, Olympia, Johnstown,
Hinwil, Ty Gard, Tacoma, Nigeria, Broomfield, Hinchingbrook, Hekinana, P‐Nilotic,
Buenos Aires, North Manchester, M‐Iwate, City of Hope, Alzette [27], New York,
Ranier [27], Twin Peaks [32], glycosylated S, Hammersmith, Sidcup, Köln (when not
denatured), Silver‐Springs, Tyne, Hounslow, Fountainebleau, Ravenscourt Park,
Chelsea, Bushwick, Barika, Middlesbrough
Between 3.11–3.29 Henri Mondor [27], G‐Georgia [27]

A and A2
A2 3.60 3.30–3.90 Hackney, Abruzzo, G‐Coushatta, Deer Lodge, D‐Iran, M‐Milwaukee [27], M‐Iwate
[27], St Louis [27], Lepore, Kenya, Spanish Town, Fort Worth, D‐Ouled Rabah, E,
Ocho Rios, Cocody [27], G‐Ferrara, Kenitra [27], Osu‐Cristiansborg, G‐Honolulu,
Paddington, G‐Copenhagen, M‐Saskatoon‡, Zurich, Korle Bu, Alabama
D 4.10 3.90–4.30 Korle Bu [32], Alabama, Dhofar, Khartoum, D‐Punjab, Oleander [27], Okazaki,
Etobicoke, Caribbean, Kempsey, G‐Philadelphia, West One, Atago, Maputo [27],
G‐Norfolk
S 4.50 4.30–4.70 G‐Norfolk [27], Stanleyville II, Russ, E‐Saskatoon, G‐Pest, G‐Waimanolo,
Shimonoseki, Savaria, S, Hornchurch, Yakima, St Luke’s, Ottawa, Handsworth,
Q‐Thailand, St Mary’s, A2′, Montgomery [32], Tarrant [27], Kokura [27], Moabit
[27], glycosylated C, G2, Manitoba, Winnipeg, Titusville, Haaglanden [33],
Vellore [34]
Between S 4.71–4.89 Manitoba [27], Winnipeg [27], Titusville [27], Arya [27],Presbyterian, Setif, Q‐
and C India, C‐New Cross, Hasharon, Q‐Iran, O‐Padova [27], S‐Antilles [27], Chad [27]
C 5.10 4.90–5.30 O‐Arab, O‐Indonesia, Constant Spring, Agenogi, Siriraj, C
* A glycosylated fraction of haemoglobin A.

† A minor peak representing other post‐translationally modified haemoglobin A.
‡ Plus a second peak in the C window.
Fig. 2.22 Photograph of isoelectric focusing (IEF) plate

showing, from left to right: (a) haemoglobins F, A and
Bart’s; (b) haemoglobins S and C; (c) haemoglobins S and
F; (d) haemoglobin S; (e) haemoglobins A and D;
(f) haemoglobins A and S; (g) haemoglobins A and E;
(h) normal (haemoglobins A and A2). (With thanks to Dr
Barbara Wild.)
(e.g. an agarose gel) according to their isoelectric

point (pI), which is the point at which they have no
net charge. Commercially available prepared plates
of polyacrylamide or cellulose acetate contain car
rier amphoteric molecules of various pIs, establish
ing a pH gradient across the plate. When a
haemolysate is applied to the prepared plate in a
strong electric field, the haemoglobin molecules
migrate through the plate until they reach the point
at which the pH corresponds to the pI of the haemo
globin. Since the haemoglobin molecule then has no
net charge it remains at that point. The various hae
moglobin bands are stained (Fig. 2.22) and can be
quantified by densitometry (Fig. 2.23). Densitometric
traces can be superimposed on traces of known var
iants to aid in their identification.
Bands on IEF are sharper than those obtained
with cellulose acetate electrophoresis. In addition,
Fig. 2.23 Densitometric scanning of one lane of an IEF
some haemoglobins that cannot be distinguished
plate from a sample showing bands with the mobilities of
from each other by electrophoresis can be separated
A and S on cellulose acetate electrophoresis at alkaline
by IEF. For example some D and G variants (such as pH. The patient’s IEF scan is stippled; comparison with
D‐Punjab/Los Angeles and G‐Philadelphia) can be an A, F, S, C control sample shows that the variant
separated from haemoglobin S and from each other haemoglobin is not S; comparison with control results for
(Fig. 2.24) [35]. Haemoglobins that can be distin haemoglobin D‐Punjab and haemoglobin G‐Philadelphia
guished from each other by IEF differ between identifies the variant as haemoglobin D‐Punjab.
62 Chapter 2
30 H
Bart’s
25 H
Bart’s
20 Bart’s, Portland
H
15
N-Baltimore
10
5
Distance in mm from haemoglobin A
A1C
0 A
Fig. 2.24 Diagram showing the

G-Philadelphia, G-Coushatta, Lepore
G-Pest mobilities of various haemoglobins on
D-Ouled Rabah
D-Punjab an IEF plate: haemoglobins with
similar mobility on haemoglobin
S G-Galveston, G-Norfolk
electrophoresis can be distinguished
G-Ferrara from each other: haemoglobins A2, C
10 D-Iran
and E can be distinguished (but E
Methaemoglobin cannot be distinguished from
C‐Harlem and O‐Arab): haemoglobins
S, D‐Punjab and G‐Philadelphia can be
O-Indonesia
15 A2 distinguished from each other (and
E,C-Harlem, O-Arab
C also from D‐Iran and G‐Galveston but
C-Ziguinchor
G‐Philadelphia has the same pI as
Constant Spring
G‐Coushatta and Lepore).
(Modified from reference 35.)
different instrument/reagent systems. Although important. In dealing with samples from adults, the
quantification by densitometry is possible, preci advantages of IEF over cellulose acetate electropho
sion at low concentrations is poor and this method resis are less important, although the ability to
is therefore not suitable for quantification of haemo characterise haemoglobins D and G more precisely
globin A2. IEF is a more expensive procedure than can be useful. IEF has the disadvantage that minor
electrophoresis on cellulose acetate, both because of components such as methaemoglobin (resulting
greater capital costs and because the cost per test is from ageing of the sample) and glycosylated haemo
greater. It has a role in diagnosis in neonates when globins are resolved as separate bands and this
the ability to use a small sample volume or an eluate adds to the complexity and difficulty in interpreta
from a dried blood spot, together with the ability tion. For routine use in adults, HPLC is a much
to get good separation of bands, is particularly more useful technique.
Capillary isoelectric focusing A sickle solubility test can be performed by pur

chasing the necessary reagents or by purchasing
In this technique IEF is carried out in a capillary
commercial kits that include the necessary reagents
tube, permitting higher voltages to be used and
[3, 36]. A number of kits have been evaluated and
shortening the run time. Haemoglobins are focused
have been found to detect haemoglobin S down to a
at their pI and are then mobilised to pass through a
concentration of 20%, and sometimes below – in
detector. Capillary IEF can be automated.
some cases as low as 8% [36]. A positive and a nega
tive control should be included whenever a patient
Sickle solubility test and other methods sample is tested (Fig. 2.25). If the patient is anaemic
for the detection of haemoglobin S it is essential to correct the haematocrit to about
0.50 in order to avoid false negative tests. This can
Sickle test be done by removing some of the plasma but it is
The presence of haemoglobin S can be demon preferable for a sickle solubility test to be performed
strated by a sickle test in which sickle cell formation on reconstituted packed red cells in order to not
is induced when blood is deoxygenated. A drop of only avoid problems from anaemia but also lessen
blood is sealed between a glass slide and a cover the possibility of false positive tests when an abnor
slip and is sealed with molten paraffin wax so that mal plasma protein is present (Fig. 2.26) or when
the metabolic activity of white cells leads to deoxy
genation. After an appropriate period of time, the
preparation is observed with a microscope. In a
common modification of the method, the drop of
blood is first mixed with a drop of 2% sodium meta
bisulphite and the preparation is observed at 24
hours. These methods are not very suitable for use
in a routine diagnostic laboratory, where their place
has largely been taken by a sickle solubility test.
Sickle solubility test

A sickle solubility test should be performed when Fig. 2.25 Sickle solubility test showing a positive control,
ever a variant haemoglobin with the electrophoretic a negative control, a positive test and a negative test.
or HPLC characteristics of haemoglobin S is
detected. The only exception is when the quantity
of the variant haemoglobin present is so low that a
positive sickle solubility test would be unlikely; in
this circumstance an alternative second technique
should be employed to strengthen the presumptive
identification of the variant haemoglobin. It is also
prudent to perform a sickle solubility test whenever
a variant haemoglobin of uncertain nature is pre
sent, since there are a number of abnormal haemo
globins that have a second amino acid substitution
in addition to the valine that replaces glutamic acid
in haemoglobin S (see Table 4.2). These variant hae Fig. 2.26 False positive sickle solubility test caused by
moglobins may have an electrophoretic mobility increased plasma proteins. From left to right, a negative
that differs from that of haemoglobin S but never control, a positive control (before and after centrifugation)
theless their presence leads to red cell sickling and a false positive result caused by a myeloma protein
in vitro and in vivo. (before and after centrifugation).
64 Chapter 2
there is hyperlipidaemia. Other causes of false posi problems. The wording of the report on such a test
tive tests include a very high count of either white must state that a negative test does not exclude the
cells or nucleated red cells or a Heinz body haemo presence of a low percentage of haemoglobin S and
lytic anaemia (see later). Most methods require that that further testing is necessary and will follow.
all negative or equivocal sickle solubility tests be
centrifuged before reading to increase sensitivity
Other tests for detection of sickle cell
and reliability. It is important that this step is not
disease or haemoglobin S
omitted if samples with a relatively low percentage
of haemoglobin S are to be detected reliably. It A modification of the gel technology used for
should be noted that not only the presence of a blood grouping has been designed for detection of
paraprotein but also the presence of large numbers haemoglobin S – the principle being that cells that
of Heinz bodies can cause a false positive sickle have sickled do not pass through the gel; this test
solubility test. The latter phenomenon can lead to a has been found to be unreliable and cannot be
false positive test in a patient with an unstable recommended [37]. Haemoglobin S can also be

haemoglobin, particularly if the patient has been detected by immunoassay (see later).
splenectomised. Other tests have been developed that are applica
Sickle solubility tests should be capable of giving ble in a resource‐poor setting and are discussed on
positive results in all cases of sickle cell trait beyond p. 360.
the period of early infancy, even when there is
coexisting α thalassaemia trait. However they
Quantification of haemoglobin A2
should not be relied on in early infancy when the
percentage of haemoglobin S may be a great deal
Choice of method
less than 20%. In this circumstance the provisional
identification of haemoglobin S should be based on Haemoglobin A2 can be quantified with acceptable
two independent methods other than a sickle solu accuracy and precision by:
bility test (e.g. IEF plus HPLC, or cellulose acetate • high performance liquid chromatography;
electrophoresis plus an immunoassay). • capillary electrophoresis;
All sickle solubility tests, whether positive, nega • microcolumn chromatography;
tive or equivocal, should be confirmed by haemo • cellulose acetate electrophoresis followed by
globin electrophoresis or an alternative technique in elution and spectrophotometry.
order to: (i) confirm the presence of haemoglobin S; Elution followed by spectrophotometry is only
(ii) distinguish sickle cell trait from sickle cell anaemia satisfactory when relatively small numbers of
and from compound heterozygous states; and (iii) samples are dealt with. Cellulose acetate electro
detect false negative results due to technical error or phoresis followed by scanning densitometry is not
an unusually low percentage of haemoglobin S. If sufficiently precise to be recommended [38, 39].
rapid results are required (e.g. before emergency This technique has a coefficient of variation (CV)
anaesthesia), the distinction between sickle cell trait of around 20%, in comparison with a CV of 3–4%
and sickle cell anaemia or compound heterozygous for the recommended techniques [38]. Similarly,
states can be made with reasonable accuracy by IEF is unsuitable since it has a CV of 20% or
combining a sickle solubility test with a blood film more [40]. However capillary IEF is satisfactory.
and a blood count and assessing these in the light of International Council for Standardization in
clinical features (see Chapter 4). Haematology (ICSH) guidelines are available [41].
In general, a sickle solubility test is not indicated For the purpose of β thalassaemia diagnosis, any
in an infant less than 6 months of age since a nega variant of haemoglobin A2 (e.g. haemoglobin A2′)
tive result may be misleading. However a sickle should be included in the quantification. When
solubility test can sensibly be performed before HPLC is the method used, it should be noted that
emergency anaesthesia since, if it is negative, it is haemoglobin A2′ cannot be quantified in the presence
unlikely that anaesthesia will cause any clinical of haemoglobin S (including traces of haemoglobin S
resulting from carry‐over from a previous sample), variation in the degree of imprecision. In one study
haemoglobin C (since glycosylated haemoglobin C of 40 samples, mean values varied from 2.7 to 3.1%
co‐elutes with A2′) or haemoglobin G‐Philadelphia between five HPLC instruments and between 2.4
(since haemoglobin G2 co‐elutes with A2′). The and 3.0% for three capillary electrophoresis instru
third possibility does not create a problem since ments [43]. The CV of duplicate measurements
the two should be summed. A split A2 band due to varied from 0.5% to 4.4% [43].
haemoglobin A2′ is otherwise detectable by HPLC
(see Fig. 1.12) and is also detectable by capillary
High performance liquid chromatography
electrophoresis (Fig. 2.27).
The standardisation of quantification of haemo High performance liquid chromatography is now
globin A2 by HPLC and capillary electrophoresis the method most often used for the quantification of
remains imperfect. There is both variation in mean haemoglobin A2. In general it is a satisfactory
values on the same samples between different method but haemoglobin A2 may be underestimated
instruments, sometimes even different instruments in the presence of haemoglobin D‐Punjab [32, 44] (as
from the same manufacturer [25, 42], and also a result of partial overlap and a high baseline) and
Fig. 2.27 Capillary electrophoresis, electropherogram of a normal sample, Sebia Capillarys 3, showing haemoglobins A,
A2 and A2′. Note that the sum of A2 and A2′ is normal.
66 Chapter 2
overestimated in the presence of haemoglobin S [45] Ultra II and Tosoh G11) co‐elution does not occur.
(because of the presence of post‐translationally Overestimation of haemoglobin A2 in the presence
modified haemoglobin S in the A2 window). Simi of haemoglobin C has been reported with some
larly, haemoglobin A2 appears as a double peak in instruments (Menarini, Tosoh and Beckman) [43].
the presence of haemoglobin Q‐India, attributable to
glycated Q‐India and likely to give rise to some
Capillary electrophoresis
overestimation; this is counterbalanced by the pres
ence of a variant A2 with a Q‐India α chain (Fig. 2.28). Capillary electrophoresis can be used to quantify
With some instruments (e.g. Bio‐Rad Variant II), haemoglobin A2. The method shows satisfactory
haemoglobin A2 cannot be measured in the presence precision and good correlation with HPLC,
of haemoglobin E, while with others (e.g. Primus although mean values show slight but statistically
Fig. 2.28 HPLC chromatogram (Bio‐Rad Variant II) in a carrier of haemoglobin Q‐India. The peaks from left to right are:
injection artefact, haemoglobin F, glycated haemoglobin A, other post‐translationally modified haemoglobin A,
haemoglobin A0, double peak representing haemoglobin A2 plus glycated Q‐India, other post‐translationally modified
haemoglobin Q‐India, unmodified haemoglobin Q‐India and a variant haemoglobin A2 with a Q‐India α chain.
significant differences. In one study the mean value Microcolumn chromatography

for haemoglobin A2 with Sebia Capillarys electro
Microcolumn chromatography is an anion‐
phoresis was higher than that with HPLC using a
exchange chromatography method, which has now
Primus instrument (mean values 2.8% and 2.3%,
largely been replaced by other methods.
respectively) [9]. Similarly, in a comparison with a
Microcolumns are prepared containing a suspen
Bio‐Rad Variant II instrument, mean values in nor
sion of an anion‐exchange resin in buffer. The resin
mal subjects were 2.95% by HPLC in comparison
is composed of small particles of cellulose cova
with 2.49 with a Sebia Capillarys II instrument [44];
lently bound to small, positively charged molecules.
in a second study with the same instruments, mean
A haemoglobin solution is applied to the column
values were 3.27 by HPLC and 3.17 by capillary
and is adsorbed onto the resin. There is then an
electrophoresis (p < 0.03) [10].
interchange of charged groups between the posi
Capillary electrophoresis is more satisfactory
tively charged resin and the negatively charged
than microcolumn chromatography in patients
haemoglobin molecules, which retards the passage
with sickle cell trait [44, 46] or haemoglobin D trait
of haemoglobin through the column. The strength
[44]. In the presence of haemoglobin S, the meas
of the association of various types of haemoglobin
ured haemoglobin A2 is lower by capillary electro
molecule to the matrix can be altered by changing
phoresis than by HPLC (mean values 3.1% and 4%,
the pH and ionic strength of the eluting solution
respectively) [9]. The reverse is seen in the case of
applied to the column. It is therefore possible to
haemoglobin D, in which haemoglobin A2 is under
elute different haemoglobins selectively by using
estimated on HPLC. Measurement is possible in the
different eluting solutions (Fig. 2.29). When this
presence of haemoglobin E but not always in the
method is used for quantification of haemoglobin
presence of haemoglobin C [10].
Haemoglobin A
Haemoglobin A2
Resin
(a) (b)
Fig. 2.29 Diagrammatic representation of the principle of microcolumn chromatography: (a) both haemoglobin A (pink)
and haemoglobin A2 (dark red) are negatively charged and are bound to the positively charged beads of the anion‐exchange
resin; (b) with application of an eluting solution that alters pH and ionic strength, haemoglobin A2 is eluted while
haemoglobin A remains attached to the matrix beads.
68 Chapter 2
A2 there is elution first of haemoglobin A2 and then, A reference range should be determined in each
using a second eluting solution, of haemoglobin A. individual laboratory, using blood samples from
The two fractions are collected separately and the healthy subjects with a normal haemoglobin con
absorbance of the eluate is read on a spectrophotom centration and normal red cell indices. Results just
eter, permitting expression of the amount of haemo above the upper limit of normal (e.g. 3.4–3.7%)
globin A2 present as a percentage of total haemoglobin. should be regarded as equivocal, should be inter
Alternatively, it is possible to elute only haemoglobin preted in the light of clinical and haematological
A2 and measure total haemoglobin in a second tube. features and should usually be repeated on the
Microcolumn chromatography is a satisfactory same sample and also on a second sample.
technique for quantification of haemoglobin A2
when relatively large numbers of samples are to be
Quantification and determination
assayed. Chromatography columns can be prepared
of distribution of haemoglobin F
by individual laboratories.
It should be noted that the standard column for
Quantification
measurement of haemoglobin A2 in suspected β
thalassaemia trait is not suitable for haemoglobin A2 Quantification of haemoglobin F is useful in the
quantification in the presence of haemoglobin S. diagnosis of δβ thalassaemia and hereditary per
Modified columns for this purpose can be prepared. sistence of fetal haemoglobin, and in monitoring
It should be noted, however, that measuring response to treatment in sickle cell anaemia. An
haemoglobin A2 in the presence of haemoglobin S ICSH guideline is available [47]. Quantification can
is not essential for making a distinction between be by HPLC, capillary electrophoresis or capillary
sickle cell/β+ thalassaemia and sickle cell trait. IEF. In a resource‐poor setting, cellulose acetate
This distinction can be more readily made by quan electrophoresis followed by scanning densitometry
tifying haemoglobin S since S is more than 50% of or elution is applicable. Two‐minute alkali dena
total haemoglobin in sickle cell/β+ thalassaemia turation followed by spectrophotometry (modified
and less than 50% in sickle cell trait. Expressed Betke method) can also give a clinically useful
in another way, there is more haemoglobin A than S result. Scanning densitometry is unreliable below
in sickle cell trait and more haemoglobin S than A in 10–15% and either HPLC or alkali denaturation [3]
sickle cell/β+ thalassaemia. Measuring haemoglobin is then preferred to densitometry. HPLC gives esti
A2 in the presence of haemoglobin S is of limited mates somewhat higher than alkali denaturation
value in distinguishing between sickle cell anaemia [26, 28]. For levels above 50%, scanning densitometry,
and sickle cell/β0 thalassaemia (see pp. 215 and elution plus spectrophotometry or HPLC is preferred
230). It should be noted, in addition, that coexisting as alkali denaturation is inaccurate. Haemoglobin F
α thalassaemia trait will influence these values. can also be quantified by an immunoassay, using
Overlap occurs between these two conditions and radial immunodiffusion, but this method has been
interpretation should be undertaken with caution. found to be inaccurate [48] and therefore cannot
In general, quantifying haemoglobin A2 in the be recommended.
presence of haemoglobin S is not a test that a labo
ratory needs to perform.
High performance liquid chromatography
Microcolumn chromatography should be com
bined with haemoglobin electrophoresis at alkaline Chromatograms should be examined carefully if
pH in order to detect any variant haemoglobin. It haemoglobin F is apparently increased on HPLC
should be noted that an increased haemoglobin A2 since an increased glycosylated haemoglobin is
percentage can occur in the presence of an unstable sometimes misidentified as haemoglobin F [4].
haemoglobin and if there is any reason to suspect It should be noted that, with some HPLC instru
that a case is not a straightforward β thalassaemia ments, acetylated haemoglobin F (F1) appears in the
trait a specific test for an unstable haemoglobin void volume and is not integrated, so that total
should be performed. haemoglobin F is underestimated. If haemoglobin F
appears to be greater than 10% on HPLC, its nature

should be confirmed by an alternative test to
exclude misidentification of haemoglobin N or
haemoglobin J as haemoglobin F [4].
Capillary electrophoresis
Acetylated and other adducts of haemoglobin F do
not separate from F0 so that all are included in the
quantification.
(a)
Distribution of haemoglobin F
The distribution of haemoglobin F between indi
vidual red cells can be determined by the Kleihauer
test or, more reliably, by flow cytometry.
Kleihauer test
A Kleihauer test is useful for confirmation when
ever there appears to be an increased percentage of
haemoglobin F. The distribution of haemoglobin F
between cells can help to distinguish δβ thalassae
mia trait, when the distribution of haemoglobin F
(b)
is usually heterocellular, from many cases of hered
itary persistence of fetal haemoglobin, in which Fig. 2.30 Kleihauer test showing: (a) heterogeneous
the distribution of haemoglobin F is pancellular distribution of haemoglobin F in hereditary persistence of
(see Table 3.14). The Kleihauer test is also applica fetal haemoglobin in contrast with: (b) a control sample
showing a mixture of normal and fetal cells.
ble to the detection of feto‐maternal haemorrhage.
In this setting, it is important to distinguish deep
pink fetal cells from the paler pink cells of a mother Immunoassay for variant
with hereditary persistence of fetal haemoglobin haemoglobins
[49] (Fig. 2.30).
Immunological detection of haemoglobin S
Haemoglobin A and certain variant haemoglobins can
Quantification of cells containing haemoglobin F
be detected immunologically. Kits were previously
Cells containing haemoglobin F (F cells) can be commercially available for the immunodetection of
quantified by flow cytometry using permeabilised haemoglobins S, C, E and A, known respectively as
red cells and a fluorochrome‐conjugated monoclo HemoCard Hemoglobin S, HemoCard Hemoglobin
nal antibody to haemoglobin F [50]. This method is C, HemoCard Hemoglobin E and HemoCard
applicable to the quantification of feto‐maternal Hemoglobin A. This technique is potentially useful
haemorrhage as well as to the study of patients with and, when functioning properly, all HemoCards
disorders of globin chain synthesis. By combining a detected the relevant variant haemoglobins at least
labelled antibody to haemoglobin F with a fluoro down to 10% and sometimes down to 5% [36, 52, 53].
chrome that binds to nucleic acids, it is possible to However there were problems with consistent
quantitate F cells, reticulocytes and F‐reticulocytes availability and quality control of kits. Potential
by flow cytometry [51]. roles, if these problems could be overcome, include
70 Chapter 2
confirmation of the presence of haemoglobin S in a to that of haemoglobin H. Whether this test should
neonate when a variant haemoglobin is detected by also be performed in suspected α0 thalassaemia
IEF or HPLC and confirmation of the presence of depends on prevalence of this condition and other
haemoglobin E or C following detection of a variant resources available for testing (see p. 102).
haemoglobin by HPLC or cellulose acetate The principle of the test is that haemoglobin H
electrophoresis. precipitates following exposure to a mild oxidant
such as brilliant cresyl blue. The appearance of hae
moglobin H inclusions differs according to whether
Immunological detection of α0 thalassaemia
or not the patient has had a splenectomy. If no sple
An immunochromatographic strip has been devel nectomy has been performed, small blue‐staining
oped in China for screening for genetic subtypes of α0 inclusions are evenly distributed through the cell,
thalassaemia in which the ζ gene is not deleted. A an appearance compared to a golf‐ball (Fig 2.31). In
study in Thailand found it to be sensitive when used patients with haemoglobin H disease who have
for screening for − −SEA/αα, with false positives occur had a splenectomy, haemoglobin H is present as
ring in some patients with −α/αα or −α/−α [54]. This preformed Heinz bodies and, in addition, typical
technique will not detect − −THAI/αα or − −FIL/αα, in ‘golf‐ball’ inclusions appear during incubation with
which the ζ gene is deleted. Immunochromatographic vital dyes (Fig. 2.32).
detection of haemoglobin Bart’s can also be used for
screening for α0 thalassaemia, and may not be influ
Detection of an unstable haemoglobin
enced by whether or not the ζ gene is deleted [55]. An
immunochromatographic strip for this purpose has Unstable haemoglobins can be detected using either
also been developed in Thailand (see p. 360). a heat test or an isopropanol test. Tests should be set
up with a positive and a negative control. A positive
result is the appearance of a precipitate at the end
Detection of haemoglobin H inclusions
point of the test that is not present in a normal
A technique for the demonstration of haemoglobin control (Fig. 2.33). The test should be performed
H inclusions should be performed for confirmation promptly, since ageing of the blood can lead to a
when haemoglobin H disease is suspected and false positive result as a result of formation of meth
when haemoglobin electrophoresis or HPLC shows aemoglobin. If any delay has occurred, the negative
a band or peak with a mobility/retention time similar control should be of the same age as the test sample.
Fig. 2.31 A haemoglobin H

preparation from a patient with
haemoglobin H disease showing
haemoglobin H inclusions (blue
arrow) and reticulocytes (red arrow).
Fig. 2.32 A haemoglobin H

preparation in a patient with
haemoglobin H disease who had had
a splenectomy showing haemoglobin
H inclusions (red arrow), reticulocytes
(blue arrow) and preformed Heinz
bodies (green arrow).
Fig. 2.33 An isopropanol test for an

unstable haemoglobin.
Detection of Heinz bodies

An aged fetal sample is suitable as a positive con
trol. The fact that haemoglobin F is less stable than Testing for Heinz bodies, by incubation with either
haemoglobin A complicates testing if an unstable methyl violet or brilliant cresyl blue, is relevant
fetal haemoglobin is suspected. In this circumstance whenever an unstable haemoglobin is suspected
a fetal sample with a similar proportion of haemo (e.g. unexplained anaemia with an increased reticu
globin F should be used as the negative control. locyte count, irregularly contracted cells on a blood
False positive tests can result from the presence of film or blurred bands or small irregular peaks on
haemoglobin S or an increased percentage of hae electrophoresis and HPLC, respectively) (Fig. 2.34).
moglobin F. If Heinz bodies are not detected in a fresh speci
The isopropanol test is less sensitive than the heat men, they may appear after incubation at 37°C for
stability test and some slightly unstable haemo 24 hours. They are more likely to be present after
globins do not give a positive result. splenectomy.
72 Chapter 2
Fig. 2.35 A 2,6‐dichlorophenolindophenol (DCIP) test

Fig. 2.34 A Heinz body preparation (incubation with
being read on a light box with a striped glass slide to
methyl violet) in a patient with haemoglobin
increase legibility.
Southampton. (With thanks to Mr David Roper.)
deficiency are a disadvantage, whereas false positive

Osmotic fragility as a screening test
results from haemoglobin E and α0 thalassaemia
for thalassaemia
are advantageous.
In resource‐poor countries without easy access to
automated blood counters, a simple visual one‐tube
Dichlorophenolindophenol test
osmotic fragility test has been used as a screening
for haemoglobin E screening
test for thalassaemia trait [56]. This test depends on
the fact that the hypochromic cells of thalassaemia When resources are limited, a 2,6‐dichlorophenolin
trait are osmotically resistant. Whereas normal cells dophenol (DCIP) test can be used for screening for
lyse in buffered saline of a certain concentration, haemoglobin E [61, 62] (Fig. 2.35). This reduces the
thalassaemia trait cells do not, so there is a cloudy number of samples that need to be referred to a
suspension rather than a clear solution. Cells from central laboratory for definitive diagnosis. The test
individuals with iron deficiency or haemoglobin E is read visually, but is harder to read than a sickle
trait may also fail to lyse. The test can be done on a solubility test.
finger‐prick sample and the number of patients
requiring phlebotomy and definitive tests is thereby
Quantification of haemoglobin A1c
reduced. In one study using 0.34% saline, all sam
ples from individuals with α0 thalassaemia were Haemoglobin A1c is a glycosylated derivative of
identified [57]. For β thalassaemia trait, 0.34% saline haemoglobin A. Since haemoglobin A1c can be
is not sufficiently sensitive, 0.36% being needed quantified by HPLC or ion‐exchange microcolumn
[58]; 0.45% glycerine‐saline is more sensitive [59]. chromatography, quantification is sometimes per
False positive tests (e.g. as a result of iron deficiency) formed by haemoglobinopathy laboratories.
are not infrequent. Coexisting South‐East Asian The percentage of haemoglobin A1c is deter
ovalocytosis, glucose‐6‐phosphate dehydrogenase mined by red cell life span and by the average
deficiency and α thalassaemia reduce the sensi blood glucose level during the life of the red cell.
tivity of the test for the detection of β thalassaemia, High levels are thus noted in diabetes mellitus,
possibly to as low as 70% in areas where all three particularly when control is poor, and also when
disorders are common [59, 60]. The specificity of there has been a recent arrest of red cell produc
the test for β thalassaemia is reduced by a high tion. Low levels are seen when there is a young
frequency of iron deficiency, haemoglobin E and red cell population, and this has been suggested
α thalassaemia. False positive results from iron as an aid to distinguishing haemolytic anaemia
from other types of anaemia [63]. Quantification Incidental detection of a variant

has also been suggested in order to make a dis haemoglobin
tinction between transient erythroblastopenia of
childhood (increased percentage) and Diamond– Incidental detection and subsequent identification
Blackfan syndrome (normal percentage) [64]. The of a variant haemoglobin, not necessarily of any
mean level is increased in iron deficiency, in one clinical significance, is quite frequent when HPLC is
study the mean being 6.15% pre‐treatment and used for quantification of haemoglobin A1c and aber
5.25% post treatment [65]. rant results are noted. Much less often, a variant
The percentage of haemoglobin A1c, as measured haemoglobin is detected because of a factitiously
by HPLC, can be erroneous in the presence of a low arterial oxygen saturation by pulse oximetry.
variant haemoglobin [8, 31]. Factitious elevation The factitiousness of the result is identified when a
can be the result of an increased percentage of hae measurement of arterial oxygen saturation is nor
moglobin F or the presence of a ‘fast’ haemoglobin mal. The explanation is that the oxy form of the
such as haemoglobin I, J or N. Haemoglobin variant haemoglobin has an absorption peak close
Hope, haemoglobin Raleigh and haemoglobin Osu‐ to the absorption trough of oxyhaemoglobin A at
Christiansborg also have retention times similar to 660 nm, giving a falsely reduced difference between
that of glycosylated haemoglobin A. If there is no the measured peak and the measured trough [66].
haemoglobin A present, haemoglobin A1c will
necessarily be zero. Haemoglobin A1c can also Other more specialised tests
appear to be low if there is a variant haemoglobin Some more specialised tests are required infre
present (e.g. S, C, D, G, E) and the percentage of quently and are generally better performed in a
haemoglobin A1c is quantified as a percentage of regional centre or reference laboratory rather than in
total haemoglobin, ignoring the fact that there will a routine diagnostic laboratory. The exception is in
also be a glycosylated component of the variant certain geographic areas where a high prevalence of
haemoglobin (Fig. 2.36). An alternative technique specific disorders of haemoglobin synthesis makes it
(affinity column chromatography) permits reliable cost effective for laboratories in large hospitals to
quantification of haemoglobin A1c despite the carry out specific specialised tests. For completeness,
presence of a variant haemoglobin. these tests will be discussed briefly in this chapter.
A high glycosylated fraction may be noted dur
ing investigation of a suspected variant haemoglo
Detection of high or low affinity
bin. If the patient is not known to suffer from
haemoglobins
diabetes, the presence of an elevated level should be
reported and definitive testing for diabetes is then An oxygen dissociation curve with determination
appropriate [20]. of P50o2 (the Po2 at which haemoglobin is 50%

Variant II) in a patient with haemoglobin C trait
showing glycosylated haemoglobin A and
glycosylated haemoglobin C: from left to right
the peaks are haemoglobin F (shaded),
glycosylated haemoglobin A (retention time 1.29
minutes), other post‐translationally modified
haemoglobin A, haemoglobin A0, haemoglobin
A2 (shaded), glycosylated haemoglobin C
(retention time 4.6 minutes) and haemoglobin C
with a shoulder on the left of the peak that
represents other post‐translationally modified
haemoglobin C.
74 Chapter 2
saturated) should be performed if a high oxygen

affinity haemoglobin is suspected (Fig. 2.37). Low
affinity haemoglobins (other than haemoglobin S)
are quite uncommon so that although they also are
diagnosed by an oxygen dissociation curve, this
investigation is rarely productive in the investigation
of an unexplained low haemoglobin concentration.
Globin chain electrophoresis

Globin chain electrophoresis (Fig. 2.38) is carried
out on a red cell lysate to which dl‐dithiothreitol
and urea have been added to dissociate the haem
groups and globin chains. Electrophoresis is then
carried out on cellulose acetate membranes using
both acid and alkaline buffer systems. Globin chain
electrophoresis permits a distinction between α and
β chain abnormalities and, when used as a supple
ment to haemoglobin electrophoresis, permits a
presumptive identification of a larger range of
variant haemoglobins. This technique has become
unimportant since the wider availability of HPLC
has provided an alternative technique that is much
more rapid and less labour intensive.
Fig. 2.37 An oxygen dissociation curve showing

Analysis of the rate of globin chain
haemoglobin A (broken line) and a high affinity synthesis
haemoglobin, haemoglobin Heathrow (complete line). The relative rates of synthesis of α and β globin
(With thanks to Mr David Roper.)
chains by reticulocytes or bone marrow cells can
be useful in the diagnosis of thalassaemias. This is
Fig. 2.38 Results of globin chain

synthesis analysis showing: (1) α and β
chain (normal); (2) α, βS and β chain
(sickle cell trait); (3) α, βC and β chain
(haemoglobin C trait); (4) α and β chain
(normal); (5) αvariant, α and β chain
(identifying an unknown variant as
being an α chain variant); (6) αvariant, α and
β chain (identifying an unknown variant
as being an α chain variant); (7) α, βS and
β chain (sickle cell trait); (8) α, βC and β
chain (haemoglobin C trait).
determined by the amount of radioactivity incorpo that do not involve the globin genes; for example,
rated into α and β chains after a fixed period of time. some cases of types I and III congenital dyserythro
This technique is critically dependent on the blood poietic anaemia have been found to have a reduced
or bone marrow sample being fresh (i.e. less than α:β ratio (e.g. 0.76 in one case of type III congenital
6 hours old). The results are usually expressed as a dyserythropoietic anaemia) [68].
ratio of α:β or α:β+γ. The results of such analysis are The more ready availability of molecular tech
shown in (Fig. 2.39) and typical globin chain ratios niques for the diagnosis of α thalassaemia has
in various conditions are shown in Table 2.4. An greatly reduced the need for analysis of the rate of
abnormal globin chain ratio does not always indicate α and β globin chain synthesis.
an abnormality in the rate of synthesis. If a globin
chain is very unstable the ratio can be abnormal
Deoxyribonucleic acid analysis
because of very rapid destruction of the unstable
chain. Very unstable α chains have been associated The most important clinical applications of deoxy
with a reduced α:non‐α ratio at 5 minutes, but a ribonucleic acid (DNA) analysis are: (i) the confir
paradoxically increased ratio at 15 minutes and mation of the diagnosis of α0 thalassaemia trait,
1 hour, which could lead to a misdiagnosis as β tha particularly for genetic counselling; (ii) confirmation
lassaemia intermedia [67]. An abnormal ratio has of the presence of haemoglobin D‐Punjab (clinically
also sometimes been detected in inherited disorders important) rather than of any other D or G group
Fig. 2.39 Globin chain synthesis

studies in a patient with haemoglobin
C/β+ thalassaemia compound
heterozygosity. (With thanks to
Professor Lucio Luzzatto.)
Table 2.4 α:β globin chain synthesis

ratio in normal subjects and in Normal or type of thalassaemia α:β or α:non‐α globin chain synthesis ratio
various thalassaemias.
Normal 0.95–1.05
α thalassaemia
One gene deletion 0.65–0.80
Two gene deletion 0.38–0.60
Haemoglobin H disease 0.20–0.30
β thalassaemia trait 1.67–2.22
β thalassaemia major No β chain production or 3.6–30

76 Chapter 2
haemoglobin; and (iii) the prenatal diagnosis of haemoglobin, since it is silent on electrophoresis and
serious disorders of haemoglobin synthesis in the chromatography [69]. For fetal diagnosis, tests are
fetus (essential for first trimester diagnosis). In usually carried out on DNA obtained by chorionic
some communities where α, β and δ thalassaemias villous sampling, a procedure that can be carried out
and complex interactions are relatively common and by 9–10 weeks of gestation. A technique for analysis
where silent β thalassaemia occurs, there is a need for of fetal DNA in maternal plasma is currently under
more extensive application of DNA analysis to per development and evaluation.
mit accurate diagnosis prior to genetic counselling. Some of the techniques that can be applied are
In Thailand and Malaysia, where haemoglobin shown in Table 2.5 [1, 70–72]. Techniques are modi
Malay represents around 15% of β thalassaemia fied according to the mutations expected in a given
alleles, DNA analysis for antenatal diagnosis should population. For β thalassaemia, the polymerase
include primers for the detection of this variant chain reaction (PCR) is the method of choice when
Table 2.5 Techniques for diagnosis of thalassaemias and haemoglobinopathies by DNA analysis [1, 70–72].
Diagnosis Test
α0 thalassaemia Gap PCR*

PCR with allele‐specific primers (ARMS‐PCR)† (including multiplex
PCR for deletions common in a specific region) and RQ‐PCR [71])
Reverse dot blot analysis using mutation‐specific probes
ELISA for detection of embryonic ζ chains
DNA sequencing
β thalassaemia
Known common mutations PCR with allele‐specific primers

Reverse dot blot analysis using mutation‐specific probes
Gap PCR*
DGGE or heteroduplex analysis‡
Unknown mutations
RFLP linkage analysis§
DNA sequencing
Hb Lepore Gap PCR, sequencing
δβ thalassaemia Gap PCR, sequencing
Deletional hereditary persistence of fetal Gap PCR, sequencing

haemoglobin
Hb S PCR, Ddel digestion
Hb C PCR with allele‐specific primers
Hb E PCR with allele‐specific primers or restriction enzymes
Hb D‐Punjab PCR, EcoRI digestion
Hb O‐Arab PCR, EcoRI digestion
ARMS, amplification refractory mutation system; DGGE, denaturing gradient gel electrophoresis; DNA, deoxyribonucleic acid; ELISA,
enzyme‐linked immunosorbent analysis; PCR, polymerase chain reaction; RFLP, restriction fragment length polymorphism; RQ‐PCR, real
time (quantitative) PCR.
* GAP PCR indicates that there is a ‘gap’ in the DNA sequence (i.e. a deletion) and the primers are chosen so that, if the deletion is present,
there is amplification across the ‘gap’. This is useful for the diagnosis of α thalassaemia and the minority of cases of β thalassaemia that result
from a relatively large deletion. Other haemoglobinopathies resulting from deletion are also susceptible to detection by this technique.
† ARMS is a PCR technique using two primer sets, one amplifying normal sequences and one abnormal sequences.
‡ DGGE or heteroduplex analysis can be used initially to locate the mutation.
§ Requires study of the family.
the likely mutations are known. Multiplex PCR without the mutation; however, since direct
and use of techniques such as the detection of the sequencing has become less costly, this is increas
PCR product by an enzyme‐linked immunoassay ingly used when PCR has not demonstrated one of
(ELISA) permit automation of the process [73]. the more common mutations.
When the likely nature of a mutation is unknown, α thalassaemia, including α0 thalassaemia, can be
detection can be based on linkage analysis using diagnosed by Southern blot analysis of genomic
restriction fragment length polymorphisms (RFLP), DNA using α and ζ probes (Fig. 2.40). However
which requires study of family members with and PCR (Fig. 2.41) is cheaper and faster and primers
A B C
0 10 20 30
ζ2 inter-ζHVR ψζ1 ψα2 ψα1 α2 α1 θ1 3’HVR
(a)
1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8
C
B
B
(b) (c)
Fig. 2.40 Southern blot analysis following hybridisation of a ζ globin gene probe to BglII digests of genomic deoxyribonucleic
acid (DNA), performed for the diagnosis of α thalassaemia: (a) Diagram of the α globin gene cluster showing sites where BglII
cleaves the DNA (↓); BglII digestion of normal DNA produces three fragments (A, B and C) that will hybridise with the ζ
probe: fragment A is a small fragment containing the ζ2 gene; fragments B and C are larger and each contains part of the ψζ1
gene. Fragment B is of variable length because it contains the inter‐ζ hypervariable region (inter‐ζHVR) and for this reason its
position on a gel is variable, with two distinct B bands often being present. (b) Gel showing: lane 1, αα/‐α3.7; lane 2, αα/‐α4.2;
lane 3, −α3.7/−α3.7; lane 4, αα/−α3.7; lane 5, −α3.7/−α3.7; lane 6, −α3.7/−α3.7; lane 7, −α3.7/−α3.7; lane 8, αα/αα. (With thanks to Dr
Tom Vulliamy.) (c) Explanatory diagram: Lane 8, with normal α genes, shows normal A, B and C fragments. Lanes 1 and 4
show three normal fragments but, in addition, there is a larger fragment that represents an abnormal C fragment consequent
on deletion of the 3′ BglII cleavage site by the −α3.7 deletion; since fragments of normal size are also present it can be seen that
these individuals are heterozygous for this deletion. Lanes 3, 5, 6 and 7 show loss of the normal C fragment and replacement
by a larger C fragment characteristic of −α3.7; these individuals are therefore homozygous for −α3.7. Lane 2 shows three normal
fragments but in addition there is a fragment that is smaller than B and C; this represents a C fragment of reduced size
consequent on a −α4.2 deletion; as fragments of normal size are also present this individual must be a heterozygote.
78 Chapter 2
0 10 20 30
ζ2 inter-ζHVR ψζ1 ψα2 ψα1 α2 α1 α1 3’HVR
8 9
(a)
(b)
Fig. 2.41 Polymerase chain reaction (PCR) in the diagnosis of α thalassaemia. (a) Explanatory diagram modified from
reference [75] showing an α gene cluster and the three primers (7, 8 and 9) described in this paper for the diagnosis of
the α0 thalassaemia determinant − −SEA. In a normal genome, primers 7 and 8 amplify a small fragment of DNA but
primers 7 and 9 are too widely separated for amplification to occur; in the presence of the large − −SEA deletion the
sequence to which primer 8 anneals is deleted and the sequences to which primers 7 and 9 bind are brought sufficiently
close together that a fragment is amplified; heterozygotes will have two fragments of different sizes whereas a hydropic
fetus with − −SEA/− −SEA will have a single abnormal fragment. (b) Gel using the above technique showing a control
sample from a subject with αα/− −SEA (far right) and 17 individuals being tested, one of whom (third from left) has
αα/− −SEA. (With thanks to Dr Tom Vulliamy.)
have now been designed to permit the diagnosis

Electrospray ionisation mass spectrometry
of − −MED, − −FIL, − −SEA, − −THAI, −α3.7, −α4.2, −(α)20.5, αNCoIα,
αHphIα and also αααanti3.7 by PCR [74–77]. α thalassaemia Electrospray ionisation mass spectrometry was pre
can also be diagnosed by a reverse dot blot method viously a research technique that was further devel
[78] in which membrane‐bound oligonucleotide oped for the identification of variant haemoglobins
allele‐specific probes are used as targets for hybridi [80, 81] and is now being applied in a modified
sation of amplified DNA (Fig. 2.42); filter strips are form to antenatal and neonatal haemoglobinopathy
prepared with a probe for the normal α gene and for screening. If a mass measurement scan of whole
the variant being sought, permitting the detection blood is followed by scanning a tryptic digest, giv
of both heterozygotes and homozygotes. ing mass:charge measurement of the peptides pro
Other techniques applicable to DNA analysis, duced, variant haemoglobins can be identified [81].
including restriction enzyme PCR and high resolu It is possible to determine whether the variant
tion melting analysis (HRMA) for nucleotide varia haemoglobin is an α or β chain variant, to estimate
tions, and MPLA (multiple ligation‐dependent the proportion of the variant and to predict the
probe amplification) for detection of larger amino acid substitution that could account for any
deletions and duplications, are summarised by
observed change in mass. When there is only a
Vrettou et al. [79]. small quantity of a variant haemoglobin present,
30 59 QS CS
A
Normal
Fig. 2.42 Reverse dot blot analysis for the detection of four
T
non‐deletional α thalassaemia determinants, codon 30, codon 59,
αQuong Sze and αConstant Spring. Amplified DNA samples were hybridised A
to strips each containing normal (A) and mutant (T) oligonucleotide Heterozygote
T
probes for the particular defect; positive signals appear as blue dots.
Samples from heterozygotes show a signal with both the normal
and thalassaemia probe whereas samples from homozygotes show a A
Homozygote
signal only with the thalassaemia probe. (With thanks to Professor T
V. Chan and the British Journal of Haematology.)
prior separation by electrophoresis may be needed. 2.1 The following variant haemoglobins have the
Highly unstable haemoglobins may not be identi same mobility as haemoglobin S on cellulose
fied. The apparatus is very expensive and consider acetate electrophoresis at pH 8.4
able skill is required in interpretation of results. (a) haemoglobin C
The original technique identified variants on the (b) haemoglobin Lepore
basis of mass differences and mass:charge analysis (c) haemoglobin D‐Punjab
of peptides produced by tryptic digestion. The cur (d) haemoglobin G‐Philadelphia
rent technique, which is automated, depends on (e) haemoglobin E
measurement of the mass:charge ratio using multi
ple reaction monitoring so that certain specific vari 2.2 The following haemoglobins have the same
ants are looked for. Hence, if the common variants mobility as haemoglobin A on agarose gel
S, C, D‐Punjab, E and O‐Arab are targeted, the pro electrophoresis at pH 6.2
cedure is useful for antenatal and neonatal screen (a) haemoglobin E
ing [82, 83]. The introduction of instruments more (b) haemoglobin S
suitable for use in a routine diagnostic laboratory (c) haemoglobin D‐Punjab
may lead to greater use of the technique. (d) haemoglobin C
(e) haemoglobin G‐Philadelphia
2.3 Satisfactory methods for the precise quantifica

Quality assurance tion of the haemoglobin A2 percentage include
All tests should be carried out by appropriately (a) scanning densitometry
trained personnel following a standard operating (b) high performance liquid chromatography
procedures (SOP) for each test. A laboratory requires (HPLC)
clearly defined protocols for different clinical situa (c) isopropanol test
tions (e.g. for antenatal testing or for the investiga (d) capillary electrophoresis
tion of neonates (see Chapter 7)). All laboratories (e) cellulose acetate electrophoresis followed
carrying out haemoglobinopathy testing should by elution
participate in an external quality assurance scheme.
2.4 The following variant haemoglobins have the
same mobility as haemoglobin C on cellulose
acetate electrophoresis at pH 8.4
Check your knowledge
(a) haemoglobin A2
One to five answers may be correct. Answers to (b) haemoglobin Lepore
almost all questions can be either found in this (c) haemoglobin O‐Arab
chapter or deduced from information given. (d) haemoglobin C‐Harlem
Answers are given on p. 84. (e) haemoglobin E
80 Chapter 2
2.5 A false positive sickle solubility test may be 2.11 An α:β chain synthesis ratio of 0.25:1 is
caused by compatible with
(a) the presence of haemoglobin C (a) α thalassaemia trait
(b) anaemia (b) normal
(c) increased plasma proteins (c) β thalassaemia trait
(d) the presence of haemoglobin D (d) haemoglobin H disease
(e) the presence of numerous Heinz bodies (e) β thalassaemia major
2.6 If an unstable haemoglobin is suspected,

relevant tests include Further reading
(a) blood film Bain BJ, Wild BJ, Stephens AD and Phelan LA.
(b) reticulocyte count Variant Haemoglobins: a Guide to Identification.
(c) isopropanol test Wiley‐Blackwell, Oxford, 2010.
(d) heat stability test Wild BJ and Bain BJ. Investigation of variant haemo
(e) alkali denaturation globins and thalassaemias. In: Bain BJ, Bates I
and Laffan MA, eds. Dacie and Lewis’s Practical
2.7 Techniques applicable in the diagnosis of α or Haematology, 12th edn. Elsevier, London, 2017.
β thalassaemia in a fetus include
(a) reverse dot blot analysis
(b) polymerase chain reaction (PCR) References
(c) isopropanol test
1 Working Party of the General Haematology Task
(d) linkage studies using restriction fragment
Force of the British Committee for Standards in
length polymorphism (RLFP) analysis
Haematology (1998) The laboratory diagnosis of
(e) microcolumn chromatography
haemoglobinopathies. Br J Haematol, 101, 783–792.
2 International Committee for Standardization in
2.8 Quantification of haemoglobin A2 can be
Haematology (1988) Recommendations for neo
inaccurate in the presence of
natal screening for haemoglobinopathies. Clin Lab
(a) a raised haemoglobin F
Haematol, 10, 335–345.
(b) diabetes mellitus
3 Wild BJ and Bain BJ. Investigation of abnormal
(c) haemoglobin D‐Punjab
haemoglobins and thalassaemias. In: Bain BJ,
(d) haemoglobin E Bates I and Laffan MA, eds. Dacie and Lewis’s
(e) haemoglobin S Practical Haematology, 12th edn. Elsevier, London,
2017, pp. 282–311.
2.9 The presence of haemoglobin S may be
4 Wild BJ and Bain BJ (2004) Detection and quantita
conclusively demonstrated by
tion of normal and variant haemoglobins: an analyti
(a) a positive sickle test
cal review. Ann Clin Biochem, 41, 355–369.
(b) a positive sickle solubility test 5 Lafferty J, Ali M, Carstairs K and Crawford L (1998)
(c) an immunoassay The effect of carbon tetrachloride on the detection
(d) PCR with appropriate restriction enzymes of hemoglobin H using various commercially
(e) electrophoresis on cellulose acetate at pH 8.3 available electrophoresis products. Am J Clin
Pathol, 109, 651–652.
2.10 A positive test for an unstable haemoglobin 6 International Committee for Standardization in
can result from Haematology (1978) Simple electrophoretic system
(a) an aged sample for presumptive identification of abnormal haemo
(b) the presence of a high concentration of globins. Blood, 52, 1058–1064.
haemoglobin F 7 International Committee for Standardization in
(c) α thalassaemia trait Haematology (1978) Recommendations for a system
(d) β thalassaemia trait for identifying abnormal hemoglobins. Blood, 52,
(e) the presence of an unstable haemoglobin 1065–1067.
8 Hoyer JD and Kroft SH, eds. Color Atlas of Hemo 20 Bain BJ, Beer P, Conroy M, Hughes C and Phelan L
globin Disorders. College of American Pathologists, (2005) The frequency of detection of unexpected
Northfield, 2003. diabetes mellitus during haemoglobinopathy
9 Keren DF, Hedstrom D, Gulbranson R, Ou CN and investigations, J Clin Pathol, 64, 898–900.
Bak R (2008) Comparison of Sebia Capillarys capil 21 Li S‐H, Harro D, Alfsen J, Bolterman J, Ketha H,
lary electrophoresis with the Primus high‐pressure Schroeder LF et al. (2018) Effect of diabetes mellitus
liquid chromatography in the evaluation of hemo on sickle hemoglobin quantitation in sickle cell
globinopathies. Am J Clin Pathol, 130, 824–831. trait. Am J Clin Pathol, 150, 105–115.
10 Borbely N, Phelan L, Szydlo R and Bain B (2013) 22 Wild BJ and Stephens AD (1997) The use of auto
Capillary zone electrophoresis for haemoglobinop mated HPLC to detect and quantitate haemoglob
athy diagnosis. J Clin Pathol, 66, 29–39. ins. Clin Lab Haematol, 19, 171–176.
11 Agouti I, Merono F, Bonello‐Palot N and Badens C 23 Mais DD, Boxer LA, Gulbranson RD and Keren DF
(2013) Analytical evaluation of the Capillarys 2 Flex (2007) Hemoglobin Ypsilanti: a high‐oxygen‐affin
piercing for routine haemoglobinopathies diagnosis. ity hemoglobin demonstrated by two automated
Int J Lab Hematol, 35, 217–221. high‐pressure liquid chromatography systems. Am
12 Grey V, Wilkinson M, Phelan L, Hughes C and J Clin Pathol, 128, 850–853.
Bain BJ (2007) Inaccuracy of high‐performance 24 Lafferty JD, McFarlane AG and Chui DHK (2002)
liquid chromatography estimation of haemoglobin F Evaluation of a dual hemoglobin A2/A1c quantita
in the presence of increased haemoglobin A1c. Int tion kit on the Bio‐Rad Variant II automated hemo
J Lab Hematol, 29, 42–44. globin analyzer. Arch Pathol Lab Med, 126, 1494–1500.
13 Waneesorn J, Panyasai S, Kongthai K, Singboottra P 25 Riou J, Godart C, Mathis M, Hurtrel D, Wajcman H,
and Pornprasert S (2011) Comparison between cap Préhu C and Bardahdjian J (2005) Evaluation of the
illary electrophoresis and high performance liquid Bio‐Rad Variant II HbA2/HbA1C dual program for
chromatography for detection and quantification of measurement of hemoglobin concentrations and
Hb Constant Spring [Hb CS; α142, Term→Gln detection of variants. Clin Chem Lab Med, 43, 237–243.
(TAA>CAA IN α2)]. Hemoglobin, 35, 338–345. 26 Waters HM, Howarth JE, Hyde K, Goldstone S,
14 Pornprasert S, Saoboontan S and Punyamung M Kadkhodaei‐Elyaderani M, Cinkotai KI and Richards
(2015) Detection of Hb Constant Spring (HBA2: JT (1996) Evaluation of the Bio‐rad Variant Beta
c.427T>C) heterozygotes in combination with β‐ Thalassaemia Short Program. MDA Evaluation Report
thalassemia or Hb E trait by capillary electrophore MDA/96/28. Medical Devices Agency, London.
sis. Hemoglobin, 39, 211–215. 27 Riou J, Godart C, Didier H, Mathis M, Bimet C,
15 Liao C, Zhou JY, Xie XM, Li R and Li DZ (2012) Bardakdjian‐Michau J et al. (1997) Cation‐exchange
Cord blood analysis for rapid prenatal confirma HPLC evaluated for presumptive identification of
tion of Hb Bart’s disease using the Sebia Capillary hemoglobin variants. Clin Chem, 43, 34–39.
electrophoresis system. Hemoglobin, 36, 186–191. 28 Bain BJ and Phelan L (1997) Evaluation of the Primus
16 Phelan L, Bain BJ, Roper D, Jury C and Bain K Corporation CLC330TM HPLC system for haemo
(1999) An analysis of relative costs and potential globinopathy screening. MDA Evaluation Report
benefits of different policies for antenatal screening MDA/97/53, Medical Devices Agency, London.
for β thalassaemia trait and variant haemoglobins. 29 Bain BJ and Phelan L (1997) Evaluation of the
J Clin Pathol, 52, 697–700. Kontron Instruments Haemoglobin System PV for
17 Chen D, Crimmins DL, Hsu FF, Lindberg FP and Scott haemoglobinopathy screening. MDA Evaluation
MG (1998) Hemoglobin Raleigh as the cause of a Report MDA/97/54, Medical Devices Agency, London.
falsely increased hemoglobin A1C in an automated ion‐ 30 Bain BJ, Wild BJ, Stephens AD and Phelan LA.
exchange HPLC method. Clin Chem, 44, 1296–1301. Variant Haemoglobins: a Guide to Identification. Wiley‐
18 NGSP (2019) Hb A1c Assay interferences (2018) Blackwell, Oxford, 2010.
www.ngsp.org/interf.asp (accessed June 2019). 31 Kapoor D, Ng JP and Jones TH (2005) Falsely high gly
19 English E, Idris I, Smith G, Dhatariya K, Kilpatrick cated haemoglobin [HbA1C] because of haemoglobin
ES and John WG (2015) The effect of anaemia and Osu‐Christiansborg. Clin Lab Haematol, 27, 334–335.
abnormalities of erythrocyte indices on HbA1c anal 32 Joutovsky Hadzi‐Nesic J and Nardi MA (2004)
ysis: a systematic review. Diabetologia, 58, 1409–1421. HPLC retention time as a diagnostic tool for
82 Chapter 2
aemoglobin variants and hemoglobinopathies: a

h the World Health Organization International
study of 60 000 samples in a clinical; diagnostic Reference Reagent for Hb A2 quantitation (89/666)
laboratory. Clin Chem, 50, 1737–1747. for continued use. Int J Lab Hematol, 41, 262–270.
33 Harteveld CL, Ponjee G, Bakker‐Verweij M, 43 Paleari R, Gulbis B, Cotton F and Mosca A (2012)
Arkesteijn SGJ, Phylipsen M and Giordano PC (2012) Interlaboratory comparison of current high‐perfor
Hb Haaglanden: a new nonsickling β7Glu>Val vari mance methods for HbA2. Int J Lab Hematol, 34,
ant. Consequences for basic diagnostics, screening, 362–368.
and risk assessment when dealing with HbS‐like 44 Higgins TN, Khajuria A and Mack M (2009)
variants. Int J Lab Hematol, 34, 551–555. Quantification of HbA2 in patients with and without
34 Edison ES, Sathya M, Rajkumar SV, Nair SC, β‐thalassemia and in the presence of HbS, HbC, HbE,
Srivastava A and Shaji RV (2012) A novel β‐globin and HbD Punjab hemoglobin variants: comparison
gene mutation HBB.c.22 G>C produces a hemo of two systems. Am J Clin Pathol, 131, 357–362.
globin variant (Hb Vellore) mimicking HbS in 45 Head CE, Conroy M, Jarvis M, Phelan L and Bain BJ
HPLC. Int J Lab Hematol, 34, 556–558. (2004) Some observations on the measurement of
35 Basset P, Beuzard Y, Garel MC and Rosa J (1978) haemoglobin A2 and S percentages by high perfor
Isoelectric focusing of human haemoglobins: its mance liquid chromatography in the presence and
application to screening, to characterization of 70 absence of thalassaemia. J Clin Pathol, 57, 276–280.
variants and to study of modified fractions of nor 46 Shihabi ZK, Hinsdale ME and Daugherty HK (2000)
mal haemoglobins, Blood, 51, 971–982. Haemoglobin A2 quantification by capillary zone
36 Bain BJ and Phelan L (1996) An evaluation of the electrophoresis. Electrophoresis, 21, 749–752.
HemoCard Hemoglobin S test and four sickle cell 47 Stephens AD, Angastiniotis M, Baysal E, Chan V,
solubility kits (Ortho Sickledex, Dade Sickle‐Sol, Davis B, Fucharoen S et al.; International Council for
Microgen Bioproducts S‐Test and Lorne Sickle‐ The Standardisation of Haematology (ICSH) (2012)
Check) for the detection of haemoglobin S and the ICSH recommendations for the measurement of
HemoCard Hemoglobin A plus S test for the detec haemoglobin F. Int J Lab Hematol, 34, 14–20.
tion of haemoglobins A and S. MDA Evaluation Report 48 Schultz JC (1999) Comparison of radial immunodif
MDA/96/56, Medical Devices Agency, London. fusion and alkaline cellulose acetate electrophoresis
37 Balasubramaniam J, Phelan L and Bain BJ (2001) for quantitating elevated levels of fetal hemoglobin
Evaluation of a new screening test for sickle cell (HbF): application for evaluating patients with
haemoglobin. Clin Lab Haematol, 23, 379–383. sickle cell disease treated with hydroxyurea. J Clin
38 International Committee for Standardization in Lab Anal, 13, 82–89.
Haematology (1975) Recommendations for selected 49 Bain BJ and Chapple L (2015) Interpreting a post
methods for quantitative estimation of Hb A2 and partum Kleihauer test. Am J Hematol, 90, 77.
for Hb A2 reference preparations. Br J Haematol, 38, 50 Hoyer JD, Penz CS, Fairbanks VF, Hanson CA and
573–578. Katzmann JA (2002) Flow cytometric measurement
39 Thalassaemia Working Party of the BCSH General of hemoglobin F in red cells. Am J Clin Pathol, 117,
Haematology Task Force (1994) Guidelines for the 857–863.
investigation of the α and β thalassaemia traits. 51 Mundee Y, Bigelow NC, Davis BH and Porter JB
J Clin Pathol, 47, 289–295. (2001) Flow cytometric method for simultaneous
40 Shephard DTS (1995) An assessment of isoelectric assay of foetal haemoglobin containing red cells,
focusing (Helena REP Hb‐IEF) for the measurement reticulocytes and foetal haemoglobin containing
of haemoglobin A2 percentage in the diagnosis of reticulocytes. Clin Lab Haematol, 23, 349–354.
β‐thalassaemia trait. BSc project, St Mary’s Hospital 52 Bain BJ and Phelan L (1996) An assessment of
Medical School, London. HemoCard Hemoglobin C and HemoCard
41 Stephens AD, Angastiniotis M, Baysal E, Chan V, Hemoglobin E kits for the detection of haemoglob
Fucharoen S, Giordano PC et al.; International Council ins C and E. MDA Evaluation Report MDA/96/57,
for the Standardisation of Haematology (ICSH) Medical Devices Agency, London.
(2012) ICSH recommendations for the measurement 53 Chapman C and Chambers K (1997) Neonatal
of haemoglobin A2. Int J Lab Hematol, 34, 1–13. haemoglobinopathy screening methods. MDA

42 De la Salle B, Stephens AD, Wild BJ, Harteveld CL Evaluation Report MDA/97/64, Medical Devices
and Hyde K (2019) Evaluation of the suitability of Agency, London.
54 Jomoui W, Fucharoen G, Sanchaisuriya K and Chinese family as a cause of spurious hypoxaemia

Fucharoen S (2017) Screening of (–SEA) α‐thalassae measured by pulse oximetry. J Clin Pathol, 63,
mia using an immunochromatographic strip assay 847–949.
for the ζ‐globin chain in a population with a high 67 Traeger‐Synodinos J, Papassotiriou I, Mataxotou‐
prevalence and heterogeneity of haemoglobinopa Mavrommati A, Vrettou C, Stamoulakatou A and
thies. J Clin Pathol, 70, 63–68. Kanavakis E (2000) Distinct phenotypic expression
55 Prayalaw P, Fucharoen G and Fucharoen S (2014) associated with a new hyperunstable alpha globin
Routine screening for α‐thalassaemia using an variant (Hb Heraklion,α1cd37(C2)Pro>): compari
immunochromatographic strip assay for haemoglo son to other α‐thalassemic hemoglobinopathies.
bin Bart’s. J Med Screen, 21, 120–125. Blood Cell Mol Dis, 26, 276–284.
56 Kattamis C, Efremov G and Pootrakul S (1981) 68 Wickramasinghe SN (1998) Congenital dyserythro
Effectiveness of one tube osmotic fragility screen poietic anaemias: clinical features, haematological
ing in detecting β‐thalassaemia trait. J Med Genet, morphology and new biochemical data. Blood Rev,
18, 266–270. 12, 178–200.
57 Panyasai S, Sringam P, Fucharoen G, Sanchaisuriya 69 Fucharoen S, Sanchaisuriya K, Fucharoen G and
K and Fucharoen S (2002) A simplified screening test Surapot S (2001) Molecular characterization of
for α‐thalassemia 1 (SEA type) using a combination thalassemia intermedia with homozygous Hb
of a modified osmotic fragility test and a direct PCR Malay and Hb Malay/HbE in Thai patients.
on whole blood lysates. Acta Haematol, 108, 74–78. Haematologica, 86, 657–658.
58 Chow J, Phelan L and Bain BJ (2005) Evaluation of 70 Galanello R, Sollaino C, Paglietti E, Barella S, Perra
single‐tube osmotic fragility as a screening test for C, Doneddu I et al. (1998) α‐Thalassaemia carrier
thalassemia. Am J Hematol, 79, 1–4. identification by DNA analysis in the screening for
59 Penman BS, Gupta S and Weatherall DJ (2015) thalassemia. Am J Hematol, 59, 273–278.
Epistasis and the sensitivity of phenotypic screens 71 Chan V, Yiop B, Lam YH, Tse HY, Wong HS and
for beta thalassaemia. Br J Haematol, 169, 117–128. Chan TK (2001) Quantitative polymerase chain
60 Fucharoen G, Fucharoen S, Singsanan S and reaction for the rapid prenatal diagnosis of homozy
Sanchaisuriya K (2007) Coexistence of Southeast Asian gous α‐thalassaemia (Hb Barts hydrops fetalis). Br J
ovalocytosis and beta‐thalassemia: a molecular and Haematol, 115, 341–346.
hematological analysis. Am J Hematol, 82, 381–385. 72 Brancaleoni V, Di Pierro E, Motta I and Cappellini
61 Issaragrisil S, Siritanaratkul N and Fucharoen S. MD (2016) Laboratory diagnosis of thalassemia. Int
Diagnosis and Management of Thalassemia: Thailand J Lab Hematol, 38, Suppl 1, 32–40.
as a Model. American Society of Hematology, 73 Ugozzoli LA, Lowery JD, Reyes AA, Lin C‐IP, Re A,
Education Program Book, American Society of Locati F et al. (1998) Evaluation of the BeTha gene 1
Hematology, Washington DC, 2001, pp. 483–488. kit for the qualitative detection of the eight most
62 Chapple L, Harris A, Phelan L and Bain BJ (2005) common Mediterranean β‐thalassaemia mutations.
Reassessment of a simple chemical method using Am J Hematol, 59, 214–222.
DCIP for screening for haemoglobin E. J Clin Pathol, 74 Bowden DK, Vickers MA and Higgs DR (1992)
59, 74–76. A PCR‐based strategy to detect common severe
63 Panzer S, Kronik G, Lechner K, Bettelheim P, determinants of α‐thalassaemia. Br J Haematol, 81,
Neumann E and Dudczak R (1982) Glycosylated 104–108.
hemoglobins (GHb): an index of red cell survival, 75 Oron‐Karni‐V, Filon D, Oppenheim A and Rund D
Blood, 59, 1348–1350. (1998) Rapid detection of the common Mediterranean
64 Karsten J, Anker AP and Odink RJ (1996) α‐globin deletions/rearrangements using PCR. Am J
Glycosylated haemoglobin and transient erythro Hematol, 58, 306–310.
blastopenia of childhood. Lancet, 347, 273. 76 Hunt JA, Lee L, Donlon TA and Hsia YE (1999)
65 El‐Agouza I, Shahla AA and Sirdah M (2002) The Determination of the breakpoint of the common α‐
effect of iron deficiency anaemia on the levels of thalassaemia deletion in Filipinos in Hawaii, Br J
haemoglobin subtypes: possible consequences for Haematol, 104, 284–287.
clinical diagnosis. Clin Lab Haematol, 24, 285–289. 77 Liu YT, Old JM, Miles K, Fisher CA, Weatherall DJ
66 So C‐C, Chan AY, Chow EY, Lin S‐Y, Yip S‐F, Szeto and Clegg JB (2000) Rapid detection of α‐thalassae
S‐C and Chan L‐C (2010) Haemoglobin Bonn in a mia deletions and α‐globin gene triplication by
84 Chapter 2
multiplex polymerase chain reactions. Br J Haematol, 81 Wild BJ, Green BN and Stephens AD (2004) The
108, 295–299. potential of electrospray ionization mass spectrome
78 Chan V, Yam I, Chen FE and Chan TK (1999) A try for the diagnosis of hemoglobin variants found in
reverse dot‐blot method for rapid detection of newborn screening. Blood Cells Mol Dis, 22, 308–317.
non‐deletion α thalassaemia. Br J Haematol, 104, 82 Daniel YA, Turner C, Haynes RM, Hunt BJ and
513–515. Dalton RN (2005) Rapid and specific detection of
79 Vrettou C, Kakourou G, Mamas T and Traeger‐ clinically significant haemoglobinopathies using
Synodinos J (2018) Prenatal and preimplantation electrospray mass spectrometry‐mass spectrome
diagnosis of hemoglobinopathies. Int J Lab Hematol, try. Br J Haematol, 130, 635–643.
40, Suppl 1, 74–82. 83 Moat SJ, Rees D, George RS, King L, Dodd A, Ifederu
80 Wild BJ, Green BN, Cooper EK, Lalloz MRA, Erten S, A et al. (2017) Newborn screening for sickle cell dis
Stephens AD and Layton DM (2001) Rapid identifica orders using tandem mass spectrometry: three years’
tion of hemoglobin variants by electrospray ionization experience of using a protocol to detect only the dis
mass spectrometry. Blood Cells Mol Dis, 27, 691–704. ease states. Ann Clin Biochem, 54, 601–611.
2.1 (a) F 2.4 (a) T 2.7 (a) T 2.10 (a) T
(b) T (b) F (b) T (b) T
(c) T (c) T (c) F (c) F
(d) T (d) T (d) T (d) F
(e) F (e) T (e) F (e) T
2.2 (a) T 2.5 (a) F 2.8 (a) F 2.11 (a) F

(b) F (b) F (b) F (b) F
(c) T (c) T (c) T (c) F
(d) F (d) F (d) T (d) T
(e) T (e) T (e) T (e) F
2.3 (a) F 2.6 (a) T 2.9 (a) T

(b) T (b) T (b) F
(c) F (c) T (c) T
(d) T (d) T (d) T
(e) T (e) F (e) F
Appendix Helena BioSciences Europe, Gateshead, Tyne and

Wear, UK.
The following methods and kits have been found High Performance Liquid Chromatography
satisfactory in the laboratory with which the author is Bio-Rad Variant II
associated. Bio-Rad Laboratories, (UK) Ltd, Watford,
Electrophoresis on cellulose acetate at alkaline pH Hertfordshire, UK.
Titan III-Hb (catologue 3022) Capillary electrophoresis
Supre Heme buffer (catologue 5802) Capillarys 2 Flex Piercing
Haemolysate reagent (catalogue 5127). Sebia (UK) Ltd, Blackwater, Surrey, UK.
Helena BioSciences Europe, Gateshead, Tyne and Sickle solubility test: Streck, Sickledex®.
Wear, UK. Distributed by Alpha Laboratories, Eastleigh,
Electrophoresis on agarose gel at acid pH Hampshire, UK.
SAS-MX Acid Hb Kit (catalogue 100700) For other recommended methods, see reference 3.
Haemolysate reagent (catalogue 5127).
3 α, β, δ and γ thalassaemias and related conditions
Thalassaemia is the name given to a globin gene dis of the α genes are affected. γ thalassaemia would
order that results in a diminished rate of synthesis of only be of potential significance in intrauterine and
one or more of the globin chains and consequently a early neonatal life when haemoglobin F is a major
reduced rate of synthesis of the haemoglobin or hae haemoglobin. However, since there are four γ genes,
moglobins of which that chain constitutes a part. significant disease is unlikely. δ thalassaemia is of
The condition was first described by Cooley and Lee no clinical significance except that its presence may
in 1925 [1] with the name ‘thalassaemia’ from the interfere with the diagnosis of coexisting β
Greek θαλαασσα, sea, via ‘thalassic anaemia’, being thalassaemia.
given by Whipple and Bradford in 1936 [2]. In the Thalassaemia usually results from mutation of a
late 1930s the hereditary nature of thalassaemia was single globin gene or from deletion of one or more
clearly identified in both Greece and Italy. It is prob globin genes. One mechanism of deletion is une
able that the failure to identify thalassaemia as a dis qual crossover between chromosomes at meiosis so
crete entity in the Mediterranean area until after its that parts of two genes are deleted and the 5′ end of
description in the USA was because malaria, as a one gene fuses with the 3′ end of another gene. This
cause of childhood anaemia and splenomegaly, was is the mechanism underlying some types of α tha
still prevalent around the Mediterranean. lassaemia trait (Fig. 3.1). Another defect of this type
In thalassaemia, a significantly reduced rate of leads to deletion of part of both a δ gene and a β
synthesis of one type of globin chain leads to unbal gene, with production of a δβ fusion gene, leading
anced chain synthesis with excess of a normal glo to synthesis of haemoglobin Lepore (named from
bin chain contributing to the pathological effects, the Italian family in whom it was first described)
causing either damage to erythroid precursors and (see Fig. 1.14a). The rate of synthesis of the abnor
ineffective erythropoiesis or damage to mature mal δβ chain is slower than the rate of synthesis of β
erythrocytes and haemolytic anaemia or both. chain so that haematological features are very simi
Thalassaemia can be classified according to the phe lar to those of β thalassaemia. Since one δ gene has
notype or the genotype. The major types of thalas effectively been lost there is also a reduced rate of
saemia, classified according to the genotype, are synthesis of δ chain and a reduced proportion of
shown in Table 3.1. Thalassaemia can result from haemoglobin A2. A thalassaemic phenotype can
deletion of a large part or all of a gene (as is usual in also result from a mutation that leads to formation
α thalassaemia) or from a small deletion or other of a very unstable haemoglobin molecule or a very
mutation of a gene (as is usual in β thalassaemia). unstable globin chain that precipitates before it is
Mutations of α and β genes are of potential clinical incorporated into haemoglobin.
significance since there is a reduced rate of synthe Imbalance of globin chain synthesis can result
sis of haemoglobin A, the major haemoglobin of from deletion or mutation of a globin gene, leading
adult life. A serious clinical disorder usually results to a reduced rate of synthesis of the relevant globin
only when both of the β genes or either three or four chain, but also from duplication of a globin gene,

85
86 Chapter 3
Table 3.1 Classification of the thalassaemias.
Type of thalassaemia Chain or chains synthesised at Haemoglobin or haemoglobins

a reduced rate synthesised at a reduced rate
Alpha: α0 or α+ α A, A2 and F
Beta: β or β
0 +
β A
Gamma: γ γ F
Delta: δ or δ0 +
δ A2
Delta beta: δβ or δβ 0 +
δ and β A and A2
A
gamma delta beta: γδβ A 0 A
γ, δ and β A and A2
Epsilon gamma delta beta*: ε γ γδβ G A 0

ε, γ, γ, δ and β
G A
A, A2 and F*
Haemoglobin Lepore δ and β A and A2†
* Often referred to as γδβ thalassaemia; in fetal life there is decreased synthesis of haemoglobins Gower 1, Gower 2 and Portland 1.
† Haemoglobin Lepore is synthesised at a reduced rate in comparison with haemoglobin A but at an increased rate in comparison with
haemoglobin A2.
α2 α1 fusion
–α3.7
Unequal cross-over
between α2 and α1
genes during meiosis
α2 α1 α2 α1 Fig. 3.1 Diagrammatic representation
of the likely mechanism for the
αααanti 3.7 occurrence of the α gene deletion
α2 α1 α2 fusion α1
(–α3.7) and triple α (αααanti3.7) by
unequal crossover between
Likely mechanism of development of –α3.7 and αααanti 3.7 by unequal cross- homologous sequences of the α2 and
over between homologous sequences of the α2 and α1 genes during meiosis
α1 genes during meiosis.
leading to an increased rate of synthesis of that glo

α thalassaemias
bin chain. Thus there may be three, four or even five
α genes on a single chromosome (instead of the nor The α thalassaemias are a group of conditions
mal two), referred to as triple α, quadruple α and resulting from a reduced rate of synthesis of α glo
quintuple α, respectively. Duplication of a gene is bin. There are more than 100 genetic defects under
not usually of any clinical significance but when it lying α thalassaemia [3]. The severity of the defect is
coexists with thalassaemia it may either lessen the very variable. At one extreme is a completely
chain imbalance and ameliorate the condition or asymptomatic condition, resulting from deletion or
aggravate the chain imbalance and thus increase dysfunction of one of the four α genes, which pro
the severity of the disorder. Thus if one chromo duces either a trivial abnormality in the blood count
some has ααα and the other has −α there is likely to and film or no abnormality at all. At the other
be negligible chain imbalance and no haematologi extreme is haemoglobin Bart’s hydrops fetalis, a
cal effect. However the coexistence of ααα/αα or condition that is generally incompatible with life,
ααα/ααα and β thalassaemia increases the severity usually resulting from deletion of all four α genes
of the β thalassaemia. and consequent total lack of α globin synthesis.
α, β, δ and γ thalassaemias and related conditions 87
Some of the clinicopathological features of α thalas normal α genes if the locus control region is
saemia syndromes result from the lack of α globin deleted.
chain while others result from damage to red cell α thalassaemia can be divided broadly into dele
precursors and mature red cells by excess non‐α tional and non‐deletional thalassaemia. Deletional α
chains. Excess β and γ chains can, in the absence of thalassaemia results in either α0 or α+ thalassaemia,
sufficient α chain, form haemoglobins with β or γ depending on the length and nature of the deletion.
chain tetramers. The resultant haemoglobins are Non‐deletional α thalassaemia usually leads to α+
haemoglobin H with β tetramers, first described by thalassaemia trait. Non‐deletional α thalassaemias
Rigas and colleagues in 1955 [4], and haemoglobin can result from mutations of the α2 gene (αTα thalas
Bart’s with γ tetramers, first described by Fessas saemia) or the α1 gene (ααT thalassaemia). More than
and Papaspyrou in 1957 [5]. Haemoglobin H was so 70 such mutations have been reported but most of
named because haemoglobin G (now known as these are rare. Recognised cases of non‐deletional
haemoglobin Korle Bu) had been described a year thalassaemia are much more often caused by muta
earlier [6]. The current name of haemoglobin Bart’s, tion of the α2 gene (HBA2) than of the α1 gene
previously designated haemoglobin Fessas and (HBA1). This is likely to be, at least in part, because
Papaspyrou or haemoglobin F and P, was given by the former has a much more severe phenotype.
Ager and Lehmann in 1958 [7], because the patient Mutation in the α2 gene has a more severe pheno
in whom they observed it was a patient of St type than deletion of the α2 gene as upregulation of
Bartholomew’s Hospital. the α1 gene occurs only in the latter instance.
When both α genes on a single chromosome are Important deletions that lead to α thalassaemia
deleted or transcriptionally completely inactive are shown diagrammatically in Fig. 3.2 and dele
the designation α0 thalassaemia (α zero thalassae tions and mutations are summarised in Tables 3.2
mia) is used. When there is some residual gene and 3.3 [8–25]. Deletions and mutations causing α
function and some production of α globin chain thalassaemia fall into nine broad categories:
directed by genes on that chromosome, for exam
ple when only one of the two α genes on a chromo
Deletional
some is deleted, the designation α+ thalassaemia (α
plus thalassaemia) is used. Function of the α glo 1 Deletion of all or part of one or both α genes:
bin gene cluster at 16p13.3 requires the presence of • deletion of an α2 gene (α+ thalassaemia) (e.g. −α4.2);
a major upstream regulatory element, the locus • deletion of an α1 gene (α+ thalassaemia);
control region alpha (LCRA), previously referred • deletion of part of α2 gene and part of α1 gene
to as HS −40 because is 40 kb upstream of the ζ2 with formation of a fusion α gene (α+ thalassaemia)
locus; thalassaemia can occur with completely (e.g. −α3.7);
Telomere Centromere
0 10Kb 20Kb 30Kb
ζ2 ψζ1 ψα2 ψα1 α2 α1 θ1
––SEA α0
–– FIL
α0
––THAI α0
––MED α0
of some common deletions that can –(α)20.5 α0
lead to α0 and α+ thalassaemia; the –α3.7 α+
shaded blocks indicate the length of –α4.2 α+
the deletion.
88 Chapter 3
Table 3.2 Classification of deletional α thalassaemia.
Type of deletion Phenotype Number of examples Examples

recognised*
Deletion involving one or both α genes

Deletion of all or part of one α gene α+ thalassaemia 7 −α4.2, −α3.7I, −α3.7II, −α3.7III,
−α3.5, α(α)5.3†, −α2.7
Deletion of all or part of both α α0 thalassaemia 20 −−SEA, −−THAI, −−MED,

genes but without deletion of LCRA −−FIL, −−BRIT, −−SPAN,
−−YEM, –(α)20.5†, –(α)5.2†
Deletion of both α genes and of α0 thalassaemia 8, without other −−DUTCHII

LCRA (100–>250 kb) phenotypic abnormality
Extensive loss of 16p13.3 (1–2 Mb) α0 thalassaemia 17, with mental −−BO
including both α genes and LCRA‡ retardation and
dysmorphism
Deletion of α1 gene and 18–20 kb α0 thalassaemia 1 (α)−ZF†

downstream of α1 gene [18]
Deletion leaving α genes intact

Deletion of upstream major α0 or very severe 12 (αα)RA§, (αα)TAT§, (αα)MM§,
regulatory element (LCRA, MCS‐R2) α+ thalassaemia (αα)IJ§, (αα)Jx§
without deletion of α genes [25]
* Likely to be an underestimate.
† (α) indicates that the gene is present but non‐functional.
‡ Loss of 16p13.3 may be the result of deletion, inversion plus deletion, formation of a ring chromosome 16 that lacks the α gene cluster or
unbalanced inheritance of a derivative (16) lacking 16p13.3 from a parent who had a balanced translocation, e.g. t(1;16), t(5;16) or t(16;20) [20].
§ (αα) indicates that both α genes are present but non‐functional.
• deletion of an α1 gene and more than 18 kb c hromosome, designated HKαα (−α3.7 and αααanti−4.2)
downstream but with inactivation of the remain and anti‐HKαα (−α4.2 and αααanti−3.7) [26]; unlike the
ing structurally normal α2 gene by a negative simple heterozygotes for −α3.7 and −α4.2, there is no
positional effect (α−ZF) [8, 18] (α0 thalassaemia); risk of haemoglobin H disease in offspring.
• deletion of adjacent α2 and α1 genes (α0 2 Deletion of the upstream major regulatory ele
thalassaemia); ment including the LCRA with marked downregu
• deletion of both α genes and LCRA (α0 lation or abrogation of expression of both
thalassaemia); structurally normal α genes [27] (it is estimated that
• deletion of the MCS‐R2 enhancer of the α genes. α chain production is less than 1% of normal so this
There are at least 36 different deletions known, is effectively α0 thalassaemia; there are at least 12
some also having deletion of the ζ gene and inter examples) (Table 3.2).
vening sequences; this category includes extensive
deletions or unbalanced translocations resulting in
Non‐deletional
loss of the telomere of chromosome 16, causing a
syndrome of α thalassaemia trait (α0 thalassaemia), 3 Mutations affecting ribonucleic acid (RNA) splic
dysmorphism and mild to moderate mental retar ing (Table 3.3). The mutation IVS1 117 G→A in the
dation referred to as the ATRX syndrome (Table 3.2). α1 gene, for example, is an acceptor splice site
The two most common α+ mutations, −α3.7 and −α4.2), mutation found in India that makes the gene non‐
occasionally coexist with triple α on the same functional; splice site mutations such as this, since
Table 3.3 Classification of non‐deletional α thalassaemia.
Type of mutation Phenotype Number of examples Examples

recognised
RNA splice site mutation in α+ thalassaemia At least 3 (α2 donor HBA2 IVS1 (–5nt) donor splice site
α1 or α2 gene (donor or site, α2 acceptor site, mutation in Mediterranean area and
acceptor site) α1 acceptor site) Middle East
RNA polyadenylation α+–α0 thalassaemia At least 5 (described HBA2 AATAAA→AATAAG (αPA 1α,
signal mutations (i.e. severe α+) or α+ only for α2 gene which also known as αTSaudiα) HBA2
is likely to account for AATAAA→AATA−−(αT Indiaα, India
the severe phenotype) and Thailand)
Impaired RNA translation α+ thalassaemia, α+–α0 At least 5 (two in α2 HBA2 ATG→ACG, GTG or A–G; −α3.7
consequent on initiation or, when the mutation gene, one in α1 gene, ATG→GTG (mutation in association
codon or initiation occurs in association two in single α gene) with deletion gives α0 phenotype)
consensus sequence with deletional α
mutation thalassaemia, α0
thalassaemia
Impaired RNA translation α+ or α0 thalassaemia At least 5 (4 frameshift Codon 30/31 (−4 nt) frameshift and α2
consequent on a frameshift plus 1 nonsense) CD116 GAG→TAG nonsense mutation
or nonsense mutation
Impaired RNA translation α+ thalassaemia At least 5 (all α2 gene) Hb Constant Spring TAA→CAA
consequent on a termination (αCSα), Hb Icaria TAA→AAA (αIcα), Hb
codon mutation leading to Koya Dora TAA→TCA, Hb Seal Rock
an elongated mRNA and α TAA→GAA, Hb Paksé TAA→TAT
globin chain
Production of highly α+ thalassaemia At least 18: 14 point Hb Agrinio (αAgrα), Hb Petah Tikvah
unstable α chain as a result mutations, 4 small (αPT), Hb Quong Sze (αQSα), Hb Suan
of point mutation or a deletions; 11 affecting Dok (αSDα), and Hb Evaston (point
small deletion α2 gene, 4 affecting α1 mutations); Hb Taybe (small deletion);
gene and 3 affecting a Hb Adana (αAdanaα or ααAdana)
single α gene
Lack of a transactivating α+ thalassaemia ATR‐X syndrome

factor encoded by the
ATRX gene
they affect only one of the two α genes, produce an 5 Mutations affecting RNA translation (Table 3.3):
α+ phenotype. • initiation codon or initiation consensus
4 Mutations affecting polyadenylation (Table 3.3). sequence mutations leading to absent or
The mutation αTSaudiα (αPA6 A→Gα), for example, which reduced translation;
is common around the Mediterranean, is one of a • frameshift mutations resulting from small
number of mutations affecting the highly conserved deletions or deletion plus insertion or non
messenger RNA (mRNA) cleavage and polyade sense mutations acting as premature termina
nylation (PA) signal; this mutation leads to a tion codons, leading to inactivation of the gene
marked reduction in α chain synthesis, giving a or synthesis of a very unstable α chain;
severe α+ phenotype which is sometimes designated • mutation of a termination codon to a coding
α+‐α0; other mutations affecting polyadenylation sequence leading to an elongated α chain that
may be less severe. is synthesised at a reduced rate, possibly
90 Chapter 3
because of instability of the mRNA; examples It is likely that the two commonest deletions, −α3.7
of mutations of the termination codon leading and −α4.2, are both the result of unequal crossover
to an elongated α chain include haemoglobin during meiosis, with the result that when this muta
Constant Spring (found in southern China, tion first arose one chromosome was left with a sin
Thailand, Cambodia, Vietnam, Laos and gle α gene while the other had a triplicated α gene.
around the Mediterranean, e.g. Greece and In the case of −α3.7 the single α gene and the central
Sicily), haemoglobin Koya Dora (found with a gene of the three α genes is a fusion gene (see
10% prevalence in the Koya Dora tribe in Fig. 3.1). Either abnormal chromosome could have
Andhra Pradesh, India), haemoglobin Icaria, passed into the gamete and thus the fetus and thus
haemoglobin Seal Rock and haemoglobin both αααanti3.7 and αααanti4.2 are known to exist. It
Paksé (found in Laos, Cambodia and appears that both these mutational events occurred
Thailand); each of these α chains is elongated a number of times in a variety of ethnic groups.
by 31 amino acids as translation continues into Further investigation of −α3.7 has established that
the 3′ untranslated region (UTR) until a down there are, in fact, three slightly different deletions,
stream termination codon is encountered occurring in different ethnic groups, which are now
within the polyadenylation signal sequence. designated −α3.7I, −α3.7II and −α3.7III. Only −α3.7I is
6 Mutations causing marked post‐translational common in many ethnic groups. −α3.7II has been
instability of a highly abnormal α chain, usually as described in India and Nepal whereas −α3.7III is con
the result of a defect in the haem pocket, in α1β1 con fined to Oceania [32].
tacts or affecting binding to α‐haemoglobin stabi A variant haemoglobin may be predicted from
lising protein (which is necessary for folding and the DNA sequence in patients with thalassaemia
solubility of free α chains) [11, 19, 28]; examples but may be undetectable (e.g. haemoglobin Quong
include haemoglobin Quong Sze (α125 Leu→Proα or Sze), or present in very low amounts (e.g. haemo
αQSα), which is found in Kurdish Jews and in globin Suan Dok), because of very marked instabil
South‐East Asia, haemoglobin Agrinio (α29 Leu→Proα ity of the α chain, the αβ dimer or the haemoglobin
or αAgrα), found in the Mediterranean area (Greece molecule. Sometimes a hyper‐unstable haemoglo
and Cyprus) and in South‐East Asia, and haemoglo bin is detectable only after splenectomy. When the
bin Adana, found in Indonesia and as occasional haematological features are those of α thalassaemia
cases in other parts of the world. rather than of an unstable haemoglobin, classifica
7 Deletion that leads to loss of the α1 gene and tion as non‐deletional α thalassaemia is appropri
truncation of a downstream gene, LUC7L, which is ate. When an α chain variant is only moderately
transcribed in the opposite direction, leading to unstable it will constitute a larger proportion of
transcription of RNA that is antisense with regard total haemoglobin and will produce the phenotype
to the α2 gene, in turn leading to methylation and of a Heinz body haemolytic anaemia and classifica
silencing of the remaining α2 gene [29]. tion as an unstable haemoglobin is then appropri
8 Gain‐of‐function mutation between the α genes ate. An unstable haemoglobin can interact with α0
and their locus control region that creates a pro determinants to cause haemoglobin H disease (see
moter‐like region that interferes with transcription Table 3.6).
of both α genes, leading to (αα)T [30]. The types of mutation most commonly found in
9 Transactivating abnormality resulting from different ethnic groups are shown in Table 3.4 and
mutation in the ATRX gene at Xq21.1 (previously the incidence of α0 and α+ thalassaemia in different
known as the XH2 locus), which encodes a deoxyri ethnic groups in Table 3.5 [17, 33–76]. Overall, the
bonucleic acid (DNA) helicase [31], resulting in a highest prevalence of α thalassaemia is found in
syndrome of moderately severe mental retardation, Oceania and the Indian subcontinent. The highest
dysmorphism and α thalassaemia in males, referred prevalence of the less common but more serious α0
to as the ATRX syndrome; more than 180 families thalassaemia is found in southern China and South‐
have been described [19]. East Asia.
Table 3.4 Types of mutation most often responsible for α thalassaemia in different ethnic groups.
Ethnic group Type of Designation Nature of mutation

thalassaemia
South‐East Asia and α0 −−SEA Deletion of both α genes

southern China −−FIL Deletion of both α genes
−−THAI Deletion of both α genes
α+ −α4.2 Deletion of α2 gene

−α3.7 Deletion of part of both α genes with
formation of an α2α1 fusion gene
αCSα Haemoglobin Constant Spring, reduced
rate of synthesis of a haemoglobin with
an elongated α chain
αNcoIα Mutation in initiation codon of α2 gene
ααNcoI Mutation in initiation codon of α1 gene
αSuan Dokα Very unstable α chain
αQuong Szeα Very unstable α chain
Mediterranean α0 −−MED I, −−MED II Deletion of both α genes

(particularly Greece −(α)20.5 Deletion of all of one α gene and part of
and Cyprus) the other
α+ −α3.7 As above
αTSaudiα Polyadenylation signal sequence
mutation
αHphα Small frameshift mutation of IVS1
donor site
Middle East α0 −−MED Deletion of both α genes
α+
αTSaudi
α Polyadenylation signal sequence
mutation
India α+ −α3.7 As above

−α4.2 (less common As above
than −α3.7)
ααIVS1 nt117 G→A Acceptor splice site mutation

αKoya Doraα Mutation of termination codon resulting
in an extended unstable α chain
Sri Lanka α+ −α3.7 As above

−α4.2 As above
Nepal α+ −α3.7 As above
African, Afro‐American α+ −α3.7 As above

and Afro‐Caribbean
Melanesia α+ −α3.7 As above

−α4.2 (less frequent As above
than −α3.7 except in
Papua New Guinea)
Polynesia α+ −α3.7 As above; −α3.7III is most characteristic of

Polynesia and almost confined to this
area
92 Chapter 3
Table 3.5 The prevalence of α0, α+ and β thalassaemia heterozygosity in different countries and ethnic groups (derived
from multiple sources including references 17, 33–76).
Country or ethnic α0 thalassaemia α+ thalassaemia β thalassaemia

group
Greece About 1.5% 7–10% 6–28% (overall 8%)
Cyprus Uncommon, averaging 26% (including 1% α α) T

14–18% (incidence is similar in
around 2% (both Greek Greek and Turkish Cypriots,
and Turkish Cypriots) overall 15%)
Turkey Uncommon (about 6% 1–37% (high incidence confined

0.6%) to Eti‐Turks on south‐eastern
coast, overall 2–3%)
Italy Rare (0.5% in Sardinia, 10% in Sicily, 28% in 1–30%, overall 4%) (highest
0.2% in Sicily, even Sardinia prevalence in Po delta, Sardinia
lower in southern Italy) (10%), southern Italy and Sicily
Spain Very rare (0.2%) 2% 1–8% (highest in Minorca,

southern Spain and Galicia);
5% in Spanish gypsies
Portugal Rare 10% 0. 5–7.5%
France – Corsica 3%
Malta 3–4% [61]
Eastern Europe 0.05% in Republic of 1.5% in Republic of Overall 2–20% (rare in former
(Romania, Bulgaria, Macedonia (former Macedonia (former Czechoslovakia, Hungary,
former Yugoslavia, Yugoslavia) [62] Yugoslavia) [62] northern former USSR; higher in
former Soviet Union) Romania, 3% in Uzbekistan, 5%
in Tajikistan, 5.5% in Azerbaijan);
1–10% (average 2.6%) in Republic
of Macedonia (former
Yugoslavia) [62]; 0.5–19.9%
(average 2.5%) in Bulgaria [63],
8.4% in one study in Albania [64]
British (white) Rare (occurs particularly <1% <0.5%, probably about 0.1%
in Lancashire and
Cheshire; 0.05% in
Wigan, UK)
Middle East (Iran, Rare (occurs rarely in <1–20% (overall 9%); 47% in Overall 2–20%: Iran 1–7%
Iraq, Syria, Lebanon, Israel, United Arab Saudi Arabia, highest in (overall 3%), Iraq 2–7% (4% in
Jordan, Bahrain, Emirates, Iran; −−YEM/ is eastern province; 18–50% in north‐east Iraq [66], 7% in Erbil
United Arab occasionally found in United Arab Emirates, 64% province [67]), Syria 1%, Lebanon
Emirates, Saudi Yemenite Jews and in Kuwait, 89% in Oman, 2–6%, Jordan <1–4%, Bahrain
Arabia, Israel, −−MED/ in Middle relatively frequent in Israeli 2–3%, Oman 1–2%, United Arab
Palestine, Yemen) Eastern Jews; occasional Arabs and Yemenite Jews, Emirates 8% [68], Central Saudi
−−MED/ in Arabs; 6% of α 9% in Ashkenazi Jews Arabia 3–4%, eastern Saudi
thalassaemia alleles in Arabia 13–18%, overall Saudi
Iran were −−MED [65]) Arabia 1–1.8% [69], Yemen 1–2%,
Palestine (Gaza strip) 4%, Israeli
Arabs 3–25%, Yemenite Jews 9%,
Kurdish Jews 20%, Ashkenazi
Israeli Jews <0.1%, other Middle
Eastern Jews 2–4%

group
North Africa and 0.6% α0 in Algeria (−−MED 5–8% overall; 4% in Algeria Libya <1–11%, Algeria <1–15%
Horn of Africa and −(α)20.5 [70]; not [70]; 2–4% in Tunisia [71]; (overall 2%), Morocco <1–7%
(Morocco, Tunisia, detected in Tunisia [71]; 10% in Egypt [72] (overall 3%), Tunisia 3.5%, Egypt
Algeria, Libya, Egypt, 5% −−/αα or <1–9%, Sudan 1–10% (overall
Sudan, Ethiopia) haemoglobin H disease 4%), Ethiopia <1–8%
in Egypt [72]
West Africa Nil Gambia 8–15%, Togo 46%, Overall 1–14%, Senegal <1–5%,
Nigeria 8–58%, Senegal Liberia <1–9%, Ivory Coast
22%, Benin and Burkina 1–12%, Mali <1%, Burkina Faso
Faso 29%, Ivory Coast 39% 2–12%, Ghana 1–11% (overall
1–2%), Togo <1–2%, Nigeria
1–4% (overall 0.8%), Cameroon
<1–2%,
East Africa Nil Kenya 19–34%, Tanzania 2% Rare in Kenya, Uganda and
Tanzania, 2% in Mauritius and
Reunion Island
Central Africa Nil Central African Republic Central African Republic <1%,
39% (23% of Pygmy Republic of the Congo <1%
population), Republic of the (Pygmy population of 6.5%)
Congo 36–40% (29% of
Pygmy population)
Southern Africa Nil in South African Zambia 20–27%, Malawi Comoros 3%
black populations 39%, Namibia 11.5%, South
African Cape Coloured
population 7%, South
African black 12 (San) to 36
(Venda)%, Mozambique
5–6%, Madagascar <1–3%,
Comoros 2% or more
Africans in UK Probably 25–30% 0.9%
Afro‐Americans 25% 1–2%
Afro‐Caribbeans Rare (but recognised in Overall 25%, Jamaica 34% 0.5–10% (overall 1%), Jamaica 1%
West Indians with [73] and 4–5% reported, Lesser
Chinese ancestry) Antilles <1–10%, Guadeloupe
0.5%,
0.9% in Afro‐Caribbeans in UK
Mexico 0.4% [74]; up to 15% reported
Afghanistan 3%
Pakistan 15–20% 5%, about 4.5% in UK Pakistanis
Nepal 6–14% (but up to 97% in 13%

some tribal populations)
India Rare 5–33% (17–99%, mainly Overall 1–16% (average around

above 50%, in tribal 3%, but up to 40% in some tribal
populations) populations), about 3.5% in UK
Indians
(Continued on pp. 94–95.)
94 Chapter 3

group
Bangladesh 3%
Indian subcontinent Overall, about 4.5% (from 3% in

populations in Britain UK Bangladeshis and Punjabi
Sikhs to about 6% in East African
Asians)
Sri Lanka 15–16% (about 13% −α3.7 1–5% (overall 2.2%); 0–16.4% in
and about 2% −α4.2); 3–20% different districts [75]
−α3.7 and 0–2.8% −α4.2 in
different districts [75]
Maldives 16% [74]
Japan <1% Rare
Korea Very low
Hong Kong 4.5% 0.5–1% 3%
Singapore 3–4% 8%
China 3–9% in southern China <1–6% in southern China 0.5% in north and north‐west,
2–6% in south, overall c 1.7%
Chinese in UK About 3%
Taiwan 3.5% (higher and lower 2% 1–3%

in different aboriginal
populations)
Myanmar (Burma) 0.5–6% (overall 3%); 8% in one

study [76]
Thailand 4% in central Thailand, 3–17%, 3% around 4–11% (overall 3%)

14% in northern Bangkok, 9% in north, 17%
Thailand in north‐east; 1–8%
Constant Spring
Cambodia 1–4% 12–28%, Constant Spring up 1–5% (overall c 2.8%)
to 2%
Laos 4% About 14%, Constant 1–9% (overall 5%)

Spring 9%
Vietnam 2–3% (southern About 8%, ~17–22% and 1–25% (overall 4%)
Vietnam) Constant Spring 0–4% in
southern Vietnam
Malaysia 3–9% Constant Spring <1–6%, 1–5% (overall 2%)

overall up to 29%
Philippines 10% 5% 1–2%
Brunei 2%
Indonesia 2.6–3.2% 6% (2–30% on different 0–11% (overall 3%)

islands)

group
Papua New Guinea 10% in highlands, 62% in 1–25%

lowlands
Solomon Islands 45%
Vanuatu (previously 45%

New Hebrides)
New Caledonia 6%
New Zealand (Maori) 5–10%
Australia 6% Rare
(Aboriginal)
Indigenous Rare
Americans
Brazil Overall 1%
α+ thalassaemia heterozygosity, compound

or α1 gene, designated αTα or ααT. However many
heterozygosity and homozygosity
cases, particularly those affecting the α1 gene, are
α+ thalassaemia trait is the commonest monogenic likely to be unrecognised so that the true frequency
disorder in the world. It usually results from dele is unknown. Non‐deletional α thalassaemia affect
tion of all or part of the α2 globin gene. The com ing the α2 gene leads to a more marked reduction of
monest mutations causing α+ thalassaemia are −α4.2 α globin production than α2 deletion, as there is no
and −α3.7. −α4.2 is a 4.2 kb deletion including the α2 upregulation of the α1 gene. Consequently the hae
gene. −α3.7 designates a group of three slightly dif matological abnormality is more marked. αTα is
ferent 3.7 kb deletions of the 3′ end of the α2 gene commonest in the Middle East, particularly in Saudi
and the 5′ end of the α1 gene with formation of an Arabia (αTSaudiα) but also in Cyprus and Sardinia.
α2α1 fusion gene. Both −α4.2 and −α3.7 result in about Non‐deletional α thalassaemia affecting the α1
a 50% reduction in α chain synthesis from the gene, ααT, is milder than αTα, usually being interme
affected chromosome. Although the α2 gene is usu diate in severity between −α/αα and −−/αα.
ally responsible for about 70% of α chain produc Certain α chain variants give rise to the pheno
tion there is some upregulation of the α1 gene when type of α+ thalassaemia (see Table 3.3). The com
the α2 gene is deleted (−α4.2), whereas the α2α1 monest non‐deletional α thalassaemia is
fusion gene is downregulated in comparison with a haemoglobin Constant Spring with a frequency of
normal α2 gene. About a quarter of individuals 1–8% in Thailand. This results from mutation of the
with African ancestry are heterozygous for α+ tha STOP codon of the α2 gene so that a further 31
lassaemia (having the genotype −α3.7/αα), while amino acids are added to the α chain; the mRNA is
1–2% are homozygotes (having the genotype very unstable and the rate of α chain synthesis
−α3.7/−α3.7). Other ethnic groups in which deletional reduced to about 1% of normal [77]. Haemoglobin
α+ thalassaemia occurs include Greeks, Cypriots, Paksé has a similar significance and cannot be dis
Turks, Sardinians, Lebanese, Saudi Arabs, Indians, tinguished from haemoglobin Constant Spring on
Thais, Filipinos, Indonesians, Melanesians and high performance liquid chromatography (HPLC),
Polynesians. Less often α+ thalassaemia is found to cellulose acetate electrophoresis or capillary
result from a mutation rather than deletion of the α2 electrophoresis.
96 Chapter 3
α+ thalassaemia is of little significance to the indi Laboratory features

vidual since it is clinically silent. Heterozygosity, and
Heterozygotes for α+ thalassaemia (−α/αα) may
to a greater extent homozygosity, for α+ thalassaemia
have a completely normal blood count and film or
offer partial protection against severe malaria
trivial anaemia and microcytosis with slight reduc
although the incidence of mild malaria in young chil
tion of the mean cell volume (MCV) and mean cell
dren is actually increased [78]. There also appears to
haemoglobin (MCH). On average, the haemoglo
be protection against other infections severe enough
bin concentration (Hb) is probably about 10 g/l
to cause hospitalisation in children [78]. The condition
lower than in subjects with four α genes. All hae
is of some genetic significance since people who are
matological variables – red cell count (RBC), Hb,
compound heterozygous for deletional α+ thalassae
haematocrit (Hct), MCV and MCH – show consid
mia and α0 thalassaemia have haemoglobin H disease
erable overlap with normal values but mean levels
(see later). Homozygotes for more severe non‐dele
differ. Homozygotes (−α/−α) have more marked
tional α+ thalassaemia (αTα/αTα) can also have the
haematological abnormalities (Figs 3.3 and 3.4),
clinical features of haemoglobin H disease.
which are usually comparable with those seen in β
Homozygosity for termination codon mutations, hae
thalassaemia heterozygotes. In a comparison of
moglobin Constant Spring, haemoglobin Koya Dora
healthy adult Afro‐Americans, those with −α/−α,
and haemoglobin Icaria leads to more anaemia than is
−α/αα or αα/αα were found to have mean MCVs
usual in homozygosity for α+ thalassaemia and a dif
of 89.6, 85 and 75.7 fl (men) and 90.3, 84.2 and
ferent haematological picture. There is usually a mild
72.3 fl (women), respectively [80]. In a large survey
haemolytic anaemia and some cases have mild jaun
of Kenyan children, the presence of sickle cell trait
dice and hepatosplenomegaly. The haemolysis is the
was found to ameliorate the effect of α+ thalassae
result of membrane damage by oxidised abnormal α
mia on the RBC, Hb, MCV and MCH [81].
chains [28]. Deletion of LCRA also leads to a severe
Haemoglobin electrophoresis or HPLC is normal
phenotype, equivalent to α0.
with the haemoglobin A2 percentage being normal
There may be an increased prevalence of neural
or reduced.
tube defects in the fetuses of women who have
During pregnancy, haematological changes in
either homozygosity for α+ thalassaemia, heterozy
women with α thalassaemia trait mirror those in
gosity for α0 thalassaemia or β thalassaemia hete
normal women. On average, the Hb falls about
rozygosity, and administration of folic acid may be
14 g/l, the MCV rises by about 5% and the MCH
particularly important in these women [79].
rises by about 7% [82].
Fig. 3.3 Blood film of an adult male

α+ thalassaemia homozygote
(genotype −α3.7/−α3.7); red cell indices
were red cell count (RBC)
5.77 × 1012/l, haemoglobin
concentration (Hb) 139 g/l,
haematocrit (Hct) 0.44, mean cell
volume (MCV) 76 fl, mean cell
haemoglobin (MCH) 24.1 pg, mean
cell haemoglobin concentration
(MCHC) 318 g/l. May–Grünwald–
Giemsa (MGG) ×100 objective.
(a)
(b)
Fig. 3.4 Red cell cytograms and histograms on a Technicon H2 instrument of (a) a male α+ thalassaemia trait homozygote
and (b) a haematologically normal control subject; α thalassaemia is associated with microcytosis that is more marked
than the associated hypochromia.
98 Chapter 3
Fig. 3.5 Blood film of a

haemoglobin Constant Spring
heterozygote showing basophilic
stippling. MGG ×100.
Heterozygotes for haemoglobin Constant Spring only mild anaemia from several months after birth
have more marked anaemia than is usual in α+ tha [86]. Compound heterozygosity for haemoglobin
lassaemia trait but the MCV is not proportionately Constant Spring and haemoglobin Paksé can simi
reduced [77] and may be normal. Basophilic stip larly cause hydrops fetalis [87].
pling is usually prominent (Fig. 3.5) [10, 83] but this In a homozygote for haemoglobin Icaria, there
is not invariable [84]. Homozygotes for this variant was no increase of the RBC, a normal MCV and a
haemoglobin are usually anaemic (Hb around reduced MCH [88].
100 g/l) with a reduction of the MCH (on average Increased red cell protoporphyrin, which has
around 26 pg) [83]. However the mean MCV is nor been used as a screening test for iron deficiency, has
mal or low‐normal (in one study averaging 88 fl) been found in 20% of cases of α thalassaemia trait
[83] (i.e. considerably less reduced than would be [89, 90]. Elevation tends to be less than in iron defi
expected given the degree of reduction of the ciency but there is some overlap so that this test is
MCH); this is because damage to red cell mem not reliable in distinguishing between these two
branes by oxidised α and αCS globin chains leads to conditions. The −α/αα and −α/−α genotypes are
cellular overhydration. The mean cell haemoglobin associated with increased soluble transferrin recep
concentration (MCHC) tends to be low (mean tor [91], indicating increased erythropoiesis.
300 g/l) [83]. Damage to the red cell membrane At birth, neonates with α+ thalassaemia trait have
leads to a considerably shortened red cell survival a lower mean Hb than other neonates. In one study
and reticulocytosis (reticulocytes usually around the mean level was 151 g/l in heterozygotes and
6–10%). Splenomegaly is common. There is usually 141 g/l in homozygotes, in comparison with a mean
2–11% haemoglobin Constant Spring and often normal of 154 g/l [92]. The MCV and MCH were
1–3% haemoglobin Bart’s but no haemoglobin H similarly reduced. The mean MCV was 100 and
[10, 83]. In occasional cases haemoglobin Constant 94 fl in comparison with a normal mean of 105 fl.
Spring is lower or even undetectable [85]. The MCH was 33 and 31 pg in comparison with a
Haemoglobin A2 tends to be low [85]. Haemoglobin mean normal of 35 pg. In one study using HPLC,
F is normal. Homozygotes for haemoglobin some but not all babies with heterozygous α+ thalas
Constant Spring generally have mild haemolytic saemia had around 1–2% of haemoglobin Bart’s, in
anaemia and splenomegaly. However fetal anaemia comparison with 0.5–1.0% in neonates with four α
requiring intrauterine transfusion for hydrops feta genes. However, a minority (less than 10%) of
lis has been described, as has neonatal jaundice babies with four α genes have haemoglobin Bart’s
necessitating phototherapy [86, 87] with there being up to 1.2–2.6% [92]. In this study there was a
difference in haemoglobin Bart’s levels between

bilirubin, with which it may be confused, is slightly
heterozygotes and homozygotes for α+ thalassae less than 0.1. If there is any difficulty making the
mia – mean values 1.8 and 6.4%, respectively – but distinction, plasma should be removed from the
overlap between heterozygotes and homozygotes blood sample before repeating the analysis.
occurred [92]; the homozygotes had around 3–10% Elevation is more likely in association with −α4.2
haemoglobin Bart’s. A second HPLC study found than in association with −α3.7. A concentration of
0.8–2.8% haemoglobin Bart’s in heterozygotes in >2% of haemoglobin Bart’s on haemoglobin electro
comparison with 2.5–8% in homozygotes, but again phoresis is suggestive of either −−/αα or −α/−α,
overlap occurred [93]. In a third HPLC study, α+ tha and in an ethnic group in which −−/αα does not
lassaemia heterozygotes had haemoglobin Bart’s of occur provides presumptive evidence of the −α/−α
1.55–5%, whereas homozygotes had 5–12% with no genotype. When the more sensitive technique of
overlap [94]. The percentage of haemoglobin Bart’s HPLC is used, the detection of increased haemoglo
is affected by the nature of the α+ thalassaemia bin Bart’s in a neonate is suggestive of α thalassae
mutation, being higher for −α4.2 than for −α3.7 [95]. mia trait but, because of the detection of
Individuals with α+ thalassaemia homozygosity haemoglobin Bart’s in some normal babies, cannot
have a lower haemoglobin Bart’s percentage than be regarded as providing a definitive diagnosis.
those with α0 thalassaemia heterozygosity – 2.5– Since many individuals with α+ thalassaemia trait
8.0% in comparison with 7.6–12.3% – even though have haematological variables falling within the
both have only two α genes [93]. Haemoglobin normal range the diagnosis is likely to be unsus
Bart’s disappears by 3–6 months of age. pected in many cases unless revealed by population
surveys or by family studies of a patient with hae
moglobin H disease.
Diagnosis
When an α+ thalassaemia phenotype is conse
A diagnosis of α thalassaemia trait is usually sus quent on the presence of an α chain variant synthe
pected when an individual is found to have micro sised at a much reduced rate, the variant
cytosis that is not explained by β or δβ thalassaemia haemoglobin may be detected by haemoglobin
trait or iron deficiency. (However it should be noted electrophoresis or, more often, HPLC, but it com
that there are alternative explanations for ‘thalas prises a very low proportion of total haemoglobin
saemic’ red cell indices with normal percentages of (e.g. haemoglobin Constant Spring is usually 0.5–
haemoglobins A2 and F; other explanations, which 1% of total haemoglobin). Haemoglobin Constant
will be discussed later, include coinheritance of β Spring can be identified on cellulose acetate electro
and δ thalassaemia, normal‐A2‐β thalassaemia, γδβ phoresis at alkaline pH, particularly if a heavy
thalassaemia and polycythaemia vera complicated application is used; it moves between carbonic
by iron deficiency.) Haemoglobin electrophoresis anhydrase and haemoglobin A2, whereas haemo
and HPLC are normal except for a tendency to a globin A2′ is even slower, moving between the
reduction of the A2 percentage. Very occasional cells application point and carbonic anhydrase. On
may contain haemoglobin H inclusions but this is HPLC, haemoglobin Constant Spring appears in
not a reliable diagnostic test even in those who are the C window as one to three very small peaks. In
homozygous for α+ thalassaemia. Definitive diag one study at least one abnormal peak was detected
nosis, if needed, requires DNA analysis. in all eight homozygotes and in 38 of 44 heterozy
In the neonate, a haemoglobin Bart’s concentra gotes [96]. In the same study, haemoglobin Paksé
tion of 1–2% on haemoglobin electrophoresis is sug was not detected by HPLC. In another study hae
gestive of −α/αα but not all neonates with this moglobin Constant Spring appeared as one to three
genotype have elevated haemoglobin Bart’s. On small peaks but sometimes there was no peak
HPLC, haemoglobin Bart’s must be distinguished apparent and homozygotes could not be distin
from a peak due to bilirubin. With the Bio‐Rad guished from heterozygotes [97]. Delay in testing
Variant II instrument, haemoglobin Bart’s has a beyond 3 days increases the chance of no peak
retention time of about 0.1 minute, whereas that of being apparent [97]. On capillary electrophoresis,
100 Chapter 3
haemoglobin Constant Spring and haemoglobin haematological differences between the different
Paksé appear in the C zone. DNA‐based techniques mutations giving rise to α0 thalassaemia. The −−BRIT
can be used for the diagnosis of non‐deletional α mutation is found in Lancashire and also in New
thalassaemia and are necessary when no variant Zealand and Newfoundland (Canada) [99]; the rare
haemoglobin is detected and diagnosis is important deletion described as −−BLACK in Afro‐Americans
(see Table 2.5). appears to be identical [99].
Coinheritance with other abnormalities of globin Laboratory features

chain synthesis
α0 thalassaemia trait leads to very mild anaemia
Coinheritance of α thalassaemia, particularly if
+
with the Hb overlapping the normal range. The
homozygous, can lessen the laboratory abnormali RBC is increased and the MCV and MCH are
ties of β thalassaemia trait and the clinical and labo reduced. The blood film shows microcytosis and a
ratory abnormalities of homozygosity or compound variable degree of hypochromia (Fig. 3.6). The retic
heterozygosity for β thalassaemia. In β thalassaemia ulocyte count is elevated to 2–3% [82]. When there
trait, the MCV and MCH are higher and the haemo is coexisting β thalassaemia trait or triple α in trans
globin A2 percentage is lower when α thalassaemia the MCV and MCH are less abnormal.
is coinherited. Coexisting α thalassaemia trait At birth, haemoglobin electrophoresis shows
reduces the proportion of the variant haemoglobin neonates with α0 thalassaemia trait to have 5–10%
in individuals with sickle cell trait, haemoglobin C haemoglobin Bart’s. Quantification by HPLC gives
trait and haemoglobin E trait. values of around 7–11% [92] (Fig. 3.7); the mean
value is higher than in homozygosity for α+ thalas
saemia but overlap occurs. Haemoglobin concen
α0 thalassaemia trait
tration, MCV and MCH are all, on average, lower
α0 thalassaemia usually results from deletion of than in neonates with four α genes or with hete
both α genes. Rarely it results from deletion of only rozygosity or homozygosity for α+ thalassaemia. In
the α1 globin gene and downstream sequences with one study mean values were 133 g/l, 86 fl and
inactivation of the remaining α2 gene. α0 thalassae 29 pg, respectively [92].
mia is relatively common in Chinese individuals
originating in south‐eastern China (including Hong
Diagnosis
Kong), in Taiwan and in South‐East Asian popula
tions, specifically in Thailand, Vietnam, Laos, α0 thalassaemia should be suspected when red cell
Cambodia, Malaysia, the Philippines, Burma and indices suggestive of thalassaemia trait are found in
Indonesia. It occurs at lower frequency in the a patient of appropriate ethnic origin with normal
Mediterranean area – in Greece, Cyprus, Turkey, percentages of haemoglobins A2 and F. In this con
Israel and certain parts of Italy (e.g. Sardinia). Three text an MCH of less than 25 pg suggests a diagnosis
types of α0 thalassaemia trait are common in South‐ of either α0 thalassaemia trait or homozygosity for
East Asia and are designated −−SEA, −−FIL and −−THAI α+ thalassaemia trait rather than α+ thalassaemia
(see Fig. 3.2). Two of these, −−FIL and −−THAI, are heterozygosity. Definitive diagnosis requires DNA
large deletions that also include the ζ gene. A rare analysis. Because tests for the diagnosis of α thalas
Indian mutation, −−KOL, also has deletion of the ζ saemia are not readily available in many hospitals it
gene [98]. In the Mediterranean area the commonest is usual to exclude iron deficiency, β thalassaemia
deletion is −−MED, which does not include the ζ gene, and δβ thalassaemia trait (see later) and the pres
followed in frequency by −(α)20.5. The −−MED deletion ence of variant haemoglobins before proceeding to
also occurs in the United Arab Emirates, Iran, DNA analysis. Iron deficiency can be excluded by
Yemen, Kuwait and Jordan. In addition, Iran has the measuring the serum ferritin concentration but it
−(α)20.5 deletion and Yemen has −−YEM. In the hete should be noted that serum transferrin receptor
rozygous state there are no important clinical or concentration is elevated in α thalassaemia trait
Fig. 3.6 Blood film of a female α0

thalassaemia heterozygote (genotype
−−SEA/αα); the red cell indices were
RBC 5.71 × 1012/l, Hb 115 g/l, Hct
0.38, MCV 66 fl, MCH 20.1 pg,
MCHC 304 g/l. MGG ×100.
AFSC
Baby
Baby
Baby
AC
AFSC
Fig. 3.7 (a) Haemoglobin (a)

electrophoresis on cellulose acetate at
alkaline pH showing a fast band that
is haemoglobin Bart’s (‘Baby’); AFSC
is a control sample containing
haemoglobins A, F, S and C; (b) high
performance liquid chromatography
(HPLC) chromatogram (Bio‐Rad
Variant II) showing haemoglobin
Bart’s, superimposed on acetylated
haemoglobin F, in a baby of Filipino
ancestry with α0 thalassaemia trait
(−−FIL/αα); HPLC analysis was done
on washed red cells so there is no
bilirubin present. From left to right,
the peaks are haemoglobin Bart’s (tall
peak) plus acetylated haemoglobin F
(complex peaks), haemoglobin F
(dark grey) and haemoglobin A0. (b)
102 Chapter 3
Fig. 3.8 Haemoglobin H preparation

in a patient with α thalassaemia trait
showing a single cell containing
haemoglobin H inclusions. Brilliant
cresyl blue ×40.
[100] as well as in iron deficiency so is not a useful homozygosity for α+ thalassaemia trait. When
test in this context. Zinc protoporphyrin may be quantitated by HPLC, the mean percentage is
elevated; values tend to be lower than in iron defi higher than in α+ thalassaemia heterozygosity (9.2%
ciency although there is some overlap [90]. The in comparison with 6.4%) but again overlap occurs
detection of rare haemoglobin H inclusions in red [92]. Haemoglobin Bart’s disappears by 3–6 months
cells can also be useful in the diagnosis of α0 thalas of age. α0 thalassaemia caused by the −−SEA muta
saemia trait (Fig. 3.8). However this test is time con tion can be detected by demonstrating an increased
suming and both false negative and false positive percentage of ζ chains (e.g. by slot blot immunob
results can occur. The number of cells with haemo inding or enzyme‐linked immunosorbent assay
globin H inclusions is reduced by concomitant β (ELISA)). The latter may be practicable for screen
thalassaemia trait or haemoglobin E trait. It is there ing purposes in areas where the majority of α0 tha
fore reasonable, in low prevalence areas, not to test lassaemia is caused by this mutation [104]. Reagents
for haemoglobin H inclusions but to proceed are commercially available [105]. This technique is
straight to DNA analysis if accurate diagnosis is only of use in countries with a significant incidence
important (e.g. in a pregnant woman and her part of this genotype and DNA analysis will still be nec
ner with suspected α0 thalassaemia trait). In high essary if both partners appear likely to have α0 tha
prevalence areas, particularly if resources are lim lassaemia trait, on the basis of red cell indices, but
ited, the test has been found useful. However an only one has detectable ζ chain.
immunochromatographic strip for the detection of
haemoglobin Bart’s (i+MED Laboratories) can be
Coinheritance with other abnormalities of globin
used for screening [101–103] and has been found to
chain synthesis
be more sensitive and specific than screening for
haemoglobin H inclusions. Some false negatives Inheritance of α0 thalassaemia trait has a similar
have been reported [101, 102] but in one study all effect to homozygosity for α+ thalassaemia on the
19 cases of heterozygous α0 thalassaemia were phenotype of coinherited β thalassaemia, sickle cell
detected [103]. trait, haemoglobin C trait and haemoglobin E trait
In the neonatal period, a significantly elevated (see earlier). In ethnic groups in which haemoglo
percentage of haemoglobin Bart’s (5–10%) on hae bin E and α0 thalassaemia are both common, it is
moglobin electrophoreses or HPLC is compatible necessary to consider the possibility of their coexist
with a diagnosis of α0 thalassaemia trait but is ence, particularly when screening pregnant women.
not diagnostic since similar levels are seen in If the haemoglobin E plus A2 (as measured by
HPLC) is less than 21.5%, testing for α0 thalassaemia a haemoglobin H band is not detected on electro
is indicated [106]. Coinheritance with α+ thalassae phoresis [112].
mia or with a hyper‐unstable α chain variant usu Haemoglobin H disease is characterised by a
ally leads to typical haemoglobin H disease (see greatly reduced rate of synthesis of α chain, leading
later) but sometimes to a more severe, transfusion‐ to a hypochromic microcytic anaemia. The excess of
dependent condition [107]. β chain production over α chain production leads to
In relation to antenatal screening of pregnant formation of an abnormal haemoglobin with β
women, it should also be noted that the presence of chain tetramers, referred to as haemoglobin H.
β thalassaemia trait does not exclude the simultane Haemoglobin H functions poorly from the point of
ous presence of α0 thalassaemia trait. Failure to view of oxygen delivery to tissues; it lacks haem–
detect both abnormalities may lead to a failure to haem interaction, so that the oxygen dissociation
predict haemoglobin Bart’s hydrops fetalis when curve is hyperbolic rather than sigmoid, and it has a
one partner has α0 thalassaemia trait and the other very high oxygen affinity. Haemoglobin H is solu
has both α0 and β thalassaemia trait. In ethnic ble but it is prone to oxidation and then becomes
groups in which α0 thalassaemia occurs, DNA anal unstable, precipitating to some extent in erythro
ysis of both partners is indicated if one partner has blasts with resultant intramedullary cell death
β thalassaemia trait and the other has probable α0 (ineffective erythropoiesis) [110]. Oxidised haemo
thalassaemia trait [108]. The increase in the MCV globin H precipitates in circulating erythrocytes,
and MCH is α0 thalassaemia trait and triple α are becoming attached to and damaging the red cell
coinherited (−−/ααα) may mean that occasional membrane with resultant membrane rigidity, red
cases of α0 heterozygosity are missed [109]. cell fragmentation and chronic haemolytic anaemia
[77, 110]. Erythrophagocytosis by splenic mac
rophages is attributable in part to membrane
Haemoglobin H disease
rigidity and in part to increased binding of
Haemoglobin H disease is a clinical syndrome immunoglobulin to damaged membranes so that
resulting from a variety of genetic abnormalities the cells are then recognised by the Fc receptors of
(Table 3.6) [15, 21–25, 30, 110, 111]. The commonest macrophages. Free and precipitated β chains cause
cause is compound heterozygosity for α+ thalassae less damage to erythroblasts than the excess α
mia and α0 thalassaemia (e.g. −−SEA/−α4.2 in South‐ chains in β thalassaemia major, so although ineffec
East Asia and −(α)20.5/−α3.7 or −−MED/−α3.7 in the tive erythropoiesis occurs it is less. Haemolysis is
Mediterranean). Alternative mechanisms are (i) thus more important than ineffective erythropoiesis
compound heterozygosity for α0 thalassaemia and a as a cause of anaemia. In one study of eight patients
non‐deletional α thalassaemia such as haemoglobin mean red cell survival was reduced to about a quar
Constant Spring, haemoglobin Paksé or the Saudi ter of normal (22.5 days in comparison with a mean
type of non‐deletional α thalassaemia (e.g. of 97 days in five normal controls) [113].
−−SEA/αCSα, −−SEA/αPakséα or −−MED/αTSaudiα); (ii) com In haemoglobin H disease due to deletion of three
pound heterozygosity for α0 thalassaemia and a genes, the disease tends to be more severe when
highly unstable α chain such as −−SEA/αQSα; and HBA2 is deleted than when HBA1 is deleted. In gen
(iii) homozygosity or compound heterozygosity eral, disease is more severe when there is a non‐
for non‐deletional α thalassaemia, mainly deletional α thalassaemia or a highly unstable α
αTSaudiα/αTSaudiα or αTSaudiα/αAgrα. Rare non‐deletional chain (e.g. −−/αTSaudiα or −−/αQSα, respectively, than
cases are caused by mutation of a gene on the X when there is simple gene deletion, −−/−α. In one
chromosome leading to the ATRX syndrome. study of more of 338 patients the mean Hb was
Whether −−SEA/αWSα, when haemoglobin Westmead 75 g/l in comparison with 90 g/l [112]. In another
is coinherited with α0 thalassaemia, should be clas study of 37 patients, non‐deletional haemoglobin H
sified as haemoglobin H disease is uncertain since disease (mainly −−SEA/αCSα) and −−SEA/αQSα had a
the condition is milder; only half the patients are similar Hb to deletional disease (mainly −−MED/−α3.7
anaemic, some have a normal MCV and MCH and or −−MED/−α3.7) but the effective haemoglobin
104 Chapter 3
Table 3.6 Examples of genetic abnormalities leading to haemoglobin H disease [15, 21–25, 30, 110, 111].
−−/−α Compound heterozygosity for α0 and α+ thalassaemia, e.g. −−SEA/−α4.2, −−SEA/−α3.7 in

China and South‐East Asia, −(α)20.5/−α3.7, −−MED/−α3.7 in the Mediterranean area and
−−SA/−α3.7 or −−SA/−α4.2 in India
−−/αTα Compound heterozygosity for α0 thalassaemia and non‐deletional α thalassaemia, e.g.

−−MED/α−5ntα and −−MED/αTSaudiα in the Mediterranean area
αTα/αTα Homozygosity for non‐deletional α thalassaemia, e.g. αTSaudiα/αTSaudiα or αNCOα/αNCOα in

the Mediterranean area
αTαTα/αTα Rare example of haemoglobin H disease with a total of 5 α genes, 3 of which carry the
αTSaudi mutation
αTSaudiα/αTSaudiα, αPA 2α/αPA Homozygosity for non‐deletional α thalassaemia affecting the dominant α2 gene or
2
α and αTSaudiα/–α3.7 compound heterozygosity for non‐deletional α thalassaemia and deletional α
thalassaemia trait in the Middle East
−−/αCSα, αCSα/αCSα or Compound heterozygosity for α0 thalassaemia and either haemoglobin Constant Spring
−−/αPakséα or haemoglobin Paksé, e.g. −−SEA/αCSα or −−SEA/αPakséα in South‐East Asia* or
homozygosity for αCSα in South‐East Asia or the Middle East
−−/αQSα Compound heterozygosity for α0 thalassaemia and haemoglobin Quong Sze (non‐
deletional α thalassaemia), e.g. −−SEA/αQSα in South‐East Asia
Genotype not certain Compound heterozygosity for α0 thalassaemia and an unstable haemoglobin,
haemoglobin Petah Tikva†
αTSaudiα/αAgrinioα Compound heterozygosity for non‐deletional α thalassaemia and haemoglobin Agrinio,

a haemoglobin with an unstable or rapidly catabolized α chain, in Greeks and Greek
Cypriots; haemoglobin H percentage is relatively low but disease may be severe [111]
−−/αAgrinioα Compound heterozygosity for α0 thalassaemia and haemoglobin Agrinio giving severe
disease in occasional Cypriots [111]
−α3.7(−AC)/−α3.7(−AC) Homozygosity for ACCATG→−−CATG mutation superimposed on −α3.7 in North

Africa and Mediterranean area
−−/−αQ Compound heterozygosity for α0 thalassaemia and αQ on a chromosome with a −α4.2

deletion, haemoglobin H/Q disease
−−/αGPhiladelphia− Compound heterozgosity for α0 thalassaemia and αPhiladelphia occurring on a chromosome

on which the other α gene is deleted; only haemoglobins G, G2 and H are present,
haemoglobin H/G‐Philadelphia disease
(αα)/−α Compound heterozygosity for deletion of the upstream regulatory region and α+
thalassaemia
(αα)T/(αα)T Mutation creating a new GATA1 binding site between the locus control region and the
α genes that out‐competes the α gene promoters for the locus control region [30];
Melanesian haemoglobin H disease
(αα)Jx/−α Compound heterozygosity for deletion of enhancer of γ and α genes (LCRA, MCS‐R2)
and α+ thalassaemia
(αα)Jx/−−SEA Compound heterozygosity for deletion of enhancer of γ and α genes (HS −40, MCS‐R2)
and α0 thalassaemia (very severe haemoglobin H disease) [25]
−α3.7/αSoulisα Compound heterozygosity for α+ thalassaemia and haemoglobin Souli (unstable

rapidly degraded α chains) [21]
αDartmouthα/αDartmouthα and Both transfusion dependent [22]

αDartmouthα/−−SEA
Homozygosity for 2 Sometimes transfusion dependent [23] but not usually [24]
nucleotide deletion in
polyadenylation signal
Mutation of the ATRX ATRX syndrome of mild haemoglobin H disease, mental retardation and facial and
gene at Xq13.3 encoding genital dysmorphism
a trans‐acting factor
Acquired somatic Acquired haemoglobin H disease in haematological neoplasms

mutation of the ATRX
gene
* Other termination codon mutations can also interact with α0 thalassaemia to cause haemoglobin H disease, e.g. haemoglobin Icaria [9].
† It is not known whether the mutation is in the α2 or α1 gene; other very unstable α chains or haemoglobins can also interact with α0
thalassaemia to produce haemoglobin H disease, e.g. haemoglobins Suan Dok [15], Evanston [15], Adana [15] and Pak Num Po [110].
concentration was lower because of a higher developing erythroblasts by precipitated unstable α

percentage of haemoglobin H (mean 29% cf. 6.9%); chain.
there was also more splenomegaly [114]. Splenec Haemoglobin H disease may present clinically
tomy is more likely to be carried out in non‐ with splenomegaly and symptoms of anaemia.
deletional cases [115]. In another study, children Reduced height, body weight and body mass index
with haemoglobin H/Constant Spring disease has also been noted [119]. It should be noted that,
(−−SEA/αCSα) had slower growth and, although the because of the high oxygen affinity of haemoglobin
MCV and MCH were higher, the Hb was lower, the H, symptoms will be more severe than would
reticulocyte count and plasma bilirubin were higher otherwise be expected from the haemoglobin con
and there was a much greater risk of severe anae centration. The anaemia is aggravated by infection,
mia requiring blood transfusion [116]. In a series of pregnancy or exposure to oxidant drugs. Sudden
patients in Iran, none of the 31 patients with dele worsening of anaemia can be due to increased
tional haemoglobin H disease had required transfu haemolysis for which the bone marrow cannot
sion whereas of 18 patients with non‐deletional compensate. Haemolytic crises can be caused by
haemoglobin H disease there were six who were infection, inflammation or drugs. Acute intravas
regularly transfused, four who were occasionally cular haemolysis can occur [120]. Rarely, acute
transfused, two who had been transfused once only haemolysis induced by infection causes very severe
and only six who had not been transfused [117]. In anaemia, shock and renal failure. A sudden wors
the case of −−/αCSα there is also a greater degree of ening of anaemia can also be due to superimposed
ineffective erythropoiesis as a result of damage to folate deficiency or aplastic crisis following parvo
developing erythroblasts by precipitated Constant virus B19 infection. Some patients are jaundiced
Spring α chains [118]. The serum ferritin is signifi and the incidence of gallstones and cholecystitis
cantly higher in non‐deletional cases [112, 114]. is increased. In one study, 11 of 72 patients had
However, the ratio of serum ferritin to hepatic iron gallstones [112] and in another 12 of 23 [121]. Some
is lower in haemoglobin H/Constant Spring than in patients have leg ulcers. There is an increased inci
deletional haemoglobin H disease, which needs to dence of osteopenia and osteoporosis. The outcome
be taken into account if serum ferritin is used to of pregnancy may be affected adversely; in one
indicate the need for chelation therapy [118]. study there was an increased incidence of intrau
Even greater severity is seen if there is a highly terine growth restriction, preterm birth and low
unstable α chain produced, as with −−/αQSα [110], birth weight [122].
−−/αSDα or −−/ααPak Num Po, so that there is not only Often the clinical features of haemoglobin H
precipitation of haemoglobin H but also damage to disease are so mild that the diagnosis is made

106 Chapter 3
incidentally. In other patients the anaemia is much genes [132, 133]. Most of the non‐deletional α chain
more severe and intermittent or, rarely, regular mutations described in haemoglobin H disease
transfusions are required (three of 251 patients in have been in the α2 gene but occasional examples
one series) [123]. Transfusion may become necessary are in the α1 gene [132, 133]. In one study the func
only later in life. In severely affected patients, there tional defect in non‐deletional haemoglobin H dis
is growth retardation and expansion of the bone ease was found to be similar whether the mutation
marrow cavity leading to bony deformity affecting was in the α1 or α2 gene; it was suggested that this
the facial bones, similar to that which is seen in β was because transcription of an abnormal gene
thalassaemia major. Clinically significant iron over interfered with transcription of the normal gene
load is not common but can lead to diabetes melli [132]. In another series, both patients with a non‐
tus, cardiomyopathy, endocrine dysfunction, deletional defect of the α1 gene had mild disease
hepatic fibrosis and even cirrhosis [110]. In one whereas those with a non‐deletional defect of the α2
study, four of 338 patients had a ferritin of greater gene were generally severe [133].
than 1000 ng/ml and required iron chelation ther In the case of the ATRX syndrome, the haemoglo
apy [112]. Some patients require splenectomy to bin H disease is relatively mild and the clinical pres
relieve hypersplenism or because of recurrent epi entation is with dysmorphism and mental
sodes of severe anaemia [116]. There is a considera retardation in young boys. There is short stature,
bly increased risk of post‐splenectomy thrombosis microcephaly, facial dysmorphism and abnormali
[3]. The incidence of pulmonary hypertension is ties of the external genitalia. Most patients with this
increased [124]. Very rarely, the phenotype of hae syndrome have been Caucasian but five Japanese
moglobin H disease is even more severe, with cases have also been reported [134–136].
hydrops fetalis leading, in some cases, to death in
utero or soon after birth [125]. Those babies who sur
Laboratory features
vive the perinatal period may have disease of mod
erate severity, but with transfusion independence in There is a moderately severe hypochromic micro
later life, or may remain transfusion dependent. cytic anaemia with the Hb varying from 30 to
This severe phenotype of haemoglobin H disease 110 g/l (usually 70–100 g/l) (Fig. 3.9). The MCV is
usually results from compound heterozygosity for of the order of 50–65 fl and the MCH usually
α0 and non‐deletional α thalassaemia affecting the α2 15–20 pg. The MCHC is reduced, usually to between
gene (e.g. −−MED/αTSaudiα) [125, 126] but can also 250 and 300 g/l. The reduction in the MCHC reflects
result from compound heterozygosity for α0 thalas not only reduced haemoglobin synthesis but also
saemia and a severe α+ thalassaemia phenotype cellular overhydration resulting from membrane
resulting from a very unstable α chain such as hae damage; this is particularly so in patients with hae
moglobin Quong Sze (−−SEA/αQSα) [127] or haemo moglobin Constant Spring [77, 110]. When one of
globin Adana [128] (−−FIL/αAdanaα or −(α)20.5/αAdanaα) the genetic abnormalities leading to haemoglobin H
or from homozygosity for non‐deletional α thalas disease is a non‐deletional α thalassaemia (e.g. αCS,
saemia (e.g. αTIndianα/αTIndianα) [129], αCSα/αCSα [130], αPaksé or αQS), the anaemia, reticulocytosis and
αAgrα/αAgrα [131] or αAdanaα/αAdanaα [60]. These geno hypochromia are more marked, the MCHC is lower
types do not necessarily lead to hydrops fetalis, and the MCV and MCH are significantly higher [77,
although this may be the usual outcome with 110, 114, 137]; the MCV can be near normal as a
αAdanaα/αAdanaα [60]. The phenotypic variation with a result of overhydration of cells [28]. In one study
single genotype is not well understood. the mean MCV was 68.8 fl in comparison with
In summary, haemoglobin H disease is most 58.3 fl in deletional cases [112]. In another compari
severe if there is homozygosity or compound hete son, cases of non‐deletional haemoglobin H disease
rozygosity for non‐deletional α thalassaemia, next had a mean MCV of 76 fl (cf. 64 fl in deletional), a
most severe if there is compound heterozygosity for mean MCH of 22.1 pg (cf. 20.0 pg) and a mean
deletional and non‐deletional thalassaemia and MCHC of 292 fl (cf. 313 fl) [114]. The RBC is
least severe if there is deletion of three of the four α increased. In patients with hypersplenism there
(a)
(b)
Fig. 3.9 Red cell cytograms and histograms on a Technicon H2 instrument in two patients with haemoglobin H disease:
(a) a patient with disease of average severity showing moderately severe anaemia and marked microcytosis and
hypochromia; (b) a patient with disease severe enough to have required splenectomy showing marked microcytosis and
an extreme degree of hypochromia.
108 Chapter 3
may be a reduction in the white blood cell count reticulocyte count are increased. Serum soluble
(WBC) and platelet count. transferrin receptor and erythropoietin concentra
The blood film (Fig. 3.10) shows striking anisocy tion are increased. Following splenectomy,
tosis, poikilocytosis, hypochromia and microcyto preformed haemoglobin H inclusions that are no
sis. Poikilocytes may include target cells, fragments longer pitted by the spleen may be apparent in
and teardrop poikilocytes. Basophilic stippling may erythrocytes [138] (see Fig. 3.17).
be present. Nucleated red blood cells can be present The bone marrow aspirate (Fig. 3.11) shows
but often they are not. The percentage and absolute erythroid hyperplasia and micronormoblastic
(a)
Fig. 3.10 Blood films of four

patients with haemoglobin H
disease showing the range of
abnormality that is observed;
(a) red cell indices were RBC
4.95 × 1012/l, Hb 96 g/l, Hct 0.30,
MCV 60.5 fl, MCH 19.4 pg,
MCHC 321 g/l, reticulocyte
count 237 × 109/l; (b) red cell
indices and haemoglobin H
percentage were RBC
5.34 × 1012/l, Hb 93 g/l, Hct 0.34,
MCV 63 fl, MCH 17.5 pg, MCHC
(b) 277 g/l and haemoglobin H 5%.
(c)
Fig. 3.10 Continued. (c) red cell

indices were RBC 5.65 × 1012/l,
Hb 92 g/l, Hct 0.33, MCV 58.2 fl,
MCH 16.4 pg, MCHC 281 g/l;
(d) red cell indices were RBC
5.34 × 1012/l, Hb 104 g/l, Hct
0.34, MCV 63.6 fl, MCH 19.5 pg,
MCHC 306 g/l. MGG ×100. (d)
Fig. 3.11 Bone marrow aspirate in

haemoglobin H disease; there is
erythroid hyperplasia and some
erythroblasts show defective
haemoglobinisation. MGG ×100.
110 Chapter 3
AFSC

Fig. 3.12 Ultrastructural examination of bone marrow acetate at alkaline pH in a patient with haemoglobin H
erythroblast in a patient with haemoglobin H disease disease showing a fast band to the left of haemoglobin A
showing an electron‐dense stellate inclusion, probably (bands b and c); AFSC is a control sample containing
consisting of precipitated β chain. (By courtesy of the late haemoglobins A, F, S and C.
Professor Sunitha N. Wickramasinghe.)
Fig. 3.14 HPLC chromatogram

(Bio‐Rad Variant) in two patients with
haemoglobin H disease; the double
peak to the left is haemoglobin H while
the major peak on the right is
haemoglobin A.
aturation. Ultrastructural examination shows

m Haemoglobin electrophoresis (Fig. 3.13), HPLC
abnormal cytoplasmic inclusions that are likely to (Fig. 3.14), capillary electrophoresis and isoelectric
represent precipitated β chain (Fig. 3.12). focusing (IEF) show that haemoglobin H comprises
Serum bilirubin concentration and urinary uro 1–40% of total haemoglobin (usually 8–10%). The
bilinogen are increased. The bilirubin is unconju haemoglobin H percentage is usually higher when
gated. Serum lactate dehydrogenase (LDH) may one of the genetic defects is a non‐deletional thalas
also be increased. Serum haptoglobin is reduced. saemia (e.g. a mean of 15% in comparison with a
Assays of glucose‐6‐phosphate dehydrogenase mean of 7% [140], a mean of 29% in comparison
(G6PD) give higher levels in subjects with haemo with 6.9% [114] and a mean of 12.2% in comparison
globin H disease, whether expressed per gram of with 7.1% [137] in three studies) and is lower when
haemoglobin or per number of red cells [139]; this there is coexisting heterozygosity for βS, βC or βE
may be the result of reticulocytosis. Reticulocytosis [110]. Rarely, no haemoglobin H is apparent [112].
is also likely to lead to a lower percentage of Conversely the haemoglobin H can be as high as
glycated haemoglobin (haemoglobin A1c). 50% [114]. On HPLC using the Bio‐Rad Variant II
system, haemoglobin H may be missed if a haemo H inclusions in a large proportion (usually 35–90%)
globin A2/A1c dual programme is used, as a result of red cells (Fig. 3.15). These inclusions, which form
of haemoglobin H appearing early in the haemoglo in vitro during incubation with vital dyes, are shown
bin F window [141]. Laboratories using the Beta on ultrastructural examination to be attached to the
Thal Short programme on this instrument need to red cell membrane (Fig. 3.16). In patients who have
inspect the chromatogram since haemoglobin H is been splenectomised there are, in addition, larger
not integrated. Haemoglobin Bart’s is present in preformed Heinz bodies, also attached to the red
some patients, usually comprising around 5% of cell membrane. Sometimes these are apparent in a
total haemoglobin; at birth it may be 20–40% [110]. routine blood film (Fig. 3.17) [138] as well as in a
It is higher in non‐deletional cases, with a mean of haemoglobin H preparation (Fig. 3.18). Both the
1.61% in comparison with 0.89% [137]. A very low percentage of haemoglobin H and the proportion of
percentage of haemoglobin H can also be present in cells containing haemoglobin H inclusions are low
the neonate. The percentage of haemoglobin Bart’s ered by concomitant iron deficiency; haemoglobin
in the neonate is higher than the percentage of hae H formation may even be completely suppressed
moglobin H later in life, probably because of insta [143]. A similar temporary suppression of haemo
bility of haemoglobin H. In neonatal screening, 10% globin H synthesis has been reported in one
haemoglobin Bart’s has been used as a threshold to patient with anaemia of chronic disease and in
investigate for haemoglobin H disease [142]. The another with alcohol‐induced sideroblastic anae
percentage of haemoglobin A2 is usually reduced mia [144]. When haemoglobin H disease is caused
(e.g. to 1–2%). It is lower in non‐deletional cases by non‐deletional α thalassaemia, haemoglobin H
(0.67% in comparison with 0.97%) [137]. is not always detectable on haemoglobin electro
Haemoglobin F may be increased to 1–3%. When phoresis and haemoglobin H inclusions may be
either haemoglobin Constant Spring or haemoglo infrequent; it has been postulated that this may
bin Paksé is present, the variant haemoglobin is a be because β chains and an abnormal α chain are
very low percentage and haemoglobin Paksé is not forming transitory and unstable dimers that are
detected in all patients in whom it is predicted from rapidly destroyed [15].
molecular data [140]. A haemoglobin H preparation Globin chain synthesis studies show an α:non‐α
in which red cells are exposed to a mildly oxidant ratio of the order of 0.26–0.60 [133]. The oxygen
dye, such as brilliant cresyl blue or new methylene dissociation curve is abnormal. Since haemoglobin
blue, shows characteristic ‘golf‐ball’ haemoglobin H comprises a relatively low percentage of total
Fig. 3.15 Haemoglobin H

haemoglobin H disease showing
both ‘golf‐ball’ haemoglobin H
inclusions and also an increased
reticulocyte count. Brilliant cresyl
blue ×100.
Fig. 3.16 Ultrastructural
examination of erythrocytes in a
patient with haemoglobin H disease
following incubation with brilliant
cresyl blue; haemoglobin H
inclusions are apparent, attached to
the red cell membrane. (By courtesy
of the late Professor Sunitha N.
Wickramasinghe.)
Fig. 3.17 Blood film of a patient with

haemoglobin H disease who has had
a splenectomy showing a
hypochromia, poikilocytosis, and a
number of preformed Heinz bodies
(solid inclusions attached to the red
cell membrane). MGG ×100.
Fig. 3.18 Haemoglobin H

haemoglobin H disease who has
been splenectomised showing
‘golf‐ball’ haemoglobin H inclusions
and preformed Heinz bodies.
Brilliant cresyl blue ×40.
haemoglobin, the P50 may be normal or only slightly Haemoglobin H is present as a low percentage.
reduced; however the lower part of the dissociation Neonatal screening for haemoglobin H disease can
curve is displaced to the left, giving a biphasic be carried out, as is done in the state of California,
curve. by further investigation of babies found to have
Haemoglobin H disease occurring as part of the more than 25% of haemoglobin Bart’s at birth [146].
ATRX syndrome is relatively mild. Not all patients
are anaemic and the MCH and MCV are sometimes
normal. The percentage of haemoglobin H is lower
chain synthesis
than is usual in haemoglobin H disease and in some
cases none is detected on electrophoresis. Coexisting β thalassaemia trait often makes haemo
Haemoglobin H inclusions are detectable in the globin H disease milder than would otherwise be
great majority of cases but in a relatively low pro the case but some instances of transfusion depend
portion of cells (e.g. 0.5–15%). ency have been reported [147]. The Hb and MCHC
In occasional cases of haemoglobin H disease, tend to be higher although the MCV and MCH are
the single remaining α gene encodes a structurally lower [148, 149]. The Hb is higher, for example a
abnormal α chain so that no haemoglobin A is syn mean of 113 g/l in comparison with 97 g/l in one
thesised. There is haemoglobin H, a significant study [149]. Haemolysis is less. The reticulocyte
proportion of an α chain variant (e.g. haemoglobin count and bilirubin concentration may be normal
G‐Philadelphia or haemoglobin Q) and small [147] or the reticulocyte count may be mildly ele
amounts of a haemoglobin A2 variant (e.g. G‐ vated [150]. Serum soluble transferrin receptor is
Philadelphia2 or Q2). Less rare is haemoglobin H elevated, indicating increased (but ineffective)
disease caused by compound heterozygosity for α0 erythropoiesis [150]. Significant iron overload can
thalassaemia and non‐deletional α+ thalassaemia occur [147]. The percentage of haemoglobin F and
consequent on an α chain variant that is synthe the proportion of cells containing haemoglobin H
sised at a greatly reduced rate (e.g. haemoglobin inclusions are reduced in comparison with those of
Constant Spring, haemoglobin Paksé or haemo uncomplicated haemoglobin H disease [145] and
globin Icaria). In this case small amounts of the sometimes no haemoglobin H can be detected [147].
variant (e.g. 1.5–2% of haemoglobin Constant In one study no haemoglobin H was detected on
Spring) will be present in addition to haemoglobin HPLC in six of 14 cases [148] and in another none
A and haemoglobin H. Haemoglobin H disease was detected on capillary electrophoresis in 28
with haemoglobin Paksé is very similar to haemo cases [149]. A trace of haemoglobin Bart’s is some
globin H disease with haemoglobin Constant times detected. The haemoglobin A2 percentage is
Spring and in the past has been misdiagnosed sometimes normal and sometimes elevated, but to a
[140]; both have a minor slowly migrating band on lesser extent than is usual in β thalassaemia trait.
electrophoresis and inconspicuous peaks on The α:non‐α chain synthesis ratios are of the order
HPLC. A multiplex allele‐specific PCR permits the of 0.5–0.7, similar to those seen in α thalassaemia
distinction to be made. Genes encoding very heterozygosity [150].
unstable α chains can also lead to haemoglobin H Coinheritance of the genotype of haemoglobin H
disease, either when there is homozygosity or disease with compound heterozygosity for β+ and β0
when they are coinherited with other α thalassae thalassaemia leads to balanced chain synthesis so
mia variants [11]. Examples include haemoglobin that although there is a moderate anaemia and
Quong Sze and haemoglobin Suan Dok. marked microcytosis the patient has a clinically
Neonates with haemoglobin H disease have a mild to moderate condition [151]. Haemoglobin H
lower Hb than other neonates, usually of the order is not detected but there may be traces of haemoglo
of 120–140 g/l, and a markedly lower MCV and bin Bart’s; haemoglobin A2 in one patient was 15%
MCH (e.g. 73 and 74 fl and 22 pg in two reported [151]. Haemoglobin Bart’s in the absence of haemo
cases [145]). Haemoglobin Bart’s comprises globin H has also been reported when there is coin
5–40% (usually 20–25%) of total haemoglobin. heritance with β0 thalassaemia heterozygosity [152].
114 Chapter 3
Atypical phenotypes occur when the genotype of undescended testes), terminal transverse limb
haemoglobin H disease is coinherited with haemo defects, pulmonary hypoplasia and retarded brain
globin S trait (see p. 198), haemoglobin C trait (see growth [9, 154–156]. In one study developmental
p. 265), heterozygosity for deletional hereditary abnormalities were present in 64% of 58 infants
persistence of fetal haemoglobin (see p. 160), non‐ with data available [154]. There is often growth
deletional hereditary persistence of fetal haemoglo retardation.
bin (see p. 161) and haemoglobin E heterozygosity Haemoglobin Bart’s hydrops fetalis is usually
or homozygosity (see p. 290). Coinheritance of het incompatible with extrauterine life. Some fetuses
erozygosity for βNew York leads to more severe disease die in utero and others within a short time of birth.
as the βNew York chain has greater affinity than the nor Rarely there is survival for a few days, even in the
mal β chain for the available α chains and haemo absence of treatment. Increasing number of
globin New York is unstable [110]. Coinheritance fetuses have been ‘rescued’ by intrauterine or
with heterozygosity for a β chain variant can lead to post‐delivery transfusion [17, 154, 157] but some
the absence of detectable haemoglobin H (e.g. with times with significant brain damage having
haemoglobin Tak or haemoglobin Hope) or the already occurred, leading to cognitive and motor
presence of both haemoglobin H and haemoglobin impairment [3].
Bart’s (e.g. with haemoglobin J‐Bangkok or haemo Deletion of all four α genes but with one or both
globin Pyrgos) [152]. ζ genes being intact (e.g. −−SEA/−−SEA, −−MED/−−MED,
−−SEA/−−THAI or −−SEA/−−FIL) leads to haemoglobin
Bart’s hydrops fetalis (Fig. 3.19). Exceptionally, the
Haemoglobin Bart’s hydrops fetalis
condition is due to uniparental disomy with two
and related conditions
copies of an α0 mutation being derived from either
Homozygosity for α0 thalassaemia leads to a total the mother or the father [158]. Almost all the hae
failure of α chain synthesis so that no synthesis of moglobin present is haemoglobin Bart’s, a haemo
haemoglobin F, A or A2 can occur. The result is the globin with γ tetramers, which is unable to deliver
clinical syndrome known as haemoglobin Bart’s oxygen to tissues since, like haemoglobin H, it has
hydrops fetalis, first described in Indonesia in 1960 a very high oxygen affinity and lacks haem–haem
[153]; the designation ‘α thalassaemia major’ is interaction, leading to a hyperbolic rather than sig
sometimes used. This condition is most often seen moid oxygen dissociation curve. The remainder of
in South‐East Asia and southern China but there is the haemoglobin is largely haemoglobin Portland
a low but significant incidence in Greece, Turkey 1 (ζ2γ2) which is capable of oxygen delivery to tis
and Cyprus. Rare cases have been observed in sues and can keep the fetus alive into the third tri
Sardinia and in families of Indian and Pakistani eth mester. There may be small amounts of
nic origin. It has been estimated that in the UK, haemoglobin Portland 2 (ζ2β2) and haemoglobin H.
20–30 pregnancies a year carry a risk of haemoglo There is severe anaemia (consequent on a reduced
bin Bart’s hydrops fetalis. rate of haemoglobin synthesis together with inef
Characteristic clinical features include severe fective haemopoiesis and shortened red cell life
anaemia, hepatosplenomegaly, cardiac failure, span caused by precipitation of haemoglobin
serous effusions and gross oedema of the fetus and Bart’s) and extramedullary haemopoiesis leading
placenta, although some fetuses are anaemic but to hepatosplenomegaly. Because haemoglobin
not hydropic. There is hypoalbuminaemia (likely to Bart’s is a high affinity haemoglobin the functional
be related to the marked extramedullary hae effects of the anaemia are much greater than would
mopoiesis in the liver). The severe anaemia and fail be expected from the haemoglobin concentration.
ure of oxygen delivery to tissues lead to abnormal At birth, the neonate is pale, faintly jaundiced,
organogenesis. Developmental abnormalities can growth retarded and usually hydropic. There may
include congenital cardiac abnormalities (atrial sep be pulmonary hypertension and respiratory
tal defect and patent ductus arteriosus), genital distress. There may also be subcutaneous blue

abnormalities (ambiguous genitalia, hypospadias, nodules of haemopoietic tissue.
Fig. 3.19 Fetus with haemoglobin

Bart’s hydrops fetalis. (With thanks
to Dr Robyn Rodwell and by
courtesy of the late Professor Harry
Smith.)
When there is homozygosity for a large deletion haemorrhage and retained placenta and a high rate
including the ζ genes as well as both α genes, no of caesarean section. If molecular diagnosis of α0
functional haemoglobin can be produced. The only thalassaemia is precluded by economic constraints,
haemoglobins synthesised are haemoglobin Bart’s consideration should be given to ultrasound exami
and a haemoglobin with tetrameric ε. The fetus dies nation commencing early in pregnancy in poten
early in gestation with consequent miscarriage. tially at‐risk pregnancies in order to detect hydrops
This syndrome can result from homozygosity for fetalis and offer termination of pregnancy before
−−FIL and −−THAI. maternal complications occur. Increased placental
Rarely the phenotype of hydrops fetalis can result thickness and an increased cardiothoracic ratio are
from heterozygosity for α0 thalassaemia and severe the earliest ultrasound signs.
non‐deletional α+ thalassaemia (e.g. −−/αTSaudiα, When a fetus is rescued by intrauterine transfu
−−/αQSα or −−/α59Gly→Aspα). The disease phenotype sion and subsequently maintained on a life‐long
may be more severe if the ζζ locus is deleted as well transfusion regime similar to that used in β thalas
as the αα locus; this can be designated by the nota saemia major (e.g. aiming for a pretransfusion Hb
tion −−−−/ζζ(αα)T. Haemoglobin A and F are pre of >90 g/l) erythropoiesis is inadequately sup
sent in addition to haemoglobin Bart’s, haemoglobin pressed, non‐functional haemoglobin H is present,
Portland 1 and haemoglobin H [159]. Proportions of and iron overload occurs – attributed to increased
the various haemoglobins are similar to those in gastrointestinal absorption as well as transfusion.
haemoglobin H disease but the haemoglobin con More intensive treatment than for patients with β
centration during intrauterine life and at birth is thalassaemia major is therefore needed in order to
lower. The genotype −−/αTα more often results in suppress erythropoiesis and prevent the haemoly
haemoglobin H disease. The genotype sis and poor oxygen delivery that results from a
αAdanaα/αAdanaα, which occurs in Indonesia, is associ haemoglobin H of 24–64% [158]. A pretransfusion
ated with hydrops fetalis [60]. Hb of >100 g/l and a haemoglobin H <15% are
Women carrying a fetus with haemoglobin Bart’s therefore recommended [158, 160].
hydrops fetalis have an increased rate of preg
nancy‐related hypertension, severe anaemia, poly
Laboratory features
hydramnios and oligohydramnios and usually
have a difficult delivery as the result of delivering a A fetus with haemoglobin Bart’s hydrops fetalis has
hydropic fetus and hydropic placenta [3, 154]; there severe anaemia (Hb usually 30–80 g/l, occasionally
is an increased rate of antepartum and postpartum up to 100 g/l or even higher) and striking
116 Chapter 3
(a)
(b)
Fig. 3.20 Blood films in two

cases of haemoglobin Bart’s
hydrops fetalis showing a
striking increase in nucleated
red blood cells: (a) MGG ×40; (b)
MGG ×100 (with thanks to Dr
Mary Frances McMullin); (c)
MGG ×100 (by courtesy of the
(c) late Professor Harry Smith).
Fig. 3.21 Haemoglobin

electrophoresis in haemoglobin Bart’s
hydrops fetalis, third and seventh
patterns from left, showing a major
haemoglobin Bart’s band and a minor
haemoglobin Portland band. (With
thanks to Dr Helen Dodsworth.)

(Bio‐Rad Variant) in (a) a normal
neonate showing post‐translationally
modified haemoglobin F,
haemoglobin F (black) and
haemoglobin A and (b) in a neonate
with haemoglobin Bart’s hydrops
fetalis showing haemoglobin Bart’s
(a) (b)
only (derived from reference 92).
anisocytosis and poikilocytosis (including target increased cardiothoracic ratio and increased rate of
cells and elongated cells). Erythrocytes appear large blood flow in the middle cerebral artery. It is desira
but markedly hypochromic. There is a reticulocyto ble that this condition be predicted early in preg
sis. Circulating nucleated red cells are greatly nancy by identification of all women at risk of α0
increased, which is a condition designated erythro thalassaemia trait, followed, when both partners are
blastosis fetalis (Fig. 3.20). The MCH and MCHC at risk of α0 thalassaemia trait, by DNA analysis.
are greatly reduced but MCV can be normal. When both parents are found to have α0 thalassaemia
Haemoglobin electrophoresis (Fig. 3.21) and trait (or have haemoglobin H disease), DNA analysis
HPLC (Fig. 3.22) show haemoglobin Bart’s (70– of the fetus is required. Fetal tissue for diagnosis can
100%) and sometimes smaller amounts of haemo be obtained by chorionic villous sampling in the first
globin Portland 1 (usually around 10–20%), or second trimester, from amniotic fluid obtained by
haemoglobin Portland 2 and haemoglobin H. amniocentesis in the second trimester and by fetal
Tetramers of δ chain have also been detected. blood sampling in the second trimester. Management
options are termination of pregnancy or intrauterine
transfusion with the knowledge that life‐long trans
Diagnosis
fusion therapy will be required.
Following delivery of a hydropic fetus, the diagnosis
of haemoglobin Bart’s hydrops fetalis can be made
Coinheritance of α thalassaemia and other
from the haematological and electrophoretic/chro
haemoglobinopathies
matographic characteristics. Sometimes the condi
tion is diagnosed in utero as a result of ultrasound As commented on earlier, coinheritance of α0 thalas
examination, which can detect an enlarged placenta, saemia and an unstable α chain variant can cause
118 Chapter 3
haemoglobin H disease. Coinheritance of α thalas varying from mild to severe. Homozygotes thus
saemia and an α chain variant haemoglobin will have disease varying from mild to very severe.
increase the percentage of the variant and if this Compound heterozygotes for β thalassaemia may
has undesirable characteristics (e.g. high affinity or have two different β0 thalassaemia genes, two
methaemoglobinaemia), the effects will be aggra different β+ thalassaemia genes or both a β0 and a
vated. Coinheritance of α thalassaemia and a β chain β+ thalassaemia gene.
variant may decrease the percentage of the variant β thalassaemia usually results from mutation in
(e.g. haemoglobin S, C or E) or increase the percent or near the β gene. A smaller proportion of cases
age (e.g. haemoglobin J‐Baltimore or J‐Iran), depend result from deletion, either of the β gene itself or of
ing on the relative affinities of the variant and controlling sequences 5′ to the gene (i.e. upstream
normal β chains for the reduced numbers of α from the gene). Disorders with the phenotype of β
chains. The effects of β thalassaemia genes are thalassaemia can also result from mutations that
ameliorated by α thalassaemia (see later). lead to structural abnormalities of the β globin
chain; the mechanism may be either reduced rate of
β chain production or production of a very unstable β
β thalassaemias
chain or a very unstable haemoglobin. Haemoglobin
The beta or β thalassaemias are a group of condi E is the most prevalent example of a structurally
tions resulting from a reduced rate of synthesis of abnormal haemoglobin in which the mutated β glo
β globin. More than 280 β thalassaemia mutations bin is synthesised at a reduced rate, in this instance
have been recognised, occurring in a wide range as a result of activation of a cryptic splice site. It can
of ethnic groups (see Table 3.5), including at least be regarded as a thalassaemic haemoglobinopathy.
230 point mutations and at last 18 large deletions. Similarly, haemoglobin Malay, another variant hae
β thalassaemia is common around the Mediter moglobin with activation of a cryptic splice site,
ranean, in the Indian subcontinent and in South‐ represents 15% of thalassaemia alleles in Malaysia
East Asia and relatively common in people of and 16% in southern Thailand and also occurs in
African ancestry. It is recognised but rare in the China and Indonesia [162]; it is likely to be
indigenous British population [161] and in other frequently unidentified since it cannot be distin
Northern European populations. The severity of guished from haemoglobin A by electrophoresis or
the defect is very variable. Since normal individuals chromatography. Haemoglobin North Shore and
have two allelic β globin genes, β thalassaemia haemoglobin Vicksburg are less common variant
can exist in a heterozygous or homozygous state. haemoglobins associated with a β thalassaemia
Since there are a large number of β thalassaemia phenotype [163]. Haemoglobin Dhohar, found in
mutations, compound heterozygosity can also Oman, is associated with a thalassaemic phenotype
occur with the individual having two mutant [164]. It results from the coexistence of two muta
genes but no normal β gene. β thalassaemia muta tions, one causing a structural alteration and the
tions are divided into two broad categories, β0 other, which creates an alternative splice site, being
(beta zero) thalassaemia and β+ (beta plus) thalas responsible for the thalassaemic features [164].
saemia. In β0 thalassaemia there is either an abnor Heterozygotes have 9–21% of the variant haemo
mal gene that is not expressed or, less often, gene globin and typical thalassaemic red cell indices
deletion. If β0 thalassaemia occurs in the homozy [164]. Homozygotes have the clinical features of
gous or compound heterozygous state (β0β0) there thalassaemia major or severe thalassaemia interme
is a total lack of β chain production and a total fail dia [164]. Mutations that can result in β thalassae
ure to produce haemoglobin A. In β+ thalassaemia mia and the phenotype usually associated with
there is reduced but not absent expression of the each are summarised in Table 3.7 [10, 165–174] and
abnormal β gene so that in homozygotes there is the sites of mutations leading to β0 and β+ thalassae
some production of haemoglobin A. There are a mia, respectively, are summarised in Figs 3.23 and
large number of different β+ thalassaemia mutations 3.24. Rarely there are two thalassaemia mutations
with the severity of the defect in chain synthesis on the one chromosome, as in a patient with both
Table 3.7 Types of mutation that can result in the phenotype of β thalassaemia (the numbers of mutations given should
be regarded as approximate since new mutations continue to be described) [10, 165, 172–174].
Type of mutation Consequence
Deletional
Large deletion involving the β gene (at least 15 Absent transcription, β0 thalassaemia
reported, only that occurring in Sind and Punjabi unusually high haemoglobin
populations is common) A2 in heterozygotes
Small deletion 5′ of the β gene (at least 3 reported) Reduced transcription β+ thalassaemia
Non‐deletional
Insertion of a transposable element into IVS2 Reduced transcription β+ thalassaemia

(1 reported)
Mutations in promoter sequences, CACCC, Reduced transcription, Silent β, mild (β++) thalassaemia
CCAAT or TATA box (at least 26 reported) increased transcription of γ or β+ thalassaemia
and δ
Mutation in 5′ untranslated region near CAP site Reduced transcription and Silent or mild (β++) thalassaemia
(at least 8 reported) translation and instability of
mRNA
Mutation of initiation codon (at least 9 reported) Absent transcription β0 thalassaemia (more severe
than most other β0 thalassaemia)
RNA splicing mutations – involving invariant Absence of properly spliced β0 thalassaemia

nucleotides of either donor or acceptor site (at mRNA
least 27 reported)
RNA splicing mutations – involving nucleotides Inefficient splicing of mRNA Silent, mild (β++) thalassaemia or
flanking splice junctions (consensus sequences) β+ thalassaemia or occasionally,
(at least 12 reported) β0 thalassaemia
RNA splicing mutations – activation of cryptic Aberrant mRNA is Mild (β++) thalassaemia, β+
splice site in an intron or an exon with or without produced in addition to thalassaemia. β0 thalassaemia or,
an alteration in the coding sequence (at least 12 normal mRNA; sometimes a occasionally, dominant β
reported) structurally abnormal β thalassaemia; haemoglobin E,
chain is produced, which haemoglobin Malay, Hb
may be highly unstable Knossus
Mutation interfering with polyadenylation and Unstable elongated RNA β+ or mild (β++) thalassaemia
therefore mRNA cleavage (at least 7 nucleotide transcript (plus some
substitutions and at least 3 small deletions normal transcript)
reported)
Other mutations interfering with mRNA Abnormal processing of Silent, mild (β++) or β+
processing (at least 4 reported) messenger RNA thalassaemia
Premature termination codon consequent on With absent translation β0 thalassaemia (exon 1 and 2
alteration of a single nucleotide (at least 16 (exon 1 and 2 mutations) or mutations) or dominantly
reported) or on a frameshift mutation (at least 72 translation of aberrant inherited β thalassaemia (exon 3
reported, including one large insertion in exon 2 mRNA (exon 3 mutations) mutations)
and other deletions and insertions) leading to a very unstable
truncated β chain
New sense mutation, alteration of STOP codon to Hb Zunyi (dominant β

a coding sequence thalassaemia with
reduced Hb A2)
120 Chapter 3
Type of mutation Consequence
Other ‘thalassaemic haemoglobinopathies’ caused With reduced transcription Hb North Shore, Hb Vicksburg
by a single nucleotide change of an abnormal mRNA
With transcription of Hb Indianapolis, Hb Geneva

abnormal RNA coding for a
very unstable β chain
‘Thalassaemic haemoglobinopathies’ associated Synthesis of an elongated Hb Tak, Hb Cranston, Hb

with mutation of the termination codon to a mRNA and β chain Saverne
coding sequence (‘anti‐termination’ mutation)
Haemoglobin Lepore* δβ fusion gene with reduced Hb Lepore Boston, Hb Lepore

rate of synthesis of aberrant Baltimore, Hb Lepore Hollandia
δβ chain
Certain unstable haemoglobins Marked instability of Hb Showa‐Yakushigi, Hb K

haemoglobin leading to Woolwich
rapid post‐translational
degradation
Mutation in the GATA1 gene, a gene at Xp11.23 Reduced transcription Thalassaemia with
encoding an erythroid transcription factor thrombocytopenia and (in one
patient) congenital
erythropoietic porphyria
Mutation in the ERCC2 (previously XPD) gene at Reduced transcription Thalassaemia with
19q13.32, which encodes a component of the trichothiodystrophy
general transcription factor TFIIH
* Haemoglobin Lepore can be regarded as a δβ+ thalassaemia.
β0 Mutations
1 30 31 104 105 146
5’ 3’
Initiation Invariant 3‘ UTR (1)
codon (6) dinucleotides (2)
False splice
Consensus Consensus site (1)
sequences (2) sequences (1)
Consensus
Invariant Invariant sequences (1)
dinucleotides (6) dinucleotides (4)
Invariant
New splice site New splice site dinucleotides (6)
in intron (1) in intron (1)
Fig. 3.23 Sites of some mutations giving rise to β0 thalassaemia, including dominant thalassaemia. The numbers above
the chromosome represent the codons. Introns are shown in blue and the 5′ and 3′ untranslated regions (UTR) in green.
The dotted lines are a diagrammatic indication of the consensus sequences flanking the invariant nucleotides at the
splice sites. In addition to the mutation sites shown, there are now several hundred reported mutations including
nonsense mutations (some dominant), frameshift mutations, unstable variants (some dominant), and insertions or
deletions of triplet codons (some dominant).
β+ Mutations
PROMOTER
Proximal CAP site

CACCC 5‘UTR
Distal CACCC TATA EXON 1 EXON 2 EXON 3 3‘UTR
Fig. 3.24 Sites of some mutations

5’ 3’
giving rise to β+ thalassaemia. Introns
1 3‘UTR(1)
are shown in blue, the 5′ and 3′ UTR in Promoter
mutations 9
green and important promoter 8 CAP+1(1) Consensus Polyadenylation
sequences (13) signal within
sequences in yellow. The dotted lines 5‘UTR(6)
3’UTR (6)
New splice site
are a diagrammatic indication of the New splice
Hb Malay in intron (1) Insertion (1)
Codon 24 New splice site in intron (1)
consensus sequences flanking the site in
Hb E
exons
invariant nucleotides at the splice Hb Knossus
sites.
IVS2 654 C→T and 3′ UTR nt1570 T→C [175]. The heterozygosity for mild β+ thalassaemia and hete
great majority of cases of β thalassaemia result from rozygous β thalassaemia aggravated by coexisting
mutations in or near the β gene (HBB). However heterozygosity or homozygosity for ααα. β thalas
there are occasional cases with no detectable saemia mutations that are characteristically associ
abnormality in the β globin cluster and in which ated with clinical abnormalities in heterozygotes
inheritance is not linked to the cluster [176, 177]. are referred to as dominant β thalassaemia (see
Such cases can result from a mutation in a gene p. 134). Although, by definition, blood transfusion
encoding a trans‐activating factor, either in ERCC2 is not essential to maintain life in thalassaemia
(previously XPD), encoding a component of a gen intermedia, patients with more severe disease may
eral transcription factor, TF11H; in GATA1, encod be treated electively with blood transfusion in order
ing an erythroid‐specific transcription factor; or in to improve quality of life.
ASH1L, encoding a histone methyltransferase (see
Table 1.4). In the case of a GATA1 mutation, the β
β thalassaemia trait
thalassaemia trait can be associated with thrombo
cytopenia [172]. β thalassaemia trait or heterozygosity for β thalas
β thalassaemia heterozygosity, also referred to as saemia is often completely asymptomatic and may
β thalassaemia trait, is usually clinically asympto therefore be referred to as β thalassaemia minor.
matic, a condition that may also be referred to as β However, a controlled study in Sri Lanka found that
thalassaemia minor. Homozygosity or compound β thalassaemia heterozygotes were significantly
heterozygosity for β thalassaemia usually leads to a more likely to have symptoms of anaemia such as
clinically severe phenotype, referred to as β thalas lethargy, fatigue and dizziness [178]. They also had
saemia major, in which the individual is dependent more febrile episodes, suggesting an increased
on blood transfusion. The term ‘thalassaemia inter susceptibility to infection [178]. In conditions of

media’ indicates that the clinical features of a case haemopoietic stress, for example during pregnancy
are intermediate between those of β thalassaemia or during intercurrent infections, the patient may
minor and of β thalassaemia major; the individual is become anaemic and may even require blood trans
symptomatic but although he or she may need fusion. A large retrospective study in Sicily found
occasional blood transfusions these are not essential an association with cholelithiasis, kidney disease,
to maintain life. The clinical conditions designated cirrhosis and mood disorders [179]. The risk of gall
thalassaemia intermedia cover a genetically hetero stones is further increased by coinheritance of
geneous group of disorders, of very variable sever Gilbert’s syndrome [180]. Renal tubular dysfunc
ity, including homozygosity or compound tion is seen in a significant minority of patients [181].
122 Chapter 3
There appears to be a higher incidence of rheuma There are conflicting data on the prevalence of
toid arthritis [182]. Women with β thalassaemia iron deficiency during pregnancy in women with β
heterozygosity are over‐represented among preg thalassaemia trait.
nancies in which the fetus has a neural tube defect,
and supplementation with folic acid may be par
Laboratory features
ticularly important in this group of patients [183].
Occasional patients have palpable splenomegaly The blood count characteristically shows a normal
and ultrasonography shows that splenic volume is or slightly reduced Hb, elevation of the RBC and
increased in comparison with controls [184]. reduction of the MCH and MCV. The MCHC is
Serum ferritin is higher in β thalassaemia usually normal when measured by impedance
heterozygotes [185]. Iron overload can occur,
counters (e.g. Coulter or Sysmex instruments) but
particularly as a result of long continued
may be reduced when measured by Siemens light
inappropriate administration of iron [186]. The clin scattering instruments. The red cell distribution
icopathological features of hereditary haemochro width (RDW), a measurement that reflects red cell
matosis are aggravated by β thalassaemia trait and anisocytosis, is usually normal. Mean values for
it is possible that the risk of iron overload is haematological variables (Hb, MCV and MCH)
increased in individuals with an HFE genotype differ significantly between β+ and β0 thalassaemia
associated with a mild risk of haemochromatosis trait but there is overlap [201, 202]. An MCV of less
[187]. A single case report suggests that iron over than 67 fl and an MCH of less than 21 pg give a
load may also be aggravated in patients with a reasonable separation of β0 from β+ thalassaemia
TRF2 mutation [188]. In patients with chronic hepa [202]. The reticulocyte count was elevated in about
titis C infection, β thalassaemia heterozygosity is a quarter of patients in one series [203] and in
associated with more hepatic iron deposition and another mean counts were 3% and 150 × 109/l
more fibrosis [189]. Similarly, there is an association [204]. The steady state Hb is lower if there is coin
with hepatic siderosis and fibrosis in patients with heritance of G6PD deficiency, a mean Hb of
non‐alcoholic fatty liver disease [190]. 125.5 g/l being observed in comparison with a
Several studies have found that β thalassaemia mean of 136.1 g/l for uncomplicated β thalassae
heterozygosity protects against myocardial infarc mia trait [205]; the reduction in MCV and MCH
tion in men [191, 192] or protects against ischaemic was marginally less when G6PD deficiency
heart disease [193] or severe coronary artery disease coexisted. In another study the MCV was again
[194], but in another study protection against coro somewhat higher in subjects with G6PD defi
nary artery disease was not found [195]. β thalas ciency, the mean difference being about 3 fl [139].
saemia trait is associated with a lower prevalence of The haematological abnormality is less if there is
hypertension and, in men, with protection against coinheritance of α thalassaemia (see later). The Hb
ischaemic stroke [196]. It may also protect against falls in pregnancy since plasma volume rises nor
peripheral vascular disease [197]. The plasma con mally but the rise in red cell mass is less than nor
centration of cholesterol and low density lipopro mal [206]. However an Hb of less than 80–90 g/l
teins is reduced, even in subjects with familial suggests a complicating factor [206]. The MCV
hypercholesterolaemia, while triglycerides and rises, on average, by about 2 fl, in comparison with
high density lipoproteins are normal [198, 199]. a rise averaging about 4 fl in haematologically nor
Beneficial effects on vascular disease could be mal subjects [207]. The percentage of reticulocytes
attributable to reduced viscosity as well as effects may be slightly elevated. Reticulocyte indices
on lipids. Epidemiological studies in Sardinia (pre (ADVIA 120 Hematology System) are significantly
viously a malaria area) and Melanesia have shown different from those of iron deficiency, MCVr
β thalassaemia to have an inverse relationship with (reticulocyte mean cell volume), CHr (mean reticu
altitude and the prevalence of malaria, consistent locyte content of haemoglobin) and MCHCr
with a protective effect [200]. Evidence from Sri (MCHC of reticulocytes), all of which are signifi
Lanka also supports a protective effect [75]. cantly lower, with CHr showing no overlap [208].
The characteristic red cell indices have been used another) is characteristic of iron deficiency and is
in a number of formulae designed to separate cases not usually a feature of uncomplicated β thalassae
of iron deficiency from cases of β thalassaemia trait mia trait. Patients with β thalassaemia trait who
[209–218]; these were tabulated in the first edition have required splenectomy for any reason may
of this book [219]. Although these formulae may be have a much more bizarre blood film (Fig. 3.26).
of some value in separating uncomplicated cases Zinc protoporphyrin is elevated in two thirds of
into two diagnostic categories, they are unreliable patients; values tend to be lower than in iron defi
in children and during pregnancy and are of no ciency although there is some overlap [90]. Serum
help in patients who have both β thalassaemia trait soluble transferrin receptor concentration is
and iron deficiency. Patients who are under treat increased. Pyrimidine 5′ nucleotidase is reduced to
ment for iron deficiency and those who have poly levels comparable with those seen in heterozygotes
cythaemia vera complicated by iron deficiency can for inherited deficiency [221]. Red cell survival is
also have results more suggestive of thalassaemia normal [222].
trait than of iron deficiency. In a study in randomly In subjects with G6PD deficiency, levels are
selected adult patients with mild microcytosis none higher in those with β thalassaemia trait when
of the four most popular formulae were found to be expressed per gram of haemoglobin [139]. However,
more effective than the MCV (cut‐off point <72 fl) in it should be noted that the results of assays for
distinguishing thalassaemia trait from other condi G6PD are altered by coexisting β thalassaemia trait,
tions [218]. These various formulae may be useful depending on how they are expressed. One study
in indicating the most likely diagnosis but they are showed that when G6PD levels were expressed in
not of use when it is necessary to either make a defi terms of grams of haemoglobin or millilitres of red
nite diagnosis of β thalassaemia trait or exclude this blood cells, concentration appeared to be increased
diagnosis. Their use is therefore not recommended. in individuals with β thalassaemia trait, whereas if
Ferrokinetic and red cell survival studies indicate expressed in terms of number of red cells, concen
an increase in ineffective erythropoiesis and a short tration was normal [205]. In another study results
ened red cell life span [220]. were higher when expressed per gram of haemo
It should be noted that when a patient with β tha globin and to a lesser extent when expressed per
lassaemia trait develops megaloblastic anaemia or number of red cells [139].
significant liver disease the MCV and the MCH The bone marrow aspirate (Fig. 3.27) shows
may rise into the normal range. The same will occur increased cellularity as a result of erythroid hyper
with administration of drugs such as zidovudine or plasia. Some erythroblasts show defective haemo
hydroxycarbamide but in these patients there is globinisation and cytoplasmic vacuolation. An iron
usually a pre‐treatment blood count showing char stain may show heavy siderotic granulation.
acteristic red cell indices. Incubation of bone marrow with methyl violet
The blood film varies from almost normal, with shows small numbers of erythroblasts with α chain
only mild microcytosis, to markedly abnormal inclusions [206]. Ultrastructural examination shows
(Fig. 3.25). Abnormal features, in addition to micro α chain precipitates in a small proportion of late
cytosis, include anisocytosis, hypochromia and erythroblasts, accompanied by evidence of dys
poikilocytosis. Individuals with a more severe phe erythropoiesis such as duplication of the nuclear
notype may have prominent basophilic stippling, membrane (Fig. 3.28a), autophagic vacuoles
target cells and small numbers of irregularly con (Fig. 3.28b), myelin figures, glycogen accumulation
tracted cells. Elliptocytes are generally more charac and iron‐laden mitochondria [223]. The presence of
teristic of iron deficiency than of thalassaemia trait α chain precipitates correlates with reduced protein
but some thalassaemic individuals have prominent synthesis [223].
elliptocytes. Target cells and basophilic stippling In the neonatal period, babies with β thalassae
are generally more common in β thalassaemia than mia trait, in contrast to those with α thalassaemia
in iron deficiency. Anisochromasia (i.e. variation in trait, have a normal Hb and normal red cell indices.
the degree of haemoglobinisation from one cell to Differences from normal start to appear around the
124 Chapter 3
(a)
Fig. 3.25 Blood films of three

patients with β thalassaemia trait
showing the range of
abnormalities observed: (a) blood
film showing microcytosis,
anisochromasia, a teardrop
poikilocyte and basophilic
stippling; the red cell indices were
RBC 4.41 × 1012/l, Hb 140 g/l, Hct
0.42, MCV 69 fl, MCH 23.2 pg,
MCHC 334 g/l; (b) from the same
patient showing unusually
(b) prominent elliptocytes.
age of 3 months. By the age of 6 months there is no percentage is due to an absolute rather than merely
appreciable overlap. a relative increase in δ chain synthesis [206], for two
reasons. An excess of α chains favours formation of
haemoglobin A2; in addition, some promoter muta
Diagnosis
tions are associated with an increased rate of δ chain
Diagnosis rests on detection of an increased haemo synthesis. Haemoglobin A2 tends to be higher in β0
globin A2 percentage (Figs 3.29 and 3.30, see also than in β+ thalassaemia but there is no clear separa
Fig. 2.7) [224, 225]. The increased haemoglobin A2 tion [202]. Other inherited and acquired causes of a
(c)
Fig. 3.25 Continued. (c) from a

patient with RBC 7.3 × 1012/l, Hb
143 g/l, Hct 0.43, MCV 59 fl,
MCH 19.9 pg, MCHC 328 g/l,
haemoglobin A2 5.6%,
haemoglobin F 0.6%; (d) from a
patient with β0 thalassaemia trait
demonstrated by family studies.
MGG ×100. (d)

β thalassaemia trait who had been
splenectomised because of a
lymphoma involving the spleen,
showing very abnormal red cell
morphology. MGG ×100.
Fig. 3.27 Bone marrow aspirate in β
thalassaemia trait showing mild
dyserythropoiesis. MGG ×100.
(a)

examination of the bone marrow in
β thalassaemia trait showing α
chain precipitates and
dyserythropoiesis: (a) cytoplasmic
and intranuclear α chain
precipitates with duplication of the
nuclear membrane in association
with some of the precipitates
(arrow); (b) small masses of
cytoplasmic α chain precipitates
associated with dilation of the space
between the two layer of the
nuclear membrane and several
autophagic vacuoles. (By courtesy
of the late Professor Sunitha N.
Wickramasinghe and by permission
(b) of the British Journal of Haematology.)
high haemoglobin A2 percentage that should be

considered in the differential diagnosis are shown
a in Tables 3.8) [226–239] and 6.3. The relatively fre
quent occurrence of a variant haemoglobin A2, des
b
ignated A2′ (or B2), must be considered when
c seeking to diagnose β thalassaemia trait. This vari
ant A2 has the same retention time on HPLC as hae
d
moglobin S and thus, if haemoglobin S is absent, it
e is readily identified (Fig. 3.31). In one study haemo
globin A2′ was found in about 4% of samples [240].
f It is found mainly in those of African ancestry, in
whom prevalence is 1–2%, but has also been
g
reported from Oman [241]. There are also haemo
ASC globin A2 variants that are present at a low level as
a result of either a reduced rate of synthesis or insta
bility; nevertheless the sum of A2 plus A2 variant
Fig. 3.29 Haemoglobin electrophoresis on cellulose may reach diagnostic levels [242]. In individuals
acetate at alkaline pH showing increased haemoglobin A2 with haemoglobin G‐Philadelphia, haemoglobin G2
in two patients with β thalassaemia trait (lanes e and f); elutes in the S window [240]. In seeking to make a
ASC is a control sample containing haemoglobins A, S diagnosis of β thalassaemia trait, haemoglobin A2′
and C. Haemoglobin A2 has the same mobility at C and E. and any other variant A2 (whether resulting from a
Fig. 3.30 Capillary electrophoresis, electropherogram (Sebia Capillarys 3) showing increased haemoglobin A2.
128 Chapter 3
Table 3.8 Some inherited causes of an increased A proportion of cases, around a third to a half,
percentage of haemoglobin A2 [226–239]. also have an increased proportion of haemoglobin
F. Usually this is not more than 2–3% but in one
Disorders of globin genes
family with deletional β0 thalassaemia levels were
β thalassaemia trait (almost all cases)
Haemoglobin E trait [226]
4–11% [243].
In screening populations for β thalassaemia it is
Vietnamese/South‐East Asian type of deletional possible either to measure the haemoglobin A2 per
hereditary persistence of fetal haemoglobin (which spares
centage in everyone or to screen by the red cell indi
the δ gene)
ces and test only those with an MCV or MCH below
Hereditary high haemoglobin A2 with autosomal a certain cut‐off point. An MCH of less than 27 pg is
dominant inheritance [227]
an indication to quantitate haemoglobin A2.
Hereditary high haemoglobin A2 resulting from a Screening by the MCV may be less satisfactory as
mutation in the promoter of the δ gene [228] there is more variation in values between different
KLF1 mutation* (also have InLu phenotype and instruments and with some instruments the meas
haemoglobin F varying from normal up to 4.1 [230] or 4.4 ured MCV rises with storage of the blood sample
[229] or 4.7% [231]) [244]. In one study an MCH of less than 27 pg was
Triple α reported to cause a mild elevation [230] but not found to be marginally more sensitive than an MCV
confirmed [232] and in a third study this was rare (two of less than 80 fl [245]. In developing countries,
cases with haemoglobin A2 of only 3.5 and 3.6%) [229] where well calibrated automated instruments suit
Unstable haemoglobin
able for the measurement of red cell indices may not
be available, osmotic fragility can be used for initial
Sickle cell trait† screening. The cells of β thalassaemia trait (and of α
Sickle cell anaemia, particularly if there is coexisting α thalassaemia trait, iron deficiency and some haemo
thalassaemia† globinopathies) are more resistant to lysis in hypo
Sickle cell/β0 thalassaemia† tonic solutions than normal cells (see p. 360). Use of
a single‐tube visual osmotic fragility test reduces
Heterozygosity for certain other β chain variants, e.g.
the number of samples that have to be referred to a
haemoglobin Leslie [233]
central laboratory for definitive diagnosis. However
GA
γ fusion (one family) [234] it should be noted that the sensitivity of the osmotic
Haemoglobin anti‐Lepore Hong Kong (17–19%) [235] fragility test is reduced by coexisting α thalassae
mia, G6PD deficiency and South‐East Asian ovalo
Disorders unrelated to globin genes
Congenital dyserythropoietic anaemia (some cases cytosis; in populations where all three are common
among Israeli Bedouins) [236] the sensitivity may be as low as 70% [246].
Suitable methods for the quantification of haemo
Hereditary spherocytosis during acute haemolysis (one
globin A2 include haemoglobin electrophoresis fol
case, associated increase in haemoglobin F) [237]
lowed by elution and spectrophotometric estimation
Down’s syndrome infants (up to 80 days of age) [238] (only suitable when a laboratory is dealing with
Pseudoxanthoma elasticum [239] small numbers of samples), microcolumn chroma
tography, capillary electrophoresis and HPLC.
* Mild increase, e.g. 3.3–4.4%. Quantification of haemoglobin A2 by scanning den
† But note that if haemoglobin A2 is measured by HPLC there is
sitometry (Fig. 3.32) is not sufficiently precise to be
also a factitious rise as post‐translationally modified haemoglobin
S has the same retention time as haemoglobin A2. used in the diagnosis of β thalassaemia trait and IEF
has not been validated for this purpose.
The percentage of haemoglobin A2 is dependent
variant δ chain or from a variant α chain) must be on the precise mutation present. In most cases of
added to normal haemoglobin A2 and the total of heterozygosity for β0 or severe β+ thalassaemia the
the two used to determine whether ‘haemoglobin haemoglobin A2 is between 4 and 5% whereas when
A2’ is elevated. there is heterozygosity for mild β+ thalassaemia

(Bio‐Rad Variant II) in a patient with
homozygosity for haemoglobin A2′;
normal haemoglobin A2 is absent,
indicating the patient has no normal δ
gene; the A2′ was 2.2% and was in the
‘S window’ with a retention time of
4.58 minutes; from left to right, the
peaks are haemoglobin F (shaded),
post‐translationally modified
haemoglobin A (two peaks),
haemoglobin A0 and haemoglobin A2′.
Haemoglobin
there is usually 3.6–4.2% of haemoglobin A2 [247].
electrophoresis Higher percentages are seen in β thalassaemia trait
pH 8.9 consequent on deletion of the 5′ part of the β globin
Carbonic Hb gene (see Table 3.7). In addition, promoter muta
anhydrase A tions (such as −88 C→T and −29 A→G) and muta
tions of the initiation codon, such as initiation
codon A→G, are associated with higher levels [247].
Hb Normal subject By definition, in silent β thalassaemia the haemo
A2 Hb A2 2% globin A2 percentage is normal. For example, with a
C→G mutation at position 6, 3′ to the termination
codon the mean level is 2.4% [247]. A normal hae
moglobin A2 can also result from coinheritance of δ
thalassaemia.
When haemoglobin F is elevated it usually com
Carbonic Hb prises between 2 and 7% of total haemoglobin. A
anhydrase A
further rise occurs in pregnancy, paralleling the rise
that occurs in haematologically normal pregnant
women [206]. Elevation is usual when β thalassae
Hb
A2
Thalassaemia mia trait is caused by deletion of the 5′ part of the β
trait Hb A2 9%
globin gene. The level of haemoglobin F is influ
enced not only by the nature of the mutation caus
ing the β thalassaemia but also by coinheritance of
Origin non‐deletional hereditary persistence of fetal hae
moglobin mutations, some of which are quite com
Fig. 3.32 Densitometric scan of an electrophoretic strip
mon. A Kleihauer test shows that distribution is
in β thalassaemia trait and a haematologically normal
subject showing increased haemoglobin A2 in β heterogeneous. Quantification of haemoglobin F is
thalassaemia trait. This technique is not sufficiently not essential for diagnosis of β thalassaemia, which
precise be used for reliable diagnosis but is used here is dependent on detection of an increased concen
for illustrative purposes. tration of haemoglobin A2. However if a patient has
130 Chapter 3
red cell indices suggestive of thalassaemia but has a deficiency, has been found to be elevated in 51% of
normal haemoglobin A2 percentage it is essential to cases of β thalassaemia trait [89]. This test is there
exclude elevation of haemoglobin F since δβ thalas fore of limited value in distinguishing between
saemia is an alternative diagnosis. Quantification of these two conditions.
haemoglobin F will be automatically available if It should be noted that β thalassaemia trait can be
haemoglobin A2 is quantified by HPLC. If cellulose simulated by heterozygous inactivating mutations
acetate electrophoresis and microcolumn chroma of KLF1, which can cause microcytosis (MCV down
tography are employed as the routine diagnostic to 68 fl and MCH down to 21 pg), increased haemo
techniques the electrophoretic strip should be care globin A2 (up to 4.9%) and increased haemoglobin F
fully inspected for a prominent haemoglobin F (up to 7.4%) [231]. When a KLF1 mutation is coin
band. A concentration above 2% can usually be herited with β0 thalassaemia, there can be a severe
detected visually and it is then possible to select deficiency of haemoglobin A synthesis in utero and
samples for quantification (e.g. by alkali denatura at birth but with improvement predicted to occur
tion). Similarly, the presence of haemoglobin Lepore thereafter [248]. Homozygosity for a KLF1 mutation
should be excluded by cellulose acetate electropho can lead to β thalassaemia intermedia [249].
resis, IEF, HPLC or capillary electrophoresis. If β
and δβ thalassaemia and ‘thalassaemic haemoglobi Problems in the diagnosis of β thalassaemia trait
nopathies’ such as haemoglobins E and Lepore
have been excluded then the most likely explana Neonatal period. β thalassaemia cannot be diagnosed
tion of ‘thalassaemic’ red cell indices, in most ethnic from the haemoglobin A2 percentage in neonates as
groups, is α thalassaemia. γδβ thalassaemia is also a levels are low. However, by 6–12 months of age the
possible explanation but is very rare. Other patients average haemoglobin A2 percentage is higher than
will have iron deficiency anaemia with atypical red in other infants [250, 251] and by 6 months of age
cell indices. Children, in particular, may have iron the A2 percentage does not overlap with values seen
deficiency anaemia with an elevated red cell count. in babies with normal globin genes (Table 3.9) [251].
In adults with polycythaemia vera and complicat The rate of decline in haemoglobin F is slower than
ing iron deficiency the red cell indices are very simi in haematologically normal infants and adult levels
lar to those of thalassaemia trait, although the RDW are not reached until well into childhood [206].
tends to be higher.
A low serum ferritin indicates that there is iron Silent and almost silent β thalassaemia. There are β
deficiency but does not exclude a diagnosis of β tha gene mutations that cause minor or no haemato
lassaemia trait. An increased red cell protoporphy logical abnormalities in heterozygotes but that
rin, which has been used as a screening test for iron nevertheless can cause clinically significant disease
Table 3.9 Haemoglobin A2 percentage in normal babies and babies with β thalassaemia heterozygosity [251].
Normal babies Babies with β thalassaemia

heterozygosity
Number Mean (95% range) (%) Number Mean (95% range) (%)
Birth 16 0.4 (0–0.8) 31 0.5 (0.1–0.9)
3 months 8 1.7 (1.1–2.3) 12 3.2 (1.8–4.6)
6 months 8 2.5 (1.9–3.1) 10 4.8 (3.4–6.2)
9–10 months 6 2.5 (1.7–3.3) 14 5.1 (4.1–6.1)
1 year 5 2.5 (1.9–3.1) 8 4.8 (4.0–5.6)

in homozygotes and compound heterozygotes. It is Mediterranean individuals. Heterozygotes for the

convenient to divide these mutations into two CAP +1 mutation have, for example, been observed
groups designated ‘silent β thalassaemia trait’ and to have the following mean values: MCV 79 fl, MCH
‘almost silent β thalassaemia trait’. In silent β thalas 24.7 pg, Hb A2 3.4% (data kindly provided by Dr
saemia trait both the red cell indices and the haemo John Old). Heterozygotes for IVS1 6 T→C had an
globin A2 percentage are normal. It is inevitable that MCV ranging from 60 to 78 fl and an A2 percentage
most of these cases will be missed in the routine ranging from 3.2 to 4.6% [257]. Individuals carrying
diagnostic laboratory. In almost silent β thalassae these mutations will sometimes have the phenotype
mia trait, red cell indices are abnormal but the hae of mild β thalassaemia trait and sometimes of almost
moglobin A2 percentage is not increased. The silent thalassaemia trait.
haematological phenotype thus resembles that of α
thalassaemia trait. Both silent and almost silent β Other causes of β thalassaemia trait with a normal hae
thalassaemia trait can be detected by studies of moglobin A2 percentage. Both inherited and acquired
rates of globin chain synthesis and by molecular abnormalities can lead to β thalassaemia trait with a
genetic analysis. Silent and almost silent β thalas normal haemoglobin A2 concentration.
saemia trait can also be referred to as ‘normal A2 β Iron deficiency has been found to lower the hae
thalassaemia’, although it should be noted that moglobin A2 percentage both in individuals with no
there are other causes of β thalassaemia trait with a defect of globin genes and in those with thalassae
normal haemoglobin A2 concentration. Some muta mia trait [261–266], although not all studies have
tions are so mild that they are always silent (e.g. the confirmed this observation [267]. In one study of
+1480 C→G mutation in Greek populations) [247, four patients the mean haemoglobin A2 percentage
252]. In other mutations with a greater reduction in increased from 3.07 to 3.81% when iron deficiency
β chain synthesis the phenotype in heterozygotes was treated [266]. This is not likely to lead to any
varies from silent or almost silent to mild β thalas diagnostic problem in individuals with a genetic
saemia trait. For example, fewer than half of hete abnormality that is usually associated with a
rozygotes for −101 C→T are truly silent [252]. In one marked elevation of haemoglobin A2. However, in
study of 45 heterozygotes, 17 subjects (38%) were those with mild β thalassaemia mutations, in which
completely silent, the same number had a normal the haemoglobin A2 is usually only slightly ele
MCH but an elevated haemoglobin A2, four (9%) vated, results may fall to within the normal range.
had an abnormal MCH but a normal A2 and in four For these reasons it is preferable not to attempt to
(9%) both the MCH and the haemoglobin A2 were exclude a diagnosis of β thalassaemia trait in
abnormal [253]. The others had only a modest ele patients with severe or moderately severe iron defi
vation of haemoglobin F. Among individuals iden ciency but rather to correct the iron deficiency and
tified only because of a child with β thalassaemia then measure A2 percentage if the red cell indices do
intermedia, half were completely silent, suggesting not return to normal. An exception to this generali
ascertainment bias in the larger group [253]. sation is in the case of pregnant women, where
Some of the mutations that may lead to a silent or diagnosis of β thalassaemia trait is required without
almost silent β thalassaemia phenotype, and the eth delay. In such circumstances it may be necessary to
nic groups in which they occur, are shown in measure haemoglobin A2 in the partner and, if he is
Table 3.10 [10, 167, 171, 176, 247, 252, 254–259]. The found to have β thalassaemia trait, consider pro
commonest mutations responsible for silent β tha ceeding to molecular analysis in women with bor
lassaemia are −101 C→T and −92 C→T. Mean values derline haemoglobin A2 percentages. Folic acid
reported for individuals carrying the latter mutation deficiency can also lower the percentage of haemo
are MCV 83.9 fl, MCH 28.6 pg and haemoglobin A2 globin A2 and interfere with the diagnosis of β tha
3.4% [260]. Almost silent β thalassaemia trait results lassaemia trait although, in individuals who do not
from a small group of mild β thalassaemia muta have β thalassaemia trait, megaloblastic anae
tions, such as CAP +1 A→C in South Asian (Indian) mia – resulting from deficiency of either vitamin B12
populations and, occasionally, from IVS1 6 T→C in or folic acid – causes a significant rise in haemoglobin
132 Chapter 3
Table 3.10 Causes of normal haemoglobin A2β thalassaemia.
Mutation Origin Usual haemoglobin A2 Usual MCH (mean Usual MCV (mean
(mean or range) (%) or range) (pg) or range) (fl)
Silent β thalassaemia trait (normal MCV, MCH and haemoglobin A2 percentage)

−101 C→T Mediterranean 3.3 28 85
−92 C→T Mediterranean 3.5 28 82
CAP +8 C→T [171] Chinese 3.0 33.7 98
CAP +10(−T) [171, 254] Greek 2.5–2.7 30–33 94–102
IVS2 844 C→G Mediterranean 3.5 28–29 85
(Italian)
CAP +33 C→G [255] Mediterranean 3.0 29 86
(Greek Cypriot)
CAP +1480 C→G Mediterranean 1.9–3.4 20–31 79–95
(termination codon +6 (Greek)
C→G) [171, 247, 252]
Almost silent β thalassaemia trait (reduced MCV and MCH, normal haemoglobin A2 percentage)
IVS1 6 T→C Mediterranean* 3.5 23 71
Codon 27G→T Mediterranean and 2.1 25 71
(haemoglobin Knossos†) Middle Eastern
IVS1 5 G→A Corfu δβ‡ Mediterranean
IVS1 128 T→G Saudi 3.5 25 70
CAP +1 A→C South Asian 3.4 25 80
Mutation not linked to β Italian 1.6§ 23.5§ 76§
globin gene cluster [176]
CAP +22 G→A [256] Turkish, Bulgarian 3.9 23.5 79
Poly A T→C African
Other β thalassaemia Mediterranean
mutations with a defective including Sardinia
δ gene in cis or in trans
Indices typical of thalassaemia trait but haemoglobin A2 percentage normal

β thalassaemia caused by Various Normal Typical of β Typical of β
deletion of the locus thalassaemia thalassaemia
control region
γδβ thalassaemia Various Normal Typical of β Typical of β
thalassaemia thalassaemia
MCH, mean cell haemoglobin; MCV, mean cell volume.

Dr John Old from the National Haemoglobinopathy Reference Laboratory, Oxford, kindly provided some of the data on which this table
is based. Further information from references 10, 167, 171, 176, 247, 252, 255, 258.
* Sometimes referred to as the Portuguese mutation but is common throughout the Mediterranean area. Most common mutation in Malta
where half of heterozygotes for this mutation were demonstrated to have a haemoglobin A2 percentage below that usually used for
diagnosis of β thalassaemia heterozygosity [257].
† Does not separate from haemoglobin A on alkaline or acid electrophoresis or HPLC but separates on isoelectric focusing; 35–40% of
haemoglobin; often has a δ0 thalassaemia mutation in cis and haemoglobin A2 is then normal in heterozygotes and absent in
homozygotes [259].
‡ Actually represent δ0 with β+ thalassaemia in cis.
§ One case reported; compound heterozygotes with this mutation and β0 thalassaemia have thalassaemia intermedia or major.
A2 percentage, proportional to the degree of anae but abnormal red cell indices having normal find
mia [262]. ings with regard to both red cell indices and haemo
The phenotype of abnormal red cell indices with globin A2 concentration (i.e. an almost silent
a normal haemoglobin A2 concentration can also thalassaemia becomes a silent thalassaemia).
result from the coinheritance of δ thalassaemia (in When an individual has both haemoglobin H dis
cis or trans) and a ‘high haemoglobin A2’ β+ or β0 tha ease and β thalassaemia heterozygosity, the haemo
lassaemia mutation. Such coinheritance of β and δ globin A2 can be normal. Seven patients reported by
thalassaemia trait occurs particularly in Sardinians Liang et al. [271] had haemoglobin A2 varying from
and in Cypriots and has also been reported in 2.8 to 3.9% with a mean value of 3.5%. In another
China. The prevalence of δ0 thalassaemia in Sardinia series of seven patients, haemoglobin A2 percentage
is more than 1% [268]. The so‐called ‘Sardinian δβ ranged from 2.1 to 4.8% with two values clearly
thalassaemia’ (see p. 153) actually represents coin being within the normal range [150].
heritance of β0 thalassaemia and non‐deletional
hereditary persistence of fetal haemoglobin result
Coinheritance of β thalassaemia and either δ
ing from a mutation in the HBG1 gene; haemoglo
thalassaemia or a δ chain variant
bin A2 averages around 2.5%. Haemoglobin
Knossus, a ‘thalassaemic haemoglobinopathy’, is The coinheritance of β and δ thalassaemia can lead
also responsible for some cases of β thalassaemia to a normal haemoglobin A2 percentage or, if there
trait with a normal haemoglobin A2. This is caused is homozygosity for δ thalassaemia, absent haemo
by a δ 0 thalassaemia gene in cis; this variant hae globin A2. The red cell indices remain typical of tha
moglobin is not detectable by haemoglobin elec lassaemia trait and the diagnosis can be made by
trophoresis under standard conditions [167]. DNA analysis.
Homozygosity for δ0 thalassaemia can lead to β tha A diagnostic problem can also occur when an A2
lassaemia heterozygosity with absent haemoglobin variant is present. Failure to detect a split A2 band
A2 [269]. or peak, indicating either an α or a δ chain variant,
It should also be noted that εγδβ thalassaemia can cause the diagnosis of β thalassaemia trait to be
and deletions leading to loss of the upstream locus missed as a result of an incorrect estimation of hae
control region will not have any elevation of hae moglobin A2 percentage.
moglobin A2 but, from the point of view of antena
tal diagnosis, have a similar significance to β
The significance of borderline haemoglobin A2
thalassaemia heterozygosity.
A borderline high haemoglobin A2 (e.g. 3.4–3.9%) or
even a high normal haemoglobin A2 (e.g. 3.0–3.5%)
Coinheritance of α and β thalassaemia
can create diagnostic problems. Giambona et al.
When β thalassaemia is coinherited with heterozy investigated 23 485 Sicilian subjects with haemoglo
gosity for α0 thalassaemia or either heterozygosity bin A2 of 3.4–3.9% and normal or abnormal red cell
or homozygosity for α+ thalassaemia, the MCV and indices and found 17% with borderline haemoglo
MCH are higher and the haemoglobin A2 percent bin A2 levels [272]. Of the 410 who were further
age is lower; the haemoglobin A2 may be normal investigated because their partners had β thalassae
[270]. The coinheritance of either heterozygous α0 mia trait, 13% had either mild β thalassaemia or
thalassaemia or homozygous α+ thalassaemia trait coinheritance of δ and β thalassaemia. Galanello
and mild β thalassaemia (e.g. mutations such as et al. had previously investigated 125 individuals of
CAP +1 A→C trait) makes it more likely that the Sardinian descent with haemoglobin A2 between 3.0
diagnosis of β thalassaemia trait will be missed. In a and 3.5% (with normal or abnormal red cell indices)
small proportion of such individuals the MCV and and found 33 individuals (26%) to have β thalassae
MCH are normal. The coinheritance of α thalassae mia heterozygosity (either a mutation typically
mia trait can lead to a mutation that would other associated with mild β thalassaemia or coinher
wise have presented with a normal haemoglobin A2 itance of δ and β thalassaemia) [273]. Both Sicily and
134 Chapter 3
Sardinia are high prevalence areas for δ and β tha a more severe thalassaemia intermedia phenotype.
lassaemia. The proportion of patients with a border An unusual phenotype is of a significant haemolytic
line haemoglobin A2 having a significant genetic anaemia with a normal MCV but a low MCH [276].
abnormality would obviously be lower in a low Coinheritance of β thalassaemia trait and quadruple
prevalence area. α can lead to thalassaemia major [277].
Coinheritance of β thalassaemia trait and non‐
deletional hereditary persistence of fetal haemoglo
β thalassaemia trait with normal red cell indices but
bin leads to a modification of the phenotype of β
elevated haemoglobin A2
thalassaemia [278]. The MCH and MCV are higher
There are occasional patients with heterozygosity and the haemoglobin A2 percentage is somewhat
for mild β thalassaemia variants who have an ele lower; the serum soluble transferrin receptor con
vated haemoglobin A2 percentage despite normal centration is lower.
red cell indices. This was found, for example, in 17 There are several variant haemoglobins that,
of 45 individuals with heterozygosity for −101 when coinherited with β thalassaemia trait, give the
C→T [253]. clinical picture of β thalassaemia intermedia (see
In addition, coinheritance of α and β thalassaemia Table 3.11). There are many others, for example hae
can lead to normal red cell indices and more bal moglobin J‐Sardegna and haemoglobin Norfolk,
anced chain synthesis but with haemoglobin A2 that show no such interaction [279]. The inheritance
being elevated [274]. This is seen mainly in those of a β thalassaemia allele together with a high affin
with deletion of two of the four α genes or with non‐ ity haemoglobin, such as haemoglobin Crete or hae
deletional α thalassaemia. Gasperini et al. [275] moglobin San Diego, does not prevent the
found that of 315 individuals with normal red cell development of polycythaemia [279].
indices and a high haemoglobin A2, 313 had coexist
ing α and β thalassaemia.
Coinheritance with other erythrocyte abnormalities
Acquired abnormalities (e.g. liver disease) can
raise the MCV and MCH of patients with β thalas Coinheritance of β thalassaemia and hereditary
saemia trait into the normal range. elliptocytosis may cause a symptomatic haemolytic
Cases of these types will be detected if estimation anaemia necessitating splenectomy [280].
of haemoglobin A2 percentage is carried out regard Coinheritance with South‐East Asian ovalocytosis
less of the red cell indices (e.g. in laboratories using alters the blood film features, raises the MCHC and
HPLC or capillary electrophoresis for screening for normalises the osmotic fragility but does not alter
haemoglobinopathies and thalassaemias) but will the clinical severity [281].
necessarily be missed if A2 measurement is per
formed only on individuals with abnormal indices.
Dominant β thalassaemia
Most individuals who are heterozygous for a β tha
lassaemia mutation have clinicopathological fea
chain synthesis
tures described as ‘thalassaemia minor’ (i.e. the
Coinheritance of haemoglobin H disease and β tha blood count and film are abnormal but there are no
lassaemia has been discussed on p. 113. Although abnormal physical findings or symptoms). However
the condition may be milder than typical haemoglo some mutations produce clinically apparent abnor
bin H disease, abnormalities are more marked than malities in heterozygotes, mainly splenomegaly,
in typical β thalassaemia trait, with significant anae anaemia, jaundice and an increased incidence of
mia, more marked microcytosis, reduced MCHC, gallstones. This is referred to as dominant β thalas
increased RDW and significant iron overload. saemia. Dominant β thalassaemias are rare but cases
Coinheritance with triple α leads to worse chain are found scattered throughout the world. The clin
imbalance with a variable phenotype. Some patients icopathological features are those of thalassaemia
have anaemia and microcytosis whereas others have intermedia with both ineffective haemopoiesis and
Table 3.11 Genotypes that can produce the phenotype of β thalassaemia intermedia [162, 167, 173, 231, 235, 259, 279,
284, 289, 291, 302–314].
With two β thalassaemia alleles

Homozygosity or compound heterozygosity for mild or very mild β+ thalassaemia alleles, e.g. CAP +1 A→C, IVS1 6
T→C, +33 C→G, −101 C→T, −88 C→T, −87 C→G, −29 A→G* and some polyadenylation signal mutations, particularly if
coinherited with α thalassaemia trait (α0 thalassaemia trait or homozygous α+ thalassaemia trait)
Compound heterozygosity for a mild or very mild β+ thalassaemia allele and a severe β+ or β0 thalassaemia allele,
particularly when ameliorated by coinheritance of α thalassaemia trait (−α/−α or −−/αα) or non‐deletional HPFH (either
−158 Gγ C→T mutation, Gγ or Aγ promoter mutations or enhanced synthesis of γ chain not linked to the β globin locus†)
Homozygosity or compound heterozygosity for β+ thalassaemia if ameliorated by α0 thalassaemia heterozygosity or α+
thalassaemia homozygosity or non‐deletional α thalassaemia or the genotype of haemoglobin H disease†; compound
heterozygosity for β+ and β0 thalassaemia if ameliorated by the genotype of haemoglobin H disease
Homozygosity or compound heterozygosity for severe β+ or β0 alleles when ameliorated by coinheritance of non‐
deletional HPFH, particularly if homozygous (e.g. either −158 Gγ C→T mutation, −196 Aγ C→T or enhanced synthesis of
γ chain not linked to the β globin locus‡) or a heterozygous inactivating mutation of KLF1 [231] or α thalassaemia
(deletion of 2 or 3 α genes or non‐deletional α thalassaemia)
Homozygosity for β0 thalassaemia caused by 5′ deletions or point mutations of the β promoter leading to enhanced
haemoglobin F and sometimes haemoglobin A2 synthesis
Homozygosity for Spanish δβ thalassaemia
Homozygosity for ‘Corfu δβ thalassaemia’ (coinheritance of δ and β thalassaemia with a β gene mutation in cis that
downregulates the gene and is associated with increased haemoglobin F synthesis) [173, 311]§
Compound heterozygosity for δβ and β+ or β0 thalassaemia or homozygosity for δβ thalassaemia
Homozygosity for haemoglobin Lepore
Homozygosity for haemoglobin Knossos [259]
Homozygosity or compound heterozygosity for β0 or β+ thalassaemia with no detectable ameliorating factors
Homozygosity for haemoglobin Malay or compound heterozygosity for haemoglobin Malay and haemoglobin E [162]
With one β thalassaemia allele and a variant haemoglobin or HPFH

Compound heterozygosity for haemoglobin E or Hb Knossus and β thalassaemia or haemoglobin Lepore
Compound heterozygosity for β0 thalassaemia and haemoglobin D‐Punjab, haemoglobin C, haemoglobin O‐Arab,
haemoglobin City of Hope, haemoglobin Siriraj, haemoglobin Beograd or the unstable haemoglobins, haemoglobin
Acharnes, haemoglobin Arta or haemoglobin Lulu Island [279, 289, 302, 304]
Compound heterozygosity for β or δβ thalassaemia and deletional HPFH
Compound heterozygosity for β thalassaemia and haemoglobin anti‐Lepore Hong Kong [235]
With one β thalassaemia allele

β+ thalassaemia or β0 thalassaemia coinherited with heterozygosity or homozygosity for triple α or quadruple α (total of
5–7 α genes), e.g. αα/ααα, ααα/ααα, −α3.7/αααα, αα/αααα, αααα/αααα [305, 306] or duplication of entire α globin cluster
αα/αα,αα [307] sometimes with triple α coinherited ααα/αα,αα [292]
Dominant β thalassaemia due to very unstable β globin chain [284] due to an initiator codon mutation [291]
Somatic deletion of one β globin locus leading to mosaicism [309, 310]
HPFH, hereditary persistence of fetal haemoglobin.

* In black populations, because the same chromosome carries −158 Gγ C→T; in Chinese populations is associated with thalassaemia
major [167].
† However the genotype of haemoglobin H disease aggravates β0β0 thalassaemia.
‡ For example, Xp22.2‐linked, 6q23‐linked or 8q11‐linked.
§ 88–90% haemoglobin F and mild anaemia as γ genes are upregulated [311].
136 Chapter 3
a haemolytic component. Occasionally, as with hae often lead to production of either a truncated or elon
moglobin Boston‐Kuwait [282] or with a highly gated β globin chain that is very unstable and may
unstable elongated β chain due to an intragenic co‐precipitate with normal α chains [285, 286]. These
duplication [283], the phenotype is that of thalas mutations have a dominant negative effect conse
saemia major. Red cell survival is shorter than in quent on the presence of this abnormal protein, in
typical β thalassaemia trait and the reticulocyte comparison with the haploinsufficiency recessive
count is increased. Patients may require occasional effect of other mutations. Haemopoiesis is not only
blood transfusions. There is extramedullary hae ineffective but also dysplastic. Molecular mecha
mopoiesis and iron overload can occur. The blood nisms include the following [167, 173, 284, 286–290]:
film (Fig. 3.33) is usually very abnormal with prom • nonsense mutations in the third exon leading to a
inent basophilic stippling and circulating nucleated truncated unstable β chain;
red cells. The bone marrow shows erythroid hyper • mis‐sense mutations, particularly in the third
plasia and dyserythropoiesis (Fig. 3.34). Red cell exon but occasionally in the first or second exon;
inclusions are detected on incubation with vital • frameshift mutations (e.g. small deletions or, less
dyes. This gave rise to an earlier designation, ‘inclu often, insertions or a combination of deletion and
sion body β thalassaemia’, but this is a less appro insertion) in the third exon, or a mutation resulting
priate term than ‘dominant β thalassaemia’ since in aberrant splicing, leading to a truncated or elon
cases of thalassaemia major also have erythroblast gated β chain;
inclusions. Erythroblast inclusions are composed of • deletion or insertion of complete codons in the
excess α chains and abnormal β chains (Fig. 3.35), second or third exons leading to destabilisation;
whereas in β thalassaemia major they are composed • deletion within the 5′ consensus splicing region
of excess α chains alone. of the second intron leading to aberrant splicing.
More than 40 dominantly inherited alleles have One possible explanation of the particular asso
now been described [167, 173, 284]. In comparison ciation between dominant β thalassaemia and exon
with the common recessive forms of β thalassaemia, 3 mutations is that the abnormal globin chain that is
dominantly inherited β thalassaemia much more synthesised is sufficiently long to bind haem (bind
often results from mutations in the 3′ third of exon 2 ing sites being mainly encoded by exon 2) but lacks
or in exon 3 rather than in exon 1 or the 5′ part of the residues necessary for αβ dimer formation
exon 2 and is associated with substantial amount of (encoded by exon 3) [284]; the aberrant chains are
mutant mRNA [167, 284]. Mutations responsible more slowly degraded than the shorter globin
Fig. 3.33 Blood film in dominant β

thalassaemia showing anisocytosis,
poikilocytosis and basophilic
stippling. MGG ×100. (With thanks
to Dr Ayed Eden.)

dominant β thalassaemia showing
erythroid hyperplasia and dysplasia.
MGG ×100. (With thanks to Dr Ayed
Eden.)

examination in dominant β
thalassaemia. (By courtesy of the late
Professor Sunitha
N. Wickramasinghe.)
chains encoded by more truncated genes. There is Mutations of the initiator codon can also lead to
therefore damage to red cell precursors. Some of the an unusually severe phenotypic abnormality in het
abnormal β chains produced when there is deletion erozygotes with significant anaemia and spleno
or insertion of an entire codon may likewise be una megaly [291].
ble to form αβ dimers [284].
Many dominant β thalassaemias have been desig
Haemoglobin Lepore trait
nated ‘haemoglobin variants’ although it is rare to
be able to detect the variant haemoglobin predicted Unequal crossover during meiosis with deletion of
from the DNA sequence. These can be regarded as the 3′ part of the δ gene and the 5′ part of the β gene
hyper‐unstable haemoglobins [289]. leads to formation of a δβ fusion gene. The fusion
138 Chapter 3
gene encodes a variant δβ fusion chain that is syn Caribbeans in the UK). Haemoglobin Lepore
thesised at a much reduced rate in comparison with Baltimore is found in Brazil, Portugal and Italy.
normal β chain. The variant haemoglobin produced Haemoglobin Lepore Hollandia is rare, having
is designated haemoglobin Lepore (from the family been reported in isolated families in Papua New
name of the first patient in whom this variant hae Guinea, Bangladesh and Thailand. Haemoglobin
moglobin was recognised). Since the extent of the Lepore is important because of the possibility of
deletion varies, there are several different haemo interaction with haemoglobin S and with β thalas
globins designated ‘haemoglobin Lepore’, the most saemias. From the functional point of view it can be
common of which is haemoglobin Lepore Boston/ regarded as a δβ+ thalassaemia.
Washington. Others include haemoglobin Lepore
Baltimore, haemoglobin Lepore Hollandia, haemo
Laboratory features
globin Lepore Leiden [292] and haemoglobin
Lepore‐ARUP [293]. Haemoglobin Lepore Leiden The blood count and blood film (Fig. 3.36) features
results from a very complex crossover so that the cannot be distinguished from those of β thalassae
sequence of gene segments is actually δβδβδβ [292]. mia trait.
Haemoglobin Parchman, however, although it is Haemoglobin electrophoresis shows 5–15% of
also the result of a complex crossover with a δβδ haemoglobin Lepore with haemoglobin A2 being
haemoglobin sequence, is without apparent haema reduced, on average, to about half of the normal
tological consequences since the β gene is intact level. The percentage of haemoglobin Lepore
[294]; haemoglobin A2 is reduced. Haemoglobin Baltimore in heterozygotes is slightly but signifi
Palencia is also a δβδ hybrid, without clinical conse cantly higher than the percentage of haemoglobin
quences and with a normal haemoglobin A2, sug Lepore Boston [296]. The Haemoglobin A2 percent
gesting that the δ gene is retained [295]. age tends to be lower with haemoglobin Lepore
Haemoglobin Lepore Boston occurs with a low fre Baltimore than with haemoglobin Lepore Boston
quency in a variety of ethnic groups including [296]. Haemoglobin F is sometimes mildly
Italians (particularly from around Naples, and increased; this may be because of linkage to a poly
also 0.6% of Sicilians), Greeks (particularly morphism that determines haemoglobin F percent
Macedonians), Turks, Spaniards, Balkan popula age [296]. At least in Spaniards, the haemoglobin F
tions and individuals with African ancestry percentage tends to be higher in association with
(Cubans, Caribbeans, Afro‐Americans and Afro‐ haemoglobin Lepore Baltimore than in association
Fig. 3.36 Blood film in haemoglobin

Lepore trait showing hypochromia,
microcytosis and mild poikilocytosis;
red cell indices were RBC
5.36 × 1012/l, Hb 120 g/l, Hct 0.351,
338 g/l. MGG ×100.
with haemoglobin Lepore Boston [296]. Haemoglobin Lepore homozygosity

Haemoglobin Lepore has the same mobility as hae and compound heterozygosity
moglobin S on cellulose acetate electrophoresis at
Haemoglobin Lepore homozygotes have the clinical
alkaline pH (Fig. 3.37) and moves with haemoglo
and haematological picture of thalassaemia major or
bin A at acid pH. On HPLC, it has the same reten
thalassaemia intermedia. Haemoglobin electropho
tion time as haemoglobin A2 (Fig. 3.38). On capillary
resis shows haemoglobins F and Lepore only.
electrophoresis, the Lepore haemoglobins separate
Compound heterozygotes for haemoglobin Lepore
from haemoglobins A, S, E, C and A2, appearing in
and β thalassaemia similarly can have the clinical
the same zone as haemoglobin D‐Punjab and hae
and haematological features of either thalassaemia
moglobin G‐Philadelphia.
major or thalassaemia intermedia. Haemoglobin
electrophoresis shows haemoglobins F, Lepore and
A2 with or without some haemoglobin A.
a β thalassaemia intermedia
b β thalassaemia intermedia, also often designated
non‐transfusion‐dependent thalassaemia, refers to
c
a clinical phenotype with diverse genetic explana
d tions. In comparison with a typical patient with β
thalassaemia trait, there are significant clinical
e problems such as anaemia, splenomegaly, leg ulcers
f
and bony deformity. The condition differs from tha
lassaemia major in that the patient is not dependent
AFSC on regular blood transfusions for survival, although
transfusions may be needed occasionally (e.g. dur
AFSC
ing intercurrent infection), or may become neces
sary later in life. The severity of β thalassaemia
intermedia varies from a condition in which sur
acetate at alkaline pH in haemoglobin Lepore trait vival without transfusion is barely possible, and
(lane d); AFSC indicates a control sample containing there is growth retardation and bony deformity, to a
haemoglobins A, F, S and C. Haemoglobin Lepore has much milder condition that resembles β thalassae
the same mobility as haemoglobin S. mia trait but has a greater degree of anaemia and

(Bio‐Rad Variant II) in haemoglobin
Lepore trait; haemoglobin Lepore
was 13.4% and its retention time was
3.49 minutes; from left to right the
peaks are haemoglobin F (shaded),
haemoglobin A0 and haemoglobin
Lepore plus haemoglobin A2
(shaded).
140 Chapter 3
Fig. 3.39 Radiograph of the hands of

a patient with β thalassaemia
intermedia showing lucent areas that
represent expanded erythropoiesis.
(With thanks to Dr Saad Abdalla.)
Fig. 3.40 Computed tomography

(CT) scan of abdomen showing a
tumour‐like mass of haemopoietic
tissue in the liver in a patient with β
thalassaemia intermedia. (With
thanks to Dr S.K. Ma and W.Y. Au
and by permission of the British
Journal of Haematology.)
splenomegaly. There is expansion of the medullary formed a tumour‐like mass, detectable on com
cavity (Fig. 3.39). There can be extramedullary hae puted tomography (CT) scanning (Fig. 3.40) [297].
mopoiesis. Some patients develop symptoms Iron overload, hypothyroidism and gonadal failure
resulting from pressure on vital organs when can occur. Hepatic iron is preferentially in hepat
extramedullary haemopoietic tissue forms tumour‐ ocytes rather than macrophages, leading to a
like masses; these are often in the mediastinum or disproportionately low serum ferritin. Hepatic

pleura or within the spinal canal, causing spinal complications, in addition to extramedullary hae
cord compression. In one reported patient mopoiesis, include fibrosis, cirrhosis and hepato
extramedullary haemopoietic tissue in the liver cellular carcinoma resulting from iron overload.
The incidence of gallstones is increased. Bone com β thalassaemia to thalassaemia intermedia as a

plications include expansion of the medullary cavity, result of progressive somatic clonal segmental dele
osteoporosis (related in part to gonadal failure) and tion that included the normal β globin gene cluster
fractures. Patients with thalassaemia intermedia [316]. In some patients the explanation for a thalas
can develop hypersplenism; splenic sequestration saemia intermedia rather than a thalassaemia major
has also been recognised. Cardiovascular complica or thalassaemia minor phenotype is not clear. Some
tions are common and include congestive cardiac genotypes are consistently associated with thalas
failure, acute pericarditis, chronic pericardial thick saemia intermedia whereas there are others that are
ening, incompetence of mitral and aortic valves and sometimes associated with thalassaemia major and
pulmonary hypertension [298]. There is an increased sometimes with thalassaemia intermedia.
incidence of venous thromboembolism and portal In some communities, β thalassaemia intermedia
vein thrombosis, the hypercoagulable state being is not uncommon. For example, in Sardinia 10% of
particularly pronounced after splenectomy. Silent patient who are homozygotes or compound hete
cerebral infarction can occur [299]. Pulmonary rozygotes for β0 thalassaemia have a thalassaemia
hypertension is attributable both to recurrent intermedia phenotype, as a result of coinheritance
venous thromboembolism and to interstitial fibrosis of homozygous α+ thalassaemia, non‐deletional α
resulting from iron deposition. Glomerular hyper thalassaemia or heterocellular hereditary persis
filtration is common, with proteinuria occurring in tence of fetal haemoglobin [317]. In countries where
a minority of patients [300]; in small minority of haemoglobin E is common, compound heterozygo
patients end stage renal failure eventually occurs. sity for haemoglobin E and β thalassaemia makes
Rarely priapism has been described [301]. Cognitive thalassaemia intermedia a common phenotype.
impairment has been reported in adults with β It has been suggested that in Mediterranean
thalassaemia intermedia. countries with a high incidence of β thalassaemia,
The causes of thalassaemia intermedia are sum prenatal diagnosis should include seeking to iden
marised in Table 3.11 [162, 167, 173, 231, 233, 259, tify ααα and silent β thalassaemia due to −101 C→T
279, 284, 289, 291, 292, 302–314]. The condition can in the partners of women with β thalassaemia trait,
occur in patients with either one or two abnormal β in order to predict the occurrence of β thalassaemia
genes. Those who are homozygotes or compound intermedia and offer informed choice with regard
heterozygotes for β thalassaemia alleles either have to termination of pregnancy [318]. Heterozygosity
mutations that are usually associated with mild or for both ααα and β thalassaemia leads to the clinical
very mild β+ thalassaemia or have ameliorating fac picture of thalassaemia intermedia in three quarters
tors such as coinheritance either of α thalassaemia of cases while compound heterozygosity for both
trait or of a mutation or polymorphism that leads to ααα and −101 C→T silent β thalassaemia leads to a
enhanced γ chain synthesis in conditions of hae mild thalassaemia intermedia phenotype in 71% of
mopoietic stress (e.g. in BCL11A, Xmn1‐HBG2 or instances [318]. Testing for ααα would be most likely
HBS1L‐MYB). A polymorphism in the transcription to give relevant information since this was detected
factor gene CEBPE can also cause disease ameliora in 3.8% of partners of women with β thalassaemia
tion [315]. Those who have only a single abnormal β trait whereas −101 C→T silent β thalassaemia was
gene either have ‘dominant β thalassaemia’ or have detected in only 0.2% [318].
coinherited mutations that aggravate the chain
imbalance, such as homozygosity or heterozygosity
Laboratory features
for triple α or quadruple α. A rare cause of β thalas
saemia intermedia is the occurrence of a somatic The Hb is very likely to decrease with age; in one
mutation during development with loss of one β study of 584 patients there was a mean of about
gene from a proportion of haemopoietic cells in an 90 g/l in infancy in contrast to a mean of about
individual who is heterozygous for a β thalassaemia 60 g/l by 50 years of age (p < 0.001) [319]. The blood
mutation [309]. In a unique individual there was film shows features similar to those of typical β tha
transition, during adult life, of typical heterozygous lassaemia trait but the abnormalities are more severe
142 Chapter 3

patients with thalassaemia
intermedia: (a) adult female with
Hb 79 g/l, heterozygosity for β
thalassaemia and triple alpha (ββ
IVS1 nt5 mutation, ααα/αα) (with
(a) thanks to Dr N. Jackson).
(Fig. 3.41). In addition to hypochromia, microcyto it should be noted that hepatic steatosis is not
sis, anisocytosis, poikilocytosis and basophilic stip uncommon in thalassaemia intermedia and leads
pling there may be polychromasia and circulating to higher serum ferritin in comparison with
erythroblasts. The findings on haemoglobin electro patients without steatosis [323].
phoresis or HPLC are dependent on the precise
underlying genetic defect (see Table 3.11). The hae
β thalassaemia major
moglobin A2 percentage is likely to be elevated
somewhat more than in β thalassaemia trait and the β thalassaemia major refers to patients with
haemoglobin F is elevated. A higher percentage of homozygosity or compound heterozygosity for β
haemoglobin F correlates with less morbidity [320]. thalassaemia who are dependent on blood transfu
Serum erythropoietin is increased, particularly sions to maintain life beyond early childhood. The
when haemoglobin F (a high affinity haemoglobin) alternative designation, transfusion‐dependent
is more than 50%. The bone marrow aspirate shows thalassaemia, is often used. It has been estimated
abnormalities of erythropoiesis that are more severe that in the UK there are 20–30 births per year of
than those of β thalassaemia trait (Fig. 3.42). babies with β thalassaemia major. Very rarely, het
The haematological features may be altered by erozygotes for β thalassaemia have the clinical
therapy. Response to hydroxycarbamide can be phenotype of β thalassaemia major as a result of
associated with a rise in Hb, MCV and haemoglobin coinheritance of extra copies of the α gene; one
F percentage and a fall in the number of circulating such patient was homozygous for ααα and another
erythroblasts [321]. was homozygous for αααα [324]. Very rarely there
Serum iron, transferrin saturation and ferritin is evolution of thalassaemia minor to thalassaemia
are elevated. If magnetic resonance imaging is not major as a result of uniparental disomy of 11p15
available, a serum ferritin of 800 ng/ml can be sequences [325]. This can be the result of mosaic
used as an indication for iron chelation therapy uniparental disomy with progressive clonal selec
with therapy being interrupted if the level falls to tion [326]. Very rarely β thalassaemia major devel
300 ng/ml and the dose of chelating agent esca ops in adult life as a result of a somatic deletion in
lated if the level rises to 2000 ng/ml [322]. However an individual with β thalassaemia heterozygosity.
(b)
(c)
Fig. 3.41 Continued. (b) 6‐year‐old

girl with 8 cm splenomegaly and
Hb 86 g/l with compound
heterozygosity for β0
thalassaemia and type 2
deletional hereditary persistence
of fetal haemoglobin; (c) adult
female with compound
heterozygosity for β0 and silent β
thalassaemia trait, Hb 89 g/l,
haemoglobin F 23%,
haemoglobin A2 6.3% (genotype
β 39 C→T, β −101 C→T); (d) adult
female with homozygosity for
very mild β+ thalassaemia, the
−29 A→G mutation. MGG ×100. (d)
144 Chapter 3
Fig. 3.42 Bone marrow film of a

patient with thalassaemia intermedia
(same patient as 3.41a) showing
erythroid hyperplasia and scanty
ragged cytoplasm.
The phenotype of β thalassaemia major can increased erythropoiesis, both in an expanded bone
also result from compound heterozygosity for a marrow compartment and at extramedullary sites.
β thalassaemia allele and a ‘thalassaemic haemo The expansion of haemopoietic bone marrow leads
globinopathy’, such as haemoglobin E or the less to bony deformity, particularly in the skull and facial
common haemoglobin Malay [162]. bones with frontal bossing, deformity of the facial
Patients with β thalassaemia major have both inef bones, displacement of the teeth and a ‘hair‐on‐end’
fective erythropoiesis and a considerably shortened appearance on skull radiography (Figs 3.43 and
red cell life span (20 days or less) leading to severe 3.44). There is bone pain, for example in the jaw and
anaemia. Ineffective haemopoiesis results from dam the vertebral column, and tenderness and an
age to erythroblasts, both by free α chains and from α increased incidence of fractures consequent on thin
chain precipitates. Free α chains are normally bound ning of cortical bone. Erythropoiesis at extramedul
by the erythroid‐specific molecular chaperone, alpha lary sites leads to gross hepatomegaly and
haemoglobin stabilising protein, but when its bind splenomegaly (Fig. 3.45). The splenomegaly is asso
ing capacity is exceeded they undergo auto‐oxida ciated with increased trapping of abnormal red cells
tion to highly damaging α‐hemichromes (α globin in the spleen and may, in turn, lead to hypersplenism
monomers containing oxidised ferric iron) and gen with an expanded plasma volume and pooling of red
erate reactive oxygen species [327]. The α‐hemi cells and platelets in the spleen. Ineffective hae
chromes cause clumping of band 3 protein, leading mopoiesis and shortened red cell life span lead to
to anti‐band 3 antibody‐mediated clearance of eryth mild jaundice and an increased incidence of gall
roblasts and erythrocytes. Reactive oxygen species stones. There is wasting of the limbs and stunting of
oxidise proteins and membrane lipids, thus contrib growth. Leg ulcers can occur. Severe anaemia can
uting to haemolysis, and also cause ineffective eryth cause high output cardiac failure. Anaemia also
ropoiesis by activation of apoptosis and by causes growth retardation. There can be a hyperco
degradation of GATA1 [327]. Haemolysis is both agulable state as a result of damage to the red cell
intravascular and extravascular, the former leading membrane and nitric oxide scavenging; prior
to nitric oxide scavenging and platelet activation. splenectomy can contribute. There is a resultant

Reduced serum haptoglobin and haemopexin increased incidence of venous thromboembolism,
reflects the intravascular element of the haemolysis. portal vein thrombosis and pulmonary hyperten
The disease usually presents in the first year of life, sion. Pulmonary fibrosis, attributable to iron deposi
from the age of 3 months onwards. Initial p
resentation tion, can also contribute to pulmonary hypertension
is usually with failure to thrive, episodes of infection and hypoxia. Cognitive impairment has been
or abdominal enlargement. There is markedly reported. There is an increased incidence of
eadaches (in a study group of β thalassaemia major

h
and intermedia and haemoglobin E/β thalassaemia),
which correlates with white matter lesions on mag
netic resonance imaging [328]. Many of the adverse
effects of β thalassaemia major can be largely avoided
by an appropriate blood transfusion programme.
This, however, leads to serious iron overload unless
chelation therapy is given. Iron overload in turn can
lead to cardiac and hepatic damage, hypopituitarism
(contributing to growth retardation), hypogonado
tropic hypogonadism with delayed puberty, diabetes
mellitus, hypothyroidism and hypoparathyroidism
(often subclinical). Adrenal insufficiency of central
origin may be subclinical and may only become
apparent under stress [329]. Cardiac damage can
manifest as left ventricular dilation with a reduced
ejection fraction or as diastolic dysfunction with
restrictive filling [330]. There can be valvular disease,
atrial flutter or fibrillation, ventricular tachycardia
Fig. 3.43 The face of a child with β thalassaemia major and heart failure. Males are more likely to have car
showing frontal bossing, prominence of the maxilla and diac complications than females with the same
displacement of the teeth. (By courtesy of the late degree of iron overload [331]. Liver disease includes
Professor Harry Smith.)
Fig. 3.44 Skull radiograph of a child

with β thalassaemia major showing a
‘hair‐on‐end’ appearance as a
consequence of marked erythroid
hyperplasia. (Reproduced from
Hoffbrand AV and Steensma DP, eds.
Hoffbrand’s Essential Haematology, 8th
edn. © 2020 John Wiley & Sons Ltd. By
kind permission of Professor Victor
Hoffbrand.)
146 Chapter 3
Fig. 3.45 An undertransfused child

with β thalassaemia major showing
abdominal distension and an everted
umbilicus caused by gross
hepatosplenomegaly; there is also
wasting of limbs.
fibrosis, cirrhosis and hepatocellular carcinoma with years of age whereas those with β+ homozygosity
possible interaction with transfusion‐transmitted may survive to late childhood [271]. With iron
hepatitis B or C in some countries. There is an chelation regimes becoming less arduous, it is more
increased incidence of papillary carcinoma of the feasible to start transfusion earlier, thus reducing
thyroid. There may also be vitamin D deficiency and the frequency of alloimmunisation.
hypercalciuria of uncertain mechanism [332]. There
is an increased prevalence of renal calculi, correlating
Laboratory features
with hypercalciuria and reduced bone density [333].
There is an association with myelolipoma, a rare The Hb, RBC, Hct, MCV, MCH and MCHC are
benign tumour composed of mature adipose tissue reduced and RDW is increased. The Hb is usually in
and islands of haemopoietic cells [334]. the range of 30–70 g/l, the MCV 50–60 fl and the
In the absence of treatment, children with MCH 12–18 pg. The blood film (Fig. 3.46) shows
homozygosity for β0 thalassaemia usually die at 3–4 marked anisocytosis, poikilocytosis (including
(a)
(b)
Fig. 3.46 Blood films of four patients

with β thalassaemia major; patients
(a) and (b) were being transfused;
patient (c) had not been transfused
for the previous 3 months because of
the development of red cell
alloantibodies; all these patients had
been splenectomised; α chain
precipitates are clearly seen in all
patients; patient (d) was a baby at
4 months of age who had never been
transfused. (Continued on p. 148.) (c)
148 Chapter 3
(d) Fig. 3.46 Continued.
fragments and teardrop poikilocytes), hypochromia which are very flat cells with little reduction in cell
and microcytosis. Basophilic stippling, Pappenheimer diameter but striking hypochromia.
bodies and target cells may be noted. Circulating The bone marrow aspirate (Fig. 3.47) shows gross
nucleated red cells showing defective haemoglobini erythroid hyperplasia. There is quite severe dyseryth
sation and dyserythropoietic features are present. ropoiesis with nuclear lobulation and fragmentation,
The total white cell count and the neutrophil count basophilic stippling, defective haemoglobinisation
are increased. In children with massive spleno and the presence of α chain precipitates. Actively
megaly, hypersplenism leads to aggravation of the phagocytic macrophages are prominent and pseudo‐
anaemia and leucopenia, neutropenia and throm Gaucher cells are present. Iron stores are increased.
bocytopenia. The percentage of reticulocytes is Ultrastructural examination shows that α chain pre
increased. The absolute reticulocyte count is stated to cipitates are often present in profiles of late polychro
be rarely high although it tends to increase after sple matic erythroblasts and in a low percentage of profiles
nectomy [206]. Serum levels of both haptoglobin and of early polychromatic erythroblasts; arrest in G1,
haemopexin are reduced, reflecting the intravascular seen quite frequently in the latter population, may be
element of the haemolysis. attributable to the presence of free α chains [336].
The concentrations of protein C, protein S and Erythroblasts, particularly those containing α chain
antithrombin are often reduced [335]. precipitates, show other structural abnormalities in
If the spleen has been removed, the usual features both cytoplasm and nucleus.
of hyposplenism are present – Howell–Jolly bodies, In the case of homozygotes or compound hete
target cells, lymphocytosis, thrombocytosis and rozygotes for β0 thalassaemia (β0β0), techniques such
giant platelets. Pappenheimer bodies are very as haemoglobin electrophoresis, IEF and HPLC
prominent and nucleated red cells are markedly show only haemoglobin F and haemoglobin A2
increased. After splenectomy the red cells may (Fig. 3.48). When there is homozygosity for β+ tha
show inclusions with the same staining characteris lassaemia (β+β+) or compound heterozygosity for β0
tics as haemoglobin; these stain supravitally with and β+ thalassaemia (β0β+) haemoglobin A is also
methyl violet. They represent α chain precipitates. present, in variable amounts, sometimes up to 35%
Such inclusions are present in much smaller num of total haemoglobin. In β thalassaemia major, the
bers in patients who have not been splenectomised. haemoglobin A2, percentage may be normal, ele
These α chain precipitates may also be detectable vated or, occasionally, reduced. A Kleihauer test
in circulating nucleated red cells. Following shows that haemoglobin F is irregularly distributed
splenectomy, the blood film may show leptocytes, between cells.
(a)
Fig. 3.47 Bone marrow aspirates

of two patients with β
thalassaemia major: (a) an MGG‐
stained film showing erythroid
hyperplasia and a debris‐laden
macrophage; (b) ultrastructural
examination showing α chain
deposits. (b, By courtesy of the
late Professor Sunitha N.
Wickramasinghe.) (b)
Biochemical tests show increased bilirubin, δβ, γδβ and εγδβ thalassaemias
increased urinary urobilinogen and hyperuricae
mia. Serum soluble transferrin receptor (an indi δβ and Aγδβ thalassaemias
cator of erythropoietic activity) is increased two‐ to δβ0 or (δβ)0 (delta beta zero) thalassaemia (some
threefold. Serum haptoglobin is greatly reduced times also designated GγAγ(δβ)0 thalassaemia)
or absent, as a result of intravascular haemolysis, results from deletion of both δ and β genes but
the level correlating inversely with serum trans with preservation of the γ genes. Aγδβ0 or (Aγδβ)0
ferrin receptor [337]. Serum haemopexin is simi thalassaemia (sometimes also designated
larly decreased. Free haemoglobin may be γ( γδβ)0 thalassaemia) results from deletions of
G A
detectable in the plasma and methaemalbumin the Aγ, δ and β genes. The phenotype of heterozy
may be present. gotes of both resembles that of β thalassaemia
150 Chapter 3

Variant II) in a baby at 4 months of age
with β thalassaemia major (same case as
Fig. 3.46d); the haemoglobin A2 was
increased at 1.4% but otherwise the
chromatogram is not distinguishable
from that of a newborn premature
neonate; from left to right, the peaks are
haemoglobin F, haemoglobin F (black)
and haemoglobin A2.
Fig. 3.49 Blood film of an adult male

with δβ thalassaemia trait; the red cell
indices were RBC 6.04 × 1012/l, Hb
140 g/l, Hct 0.42, MCV 69 fl, MCH
23.2 pg, MCHC 334 g/l. MGG ×100.
trait but the haemoglobin A2 percentage is not have thalassaemia intermedia rather than thalas
increased; since one δ gene has been lost it might saemia major. Homozygotes for δβ or Aγδβ thalas
be expected that haemoglobin A2 would be saemia generally have the clinical and
reduced but in fact it is often normal [338]. The haematological features of thalassaemia interme
blood film features (Fig. 3.49) are very similar to dia (Fig. 3.52); they have 100% haemoglobin F
those of β thalassaemia trait. Haemoglobin F is (Fig. 3.53) and necessarily a pancellular distribu
consistently elevated, usually between 5 and 20% tion [340]. In δβ thalassaemia homozygotes, Gγ
(Figs 3.50 and 3.51) [339]. The distribution of hae and Aγ globin chains are present in similar
moglobin F, best observed by flow cytometry, is amounts whereas homozygotes for Aγδβ thalas
heterocellular. Because of the increased synthesis saemia have only Gγ globin chains. Heterozygotes
of haemoglobin F, homozygotes and compound for either of these types of thalassaemia usually
heterozygotes with a severe β+ or β0 mutation may have splenomegaly and an Hb of 80–130 g/l. The
Fig. 3.50 Haemoglobin electrophoresis on cellulose acetate at

g
alkaline pH (lane b) of an adult male with δβ thalassaemia (same
patient as Fig. 3.49) showing increased haemoglobin F; AFSC AFSC
indicates a control sample containing haemoglobins A, F, S and C.
Fig. 3.51 HPLC (Bio‐Rad

Variant II) of a 10‐year‐old
Iraqi girl with heterozygosity
for δβ thalassaemia. The red
cell indices were RBC
5.41 × 1012/l, Hb 105 g/l, Hct
0.38, MCV 64 fl, MCH
19.4 pg, MCHC 303 g/l, red
cell distribution width
(RDW) 32%. Three other
heterozygotes in the family
had haemoglobin A2 of 2.7%,
2.9% and 2.9% with
haemoglobin F of 9.8%, 6%
and 12.6%. (With thanks to
Dr Hadeel Ibrahim.)
152 Chapter 3
Fig. 3.52 Blood film of an 11‐year‐old Iraqi girl

with homozygosity for δβ thalassaemia. The red
cell indices were RBC 5.30 × 1012/l, Hb 105 g/l,
Hct 0.313, MCV 59 fl, MCH 19.8 pg, MCHC
335 g/l, RDW 32.9%. MGG ×100. (With thanks
to Dr Hadeel Ibrahim.)
Fig. 3.53 HPLC (Bio‐Rad

Variant II) of an 11‐year‐old
Iraqi girl with homozygosity
for δβ thalassaemia (same
patient as Fig. 3.52) showing
an increase of acetylated
haemoglobin F and F0. (With
thanks to Dr Hadeel
Ibrahim.)
MCV may be reduced or low‐ normal and the globin also has α thalassaemia (with deletion of two
MCH reduced or normal. α genes) the indices will be significantly abnormal.
There are at least ten mutations giving rise to δβ0 DNA analysis is needed to make a reliable
thalassaemia. This type of thalassaemia is observed distinction.
in many ethnic groups including some Mediterranean
(Italians, Greeks and Turks), black, Spanish and
εγδβ thalassaemias
Japanese populations [173]. There are at least 12
mutations giving rise to Aγδβ thalassaemia. This type There are at least 20 mutations that either delete the
of thalassaemia also occurs in many ethnic groups, entire β gene cluster (at least 14 examples) or inacti
including Indian and Chinese populations. vate all genes of the cluster because the upstream
An unusual molecular mechanism underlying δβ regulatory LCRB is deleted (at least six examples)
thalassaemia is a δβ fusion gene, observed in a [343, 344]. When LCRB is deleted, all genes of the β
Senegalese family, that results in a δ0β+ thalassaemia cluster may be intact (Hispanic deletion) or there
with the δ promoter controlling β chain synthesis may also be deletion of ε and Gγ, leaving Aγ, δ and β
[341]. The heterozygote described had thalassaemia intact but inactivated [345] or deletion of ε only,
trait with a normal haemoglobin A2 percentage and leaving Gγ, Aγ, δ and β intact but inactivated (English
2.7% haemoglobin F. It is more usual for δβ fusion II) [343]. In the case of one large deletion, there was
genes to lead to synthesis of haemoglobin Lepore also dysmorphism and mild mental retardation,
(see earlier). possibly related [346]. This type of thalassaemia is
The Sardinian type of ‘δβ thalassaemia trait’ is correctly referred to as εγδβ or as εGγAγδβ thalassae
actually a phenocopy of δβ thalassaemia trait mias but more commonly the term γδβ (gamma
caused by coinheritance (in cis) of the codon 39 non delta beta) thalassaemia is used. Some cases result
sense mutation that is a common cause of β0 thalas from de novo mutation. All are rare and are recog
saemia in the Mediterranean area, plus a mutation nised only in heterozygotes. The homozygous state
of the Aγ promoter leading to overproduction of γ would be incompatible with fetal life. In the neona
chain. There are thalassaemic indices with normal tal period, heterozygotes are characterised by
haemoglobin A2 and 15–20% haemoglobin F [10], haemolysis, possibly with erythroblastosis, hepato
which is very largely α2Aγ2. In one described megaly and splenomegaly; there may be a need for
homozygote, there was microcytosis but the Hb blood transfusion at birth and during the first 6
was normal and the condition was clinically silent; months of life. Occasional cases have had life‐
there was 99.8% haemoglobin F and 0.2% haemo threatening neonatal anaemia or have required
globin A2 [342]. Another phenocopy, initially intrauterine transfusion [345, 347]. The blood film
described in Corfu, is caused by coinheritance of a shows microcytosis and sometimes basophilic stip
deletion extending downstream from the ψβ gene pling (Fig. 3.54) and the reticulocyte count may be
and encompassing the δ gene, and a point mutation increased. Thereafter the phenotype resembles that
in the β gene. Heterozygotes have the phenotype of of β thalassaemia trait, although there may be anae
δβ thalassaemia trait while homozygotes have mia, and microcytosis and hypochromia tend to be
almost 100% F, traces of haemoglobin A and no hae more severe [343, 344]. Anaemia may become severe
moglobin A2 [10]. during pregnancy [348]. There can be mild haemol
Haemoglobin Lepore (see earlier) can be regarded ysis [344]. There is no elevation of haemoglobin F or
as a type of δβ+ thalassaemia since there is a reduced A2. The diagnosis can be made only by chain syn
rate of synthesis of both δ and β chains. thesis studies and DNA analysis. Deletion of the β
The red cell indices are important in distinguish gene can also be detected by fluorescence in situ
ing between δβ thalassaemia trait and hereditary hybridisation [347].
persistence of fetal haemoglobin. However, if a Coinheritance of εγδβ and triple α can lead to
patient with hereditary persistence of fetal haemo hydrops fetalis [348].
154 Chapter 3
Fig. 3.54 Blood film in a neonate

with γδβ thalassaemia trait showing
microcytosis and basophilic
stippling. MGG ×100. (With thanks
to Dr Manju Bhavnani.)

(Bio‐Rad Variant II) in δ0
thalassaemia homozygosity showing
a total absence of haemoglobin A2.
δ thalassaemia
causes of reduced haemoglobin A2, which should
δ thalassaemia is of no clinical significance, since be considered in the differential diagnosis, see
it only affects synthesis of haemoglobin A2. It is, Tables 3.12 and 6.3.
however, of significance in relation to diagnosis Mutations responsible for δ thalassaemia include
of β thalassaemia trait since inheritance of δ tha a large deletion, point mutations and frameshift
lassaemia in cis or in trans to β thalassaemia mutations. The non‐deletional mutations can pro
means that the haemoglobin A2 will not be ele duce a premature STOP codon or interfere with
vated and the diagnosis of β thalassaemia hete either transcription or RNA processing or transla
rozygosity may be missed. Both δ0 and δ+ tion. There are also structural haemoglobin A2 vari
mutations exist. Heterozygotes and homozygotes ants synthesised at a reduced rate, thus leading to a
for δ+ thalassaemia have a reduced percentage of ‘thalassaemic haemoglobinopathy’. The so‐called
haemoglobin A2, while haemoglobin A2 is reduced Corfu δβ thalassaemia (see earlier) is actually a phe
in δ0 heterozygotes and absent in homozygotes nocopy of δβ thalassaemia caused by coexistence of
(Fig. 3.55). For other inherited and acquired δ0 and β+ thalassaemia.
Table 3.12 Inherited causes of a low haemoglobin A2.
δ+ thalassaemia heterozygosity or homozygosity
δ0 thalassaemia heterozygosity or homozygosity*
Very unstable δ chain variant
Some cases of δβ0 thalassaemia heterozygosity and all cases of δβ0 thalassaemia homozygosity* [338]
Some cases of Aγδβ0 thalassaemia heterozygosity [338] and all cases of Aγδβ0 thalassaemia homozygosity*
Deletional hereditary persistence of fetal haemoglobin heterozygosity and homozygosity* (except Vietnamese/South‐East
Asian type, which spares the δ gene and is associated with an increased haemoglobin A2)
α thalassaemia trait and haemoglobin H disease
Haemoglobin Lepore heterozygosity, compound heterozygosity† and homozygosity† (since one or both δ genes are lacking)
Haemoglobin Kenya heterozygosity
Heterozygosity for δ chain variants (but the total of A2 and variant A2 will generally be normal)
Trisomy D syndrome [361]
Bi‐allelic null KLF1 mutations [368]
* Absent haemoglobin A2 in homozygotes (and compound heterozygotes with haemoglobin Lepore).

† Absent haemoglobin A2 in homozygotes (and compound heterozygotes with δβ thalassaemia).
γ thalassaemia haemoglobin F, persisting beyond infancy, with lit

tle or no imbalance of chain synthesis and normal
γ thalassaemia refers to a reduced rate of synthesis red cell indices. Definitions of an increased haemo
of γ chain and therefore of haemoglobin F. This con globin F percentage in this context have included
dition is manifest maximally during intrauterine 1% or above [350] and 2% or above [351]. HPFH
life and, since there are normally four γ genes, clinical has also been defined in terms of the proportion of
sequelae are likely to be minor. One recognised F cells with less than 4% F cells being classified as
cause is deletion of the 3′ part of the Gγ gene and the normal, 4–8% as equivocal and more than 8% as
5′ part of the Aγ gene with production of a GγAγ diagnostic [350]. There are other inherited
fusion gene under the control of a Gγ promoter [349]. (Table 3.13) [231, 352–360] and acquired (Table 6.2)
causes of an increased proportion of haemoglobin
F. Conversely, a reduced percentage of haemoglo
Hereditary persistence of fetal bin F is observed in infants with Down’s syndrome
haemoglobin (HPFH) and other up to 60 days of life [361]. HPFH can result from
inherited causes of an increased deletions within the β globin cluster or from muta
proportion of haemoglobin F tions or polymorphisms of regulatory genes, either
In adult life haemoglobin F is usually quite a low on chromosome 11 or elsewhere. The deletional
percentage of total haemoglobin and is confined to HPFHs result from relatively large deletions that
a small proportion of cells, designated F cells. include the δ and β genes. The critical difference
However in many ethnic groups 10–15% of between deletional HPFH and similar deletions
individuals have a slight increase in the percentage that cause β0 thalassaemia may be that in the former
of haemoglobin F and the percentage of F cells. This there is deletion of a globin gene silencer located 5′
is the commonest form of hereditary persistence of to the δ globin gene [362]. Since there is no β gene
fetal haemoglobin (HPFH). on the affected chromosome, deletional HPFH
Hereditary persistence of fetal haemoglobin can behaves as if it were an allele of the β globin gene;
be defined as an inherited characteristic in which both homozygotes for HPFH and compound hete
heterozygotes show an increased proportion of rozygotes for HPFH and a β chain variant totally
156 Chapter 3
Table 3.13 Inherited conditions associated with high haemoglobin F percentage [231, 352–360].
Inherited abnormalities of globin genes

Heterozygotes and homozygotes for hereditary persistence of fetal haemoglobin
Deletional
Non‐deletional
γ gene mutation
Mutation in the promoter of the Gγ or Aγ gene
β thalassaemia
Heterozygotes for β thalassaemia (some cases, particularly those with deletions of the 5′ part of the β gene or promoter
mutations)
Compound heterozygotes and homozygotes for β thalassaemia (β thalassaemia intermedia and major)
Heterozygotes and homozygotes for δβ and Aγδβ thalassaemia
Heterozygotes and homozygotes for haemoglobin Lepore
Heterozygotes for haemoglobin Kenya
Sickle cell trait
Some cases, particularly with the Saudi/Indian haplotype
Sickle cell anaemia and other forms of sickle cell disease
Some cases, particularly during treatment with hydroxycarbamide or with certain haplotypes: higher in Senegal haplotype
than in Benin and Bantu haplotypes; particularly high in Saudi/Indian haplotype found in eastern province of Saudi
Arabia and India
Unstable haemoglobins
Inherited abnormalities and polymorphisms other than those of globin genes
Genes controlling haemoglobin F synthesis
BCL11A haploinsufficiency due to microdeletion* or downregulation due to deletion of downstream regulatory
element [352]
Heterozygous inactivating mutations of KLF1 [231] and compound heterozygosity for KLF1 mutations [353]
HBFQTL2 at 6q22.3–23.1, HBS1L −MYB intergenic region)
HBFQTL3 at Xp22.2
HBFQTL4 on 8q
Haematological disorders
Congenital aplastic anaemia (Fanconi anaemia)
Blackfan–Diamond syndrome, particularly during corticosteroid administration
Dyskeratosis congenita [354]
Congenital dyserythropoietic anaemia [355]
Shwachman–Diamond syndrome [356]
Hereditary spherocytosis [357]
Metabolic disorders
β‐ketothiolase deficiency (high levels of butyric acid) [358]
Disorders of proprionate metabolism [359]
Other
Osteopetrosis [360]
HBFQTL, haemoglobin F quantitative trait locus.

* There may be associated neurodevelopmental delay
lack haemoglobin A. Synthesis of haemoglobin F is and some of which are not. The difference between
upregulated because γ genes are brought into prox deletional HPFH and δβ0 thalassaemia is one of
imity to an enhancer 3′ to the deletion [311]. The degree. The former has 15–30% haemoglobin F and
non‐deletional HPFHs are a heterogeneous group almost balanced α and non‐α chain synthesis
of disorders, some of which are allelic to the β gene whereas the latter has 5–15% haemoglobin F and
Table 3.14 Distribution of haemoglobin F in various conditions associated with an increased haemoglobin F percentage.
Heterocellular Pancellular
Some types of non‐deletional HPFH Some types of non‐deletional HPFH
δβ thalassaemia Deletional HPFH
Haemoglobin Lepore trait Haemoglobin Kenya trait
Sickle cell anaemia, sickle cell/haemoglobin C disease Sickle cell/deletional HPFH compound heterozygosity
and sickle cell/β thalassaemia
β thalassaemia heterozygosity and haemoglobin E/β β thalassaemia homozygosity and compound heterozygosity
thalassaemia compound heterozygosity
HPFH, hereditary persistence of fetal haemoglobin.
unbalanced chain synthesis. The distribution of persistence of haemoglobin Portland and low
haemoglobin tends to be pancellular in the former haemoglobin A2 [368].
and heterocellular in the latter (Table 3.14).
However, there is actually a continuous spectrum
Deletional hereditary persistence of fetal
of disorders rather than two distinct groups and
haemoglobin
one condition that was previously designated type
6 HPFH [363, 364] is now considered to be more Deletional HPFH can result from a number of
correctly characterised as Aγδβ0 (i.e. Gγ(Aγδβ)0) tha deletions of the β globin gene cluster, which are
lassaemia [338]. The molecular basis of the differ shown diagrammatically in Fig. 3.56. The haema
ence is not clearly understood; juxtaposition of a tological features of heterozygous subjects are
downstream enhancer to the γ genes in HPFH but summarised in Table 3.15 [338, 369–380].
not in δβ thalassaemia has been proposed as a Deletional HPFH is quite common in some ethnic
mechanism [365]. groups. Its prevalence in Afro‐Americans is about
The distribution of haemoglobin F between cells 1 in 1000.
in various conditions associated with an increased The first type of deletional HPFH to be recog
percentage of haemoglobin F is best determined by nised was an African type of δβ0 HPFH described in
flow cytometry. an Afro‐American child from Baltimore, USA. An
A unique family has been reported in HPFH alternative terminology is (δβ)0 or GγAγ(δβ)0 HPFH.
(haemoglobin F 22% and 31%), which was due to There are now known to be at least six different
compound heterozygous mutation for a mis‐ deletions classified as δβ0 HPFH, two occurring in
sense and a nonsense mutation in a transcription subjects of African descent (HPFH‐1 and HPFH‐2),
factor gene, KLF1. There was also an increase in one in Indians (HPFH‐3), two in Italians (HPFH‐4
red cell p
rotoporphyrin and a mild normocytic and HPFH‐5) and one in Vietnamese/South‐East
or microcytic haemolytic anaemia [353]. Asians [338, 369, 370]. In the first five of these, both
Haploinsufficiency of KLF1 due to microdeletion the δ and the β gene are deleted but the two γ genes
can also be causative, with values of 7% and 17% are intact. Since homozygotes have no β or δ genes
in two syndromic patients [366]. A very high hae they cannot synthesise haemoglobin A or A2.
moglobin F (37%) and the persistence of haemo Haemoglobin F comprises 100% of haemoglobin.
globin Portland has similarly been described in Both Gγ and Aγ chains are synthesised but the pro
congenital dyserythropoietic anaemia type IV, portion varies in the different subtypes. The synthe
associated with a KLF1 mutation [367]. sis of γ chain is almost sufficient to compensate for
Compound heterozygosity for two KLF1‐null the lack of β chain synthesis so that there is no anae
mutations is associated with severe haemolytic mia. In the Vietnamese/South‐East Asian type the δ
anaemias with more than 70% haemoglobin F, gene is retained.
158 Chapter 3
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140

kb
LCR
ε Gγ Aγ ψβ δ β
Hb F
HPFH 1 Black 18.6–13.1

HPFH 2 Ghanaian 22.4–26.6
GγAγ(δβ)0 HPFH 3 Indian 17–25
HPFH 4 Italian 1 14–30
HPFH 5 Italian 2 16–20
Vietnamese/SEA 14.1–26.2
Mediterranean 5.9–19
SEA 9.9–20
GγAγ(δβ)0 Eastern European 13–24
thalassaemia Black 25
HbF 4–25% Macedonian/Turkish 4.2–13.5
Indian* 16.6
Spanish 5–13
Japanese 7–8
Black 4–16%
Chinese 9.3–23
Belgian 14.2–23
Gγ(Aγ δβ)0 Indian 9.7–18.1
Yunnanese
thalassaemia
Malaysian 2 9.9–12.5
HbF 4–23%
German 10–13.5
Turkish 17.2–22.9
†SE Asian (Thai) 12–17
Italian
* a Turkish GγAγ(δβ)0 thalassaemia is similar to Indian GγAγ(δβ)0 thalassaemia

† “HPFH 6”
Fig. 3.56 Deletions resulting in hereditary persistence of fetal haemoglobin or in δβ thalassaemia (modified from
reference [338]); GγAγ(δβ)0 = δβ0 thalassaemia; Gγ(Aγδβ) thalassaemia = Aγδβ0 thalassaemia.
Heterozygotes for δβ0 HPFH have a variable hae to cell. The globin chain synthesis ratio (α:non‐α) is
moglobin F percentage, depending on the precise approximately normal.
deletion (see Table 3.15). The haemoglobin A2 per Homozygotes are not anaemic. In fact, because
centage is either mildly reduced or normal, averag haemoglobin F has a higher oxygen affinity than
ing around half of the normal mean level in most haemoglobin A, there may be mild polycythae
subtypes. The Hb is normal but the MCV and MCH mia. In some homozygotes the red cell indices
may be somewhat reduced. The mean MCH varies resemble those of β thalassaemia trait with an
from about 26 pg in HPFH‐1 to about 28 pg in increased RBC and reduced MCV and MCH. In
HPFH‐3 [338]. The MCV shows similar variation others the RBC is towards the top and the MCV
between subtypes from a mean that is below the and MCH towards the bottom of their respective
lower limit of normal to a mean that is clearly nor normal ranges. The reticulocyte count is normal.
mal [338]. The blood film (Fig. 3.57) may be normal The blood film may show anisocytosis, poikilocy
or show an occasional target cell. A Kleihauer test or tosis, mild hypochromia, mild microcytosis and
flow cytometry shows pancellular distribution of target cells. ‘Cells resembling spherocytes’ (prob
haemoglobin F but there is some variation from cell ably irregularly contracted cells), have been
Table 3.15 Haematological features of heterozygosity for deletional hereditary persistence of fetal haemoglobin (HPFH)
[338, 369, 378–380].
Type of HPFN Usual Usual Hb A2 (%) Usual Gγ:Aγ ratio Molecular defect Reference
Hb F (%)
Afro‐American δβ0 15–30* 1.2–2.7 50:50 Deletion including δ and β 372, 379
(HPFH‐1) genes
African (Ghanaian) 20–30† Reduced 30:70 Deletion including δ and β 379

δβ0 (HPFH‐2) genes
Indian δβ0 (HPFH‐3) 17–25 1.6–2.2 70:30 Deletion including δ and β 373, 374
genes
Italian 1 (southern 14–30 1.7–2.0 35:65 Deletion including δ and β 375

mainland Italy) δβ0 genes
(HPFH‐4)
Italian 2 (Sicilian) δβ0 16, 20 2, 2.1 15:85 Deletion including δ and β 376
(HPFH‐5) genes
Vietnamese/ 20.7 ± 3.8 Increased 60:40 Deletion of β gene 338

South‐East Asian (3.8 ± 0.6)
Haemoglobin Kenya 5–28‡ 1.4–1.8 Mainly Gγ Deletion of part of the Aγ

gene, all of the δ gene and
part of the β gene with Aγ‐β
fusion
* The percentage of haemoglobin F is very significantly reduced by coexisting iron deficiency [377].
† The percentage of haemoglobin F is very significantly reduced by coexisting iron deficiency [378].
‡ Plus 5–28% (usually 7–12%) of haemoglobin Kenya (α2Aγβ2) [380].
Fig. 3.57 Blood film of an adult

African woman with hereditary
persistence of fetal haemoglobin.
Red cell indices were RBC
4.2 × 1012/l, Hb 120 g/l, MCV 88 fl,
MCH 28.6, MCHC 324 g/l; there was
23% haemoglobin F and 1.6%
haemoglobin A2. MGG ×100.
160 Chapter 3
described [371]. Globin chain synthesis shows an because it creates a cleavage site for the enzyme
imbalance similar to that in β thalassaemia trait Xmn1. For the sake of brevity, −158 Gγ C→T is used
with an α:non‐α ratio of about 1.4–3.0. A Kleihauer here to indicate this mutation and a similar notation
test shows a pancellular distribution of haemoglo is used for other mutations leading to non‐dele
bin F, an inevitable feature since only haemoglo tional HPFH. A further polymorphism also influ
bin F is present. ences the percentage of haemoglobin F, although it
Heterozygosity for haemoglobin Kenya (found in is not usually categorised with the HPFHs. It is
Kenyans and Ugandans) produces a variant of dele based on repeat sequences within HS2 (hypersensi
tional, pancellular HPFH. Since one δ gene has been tive site 2) of LCRB. The sequence is designated
deleted there is a reduced proportion of haemoglo (AT)x N12GT(AT)y when x and y are variable num
bin A2. Reported levels of haemoglobin Kenya have bers of repeats of a sequence. There are at least eight
varied from 6 to 23% (mean 13%) and of haemoglo different combinations of repeat sequences of which
bin F from 5 to 28% (mean 10%) [380]. Heterozygotes (AT)9N12GT(AT)10 is associated with an increased
may be haematologically normal or show slight synthesis of haemoglobin F. Both C→T at −158 and
anaemia and occasional target cells. Globin chain (AT)9N12GT(AT)10 are associated with an increased
synthesis is balanced. number of F cells (and a small increase in haemo
globin F percentage). It has been suggested that the
association −158 Gγ C→T with increased synthesis
Interactions with other haemoglobinopathies
of haemoglobin F is consequent only on its linkage
Deletional HPFH coinherited with βS leads to a very disequilibrium with (AT)9N12GT(AT)10, the latter
mild sickling disorder with about 30% F and about polymorphism being much more strongly linked to
70% S [372, 381]. Haemoglobin A2 may be low‐nor an increased production of haemoglobin F in one
mal or decreased. study [383]. However this seems unlikely in view of
Compound heterozygotes for deletional HPFH a considerable number of other studies that have
and β thalassaemia have about 70% haemoglobin F. linked −158 Gγ C→T to increased haemoglobin F in
The phenotype is variable. In the case of the two a variety of contexts.
African types, the compound heterozygous state is Haemoglobin F concentration is also affected by
phenotypically very mild whereas in the Indian type, genes encoding trans‐acting factors. One is the
although the heterozygotes do not have thalassaemic determinant at Xp22.2, which influences F‐cell
features, the compound heterozygotes have the clini production, and another has been mapped to 6q23.3
cal picture of thalassaemia intermedia [374, 381]. [167]. These give rise to increased synthesis of both
When coinherited with haemoglobin H disease in G
γ and Aγ globin chains and to GγAγ HPFH.
one family, HPFH was associated with some Non‐polymorphic non‐deletional HPFH results
improvement in the Hb, a reduced proportion of mainly from point mutations (or occasionally a
haemoglobin H and 11% haemoglobin Bart’s, sug small deletion or insertion) involving the β globin
gesting that the reduced number of α chains were gene cluster but not the δ and β globin genes them
demonstrating a greater affinity for β chains than selves. The mutations are in and around highly con
for γ chains [382]. served promoter motifs 5′ to either the Gγ or the Aγ
gene. They are at −114, −117 or −175 from the tran
scription initiation sites of these genes or clustered
Non‐deletional hereditary persistence
around −158 to −161 or −195 to −202. These muta
of fetal haemoglobin
tions are likely to alter the binding of trans‐acting
Non‐deletional HPFH is a heterogeneous group of factors to the promoter. This type of non‐deletional
disorders. The commonest form is associated with a HPFH can be further categorised according to
polymorphism in regulatory sequences of the β glo whether there is increased synthesis of Gγ or Aγ
bin gene cluster. There is a C→T change at position chain. Increased synthesis of Gγ chains results from
−158 from the Gγ gene, which is readily detected mutation upstream of the Gγ gene and increased
synthesis of Aγ chain from mutations upstream of described in black, white (British, Australian,
the Aγ gene. Interestingly, the same mutations have Italian) and Chinese individuals and in Brazilians
often been observed in the same position 5′ to one of various ethnic origins (see Table 3.16). They gen
or other gene. erally result from point mutations but in one muta
In non‐deletional HPFH there is increased syn tion, described in two individuals with sickle cell
thesis of haemoglobin F but haemoglobins A and A2 trait, there was a 13 bp deletion involving the distal
continue to be synthesised, although at a reduced CCAAT box at −115 to −111 [406].
rate, so that α:non‐α chain synthesis is fairly bal
anced. Whether the distribution of haemoglobin F
is pancellular or heterocellular is partly a function
chain synthesis
of the proportion of haemoglobin F present and the
sensitivity of the method used for its detection. Coinheritance of non‐deletional HPFH amelio
The reported types of non‐deletional HPFH for rates sickle cell anaemia and homozygous and
which a molecular mechanism has been defined are heterozygous β thalassaemia. For example, in
shown in Table 3.16 [338, 350, 351, 369, 381–409]. It both conditions −158 Gγ C→T and the
will be noted that, although the same mutation can (AT)9N12GT(AT)10 polymorphism are associated
occur 5′ to either the Gγ or the Aγ gene, the percent with an increase in haemoglobin F synthesis.
age of haemoglobin F characteristically seen may Homozygosity for −158 Gγ C→T is also associated
differ considerably. with an increased haemoglobin F percentage in
Typical of pancellular Gγ HPFH are the two point haemoglobin E/β thalassaemia and haemoglobin
mutations at −202 and −175 from the Gγ gene E disease [410].
observed in subjects of African ancestry. They show A child with both −117 Aγ G→A non‐deletional
15–25% and 17–30%, respectively, of haemoglobin HPFH and the genotype of haemoglobin H disease
F, almost all of Gγ type. The compound heterozy has been reported [382]. The percentage of haemo
gous state with haemoglobin S indicates that the β globin F was what was expected for the genotype,
gene in cis of the HPFH determinant continues to be 9.5%, and haemoglobin A2 was 1.3%. There was no
expressed, albeit at a reduced rate. The second of haemoglobin H but haemoglobin Bart’s was 11%,
these mutations has also been observed in indicating that the reduced amount of α chain was
Sardinians and white British individuals. Other Gγ combining preferentially with β chains rather than
promoter mutations, leading to either heterocellu with γ chains.
lar or pancellular HPFH, have been observed in The findings in other interactions of non‐dele
black, Yugoslavian, white Australian and Japanese tional HPFH and other haematological abnormali
populations (see Table 3.14). ties are summarised in Table 3.16.
Typical of pancellular Aγ HPFH is −117 Aγ G→A,
observed initially in Greeks but subsequently in
Sardinians, Chinese and people of African ances
Other inherited abnormalities
try. The percentage of haemoglobin F, of mainly Aγ
and haemoglobin F level
type, has generally been around 10–20%. Globin
chain synthesis is balanced and there is no haema An increased proportion of haemoglobin F is
tological abnormality. Two homozygotes have observed in a variety of inherited conditions, related
been described with approximately 75% haemo and unrelated to the β globin gene cluster (see
globin A, 24% haemoglobin F and 0.8% haemoglo Table 3.13).
bin A2, indicating that the δ and β genes in cis to the A number of haemoglobinopathies are associ
HPFH determinant are expressed, albeit at a ated, in a proportion of cases, with an increased
reduced level. percentage of haemoglobin F. The increase of hae
Other Aγ promoter mutations, leading to either moglobin F can sometimes be linked to the nature
pancellular or heterocellular HPFH, have been of the β globin gene mutation. Heterozygotes for
162 Chapter 3
Table 3.16 Haematological features of non‐deletional hereditary persistence of fetal haemoglobin. (Based on references
[338, 350, 351, 369, 381–409].)
Type of Ethnic group Molecular defect Percentage of Percentage of Reference

HPFH haemoglobin F in haemoglobin F during
heterozygotes erythropoietic stress
γ
G
African −202 Gγ C→G 15–25%* 19.9% and 23.5% with 381
βS in trans
Tunisian −200 Gγ +C 18–27% (48 and 49% in 338, 390

homozygotes)
Black/Sardinian/ −175 Gγ T→C 17–30* 29.5% with βS in trans 385–387

British
African −161 Gγ G→A 1–2% More marked increase 351, 409
Many ethnic groups, −158 Gγ C→T 2–3% or less (not always More marked increase 388
frequency of 0.32–0.35 elevated if otherwise with β thalassaemia in
genetically normal) trans; 10% with βS in
trans; 15% with βSβS
Afro‐American −158 GγGγ C→T† 2.3–3.8% 388

(XmnI Gγ
polymorphism)
Australian −114 Gγ C→G 8.5% 389
Japanese −114 Gγ C→T 11–14%
Czechoslovakian −110 Gγ A→C 1% 409
Yugoslav γ γ γ‡
G G A
About 5% 391
Portuguese (1 family) γ γ γ γ AGγAγ

G AG AG AG
0.3, 8% 391
A
γ African −202 Aγ C→T 1.6–3.9% 1.6–9% with βS in cis; 351
25% with βSβS
British/Australian −198 Aγ T→C 3.5–10%* (20% in

homozygote)
Italian −197 Aγ C→T 6% 409
Italian/Chinese −196 γ C→T§

A
12–21%* 38 and 40% with β 393, 394
thalassaemia in trans;
20% with β0 in cis¶
Brazilian (black, white −195 Aγ C→G 4.5–8.5% 7% with hereditary 395

or mixed) spherocytosis
Afro‐American −175 Aγ T→C 17–38% 38% with βC in trans 393, 409

and −158 Gγ C→T in cis
Cretan −158 Aγ C→T** Slight increase 338
Greek/Sardinian/ −117 Aγ G→A 7–20* (mean 13% in one 20–50% with β 397–403
black series, 9.7% in another); thalassaemia in trans
24% in two homozygotes
and 37.6 in another; 13.5%
in compound
heterozygosity with −158
G
γ C→T
Type of Ethnic group Molecular defect Percentage of Percentage of Reference

HPFH haemoglobin F in haemoglobin F during
heterozygotes erythropoietic stress
Afro‐American −114 Aγ C→T 3–5 404
Italian −113 Aγ A→G 6.5% 409
Black Del Aγ −114 to −102 30% and 31% with 405

sickle cell trait
Chinese Del Aγ −226 to −223 5.4% in homozygote, 406

(AAGC del) 3.2% in heterozygote
African Del Aγ −225 to −222 6–7% 409
γ γ
G A
Many ethnic groups Locus at 0.7–8.0
(‘Swiss type’) Xp22.2–22.3
Indian Locus at 6q22.3– (interaction with −158 Gγ 407

23.1, designated C→T)
haemoglobin F
quantitative trait
locus 2 (HBFQTL2)
British Unknown 1–10.8% (interaction with 350

autosomal locus −158 Gγ C→T)
other than
6q22.3–23.1
Maltese†† KLF1 at 19p13.12 3–19.5% 408
* Distribution of haemoglobin F is pancellular.

† Both γ genes are Gγ.
‡ However triplication of the γ gene does not necessarily lead to increased haemoglobin F.
§ In Sardinia occurs with a β0 mutation in cis producing the phenotype of δβ thalassaemia.
¶ So‐called Sardinian δβ thalassaemia – actually this mutation plus β0 thalassaemia.
** This mutations occurs in cis to −158 Gγ C→T.
†† Associated hypochromia, microcytosis and poikilocytosis.
non‐deletional β thalassaemia and 3′ deletional β at −540 from the β gene [411]. Both of these poly
thalassaemia have been found to have a slight ele morphisms were found to exert an influence on
vation of haemoglobin F (e.g. 1.5% in comparison haemoglobin F levels when found either in cis or in
with a normal of 0.7%), whereas haemoglobin trans to the abnormal β or δβ globin gene. The
Lepore heterozygotes had around 3% haemoglo (AT)9T5 polymorphism was also linked to an
bin F and heterozygotes for 5′ deletional β thalas increased percentage of haemoglobin F in β thalas
saemia had around 3.5% [381]. Elevation of saemia major [411].
haemoglobin F in individuals with abnormalities The haemoglobin F level in sickle cell anaemia is
of the β globin gene can also often be linked either determined by many factors (see p. 217), including
to polymorphisms within the β globin gene cluster age, sex, possibly the X chromosome F‐cell deter
that affect binding of transcription factors or to mining locus, various determinants in the β globin
other genetic factors. Heterozygotes for β thalas gene cluster and the number of α genes. However, it
saemia or haemoglobin Lepore with a high F per should be noted that interpretation of the F percent
centage have been found to have either the age in sickle cell anaemia is complicated by the
common −158 Gγ C→T or (AT)9T5 instead of (AT)7T7 preferential survival of F‐containing cells.
164 Chapter 3
In β thalassaemia major the haemoglobin F percent (d) high performance liquid chromatography
age is greatly elevated, in homozygotes for β0 thalas (e) cellulose acetate electrophoresis followed
saemia constituting almost all of the haemoglobin. by elution and spectrophotometry
Inherited metabolic disorders can affect haemoglo
bin F synthesis when gene expression is altered by an 3.5 Haemoglobin Bart’s hydrops fetalis
abnormal metabolite. Very abnormal h aemopoiesis, (a) is expected in about 50% of fetuses if both
for example in congenital dyserythropoietic or con parents have α0 thalassaemia
genital aplastic anaemias, can also be associated with (b) is a likely outcome in West Africans if
an elevation of haemoglobin F percentage. both parents have α thalassaemia trait
(c) is associated with an increased incidence
of pregnancy‐associated hypertension
Check your knowledge
(d) can cause developmental abnormalities in
One to five answers may be correct. Answers to limbs
most questions can be either found in this chapter (e) is associated with good oxygen delivery
or deduced from information given. Answers are to tissues
given on p. 184.
3.6 A decreased percentage of haemoglobin A2
3.1 An increased percentage of haemoglobin A2 is can be a feature of
expected in (a) α thalassaemia trait
(a) α thalassaemia trait (b) β thalassaemia trait
(b) β thalassaemia trait (c) haemoglobin Lepore trait
(c) δβ thalassaemia trait (d) iron deficiency anaemia
(d) γδβ thalassaemia trait (e) δβ thalassaemia trait
(e) silent β thalassaemia trait
3.7 α0 thalassaemia
3.2 Genes forming part of the β gene cluster (a) is common in Afro‐Caribbeans
include (b) in its homozygous form leads to
(a) α haemoglobin H disease
(b) β (c) is most often caused by deletion of both α
(c) γ genes on a single chromosome
(d) δ (d) cannot usually be suspected from the red
(e) ε cell indices
(e) can be diagnosed by haemoglobin
3.3 Moderate to marked microcytosis is usually a electrophoresis
feature of
(a) haemoglobin H disease 3.8 Hereditary persistence of fetal haemoglobin
(b) α0 thalassaemia trait can be caused by
(c) β thalassaemia trait (a) deletions that include the β and δ genes
(d) heterozygosity for hereditary persistence (b) deletion of an α gene
of fetal haemoglobin (c) deletion of a γ gene
(e) haemoglobin Lepore trait (d) mutation in a gene on chromosome 16
(e) point mutations upstream of either the Gγ
3.4 Suitable methods for quantifying haemoglobin or the Aγ globin gene
A2 for the diagnosis of β thalassaemia trait
include 3.9 δβ thalassaemia
(a) inspection of an electrophoretic strip (a) leads to an increased percentage of
(b) capillary electrophoresis haemoglobin F
(c) cellulose acetate electrophoresis followed (b) leads to an increased percentage of
by densitometric scanning haemoglobin A2
(c) usually results from deletion of the δ and Further reading

β genes
Steinberg MH, Forget BG, Higgs DR and Nagel RL, eds.
(d) when homozygous may lead to the
Disorders of Hemoglobin: Genetics, Pathophysiology, and
phenotype of thalassaemia intermedia
Clinical Management, 2nd edn. Cambridge University
(e) may be simulated by the coinheritance of
Press, Cambridge, 2009.
δ and β thalassaemia
Weatherall DJ. The thalassemias. In:
Stamatoyannopoulos G, Nienhuis AW, Majerus PW
3.10 Haemoglobin Bart’s hydrops fetalis
and Varmus H, eds. The Molecular Basis of Blood
(a) usually results from homozygosity for α+ Diseases, 3rd edn. WB Saunders, Philadelphia, 2001,
thalassaemia pp. 183–226.
(b) is mainly seen in those of Chinese or Weatherall D and Clegg JB, The Thalassaemia Syndromes,
South‐East Asian origin 4th edn. Blackwell Science, Oxford, 2001.
(c) is characterised by worse tissue hypoxia
than would be predicted from the
haemoglobin concentration References
(d) is characterised by low serum albumin
1 Cooley TB and Lee P (1925) A series of cases of sple
and generalised oedema nomegaly in children with anemia and peculiar
(e) occurs occasionally in Greeks and Cypriots bone changes. Trans Am Pediatr Soc, 37, 29.
2 Whipple GH and Bradford WL (1936) Mediterranean
3.11 The phenotype associated with β thalassaemia disease‐thalassemia (erythroblastic anemia of
homozygosity or compound heterozygosity Cooley); associated pigment abnormalities simulat
may be ameliorated by ing hemochromatosis. J Pediatr, 9, 279–311.
(a) coinheritance of ααα (triple alpha) 3 Piel FB and Weatherall DJ (2014) The α‐thalas
(b) coinheritance of γ thalassaemia semias. N Engl J Med, 371, 1908–1916.
(c) coinheritance of non‐deletional hereditary 4 Rigas DA, Koler RD and Osgood EE (1955) New
persistence of fetal haemoglobin hemoglobin possessing a higher electrophoretic
(d) coinheritance of α0 thalassaemia mobility than normal adult hemoglobin. Science,
heterozygosity 121, 372.
(e) coinheritance of α+ thalassaemia 5 Fessas P and Papaspyrou A (1957) New “fast”
homozygosity hemoglobin associated with thalassemia. Science,
126, 1119.
3.12 The following variant haemoglobins lead to 6 Edington GM and Lehman H (1954) A new haemo
an α or β thalassaemia phenotype globin found in a West African. Lancet, 267,
173–174.
(a) haemoglobin S
7 Ager JAM and Lehmann H (1958) Observations on
(b) haemoglobin C
some ‘fast’ haemoglobins: K, J, N and Bart’s. Br Med
(c) haemoglobin E
J, i, 929–931.
(d) haemoglobin Lepore
8 Higgs DR, Sharpe JA and Wood WG (1998)
(e) haemoglobin Constant Spring
Understanding α globin gene expression: a step
towards effective gene therapy. Semin Hematol, 356,
3.13 β thalassaemia trait with a normal haemoglobin 93–104.
A2 percentage may be caused by 9 Bernini LF and Harteveld CL (1998) α‐Thalassaemia.
(a) coexisting δ thalassaemia Bailliere’s Clin Haematol, 11, 53–90.
(b) dominant β thalassaemia 10 Weatherall DJ, The thalassemias. In:
(c) a β thalassaemia mutation association Stamatoyannopoulos G, Nienhuis AW, Majerus PW
with only a minor reduction of β chain and Varmus H, eds. The Molecular Basis of Blood
synthesis Diseases, 3rd edn. WB Saunders, Philadelphia, 2001,
(d) coinherited triple α pp. 183–226.
(e) β thalassaemia caused by deletion of 11 Morlé F, Francina A, Ducrocq R, Wajcman H,
LCRB Gonnet C, Philippe N et al. (1995) A new α chain
166 Chapter 3
variant Hb Sallanches [α2 104{G11) Cys→Tyr] asso disease in one homozygous patient. Hemoglobin, 39,
ciated with HbH disease in one homozygous 152–155.
patient. Br J Haematol, 91, 608–611. 23 Farashi S, Garous NF, Ashki M, Vakili S, Zeinali F,
12 Hill AVS (1992) Molecular epidemiology of the tha Imanian H et al. (2015) Homozygosity for the
lassaemias (including haemoglobin E). Bailliere’s AATAAA > AATA‐ ‐ polyadenylation site mutation
Clin Haematol, 5, 209–238. on the α2‐globin gene causing transfusion‐depend
13 Çürük MA, Baysal E, Gupta RB, Sharma S and ent Hb H disease in an Iranian patient: a case report.
Huisman THJ (1993) An IVS‐I‐117 (G A) acceptor Hemoglobin, 39, 355–358.
splice site mutation in the α1‐globin gene is a non‐ 24 Deshpande P, Kamalanathan N, Sampath E, George
deletional α‐thalassaemia‐2 determinant in an B, Shaji RV and Edison ES (2015) Characterization
Indian population. Br J Haematol, 83, 148–153. of clinical and laboratory profiles of the deletional
14 Higgs DR (1993) α‐Thalassaemia. Bailliere’s Clin α2‐globin gene polyadenylation signal sequence
Hematol, 6, 117–150. (AATAAA > AATA‐ ‐) in an Indian population.
15 Hall GW, Thein SL, Newland AC, Chisholm M, Hemoglobin, 39, 415–418.
Traeger‐Synodinos J, Kanavakis E et al. (1993) A 25 Huang LY, Yan JM, Zhou JY, Li J, Xie XM and Li DZ
base substitution (T→G) in codon 29 of the α2‐glo (2017) A severe case of hemoglobin H disease due
bin gene causes α thalassaemia. Br J Haematol, 85, to compound heterozygosity for deletion of the
546–552. major α‐globin regulatory element (MCS‐R2) and
16 Kattamis AC, Camaschella C, Sivera P, Surrey S and α0‐thalassemia. Acta Haematol, 138, 61–64.
Fortina P (1996) Human α‐thalassemia syndromes: 26 Wu M‐Y, Li J, Li S‐C, Li Y and Li D‐Z (2015)
detection of molecular defects. Am J Hematol, 53, Frequencies of HKαα and anti‐HKαα alleles in
81–91. Chinese carriers of silent deletional α‐thalassemia.
17 Chui DHK and Waye JS (1998) Hydrops fetalis Hemoglobin, 39, 407–411.
caused by α‐thalassemia: an emerging health care 27 Viprakasit V, Kidd AMJ, Ayyub H, Horsley S,
problem. Blood, 91, 2213–2222. Hughes J and Higgs DR (2003) De novo deletion
18 Barbour VM, Tufarelli C, Sharpe JA, Smith ZE, within the telomeric region flanking the human α
Ayyub H, Heinlein CA et al. (2000) α‐Thalassemia globin locus as a cause of α thalassaemia. Br J
resulting from a negative chromosomal position Haematol, 120, 867–875.
effect. Blood, 96, 800–807. 28 Wajcman H, Traeger‐Synodinos J, Papassotiriou I,
19 Higgs DR. The molecular basis of α thalassemia. Giordano PC, Harteveld CL Baudin‐Creuza V and
In: Steinberg MH, Forget BG, Higgs DR and Nagel Old J (2008) Unstable and thalassemic α chain hae
RL, eds. Disorders of Hemoglobin: Genetics, Patho moglobin variants: a cause of Hb H disease and
physiology, and Clinical Management, 2nd edn. thalassemia intermedia. Hemoglobin, 32, 327–340.
Cambridge University Press, Cambridge, 2009, 29 Tufarelli C, Stanley JA, Garrick D, Sharpe JA, Ayyub
pp. 241–265. H, Wood WG and Higgs DR (2003) Transcription of
20 Gibbons RJ and Higgs DR. The alpha thalassemia/ antisense RNA leading to gene silencing and meth
mental retardation syndrome. In: Steinberg MH, ylation as a novel cause of human genetic disease.
Forget BG, Higgs DR and Nagel RL, eds. Disorders Nat Genet, 34, 157–165.
of Hemoglobin: Genetics, Pathophysiology, and Clinical 30 De Gobbi M, Viprakasit V, Hughes JR, Fisher C,
Management. Cambridge University Press, Buckle VJ, Ayyub H et al. (2006) A regulatory
Cambridge, 2001, pp. 470–488. SNP causes a human genetic disease by creating
21 Kattamis A, Delaporta P, Fyaktou I, Vrettou C, a new transcriptional promoter. Science, 312,
Kyriakopoulou D, Stamoulakatou A et al. (2015) Hb 1215–1217.
Souli, a 6 bp in‐frame deletion on the HBA2 Gene 31 Weatherall DJ. Genetic disorders of haemoglobin.
(HBA2: c.[41‐46delCCTGGG]) leads to α‐thalas In: Hoffbrand AV, Lewis SM and Tuddenham EGD,
semia intermedia, when in trans to a single α‐globin eds. Postgraduate Haematology, 5th edn. Blackwell
gene deletion. Hemoglobin, 39, 55–57. Publishing, Oxford, 2005, pp. 85–103.
22 Farashi S, Garous NF, Ashki M, Vakili S, Zeinali F, 32 Flint J, Harding RM, Boyce AJ and Clegg JB (1998)
Imanian H et al. (2015) Hb Dartmouth (HBA2: The population genetics of the haemoglobinopa
c.200T>C): an α2‐globin gene associated with Hb H thies. Bailliere’s Clin Haematol, 11, 1–51.
33 Hurst D, Tittle B, Kleman KM, Embury SH and 45 Ruiz‐Reyes G (1998) Abnormal hemoglobins and
Lubin BH (1982) Anemia and hemoglobinopathies thalassemias in Mexico. Rev Invest Clin, 50,
in South East Asian refugee children. J Pediatr, 102, 163–170.
692–697. 46 Ko TM, Caviles AP, Hwa HL, Liu CW, Hsu PM and
34 Cabrera A and de Pablos JM (1984) Beta‐thalassae Chung YP (1998) Prevalence and molecular charac
mia minor in gypsies from the south of Spain. Br J terization of β‐thalassaemia in Filipinos. Ann
Haematol, 58, 377. Hematol, 77, 257–260.
35 Livingstone FB. Frequencies of Hemoglobin Variants. 47 Ko TM, Hwa HL, Liu CW, Li SF, Chu JY and Cheung
Oxford University Press, Oxford, 1985. YP (1999) Prevalence study and molecular charac
36 Yenchitsomanus PT, Summers KM, Bhatia KK, terization of α‐thalassaemia in Filipinos. Ann
Cattani J and Board PG (1985) Extremely high fre Hematol, 78, 355–357.
quencies of α‐globin gene deletion in Madang and 48 Gill PS and Modell B (1999) Thalassaemia among
on Kar Kar Island, Papua New Guinea. Am J Hum Asians in Britain. Lancet, 318, 873.
Genet, 37, 778–784. 49 de Silva S, Fisher CA, Premawardhena A,
37 Petrou M, Brugiatelli M, Old J, Hurley P, Ward RHT, Lamabadusuriya SP, Peto TEA, Perera G et al. (2000)
Wong KP et al. (1992) Alpha thalassaemia hydrops Thalassaemia in Sri Lanka: implications for the
fetalis in the UK: the importance of screening preg future health burden of Asian populations. Lancet,
nant women of Chinese, other South East Asian and 355, 786–791.
Mediterranean extraction for alpha thalassaemia 50 Badens C, Martinez di Montmuros F, Thuret I,
trait. Br J Obstet Gynaecol, 99, 985–989. Michel G, Mattei J‐F, Cappellini M‐D and Lena‐
38 Mukherjee MB, Lu CY, Ducrocq R, Gangakhedkar Russo D (2000) Molecular basis of haemoglobi
RR, Colah RB, Kadam MD et al. (1997) Effect of α‐tha nopathies and G6PD deficiency in the Comorian
lassaemia on sickle cell anemia linked to the Arab‐ population. Hematol J, 1, 264–268.
Indian haplotype in India. Am J Hematol, 55, 104–109. 51 Mouélé R, Bodo JM, Mpele DM, Feingold J and
39 Van den Broek NR, Letsky EA, White SA and Galacteros F (2000) β‐globin gene haplotypes and
Shenkin A (1998) Iron status of pregnant women: α‐thalassemia analysis in Babinga pygmies from
which measurements are valid? Br J Haematol, 103, Congo‐Brazzaville. Human Biol, 72, 379–383.
817–824. 52 Mouélé R, Pambou O, Feingold J and Galacteros F
40 Segbena AY, Prehu C, Wajcman H, Bardakdjian‐ (2000) α‐thalassemia in Bantu population from
Michau J, Messie K, Feteke L et al. (1998) Congo‐Brazzaville: its interaction with sickle cell
Hemoglobins in Togolese newborns: Hb S, Hb C, anemia. Hum Hered, 50, 118–325.
Hb Bart’s, and α‐globin gene status. Am J Hematol, 53 Diejomaoh FME, Haider MZ, Dalal H, Abdulaziz
59, 208–213. A, D’Souza TM and Adekile AD (2000) Influence of
41 Sirdah M, Bilto YY, El Jabour S and Najjar KH (1998) α‐thalassemia trait on the prevalence and severity
Screening secondary school students in the Gaza of anemia in pregnancy in women in Kuwait. Acta
strip for β‐thalassaemia trait. Clin Lab Haematol, 20, Haematol, 104, 92–94.
279–283. 54 Rund DG, Filon D, Jackson N, Asher N, Oron‐Kami
42 Daar S, Hussein HM, Merghoub T and V, Sacha T et al. (2004) An unexpected high fre
Krishnamoorthy R (1998) Spectrum of β‐thalas quency of heterozygosity for α‐thalassemia in
semia mutations in Oman, Ann NY Acad Sci, 850, Ashkenazi Jews. Blood Cells Mol Dis, 33, 1–3.
404–406. 55 Loukopoulos D and Kollia P. Worldwide distribu
43 Romana M, Kéclard L, Froger A, Berchel C and tion of β thalassemia. In: Steinberg MH, Forget BG,
Mérault G (1998) Spectrum of β‐thalassemia muta Higgs DR and Nagel RL, eds. Disorders of
tions in Guadeloupe (French West Indies) and inter Hemoglobin: Genetics, Pathophysiology, and Clinical
actions with other hemoglobinopathies. Ann NY Management. Cambridge University Press,
Acad Sci, 850, 423–425. Cambridge, 2001, pp. 861–877.
44 Casas‐Castaneda M, Hernandez‐Lugo I, Torres O, 56 Bernini LF, Geographic distribution of α thalas
Barajas H, Cibrian S, Zamudio G et al. (1998) Alpha‐ semia. In: Steinberg MH, Forget BG, Higgs DR and
thalassemia in a selected population of Mexico. Rev Nagel RL, eds. Disorders of Hemoglobin: Genetics,
Invest Clin, 50, 395–398. Pathophysiology, and Clinical Management.
168 Chapter 3
Cambridge University Press, Cambridge, 2001, pp. thalassemia in Saudi Arabia. Ann Saudi Med, 31,
878–894. 229–235.
57 Karimi M, Yavarian M, Afrasiabi A, Dehbozorgian J 70 Mesbah‐Amroun H, Rouabhi F, Ducrocq R and
and Rachmilewitz E (2008) Prevalence of β‐thalas Elion J (2008) Molecular basis of alpha‐thalassemia
semia trait and glucose‐6‐phosphate dehydroge in Algeria. Hemoglobin, 32, 273–278.
nase deficiency in Iranian Jews. Arch Med Res, 39, 71 Zorai A, Harteveld CL, Bakir A, van Delft P, Falfoul
212–214. A, Dellagi K et al. (2002) Molecular spectrum of α‐
58 Cabral M, Baião I, Valente I, Eloy M and Freitas M thalassemia in Tunisia: epidemiology and detection
(2008) Beta‐thalassemia: prevalence in Algarve, at birth. Hemoglobin, 26, 353–362.
Portugal. Haematologica, 93, Suppl. 1, 403. 72 Youssry I, El Badawy A, Samy RM, Salama N, Abd
59 O’Riordan S, Hien TT, Miles K, Allen A, Quyen NN, Elaziz D and Rizk S (2018) Prevalence of α‐thalas
Hung NQ et al. (2010) Large scale screening for hae semia in the Egyptian population. Hemoglobin, 42,
moglobin disorders in southern Vietnam: implica 243–246.
tions for avoidance and management. Br J Haematol, 73 Mason K, Gibson F, Higgs D, Fisher C, Thein SL,
150, 359–364. Clark B et al. (2016) Haemoglobin variant screening
60 Nainggolan IM, Harahap A and Setianingsih I in Jamaica: meeting Student’s request. Br J Haematol,
(2010) Hydrops fetalis associated with homozygo 172, 634–636.
sity for Hb Adana [alpha59(E8)Gly‐‐>Asp (alpha2)]. 74 Kumar R, Sagar C, Sharma D and Kishor P (2015)
Hemoglobin, 34, 394–401. β‐globin genes: mutation hot‐spots in the global
61 Cauchi MN (1970) The incidence of glucose‐6‐phos thalassemia belt. Hemoglobin, 39, 1–8.
phate dehydrogenase deficiency and thalassaemia 75 Premawardhena A, Allen A, Piel F, Fisher C, Perera
in Malta. Br J Haematol, 18, 101–106. L, Rodrigo R et al. (2017) The evolutionary and clini
62 Efremov GD (2007) Thalassemias and other hemo cal implications of the uneven distribution of the
globinopathies in the Republic of Macedonia. frequency of the inherited haemoglobin variants
Hemoglobin, 31, 1–15. over short geographical distances. Br J Haematol,
63 Petkov GH and Efremov GD (2007) Molecular basis 176, 475–484.
of β‐thalassemia and other hemoglobinopathies in 76 Wah ST, Yi YS, Khin AA, Plabplueng C and Nuchnoi
Bulgaria: an update. Hemoglobin, 31, 225–232. P (2017) Prevalence of anemia and hemoglobin dis
64 Baghernajad‐Salehi L, D’Apice MR, Babameto‐ orders among school children in Myanmar.
Laku A, Biancolella M, Mitre A, Russo S et al. (2009) Hemoglobin, 41, 26–31.
A pilot beta‐thalassaemia screening program in the 77 Schrier SL, Bunyaratvej A, Khuhapinant A,
Albanian population for a health planning pro Fucharoen S, Aljurf M, Snyder LM et al. (1997) The
gram. Acta Haematol, 121, 234–238. unusual pathobiology of hemoglobin Constant
65 Hadavi V, Taromchi AH, Malekpour M, Gholami B, Spring in red blood cells. Blood, 89, 1762–1769.
Law H‐Y, Almadani N et al. (2007) Elucidating the 78 Allen SJ, O’Donnell A, Alexander ND, Alpers MP,
spectrum of α‐thalassemia mutations in Iran. Peto TE, Clegg JB and Weatherall DJ (1997) alpha+‐
Haematologica, 92, 992–993. Thalassemia protects children against disease
66 Al‐Allawi NA, Jalal SD, Ahmed NH, Faraj AH, caused by other infections as well as malaria. Proc
Shalli A and Hamamy H (2013) The first five years Natl Acad Sci U S A, 94, 14736–14741.
of a preventive programme for haemoglobinopa 79 Lam YH and Tang MH (1999) Risk of neural tube
thies in Northeastern Iraq. J Med Screen, 20, defects in the offspring of thalassaemia carriers in
171–176. Hong Kong Chinese. Prenat Diagn, 19, 1135–1137.
67 Polus RK (2017) Prevalence of hemoglobinopathies 80 Beutler E and West C (2005) Hematologic differ
among marrying couples in Erbil province of Iraq. ences between African‐Americans and whites; the
Iraqi J Hematol, 6, 90–93. role of iron deficiency and α‐thalassemia on hemo
68 Baysal E (2001) Hemoglobinopathies in the United globin levels and mean corpuscular volume. Blood,
Arab Emirates, Hemoglobin, 25, 247–253. 106, 740–745.
69 Memish ZA and Saeedi MY (2011) Six‐year out 81 Wambua S, Mwacharo J, Uyoga S, Macharia A and
come of the national premarital screening and Williams TN (2006) Co‐inheritance of α+ thalassae
counselling program for sickle cell disease and β‐ mia and sickle trait results in specific effects on
haematological parameters. Br J Haematol, 133, 92 Fucharoen S, Winichagoon P, Wisedpanichkij R,

206–209. Sae‐Ngow B, Sriphanich R, Oncoung W et al.
82 Higgs DR and Bowden DK. Clinical and laboratory (1998) Prenatal and postnatal diagnosis of thalas
features of the α‐thalassemia syndromes. In: semias and hemoglobinopathies by HPLC. Clin
Steinberg MH, Forget BG, Higgs DR and Nagel RL, Chem, 44, 740–748.
eds. Disorders of Hemoglobin: Genetics, Pathophysiology, 93 McFarlane AG, Barth D, Crowther MA and
and Clinical Management. Cambridge University Lafferty JD (2003) The percentage of haemoglo
Press, Cambridge, 2001, pp. 431–469. bin Barts as a predictor for the alpha thalas
83 Pootrakul P, Winichagoon P, Fucharoen S, semia genotype in cord blood samples. Blood,
Pravatmuang P, Piankijagum A and Wasi P (1981) 102, 33b.
Homozygous haemoglobin Constant Spring: a need 94 Rugless MJ, Fisher CA, Stephens AD, Amos RJ,
for revision of concept. Hum Genet, 59, 250–255. Mohammed T and Old JM (2006) Hb Bart’s in cord
84 So C‐C, Liu AK, Tsang MH, Ngai DY, Leung K‐S blood: an accurate indicator of α‐thalassemia.
and Chan AY (2015) Genetic basis of persistent red Hemoglobin, 30, 57–62.
blood cell microcytosis in the Chinese unex 95 Bowden DK, Hill AV, Higgs DR, Oppenheimer SJ,
plained by phenotypical testing. J Clin Pathol, 68, Weatherall DJ and Clegg JB (1987) Different hema
69–72. tologic phenotypes are associated with the left
85 Krishnamurti L and Little JA (1998) Homozygous ward (‐α4.2) and rightward (‐α3.7) α+‐thalassemia
hemoglobin Constant Spring with normal electro deletions. J Clin Invest, 79, 39–43.
phoresis: a possible cause of under diagnosis. Ann 96 Tangvarasittichai O, Jeenapongsa R, Sitthiworanan
NY Acad Sci, 805, 415–419. C and Sanguansermsri T (2005) Laboratory inves
86 Charoenkwan P, Sirichotiyakul S, Chanprapaph P, tigations of Hb Constant Spring, Clin Lab Haematol,
Tongprasert F, Taweephol R, Sae‐Tung R and 27, 47–79.
Sanguansermsri T (2006) Anemia and hydrops in a 97 Wisedpanichkij R, Jindadamrongwech S and
fetus with homozygous hemoglobin Constant Butthep P (2015) Identification of Hb Constant
Spring. J Pediatr Hematol Oncol, 28, 827–830. Spring (HBA2: c.427T > C) by an automated high
87 Komvilaisak P, Komvilaisak R, Jetsrisuparb A, performance liquid chromatography method.
Wiangnon S, Jirapradittha J, Kiatchoosakun P and Hemoglobin, 39, 190–195.
Fucharoen G (2017) Fetal anemia causing hydrops 98 Sarkar AA, Banerjee S, Chandra S, Ghosh M,
fetalis from an alpha‐globin variant: homozygous Banerjee D, Choudhury MD et al. (2005) A novel
hemoglobin Constant Spring. J Pediatr Hematol/ 33.3 kb deletion (‐ ‐KOL) in the alpha‐globin gene
Oncol, 40, 405–408. cluster: a brief report on deletional alpha‐thalas
88 Henderson SJ, Sykes H, McCarthy J, Molyneux saemia in the heterogeneous eastern Indian popu
AT, Gallienne A and Old JM (2007) Mild haemo lation. Br J Haematol, 130, 454–457.
lytic anaemia due to homozygosity for Hb Icaria. 99 Eng B, Greenlay B and Waye JS (2009)
Alpha 2 stop codon TAA‐AAA. Br J Haematol, 137, Characterisation of the British α0‐thalassaemia
Suppl. 72. deletion: evidence of a founder effect in
89 Graham EA, Felgenhauer J, Detter JC and Labbe RF Newfoundland, Canada. Br J Haematol, 147,
(1996) Elevated zinc protoporphyrin associated 150–152.
with thalassemia trait and hemoglobin E. J Pediatr, 100 Rees DC, Williams TN, Maitland K, Clegg JB and
129, 105–110. Weatherall DJ (1998) Alpha thalassaemia is associ
90 Harthoorn‐Lasthuizen EJ, Lindemans J and ated with increased soluble transferrin receptor
Langenhuijsen MMAC (1998) Combined use of levels. Br J Haematol, 103, 365–369.
erythrocyte zinc protoporphyrin and mean corpus 101 Winichagoon P, Kumpan P, Holmes P, Finlayson J,
cular volume in differentiation of thalassemia from Newbound C, Kabral A et al. (2015) Validation of
iron deficiency anemia. Eur J Haematol, 60, 245–251. the immunochromatographic strip for α‐thalas
91 Rees DC, Williams TN, Maitland K, Clegg JB and semia screening: a multicenter study. Transl Res,
Weatherall DJ (1998) Alpha thalassaemia is associated 165, 689–695.
with increased soluble transferrin receptor levels. 102 Bunkall C, Ghallyan N, Elliott C, Van de Water N
Br J Haematol, 103, 365–369. and Chan G (2018) Evaluation of an
170 Chapter 3
immunochromatographic strip test for alpha‐tha Guangxi province, southern China: clinical review
lassaemia screening. Int J Lab Hematol, 40, of 357 patients. Acta Haematol, 124, 86–91.
691–696. 113 Tso SC (1972) Red cell survival studies in haemo
103 Nelson AC, Motum PI and Emeto TI (2019) globin H disease using [51Cr]chromate and [32P]
Evaluation of an immunochromatographic test for di‐isopropyl phosphofluoridate. Br J Haematol, 23,
alpha thalassaemia screening in a multi‐ethnic 521–629.
population. Int J Lab Haematol, 41, 397–403. 114 Sheeran C, Weekes K, Shaw J and Pasricha S‐R
104 Ausavarungnirum R, Winichagoon P, Fucharoen (2014) Complications of haemoglobin H disease in
S, Epstein N and Simkins R (1998) Detection of γ‐ adulthood. Br J Haematol, 167, 136–139.
globin chains in cord blood by ELISA (enzyme‐ 115 Lu MY, Kuo MC, Wang S‐C, Chen SH, Ko BS and
linked immunosorbent assay): rapid screening for Hang C‐S (2017) Clinical features and morbidities
α‐thalassemia 1 (Southeast Asian type). Am J of of Hb H disease in Taiwan. Proceedings of 14th
Hematol, 57, 283–286. International Conference on Thalassaemia and
105 Lafferty JD, Barth DS, Sheridan BL, McFarlane other Haemoglobinopathies & 16th TIP
AG, Halchuk LM, Raby A and Crowther MA Conference for Patients and Parents, p. 136.
(2008) A multicenter trial of the effectiveness of ζ‐ 116 Lal A, Goldrich ML, Haines DA, Azimi M, Singer
globin enzyme‐linked immunosorbent assay and ST and Vichinsky EP (2011) Heterogeneity of
hemoglobin H inclusion body screening for the hemoglobin H disease in childhood. N Engl J Med,
detection of alpha0‐thalassemia trait. Am J Clin 364, 710–718.
Pathol, 129, 309–315. 117 Zeinali S, Fallah M‐S and Bagherian H (2011)
106 Pornprasert S, Treesuwan K, Punyamung M and Heterogeneity of haemoglobin H disease in child
Kongthai K (2012) Hb A2/E levels found in co‐ hood. N Engl J Med, 364, 2070–2071.
inheritance with the α‐thalassemia‐1 ‐ ‐SEA/type 118 Ang AL, Le TT and Tan RS (2017) HbH Constant
deletion and either Hb E or β‐thalassemia. Spring disease has lower serum ferritin relative to
Hemoglobin, 36, 381–387. liver iron concentration (LIC): importance of LIC
107 Fairweather RB, Chaffee S, McBride KL, Snow K, measurement and potential impact on serum fer
Kubik KS, Hoyer JD et al. (1999) Hyperunstable ritin thresholds for iron chelation. Br J Haematol,
Hb Dartmouth [α2‐66(E15)Leu→Pro (CTG→CCG) 176, 986–988.
in association with α‐thalassemia‐1 (SE Asian) 119 Lin P‐C, Chang T‐T, Liao Y‐M, Tsai S‐P, Chen YC,
causes transfusion‐dependent α thalassemia. Hsu W‐Y et al. (2019) Clinical features and geno
Blood, 94, Suppl. 1, 24b. types of patients with hemoglobin H disease in
108 Lam TH, Ghosh A, Tang MHY and Chan V (1997) Taiwan. Lab Med, 50, 168–173.
The risk of α‐thalassaemia in offspring of β‐thalas 120 Joly P, Colmant GR, Varennes A, Francina A,
saemia carriers in Hong Kong. Prenatal Diagnosis, Coppéré B and Delacour H (2013) Massive hae
17, 733–736. molysis and methaemalbuminaemia in a patient
109 Ma SK, Chan AYY, Chan LC, Chui DHK and Waye with decompensated haemoglobin H disease. Br J
JS (2000) Compound heterozygosity for triplicated Haematol, 163, 2.
α‐globin gene and (‐‐SEA) α‐globin gene deletion: 121 Ekwattanakit S, Siritanaratkul N and Viprakasit V
implication for thalassaemia screening. Br J (2018) A prospective analysis for prevalence of
Haematol, 110, 498–499. complications in Thai nontransfusion‐dependent
110 Chui GHK, Fucharoen S and Chan V (2003) Hb E/β‐thalassemia and α‐thalassemia (Hb H dis
Hemoglobin H disease is not necessarily a benign ease). Am J Hematol, 93, 623–629.
disorder. Blood, 101, 791–800. 122 Tongsong T, Srisupundit K and Luewan S (2009)
111 Felekis X, Phylactides M, Drousiotou A, Christou Outcomes of pregnancies affected by hemo
S, Kyrri A, Kyriakou K et al. (2008) Hb Agrinio globin H disease. Int Gynaecol Obstet, 104,
[α29(B10)Leu>Pro (α2)] in combination with ‐‐ 206–208
(MED I) results in a severe form of Hb H disease. 123 Origa R, Sollaino MC, Giagu N, Barella S, Campus
Hemoglobin, 32, 237–246. S, Mandas C et al. (2007) Clinical and molecular
112 Yin XL, Zhang XH, Zhou TH, Zhang TL, Luo RG, analysis of haemoglobin H disease in Sardinia:
Wang L et al. (2010) Hemoglobin H disease in haematological, obstetric and cardiac aspects in
patients with different genotypes. Br J Haematol, Stamoulakatou A et al. (2000) Phenotypic and
136, 326–332 molecular diversity of haemoglobin H disease: a
124 Yin XL, Zhang XH, Wu ZK, Zhao DH, Zhou YL, Greek experience. Br J Haematol, 111, 915–923.
Yu YH et al. (2013) Pulmonary hypertension risk in 134 Kurosawa K, Asoh M, Akatsuka A, Matsuo T,
patients with hemoglobin H disease: low inci Ochiai Y and Maekawa K (1996) A Japanese
dence and absence of correlation with splenec patient with X‐linked α‐thalassaemia/mental
tomy. Acta Haematol, 130, 153–159. retardation syndrome: an additional case report.
125 Lorey F, Charoenkwan P, Witkowska HE, Lafferty Jpn J Hum Genet, 41, 329–332.
J, Patterson M, Eng B et al. (2001) Hb H hydrops 135 Kuno T, Ideguchi H, Yoshida N, Masuyama T,
foetalis syndrome: a case report and review of the Ohta M, Nishimura S et al. (1997) A case of X‐
literature and review of the literature. Br J linked α‐thalassaemia/mental retardation syn
Haematol, 115, 72–78. drome: analysis of hemoglobin by an automated
126 Viprakasit V, Green S, Height S, Ayyub H and glycated hemoglobin analyzer. Acta Pediatr Jpn, 39,
Higgs DR (2002) Hb H hydrops fetalis syndrome 615–618.
associated with the interaction of two common 136 Wada T, Nakamura M, Matsushita Y, Yamada M,
determinants of α thalassaemia (‐‐MED/αTSAUDIα). Br Yamashita S, Iwamoto H et al. (1998) Three
J Haematol, 117, 759–762. Japanese children with X‐linked α‐thalassemia/
127 Li DZ, Liao C, Li J, Xie XM, Huang YN, and Wu mental retardation syndrome (ATR‐X) No To
QC (2005) Hemoglobin H hydrops fetalis syn Hattatsu, 30, 283–289.
drome resulting from the association of the – –SEA 137 Pornprasert S, Salaeh NA, Tookjai M, Punyamung
deletion and the αQuong Szeα mutation in a Chinese M, Pongpunyayuen P and Treesuwan K (2018)
woman. Eur J Haematol, 75, 259–261. Hematological analysis in Thai samples with dele
128 Hewes LB, Alvarado CS, Holley LL, Gomez K, tional and nondeletional HbH diseases. Lab Med,
Dykes F, Kutlar A and Kutlar F (2001) Hemoglobin 49, 154–159.
H hydrops fetalis due to compound heterozygo 138 Spencer‐Chapman, Luqmani A, Layton DM and
sity for double α gene deletion and point mutation Bain BJ (2018) Unusual inclusions in hemoglobin
at codon 59, GGC≡GAC (Gly≡Asp) of an α2 gene H disease post‐splenectomy. Am J Hematol, 93,
in an infant of Asian descent. Blood, 98, 22b. 963–964.
129 Henderson S, Chapple M, Rugless M, Fisher C, 139 Sanna G, Frau F, Melis MA, Galanello R, de
Kinsey S and Old J (2006) Haemoglobin H hydrops Virgiliis S and Cao A (1980) Interaction between
fetalis syndrome associated with homozygosity the glucose‐6‐phosphate dehydrogenase defi
for the alpha2‐globin gene polyadenylation signal ciency and thalassaemia genes at phenotype level.
mutation AATAAA‐‐>AATA‐ ‐. Br J Haematol, 135, Br J Haematol, 44, 555–561.
743–745. 140 Viprakasit V, Tanphaichitr VS, Pung‐Amritt P,
130 He Y, Zhao Y, Lou J‐W, Liu Y‐H and Li D‐Z (2016) Petrarat S, Suwantol L, Fisher C and Higgs DR
Fetal anemia and hydrops fetalis associated with (2002) Clinical phenotypes and molecular charac
homozygous Hb Constant Spring (HBA2: terization of Hb H‐Paksé disease. Haematologica,
c.427T > C). Hemoglobin, 40, 97–101. 87, 117–125.
131 Traeger‐Synodinos J, Douna V, Papassotiriou I, 141 Turley E, McFarlane A, Halchuk L and Verhovsek M
Stamoulakatou A, Ladis V, Siahanidou T et al. (2015) Hemoglobin H identification by high‐perfor
(2010) Variable and often severe phenotypic mance liquid chromatography in confirmed hemo
expression in patients with the α‐thalassemic vari globin H disease. Int J Lab Hematol, 37, 668–672.
ant Hb Agrinio [α29(B10)Leu→Pro (α2)]. 142 Liao C, Zhou JY, Xie XM, Tang HS, Li R and Li DZ
Hemoglobin, 34, 430–438. (2014) Newborn screening for Hb H disease by
132 Chen FE, Ooi C, Ha SY, Cheung BMY, Todd D, determination of Hb Bart’s using the Sebia capil
Liang R et al. (2000) Genetic and clinical features of lary electrophoresis system in southern China.
hemoglobin H disease in Chinese patients. N Engl Hemoglobin, 38, 73–75.
J Med, 343, 544–550. 143 O’Brien RT (1973) The effect of iron deficiency on
133 Kanavakis E, Papassotirou I, Karagiorga M, the expression of hemoglobin H. Blood, 41,
Vrettou C, Metaxotou‐Mavrommati A, 853–856.
172 Chapter 3
144 Saleem A, Irani DR, Bart JB and Alfrey CP (1987) hypospadias in three survivors. Am J Med Genet,
Suppression of hemoglobin H in disorders of iron 82, 225–227.
metabolism. Acta Haematol, 77, 34–37. 156 Goulden NJ, Kyle P, Souhail K and Pamphilon DH
145 Fucharoen S, Winichagoon P, Siritanaratkul N, (2002) Normal developmental outcome in a 2‐
Chowthaworn JEW and Pootrakul P (1998) α‐ and year‐old child born after successful salvage of
β‐thalassemia in Thailand. An NY Acad Sci, 850, haemoglobin Barts hydrops foetalis syndrome
412–414. with multiple in utero transfusions. Br J Haematol,
146 Michlitsch J, Azimi M, Hoppe C, Walters MC, 117, Suppl. 1, 12.
Lubbin B, Lorey F and Vichinsky E (2012) 157 Amid A, Chen S, Brien W, Kirby‐Allen M and
Newborn screening for hemoglobinopathes in Odame I (2016) Optimizing chronic transfusion
California. Pediatr Blood Cancer, 52, 486–490. therapy for survivors of hemoglobin Barts
147 Ma ESK, Chan AYY, Au WY, Yeung YM and Chan hydrops fetalis. Blood, 127, 1208–1211.
LC (2001) Diagnosis of concurrent hemoglobin H 158 Au PKC, Kan ASY, Tang MHY, Leung KY, Chan
disease and heterozygous β‐thalassemia. KYK, Tang TWF and Lau ET (2016) A fetus with
Haematologica, 86, 432–433. Hb Bart’s disease due to maternal uniparental dis
148 Yin X‐L, Wu Z‐K, Zhou XY, Zhou T‐H, Zhou Y‐L, omy for chromosome 16. Hemoglobin, 40, 60–69.
Wang L et al. (2012) Co‐inherited β‐thalassemia 159 Chan V, Chan W‐Y, Tang M, Lau K, Todd D and
trait and HbH disease: clinical characteristics and Chan TK (1997) Molecular defects in haemoglobin
interference in diagnosis of thalassemia by high‐ H hydrops. Br J Haematol, 96, 224–226.
performance liquid chromatography. Int J Lab 160 Amid A, Chen S, Athale U, Charpentier K,
Hematol, 34, 427–431. Merelles‐Pulcini M, Odame I and Kirby‐Allen M
149 Chen GL, Jiang F, Li J, Zhou JY and Li DZ (2018) (2018) Iron overload in transfusion‐dependent
Results of coexistence of β‐thalassemia minor in survivors of hemoglobin Bart’s hydrops fetalis.
Hb H disease patients. Hemoglobin, 42, Haematologica, 103, e184–e187.
306–309. 161 Hall GW, Barnetson RA and Thein SL (1992) Beta
150 Traeger‐Synodinos J, Papassotiriou I, Vrettou C, thalassaemia in the indigenous British population,
Skarmoutsou C, Stamoulakatou A and Kanavakis Br J Haematol, 82, 584–588.
R (2001) Erythroid marrow activity and functional 162 Fucharoen S, Sanchaisuriya K, Fucharoen G and
anemia in patients with the rare interaction of a Surapot S (2001) Molecular characterization of
single functional α‐globin and β‐globin gene. thalassemia intermedia with homozygous Hb
Haematologica, 86, 363–367. Malay and Hb Malay/HbE in Thai patients.
151 Kanavakis E, Traeger‐Synodinos J, Lafioniatis S, Haematologica, 86, 657–658.
Lazaropoulou C, Liakopoulou T, Paleologos G 163 Smith CM, Hedlund B, Cich JA, Tukey DP, Olson
et al. (2004) A rare example that coinheritance of a M, Steinberg MH and Admas JG (1983)
severe form of β‐thalassemia and α‐thalassemia Hemoglobin North Shore: a variant hemoglobin
interact in a “synergistic” manner to balance the associated with the phenotype of β‐thalassemia.
phenotype of classic thalassemic syndromes. Blood Blood, 61, 378–383.
Cells Mol Dis, 32, 319–324. 164 Daar S, Gravell D, Hussein HM, Pathare AV, Wali
152 Fucharoen S and Fucharoen G (2012) Hb H dis Y and Krishnamoorthy R (2008) Haematological
ease with various β hemoglobinopathies: molecu and clinical features of β‐thalassaemia associated
lar, hematological and diagnostic aspects. with Hb Dhofar. Eur J Haematol, 80, 67–70.
Hemoglobin, 36, 18–24. 165 Russell JE and Liebhaber SA (1996) The stability of
153 Lie‐Injo LE and Jo BH (1960) A fast moving hae human β‐globin mRNA is dependent on structural
moglobin in hydrops foetalis. Nature, 185, 698. determinants positioned within its 3′ untranslated
154 Songdej D, Babbs C and Higgs DR; BHFS region. Blood, 87, 5314–5322.
International Consortium (2017) An international 166 Hopmeier P, Krugluger W, Gu L‐H, Smetanina NS
registry of survivors with Hb Bart’s hydrops feta and Huisman THJ (1996) A newly discovered
lis syndrome. Blood, 129, 1251–1259. frameshift at codons 120‐121 (+A) of the β gene is
155 Fung TY, Kin LT, Kong LC and Keung LC (1999) not associated with a dominant form of β‐thalas
Homozygous alpha‐thalassaemia associated with saemia. Blood, 87, 5393–5394.
167 Thein SL (1998) β‐thalassaemia. Bailliere’s Clin (A→G) acceptor splice site mutation. Br J Haematol,
Haematol, 11, 91–126. 103, 370–376.
168 Cabeda JM, Correia C, Estevinho A, Simöes C, 178 Premawardhena A, Arambepola M, Katugaha N
Amorim ML, Pinho L and Justiça B (1999) and Weatherall DJ (2008) Is the β thalassaemia trait
Unexpected pattern of β‐globin gene mutations in of clinical importance? Br J Haematol, 141,
β‐thalassaemia patients from Northern Portugal, 407–410.
Br J Haematol, 105, 68–74. 179 Graffeo L, Vitrano A, Giambona A, Scondotto S,
169 Ho PJ and Thein SL (2000) Gene regulation and Dardanoni G, Gluud C and Maggio A (2017) The
deregulation: a β globin perspective. Blood Rev, 14, heterozygote state for β‐thalassemia detrimentally
78–93. affects health outcomes. Am J Hematol, 92,
170 Rashkind WH, Niakan KK, Wolff J, Matsushita M, E23–E25.
Vaughan T, Stamatoyannopoulos G et al. (1999) 180 Borgna‐Pignatti C, Rigon F, Merlo L, Chakrok R,
Mapping of a syndrome of X‐linked thrombocyto Micciolo R, Perseu L and Galanello R (2003)
penia with thalassemia to band Xp11‐12: further Thalassemia minor, the Gilbert mutation, and the
evidence of genetic heterogeneity. Blood, 95, risk of gallstones. Haematologica, 88, 1106–1109.
2262–2268. 181 Cetin T, Oktenli C, Ozgurtas T, Yenicesu M,
171 Weatherall D and Clegg JB. The Thalassaemia Sanisoglu SY, Oguz Y et al. (2003) Renal tubular
Syndromes, 4th edn. Blackwell Science, Oxford, dysfunction in beta‐thalassemia minor. Am J
2001. Kidney Dis, 42, 1164–1168.
172 Yu C, Niakan KK, Martsushita M, Stamatoy 182 Montecucco C, Caporali R, Rossi S and Epis O
annopoulos G, Orkin SH and Raskind WH (2002) (1999) Rheumatoid arthritis in beta‐thalassaemia
X‐linked thrombocytopenia with thalassaemia trait. Rheumatology, 38, 1021–1022.
from a mutation in the amino finger of GATA‐1 183 Ibba RM, Zoppi MA, Floris M, Manca F, Axiana C,
affecting DNA binding rather than FOG‐1 interaction. Cao A and Monni G (2003) Neural tube defects in
Blood, 100, 2040–2045. the offspring of thalassemia carriers. Fetal Diagn
173 Thein SL and Wood WG. The molecular basis of β Ther, 18, 5–7.
thalassemia, δβ thalassaemia, and hereditary per 184 Karimi M, Bagheri MH, Tahmtan M, Shakibafard
sistence of fetal hemoglobin. In: Steinberg MH, A and Rashid M (2007) Prevalence of hepatosple
Forget BG, Higgs DR and Nagel RL, eds. Disorders nomegaly in beta thalassemia minor subjects in
of Hemoglobin: Genetics, Pathophysiology, and Clinical Iran. Eur J Radiol, 69, 120–122.
Management, 2nd edn. Cambridge University Press, 185 Hoorfar H, Sadrarhami S, Keshteli AH, Ardestani
Cambridge, 2009, pp. 323–356. SK, Ataei M and Moafi A (2008) Evaluation of iron
174 Su Q, Chen S, Wu L, Tian R, Yang X, Huang X et al. status by serum ferritin level in Iranian carriers of
(2019) Severe thalassemia caused by Hb Zunyi beta thalassemia minor. Int J Vitam Nutr Res, 78,
[β147(HC3)Stop→Gln; HBB: c.442T>C)] on the β‐ 204–207.
globin gene. Hemoglobin, 43, 7–11. 186 Van der Weyden MB, Fong H, Hallam LJ and
175 Castro O, Kutlar F, Kutlar A, Holley L, Hassan S Harrison C (1989) Red cell ferritin and iron over
and Elshowaia S (2001) Sβ0 thalassemia in a 91‐ load in heterozygous beta‐thalassemia. Am J
year‐old woman with mild sickle cell disease with Hematol, 30, 201–205.
two thalassemia mutations on the same chromo 187 Piperno A, Mariani R, Arosio C, Vergani A, Bosio
some. Blood, 98, 14b–15b. S, Fargion S et al. (2000) Haemochromatosis in
176 Murru S, Loudianos G, Porcu S, Sciarratta GV, patients with beta‐thalassaemia trait. Br J Haematol,
Agosti S, Parodi MI et al. (1992) A β‐thalassaemia 111, 908–914.
phenotype not linked to the β‐globin cluster in an 188 Riva A, Mariani R, Bovo G, Pelucchi S, Arosio C,
Italian family. Br J Haematol, 81, 283–287. Salvioni A et al. (2004) Type 3 hemochromatosis
177 Giordano PC, Harteveld CL, Haak HL, Batelaan and β‐thalassemia trait. Eur J Haematol, 72,
D, van Delft P, Plug RJ et al. (1998) A case of non‐β‐ 370–374.
globin gene linked β thalassaemia in a Dutch fam 189 Sartori M, Andorno S, Pagliarulo M, Rigamonti C,
ily with two additional α‐gene defects: the Bozzola C, Pergolini P et al. (2007) Heterozygous
common ‐α3.7 deletion and the rare IVS1‐116 beta‐globin gene mutations as a risk factor for
174 Chapter 3
iron accumulation and liver fibrosis in chronic 200 Hill AV, Bowden DK, O’Shaughnessy DF,
hepatitis C. Gut, 56, 693–698. Weatherall DJ and Clegg JB (1988) Beta thalas
190 Valenti L, Canavesi E, Galmozzi E, Dongiovanni P, semia in Melanesia: association with malaria and
Rametta R, Maggioni P et al. (2010) Beta‐globin characterization of a common variant (IVS‐1 nt 5
mutations are associated with parenchymal G‐‐‐‐C). Blood, 72, 9–14.
siderosis and fibrosis in patients with non‐alco 201 Millard DP, Mason K, Serjeant BE and Serjeant GR
holic fatty liver disease. J Hepatol, 53, 927–933. (1977) Comparison of haematological features in
191 Crowley JP, Sheth S, Capone RJ and Schilling RF β0 and β+ thalassaemia traits in Jamaican Negroes.
(1987) A paucity of thalassemia trait in Italian men Br J Haematol, 36, 161–170.
with myocardial infarction. Acta Haematol, 78, 202 Rund D, Filon D, Strauss N, Rachmilewitz EA and
249–251. Oppenheim A (1992) Mean corpuscular volume of
192 Gallerani M, Scapoli C, Cicognani I, Ricci A, heterozygotes for β‐thalassemia correlates with
Martinelli L, Cappato R et al. (1991) Thalassaemia the severity of mutations. Blood, 79, 238–243.
trait and myocardial infarction: low infarction 203 Pootrakul P, Wasi P and Na‐Nakorn S (1973)
incidence in male subjects confirmed. J Intern Med, Haematological data in 312 cases of β‐thalassae
230, 109–111. mia trait in Thailand. Br J Haematol, 24, 703–712.
193 Bozdar M, Ahmed S and Anwar J (2010) Relative 204 Mazza U, Saglio G, Cappio FC, Camaschella C,
protection from ischaemic heart disease in beta‐ Neretto G and Gallo E (1976) Clinical and haema
thalassaemia carriers. J Coll Physicians Surg Pak, tological data in 254 cases of beta‐thalassaemia
20, 653–656. trait in Italy. Br J Haematol, 33, 91–99.
194 Tassiopoulos S, Deftereos S, Konstantopoulos K, 205 Piomelli S and Siniscalco M (1969) The haemato
Farmakis D, Tsironi M, Kyriakidis M et al. (2005) logical effects of glucose‐6‐phosphate dehydro
Does heterozygous beta‐thalassemia confer a pro genase deficiency and thalassaemia trait:
tection against coronary artery disease? Ann NY interaction between the two genes at the pheno
Acad Sci, 1054, 467–470. type level. Br J Haematol, 16, 537–549.
195 Hashemi M, Shirzadi E, Talaei Z, Moghadas L, 206 Olivieri NF and Weatherall DJ, Clinical aspects of
Shaygannia I, Yavari M et al. (2007) Effect of hete beta thalassemia. In: Steinberg MH, Forget BG,
rozygous beta‐thalassaemia trait on coronary ath Higgs DR and Nagel RL, eds. Disorders of
erosclerosis via coronary artery disease risk Hemoglobin: Genetics, Pathophysiology, and Clinical
factors: a preliminary study. Cardiovasc J Afr, 18, Management. Cambridge University Press,
165–168. Cambridge, 2001, pp. 342–355.
196 Karimi M, Borhani Haghighi A, Yazdani M, Raisi 207 Lewis D, Stockley RJ and Chanarin I (1982)
H, Giti R and Namazee MR (2008) Is beta‐thalas Changes in mean corpuscular red cell volume in
semia trait a protective factor against ischemic cer women with β‐thalassaemia trait during preg
ebrovascular accidents? J Stroke Cerebrovasc Dis, nancy. Br J Haematol, 50, 423–425.
17, 79–81. 208 Vicinanza P, Vicinanza M, Cosimato V, Terracciano
197 Vanini A, Salmi R, Manfredini R, Gallerani M, D, Cancellario S, Massari A and Danise P (2017)
Zappi M and Malacarne P (1993) Is heterozygous Mean reticolocyte hemoglobin content index
beta‐thalassemia a protective factor for peripheral plays a key role to identify children who are carri
arteriopathic disease?: A preliminary ultrasono ers of β‐thalassemia. Trans Med, 17, 31–36.
graphic study. Curr Ther Res, 54, 540–543. 209 Mentzer WC (1973) Differentiation of iron defi
198 Maioli M, Pettinato S, Cherchi GM, Giraudi D, ciency from thalassaemia trait. Lancet, i, 882.
Pacifico A, Pupita G and Tidore MG (1989) Plasma 210 England JM and Fraser PM (1973) Differentiation
lipids in beta‐thalassemia minor. Atherosclerosis, of iron deficiency from thalassaemia trait by rou
75, 245–248. tine blood count, Lancet, i, 449–452.
199 Calandra S, Bertolini S, Pes GM, Deiana L, Tarugi 211 Shine I and Lal S (1977) A strategy to detect β tha
P, Pisciotta L et al. (2004) Beta‐thalassemia is a lassaemia trait, Lancet, i, 692–694.
modifying factor of the clinical expression of 212 Srivastava PC (1973) Differentiation of thalassae
familial hypercholesterolemia. Semin Vasc Med, 4, mia minor from iron deficiency, Lancet, ii,
271–278. 154–155.
213 Ricerca BM, Storti S, d’Onofrio G, Mancini S, 225 Bain BJ, Amos RJ, Bareford D, Chapman C, Davies
Vittori M, Campisi S et al. (1987) Differentiation of SC, Old JM and Wild BL (Working Party of the
iron deficiency from thalassaemia trait: a new General Haematology Task Force of the British
approach. Haematologica, 72, 409–413. Committee for Standards in Haematology) (1998)
214 Green R, and King RA (1989) A new red cell discri The laboratory diagnosis of haemoglobinopathies.
minant incorporating volume dispersion for dif Br J Haematol, 101, 783–792.
ferentiating iron deficiency anemia from 226 Borbely N, Phelan L, Szydlo R and Bain B (2013)
thalassaemia minor, Blood Cells, 15, 481–485. Capillary zone electrophoresis for haemoglobi
215 D’Onofrio G, Zini G, Ricerca BM, Mancini S and nopathy diagnosis. J Clin Pathol, 66, 29–39.
Mango G (1992) Automated measurement of red 227 Gasperini D, Cao A, Paderi L, Barella S, Paglietti E,
blood cell microcytosis and hypochromia in iron Perseu L et al. (1993) Normal individuals with
deficiency and β‐thalassaemia trait. Arch Pathol high Hb A2 levels. Br J Haematol, 84, 166–168.
Lab Med, 116, 84–89. 228 Nagel RL and Steinberg MH, Hemoglobins of the
216 Jiminez CV, Minchinela J and Ros J (1995) New embryo and fetus and minor hemoglobins of
indices from the H*2 analyser improve differentia adults. In: Steinberg MH, Forget BG, Higgs DR
tion between heterozygous β and δβ thalassaemia and Nagel RL, eds. Disorders of Hemoglobin:
and iron deficiency anaemia. Clin Lab Haematol, Genetics, Pathophysiology, and Clinical Management,
17, 151–155. Cambridge University Press, Cambridge, 2001,
217 Eldibany MM, Totonchi KF, Joseph NJ and Rhone pp. 197–230.
D (1999) Usefulness of certain red blood cell indi 229 Jiang F, Chen GL, Li J, Zhou JY, Liao C and Li DZ
ces in diagnosing and differentiating thalassaemia (2018) Analysis of the genotypes in a Chinese pop
trait from iron‐deficiency anemia. Am J Clin Pathol, ulation with increased Hb A2 and low hematologi
111, 676–682 and Eldibany MM (1999) Correction, cal indices. Hemoglobin, 42, 161–165.
Am J Clin Pathol, 112, 428. 230 Perseu L, Satta S, Moi P, Demartis FR, Manunza L,
218 Lafferty JD, Crowther MA, Ali MA and Levine M Sollaino MC et al. (2011) KLF1 gene mutations
(1996) The evaluation of various mathematical cause borderline HbA2. Blood, 118, 4454–4458.
RBC indices and their efficacy in discriminating 231 Liu D, Zhang X, Yu L, Cai R, Ma X, Zheng C et al.
between thalassemic and non‐thalassemic micro (2014) KLF1 mutations are relatively more com
cytosis. Am J Clin Pathol, 106, 201–205. mon in a thalassemia endemic region and amelio
219 Bain BJ. Haemoglobinopathy Diagnosis, 1st edn. rate the severity of β‐thalassemia. Blood, 124,
Blackwell Publishing, Oxford, 2001. 803–811.
220 Malamos B, Belcher EH, Gyftaki E and Binopoulos 232 Stephens AD, Angastiniotis M, Baysal E, Chan V,
D (1961) Simultaneous radioactive tracer studies Fucharoen S, Giordano PC et al; International
of erythropoiesis and red‐cell destruction in tha Council for the Standardisation of Haematology
lassaemia. Br J Haematol, 7, 411–429. (ICSH) (2012) ICSH recommendations for the
221 Rees DC, Duley DA and Marinaki AM (2003) measurement of haemoglobin A2. Int J Lab
Pyrimidine 5′ nucleotidase deficiency. Br J Hematol, 34, 1–13.
Haematol, 120, 375–383. 233 Huisman THJ (1977) Trimodality in the percent
222 Bernini L, Latte B, SiniscalcoM, Piomelli S, Spada ages of β chain variants in heterozygotes: the effect
U, Adinolfi M and Mollison PL (1964) Survival of of the number of active Hbα structural loci.
51
Cr‐labelled red cells in subjects with thalassae Hemoglobin, 1, 349–382.
mia‐trait or G6PD deficiency or both abnormali 234 Parmeggiani G, Gualandi F, Selvatici R, Rimessi P,
ties. Br J Haematol, 10, 171–180. Bigoni S, Taddei Masieri M et al. (2016) A family
223 Wickramasinghe SN and Hughes M (1980) with γ‐thalassemia and high Hb A2 levels.
Ultrastructural studies of erythropoiesis in β‐tha Hemoglobin, 40, 187–190.
lassaemia trait. Br J Haematol, 46, 401–407. 235 So CC, Chan AY, Tsang ST, Lee AC, Au WY, Ma ES
224 Globin Gene Disorder Working Party of the BCSH and Chan LC (2007) A novel beta‐delta globin gene
General Haematology Task Force (1994) fusion, anti‐Lepore Hong Kong, leads to overex
Guidelines for the fetal diagnosis of globin gene pression of delta globin chain and a mild thalas
disorders. J Clin Pathol, 47, 199–204. saemia intermedia phenotype when co‐inherited
176 Chapter 3
with beta(0)‐thalassaemia. Br J Haematol, 136, 248 Huang L‐Y, Li J, Zhang Y and Li D‐Z (2018) A
158–162. KLF1 gene mutation causes β‐thalassemia minor
236 Tamary H, Shalev H, Luria D, Shaft D, Zoldan M, in a Chinese family. Int J Lab Hematol, 40, e35–e37.
Shalmon L et al. (1996) Clinical features and stud 249 Tamaddoni A, Khabaz Astaneh S, Tabaripour R
ies of erythropoiesis in Israeli Bedouins with con and Akhavan‐Niaki H (2019) Krüppel‐Like Factor 1
genital dyserythropoietic anaemia type I. Blood, gene mutations in thalassemia patients from
87, 1763–1770. North Iran: report of a new mutation associated
237 Harmerling JG and Moquin RB (1967) An abnor with β‐thalassemia intermedia. Hemoglobin, 43,
mal elevation of haemoglobin A2 in hereditary 12–17.
spherocytosis, Am J Clin Pathol, 47, 454–458. 250 Serjeant BE, Mason KP and Serjeant GR (1978) The
238 Wilson MG, Schroeder WA, Graves DA and Kach development of haemoglobin A2 in normal Negro
VD (1968) Postnatal changes of haemoglobins F infants and in sickle cell disease. Br J Haematol, 39,
and A2 in infants with Down’s syndrome (G tri 259–263.
somy). Pediatrics, 42, 349–352. 251 Mosca A Paleari R, Ivaldi G, Galanello R and
239 Martin L, Pissard S, Blanc P, Chassaing N, Legac E, Giordano PC (2009) The role of haemoglobin A2
Briault S et al. (2006) Augmentation de testing in the diagnosis of thalassaemias and
l’hémoglobine A2 au cours du pseudoxanthome related haemoglobinopathies. J Clin Pathol, 62,
élastique. Ann Dermatol Venereol, 133, 645–651. 13–17.
240 Van Kirk R, Sandhaus LM and Hoyer J (2004) 252 Maragoudaki E, Vrettou C, Kanavakis E, Traeger‐
Diagnosis of hemoglobin (Hb) A2′ trait by high‐ Synodinos J, Metaxotou‐Mavrommati A and
performance liquid chromatography (HPLC). Am Kattamis C (1998) Molecular, haematological and
J Clin Pathol, 122, 630. clinical studies of a silent β‐gene C→G mutation at
241 Daar S, Al Zadjali S, Gravell D, Al Haddabi H, Al 6 bp 3′ to the termination codon (+1480 C→G) in
Riyami M, Al Belushi A et al. (2011) Hb A2′ (Hb B2) twelve Greek families. Br J Haematol, 103, 45–51.
in the Omani population and diagnostic signifi 253 Maragoudaki E, Kanavakis E, Traeger‐Synodinos
cance. Hemoglobin, 35, 117–124. J, Vrettou C, Tzetis M, Metaxotou‐Mavrommati A
242 Chen G‐L, Huang L‐Y, Zhou J‐Y and Li D‐Z (2017) and Kattamis C (1999) Molecular, haematological
Hb A2‐Tianhe (HBD: c.323G>A): first report in a and clinical studies of the –101 C→T substitution
Chinese family with normal Hb A2‐β‐thalassemia of the β‐globin gene promoter in 25 β‐thalassaemia
trait. Hemoglobin, 41, 291–292. intermedia patients and 45 heterozygotes. Br J
243 Gilman JG, Huisman TH and Abels J (1984) Dutch Haematol, 107, 699–706.
beta0‐thalassaemia: a 10 kilobase DNA deletion 254 Athanassiado A, Papchatzopoulou A, Zoumbos
associated with significant gamma‐chain produc N, Maniatis GM and Gibbs R (1994) A novel β‐tha
tion. Br J Haematol, 56, 339–348. lassaemia mutation in the 5′ untranslated region
244 Rogers M, Phelan Land Bain B (1995) Screening of the β‐globin gene. Br J Haematol, 88, 307–310.
criteria for β thalassaemia trait in pregnant 255 Ho PJ, Rochette J, Fisher CA, Wonke B, Jarvis MK,
women. J Clin Pathol, 48, 1054–1056. Yardumian A and Thein SL (1996) Moderate
245 Karimi M and Rasekhi AR (2002) Efficiency of pre reduction of β‐globin gene transcript by a novel
marital screening of beta‐thalassemia trait using mutation in the 5′ untranslated region: a study of
MCH rather than MCV in the population of Fars its interaction with other genotypes in two fami
Province, Iran. Haematologia, 32, 129–133. lies. Blood, 87, 1170–1178.
246 Penman BS, Gupta S and Weatherall DJ (2015) 256 Öner C, Agarwal S, Dimovski AJ, Efremov GD,
Epistasis and the sensitivity of phenotypic Petkov GH, Altay C et al. (1991) The G‐‐‐‐A mutation
screens for beta thalassaemia. Br J Haematol, 169, at position +22 3′ to the Cap site of the beta‐globin
117–128. gene and a possible cause for a beta‐thalassaemia.
247 Huisman THJ (1997) Levels of Hb A2 in heterozy Hemoglobin, 15, 67–76.
gotes and homozygotes for beta‐thalassaemia 257 Scerri CA, Abela W, Galdies R, Pizzuto M, Grech
mutations: influence of mutations in the CACCC JL and Felice AE (1993) The β+ IVS, I‐NT no. 6
and ATAAA motifs of the beta‐globin promoter. (T→C) thalassaemia in heterozygotes with an
Acta Haematol, 98, 187–194. associated Hb Valetta or Hb S heterozygosity in
homozygotes from Malta. Br J Haematol, 83, 270 Saleh‐Gohari N, Khademi Bami M, Nikbakht R and
669–671. Karimi‐Maleh H (2015) Effects of α‐thalassaemia
258 Old J (2003) Screening and genetic diagnosis of mutations on the haematological parameters of β‐
haemoglobin disorders. Blood Rev, 17, 43–54. thalassaemia carriers. J Clin Pathol, 68, 562–566.
259 Baklouiti F, Dorléac E, Morlé L, Laselve P, 271 Liang R, Liang S, Jiang NH, Wen X‐J, Zhao J‐B,
Peyramond D, Aubry M et al. (1986) Homozygous Nechtman JF et al. (1994) α and β thalassaemia
hemoglobin Knossos (alpha 2 beta 227(B9) among Chinese children in Guangxi Province, P.R.
Ala→Ser): a new variety of beta+‐thalassemia China: molecular and haematological characteri
intermedia associated with delta0‐thalassemia. zation. Br J Haematol, 86, 351–354.
Blood, 67, 957–961. 272 Giambona A, Passarello C, Vinciguerra M, Li Muli
260 Pagano L, Desicato S, Viola A, de Rosa C and R, Teresi P et al. (2008) Significance of borderline
Floretti G (1995) Identification of the –92 (C→T) hemoglobin A2 values in an Italian population
mutation by the amplification refractory mutation with a high prevalence of beta‐thalassemia.
system in southern Italy. Hemoglobin, 19, 307–310. Haematologica, 93, 1380–1384.
261 Wasi P, Disthasongchan P and Na‐Nakorn S (1968) 273 Galanello R, Barella S, Ideo A, Gasperini D,
The effect of iron deficiency on the levels of hae Rosatelli C, Paderi L et al. (1994) Genotype of sub
moglobin A2. J Lab Clin Med, 71, 85–91. jects with borderline hemoglobin A2 levels: impli
262 Alperin JB, Dow PA and Petteway MB (1972) cation for beta‐thalassemia carrier screening. Am J
Haemoglobin A2 levels in health and various hae Hematol, 46, 79–81.
matological disorders. Am J Clin Pathol, 67, 274 Kanavakis E, Wainscoat JS, Wood WG, Weatherall
219–226. DJ, Cao A, Furbetta M et al. (1982) The interaction
263 Kattamis C, Lagos P, Metaxotou‐Mavromati A and of α thalassaemia with heterozygous β thalassae
Matsaniotis NJ (1972) Serum iron and unsaturated mia. Br J Haematol, 52, 465–473.
iron‐binding‐capacity in β thalassaemia trait: their 275 Gasperini D, Cao A, Paderi L, Barella S, Paglietti E,
relation to the levels of haemoglobins A, A2 and F. Perseu L et al. (1993) Normal individuals with
J Med Genet, 9, 154–159. high Hb A2 levels. Br J Haematol, 84, 166–168.
264 Galanello R, Ruggeri R, Addis M, Paglietti E and 276 Wang C and Amato D (2007) Chronic and severe
Cao A (1981) Hemoglobin A2 in iron‐deficient β‐ haemolytic anaemia caused by co‐inheritance of
thalassaemia heterozygotes. Hemoglobin, 5, β‐thalassaemia and triplicated α‐globin genes. Br J
613–618. Haematol, 137, 489.
265 Alli NA and Mendelow BV (1998) Distinguishing 277 Liu S, Huang LY, Jiang F, Sun X‐F and Li D‐Z
β‐thalassaemia carriers from normal individuals (2018) Complex interactions between thalassemia
‐ a comparison of two methods. Br J Haematol, defective alleles compromise screening and cause
102, 179. severe anemia in a Chinese family. Int J Lab
266 El‐Agouza I, Shahla AA and Sirdah M (2002) The Hematol, 40, e55–e58.
effect of iron deficiency anaemia on the levels of 278 Garner C, Dew TK, Sherwood R, Rees D and
haemoglobin subtypes: possible consequences for Thein SL (2003) Heterocellular hereditary persis
clinical diagnosis. Clin Lab Haematol, 24, 285–289. tence of fetal haemoglobin affects the haemato
267 Sharma P, Das R, Trehan A, Bansal D, Chhabra S, logical parameters of β‐thalassaemia trait. Br J
Kaur J et al. (2015) Impact of iron deficiency on Haematol, 123, 353–358.
hemoglobin A2% in obligate β‐thalassemia hete 279 Dacie JV. The Haemolytic Anaemias, Volume 2, The
rozygotes. Int J Lab Hematol, 37, 105–111. Hereditary Haemolytic Anaemias, 3rd edn. Churchill
268 Guiso L, Frogheri L, Pistidda P, Angioni L, Dore F, Livingstone, Edinburgh, 1988.
Pardini S and Longinotti M (1996) Frequency of δ+ 280 Maehara T, Tsuukamoto N, Nijima Y, Karasawa
27‐thalassaemia in Sardinians. Clin Lab Haematol, M, Murakami H, Hattori Y and Ideguchi H (2002)
18, 241–244. Enhanced haemolysis with β‐thalassaemia trait
269 Yan JM, Wu MY, Xie XM and Li DZ (2015) due to unstable β chain variant, Hb Gunma,
Heterozygous β‐thalassemia with complete accompanied by hereditary elliptocytosis due to
absence of hemoglobin A2 in a Chinese adult. Int J protein 4.1 deficiency in a Japanese family. Br J
Lab Hematol, 37, e147–e149. Haematol, 117,193–197.
178 Chapter 3
281 Fucharoen G, Fucharoen S, Singsanan S and associated with a β‐thalassemia minor phenotype.
Sanchaisuriya K (2007) Coexistence of Southeast Haemoglobin, 32, 446–453.
Asian ovalocytosis and beta‐thalassemia: a molec 293 Nussenzveig RH, Vanhille DL, Hussey D, Reading
ular and hematological analysis. Am J Hematol, 82, NS and Agarwal AM (2012) Development of a
381–385. rapid multiplex PCR assay for identification of the
282 Croteau SE, Luo H‐Y, Lehmann LE, Chui DHK three common Hemoglobin‐Lepore variants
and Neufeld EJ (2013) Novel dominant β‐thalas (Boston‐Washington, Baltimore, and Hollandia)
semia: Hb Boston‐Kuwait [Codon 139/140(þT)]. and identification of a new Lepore variant. Am J
Pediatr Blood Cancer, 60, E131–E134. Hematol, 87, E74–E75.
283 Farashi S, Rad F, Shahmohammadi B, Imanian H, 294 Adams JG, Morrison WT and Steinberg MH (1982)
Azarkeivan A and Najmabadi H (2016) First report Hemoglobin Parchman: double crossover within a
of a dominantly inherited β‐thalassemia caused by single human gene. Science, 218, 291–293.
a novel elongated β‐globin chain. Hemoglobin, 40, 295 Nieto JM, González FA, Alonso JM, Golvano E,
102–107. Guerrero L, Albarrán B et al. (2019) Hb Palencia: a
284 Thein SL (1999) Is it dominantly inherited β thalas novel δβδ‐type two‐way fusion variant with β‐glo
saemia or just a β‐chain variant that is highly bin‐like expression levels. J Clin Pathol, 72, 46–51.
unstable? Br J Haematol, 107, 12–21. 296 Ropero P, Gonzalez FA, Sanchez J, Anguita E,
285 Ho PJ, Wickramasinghe S, Rees DC, Lee MJ, Eden Asenjo S, Del Arco A et al. (1999) Identification of
A and Thein SL (1997) Erythroblastic inclusions in the Hb Lepore phenotype by HPLC. Haematologica,
dominantly inherited β thalassaemia trait. Blood, 84, 1081–1084.
89, 322–328. 297 Ma SK and Au WY (1999) Images in haematology:
286 Faustino P, Os¢rio‐Almeida L, Romao L, Barbot J, hepatic haemopoiesis in β‐thalassaemia trait. Br J
Fernandes B, Justica B and Lavinha J (1998) Haematol, 107, 1.
Dominantly transmitted β‐thalassaemia arising 298 Aessopos A, Farmakis D, Karagiorga M,
from the production of several aberrant mRNA Voskaridou E, Loutradi A, Hatziliami A et al (2001)
species and one abnormal peptide. Blood, 91, Cardiac involvement in thalassaemia intermedia:
685–690. a multicenter study. Blood, 97, 3411–3416.
287 Thein SL, Hesketh C, Taylor P, Temperley IJ, 299 Musallam KM, Taher AT, Karimi M and
Hutchinson RM, Old JM et al. (1990) Molecular Rachmilewitz EA (2012) Cerebral infarction in β‐
basis of dominantly inherited inclusion body β‐ thalassemia intermedia: breaking the silence.
thalassemia. Proc Natl Acad Sci U S A 87, Thromb Res, 130, 695–702.
3924–3928. 300 Ziyadeh FN, Musallam KM, Mallat NS, Mallat S,
288 Thein SL (1992) Dominant β thalassaemia: molec Jaber F, Mohamed AA et al. (2012) Glomerular
ular basis and pathophysiology. Br J Haematol, 80, hyperfiltration and proteinuria in transfusion‐
273–277. independent patients with β‐thalassemia interme
289 Galanello R and Cao A (1998) Relationship dia. Nephron Clin Pract, 121, c136–c143.
between genotype and phenotype: thalassemia 301 Mallat NS, Wehbe D, Haddad A, Cappellini MD,
intermedia. Ann NY Acad Sci, 850, 325–333. Marcon A, Koussa S et al. (2014) Priapism, an
290 Vetter B, Neu‐Yilik G, Kohne E, Arnold R, Sinha P, emerging complication in β‐thalassemia interme
Gaedicke G et al. (2000) Dominant β thalassaemia; dia patients. Hemoglobin, 38, 351–354.
a highly unstable haemoglobin is caused by a 302 Vassilopoulos G, Papassotiriou I, Voskaridou E,
novel 6 bp deletion of the β‐globin gene. Br J Stamoulakatou A, Premetis E, Kister J et al. (1995)
Haematol, 108, 176–181. Hb Arta [β45 (CD4) Phe→Cys]: a new unstable hae
291 Waye JS, Eng B, Patterson M, Barr RD and Chui moglobin with reduced oxygen affinity in trans
DHK (1997) De novo mutation of the β‐globin with β‐thalassaemia. Br J Haematol, 91, 595–601.
gene initiation codon (ATG→AAG) in a Northern 303 Huisman THJ (1997) Combinations of β chain
European boy. Am J Hematol, 56, 179–182. abnormal hemoglobins with each other and with
292 Harteveld CL, Wijermans PW, Arkesteijn DGJ, β‐thalassemia determinants with known muta
Van Delft P, Kerkhoffs JL and Giordano PC (2008) tions: influence on phenotype. Clin Chem, 43,
Hb Lepore‐Leiden: a new δ/β rearrangement 1850–1856.
304 Papassotiriou I, Traeger‐Synodinos J, Prom‚ D, 314 Mohamed N and Jackson N (1998) Severe thalas
Kister J, Premetis E, Stamoulakatou A et al. (1998) saemia intermedia: clinical problems in the absence
Hb Acharnes or β53(D4) Ala→Thr: an unstable of hypertransfusion. Blood Rev, 12, 163–170.
electrophoretically silent haemoglobin variant in 315 Mejri A, Mansri M, Hadj Fredj S, Ouali F, Bibi A,
association with β‐thalassaemia. Br J Haematol, Hafsia R et al. (2016) First description of the
102, 178. rs45496295 polymorphism of the C/EBPE gene in
305 Premawardhena A, Fisher CA, Olivieri NF, de β‐thalassemia intermedia patients. Hemoglobin, 40,
Silva S, Sloane‐Stanley J, Wood WG and Weatherall 411–416.
DJ (2005) A novel molecular basis for β thalas 316 Traeger‐Synodinos J, Bosch CAJ, Vrettou C,
semia intermedia poses new questions about its Maragoudaki L, Apostolidis J, Kanavakas E et al.
pathophysiology. Blood, 106, 3251–3255. (2019) Adult‐onset beta‐thalassaemia intermedia
306 Origa R, Sollaino MC, Borgna‐Pignatti C, Piga caused by a 5‐Mb somatic clonal segmental dele
A, Torres AF, Masile V and Galanello R (2014) tion in haemopoietic stem cells involving the β‐
α‐globin gene quadruplication and heterozy globin locus. Br J Haematol, 186, e165–e170.
gous β‐thalassemia: a not so rare cause of thalas 317 Chang Y‐P C, Littera R, Garau R, Smith KD, Dover
semia intermedia. Acta Haematologica, 131, GJ, Iannelli S et al. (2001) The role of heterocellular
162–164. hereditary persistence of fetal haemoglobin in β0‐
307 Clark B, Shooter C, Smith F, Brawand D, Steedman thalassaemia intermedia. Br J Haematol, 114,
L, Oakley M et al. (2018) Beta thalassaemia inter 899–906.
media due to co‐inheritance of three unique alpha 318 Maffei L, Foglietta E, Cappabianca MP, Lerone M,
globin cluster duplications characterised by next Mastropietro F and Bianco I (2002) The prevention
generation sequencing analysis. Br J Haematol, 180, of thalassemia major: emerging problems.
160–164. Haematologica, 87, ELT44.
308 Harteveld CL, Refaldi C, Cassinerio E, Cappellini 319 Taher AT, Musallam KM, El‐Beshlawy A, Karimi
MD and Giordano PC (2008) Segmental duplica M, Daar S, Belhoul K et al. (2010) Age‐related com
tions involving the alpha‐globin gene cluster are plications in treatment‐naïve patients with thalas
causing beta‐thalassemia intermedia phenotypes saemia intermedia. Br J Haematol, 150, 486–489.
in beta‐thalassemia heterozygous patients. Blood 320 Musallam KM, Sankaran VG, Cappellini MD,
Cells Mol Dis, 40, 312–316. Duca L, Nathan DG and Taher AT (2012) Fetal
309 Badens C, Mattei MG, Imbert AM, Lapouméroulie hemoglobin levels and morbidity in untransfused
C, Martini N, Michel G and Lena‐Russo D (2002) patients with β‐thalassemia intermedia. Blood, 119,
A novel mechanism for thalassaemia intermedia. 364–367.
Lancet, 359, 132–133. 321 Saxon BR, Rees D and Olivieri NF (1998)
310 Galanello R, Perseu L, Perra C, Maccioni L, Barella Regression of extramedullary haemopoiesis and
S, Longinotti M et al. (2004) Somatic deletion of the augmentation of fetal haemoglobin concentration
normal beta‐globin gene leading to thalassaemia during hydroxyurea therapy in β thalassaemia. Br
intermedia in heterozygous beta‐thalassaemic J Haematol, 101, 416–419.
patients. Br J Haematol, 127, 604–606. 322 Taher AT, Porter JB, Viprakasit V, Kattamis A,
311 Bank A (2006) Regulation of human fetal hemo Chuncharunee S, Sutcharitchan P et al. (2015)
globin: new players, new complexities. Blood, 107, Defining serum ferritin thresholds to predict clini
435–443. cally relevant liver iron concentrations for guiding
312 Rund D, Oron‐Karni V, Filon D, Goldfarb A, deferasirox therapy when MRI is unavailable in
Rachmilewitz E and Oppenheim A (1997) Genetic patients with non‐transfusion‐dependent thalas
analysis of β‐thalassaemia intermedia in Israel: saemia. Br J Haematol, 168, 284–290.
diversity of mechanisms and unpredictability of 323 Ricchi P, Meloni A, Spasiano A, Costantini S, Pepe
phenotype. Am J Hematol, 54, 16–22. A, Cinque P and Filosa A (2018) The impact of
313 Ho PJ, Hall GW, Luo LY, Weatherall DJ and Thein liver steatosis on the ability of serum ferritin levels
SL (1998) Beta‐thalassaemia intermedia: is it pos to be predictive of liver iron concentration in non‐
sible consistently to predict phenotype from geno transfusion‐dependent thalassaemia patients. Br J
type? Br J Haematol, 100, 70–78. Haematol, 180, 721–726.
180 Chapter 3
324 Fisher CA, Premawardhena A, de Silva S, Perera β‐thalassaemia major. Intern Med J, 47,
G, Rajapaksa S, Olivieri NA, Old JM, Weatherall 1064–1067.
DJ and the Sri Lanka Thalassaemia Study Group 334 Ricchi P, Costantini S, Spasiano A, Di Matola T,
(2003) The molecular basis of the thalassaemias in Cinque P and Filosa A (2017) Myelolipoma among
Sri Lanka. Br J Haematol, 121, 662–671. patients with thalassemia major and rare anemia
325 Chang J‐G, Tsai W‐C, Chong IW, Chang C‐S, Lin with iron loading: a not so rare entity. Am J
C‐C and Liu TC (2008) β‐thalassemia major evolu Hematol, 92, E25–E26.
tion from β‐thalassemia minor is associated with 335 Mustafa NY, Marouf R, Al‐Humood S and Al‐
paternal uniparental isodisomy of chromosome Fadhli SM (2010) Hypercoagulable state and
11p15. Haematologica, 93, 913–916. methylenetetrahydrofolate reductase (MTHFR)
326 Harteveld CL, Refaldi C, Giambona A, C677T mutation in patients with beta‐thalassemia
Ruivenkamp CA, Hoffer MJ, Pijpe J, De Knijff P major in Kuwait. Acta Haematol, 123, 37–42.
et al. (2013) Mosaic segmental uniparental isodis 336 Wickramasinghe SN and Bush V (1975)
omy and progressive clonal selection: a common Observations on the ultrastructure of erythropoi
mechanism of late onset β‐thalassemia major. etic cells and reticulum cells in the bone marrow
Haematologica, 98, 691–695. of patients with homozygous β thalassaemia. Br J
327 Mettananda S, Gibbons RJ and Higgs DR (2015) Haematol, 30, 395–399.
α‐globin as a molecular target in the treatment of 337 Ragab SM, Safan MA and Badr EA (2105) Study of
β‐thalassemia. Blood, 125, 3694–3701. serum haptoglobin level and its relation to eryth
328 Premawardhena A, Ranawaka U, Pilapitiya T, ropoietic activity in beta thalassemia children.
Weerasinghe G, Hapangama A, Hettiarachchi S Med J Hematol Infec Dis, 7, e2015019.
et al. (2019) Headache: an important symptom 338 Wood WG, Hereditary persistence of fetal hemo
possibly linked to white matter lesions in thalas globin and δβ thalassemia. In: Steinberg MH,
saemia. Br J Haematol, 185, 541–548. Forget BG, Higgs DR and Nagel RL, eds. Disorders
329 De Sanctis V, Elsedfy H, Soliman AT, Elhakim IZ, of Hemoglobin: Genetics, Pathophysiology, and
Soliman NA, Karimi M and Elalaily R (2016) The Clinical Management. Cambridge University Press,
diagnostic approach to central adrenocortical Cambridge, 2001, pp. 356–388.
insufficiency (CAI) in thalassemia. Mediterr J 339 Ibrahim HR (2016) A family with heterozygous δβ
Hematol Infect Dis, 8, e2016026. thalassaemia. www.bloodmed.com (accessed 7th
330 Kremastinos DT and Farmakis D (2011) Iron over November 2016).
load cardiomyopathy in clinical practice. 340 Ibrahim HR (2016) Rare homozygous delta beta
Circulation, 124, 2253–2263. thalassaemia critically analysed. www.bloodmed.
331 Pepe A, Gamberini MR, Missere M, Pistoia L, com (accessed 7th November 2016).
Mangione M, Cuccia L et al.; Working Group of 341 Zertal‐Zidani S, Ducrocq R, Weil‐Oliver C, Elion J
Gender Medicine of Italian Society of Cardiology and Krishnamoorthy R (2001) A novel δβ fusion
(2018) Gender differences in the development of gene expresses hemoglobin A (HbA) not Hb
cardiac complications: a multicentre study in a Lepore: Senegalese δ0β+ thalassaemia. Blood, 98,
large cohort of thalassaemia major patients to 1261–1263.
optimize the timing of cardiac follow‐up. Br J 342 Galanello R, Barella S, Satta S, Maccioni L, Pintor
Haematol, 180, 879–888. C and Cao A (2002) Homozygosity for non‐dele
332 Vogiatzi MG, Macklin EA, Trachtenberg FL, Fung tional δ‐β0 thalassemia resulting in a silent clinical
EB, Cheung AM, Vichinsky E et al.; Thalassemia phenotype. Blood, 100, 1913–1914.
Clinical Research Network (2009) Differences in 343 Rooks H, Bergounioux J, Game L, Close JP,
the prevalence of growth, endocrine and vitamin Osborne C, Best S et al. (2005) Heterogeneity of the
D abnormalities among the various thalassaemia εγδβ‐thalassaemias: characterization of three novel
syndromes in North America. Br J Haematol, 146, English deletions. Br J Haematol, 128, 722–729.
546–556. 344 Zebisch A, Schulz E, Grosso M, Lombardo B,
333 Wong P, Milat F, Fuller PJ, Kerr PG, Doery JCG, Acierno G, Sill H and Iolascon A (2015)
Oh DH et al. (2017) Urolithiasis is prevalent and Identification of a novel variant of epsilon‐gamma‐
associated with reduced bone mineral density in delta‐beta thalassemia highlights limitations of
next generation sequencing. Am J Hematol, 90, 356 Dror Y and Freedman MH (2002) Shwachman‐
E52–E54. Diamond syndrome. Br J Haematol, 118, 701–713.
345 Harteveld CL, Osborne CS, Peters M, van der 357 Kar R, Saxena R and Pati HP (2008) Hereditary
Werf S, Plug R, Fraser P and Giordano PC (2003) spherocytosis with high fetal hemoglobin: an
Novel 112 kb (εGγAγ) δβ–thalassaemia deletion in interesting case. Hemoglobin, 32, 520–523.
a Dutch family. Br J Haematol, 122, 855–858. 358 Peters F, Rohloff D, Kohlmann T, Renner F,
346 Von Kanel T, Röthlisberger B, Schanz U, Dutly F, Jantschek G, Kerner W and Fehm HL (1998) Fetal
Huber AR and Saller E (2013) A Swiss (εγδβ)⁰‐ hemoglobin in starvation ketosis in young women.
thalassemia patient with a novel 3‐Mb deletion Blood, 91, 691–694.
associated with mild mental impairment. Am J 359 Little JA, Dempsey NJ, Tuchman M and Ginder
Hematol, 88, 158–159. GD (1995) Metabolic persistence of fetal hemo
347 Game L, Bergounioux J, Close JP, Marzouka BE globin. Blood, 85, 1712–1718.
and Thein SL (2003) A novel deletion causing 360 Schilirò G, Musumeci S, Pizzarelli G, Russo A,
(εγδβ)0 thalassaemia in a Chilean family. Br J Marinucci M, Tentori L and Russo G (1978) Fetal
Haematol, 123, 154–159. haemoglobin in early malignant osteopetrosis. Br J
348 Rose C, Rossignol J, Labilliotte A, Depret S, Le Haematol, 38, 339–344.
Metayer N and Pissard S (2009) A novel (εγδβ)°‐ 361 Wilson MG, Schroeder WA, Graves DA and Kach
thalassemia deletion associated with an alpha glo VD (1967) Haemoglobin variations in D‐trisomy
bin gene triplication leading to a severe transfusion syndrome. N Engl J Med, 277, 953–958.
dependent fetal thalassemic syndrome. 362 Sankaran VG, Xu J, Byron R, Greisman HA, Fisher
Haematologica, 94, 593–594. C, Weatherall DJ et al. (2011) A functional element
349 Huisman THJ (1997) Gamma chain abnormal necessary for fetal hemoglobin silencing. N Engl J
human fetal hemoglobin variants. Am J Hematol, Med, 365, 807–814.
55, 159–163. 363 Kosteas T, Palena A and Anagno NP (1997)
350 Craig JE, Rochette J, Sampietro M, Wilkie AOM, Molecular cloning of the breakpoints of the heredi
Barnetson R, Hatton CSR et al. (1997) Genetic het tary persistence of fetal hemoglobin type 6
erogeneity in heterocellular hereditary persistence (HPFH‐6) deletion and sequence analysis of the
of fetal hemoglobin. Blood, 90, 428–434. novel juxtaposed region from the 3′ end of the of
351 Gilman JG, Mishima N, Wen XJ, Kutlar F and the β‐globin gene cluster. Hum Genet, 100, 441–445.
Huisman THJ (1988) Upstream promoter mutation 364 Fucharoen S, Fucharoen G, Sanchaisuriya K and
associated with a modest elevation of fetal hemo Surapot S (2002) Molecular characterization of
globin expression in human adults. Blood, 72, 78–81. thalassaemia intermedia associated with
352 Funnell AP, Prontera P, Ottaviani V, Piccione M, HPFH‐6/β‐thalassaemia and HPTH‐6/Hb E in
Giambona A, Maggio A et al. (2015) 2p15‐p16.1 Thai patients. Acta Haematol, 108, 157–161.
microdeletions encompassing and proximal to 365 Forget BG (1998) Molecular basis of hereditary
BCL11A are associated with elevated HbF in addi persistence of fetal hemoglobin. Ann NY Acad Sci,
tion to neurologic impairment. Blood, 126, 89–93. 850, 38–44.
353 Satta S, Perseu L, Moi P, Asunis I, Cabriolu A, 366 Natiq A, Lysy PA, Gillemans N, Schaap R, Sefiani
Maccioni L et al. (2011) Compound heterozygosity A, Amzazi S et al. (2017) Hereditary persistence of
for KLF1 mutations associated with remarkable fetal hemoglobin in two patients with KLF1 hap
increase of fetal hemoglobin and red cell proto loinsufficiency due to 19p13.2‐p13.12/13 deletion.
porphyrin. Haematologica, 96, 767–770. Am J Hematol, 92, E2–E3.
354 Islam A, Rafiq S, Kirwan M, Walne A, Cavenagh J, 367 Arnaud L, Saison C, Helias V, Lucien N,
Vulliamy T and Dokal I (2013) Haematological Steschenko D, Giarratana MC et al. (2010) A domi
recovery in dyskeratosis congenita patients nant mutation in the gene encoding the erythroid
treated with danazol. Br J Haematol, 162, 854–856. transcription factor KLF1 causes a congenital dys
355 Wickramasinghe SN (1998) Congenital dyseryth erythropoietic anemia. Am J Hum Genet, 87,
ropoietic anaemias: clinical features, haematologi 721–727.
cal morphology and new biochemical data. Blood 368 Magor GW, Tallack MR, Gillinder KR, Bell CC,
Rev, 12, 178–200. McCallum N, Williams B and Perkins AC (2015)
182 Chapter 3
KLF1‐null neonates display hydrops fetalis and a 380 Wilcox I, Boettger K, Greene L, Malek A, Davis L,
deranged erythroid transcriptome. Blood, 125, Steinberg MH et al. (2009) Hemoglobin Kenya
2405–2417. composed of α‐ and (Aγβ)‐fusion‐globin chains,
369 Bollekens JA and Forget BG (1991) δβ thalassemia associated with hereditary persistence of fetal
and hereditary persistence of fetal hemoglobin. hemoglobin. Am J Hematol, 84, 55–58.
Hematol Oncol Clin North Am, 5, 399–422. 381 Wood WG (1993) Increased Hb F in adult life.
370 Stamatoyannopoulos G and Nienhuis AW. Bailliere’s Clin Haematol, 6, 177–213.
Hemoglobin switching. In: Stamatoyannopoulos 382 Chui DHK, Patterson M, Dowling CE, Kazazian
G, Nienhuis AW, Majerus PW and Varmus H, eds. HH and Kendall AG (1990) Hemoglobin Bart’s
The Molecular Basis of Blood Diseases, 2nd edn. WB disease in an Italian boy: interaction between α‐
Saunders, Philadelphia, 1994, pp. 107–156. thalassaemia and hereditary persistence of fetal
371 Jane SM and Cunningham JM (1998) hemoglobin. N Engl J Med, 323, 179–182.
Understanding fetal globin gene expression: a 383 Merghoub T, Perichon B, Maier‐Redelsperger M,
step towards effective Hb F reactivation in haemo Dibenedetto SP, Samperi P, Ducrocq R et al. (1997)
globinopathies. Br J Haematol, 102, 415–422. Dissection of the association status of two poly
372 Conley CL, Weatherall DL, Richardson SN, morphisms in the β‐globin gene cluster with vari
Shepard MK and Charache S (1963) Hereditary ations in F‐cell number in non‐anemic individuals.
persistence of fetal hemoglobin. A study of 79 Am J Hematol, 56, 239–243.
affected persons in 15 Negro families in Baltimore. 384 Collins FS, Stoeckert CJ, Serjeant GR, Forget BG
Blood, 21, 261–281. and Weissman SM (1984) Gγβ+ hereditary per
373 Schroeder WH, Huisman THJ and Sukumaran PK sistence of fetal hemoglobin: cosmid cloning
(1973) A second type of hereditary persistence of and identification of a specific mutation 5′ to
foetal haemoglobin. Br J Haematol, 25, 131–135. the Gγ gene. Proc Natl Acad Sci U S A, 81,
374 Wainscoat JS, Old JM, Wood WG, Trent RJ and 4894–4898.
Weatherall DJ (1984) Characterization of an Indian 385 Surrey S, Delgrosso K, Malladi P and Schwartz E
(δβ)0 thalassaemia. Br J Haematol, 58, 353–360. (1988) A single base change at position –175 in the
375 Saglio G, Camaschella C, Serra A, Bertero T, Rege 5′‐flanking region of the Gγ‐globin gene from a
Cambrin G, Guerrasiqo A et al. (1986) Italian type black with Gγβ+HPFH. Blood, 71, 807–810.
of deletional hereditary persistence of fetal hemo 386 Ottolenghi S, Nicolis S, Taramelli R, Malgaretti N,
globin. Blood, 68, 646–651. Mantovani R, Comi P et al. (1988) Sardinian Gγ‐
376 Camaschella C, Serra A, Gottardi E, Alfarano A, HPFH: A T‐>C substitution in a conserved “octo
Revello D, Mazza U and Saglio G (1990) A new mer” sequence in the Gγ‐globin promoter. Blood,
hereditary persistence of fetal hemoglobin dele 71, 815–817.
tional has the breakpoint within the 3′ β‐globin 387 Craig JE, Sheerin SM, Barnetson R and Thein SL
gene enhancer. Blood, 75, 1000–1005. (1993) The molecular basis of HPFH in a British
377 Adams JG, Coleman MB, Hayes J, Morrison WT family identified by heteroduplex formation. Br J
and Steinberg MH (1985) Modulation of fetal Haematol, 84, 106–110.
hemoglobin synthesis by iron deficiency. N Engl J 388 Gilman JG and Huisman THJ (1985) DNA
Med, 313, 1402–1405. sequence variation associated with elevated fetal
378 Landman H and Huisman TH (1998) Persistent Gγ globin production. Blood, 66, 783–787.
iron and folate deficiency in a patient with dele 389 Motum PL, Deng Z‐M, Huong L and Trent RJ
tional hereditary persistence of fetal hemoglobin: (1994) The Australian type of nondeletional Gγ‐
the effect of the relative levels of Hb F and Gγ HPFH has a C→G substitution at nucleotide –114
chains and the corresponding mRNAs. Hemoglobin, of the Gγ gene. Br J Haematol, 86, 219–221.
22, 53–63. 390 Pissard S, M’rad A, Beuzard Y and Roméo P‐H
379 Kutlar A, Gardiner MB, Headlee MG, Reese AL, (1996) A new type of hereditary persistence of
Cleek MP, Nagle S et al. (1984) Heterogeneity in fetal haemoglobin (HPFH): HPFH Tunisia β+
the molecular basis of three types of hereditary (+C‐200)Gγ. Br J Haematol, 95, 67–72.
persistence of fetal hemoglobin and the relative 391 Efremov GD, Filipce V, Gjorgovski I, Juricic D,
synthesis of the G gamma and A gamma types of Stojanoski N, Harano T et al. (1986) GγAγ(δβ)0‐tha
gamma chain. Biochem Genet, 22, 21–35. lassaemia and a new form of γ globin gene
triplication identified in the Yugoslavian popula 402 Dedoussis GV, Sinopoulou K, Gyparaki M and
tion. Br J Haematol, 63, 17–29. Loutradis A (1999) Fetal hemoglobin expression in
392 Gonçalves I, Lavinha J, Ducrocq R and Osório‐ the compound heterozygote state for –117(G→A)
Almeida L (2002) A novel rearrangement of the Aγ HPFH and IVSII‐745 (C→G) β+ thalassemia: a
human fetal globin genes leading to a six γ‐globin case study. Am J Hematol, 61, 139–143.
gene haplotype. Br J Haematol, 116, 454–457. 403 Papadakis NM, Patrinos GP, Tsaftaridis P and
393 Giglioni B, Casini C, Mantovani R, Merli S, Comi Loutradi‐Anagnostou A (2000) Foetal hemoglobin
P, Ottolenghi S et al. (1984) A molecular study of a expression in compound heterozygotes for the
family with Greek hereditary persistence of fetal Greek type of nondeletional HPFH and beta‐
hemoglobin and β‐thalassemia. EMBO J, 3, thalassemia. Hematol J, 1, Suppl. 1, 36.
2641–2645. 404 Öner R, Kutlar F, Gu L‐H and Huisman THJ (1991)
394 Ottolenghi S, Camaschella C, Comi P, Giglioni B, The Georgia type of nondeletional hereditary per
Longinotti M, Oggiani L et al. (1988) A frequent Aγ‐ sistence of fetal hemoglobin has a C→T mutations at
hereditary persistence of fetal hemoglobin in nucleotide –114 of the Aγ globin gene. Blood, 77, 1124.
northern Sardinia: its molecular basis and hema 405 Gilman JG, Mishima N, Wen XJ, Stoming TA,
tologic phenotype in heterozygotes and com Lobel J and Huisman THJ (1988) Distal CCAAT
pound heterozygotes with β‐thalassaemia. Hum box deletion in the Aγ globin gene of two black
Genet, 79, 13–17. adolescents with elevated fetal Aγ globin. Nucleic
395 Bordin S, Martins JT, Gonçalves MS, Melo MB, Acid Res, 16, 10635–10642.
Saad STO and Costa FF (1998) Haplotype analysis 406 Huang X‐D, Yang XO, Huang R‐B, Zhang H‐Y,
and Aγ gene polymorphism associated with Zhao H‐L, Zhao Y‐J et al. (2000) A novel four base‐
Brazilian type of hereditary persistence of fetal pair deletion within the Aγ‐globin gene promoter
hemoglobin. Am J Hematol, 58, 49–54. associated with slight increase of Aγ expression in
396 Stoming TA, Stoming GS, Lanclos KD, Fei YJ, adult. Am J Hematol, 63, 16–19.
Altay F, Kutlar F and Huisman THJ (1989) An Aγ 407 Craig JE, Rochette J, Fisher CA, Weatherall DJ,
type of nondeletional hereditary persistence of Marc S, Lathrop GM et al. (1996) Dissecting the
fetal hemoglobin with a T → C mutation at posi loci controlling fetal hemoglobin production on
tion –175 to the cap site of the Aγ globin gene. chromosome 11p and 6q by the regressive
Blood, 73, 329–333. approach. Nat Genet, 12, 58–64.
397 Fessas P and Stamatoyannopoulos G (1964) 408 Borg J, Papadopoulos P, Georgitsi M, Gutiérrez L,
Hereditary persistence of fetal hemoglobin in Greece: Grech G, Fanis P et al. (2010) Haploinsufficiency
a study and a comparison. Blood, 24, 223–240. for the erythroid transcription factor KLF1 causes
398 Sofroniadou K, Wood WG, Nute PE and hereditary persistence of fetal hemoglobin. Nat
Stamatoyannopoulos G (1975) Globin gene syn Genet, 42, 801–805.
thesis in the Greek type (Aγ) of hereditary persis 409 Amato A, Cappabianca MP, Perri M, Zaghis I,
tence of fetal haemoglobin. Br J Haematol, 29, Grisanti P, Ponzini D and Di Biagio P (2014)
137–147. Interpreting elevated fetal hemoglobin in pathol
399 Collins FS, Metherall JE, Yamakawa M, Pan J, ogy and health at the basic laboratory level: new
Weissman SM and Forget BG (1985) A point muta and known γ‐gene mutations associated with
tion in the Aγ‐globin gene promoter in Greek hereditary persistence of fetal hemoglobin. Int J
hereditary persistence of fetal haemoglobin. Lab Hematol, 36, 13–19.
Nature, 313, 325–326. 410 Winichagoon P, Thonglairoam V, Fucharoen S,
400 Gelinas R, Endlich B, Pfeiffer C, Yagi M and Wilairat P, Fukumaki Y and Wasi P (1993) Severity
Stamatoyannopoulos G (1985) G to A substitution differences in β‐thalassaemia/haemoglobin E syn
in the distal CCAAT box of the Aγ‐globin gene in dromes: implication of genetic factors. Br J
Greek hereditary persistence of fetal haemoglo Haematol, 83, 633–639.
bin. Nature, 313, 323–324. 411 Gonçalves I, Ducrocq R, Lavinha J, Nogueira PJ,
401 Camaschella C, Oggiano L, Sampietro M, Gottardi Peres MJ, Picanco I et al. (1998) Combined effect of
E, Alfarano A, Pistidda P et al. (1989) The homozy two different polymorphic sequences within the β
gous state of G to A –117 Aγ hereditary persistence globin gene cluster on the level of Hb F. Am J
of fetal hemoglobin. Blood, 73, 1999–2002. Hematol, 57, 269–276.
184 Chapter 3
3.1 (a) F 3.5 (a) F 3.9 (a) T 3.13 (a) T
(b) T (b) F (b) F (b) F
(c) F (c) T (c) T (c) T
(d) F (d) T (d) T (d) F
(e) F (e) F (e) T (e) T
3.2 (a) F 3.6 (a) T 3.10 (a) F

(b) T (b) F (b) T
(c) T (c) T (c) T
(d) T (d) T (d) T
(e) T (e) T (e) T
3.3 (a) T 3.7 (a) F 3.11 (a) F

(b) T (b) F (b) F
(c) T (c) T (c) T
(d) F (d) F (d) T
(e) T (e) F (e) T
3.4 (a) F 3.8 (a) T 3.12 (a) F

(b) T (b) F (b) F
(c) F (c) F (c) T
(d) T (d) F (d) T
(e) T (e) T (e) T
4 Sickle cell haemoglobin and its interactions with other
variant haemoglobins and with thalassaemias
The first description of sickle cell anaemia is generally copolymerise with haemoglobin S and the hybrid
attributed to James Herrick who, in 1910 reported tetramer, α2βSγ, is similarly unable to polymerise.
that a patient with severe anaemia had ‘peculiar elon- In comparison with haemoglobin A, sickling is facili-
gated and sickle shaped red blood corpuscles’ – an tated by the presence of haemoglobin C, haemoglobin
observation that had been initially made by his intern, D‐Punjab/D‐Los Angeles or haemoglobin O‐Arab.
Ernest Lyons [1]. The term ‘sickle cell anaemia’ was Because acidosis and a rise in temperature shift the
first used by Verne Mason in 1922 [2]. Several decades oxygen dissociation curve to the right they favour
later, Linus Pauling and colleagues found that the sickling. However, in clinical practice, exposure to cold
sickling phenomenon was caused by haemoglobin can also provoke sickling because of slowed c irculation
with unusual characteristics [3] and subsequently through capillaries.
Vernon Ingram and colleagues identified the causa- The sickle cell mutation appears to have arisen
tive amino acid in the β chain of haemoglobin [4]. spontaneously at least five times in the history of
Sickle cell haemoglobin, haemoglobin S, has a valine mankind (Fig. 4.2). Such independent mutations
for glutamic acid substitution at position 6 of the β can be recognised by their association with different
chain. The haemoglobin can be designated α2β26Glu→Val. β globin gene haplotypes, demonstrated by analysis
Sickle cell haemoglobin can produce deleterious of restriction fragment length polymorphisms
effects because, on deoxygenation, its solubility is (RFLP). There are three main foci of haemoglobin S
reduced and polymerisation occurs (Fig. 4.1). Both in Africa, associated with different haplotypes, the
partially and fully deoxygenated haemoglobin S can haplotypes being defined by RFLP analysis. They
be incorporated into a polymer. Long polymers dis- are in Senegal (Senegal type), the Central African
tort the red cell into a holly leaf, crescent or sickle Republic and southern Africa (Bantu or Central
shape that hinders blood flow through capillaries, African Republic type) and Benin, Central, West
both because of reduced deformability and because and North Africa (Benin type) [5]. The Benin type
of increased adhesion to endothelial cells resulting has also spread to Spain, Portugal, Sicily (perhaps
from secondary changes in the red cell membrane. from Greece, perhaps from Sudanese soldiers in
The lag phase before polymerisation occurs is deter- Arab armies) and southern mainland Italy, Greece
mined by the partial pressure of O2, temperature, pH (particularly Macedonia), Albania, Turkey, northern
and haemoglobin S concentration within the red cell. and south‐western Saudi Arabia, Yemen and Oman.
When fully oxygenated, haemoglobin S is as soluble The Bantu type has spread to Kenya, Zambia and
as haemoglobin A. Although haemoglobin A can the Sudan. In addition to the three major foci, there
copolymerise with haemoglobin S, the presence of may have been a further independent mutation
haemoglobin A in a red cell slows polymerisation among the Eton people in southern Cameroon. A
because the concentration of haemoglobin S is less. fifth mutation is associated with further foci in east-
Haemoglobin F and haemoglobin A2 are even more ern Saudi Arabia, Kuwait, Bahrain, Iran and Oman
effective at retarding sickling since they do not and in extensive areas of central and southern India,

185
186 Chapter 4
Fig. 4.1 Transmission electron

micrography showing
polymerisation of haemoglobin S in
a patient with compound
heterozygosity for haemoglobin S
and haemoglobin D‐Punjab.
(With thanks to Mr S. Ladva.)
?
Arab-Indian
Senegal
Benin
Bantu
Fig. 4.2 The multifocal origin and

spread of the βS gene.
particularly among the scheduled tribes (a group central India. The Indian/Saudi Arabian haplotype
living outside the caste system). It is designated the has also been found in Afghanistan and among
Arab‐Indian haplotype. It may have arisen initially Bedouin Arabs in Israel.
in the Indus valley. The prevalence of the haemo- Migration from Africa has led to the sickle cell
globin S gene is up to 25% in eastern Saudi Arabia gene occurring also in Central and South America,
and as high as 30% in some tribal populations in in Afro‐Americans and in Afro‐Caribbeans in
Sickle cell haemoglobin and its interactions with other variant haemoglobins 187
Canada and in the UK and other European countries. electrophoresis, and in their retention time on high
For example the prevalence of the βS gene at birth in performance liquid chromatography (HPLC).
England has been estimated at 0.39 per 1000 [6] and There are other haemoglobins unrelated to hae-
in parts of Belgium (Brussels and Liege) at 1.5% [7]. moglobin S that can polymerise in vitro, such as
There is a high prevalence in some populations in haemoglobin I (α16 Lys→Glu) and haemoglobin
Mexico, Colombia, Venezuela, Guyana, Suriname, Setif (α22 Asp→Tyr). Although they are not associ-
French Guiana, Brazil and Peru. All three major ated with any relevant clinical abnormality, hae-
African haplotypes are represented in the USA, the moglobin Setif can cause a false positive sickle
Caribbean and the UK. The sickle cell gene is also solubility test [49]. There are also other haemo-
found in Madagascar, Mauritius (both Bantu and globins that have the same amino acid substitution
Arab‐Indian haplotypes), Abu Dhabi, United Arab at a different site and which have the same charac-
Emirates, Lebanon, Iraq, the southern part of the teristics as haemoglobin S on HPLC but which do
former USSR and among North African Arabs. not lead to sickling; for example, haemoglobin
The wide geographic spread of a potentially Haaglanden has the same characteristics on HPLC
deleterious gene has been attributed to protection although mobility is slightly different on capillary
of heterozygotes from premature death from falci- electrophoresis [50]; a sickle solubility or equiva-
parum malaria. In areas where malaria is endemic, lent test is important in making the distinction.
the βS and βA genes may exist as balanced polymor- Homozygosity for haemoglobin S (βSβS) causes a
phism (i.e. death or serious disability from sickle serious condition referred to as ‘sickle cell anaemia’.
cell anaemia before the age of reproduction is Heterozygosity for haemoglobin S (ββS), referred to
balanced by a decreased death rate from malaria as sickle cell trait, is usually asymptomatic. The βS
among heterozygotes). The prevalence of haemo- gene may also be coinherited with another β chain
globin S in various populations is shown in variant. When there is deleterious interaction
Table 4.1 [8–40]. between the sickle cell haemoglobin and the second
A second mutation is occasionally present in a variant haemoglobin, as is the case, for example,
gene that carries the βS mutation leading to synthe- with haemoglobin C and haemoglobin D‐Punjab,
sis of a variant haemoglobin with two amino acid a clinically significant sickling disorder occurs.
substitutions. This could arise either through a sec- Subjects who are heterozygous for β thalassaemia
ond mutation occurring in a mutant gene or through and haemoglobin S likewise suffer from the clinico-
crossover between genes encoding two different pathological effects of sickle cell formation and con-
variant haemoglobins; for example, haemoglobin sequent vascular occlusion. The term ‘sickle cell
C‐Harlem could have arisen through crossover disease’ is sometimes used as a synonym for sickle
between βS and βKorle Bu. Such variant haemoglobins cell anaemia and sometimes used as a generic term
retain their ability to sickle, and in heterozygous, to include sickle cell anaemia and other conditions
homozygous and compound heterozygous states in which a clinically significant disorder results
have a similar, although not necessarily identical, from sickle cell formation and the associated patho-
significance to haemoglobin S. At least 14 such logical processes. Usage of the term ‘sickle cell dis-
double mutations are known (Table 4.2) [12, 41–48]. ease’ to refer to sickle cell anaemia is not recommended
The commonest, haemoglobin C‐Harlem (initially since this can cause confusion. To avoid ambiguity,
described under the name of haemoglobin C‐ the term should be defined whenever used. Some sig-
Georgetown), α2β26Glu→Val,73Asp→Asn, is less prone to nificant compound heterozygous states are shown in
polymerisation than haemoglobin S itself. The rare Table 4.3 [41, 43, 51–54].
double substitution haemoglobin, haemoglobin
S‐Antilles is even more prone to sickle than haemo-
Sickle cell trait
globin S itself as are haemoglobins Jamaica Plain
and S‐Oman. Some double substitution haemo- The term sickle cell trait indicates heterozygosity
globins differ from haemoglobin S in their mobility for the sickle cell gene (ββS). Sickle cell trait is
on cellulose acetate electrophoresis and capillary asymptomatic in the great majority of individuals
188 Chapter 4
Table 4.1 Prevalence (%) of haemoglobins S and C in different populations. (From references 8–40 and other sources.)
Country or people Variant haemoglobin
Haemoglobin S Haemoglobin C
West Africa
Senegal 3–15 <1–6
Gambia 6–28 <1–2
Guinea Bissau <1–25 <1–1.5
Guinea 13–33
Sierra Leone 22–30
Liberia <1–29 (10% in Monrovia, 1–3 (<1% in Monrovia) [36]
1% SS at birth) [36]
Ivory Coast 2–26 <1–50
Mali 5–1.7
Burkina Faso 2–34 15–40
1.75% sickle cell disease
(SC and SS) at birth
Ghana 3–25 8–40
Togo 6–28 7–17
Benin 5–31 (30% of pregnant 7–27 (8% of pregnant
women AS, SC or SS) women AC, SC or CC)
Niger 5–23 1–8
Nigeria 10–41 (2% birth <1–9
prevalence of sickle cell
disease)
Central Africa
Gabon 8–32
Cameroon <1–31 <1
Central African Republic 2–24
The Republic of the Congo (Congo‐Brazzaville) 7–32
Democratic Republic of the Congo (formerly 1–46 (17% AS, 1.4% SS at
Zaire) birth)
East Africa
Kenya <1–34
Uganda 1–39 (nationwide 13.3%)
[34]
Tanzania 1–38 (overall 13%) [35],
21% birth prevalence in
north‐west [40]
Rwanda
Tutsi <1–5
Hutu 5–15
Burundi 1.5–26
Southern Africa
Angola 8–40 <0.1%
Zambia <1–30
Zimbabwe <1–11
Malawi 3–18
Mozambique <1–40
Madagascar <1–23 (7% on coast, 1.5%
in highlands)
Botswana <1
Namibia 0–15
South Africa
Bantu <1–4
Indian 2–10
Cape Coloured <1 <1
North Africa
Morocco <1–7 <1–6
Algeria <1–15 <1–13
Tunisia <1–2
Libya <1–6
Egypt <1*
Sudan <1–17 ~0.4%
Horn of Africa
Ethiopia 0–1
Djibouti ≈0
Somalia ≈0
USA
Afro‐Americans 7–8% 1–3.5
US Hispanic 0.7%
US residents overall 1.6%
Afro‐Caribbeans
Jamaica 3.5–12 2–4
Bahamas 14 3
Barbados 4 3–5
Cuba 0–23 0–2.5
Haiti 7–17 1–3
Dominican Republic 6–12 3
Puerto Rica <1–8 <1–2
Lesser Antilles 1–14 1–4.5
Guadeloupe 0.2–4.4
Martinique 8
Curacao 2
Central America
Mexico <1–9 <1
Guatemala <1–17
Belize 0–25
El Salvador <1–2
Honduras <1–16
Nicaragua ≈0
Costa Rica <1–8
Panama 0–21 0–2.5
(Continued on pp. 190–191.)
190 Chapter 4
South America
Colombia 0–15 0–6
Venezuela 0–9 0–3
Guyana <1
Suriname 0–22 0–6
French Guiana 0–18 0–7
Ecuador ≈0 ≈0
Peru <1 ≈0
Bolivia ≈0 ≈0
Brazil 0–16 0–4
Paraguay ≈0
Argentina <1
Uruguay ≈0
Chile <1 <1
Europe
Greece 0–32
Turkey <1–34 0.5–1
Cyprus <1
Italy
Sicily <1–13% 0.6%
Sardinia ≈0
Mainland southern 0.5–1
Portugal <1–5
Spain 0.1 (indigenous 0.12 (southern Spain)
Spaniards); SCD
3/100 000 births [37]
Eastern Europe
Albania 3.2% (one study)
Middle East
Turkey <1–34
Syria <5–25
Lebanon <1
Jordan 4–6
Israel†
Arabs 1–38
Jews ≈0
Iraq 0–25
Iran ≈0.03
Saudi Arabia <1–36 (overall 4.2%
heterozygosity, 0.25%
SCD)
Kuwait 2
Bahrain 11–16
Oman 5 Rare
Yemen 1–2‡
Abu Dhabi 2
United Arab Emirates 2–5%
Qatar 7.5%
Asia
India 0–35§
Pakistan 0.5–1
Sri Lanka 0–1.05 in different Very rare
districts [38]
Thailand Rare
Nepal ~9% in the Tharu
population [39]
AC, haemoglobin C trait; AS, sickle cell trait; CC, homozygosity for haemoglobin C; SC, compound heterozygosity for haemoglobins S
and C; SCD, sickle cell disease; SS, sickle cell anaemia,
* 5–22% in various oases.
† Highest in Bedouin tribes.
‡ 23% in Western province.
§ 5–35% in various tribal populations [20]; 15% in Orissa, Madhya Pradesh and Maharashtra states [21].
Table 4.2 Variant haemoglobins in which the mutation of haemoglobin S is one of two mutations [12, 41–48].
Variant haemoglobin Second substitution Mobility on cellulose acetate at alkaline pH
C‐Harlem* β73 Asp→Asn C
C‐Ziguinchor β58 Pro→Arg C
S‐Travis β142 Ala→Val S
S‐Antilles† β23 Val→Ile S [41]
S‐Providence β82 Lys→Asn A
S‐Oman† β121 Glu→Lys With C or slower than C
S‐Wake* β139 Asp→Ser [42]
S‐Cameroon β90 Glu→Lys A2

Jamaica Plain† β68 Leu→Phe S [43]
S‐South End β132 Lys→Asn A [44]
C‐Ndjamena* β37 Trp→Gly C
S‐Clichy β8 Lys→Thr Between A and F [45]
S‐San Martin‡ β85 Phe→Leu S [46]
C‐New Cross β83 Gly→Asp Slightly slower than S (between S and C on

acid agarose) [47]
S‐São Paulo§ Β65 Lys→Glu Faster than A [48]
* Similar disease to sickle cell anaemia in compound heterozygote.

† Sickling can occur in heterozygotes and S‐Antilles/C compound heterozygotes; more severe disease than sickle cell anaemia occurs in
compound heterozygote with haemoglobin S.
‡ Unstable, chronic haemolytic anaemia with exacerbations in two heterozygotes, microcytosis, variant haemoglobin 16 and 18%.
§ Mildly unstable, moderate anaemia in a heterozygote.
192 Chapter 4
Table 4.3 Causes of sickle cell disease [41, 43, 51–54]. diabetes mellitus may be made in a patient being
investigated for sickle cell trait. It should also be
Sickle cell anaemia (homozygosity for haemoglobin S) noted that underestimation of prior glycaemia by
Compound heterozygous states haemoglobin A1c measurement has been reported in
Sickle cell/haemoglobin C disease patients with sickle cell trait, but analytical interfer-
Sickle cell/β thalassaemia ence rather than a biological difference has not been
Sickle cell/haemoglobin D‐Punjab excluded [58].
Sickle cell/haemoglobin C‐Harlem
Sickle cell/haemoglobin S‐Antilles
Sickle cell/haemoglobin O‐Arab Clinical features
Sickle cell/haemoglobin Quebec‐Chori [51]
Sickle cell/haemoglobin S‐Oman People with sickle cell trait are usually asympto-
Sickle cell/haemoglobin O‐Tibesti matic. However sickle cell formation leading to vas-
Sickle cell/haemoglobin Monroe cular occlusion can occur during high fever and
Haemoglobin S‐Antilles /haemoglobin C under conditions of significant hypoxia, such as
Sickle cell/haemoglobin Lepore during travel by air (particularly but not only in
Haemoglobin C/haemoglobin C‐Harlem [52]
unpressurised aircraft), during mountain climbing,
Mutations leading to sickle cell disease in ascending in a cable car and skiing at altitude, dur-
heterozygotes ing vigorous exercise and during anaesthetic mis-
Haemoglobin S‐Antilles [41] adventures. Five percent of cells may be sickled
Haemoglobin S trait plus haemoglobin Conakry trait (an
during high altitude flying [59]. Vascular occlusion
α chain variant) [53]
in such circumstances can lead to splenic, pulmo-
Haemoglobin S‐Oman [54]
Haemoglobin Jamaica Plain [43] nary, pituitary, cerebral, retinal, renal and bone
infarcts and also to priapism (persistent erection
caused by sickling within blood vessels of the
but is of genetic importance. It gives partial protec- penis). Bone infarcts can lead to avascular necrosis.
tion against uncomplicated malaria, cerebral There have been at least 47 reports of splenic infarc-
malaria, severe malaria and death from Plasmodium tion in individuals with sickle cell trait related to
falciparum malaria; however the incidence of altitude [60]. Exceptionally, splenic infarction neces-
asymptomatic malaria is not decreased [55]. This sitating splenectomy was reported in a father and
protection is lost if there is coexistence of α thalas- son who ascended by road to 7200 feet [61]. Rarely,
saemia heterozygosity [56]. spontaneous unprovoked splenic infarction occurs
If a patient with symptoms suggestive of sickle [62]. Splenic infarction has been reported in a num-
cell disease appears to have sickle cell trait on hae- ber of patients with coexisting hereditary spherocy-
moglobin electrophoresis or HPLC, further detailed tosis, being attributed to sickling induced by
investigation is indicated since this can be the result hypoxia due to slow passage of spherocytes through
of a second mutation in a βS gene, such as haemo- the spleen [63]. Splenic infarction was also been
globin Jamaica Plain (see later), or a second muta- reported in one patient with coexisting sickle cell
tion in a βC gene, such as haemoglobin Arlington trait and severe pyruvate kinase deficiency [53].
Park (see later). The second mutation alters the Spontaneous splenic rupture has been reported
characteristics of the variant haemoglobin so that it [64], in one instance in association with bacterial
can be confused with haemoglobin A. endocarditis [65]. Splenic sequestration has like-
Sickle cell trait is sometimes an opportunistic wise been described, but very rarely, in sickle cell
diagnosis following detection of a variant haemo- trait [66]. There is a low risk of sudden death associ-
globin during measurement of haemoglobin A1c in ated with vigorous exercise, particularly exercise at a
patients with diabetes mellitus. Confirmation of the high altitude and exercise complicated by dehydra-
suspected abnormality without specific consent has tion and acidosis [67]. Such circumstances can also
been found to be acceptable to the great majority of lead to exertional rhabdomyolysis, disseminated
patients [57]. Conversely, an incidental diagnosis of intravascular coagulation and renal failure [68].
Rhabdomyolysis and acute compartment syn- carcinoma having also been reported in sickle cell
drome have also been reported following vigorous anaemia and sickle cell/haemoglobin C disease
exercise in the absence of complicated factors such [86]. Sickled cells have been observed in the urine in
as a high environmental temperature or dehydra- a patient with sickle cell trait being investigated for
tion [69]. In a study of US Air Force personnel, the post‐partum fever [87].
rate of non‐traumatic deaths in airmen was very During pregnancy, women with sickle cell trait may
low but was 25‐fold higher in those with sickle trait have an increased incidence of bacteriuria and pyelo-
than in those without sickle trait [70]. Similar obser- nephritis [88] and pregnancy‐associated hypertension
vations have been made in US army recruits. [89]. Other pregnancy‐associated complications
Because of the possibility of exertional sickling, the include a more than doubled incidence of post‐par-
US National Athletic Trainers Association recom- tum endometritis and a statistically, but probably not
mended that colleges and universities test student‐ clinically, significant decrease in the period of gesta-
athletes for sickle cell trait. However the US Armed tion at delivery and the birth weight [89].
Forces and the American Society for Hematology If the kidney is excluded, spontaneous episodes
prefer a policy of risk reduction in all military per- of vascular occlusion (i.e. episodes occurring in the
sonnel and athletes [71]; with this policy being fol- absence of fever, dehydration, hypoxia or acidosis)
lowed in US soldiers, the risk of exertional are very rare but do occur. One instance of a cere-
rhabdomyolysis is higher in those with sickle cell brovascular accident associated with moya moya
trait but there is no significant increase in the risk of has been reported [90]. In one study the overall risk
death [72]. Sickle cell trait has been related to a of coronary artery disease, congestive heart failure
higher rate of vascular complication in Africans and stroke was found not to be increased in a pre-
with diabetes mellitus [73]. Sickle cell trait was dominantly young population [79], but another
associated with intraflap thrombosis and sickling in long term prospective study found a significantly
a patient having breast reconstruction, cooling of the higher risk of ischaemic stroke [91].
flap being suspected as a precipitating factor [74]. In one study the incidence of deep vein thrombo-
Rarely, proliferative retinopathy has been reported [75]. sis was increased about twofold and the incidence
In sickle cell trait, spontaneous sickle cell forma- of pulmonary embolism about fourfold [92]. In
tion can occur in renal papillae where oxygen ten- another study the incidence of pulmonary embo-
sion is normally low, leading to renal papillary lism was increased 1.37‐fold [79]. D‐dimer concen-
necrosis, episodes of haematuria and impairment of tration is higher in Afro‐Americans with sickle cell
renal concentrating ability (hyposthenuria), which trait than in those without it [78].
can be associated with enuresis in children. An increased incidence of invasive pneumococ-
Haematuria can be the result of renal papillary cal disease has been reported [93].
necrosis [76]. Loss of renal concentrating ability is Because of the low pH of the anterior chamber of
less if α thalassaemia trait coexists [77]. In Afro‐ the eye, post‐traumatic hyphaemia can lead to local
Americans, the estimated glomerular filtration rate sickling, impaired outflow and increased intraocu-
is lower in individuals with sickle cell trait than in lar pressure, necessitating surgical evacuation.
those without [78]. There is predisposition to pro- Screening for sickle cell trait is therefore indicated
teinuria [76] and an increased risk of chronic kidney in any person of appropriate ethnic group present-
disease [76, 79–81], with end stage renal disease ing with a hyphaemia. Proliferative retinopathy has
being about twice as common as in Caucasians [81]. been reported in individuals with coexisting ocular
In Afro‐American US army soldiers, sickle cell trait or systemic disease [94] but not otherwise.
is associated with a higher incidence of acute kid- Despite this list of potential complications the
ney injury as well as chronic kidney disease [82]. great majority of patients with sickle cell trait are
Coinheritance of alpha thalassaemia protects asymptomatic and the main reason for seeking to
against chronic kidney disease [83]. There is also a identify the heterozygous state is the genetic impli-
quite strong association between sickle cell trait and cations. If both parents have sickle cell trait there is
medullary carcinoma of the kidney [84, 85], this rare a 25% probability of sickle cell anaemia in a child.
194 Chapter 4
Laboratory features are seen; this has been reported as an in vitro

a rtefact in a child with acute lymphoblastic leukae-
Blood count. The haemoglobin concentration (Hb)
mia with a high white cell count (WBC), being
is normal except in those with coexisting α thalas-
attributable to consumption of oxygen by the leu-
saemia trait who may be slightly anaemic [95, 96].
kaemic cells [103].
The mean cell volume (MCV) and mean cell
During P. falciparum malaria, subjects with sickle
haemoglobin (MCH) are reduced in those with
cell trait have a blood film showing a lower percent-
coexisting α thalassaemia trait (Table 4.4) [12, 95,
age of parasitised cells than is seen in subjects with-
97–100]. However, in a large survey carried out in
out a haemoglobinopathy [104].
Kenyan children, MCV and MCH were also
reduced and the red cell count (RBC) was Other investigations. It might be expected that hete-
increased in those with four α genes [96]. In this rozygotes for haemoglobin S (ββS) would have
same Kenyan study the presence of sickle cell trait equal amounts of haemoglobin S and haemoglobin
was found to ameliorate the effect of α thalassae- A. In fact haemoglobin A is somewhat more than
mia trait on the RBC, Hb, MCV and MCH [96]. α 50% and haemoglobin S is somewhat less, usually
thalassaemia trait is somewhat more prevalent in around 40%; this is because normal β chain has a
Africans and Afro‐Americans with sickle cell trait greater affinity than βS for α chains. Haemoglobin
than in those with normal β globin genes [101], so electrophoresis at alkaline or acid pH shows a vari-
it is not rare for s ubjects with sickle cell trait to ant haemoglobin with characteristic mobility
have borderline anaemia or reduction of the MCV (Fig. 4.4). Haemoglobins D and G have the same
and MCH. electrophoretic mobility at alkaline pH but can be
distinguished by electrophoresis at acid pH, which
Blood film. The blood film may be completely nor- usually shows mobility that is the same as, or very
mal or may show microcytosis or target cells similar to, that of haemoglobin A. Haemoglobin S
(Fig. 4.3). If a subject with sickle cell trait develops can also be distinguished from haemoglobin A, D
iron deficiency, target cells are often prominent. and G by isoelectric focusing and by HPLC (Fig. 4.5,
Although classical sickle cells are only very rarely see also Fig. 2.19) and capillary electrophoresis
seen, small numbers of plump cells that are pointed (Fig. 4.6). A sickle solubility test (see Fig. 2.25)
at both ends have been reported [102]; such cells should always be performed when the presence of a
were described in about 96% of individuals with significant proportion of haemoglobin S is sus-
sickle cell trait in comparison with 4% of normal pected. It will be positive except in the early neona-
subjects. In very unusual circumstances, sickle cells tal period when the percentage will be below the
Table 4.4 The red cell indices

and percentage of haemoglobin S 𝝰𝝰/𝝰𝝰 −𝝰/𝝰𝝰 −𝝰/−𝝰 Method Reference
reported in sickle cell trait
with and without deletion of α Haemoglobin >38 31–38 <31 DEAE‐cellulose 37
S percentage Chromatography
globin genes [12, 95–99].
35–45* 30–35* 25–30* Not stated* 95
34–38 28–34 20–28 12
31–43 29–35 22–29 HPLC 98
MCV (fl), 80–90 75–85 70–75 Not stated* 95

range or
84.9 (8.26) 79.1 (5.8) 67.8 (5.81) HPLC 98
mean (SD)
* These figures are averages derived from published series and in which the α chain deletion
was −α3.7; the quantification is likely to have been by electrophoresis; the deletion −α4.2 leads
to a greater reduction in haemoglobin S percentage [99].
(a)
(b)
Fig. 4.3 Blood films of three patients

with sickle cell trait showing the
range of features observed:
(a) normal film; (b) minimal
anisocytosis and poikilocytosis with
occasional target cells;
(c) hypochromia with occasional
target cells and other poikilocytes.
May–Grünwald–Giemsa (MGG)
×100 objective. (c)
196 Chapter 4
excluded from the estimate. The percentage

shows a trimodal distribution, depending on
whether there is coexisting α thalassaemia trait
(see Table 4.4). If there is coexisting haemoglobin
H disease the percentage of haemoglobin S is
even lower, around 10–25% [99]. Conversely, if
there are five α genes (ααα/αα) the haemoglobin
S percentage is somewhat higher than in those
with a normal complement of α genes [106],
about 45% rather than about 40%. A rare cause of
a low haemoglobin S percentage (11%) in sickle
cell trait is coinheritance of a β thalassaemia
determinant in cis (i.e. on the same chromosome);
haemoglobin S may then be as low as 10% [107,
108]. A similarly low haemoglobin S percentage
in a heterozygote (12%) has been reported as the
Fig. 4.4 Haemoglobin electrophoresis on cellulose result of somatic deletion of the βS allele of the
acetate at alkaline pH showing haemoglobins A and S in
HBB gene in about 50% of cells [109]. A further
a patient with sickle cell trait (lane a); AFS indicates a
rare cause of a low S percentage (13.5%), observed
control sample containing haemoglobins A, F and S.
in a patient with deletion of a single α gene, was
duplication of the entire β globin cluster in trans
[110]. The percentage of haemoglobin S corre-
detection limit. It follows that a negative sickle solu- lates with the MCV and MCH since all these vari-
bility test in a neonate with a variant haemoglobin ables are influenced by coexisting α thalassaemia
consistent with haemoglobin S by no means trait (see p. 194). The proportion of haemoglobin
excludes a diagnosis of sickle cell trait. Reagents S is also reduced if there is coexisting iron defi-
suitable for a sickle solubility test are commercially ciency [111], has been observed to fall markedly
available in kit form [105]. in megaloblastic anaemia [112] and may be low
Haemoglobin S can also be distinguished from in lead poisoning [113].
haemoglobin A and from haemoglobin C by immu- The percentage of haemoglobin A2 may be
nological techniques based on monoclonal antibod- slightly elevated in sickle cell trait. In addition, the
ies to sequences including the amino acids that are A2 fraction on HPLC is artefactually increased
substituted in haemoglobin S and haemoglobin C. by the presence of carbamylated [114] or other
It is thus possible to distinguish sickle cell trait from post‐translationally modified haemoglobin S.
sickle cell anaemia and from sickle cell/haemoglo- Quantification of haemoglobin A2 is not a particu-
bin C disease. However sickle cell/β+ thalassaemia larly useful investigation to perform in an individ-
may not be distinguished from sickle cell trait. Such ual with sickle cell trait so this is not of practical
monoclonal antibodies were previously commer- importance. In one study, haemoglobin F was
cially available in kit form as HemoCard A plus S slightly increased with a mean of 1.4%; F cells were
and HemoCard C [105]. a mean of 14.1% in sickle cell trait in comparison
The proportion of haemoglobin S can be quanti- with a mean of 2.8% in haematologically normal
fied by scanning densitometry of an electrophoretic Afro‐Americans [115].
strip, by elution from an electrophoretic strip, by In the neonatal period haemoglobin F will be pre-
capillary electrophoresis or by HPLC. It should be sent in large amounts. There will be more haemo-
noted that when the haemoglobin S percentage is globin A than haemoglobin S. However if
estimated by HPLC rather than electrophoresis, haemoglobins S and A are present in small amounts
there is underestimation as glycated haemoglobin S then precise quantification may be difficult. Unless
and other post‐translationally modified S is there is clearly more haemoglobin A than S then
(a)
Fig. 4.5 High performance liquid chromatography (HPLC) chromatogram (Bio‐Rad Variant II): (a) sickle cell trait in an
adult showing, from left to right, haemoglobin F (shaded), two peaks of post‐translationally modified haemoglobin A,
haemoglobin A0, haemoglobin A2 (shaded and apparently increased) and haemoglobin S.; (Continued on p. 198.)
sickle cell/β+ thalassaemia is a possible alternative

Diagnosis
diagnosis. If necessary, the test should be repeated
when the infant is a few months old. Diagnosis rests on demonstration of the presence
Electrophoretic features suggestive of sickle cell of haemoglobin S and haemoglobin A with the
trait despite clinical features of sickle cell disease percentage of haemoglobin S being less than the
should lead to investigation for an electrophoreti- percentage of haemoglobin A. The haemoglobin S
cally silent variant haemoglobin that may be inter- identification must be supported by two independ-
acting with haemoglobin S [51]. ent tests.
198 Chapter 4
(b)
Fig. 4.5 Continued. (b) sickle cell trait in a neonate, showing haemoglobin F0 82.1%, haemoglobin A0 10.5%,
haemoglobin A2 0.4% (the low level being expected in a neonate) and haemoglobin S 8.9%. The post‐translationally
modified F (complex peaks at left) has not been integrated.
Interactions of haemoglobin S count [116]. The blood film shows hypochromia,

heterozygosity with thalassaemias microcytosis and marked poikilocytosis, including
and haemoglobinopathies target cells (Fig. 4.7). Haemoglobin S is lower than is
The interaction of sickle cell trait and α thalassaemia usual in sickle cell trait with coexisting α thalassae-
trait has been discussed earlier. The coexistence of mia trait (Fig. 4.8) [52, 116]. Haemoglobin Bart’s may
sickle cell trait and the genotype of haemoglobin H be present in infancy but haemoglobin H is not
disease leads to a modification of the phenotype of detected and only occasional inclusion‐containing
the haemoglobin H disease. There is a hypochromic cells are found on a haemoglobin H preparation.
microcytic anaemia with splenomegaly and eryth- Inclusions have been reported in bone marrow
roid hyperplasia but with a normal reticulocyte erythroblasts. It could be speculated that these
Fig. 4.6 Capillary electrophoresis electropherogram (Sebia Capillarys 3) from a patient with sickle cell trait, showing
haemoglobins A, S and A2.

sickle cell trait and the genotype of
haemoglobin H disease
(homozygosity for Saudi non‐deletional
α thalassaemia (AATAAA→AATAAG at
the polyadenylation site, αTSaudiα/αTSaudiα/)).
Full blood count (FBC) showed: red
blood cell count (RBC) 5.83 × 1012/l,
haemoglobin concentration (Hb)
97 g/l, haematocrit (Hct) 0.31, mean
cell volume (MCV) 54 fl, mean cell
haemoglobin (MCH) 16.6 pg, mean cell
haemoglobin concentration (MCHC)
313 g/l. MGG ×100.
200 Chapter 4
Fig. 4.8 HPLC chromatogram (Bio‐Rad Variant II) in sickle cell trait plus haemoglobin H disease showing the typical
double peak of haemoglobin H and a low haemoglobin S percentage (21%). (Same patient as Fig. 4.7.)
represent βS precipitates, βA having combined pref- coinheritance of sickle cell trait and αG‐Philadelphia is
erentially with the reduced numbers of α chains. associated, on electrophoresis at alkaline pH,
The coinheritance of βS and various β chain var- with the presence of three bands with the mobil-
iants, β and δβ thalassaemias leading to a clinical ity of A, S (representing S and G‐Philadelphia)
abnormality is discussed later. Other compound and C (representing an S–G hybrid) (Fig. 4.9). On
heterozygous states may be asymptomatic. The agarose gel at acid pH there are two bands, a
interaction of sickle cell trait and α chain variants band with the mobility of haemoglobin A (repre-
is generally clinically silent but leads to extra bands senting A plus G‐Philadelphia) and a band with
on haemoglobin electrophoresis. For example, the mobility of haemoglobin S (representing S

on agarose gel at pH 8.6 showing three
bands with the mobilities of A, S and
C, in a patient with sickle cell trait and
heterozygosity for haemoglobin
G‐Philadelphia (fifth lane from left);
AFSC indicates a control sample
containing haemoglobins A, F, S and C.
Fig. 4.10 HPLC chromatogram in a

patient with heterozygosity for both
haemoglobin S and haemoglobin
G‐Philadelphia, showing, from left
to right, post‐translationally
modified haemoglobin A (two
peaks) and haemoglobins A0, A2, S,
G‐Philadelphia and hybrid (ααGββS).
and S–G hybrid). On HPLC there are four fractions splenic sequestration or splenic infarction [118]. It
distinguished by their retention times (Fig. 4.10). is likely that the increased haemoglobin concen-
A β thalassaemia mutation occurring in cis to a βS tration within red cells as a result of the hereditary
mutation differs considerably from sickle cell trait. spherocytosis favours sickling within the spleen.
Two individuals with this combination had haemo- Sickling can also occur when severe pyruvate
globin S of 10–11%, haemoglobin A2 of 6–7%, hae- kinase deficiency coexists with sickle cell trait;
moglobin F of around 3% and a mild microcytic this is likely to be the result of a high intracellular
anaemia with a reticulocyte count of around 3% concentration of 2,3‐diphosphoglycerate (2,3‐DPG),
[108, 117]. favouring deoxygenation [53].
Interactions of haemoglobin S heterozygosity Sickle cell anaemia

with other haematological conditions
Sickle cell anaemia is the disease resulting from
The coexistence of sickle cell trait and hereditary homozygosity for haemoglobin S. Individuals
spherocytosis has been reported in at least 19 with sickle cell anaemia have haemoglobin S as
instances. Four of these individuals had either the major haemoglobin component, with a small
202 Chapter 4
proportion of haemoglobin A2 and a variable Democratic Republic of the Congo and India
proportion of haemoglobin F. Since there is no [120]. Median survival in the UK has been esti-
synthesis of normal β chain there is a total mated at 67 years [121] and in the USA at 48 years
absence of haemoglobin A. Red cells sickle due or 66 years, depending on the statistical meth-
to polymerisation of haemoglobin S under condi- ods employed [122]. Large cohort studies from
tions of low oxygen tension but initially this pro- birth will be needed to establish accurate
cess is reversible as the cells pass through the figures.
lungs. The process is cyclical but eventually A study in California found an increased inci-
membrane damage leads to the red cell becom- dence of leukaemia in patients with sickle cell dis-
ing irreversibly sickled. The irreversibly sickled ease (not all patients necessarily having sickle cell
cell has an increased calcium content, which trig- anaemia), with a reduced incidence of breast cancer
gers Ca‐dependent potassium transport and loss and male genital cancer [123].
of potassium and water. K+/Cl− co‐transport is
also increased. The dehydrated cell becomes
Clinical features
even more rigid. The red cell membrane is dam-
aged by oxidation and the effects of repeated The clinicopathological features of sickle cell anae-
polymerisation with clustering of band 3 protein mia result directly or indirectly from haemolysis
and externalisation of phosphatidyl serine. There and vascular obstruction by sickled red cells with
is reduced red cell deformability, increased fra- consequent tissue infarction. In addition to the
gility, microvesiculation and intravascular hae- shape change, erythrocytes show increased adhe-
molysis. Intravascular haemolysis in turn leads sion to endothelium, which contributes to vascular
to depletion of nitric oxide (NO), which is fur- occlusion. Neonates are asymptomatic since a
ther aggravated by release of red cell arginase, major part of the total haemoglobin is haemoglo-
which reduces availability of arginine, the sub- bin F. As synthesis of haemoglobin F decreases and
strate for NO synthesis; there is resultant loss of synthesis of haemoglobin S increases, symptoms
NO vasodilator function. Immunoglobulin G start to appear, usually from 6 months of age. In
binds to the damaged red cell membrane and infants, bony infarction leads to avascular necrosis
this, plus red cell rigidity, leads to phagocytosis of small bones of the hands and feet which pre-
by macrophages (extravascular haemolysis). sents clinically as painful swelling of fingers and
Sickle cells show increased interaction with toes (dactylitis or ‘hand‐foot syndrome’). This can
endothelium, particularly when adhesion mole- lead to failure of growth of a phalanx and later
cules are upregulated, and also increased bind- shortening of a digit (Fig. 4.11). In children there
ing to neutrophils and platelets. There is a can be splenomegaly and occasionally there is
prothrombotic state mediated by the altered red hypersplenism. Young children can also suffer
cell membrane, damaged endothelium, activated from splenic sequestration in which pooling of red
neutrophils, monocytes and platelets, and NO cells in a rapidly enlarging spleen leads to acute
depletion. There is an increased incidence not anaemia. Splenic sequestration is not uncommonly
only of arterial thrombosis but also of venous followed by hypersplenism [124]. Atraumatic
thromboembolism. splenic rupture has been reported [125–127].
Red cell survival is around 20 days in patients, Hepatic sequestration also occurs but is less com-
with no increase in haemoglobin F but can be con- mon. Cerebral haemorrhage and infarction is a
siderably longer when haemoglobin F is increased; particular feature of children with sickle cell anae-
for example, in six patient with haemoglobin F mia, as a result of prior endothelial damage. Silent
of more than 10% as a result of hydroxycarbamide cerebral infarcts also occur, with the frequency
(hydroxyurea) therapy, red cell survival was being higher with a lower Hb and also correlating
16–54 days (mean 34 days) [119]. with previous parvovirus B19 infection [128].
Sickle cell anaemia occurs worldwide but the There is also an increased incidence of intracranial
majority of affected births occur in Nigeria, the aneurysms. Sudden deafness can occur.
(a) (b)
Fig. 4.11 Long term result of ‘dactylitis’ in sickle cell anaemia: (a) the hand of an 18‐year‐old Nigerian man;
(b) radiograph of the hand. (Reproduced from Hoffbrand AV and Pettit JE. Essential Haematology, 3rd edn. Blackwell
Scientific Publications, Oxford, 1993, by kind permission of Professor Victor Hoffbrand.)
In older children and adults there continues to be Pulmonary fibrosis aggravates hypoxia. Diffuse
infarction of bones such as ribs, vertebrae and long myocardial fibrosis can occur, leading to diastolic
bones; osteonecrosis can be detected radiologically dysfunction [133]. Other cardiac complications
(Fig. 4.12a). In addition, there is infarction of internal include atrial and ventricular dilation, atrial arrhyth-
organs including the lungs (Fig. 4.12b), abdominal mias, mitral incompetence and right to left shunting,
organs and brain. The incidence of painful crises cor- either in the lung or through a patent foramen ovale,
relates with a higher haematocrit and the incidence with associated hypoxia [134]. Bone marrow infarc-
of acute chest crisis with cigarette smoking [129]. tion can be extensive and may be complicated by
Retinopathy can occur, a haemoglobin F of at least embolism of necrotic bone marrow to the lungs.
15% being protective [130]. Some clinical features of Infarction of bones can also be complicated by osteo-
sickle cell anaemia are linked specifically to haemol- myelitis, usually caused by infection of the infarcted
ysis while others are linked more to blood viscosity bone by salmonella or staphylococcus. Recurrent
(Table 4.5). Inactivation of NO by reactive oxygen infarction of the spleen leads to hyposplenism, which
species and by free haemoglobin in the plasma is in turn causes increased severity of various infec-
likely to be a major cause of pulmonary hypertension tions, including malaria and pneumococcal septi-
[131], the prevalence of which correlates with the caemia. Splenic phagocytic function is lost first and
severity of haemolysis [132]. Pulmonary infarction, then splenic filtering function [135]. In one study,
resulting from acute chest syndrome and thrombo- loss of splenic function began before 12 months of
embolism, can be associated with pulmonary seques- age in 86% of 193 babies [136]. In developed coun-
tration of red cells and platelets and, if recurrent, can tries, splenomegaly is seen mainly in young chil-
contribute to lung fibrosis, pulmonary hypertension dren (in the steady state due to congestion and
and right heart failure with tricuspid incompetence. extramedullary haemopoiesis) but in sub‐Saharan
204 Chapter 4
(a)
(b)
Fig. 4.12 Radiography in sickle cell anaemia: (a) radiograph of the head of the humerus showing areas of reduced
radiodensity, consequent on previous infarction; (b) chest radiograph showing opacities in the lower half of both lung
fields representing pulmonary infarction as a result of sickle cell formation and vascular occlusion. (With thanks to
Professor Irene Roberts.)
Table 4.5 Features of sickle cell anaemia linked to either results in an increased incidence of pancreatitis. The
haemolysis or hyperviscosity. coinheritance of Gilbert’s syndrome aggravates
hyperbilirubinaemia. Acute hepatic crisis can occur,
Linked to haemolysis Linked to hyperviscosity characterised by pain and tenderness in the right
upper quadrant, fever and vomiting [139]. Acute
Leg ulcers Painful crises intrahepatic cholestasis is an uncommon condition
Priapism Acute chest crises
with similar clinical features to acute hepatic crisis; it
Pulmonary hypertension Retinopathy
Albuminuria and chronic Osteonecrosis
can be complicated by coagulopathy and liver failure
kidney disease [139]. Increased erythropoiesis occurs as a response
Gallstones to haemolytic anaemia, leading to overexpansion of
the bone marrow cavity and a tendency to reduced
bone mineral density. In some patients erythroid
Africa splenomegaly is common in the first 10 years hyperplasia causes frontal bossing of the skull and
of life and is observed even in 10% of adolescents, malpositioned teeth. On skull radiology there can be
without there being any clear relationship to malaria thickening of the cranial bones (Fig. 4.15) and a hair‐
[137]. Persistent splenomegaly is also often observed on‐end appearance. Patients with sickle cell anaemia
with the Arab‐Indian haplotype [138]. Infarction of can suffer rapid worsening of anaemia, with Hb
the skin can result in ulceration of the legs; this is sometimes falling to 40 g/l or lower, during infection
more common in low and middle income countries by parvovirus B19. The mechanism is pure red cell
so other factors are clearly also operating. The aplasia, which is transient but, because of the short-
increased breakdown of red cells means that patients ened red cell life span, rapidly leads to anaemia. In
are intermittently jaundiced (Fig. 4.13). There is a some countries, patients with sickle cell anaemia
high incidence of pigment gallstones (Fig. 4.14) show an increased incidence of megaloblastic anae-
resultant on this chronic haemolysis, which in turn mia, which has been attributed to inadequate intake
Fig. 4.13 Face of a child with sickle cell anaemia

showing pallor and jaundice. (With thanks to Professor
Irene Roberts.)
Fig. 4.15 Skull radiograph in sickle cell anaemia
showing expansion of the bony cavity resulting from
hyperplastic erythropoiesis.
of folic acid in the face of an increased need for this

vitamin. Transfusion‐transmitted babesiosis can
cause severe haemolysis [140].
Patients with sickle cell anaemia have an increasing
incidence of renal disease with age. Features include
hyperfiltration with an increased glomerular filtra-
tion rate early in the disease course (from early child-
hood), glomerular hypertrophy, glomerulopathy
(focal segmental or membranoproliferative glomeru-
lonephritis or microangiopathic glomerulopathy)
and papillary necrosis or other renal infarction [141].
Functional consequences include haematuria, micro-
albuminuria (loss of 30–300 mg of albumin per day),
albuminuria (loss of more than 300 mg per day,
reported in 26–68% of patients [141]), nephrotic syn-
drome, loss of concentrating ability and chronic renal
failure. Hyperfiltration is predictive of the develop-
ment of microalbuminuria and it is possible that it
contributes to glomerular damage [142, 143]. An Hb
of less than 80 g/l correlates with earlier develop-
ment of microalbuminuria [144]. Leucocytosis and
the degree of haemolysis also correlate with earlier
onset of albuminuria [143].
Patients with sickle cell anaemia living at a high
altitude have a somewhat higher mean Hb and
Fig. 4.14 Cholecystogram showing gallstones (negative higher 2,3‐DPG and are considerably more sympto-
images), caused by increased bilirubin production matic than those living at low altitudes [145].
resulting from haemolysis in a child with sickle cell Patients with sickle cell anaemia with a high
anaemia. (With thanks to Professor Irene Roberts.) haemoglobin F (e.g. those with the Arab‐Indian
206 Chapter 4
haplotype) generally have a milder clinical course Table 4.6 Causes of anaemia in sickle cell anaemia.
but severe disease can occur. There is significant
retention of splenic function but the associated Causes of steady state anaemia
Haemolysis
persisting splenomegaly means that they can
Reduced oxygen affinity leading to reduced
develop hypersplenism and remain susceptible to erythropoietic drive
splenic sequestration, splenic infarction and
Causes of worsening of anaemia
splenic abscess formation into adult life [146];
Splenic, hepatic or pulmonary sequestration
splenectomy may be necessary for hypersplenism.
Hypersplenism (usually only in infants and children)
Because there is less haemolysis, there is a lower Parvovirus B19 infection
incidence of pulmonary hypertension, priapism, Suppression of erythropoiesis in other infections
leg ulcers, stroke and nephropathy [115, 147], Malaria and, rarely, babesiosis
whereas the incidence of vaso‐occlusive events is Megaloblastic anaemia resulting from folic acid
similar or higher [31, 147]. In one study, avascular deficiency
Bone marrow infarction
necrosis of the femoral head occurred in around
Hyperhaemolysis following blood transfusion
27% of patients in comparison with 8–12% in
Renal failure
patients with a lower haemoglobin F level [31].
There is some evidence that a higher haemoglobin
F percentage is associated with a higher rate of
death from falciparum malaria [148]. organ failure due to recurrent tissue infarction (e.g.
Patients who require regular or intermittent from renal or hepatic failure).
blood transfusions often develop red cell alloanti- Patients with sickle cell anaemia can have com-
bodies and patients are thus at risk of delayed plications related to treatment as well as those due
haemolytic transfusion reactions with intravascu- to the disease itself. These include transfusion‐
lar haemolysis. Such reactions may be accompa- transmitted infections and iron overload.
nied by veno‐occlusive crisis, acute chest syndrome, In many countries the survival of individuals
pulmonary hypertension and multiorgan failure with sickle cell anaemia has improved greatly in
[149]. When a delayed transfusion reaction occurs, recent decades. In one US study, published in 1994,
the Hb may fall rapidly to levels below the pre‐ median life expectancy was 42 years for men and
transfusion level. This is due mainly to destruction 48 years for women [152] and in a Jamaican study
of transfused red cells while haemopoiesis is from a similar period it was 58 years for men and 66
suppressed, but in some patients there is also
years for women [153]. By 2014, the estimated
‘bystander’ destruction of the patient’s own red median survival in a US cohort was 58 years [154].
cells [150]. In addition, transfusion may be followed Despite the severity of the disease in many patients,
by hyperhaemolysis without any evidence of red there are a significant minority who are asympto-
cell incompatibility [151]. matic for prolonged periods. For example, in one
The causes of anaemia in homozygotes for hae- French study of 299 patients, 9% were asympto-
moglobin S are summarized in Table 4.6. matic for 3 years or more [155]. In sub‐Saharan
Death in sickle cell anaemia is most often attrib- Africa the situation is very different. It has been
utable to infection, cerebrovascular accidents or res- estimated that there is a 50–90% mortality by the
piratory failure, the latter being the result of age of 5 years as a result of infection and severe
extensive sickling within pulmonary blood vessels. anaemia [156].
In the absence of parental education and vigilance,
death of infants can result from splenic sequestra-
Ameliorating factors and interaction with α
tion. In African countries it is likely that many
thalassaemia trait
deaths in infants and children with sickle cell anae-
mia are attributable to malaria and some to severe The clinical course of sickle cell anaemia is very
anaemia. Patients with sickle cell anaemia who sur- variable. This is largely unexplained although some
vive childhood and adolescence may die from end factors have been identified that appear to ameliorate
Table 4.7 Factors ameliorating or aggravating features irreversibly sickled cells and improved red cell
of sickle cell anaemia [5, 135, 152, 155, 157–163]. survival and (ii) higher Hb, leading to increased
blood viscosity. Features that are ameliorated by
Coinheritance of hereditary persistence of fetal haemoglobin,
coinheritance of α thalassaemia are those related
or other factors either linked or unlinked to the β globin
locus, leading to a high percentage of haemoglobin F
to haemolysis (priapism, leg ulceration and albu-
ameliorates; haemoglobin F level is highest in the Saudi minuria), while the increased haematocrit and
Arabian/Indian and Senegal haplotypes, lowest in the resultant increase in blood viscosity can aggra-
Central African Republic haplotype and intermediate in the vate features resulting from microvascular occlu-
Benin haplotype; haemoglobin F in the Cameroon haplotype sion. In a US study, deletion of two α genes was
is similar to that in the Benin haplotype associated with an increased prevalence of avas-
Coinheritance of certain α chain variant haemoglobins, cular necrosis, retinopathy and splenomegaly
e.g. haemoglobin Memphis or haemoglobin Hopkins II and a decreased prevalence of leg ulcers and cer-
ameliorates ebrovascular accidents [162]. In a study of sickle
Coinheritance of α thalassaemia trait – ameliorates cell anaemia associated with the Arab‐Indian
haemolysis [158]; is associated with a higher Hb [158, 159]; haplotype, a tribal Indian group with a very high
ameliorates soft tissue end organ damage [106], reduces incidence of α thalassaemia trait had signifi-
leg ulcers [157, 162], reduces the frequency of stroke [155, cantly fewer painful crises, infections and epi-
162], is associated with longer preservation of splenic
sodes of hospitalisation than a non‐tribal group
function [135] but with more splenomegaly [157, 162],
with a much lower incidence of α thalassaemia
reduces the incidence of priapism [163] – however, does
not ameliorate painful crises [159] and probably increases trait [164]. In a study in Cameroon, coinheritance
their frequency [155, 158]; decreased the frequency of acute of −α3.7 heterozygosity or homozygosity was
chest crisis in one study [157]; increases the frequency of found to be associated with later diagnosis of
retinopathy [162] and osteonecrosis [5, 162]; does not sickle cell anaemia and possibly with longer sur-
improve survival [152] vival [165]. However, in a large study of sickle
Iron deficiency [160] (ameliorates haemolysis) cell anaemia in a population with the βS gene
associated with a variety of haplotypes, overall
life expectancy was not altered by coexisting
α thalassaemia [152] so it appears likely that the
the condition and lead to later presentation, milder beneficial and adverse effects of coexisting
symptoms and a better life expectancy. Some of thalassaemia trait balance out.
these are shown in Table 4.7 [5, 135, 152, 155, 157–
163]. As detailed earlier, higher haemoglobin F per-
Laboratory features
centages protect against the complications of sickle
cell anaemia attributable to haemolysis but not Blood count. The blood count is normal at birth.
against those attributable to vascular occlusion During the first year, as haemoglobin F is replaced
[115, 147]. Elevation of haemoglobin F to more than by haemoglobin S, there is a fall of haemoglobin
20% is usually sufficient to render sickle cell anae- concentration and a rise in the reticulocyte count.
mia largely asymptomatic. Mean values differ from controls by 1–2 months of
Coexisting α thalassaemia trait is common in age [166, 167]. Anaemia and reticulocytosis continue
sickle cell disease. For example, 30% of Afro‐ throughout childhood, adolescence and adult life.
Americans with homozygosity for haemoglobin The Hb reported in adults is most often between 60
S have a single α gene deletion and 5% have two and 100 g/l but can range from 50 to 120 g/l or even
α genes deleted [161]. The effect of coexisting α higher. In a personally observed series of 29 mainly
thalassaemia trait is complex, with some features Afro‐Caribbean patients the Hb ranged from 76 to
being ameliorated and other being worsened. 138 g/l with a mean of 90 g/l. In males there is a
The complexity may be the result of two conflict- significant post‐pubertal rise in the Hb averaging
ing effects: (i) reduced polymerisation leading to between 10 and 20 g/l [12]. Patients with a higher
less membrane damage, fewer dehydrated and percentage of haemoglobin F tend to have a
208 Chapter 4
higher Hb [12]. The Hb is of some prognostic sig- [174]. The total nucleated cell count may be
nificance [152]. In infants, anaemia and a higher increased as a consequence of significant num-
reticulocyte count are predictive of a higher rate bers of circulating erythroblasts. The neutrophil
of death and stroke (increasing risk with absolute count may be increased between as well as dur-
reticulocyte counts in quartiles of <105, 105–193, ing crises. The baseline WBC has been found to
194–307, >307 × 109/l), but anaemia lost its sig- correlate with the frequency of acute chest syn-
nificance in multivariate analysis [168]. In this drome [175] and to be predictive of earlier death
study a higher WBC (which is usually only noted from sickle cell disease [152]. The baseline WBC
after the age of 6 months) was not predictive of and neutrophil count are predictive of future
outcome [168]. During complications such as deterioration of pulmonary function [176]. An
splenic sequestration, parvovirus infection or increased WBC, partly hereditable and partly
megaloblastic anaemia the Hb may fall to as low related to cigarette smoking, correlates with the
as 15–30 g/l. Bacterial infection is also associated incidence of priapism [129]. The monocyte count
with some worsening of the anaemia. In older and the lymphocyte count are also increased
patients with sickle cell anaemia a slow fall in the [177], the latter possibly as a feature of hypo-
Hb without any alteration in the red cell indices splenism. The platelet count is increased and
may be found to be the result of the onset of renal there is an increased proportion of large platelets.
failure. A fall in both the Hb and the reticulocyte Both these features are attributable to hyposplen-
count, associated with a fall in erythropoietin ism. Particularly high platelet counts are likely to
concentration, can be an early sign of the onset of be seen in the 7–8% of Afro‐Americans who are
renal insufficiency [169]. Although the reticulo- homozygous for MPL‐Baltimore, a common poly-
cyte count is elevated in sickle cell anaemia, usu- morphism in the thrombopoietin receptor in this
ally to 5–20%, it is not increased in proportion to ethnic group [178].
the reduction in Hb. This is because haemoglobin Homozygosity for α+ thalassaemia is associ-
S has a lower oxygen affinity than haemoglobin A ated with a higher RBC, Hb and haemoglobin A2
(P50 of about 35.4 mmHg [170] in comparison with percentage and a lower MCV, MCH, MCHC,
a P50 for haemoglobin A of about 26.8 mmHg) and reticulocyte count, irreversibly sickled cell
the drive to erythropoiesis is therefore less than count and haemoglobin F percentage, with hete-
would be anticipated from the Hb. For the same rozygotes having intermediate values [157].
reason, serum erythropoietin concentration is Heterozygosity for α+ thalassaemia and, even
lower than would be expected for the degree of more so, homozygosity is associated with a sig-
anaemia [171]. The increased P50 of haemoglobin nificantly lower WBC [165]. The Hb is, on aver-
S is dependent on polymerisation of the haemo- age, 10–20 g/l higher [99, 179]. The percentage of
globin, non‐polymerised haemoglobin S having a hyperdense cells is reduced. Patients with sickle
normal oxygen affinity [172]. In addition, the con- cell anaemia with a high haemoglobin F percent-
centration of 2,3‐DPG is increased in sickle cell age tend to have a higher Hb and MCV and a
anaemia [147]. In patients with no associated α lower percentage of hyperdense cells. Those with
thalassaemia the red cell indices are normal [157, the highest F levels (e.g. patients with the Saudi/
173]. However the MCV and MCH are not ele- Indian haplotype) also have a lower reticulocyte
vated in keeping with the reticulocyte count, sug- count.
gesting a relative microcytosis. The MCV tends to Coexisting iron deficiency leads to a lower Hb,
be higher in patients with a higher haemoglobin F MCV, MCH and MCHC. There is an associated
percentage [12]. The mean cell haemoglobin con- amelioration of haemolysis [160]. Patients who are
centration (MCHC) may be slightly increased and maintained on folic acid have, on average, an MCV
the proportion of cells with a high haemoglobin 4 fl lower than patients not so maintained [180]. A
concentration is increased (Fig. 4.16). The red cell high MCV is usually due to administration of
distribution width (RDW) is generally markedly hydroxycarbamide, which may not be known to the
increased and correlates with disease severity laboratory. However the possibility of coincidental
Fig. 4.16 Red cell cytogram and histograms from a Technicon H2 instrument showing an increase of hyperchromic cells
(arrows), representing irreversibly sickled cells, and increased hypochromic macrocytes, representing reticulocytes.
vitamin B12 deficiency must not be overlooked, par- pulmonary fat embolism there is leucocytosis and
ticularly in patients who are being maintained on usually a marked fall in Hb and platelet count
folic acid [181]. [183]. Irregularly contracted cells including hemi‐
Changes occur in the blood count quite early in ghosts can appear in quite significant numbers,
sickle cell crisis. There is a fall in the Hb, a rise in being indicative of hypoxia [184]. In a study of 257
the reticulocyte percentage and a rise in the MCHC, patients with sickle cell disease, 218 of whom had
RDW, haemoglobin distribution width (HDW) and sickle cell anaemia, presenting with acute chest
percentage of hyperdense cells [182]. The HDW is a syndrome, a baseline Hb above 82 g/l and a plate-
measurement of the variation in haemoglobin con- let count of above 440 × 109/l were found to be pre-
centration between individual red cells; its increase dictive of pulmonary artery thrombosis when
is a reflection of the increased number of hyper- combined with a low Paco2 and the absence of a
dense cells. Later in a crisis there is a return of precipitating factor [185]. In addition to an acute
RDW, HDW and percentage of hyperdense cells fall in the Hb, splenic sequestration is associated
towards baseline values; the percentage of hyper- with a fall in the platelet count, reticulocytosis and
dense cells may fall below baseline values, proba- sometimes normoblastaemia [124].
bly because the densest, least deformable cells are Patients whose sickle cell anaemia is treated
being preferentially trapped in the spleen and with hydroxycarbamide with a resultant increase
destroyed. The WBC and the neutrophil count in the haemoglobin F percentage show character-
increase during painful crises and the platelet istic changes in the Hb and red cell indices. The
count may also increase. When a sickle cell crisis is Hb and the MCV rise while the MCHC, percent-
complicated by an acute chest syndrome caused by age of dense cells and reticulocyte count fall. The
210 Chapter 4
WBC, neutrophil count and platelet count can specifically Howell–Jolly bodies, target cells,
fall as a consequence of the cytotoxic effect of Pappenheimer bodies, an increased platelet count,
hydroxycarbamide. increased platelet anisocytosis and sometimes an
increased lymphocyte count. Acanthocytes, which
Blood film. The blood film is usually normal at birth are usually present in small numbers in hypo-
and in the early neonatal period since the haemo- splenic individuals, are not usually a feature of
globin S percentage is relatively low, but this is not hyposplenism in sickle cell disease. There are vari-
necessarily so (Fig. 4.17). Abnormalities are usu- able numbers of nucleated red blood cells (NRBC).
ally detectable around 6 months of age (Fig. 4.18) The neutrophil count may be increased.
when occasional sickle cells, target cells and Phagocytosis of erythrocytes by monocytes or
Howell–Jolly bodies start to appear [166]. The neutrophils may be observed but is quite uncom-
majority of infants have features of hyposplenism mon. Following recovery from parvovirus‐induced
by 1 year of age [166] and circulating erythroblasts, red cell aplasia, there is a rise in the reticulocyte
sickle cells and Howell–Jolly bodies are much count; there can also be an outpouring of erythro-
more common thereafter. In an adult with sickle blasts into the peripheral blood, which can show
cell anaemia, the blood film shows a variable num- dyserythropoietic features as a result of ‘stress
ber of crescent or sickle‐shaped sickle cells erythropoiesis’ [190] (Fig. 4.20).
(Fig. 4.19a). These represent irreversibly sickled In patients with sickle cell anaemia with a high
cells that have not corrected their shape on expo- haemoglobin F, the abnormalities in the blood
sure to atmospheric oxygen. The number of sickle film are much less (Fig. 4.21). Sickle cells are less
cells is very variable, ranging from only occasional frequent and polychromasia and anaemia are less.
cells to 30–40%. They are less numerous in those The onset of features of hyposplenism is delayed.
with a lower MCHC [12]. The percentage of sick- Coexisting α thalassaemia trait has also been
led cells plus boat‐shaped cells increases with age, observed to protect against the loss of splenic func-
through childhood and adolescence, and corre- tion [135, 191] although, surprisingly, in this study
lates with severity of symptoms [186]. In addition hyposplenism was not found to be related to age or
to classical sickle cells, there are elongated cells haemoglobin F concentration [191]. Coexisting α
pointed at one or both ends (Fig. 4.19b) [187]; these thalassaemia trait (particularly homozygous α+ tha-
have been referred to as boat‐shaped or oat‐shaped lassaemia trait) is associated with a blood film
cells or as plump sickle cells. There is polychroma- showing more target cells but fewer sickle cells [12].
sia and, in some patients microcytosis and Therapy with hydroxycarbamide leads to macro-
hypochromia. Small numbers of irregularly con- cytosis, a reduction in the number of sickle cells and
tracted cells may be seen (Fig. 4.19c) and some- boat‐shaped cells and lessening of polychromasia.
times there are cells in which the haemoglobin During sickle cell crisis there is a slight worsen-
appears to have retracted into one half of the cell ing of anaemia. There may be a further elevation of
(‘hemi‐ghosts’ or ‘blister cells’); both these features the neutrophil count, left shift and an increase in the
are particularly common in patients with wide- numbers of NRBC. An increase in the number of
spread pulmonary infarction and hypoxia. These sickle cells, in comparison with the same individu-
abnormal red cells have increased density; their al’s blood film when not in crisis, has been reported
formation has been attributed to oxidant damage but not all investigators confirm this observation
leading to transcellular bonding of damaged [192]. An increase in the number of spiculated or
regions of the red cell membrane with trapping of echinocytic sickle cells has also been noted [192].
haemoglobin within pseudovacuoles [188]. Linear Irregularly contracted cells and hemi‐ghosts can
red cell fragments may be present (see Fig. 4.19c); increase in number, particularly in those with
these were first described in a patient with cold severe hypoxia and extensive sickling within the
agglutinins and a positive antiglobulin test [189] pulmonary vasculature (see Fig. 4.19c). In one scan-
but in fact they are not rare, if specifically looked ning electron microscopy study, sickle cell crisis
for. There are features of hyposplenism (Fig. 4.19d), was associated with the presence of echinocytes,
(a)
Fig. 4.17 Blood film of a neonate

with sickle cell anaemia showing:
(a) one sickle cell; (b) other
poikilocytes consistent with
reversibly sickled cells.
MGG ×100. (b)
echinocytic sickle cells, ‘blister cells’ and macro- Hb and platelet count together with the appearance
cytes but there was no increase in the number of of increasing numbers of NRBC.
non‐echinocytic irreversibly sickled cells [193]. Various other complications of sickle cell anae-
When sickle cell crisis is complicated by extensive mia may be apparent from the full blood count
bone marrow infarction there is a greater fall in the (FBC) and the blood film. Parvovirus B19 infection
212 Chapter 4
(a)
Fig. 4.18 Blood film of a child

with sickle cell anaemia: (a) at
the age of 5 months, showing
mild anisopoikilocytosis and one
sickle cell; (b) at the age of 13
years, showing more marked
anisocytosis and sickle cell
(b) formation. MGG ×100.
may be suspected when there is worsening anae- macrocytes, oval macrocytes and hypersegmented
mia with a lack of polychromasia. The platelet neutrophils; polychromasia is less than would oth-
count is also often reduced. When recovery occurs erwise be expected in a patient with sickle cell
there is initially the appearance of numerous NRBC anaemia. Patients with coexisting sickle cell anae-
in the peripheral blood followed by reticulocytosis mia and homozygosity for α+ thalassaemia who
and thrombocytosis. In splenic sequestration there develop megaloblastic anaemia may show an
is an acute fall in the Hb with subsequent reticulo- increase in the MCV and MCH but with both val-
cytosis and increasing numbers of NRBC. The ues remaining within the normal range rather than
platelet count may also be reduced. In chronic exceeding it. Because of the shortened red cell life
hypersplenism there is worsening anaemia, throm- span, megaloblastic anaemia can also have an
bocytopenia and reticulocytosis. When there is acute onset with pancytopenia and a rapidly fall-
complicating megaloblastic anaemia there may be ing Hb without macrocytosis. In patients with a
(a)
(b)
patients with sickle cell anaemia
showing the range of abnormality
observed: (a) sickle cells and
other poikilocytes; (b) sickle
cells, a boat‐shaped cell and
nucleated red blood cells;
(c) blood film during sickle cell
crisis with pulmonary infarction
and severe hypoxia showing one
sickle cell, irregularly contracted
cells, a hemi‐ghost and several
linear fragments (detached
spicules of sickle cells);
(d) minimal sickling but features
of hyposplenism – a Howell–Jolly
body, a large platelet and a target
cell. MGG ×100. (Continued on p. 214.) (c)
214 Chapter 4
(d) Fig. 4.19 Continued.
Fig. 4.20 Blood film showing

dyserythropoiesis as a feature of
‘stress erythropoiesis’ during recovery
from an aplastic crisis. MGG ×100.
delayed transfusion reaction, some spherocytes by activated macrophages. During intercurrent

are seen but it can be difficult to recognise the infections the patient with sickle cell anaemia is
morphological features of a transfusion reaction likely to show neutrophilia, left shift, toxic granu-
in a patient with sickle cell anaemia. The direct lation and sometimes an increase in the platelet
antiglobulin test is positive. Patients with sickle count; rarely, organisms (e.g. pneumococci) are
cell disease sometimes develop hyperhaemolysis found within neutrophils.
following blood transfusion, with both homolo-
gous and autologous cells being destroyed; some Other investigations. In the adult, haemoglobin
such cases have a negative direct antiglobulin electrophoresis, HPLC and capillary electrophore-
test, low reticulocyte count and a high ferritin sis show haemoglobins S, F and A2 (Figs 4.22–4.24).
and are attributable to erythrophagocytosis Haemoglobin S is the major haemoglobin present,
Fig. 4.21 Blood film of an Arab

patient with a high haemoglobin F
percentage and clinically mild
disease; there were numerous target
cells but only occasional sickle cells.
MGG ×100.
Higher percentages are seen in those who also

have α thalassaemia trait, although there is con-
siderable variation in the mean levels reported in
different series of patients (Table 4.8) [12, 161,
195–198]. Haemoglobin F is usually only slightly
elevated (see later). Inevitably, the percentage of
haemoglobin S correlates inversely with the per-
centage of haemoglobin F. The percentage of hae-
moglobin S also varies with the number of α genes
and with the MCV and MCH. Looked at in another
way, the coexistence of α thalassaemia trait with
sickle cell anaemia leads to reduction of the MCV
and MCH and a slight reduction of haemoglobin S
percentage. Free α globin chain is increased; this
correlates with a higher lactate dehydrogenase
acetate at alkaline pH showing a patient with sickle cell
(LDH), bilirubin, dense red cell percentage, MCV,
anaemia (third from bottom) with almost all the
haemoglobin being haemoglobin S; AFSC indicates a MCH and reticulocyte percentage and inversely
control sample containing haemoglobins A, F, S and C, with haemoglobin A2 percentage [199].
while other lanes show patients with either sickle cell The sickle solubility test is positive and immuno-
trait (AS) or haemoglobin C trait (AC). assays [200] demonstrate the presence of haemoglo-
bin S with no haemoglobin A.
Neonates with sickle cell anaemia usually have
usually comprising 90–95% of total haemoglobin. predominantly haemoglobin F with haemoglo-
Haemoglobin A is totally absent. Haemoglobin A2, bin S comprising only a low percentage of total
which inhibits the polymerisation of haemoglobin haemoglobin (Fig. 4.25). Sometimes only haemo-
S, can be present in normal amounts or be slightly globin F is present and repeat testing when the
elevated (usually 2–4%) [12, 157, 194]. The level is baby is a few months of age is then necessary for
influenced by polymorphism in BCL11A and the diagnosis. In the neonatal period, diagnostic con-
HBSIL‐MYB interval and shows an inverse fusion can occur not only with sickle cell/β0 tha-
relationship with haemoglobin F percentage [194]. lassaemia but also with sickle cell/β+ thalassaemia
216 Chapter 4
Fig. 4.23 Capillary electrophoresis in sickle cell anaemia (Sebia Capillarys 3) showing haemoglobins F, S and A2.
Haemoglobin F is slightly increased and haemoglobin A2 is normal.
[201] (see later). The postnatal fall in haemoglo- Bilirubin concentration is increased, the biliru-
bin F is slower in babies with sickle cell anaemia bin being mainly unconjugated. Lactic dehydro-
than in normal babies, with mean levels of about genase is increased approximately twofold; high
20% at 1 year of age [166]. levels correlate with other evidence of haemoly-
The oxygen dissociation curve shows reduced sis and with end organ vasculopathy, proba-
oxygen affinity (i.e. a right‐shifted curve and an bly because a greater degree of haemolysis is
increased P50) [202]. The right shift is less in those associated with more resistance to NO [203].

with a high haemoglobin F percentage, either as a Hyperuricaemia is common. Serum haptoglobin
feature of the disease or as a result of hydroxycarba- is usually absent and Schumm’s test for methae-
mide therapy. Resting arterial oxygen saturation malbumin can be positive. Red cell survival stud-
when not in crisis is usually greater than 95% but in ies show a half‐life of about 7–14 days, less if
patients with significant pulmonary damage may there is splenomegaly. Heterozygous α+ thalas-
be reduced (e.g. to 80–95%). saemia is associated with longer red cell survival.
Studies of globin chain synthesis show balanced Coexisting iron deficiency leads to a considerable
synthesis of α and βS globin chains unless there is improvement in red cell survival, associated with
coexisting α thalassaemia trait. a fall in bilirubin concentration and LDH [160].
Fig. 4.24 Capillary electrophoresis in sickle cell anaemia (Sebia Capillarys 3) showing haemoglobins F, S and A2.
Haemoglobin F is moderately increased as a result of hydroxycarbamide therapy while haemoglobin A2 is normal.
Hydroxycarbamide therapy also leads to improved

Haemoglobin F percentage
red cell survival and reduced biochemical evidence
of haemolysis [204]. In sickle cell anaemia, haemoglobin F is usually
A bone marrow aspirate shows erythroid around 5–10% but may be higher, sometimes
hyperplasia and the presence of sickle cells comprising up to 40% of total haemoglobin (see
(Fig. 4.26). Dyserythropoiesis can be present. Table 4.8). The level is higher in infancy and tends
Macrophages are increased and may contain sick- to be higher in women than in men [174]. The switch
led cells (Fig. 4.27). Vaso‐occlusive crises can be from γ to βS synthesis is delayed compared with the
complicated by haemophagocytic lymphohistio- γ to β switch in normal subjects. The percentage of
cytosis and the bone marrow then shows hae- haemoglobin F falls most rapidly in the first 3 years
mophagocytosis [205]. As a result of increased of life and then more slowly till the age of 10 years;
cell turnover, foamy macrophages and sea‐blue by this time levels approximate those in adult life
histiocytes can be increased and pseudo‐Gaucher although there may be a continued slow fall up to
cells are sometimes numerous [206] (Fig. 4.28). A the age of 20 years. The percentage of haemoglobin
trephine biopsy section likewise shows erythroid F in an individual is determined by factors related
hyperplasia and sickle cells inside macrophages and unrelated to the β globin gene cluster. There is a
and within blood vessels (Fig. 4.29). clear relationship to the haplotype associated with
218 Chapter 4
Table 4.8 Haematological characteristics and percentages of various haemoglobins in adults with sickle cell anaemia
and other conditions with haemoglobins S, F and A2 only. (Derived from references 12, 161, 195–198, and other sources.)
Genotype Usual Hb (g/l) Usual MCV (fl) Usual reticulocyte Usual Hb F Usual Hb A2
count (%) percentage percentage
SS 60–100 70–100 5–20 Usually 5–10 but 1.6–3.6† (occasionally

up to 40* up to 5%)‡
2.4–4.9§
Sβ0 70–110 60–80 8–9 5–15 4–5.6
4.0–5.3§
S/δβ0 100–120 76–83 2–4 15–25 1.9–2.3
S/HPFH >120 68–88¶ Normal 20–30 1.1–2.2 in the majority
Hb, haemoglobin concentration; HPFH, hereditary persistence of fetal haemoglobin; MCV, mean cell volume.
* Influenced by coinheritance of non‐deletional HPFH as well as by the haplotype associated with the βS gene [196, 197]: Saudi‐Indian
haplotype, 10–25% Hb F; Senegal haplotype, 7–10% Hb F; Benin or Bantu haplotype, 6–7% Hb F; Cameroon haplotype, 5–6% Hb F.
Mean (and SD) of 6.06 (± 4.23) for 120 SS adults in the UK [195].
† Some overlap occurs, particularly when coexisting homozygous α+ thalassaemia raises the A2 percentage in cases of SS [12]; in one
series reported mean A2 levels were 2.8% with four α genes, 3.3% with three α genes and 3.8% with two α genes [196]; in another series
reported levels were higher – 3.5%, 3.7% and 4.9%, respectively [161].
‡ High levels are characteristically seen in the Arab‐Indian mutation.
§ When measured by capillary electrophoresis [198].
¶ Normal if there is no coexisting α thalassaemia trait.

(Bio‐Rad Variant II) in a neonate
with sickle cell anaemia showing
mainly haemoglobin F0 (dark grey),
total absence of haemoglobin A and
presence of haemoglobin S;
haemoglobin S was 5.5% and the
retention time was 4.46 minutes; the
peaks to the left of haemoglobin F0
represent post‐translationally
modified haemoglobin F.
the βS gene, averaging 10%, 7%, 7% and 5%, respec- in adults whereas adults with the Bantu and Benin
tively, in the Senegal, Benin, Cameroon and Bantu haplotypes averaged around 6–7%, but with a wide
haplotypes and 15–23% in the Arab/Indian haplo- range; the Cameroon haplotype had haemoglobin F
type [207]. In other studies, the Senegal haplotype averaging 5–6% [196, 197]. Similarly, in two further
was associated with a haemoglobin F level of 7–10% studies, the Arab‐Indian haplotype was associated

sickle cell anaemia showing
erythroid hyperplasia and two sickle
cells. MGG ×100.

sickle cell anaemia showing a
sea‐blue histiocyte packed with
sickle cells. MGG ×100.
with a haemoglobin F level of 10–25% [197] and haplotype are linked with the common −158 Gγ
18–41% [115], respectively. Saudi Arabs who have C→T polymorphism [208], whereas the Benin hap-
African haplotypes (74% Benin, 22% Bantu) have a lotype, which has a low percentage of haemoglo-
haemoglobin F about twice that of Afro‐Americans bin F, is not linked with −158 Gγ C→T. The Senegal
[115]. Arabs with the Senegal haplotype also have a haplotype is also associated with polymorphism
higher haemoglobin F percentage than individuals in the promoter of HBG2 [120]. A polymorphism at
of African ancestry [207]. A
γ IVS2 has also been linked to the high haemoglo-
The relationship of haplotype to haemoglobin F bin F level observed when the βS gene is associated
percentage appears to result from an association with the Senegal and Arab‐Indian haplotypes
between haplotype and determinants of non‐dele- [209]. In addition, the Arab‐Indian haplotype
tional hereditary persistence of fetal haemoglobin. is associated with a polymorphism at −530 bp
Both the Arab‐Indian haplotype and the Senegal (where there is (AT)9T5 rather than (AT)7T7, causing
220 Chapter 4
Fig. 4.28 Trephine biopsy section in

sickle cell anaemia showing numerous
pseudo‐Gaucher cells. Periodic
acid–Schiff stain ×40.
Fig. 4.29 Trephine biopsy section in

sickle cell anaemia showing
erythroid hyperplasia and two
vessels packed with sickle cells.
Haematoxylin and eosin ×40.
increased affinity for a BP‐1 (a negative trans‐act- with a high haemoglobin F in association with the
ing factor) and repression of βS synthesis). The Saudi Arabian or Senegal haplotype [192], while
higher haemoglobin F in sickle cell anaemia asso- the Gγ promoter associated with the Bantu haplo-
ciated with the Arab‐Indian haplotype, in compar- type has been shown to be associated with low Gγ
ison with that in the Senegal haplotype, may be synthesis [210]. In Saudi but not Indian subjects,
related to the combined effect of the (AT)x(T)y poly- increased haemoglobin F is partly related to
morphism and the −158 Gγ C→T and Aγ IVS2 poly- polymorphism of the ANTXR1 gene acting in

morphisms. The Gγ:Aγ ratio is increased in those trans [211]. Increased haemoglobin F in Brazilian
patients, associated with a reduced rate of sickle‐ When haemoglobin electrophoresis is the pri-
related complications, has been linked to enhancer mary technique, it is important not to misdiagnose
polymorphisms in BCL11A and MYB [212]. compound heterozygous states for S and D, G,
Polymorphisms in the HBS1L‐MYB intergenic Korle Bu or Lepore as sickle cell anaemia.
region also contribute. Recognising the presence of D, G, Korle Bu or
The percentage of F cells (i.e. of cells containing Lepore in compound heterozygous states with
haemoglobin F) is increased in sickle cell anae- haemoglobin S is more complex than recognising
mia. In one study the mean count was 55% (range the simple heterozygous state since all of these
17–94%), the normal level being 0.5–7% [213]; the have a single band on cellulose acetate electropho-
log of the haemoglobin F concentration correlated resis and a positive sickle solubility test (whereas
with the percentage of F cells. In another study the simple heterozygous states for any of these
the X‐linked F cell production locus was found to variant haemoglobins may simulate sickle cell trait
be the major determinant of haemoglobin F per- on electrophoresis at alkaline pH but are easily
centage in patients with sickle cell anaemia in distinguished since the sickle solubility test is
association with the three major African haplo- negative).
types [214]. Factors linked to the β gene haplo-
type were next most important. The effect of the
Interactions of haemoglobin S
X‐linked F cell locus may be the reason that
homozygosity with other thalassaemias,
women with sickle cell anaemia, like haemato-
haemoglobinopathies and other inherited
logically normal women, tend to have a higher
erythrocyte abnormalities
haemoglobin F level than men.
Individuals with coexisting α thalassaemia The modification of sickle cell anaemia by coin-
trait have been observed to have significantly heritance of α thalassaemia trait or non‐deletional
higher proportion of haemoglobin F in the first hereditary persistence of fetal haemoglobin has
decade of life [179] but thereafter have a some- been discussed earlier.
what lower proportion than those with four α Coinheritance of certain α chain variants, includ-
genes [157, 215]. ing haemoglobin Korle Bu, haemoglobin Memphis
The haemoglobin F percentage in sickle cell and haemoglobin Hopkins II, ameliorates sickle cell
anaemia is of prognostic significance [152], the anaemia.
prognosis being more favourable when the Coinheritance of other α chain variants, for
percentage is high. The haemoglobin F percent- example haemoglobin G‐Philadelphia and hae-
age is increased 2‐ to 16‐fold by hydroxycarba- moglobin Stanleyville II, has no significant effect
mide therapy [204]. on the clinical or haematological features of
sickle cell anaemia [216, 217]. The results of hae-
moglobin electrophoresis and HPLC may be
complex. With sickle cell anaemia and haemo-
Diagnosis
globin G‐Philadelphia there are two bands: an S
Diagnosis rests on the demonstration of haemo- band and a G‐Philadelphia/S hybrid band,
globins S, F and A2 only, with the presence of hae- which has the same mobility on alkaline pH as
moglobin S as the sole variant haemoglobin being haemoglobin C. The proportion of haemoglobin
confirmed by at least two independent tech- S is greater than the proportion of the hybrid
niques. In patients with microcytosis or with a band [216]. At acid pH there is a single band
significant increase of haemoglobin F, the possi- with the mobility of S, since at this pH the hybrid
bility of compound heterozygosity for S and β0 or has the same mobility as S. Coinheritance with
δβ0 thalassaemia or S and deletional hereditary the α chain v ariant Hb Montgomery also pro-
persistence of fetal haemoglobin, respectively, duces a hybrid band that has characteristics
must be considered before a diagnosis of sickle resembling those of h aemoglobin C both on cel-
cell anaemia is made. lulose acetate electrophoresis and HPLC [218].
222 Chapter 4
HPLC can show not only hybrid peaks but also, Variant haemoglobins in which there is a s econd
in the case of an α chain variant, a haemoglobin mutation in the βC gene are likely to interact with
A2 variant. haemoglobin S in a similar manner to haemoglo-
Glucose‐6‐phosphate dehydrogenase (G6PD) bin C itself. One such haemoglobin is haemoglo-
deficiency is common in many of the ethnic bin Arlington Park, β6Glu→Lys, 95Lys→Glu, which will
groups who carry the βS gene. Coinheritance of be missed on cellulose acetate electrophoresis
G6PD deficiency is associated with a lowering of at alkaline electrophoresis since there is no net
Hb on average by 10 g/l [158, 219], with infants charge change in comparison with haemoglobin A
showing a higher reticulocyte count [219] and a [222, 223].
higher rate of acute anaemic events [219]; effects
lessen after 2 years of age, probably because a fall
Clinical features
of haemoglobin F percentage is associated with a
higher reticulocyte count and therefore a higher Sickle cell/haemoglobin C disease leads to a chronic
concentration of G6PD [219]. haemolytic anaemia and to intermittent sickle cell
crises, similar to those of sickle cell anaemia but less
frequent. Dactylitis is quite uncommon [224]. The
Sickle cell/haemoglobin C disease
Hb is higher than in sickle cell anaemia and the
Sickle cell/haemoglobin C disease is consequent degree of haemolysis is less; the higher Hb is mainly
on coinheritance of βS and βC genes. There is no due to a smaller reduction in the oxygen affinity
normal β gene and therefore no haemoglobin A. rather than to less severe haemolysis. Aseptic necro-
This compound heterozygous state leads to a sis (Fig. 4.30) and bone marrow infarction, with
sickling disorder that is similar to sickle cell anae- embolism of necrotic bone marrow to the lungs, are
mia but on average is somewhat less severe. more common than in sickle cell anaemia. In one
Although most diagnoses are made in childhood series of patients, 15% suffered osteonecrosis of
or adolescence, some patients have few symp- femoral or humoral heads or vertebral bodies and
toms and around a quarter of cases are diagnosed two of 284 patients died of bone marrow embolism
in adult life [220]. The degree of haemolysis is (two of 25 deaths) [224]. Bone marrow necrosis and
less than in sickle cell anaemia, with red cells sur- fat embolism syndrome is associated with bone
viving around 27–29 days, in comparison with pain and can lead to acute renal failure [225]. Retinal
around 15–17 days. Life expectancy is considera- disease (retinitis proliferans and vitreous haemor-
bly better than that for sickle cell anaemia. In the rhages) is more frequent and more severe; in one
USA the average survival reported in 1994 was 60 series of patients it was seen in 21% and in another
years for men and 68 years for women [152], and in 23% [224]. In a third series of 15 patients, retinal
by 2014 median survival was estimated at 66 disease was seen in 34% of patients; retinopathy
years [154]. Another US study in 2019, estimated correlated with a higher Hb in this cohort [226] and
median survival at 55 or 62 years, depending on in a fourth series of patients [227]. Sensorineural
the statistical method applied [122]. deafness and vestibular symptoms can occur [220,
Sickle cell haemoglobin C disease is character- 226]. Splenomegaly persists for longer so that
ised by increased density of red cells, which is splenic infarction and splenic sequestration can
attributable to an increased K+/Cl− co‐transport occur in adults as well as children while the onset
with loss of intracellular potassium and cellular of hyposplenism, resulting from recurrent splenic
dehydration [221]. This, in turn, increases the infarction, is delayed. In contrast to patients with
likelihood of polymerisation of haemoglobin S sickle cell anaemia, in whom the spleen is atrophic
and, together with the higher haemoglobin S per- as the result of infarction, patients with sickle cell/
centage (averaging 50% rather than 40%), helps to haemoglobin C disease occasionally have splenic
explain why the compound heterozygous state infarction during aeroplane flights [228]. As a
generally causes significant disease, whereas consequence of the delay in the development of
sickle cell trait does not [221]. hyposplenism, life‐threatening infections are less
Fig. 4.30 Radiograph of hips and

pelvis in a patient with sickle cell/
haemoglobin C compound
heterozygosity showing osteonecrosis
of one hip resulting from vascular
occlusion by sickle cells; the other
hip has already been replaced
because of the same process. (With
thanks to Professor Irene Roberts.)
common than in sickle cell anaemia. However,

Laboratory features
splenomegaly can lead to hypersplenism with
chronic thrombocytopenia [229]. Priapism is com- Blood count. The Hb is higher than in sickle cell
mon [227] but is less frequent than in sickle cell anaemia, ranging from about 80 g/l up to the top of
anaemia [230]. Painful crises, leg ulcers and renal the normal range [104]. In a personally observed
impairment [220] are less common than in sickle series of 29 patients the range was 89–156 g/l with
cell anaemia. Growth retardation, which can be a a mean of 122 g/l. In one large Brazilian series the
feature of sickle cell anaemia, is not seen in sickle mean Hb was 109 g/l in women (102–120 g/l) and
cell/haemoglobin C disease [231]. Because the 127 g/l (117–137 g/l) in men [226]. In another
red cell life span is considerably longer than in Brazilian series, the mean Hb (and interquartile
sickle cell anaemia, clinically apparent parvovirus‐ range) in 361 children and adults with sickle cell/
induced aplastic crises are uncommon and gall- haemoglobin C disease was 110 g/l (104–118 g/l) in
stones are less common. Fat embolism, which is comparison with 81 g/l (74–89 g/l) in 638 patients
more common than in sickle cell anaemia, can lead with sickle cell anaemia who were not taking
to worsening neurological, pulmonary and renal hydroxycarbamide [230]. In a Canadian cohort of
dysfunction, worsening anaemia, thrombocytope- 104 adults, median Hb was 119 g/l [227]. A concen-
nia, an increased LDH and the presence of a tration of 100 g/l gives a fairly good separation
leucoerythroblastic blood film with schistocytes; between sickle cell anaemia and sickle cell/haemo-
magnetic resonance imaging of the brain shows globin C disease. In one large study of adult patients
multiple hypodense punctate lesions [232]. originating in North Africa, West Africa and the
Sickle cell/haemoglobin C disease is a hyper- Caribbean area, there were no individuals with
coagulable state, with an increased incidence sickle cell/haemoglobin C disease with an Hb less
of venous thromboembolism as well as arterial than 110 g/l [234]. In children, splenomegaly has
thrombosis and pregnancy‐associated complica- been associated with a lower Hb and platelet count
tions [233]. [229]. The MCV is lower than in sickle cell anaemia
224 Chapter 4
with a mean level around the lower limit of the patients with sickle cell/haemoglobin C disease. In
normal range [226, 235, 236]. The MCH is similar, contrast to sickle cell anaemia, concomitant α tha-
whereas the MCHC is more often elevated and lassaemia trait does not alter the Hb [99, 224, 234,
the percentage of hyperdense cells is higher. On 239] (but a lower reticulocyte count and lower LDH
density gradient analysis, red cells of compound indicates that there is less haemolysis [234]).
heterozygotes (SC) are denser than those in sickle The WBC, neutrophil count and monocyte count
cell anaemia (SS) and only slightly less dense are elevated in sickle cell/haemoglobin C disease
than those in haemoglobin C disease (CC) [237]. but less so than in sickle cell anaemia [177].
The RDW is increased but generally less than in When sickle cell/haemoglobin C disease is treated
sickle cell anaemia [174, 238]. The HDW is with hydroxycarbamide there is an increase in the
increased [238]. The reticulocyte count is less MCV and a fall in the MCHC and the proportion of
markedly elevated than in sickle cell anaemia, hyperdense cells. The reticulocyte count falls.
with a mean level around 3–6%. The accuracy of
measurement of red cell indices in sickle cell/ Blood film. The peripheral blood features of sickle
haemoglobin C disease is dependent on the auto- cell/haemoglobin C disease are compared with
mated instrument used; cells in this disease are those of sickle cell anaemia and haemoglobin C
less deformable than normal, leading to a false disease in Table 4.9 [186]. In contrast to sickle cell
elevation of the MCV and reduction of the MCHC anaemia, the blood film less often shows classical
on impedance counters and on some earlier light‐ sickle cells. Boat‐shaped cells are more common
scattering instruments [238]. than classical sickle cells, but they are also less
Splenic sequestration is associated with not only common than in sickle cell anaemia. Occasional
a fall in the Hb but also a fall in the platelet count cells may contain straight‐edged six‐sided haemo-
[229]. Bone marrow necrosis with fat embolism syn- globin C crystals. Around half of patients with
drome is associated with worsening anaemia and sickle cell/haemoglobin C disease show character-
thrombocytopenia [225]. istic poikilocytes (Fig. 4.31), which are not seen in
Individuals of African descent with sickle cell/ either sickle cell anaemia or haemoglobin C dis-
haemoglobin C disease show a similar prevalence ease [177, 240]. These misshapen cells have com-
of α thalassaemia trait to those without this condi- plex forms. Some have crystals of varying shape
tion. The prevalence has varied between 20 and and size jutting out at various angles. Others are
35% in different series of patients [224]. Coexisting curved, thus resembling sickle cells, but also
α thalassaemia trait leads to a higher RBC and a appear to contain crystals with straight edges or
lower MCV and MCH in comparison with other with blunt‐angled rather than pointed ends.
Table 4.9 The blood film

features of sickle cell anaemia, Genotype SS SC CC or Cβ0 thalassaemia
sickle cell/haemoglobin C
disease and haemoglobin C Number of cases 29 29 10
Sickle cells 24 6 0
disease or Cβ0 thalassaemia.
Boat‐shaped cells 24 16 1
(From reference 186.)
Haemoglobin C crystals 0 4 5
SC poikilocytes 0 16 0
Irregularly contracted cells 5 25 9
Howell–Jolly bodies 29 5 0
Pappenheimer bodies 25 7 1
Target cells 27 29 9
Spherocytes 10 3 1
Polychromasia 24 9 3
NRBC 26 15 7
NRBC, nucleated red blood cells.

(a)

patients with sickle cell/
haemoglobin C compound
heterozygosity showing the
range of abnormalities that may
be seen: (a) target cells and SC
poikilocytes; (b) SC poikilocytes
and a boat‐shaped cell;
(Continued on p. 226.) (b)
Haemoglobin C will copolymerise with haemoglo- Features of hyposplenism such as Howell–Jolly

bin S (as occurs in the rare sickle cells and in the bodies and Pappenheimer bodies are less com-
commoner boat‐shaped cells in the compound het- mon than in sickle cell anaemia. Polychromasia
erozygous state) [241]. Haemoglobin S will co‐ and NRBC are likewise less common whereas tar-
crystallise with haemoglobin C (as occurs in the get cells (Fig. 4.31c) show a similarly high fre-
less common cells containing haemoglobin C crys- quency and irregularly contracted cells (Fig. 4.31d)
tals) [242]. Deoxygenation favours S‐like polymer- are much more common. Hemi‐ghosts may be
isation whereas oxygenation favours C‐like present (Fig. 4.31d). Assessment of the blood
crystallisation [241, 243]. It seems likely that the count and film usually permits the distinction
formation of SC poikilocytes is consequent on both between sickle cell/haemoglobin C disease and
processes occurring simultaneously in the one cell. sickle cell anaemia. However those cases that lack
On scanning electron microscopy, forms seen sickle cells, boat‐shaped cells and SC poikilocytes
include folded cells (compared to a taco), tricon- (Fig. 4.31d) can be difficult to distinguish from hae-
cave cells and stomatocytes [221]. moglobin C disease.
226 Chapter 4
(c)

(c) hypochromia and target
cells; (d) irregularly contracted
cells, a hemi‐ghost, a target cell
(d) and a stomatocyte. MGG ×100.
Bone marrow necrosis with a fat embolism syn- and 1.5 and 1.4%, respectively, with the Benin and
drome is associated with a leucoerythroblastic Bantu haplotypes [234]. In 98 subjects in the UK, the
blood film with small numbers of schistocytes [225]. mean haemoglobin F was 1.46% (SD 1.81) with
adult levels being reached by 9 years of age [185].
Other investigations. Haemoglobin S and C are pre- The percentage of F cells is increased; in one study
sent in similar proportions (Figs 4.32–4.34). The the mean level was 27% (range 5–73%), in compari-
haemoglobin F percentage ranges from normal to son with normal levels of 0.5–7% [213]. Little
slightly elevated with mean values of 1.1 to 3.3% information is available on the haemoglobin A2

having been reported in different studies. The F percentage in sickle cell/haemoglobin C disease

percentage is significantly higher in females than in since on cellulose acetate electrophoresis haemoglo-
males [234]. As in sickle cell anaemia, the haemo- bin A2 comigrates with haemoglobin C and on
globin F percentage is affected by the β gene haplo- HPLC some post‐translationally modified haemo-
type, averaging 3.2% with the Senegal haplotype globin S falls into the A2 window.
on agarose gel at pH 6.2 showing a
patient with sickle cell/haemoglobin
C disease (second lane from left);
lanes 1 and 10 show a control sample
with, from below up, haemoglobins
F, A, S and C.
Fig. 4.33 HPLC chromatogram (Bio‐Rad Variant II) in sickle cell/haemoglobin C disease showing small early peaks
representing acetylated haemoglobin F, haemoglobin F0 (shaded), post‐translationally modified haemoglobin S in the A0
window, haemoglobin A2 (shaded), haemoglobin S and haemoglobin C (with a shoulder representing post‐translationally
modified haemoglobin C).
228 Chapter 4
Fig. 4.34 Capillary electrophoresis in sickle cell/haemoglobin C disease showing haemoglobins F, S, A2 and C. Note
that the A2 and C peaks overlap.
Diagnosis
The sickle solubility test is positive and immuno-
assays demonstrate the presence of haemoglobins S Diagnosis rests on demonstrating the presence of hae-
and C with no haemoglobin A. moglobin S and haemoglobin C with haemoglobin A
Bilirubin is normal or mildly elevated. LDH is being absent. The identity of the two variant hae-
elevated in comparison with control subjects but is moglobins must be confirmed by at least two inde-
less elevated than in sickle cell anaemia. Red cell life pendent techniques. When electrophoresis is the
span ranges from moderately shortened to slightly primary technique, it is important not to confuse
less than normal. The oxygen dissociation curve compound heterozygous states for S and C‐
shows reduced oxygen affinity (i.e. a right‐shifted Harlem, O‐Arab or E (see pp. 234, 233 and 238)
curve and a higher P50). The reduction in oxygen with sickle cell/haemoglobin C disease since all
affinity is less than is seen in sickle cell anaemia have two bands in the same positions on cellulose
[202]. acetate at alkaline pH. In compound heterozygo-
As for sickle cell anaemia, vaso‐occlusive crises sity for haemoglobins S and E, the band in the C
can be complicated by haemophagocytic lympho- position constitutes a lower percentage than
histiocytosis, and the bone marrow then shows hae- the S band. Homozygous S with coexisting G‐
mophagocytosis [205]. Philadelphia will also have bands in the positions
of S and C but the band in the C position, which amount of haemoglobin A is present. The reduced
represents the hybrid αG‐PhiladelphiaβS haemoglobin, concentration of haemoglobin S within the red
constitutes an appreciably lower percentage than cell, together with the greater or lesser increase in
the band representing S plus G‐Philadelphia. percentages of haemoglobins A2 and F, lessens the
likelihood of sickling and lessens the haemolysis
(in comparison with sickle cell anaemia) but this is
Interactions with other
counterbalanced by the higher Hb and increased
haemoglobinopathies and other
blood viscosity.
haematological diseases
There is conflicting evidence as to the effect of coex-
Clinical features
isting α thalassaemia trait. In most series of patients
α thalassaemia trait has been associated with a Patients with sickle cell/β0 thalassaemia have less
lower risk of osteonecrosis, retinopathy, gallbladder evidence of haemolysis than patients with sickle cell
disease and painful crises [224, 233]. The risk of anaemia but despite this the frequency of painful
splenic sequestration is greatly reduced. crises is, if anything, greater [159]. The explanation
Individuals with sickle cell/haemoglobin C may lie in the higher haemoglobin concentration.
disease who are also heterozygous for the α chain Patients with sickle cell/β+ thalassaemia may have
variant haemoglobin G‐Philadelphia have dis- both less haemolysis and a reduced incidence of
ease of variable severity. One reported case was painful crises in comparison with sickle cell anae-
more severe than is usual in sickle cell/haemo- mia. The amelioration of the disease is proportional
globin C disease [244], while another had a mild to the percentage of haemoglobin A present. They
clinical course with abundant crystals in circulat- may, however, have a higher incidence of prolifera-
ing cells and numerous folded cells [245]. The lat- tive retinopathy as a result of the higher Hb [192].
ter is considered the more typical clinical picture, Splenomegaly persists longer than in sickle cell
attributable to the presence of the G‐Philadelphia anaemia, particularly in those with sickle cell/β+
α chain both increasing the likelihood of crystal- thalassaemia. Patients with sickle cell/β+ thalassae-
lisation of haemoglobin C and decreasing the mia and persisting splenomegaly remain suscepti-
likelihood of polymerisation of haemoglobin S ble to splenic infarction during aeroplane flights,
[221]. Haemoglobin electrophoresis is complex. whereas those with sickle cell/β0 thalassaemia
At alkaline pH there are bands with the mobility resemble patients with sickle cell anaemia since they
of S (about 35%), C (about 47%) and a slow G/C are likely to have had recurrent splenic infarction
hybrid (about 15%) [244]. The ‘S’ band represents and consequent atrophy and therefore do not remain
S and G‐Philadelphia. The ‘C’ band represents C susceptible [228]. Sometimes massive splenomegaly
and S/G hybrid. At acid pH there are two bands leads to hypersplenism. Overexpansion of the bone
with the mobility of S and C. A similar complex- marrow cavity in the skull can cause frontal bossing.
ity is seen on HPLC. Sickle cell/β thalassaemia is generally more severe
A severe phenotype has been observed with in Mediterranean populations than in those of
coincidental hereditary spherocytosis [192]. African descent because of the greater prevalence of
β0 thalassaemia in the former group. Transfusion‐
transmitted babesiosis can cause severe haemolysis
Sickle cell/β thalassaemia
in S/β0 thalassaemia, as well as in sickle cell anaemia
Sickle cell/β thalassaemia is a compound hete- [140]. S/β0 thalassaemia has a similar survival to
rozygous state for βS and either β+ thalassaemia or sickle cell anaemia (estimated at 58 years in one US
β0 thalassaemia [246, 247]. A rare cause of an S/β0 study [154]), while S/β+ thalassaemia has a longer
phenotype is coinheritance of βS with deletion of survival, similar to that of compound heterozygo-
the locus control region β in trans [76]. In sickle sity for S and C (estimated at 66 years) [154].
cell/β0 thalassaemia there is no haemoglobin A, However estimates depend on the precise statistical
whereas in sickle cell/β+ thalassaemia a variable techniques used [122].
230 Chapter 4
Laboratory features sickle cell/β0 thalassaemia. In patients with hypo-

splenism, Pappenheimer bodies are often very
Blood count. Anaemia is milder than in sickle cell
prominent. Polychromasia is present unless there
anaemia, the Hb varying from about 50 g/l to
is associated erythropoietic failure caused by
within the normal range. The distribution of Hb is
infection or megaloblastosis.
bimodal, being higher in those with sickle cell/β+
thalassaemia than in those with sickle cell/β0 tha-
Other investigations. Haemoglobin S comprises
lassaemia; mean values in one study were 107
more than 50% of total haemoglobin, in contrast to
and 81 g/l, respectively [12]. The MCV, MCH and
sickle cell trait when it is less than 50%. In patients
MCHC are reduced, again showing a bimodal
with sickle cell/β0 thalassaemia there is no haemo-
distribution. Mean values for sickle cell/β+ tha-
globin A, whereas in those with sickle cell/β+ tha-
lassaemia and sickle cell/β0 thalassaemia, respec-
lassaemia (Figs 4.37–4.39) the amount of
tively, were 72 and 69.8 fl for MCV, 22.6 and
haemoglobin A varies from almost undetectable
20.1 pg for MCH and 315 and 288 g/l for MCHC
to, rarely, as high as 45% (Table 4.10) [246, 247,
[247]. For both groups mean values for MCV,
249–252]. Haemoglobin F is usually 5–15% and the
MCH and MCHC are lower than those in sickle
percentage of F cells is considerably increased.
cell anaemia, but overlap occurs. The RDW is
When the βS gene is associated with the Arab‐
markedly increased in sickle cell/β0 thalassaemia
Indian haplotype, the mean haemoglobin F is
and moderately increased in sickle cell/β+ thalas-
about 18%, with some values approaching 40%
saemia [174]. It should be noted that in patients
[253]. Compound heterozygosity for βS and the
with sickle cell/β thalassaemia who develop
African type of (Aγδβ)0 thalassaemia has haemoglo-
megaloblastic anaemia the MCV and MCH,
bin F of 22–34% [254]. As for sickle cell anaemia,
although elevated in comparison with baseline
the haemoglobin F concentration is influenced by
values, may be within the normal range.
the β gene haplotype associated with the βS muta-
The reticulocyte count is elevated in sickle cell/β0
tion, being higher with the Senegal and Arab‐
thalassaemia to around 8–9% on average and in
Indian haplotypes. Because of the overlap in
sickle cell/β+ thalassaemia to around 3% on average
values, the haemoglobin F percentage is not very
[235]. During complicating bacterial or parvovirus
useful in separating S/β0 thalassaemia from sickle
infection or megaloblastic anaemia the usual eleva-
cell anaemia; in one series of Jamaican patients the
tion of the reticulocyte count is lacking.
F percentage tended to be higher in the compound
Coexisting α thalassaemia increases the Hb, MCV
heterozygotes but the difference was not signifi-
and MCH and reduces the reticulocyte count [12].
cant [247]. Haemoglobin A2 tends to be somewhat
The RBC and haemoglobin A2 percentage are low-
elevated, usually 3.5–5.5%, with the level being
ered [248]. This alteration in the red cell indices and
higher when the β thalassaemia gene is a β0 rather
haemoglobin A2 percentage not infrequently leads
than a β+ [255]. Higher levels of haemoglobins F
to misdiagnosis of S/β0 thalassaemia as sickle cell
and A2 (and a milder clinical course) have been
anaemia, particularly but not only in those with
observed when the β thalassaemia mutation is a
deletion of two α genes [248].
large (290 bp) deletion [256]. The higher level of
Blood film. The blood film abnormalities are more haemoglobin A2 in sickle cell/β0 thalassaemia can
severe in sickle cell/β0 thalassaemia (Fig. 4.35) be useful in helping to make a distinction between
than in sickle cell/β+ thalassaemia (Fig. 4.36). the compound heterozygous state and sickle cell
Classical sickle cells are quite uncommon, particu- anaemia, with microcytosis consequent on coexist-
larly in sickle cell/β+ thalassaemia. There are some ing α thalassaemia trait, in which haemoglobin A2
boat‐shaped cells. There is hypochromia and is usually in the range of 2–4%. Although there is
microcytosis and circulating NRBC show defective some overlap in haemoglobin A2 percentages this
haemoglobinisation. Target cells are prominent is the most useful variable for making the distinc-
and basophilic stippling may be apparent. Features tion; the Hb, reticulocyte count and haemoglobin F
of hyposplenism may be present, particularly in percentage show more overlap (see Table 4.8).
(a)
Fig. 4.35 Blood films of two

patients with sickle cell/β0
thalassaemia showing: (a) target
cells, a nucleated red blood cell
and a number of partly sickled
cells; (b) hypochromia,
microcytosis, target cells and a
number of partly sickled cells.
MGG ×100. (b)
Red cell life span is reduced, particularly in sickle ins S and F detected so that confusion with sickle cell
cell/β0 thalassaemia, but not to the same extent as in anaemia and sickle cell/β0 thalassaemia is possible.
sickle cell anaemia. The α:β chain synthesis ratio in Only a provisional diagnosis can be made in this cir-
peripheral blood reticulocytes is increased in sickle cumstance. Family studies and follow‐up are needed
cell/β thalassaemia, whereas it is normal in sickle for a definitive diagnosis.
cell anaemia. The bone marrow aspirate (Fig. 4.40) shows
In the neonatal period the diagnosis of sickle cell/β+ erythroid hyperplasia, sickle cells and a variable
thalassaemia can be difficult [201]. Confusion with degree of iron overload.
sickle cell trait can occur if almost all the haemoglobin
present is haemoglobin F and the proportions of hae-
Diagnosis
moglobins S and A are so low that it is not clear which
is present in the greater amount. Neonates with sickle Diagnosis of compound heterozygosity for haemo-
cell/β+ thalassaemia may also have only haemoglob- globin S and β+ thalassaemia is straightforward,
232 Chapter 4

patients with sickle cell/β+
(a) thalassaemia showing:
(a) hypochromia, poikilocytosis
and a probable sickle cell;
(b) hypochromia, microcytosis
and one partly sickled cell. The
first patient had 70%
haemoglobin S, 24%
haemoglobin A and 6%
haemoglobin A2; red cell indices
were RBC 5.07 × 1012/l, Hb
106 g/l, MCV 64 fl, MCH 21 pg
and MCHC 329 g/l. The second
patient had 59% haemoglobin S,
25% haemoglobin A plus A2 and
13% haemoglobin F; the red cell
indices were RBC 4.71 × 1012/l,
Hb 106 g/l, Hct 0.32, MCV 68 fl,
MCH 22.5 pg and MCHC
(b) 330 g/l. MGG ×100.
merely requiring the demonstration of both hae- to be made from compound heterozygosity for
moglobin A and haemoglobin S by two independ- haemoglobin S and deletional hereditary persis-
ent techniques and the demonstration that tence of fetal haemoglobin, particularly with
haemoglobin S is present as a larger proportion coexisting α thalassaemia trait; in this instance
than haemoglobin A. Diagnosis of compound clinical features, Hb, MCV and haemoglobin F
heterozygosity for haemoglobin S and β0 thalas- percentage are useful (see Table 4.8). When a
saemia is more difficult since a distinction has precise diagnosis is important (e.g. for genetic
to be made from sickle cell anaemia with micro- counselling) and the diagnosis is not clear from
cytosis (e.g. due to coexisting α thalassaemia) (see family studies and from a consideration of
earlier). The coexistence of α thalassaemia can
the proportions of various haemoglobins, deox-
lead to misdiagnosis of S/β0 thalassaemia as yribonucleic acid (DNA) analysis should be
sickle cell anaemia [248]. A d istinction also needs c arried out.
cells (Fig. 4.42). On cellulose acetate electrophore-

sis at alkaline pH, haemoglobins S and D‐Punjab
show the same electrophoretic mobility, but HPLC
(Fig. 4.43) and electrophoresis at acid pH sepa-
rate these two haemoglobins from each other.
Haemoglobin D forms a somewhat higher propor-
tion of total haemoglobin than does haemoglobin
S [261]. In a few cases haemoglobin F has been sig-
nificantly elevated (e.g. 13–20% [260]), but usually
haemoglobin F is present in only small amounts.
Haemoglobin A2 may be slightly elevated [261].
This compound heterozygous state is not amelio-
Fig. 4.37 Haemoglobin electrophoresis on cellulose rated if the βS gene is associated with the Arab‐
acetate at alkaline pH in a patient with sickle cell/β+ Indian haplotype and a high haemoglobin F [263].
thalassaemia compound heterozygosity; ASC indicates a It should be noted that coinheritance of haemo-
control sample containing haemoglobins A, S and C.
globin S and haemoglobin D variants other than
haemoglobin D‐Punjab does not have the same
adverse effects as sickle cell/haemoglobin D‐Punjab
Other causes of sickle cell disease compound heterozygosity. For example, two
Nigerians with haemoglobin S/haemoglobin
Sickle cell/haemoglobin D‐Punjab/Los
D‐Ibadan were asymptomatic [259]. Similarly,
Angeles disease
haemoglobin D‐Iran does not interact adversely

Compound heterozygosity for sickle cell haemo- with haemoglobin S.
globin and haemoglobin D‐Punjab (D‐Los Angeles) Median survival is similar to that for sickle cell
leads to sickle cell disease which is, on average, anaemia.
slightly milder than sickle cell anaemia [12, 217,
257–261]. Some patients are asymptomatic [262].
Sickle cell/haemoglobin O‐Arab disease
This compound heterozygous state has been
observed in Afro‐Americans, Afro‐Caribbeans, Compound heterozygosity for sickle cell haemoglo-
Central and South Americans (Mexicans and bin and haemoglobin O‐Arab (α2β2121Glu→Lys) leads to
Venezuelans), Arabs and Turks and, in addition, in sickle cell disease that is generally severe [12, 217,
a number of individuals of mixed ancestry 264–266]. Sickle cell/haemoglobin O‐Arab has been
(Northern European/American Indian, English/ observed in Arabs, Africans (Sudanese and Kenyans),
African, English/Afro‐Caribbean), including sev- Afro‐Caribbeans, Afro‐Americans and Americans
eral individuals who appeared to have only who appeared to be of Caucasian ancestry. The Hb in
Mediterranean or Northern European ancestry. adults varies between 61 and 99 g/l. The reticulocyte
The clinical features are of a mild or moderate count is usually between 8 and 10% (1–15% reported).
haemolytic anaemia with sickling crises. Persisting Reported MCVs have been quite variable, from nor-
splenomegaly is commoner than in sickle cell mal to moderately macrocytic levels (82–110 fl in
anaemia. The Hb is usually between 50 and 100 g/l adults). The blood film (Fig. 4.44) is similar to that in
and the reticulocyte count between 5 and 20% sickle cell anaemia. Oxygen affinity is reduced,
(occasionally higher). The MCV is very variable comparable to what is seen in sickle cell anaemia.

but macrocytosis is quite common with some indi- On electrophoresis on cellulose acetate at alkaline
viduals having an MCV of 110–120 fl. The blood pH, haemoglobin O‐Arab has similar mobility to hae-
film (Fig. 4.41) shows anisocytosis, poikilocytosis, moglobin C (Fig. 4.45) but at acid pH mobility
target cells, sickle cells, boat‐shaped cells, nucle- depends on the electrophoresis medium. On aga-
ated red cells and sometimes macrocytes. The bone rose gel it is slightly slower than S (Fig. 4.46). On
marrow shows erythroid hyperplasia and sickle HPLC there are two abnormal peaks, one in the
234 Chapter 4
Fig. 4.38 HPLC chromatogram (Bio‐Rad Variant II) of haemoglobin S/β+ thalassaemia showing a triple peak of
ost‐translationally modified haemoglobin F, haemoglobin F0, low peaks of modified haemoglobin A, haemoglobin
p
A0, haemoglobin A2 (shaded) and haemoglobin S. Note the shoulder on the left of the haemoglobin A0 peak, which
represents glycated haemoglobin S.
position of S and the other between S and C Variant II instrument [267]. Haemoglobin S forms a
(Fig. 4.47). Haemoglobin O‐Arab and haemoglobin somewhat higher proportion of total haemoglobin
C‐Harlem can be easily confused with each other than does haemoglobin O‐Arab [261].
when present in the compound heterozygous state
with haemoglobin S. The difference in mobility on
Sickle cell/haemoglobin C‐Harlem
citrate agar at acid pH is most useful in making the
compound heterozygosity
distinction (Table 4.11). Retention times on HPLC
are quite similar, 4.9–4.93 minutes for O‐Arab and This condition is slightly milder than sickle cell
4.89 minutes reported for C‐Harlem on a Bio‐Rad anaemia. The blood film shows similar features.
Fig. 4.39 Capillary electrophoresis (Sebia Capillarys 3) in sickle cell/β+ thalassaemia showing haemoglobins A (10%),
F (32%), S (53%) and A2 (4.8%).
Table 4.10 Percentage of

haemoglobin A in compound Mutation and ethnic group Percentage of haemoglobin A
heterozygosity for S and β+
thalassaemia [246, 247, 249–252]. C→G at IVS2, position 745 (Greek/Turkish) 3–5
G→C at IVS1, position 5 (Indian) 3–5
G→A at IVS1, position 110 (Greek/Turkish) 8–14
C→T at –88 (black) 18–25
A→G at –29 (black) 18–25
G→T at IVS1, position 5 (Greek/Turkish) 18–25
Haemoglobin electrophoresis at alkaline pH resembles

Sickle cell/haemoglobin Lepore
that of sickle cell/haemoglobin C disease whereas at
acid pH on citrate agar (but not agarose gel) there is Compound heterozygosity for sickle cell haemo-
a single band with the mobility of haemoglobin S globin and haemoglobin Lepore‐Boston [268, 269]
(see Table 4.11). has been reported in Mediterranean (Greek and
236 Chapter 4
Fig. 4.40 Bone marrow aspirate

in sickle cell/β0 thalassaemia
compound heterozygosity showing
erythroid hyperplasia and one sickle
cell. MGG ×100.
Italian), Afro‐Caribbean, Afro‐American and Sickle cell/δβ0 thalassaemia

Indian populations. Sickle cell/haemoglobin
Sickle cell/δβ0 thalassaemia has been observed
Lepore leads to sickle cell disease of variable
in Mediterranean populations (Greek, Sicilian,
severity but resembling sickle cell/β thalassaemia
other Italian), in Arabs and in Afro‐Americans
more closely than sickle cell anaemia. Of 10 cases
[12, 217]. This compound heterozygous state is
reported up to 1997 three were severe and seven
generally much milder than sickle cell anaemia
were mild [269].
because the high percentage of haemoglobin F
Haematological variables reported in adults
protects against sickling. There is mild anaemia
[12, 217, 269] have been Hb 80–133 g/l, MCV
and splenomegaly.
66.5–83 fl, MCH 24.3–27.6 pg and reticulocyte
The blood count shows an Hb of around 100–
count 3–13% (33% in one case). The blood film
120 g/l and an MCV that is slightly reduced (76–
shows anisocytosis, hypochromia, microcytosis
83 fl). The reticulocyte count is mildly elevated,
and some sickle cells.
usually 2–4%. The blood film shows anisocytosis,
Since haemoglobin Lepore has the same
poikilocytosis and hypochromia. Haemoglobin S is
mobility as haemoglobin S on electrophoresis at
the major haemoglobin component, with haemo-
alkaline pH the only bands apparently present
globin F being 15–37% of total haemoglobin. The
are haemoglobins F, S and A 2 and diagnostic
proportion of haemoglobin A2 is normal or low
confusion with sickle cell anaemia and sickle
(1.5–3.1%). Sickle cell/δβ0 thalassaemia differs from
cell/β 0 thalassaemia can therefore occur.
microcytic cases of sickle cell anaemia, having a
However other techniques such as HPLC
higher Hb, lower reticulocyte count and lower hae-
(Fig. 4.48) show that haemoglobin Lepore is
moglobin A2 percentage (see Table 4.8). However
usually around 10–12% of total haemoglobin
definitive diagnosis requires DNA analysis.
(20% in one case) while h
aemoglobin S is 63–90%
and haemoglobin F 5–25%. Electrophoresis at
acid pH shows two bands, one with the mobility
Sickle cell/hereditary persistence of fetal
of haemoglobin A, which represents the haemo-
haemoglobin
globin Lepore. The proportion of haemoglobin
A 2 is variable, having been reported to be Compound heterozygosity for haemoglobin S and
reduced, normal or slightly elevated in different deletional or pancellular hereditary persistence of
cases (0.9–4%) [12, 269]. fetal haemoglobin (HPFH) is either asymptomatic
Fig. 4.41 Blood film in sickle cell/

haemoglobin D‐Punjab compound
heterozygosity. MGG ×100.

sickle cell/haemoglobin D‐Punjab
compound heterozygosity showing
prominent sickle cell formation.
MGG ×100.
or produces quite mild sickle cell disease [217]. slightly reduced and haemoglobin S comprising
Haemoglobin S/HPFH has been reported in Africans, around 60–80% of total haemoglobin. Virtually all
Afro‐Caribbeans and Afro‐Americans. There can be cells are F cells [213]. Haemoglobin A is absent. In
mild haemolytic anaemia and splenomegaly or infancy haemoglobin F is 50–90% with a decline
minor clinical features consequent on sickling. The and stabilisation at adult levels by 3–5 years [270].
Hb and reticulocyte count are usually normal but The proportions of various haemoglobins in S/
microcytosis is common and occasionally there is HPFH are similar to those of S/β0 thalassaemia.
mild anaemia and reticulocytosis of 2–4%. The blood Distinction between the two is aided by the usual
film (Fig. 4.49) may show anisocytosis, microcytosis lack of symptoms and by the fact that the Hb and
and target cells. Haemoglobin electrophoresis or reticulocyte count are often normal in S/HPFH
HPLC shows haemoglobin F of 15–35% (usually (see Table 4.8). Definitive diagnosis requires family
20–30%), haemoglobin A2 which is low normal or studies or DNA analysis.
238 Chapter 4
Fig. 4.43 HPLC chromatogram (Bio‐Rad Variant II) in sickle cell/haemoglobin D‐Punjab compound heterozygosity.
The peaks, from left to right, are post‐translationally modified haemoglobin F (three peaks), haemoglobin F0 (shaded),
post‐translationally modified haemoglobins S and D‐Punjab (three peaks), haemoglobin A2 (shaded), haemoglobin D
and haemoglobin S.
Sickle cell/haemoglobin E compound

fro‐Caribbeans, Indians, Pakistanis, Haitians, in
A
heterozygosity
the Middle East (including Saudi Arabia and
Compound heterozygosity for haemoglobin S Oman), in South‐East Asia and in individuals of
and haemoglobin E produces a condition that is mixed ancestry. There are sometimes complica-
usually either asymptomatic or clinically mild tions related to sickling, particularly in adults; these
[12, 217, 271–277]. Sickle cell/haemoglobin E have included painful crises, acute chest syndrome,
has been observed in Turks, Afro‐Americans, splenomegaly, splenic sequestration (which can

haemoglobin O‐Arab compound
heterozygosity showing
hypochromia, target cells and
partially sickled cells. (Note: O‐Arab
in this patient was misidentified as
C‐Harlem in the first edition of this
book; the correct identity was
subsequently confirmed by family
studies, citrate agar electrophoresis
and mass spectrometry.) MGG ×100.
1 2 3 4 5 6 7
PATIENT
Fig. 4.46 Haemoglobin electrophoresis on agarose gel at

acid pH in sickle cell/haemoglobin O‐Arab compound
heterozygosity (lanes 3 and 5) showing a faint F band
and broadening of the S band in the direction of C by the
presence of O‐Arab. From left to right, lanes are: (1) F, A
and C; (2) F and A; (3) S plus O‐Arab; (4) F and A; (5) S
Fig. 4.45 Haemoglobin electrophoresis on cellulose plus O‐Arab; (6) F; (7) A and S. The mobility of O‐Arab
acetate at alkaline pH in sickle cell/haemoglobin O‐Arab on this medium is more readily apparent in the absence
compound heterozygosity; by this technique the pattern of haemoglobin S (see Fig. 5.39).
cannot be distinguished from that of sickle cell/
haemoglobin C compound heterozygosity.
blood film (Fig. 4.50) can show target cells which
are sometimes numerous. Sickle cells have been
occur in adults), splenic infarction (recurrent dur- observed [271] but this is not usual. Haemoglobin
ing flights in one patient), hepatomegaly, osteone- S is a larger proportion of total haemoglobin than
crosis and retinal disease [272, 275, 278–280]. Death haemoglobin E (e.g. 55–63%) with around 33% of
from embolisation of infarcted bone marrow, asso- haemoglobin E plus A2 and 0–7.5% haemoglobin
ciated with parvovirus B19 infection, has been F [261, 272–275, 277 ] (Fig. 4.51). HPLC (Bio‐Rad
reported in a child [280]. There can be mild haemol- Variant II) shows two major peaks, one represent-
ysis (often compensated), jaundice and gallstones. ing haemoglobin E plus A2 and the other haemo-
The Hb is often normal but may be reduced (80– globin S (Fig. 4.52); on other HPLC instruments,
146 g/l) with sometimes a slight increase in the haemoglobins E and A2 may form separate peaks.
reticulocyte count (1.5–5.3%). The MCV can be nor- On capillary electrophoresis, there are two major
mal but is usually reduced (68–97 fl in adults). The peaks, haemoglobins E and S, plus a haemoglobin
240 Chapter 4

haemoglobin S/O‐Arab compound
heterozygosity; the haemoglobin
O‐Arab in this and another patient
had retention times on a Bio‐Rad
Variant II of 4.89 and 4.90 minutes,
respectively; peaks, from left to
right, are altered F, F0 (shaded),
altered S (double peak), A2 (shaded),
S and O‐Arab. Note that the S peak
is flanked by two small peaks that
represent altered O‐Arab.
Table 4.11 Making a distinction between haemoglobin O‐Arab and haemoglobin C‐Harlem.
Haemoglobin O‐Arab Haemoglobin C‐Harlem
Frequency Uncommon Rare
Clinical severity of compound As severe as sickle cell anaemia Somewhat milder than
heterozygous state with haemoglobin S sickle cell anaemia
Sickle solubility test Negative Positive
Mobility on cellulose acetate Mobility of C Mobility of C

electrophoresis at alkaline pH
Mobility on agarose gel electrophoresis Slightly slower than S (i.e. slightly With S
at acid pH towards C)
Mobility on citrate agar electrophoresis Somewhat faster than S (i.e. slightly With S
at acid pH on the A side of S)
High performance liquid Between S and C Between S and C

chromatography
Isoelectric focusing With E With E
A2 peak (Fig. 4.53). With both techniques it is and haemoglobin F from 12 to 30% [281]. The
apparent that haemoglobin E (with or without propositus was asymptomatic.
haemoglobin A2) is a much lower percentage than
haemoglobin S.
Other rare compound heterozygous
and related states
Sickle cell/δ0β+ thalassaemia
Compound heterozygosity for haemoglobin C
Sickle cell/δ0β+ thalassaemia, as the result of for- and haemoglobin C‐Harlem appears to produce a
mation of a δβ fusion gene, was reported in four much milder disease than sickle cell/haemoglo-
brothers in a Senegalese family. Hb varied from bin C disease. One reported patient who pre-
109 to 134 g/l, MCV from 76 to 85 fl, haemoglobin sented with haematuria had anaemia and
S from 58 to 70%, haemoglobin A from 12 to 16% splenomegaly but no symptoms suggestive of
Fig. 4.48 HPLC chromatogram (Bio‐Rad Variant II) in a patient with compound heterozygosity for haemoglobin S and
haemoglobin Lepore showing triple peak of post‐translationally modified haemoglobin F, haemoglobin F0 (shaded),
altered haemoglobin Lepore and S in the A0 window, haemoglobin A2 plus Lepore (shaded) and haemoglobin S.
sickling [10]; the blood film showed many target Compound heterozygosity for haemoglobin S
cells and occasional sickle cells. Compound hete- and the electrophoretically silent variant, haemo-
rozygosity for haemoglobin S and haemoglobin globin Quebec‐Chori, causes sickle cell disease
S‐Antilles produces a very severe form of sickle [51]. Compound heterozygosity for haemoglobin
cell disease [41]. Compound heterozygosity for S and the electrophoretically silent unstable vari-
haemoglobin S and haemoglobin S‐Oman has ant haemoglobin Volga caused splenic sequestra-
been described, presenting at the age of 1 year tion and mild disease thereafter in one patient
[282]; it is likely that the phenotype would [283]. Compound heterozygosity for haemoglo-
be severe since this double substitution haemo- bin S and mildly unstable haemoglobins such
globin can cause disease in heterozygotes. as haemoglobin Hope and haemoglobin Siriraj
242 Chapter 4

hereditary persistence of fetal
haemoglobin compound
heterozygosity showing mild
poikilocytosis and target cell
formation. MGG ×100.

haemoglobin E compound
heterozygosity showing microcytosis
and poikilocytosis. (With thanks to
Dr Rajeev Gupta.) MGG ×100.
(Figs 4.54 and 4.55) can cause mild haemolysis Kenya, has an interestingly mild phenotype,
[217, 284]. A compound heterozygote for haemo- given that S is 60–70% of total haemoglobin [286];
globin S and the mildly unstable variant, haemo- compound heterozygotes appear to be generally
globin Tyne, had the clinical features of sickle cell asymptomatic with a mild microcytic anaemia
disease [285]. A compound heterozygote for S [115, 287]. Haemoglobin Kenya is about 18% and
and haemoglobin Hofu (a fast moving haemoglo- haemoglobin F around 8–10% with a pancellular
bin) had a significant anaemia (Hb 96 g/l), 73% distribution [115, 287]; haemoglobins F and A2
haemoglobin S and apparently clinical features of will inhibit sickling and one might be tempted to
sickling [261]. Compound heterozygosity for hae- postulate that this could also be true of haemo-
moglobin S and the Aγβ fusion variant, haemoglobin globin Kenya. Splenic infarction and chest crisis
Two individuals have been reported with inherit-

ance of βS from one parent and β from the other but
with post‐zygotic mitotic recombination leading to
mosaic isodisomy and a mixture of AS and SS cells;
this led to late presentation with a sickle cell disease
phenotype [76].
There are many β chain variants that do not interact
with haemoglobin S so that compound heterozygotes
have clinical and haematological features that
resemble those of sickle cell trait. These include hae-
moglobins Camden, Caribbean, D‐Ouled Rabah,
D‐Ibadan, Detroit, E‐Saskatoon, G‐Galveston, G‐San
Jose, G‐Szuhu, J‐Amiens, J‐Baltimore, J‐Bangkok,
K‐Ibadan, K‐Matupo, K‐Woolwich, Mobile,
N‐Baltimore, Ocho‐Rios, Osu‐Christiansborg,
Pyrgos and Richmond [131, 173].
Fig. 4.51 Haemoglobin electrophoresis at acid and
alkaline pH in sickle cell/haemoglobin E compound
heterozygosity; the first three lanes from the left are Sickle cell disease in heterozygotes
cellulose acetate electrophoresis at alkaline pH; the
two lanes on the right are agarose gel electrophoresis Three variant haemoglobins in which the β6Glu→Val
at acid pH with the ASC control showing, from below substitution is one of two substitutions are
up, haemoglobins A, S and C. (With thanks to capable of producing sickle cell disease in

Dr Rajeev Gupta, Mr Martin Jarvis and Dr Anne heterozygotes. They are haemoglobin S‐Antilles,
Yardumian.) haemoglobin S‐Oman and haemoglobin Jamaica
Plain (see Table 4.2).
Haemoglobin S‐Oman is usually coinherited
can occur in compound heterozygotes for haemo- with either −α/αα or −α/−α and microcytosis is
globin S and haemoglobin New York, a mildly therefore common [290]. Heterozygotes have
unstable haemoglobin with low oxygen affinity, from 4 to 25% haemoglobin S‐Oman with the per-
which has the same retention time as haemoglo- centage tending to be lower in those with −α/−α
bin S on HPLC [288]. Compound heterozygosity [290]. Disease in heterozygotes varies from mild
for haemoglobin S and haemoglobin Monroe to severe. Strangely, severe disease is more likely
leads to a clinical syndrome resembling haemo- in those with a higher haemoglobin F level; sever-
globin S/β0 thalassaemia, since haemoglobin ity tends to be less in those with −α/−α than in
Monroe is unstable and constitutes only about 2% those with −α/αα but does not show a close rela-
of total haemoglobin [289]. Compound heterozy- tionship to the percentage of the variant haemo-
gosity for haemoglobin S and various other vari- globin [290]. The morphology of sickle cells in
ant haemoglobins can cause haematological patients who are simple or compound heterozy-
abnormalities as a result of the characteristics of gotes for haemoglobin S‐Oman differs from the
the second variant, rather than as a result of any morphology of classical sickle cells. There are
interaction between the two variant haemoglob- cells that are pointed at both ends but fat in the
ins; this appears to be true of haemoglobins middle; they have been compared to a yarn/knit-
I‐Toulouse (unstable), San Diego (high affinity), ting needle or to Napoleon hats [282] (Fig. 4.56).
Shelby (mildly unstable), Lufkin (mildly unsta- Affected heterozygotes also differ clinically from
ble), Hope (unstable and low oxygen affinity) and sickle cell anaemia in that splenomegaly can per-
North Shore (‘thalassaemic’) [117]. sist into adult life.
244 Chapter 4
Fig. 4.52 HPLC chromatogram (Bio‐Rad Variant II) in sickle cell/haemoglobin E compound heterozygosity. The peaks
from left to right are: injection artefact, haemoglobin F (shaded), three peaks representing post‐translationally modified
haemoglobin S and A, a complex peak representing haemoglobin A2 plus E (shaded) and haemoglobin S.
Sickle cell disease resulting from uniparental

Sickle cell disease can also occur in heterozygotes
disomy
if there is coinheritance of another condition leading
to a high concentration of 2,3‐DPG and reduced oxy- A unique patient has been described with sickle cell
gen affinity. For example, a patient who had coinher- disease resulting from post‐zygotic mitotic recom-
ited a severe pyruvate kinase deficiency had a twofold bination leading to uniparental disomy [291]. The
increase in 2,3‐DPG leading to reduced oxygen affin- child had 52% haemoglobin S and a mixture of AS
ity and symptomatic sickling crises [53]. and SS cells.
Fig. 4.53 Capillary electrophoresis in sickle cell/haemoglobin E compound heterozygosity. The peaks from left to right
are haemoglobin F, unidentified, haemoglobin S, haemoglobin E and haemoglobin A2.
Fig. 4.54 Blood film of a 1‐year‐old

child with haemoglobin S/
haemoglobin Siriraj compound
heterozygosity showing anisocytosis
and hypochromia. The red cell
indices were RBC 4.87 × 1012/l, Hb
112 g/l, MCV 67 fl, MCH 23.3 pg
and MCHC 344 g/l. MGG ×100.
Fig. 4.55 HPLC chromatogram (Bio‐Rad Variant II) in a patient with compound heterozygosity for haemoglobin S and
haemoglobin Siriraj showing low peaks of altered haemoglobin F, haemoglobin F0 (shaded), glycated haemoglobin S in
the A0 window, haemoglobin A2 plus other post‐translationally modified haemoglobin S (shaded), haemoglobin S, and
haemoglobin Siriraj in the haemoglobin C window. Note the two small sharp peaks between the peaks for haemoglobins
S and Siriraj, which represent post‐translationally modified haemoglobin Siriraj.

compound heterozygosity for haemoglobin
S and haemoglobin S‐Oman compound
heterozygosity showing the ‘Napoleon hat’
red cells that are characteristic of
haemoglobin S‐Oman. (With thanks to
Dr Samir Al Azzawi, Dr Raya Al Jahdhamy
and colleagues, Muscat, Sultanate of Oman.)
Check your knowledge 4.6 In comparison with sickle cell anaemia,

patients with compound heterozygosity
One to five answers may be correct. Answers for haemoglobin S and haemoglobin C
to most questions can be either found in this usually have
chapter or deduced from information given.
(a) a higher percentage of haemoglobin A
Answers are given on p. 260. (b) more severe anaemia
(c) a higher incidence of proliferative
4.1 The coinheritance of haemoglobin S and the retinopathy
following haemoglobins usually produces a (d) a higher incidence of ischaemic necrosis
clinically significant sickling disorder of the femoral head
(a) haemoglobin C (e) earlier onset of blood film features of
(b) haemoglobin G‐Philadelphia hyposplenism
(c) haemoglobin D‐Punjab
(d) haemoglobin Lepore 4.7 Significant disease would be predicted in
(e) haemoglobin A 25% of offspring if the partner of a pregnant
woman with sickle cell trait had
4.2 Haemoglobin S occurs in a significant (a) α thalassaemia trait
proportion of individuals of the following (b) β thalassaemia trait
ethnic groups (c) δ thalassaemia trait
(a) Australian aboriginals (d) δβ thalassaemia trait
(b) Greeks (e) non‐deletional hereditary persistence of
(c) Southern Italians and Sicilians fetal haemoglobin
(d) Saudi Arabs
(e) Nigerians 4.8 The disease phenotype is usually appreciably
less severe than that of homozygosity for
4.3 Recognised features of sickle cell trait include haemoglobin S in
(a) a defect in urine concentrating ability (a) sickle cell/haemoglobin C disease
(b) an increased incidence of gallstones (b) sickle cell/β+ thalassaemia
(c) an increased reticulocyte count (c) sickle cell/deletional hereditary persis-
(d) leg ulcers tence of fetal haemoglobin
(e) susceptibility to clinically significant (d) sickle cell/δβ0 thalassaemia
sickling in conditions of severe hypoxia (e) sickle cell/haemoglobin E
4.4 The following variant haemoglobins have the 4.9 On haemoglobin electrophoresis at alkaline
same mobility as haemoglobin S on cellulose pH homozygosity for haemoglobin S cannot
acetate electrophoresis at alkaline pH be distinguished from
(a) haemoglobin C (a) sickle cell/haemoglobin C disease
(b) haemoglobin D (b) sickle cell/β0 thalassaemia
(c) haemoglobin E (c) sickle cell/haemoglobin D‐Punjab
(d) haemoglobin F (d) sickle cell/haemoglobin Lepore
(e) haemoglobin G (e) heterozygosity for both haemoglobin S
and haemoglobin G‐Philadelphia
4.5 The likelihood of red cell sickling occurring is
increased by 4.10 A higher mortality rate in sickle cell anaemia
(a) acidosis correlates with
(b) a lower partial pressure of oxygen (a) higher white cell count
(c) an increased percentage of haemoglobin F (b) coexisting α thalassaemia trait
(d) reduced blood flow through tissues (c) lower percentage of haemoglobin F
(e) a lower mean cell haemoglobin concentration (d) male gender
(MCHC) (e) previous cerebrovascular accident
248 Chapter 4
4.11 The blood count in sickle cell/haemoglobin 10 de Pablos JM (1985) Incidence of Hb C trait in an
C disease is characterised by area of southern Spain. Br J Haematol, 60, 584–585.
(a) generally mild anaemia 11 Nagel RL and Fleming AF (1992) Genetic epidemi-
(b) reticulocytosis ology of the βS gene. Bailliere’s Clin Haematol, 5,
(c) increased mean cell volume (MCV) 331–365.
(d) increased mean cell haemoglobin 12 Serjeant GR. Sickle Cell Disease, 2nd edn. Oxford
concentration (MCHC) University Press, Oxford, 1992.
(e) increased red cell distribution width 13 Talafih K, Hunaiti AA, Gharaibeh N, Gharaibeh M
and Jaradat S (1996) The prevalence of hemoglobin
(RDW) and haemoglobin distribution
S and glucose‐6‐phosphate dehydrogenase defi-
width (HDW)
ciency in Jordanian newborn. J Obstet Gynaecol Res,
22, 417–420.
4.12 The haemoglobin F percentage in sickle cell
14 Wurie AT, Wurie JM, Gevao SM and Robbin‐Coker DJ
anaemia is affected by
(1996) The prevalence of sickle cell trait in Sierra Leone:
(a) age
a laboratory profile. West Afr J Med, 15, 201–203.
(b) gender 15 Flint J, Harding RM, Boyce AJ and Clegg JB (1998)
(c) haplotype of the β globin gene cluster The population genetics of the haemoglobinopa-
(d) the F cell locus on the X chromosome thies. Bailliere’s Clin Haematol, 11, 1–51.
(e) hydroxycarbamide (hydroxyurea) 16 Segbena AY, Prehu C, Wajcman H, Bardakdjian‐
therapy Michau J, Messie K, Feteke L et al. (1998)
Hemoglobins in Togolese newborns: Hb S, Hb C,
Hb Bart’s, and α‐globin gene status. Am J Hematol,
References
59, 208–213.
1 Herrick JB (1910) Peculiar elongated and sickle 17 St John MA and Lungu FN (1999) Haemoglobin
shaped red blood corpuscles in a case of severe electrophoresis patterns in Barbados. West Indian
anaemia. Ann Intern Med, 6, 517–521. Med J, 48, 221–222.
2 Mason VR (1922) Sickle cell anemia. JAMA, 79, 18 de Silva S, Fisher CA, Premawardhena A,
1318–1320. Lamabadusuriya SP, Peto TEA, Perera G et al. (2000)
3 Pauling L, Itano HA, Singer SJ and Wells IC (1949) Thalassaemia in Sri Lanka: implications for the
Sickle cell anaemia, a molecular disease. Science, future health burden of Asian populations. Lancet,
110, 543–548. 355, 786–791.
4 Ingram VM (1956) A specific chemical difference 19 Serjeant GR (2001) Historical review: the emerging
between the globins of normal human and sickle understanding of sickle cell disease. Br J Haematol,
cell anaemia haemoglobin. Nature, 178, 792–194. 112, 3–18.
5 Powars DR (1991) βS‐gene‐cluster in sickle cell ane- 20 Kaur M, Das GP and Verma IC (1997) Sickle cell
mia: clinical and hematologic features. Hematol/ trait and disease among tribal communities in
Oncol Clin North Am, 5, 475–493. Orissa, Madhya Pradesh and Kerala. Indian J Med
6 Streetly A, Sisodia R, Dick M, Latinovic R, Hounsell Res, 105, 111–116.
K and Dormandy E (2018) Evaluation of newborn 21 Serjeant GR and Sergeant BE. Sickle Cell Disease, 3rd
sickle cell screening programme in England: 2010– edn, Oxford University Press, Oxford, 2001.
2016. Arch Dis Child, 103, 648–653. 22 Elderdery AY, Mohamed BA, Karsani ME,
7 Gulbis B, Cotton F, Ferster A, Ketelslegers O, Dresse Ahmed MH, Knight G and Cooper AJ (2008)
MF, Rongé‐Collard E et al. (2009) Neonatal haemo- Hemoglobinopathies in the Sudan. Hemoglobin,
globinopathy screening in Belgium. J Clin Pathol, 32, 323–326.
62, 49–52. 23 Miller CJ, Dunn EV, Berg B and Abdouni SF (2003)
8 White JM (1983) The approximate gene frequency A hematological survey of preschool children of the
of sickle haemoglobins in the Arabian peninsula. Br United Arab Emirates. Saudi Med J, 24, 609–613.
J Haematol, 55, 563–564. 24 Al‐Arayyed S, Hamza AA, Sultan B, Shome DK and
9 Livingstone FB. Frequencies of Hemoglobin Variants. Bapat JP (2007) Neonatal screening for genetic
Oxford University Press, Oxford, 1985. blood diseases. Bahrain Med Bull, 29, 88–90.
25 Pereira MM and Vives Corrons J‐L (2009) Neonatal Oncología Pediátricas) (2017) National registry of
haemoglobinopathy screening in Spain. J Clin hemoglobinopathies in Spain (REPHem). Pediatr
Pathol, 62, 22–25. Blood Cancer, 64, e26322.
26 Tshilolo L, Aissi LM, Lukusa D, Kinsiama C, 38 Premawardhena A, Allen A, Piel F, Fisher C, Perera
Wembonyama S, Gulbis B and Vertongen F (2009) L, Rodrigo R et al. (2017) The evolutionary and clini-
Neonatal screening for sickle cell anaemia in the cal implications of the uneven distribution of the
Democratic Republic of the Congo: experience from frequency of the inherited haemoglobin variants
a pioneer project on 31 204 newborns. J Clin Pathol, over short geographical distances. Br J Haematol,
62, 35–38. 176, 475–484.
27 Kafando E, Nacoulma E, Ouattara Y, Ayéroué J, 39 Marchand M, Gill C, Malhotra AK, Bell C, Busto E,
Cotton F, Sawadogo M and Gulbis B (2009) Neonatal McKeown MD et al. (2017) The assessment and sus-
haemoglobinopathy screening in Burkina Faso. tainable management of sickle cell disease in the
J Clin Pathol, 62, 39–41. indigenous Tharu population of Nepal. Hemoglobin,
28 Rahimy MC, GangboA, Ahouignan G and Alihonou 41, 278–282.
E (2009) Newborn screening for sickle cell disease 40 Ambrose EE, Makani J, Chami N, Masoza T,
in the Republic of Benin. J Clin Pathol, 62, 46–48. Kabyemera R, Peck RN et al. (2018) High birth prev-
29 Ashtiani MT, Monajemzadeh M, Sina AH, Berenji F, alence of sickle cell disease in Northwestern
Abdollahi M, Said MG and Alam M (2009) Tanzania. Pediatr Blood Cancer, 65, e26735.
Prevalence of haemoglobinopathies in 34,030 41 Monplaisir N, Merault G, Poyart C, Rhoda MD,
healthy adults in Tehran, Iran. J Clin Pathol, 62, Craescu C, Vidaud M et al. (1986) Hb S‐Antilles: a
924–925. new variant with a lower solubility than Hb S and
30 Baghernajad‐Salehi L, D’Apice MR, Babameto‐ producing sickle cell disease in heterozygotes. Proc
Laku A, Biancolella M, Mitre A, Russo S et al. (2009) Natl Acad Sci U S A, 83, 9363–9367.
A pilot beta‐thalassaemia screening program in the 42 Kutlar F, Redding‐Lallinger R, Meiler SE, Bakanay
Albanian population for a health planning pro- SM, Borders L and Kutlar A (2010) A new sickling
gram. Acta Haematol, 121, 234–238. variant ‘Hb S‐Wake β[(Glu6Val‐Asn139 Ser)]’ found
31 Jastaniah W (2011) Epidemiology of sickle cell dis- in a compound heterozygote with Hb S β(Glu6Val)
ease in Saudi Arabia. Ann Saudi Med, 31, 289–293. coinherited with homozygous α‐thalassemia‐2:
32 Giambona A, Vinciguerra M, Cannata M, Cassarà F, phenotype and molecular characteristics. Acta
Fiorentino G, Leto F et al. (2011) The genetic hetero- Haematol, 124, 120–124.
geneity of β‐globin gene defects in Sicily reflects the 43 Geva A, Clark JJ, Zhang Y, Popowicz A, Manning
historic population migrations of the island. Blood JM and Neufeld EJ (2004) Hemoglobin Jamaica
Cells Mol Dis, 46, 282–287. Plain – a sickling hemoglobin with reduced oxygen
33 MMWR (2014) Incidence of sickle cell trait – United affinity. N Engl J Med, 351, 1532–1538.
States, 2010. Morb Mortal Wkly Rep, 63, 1155–1158. 44 Luo HY, Adewoye AH, Eung SH, Skelton TP,
34 Schaefer BA, Kiyaga C, Howard TA, Ndeezi G, Quillen K, McMahon L et al. (2004) A novel sickle
Hernandez AG, Ssewanyana I et al. (2016) hemoglobin: hemoglobin S‐South End. J Pediatr
Hemoglobin variants identified in the Uganda Hematol Oncol, 26, 773–776.
Sickle Surveillance Study. Blood Adv, 1, 93–100. 45 Zanella‐Cleon I, Prehu C, Joly P, Riou J, Becchi M,
35 Tluway F and Makani J (2017) Sickle cell disease in Wajcman H and Francina A (2009) Strategy for
Africa: an overview of the integrated approach to identification by mass spectrometry of a new
health, research, education and advocacy in human hemoglobin variant with two mutations in
Tanzania, 2004–2016. Br J Haematol, 177, 919–929. Cis in the β‐globin chain: Hb S‐Clichy [β6(A3)
36 Tubman VN, Marshall R, Jallah W, Guo D, Ma C, Glu→Val; >8(A5)Lys→Thr]. Hemoglobin, 33,
Ohene‐Frempong K et al. (2016) Newborn screening 177–187.
for sickle cell disease in Liberia: a pilot study. Pediatr 46 Feliu‐Torres A, Eberle SE, Bragós IM, Sciuccati G,
Blood Cancer, 63, 671–676. Ojeda MJ, Calvo KL et al. (2010) Hb S‐San Martin: a
37 Cela E, Bellón JM, de la Cruz M, Beléndez C, new sickling hemoglobin with two amino acid sub-
Berrueco R, Ruiz A et al; SEHOP‐Hemoglobinopathies stitutions [β6(A3)Glu→Val;β105(G7)Leu→Pro] in
Study Group (Sociedad Española de Hematología y an Argentinean family. Hemoglobin, 34, 500–504.
250 Chapter 4
47 Wild BJ, Green BN, Cooper EK, Lalloz MR, Erten S, alpha+‐thalassemia and the sickle cell trait. Nat
Stephens AD and Layton DM (2001) Rapid identifi- Genet, 11, 1253–1257.
cation of hemoglobin variants by electrospray ioni- 57 Ruparelia M, Chambers K and Chapman C (2015)
zation mass spectrometry. Blood Cells Mol Dis, 27, Haemoglobinopathy screening pathways and atti-
691–704. tudes to opportunistic screening for haemoglobi-
48 Jorge SE, Petruk AA, Kimura EM, Oliveira DM, nopathies. Br J Haematol, 169, Suppl. 1, 103.
Caire L, Suemasu CN et al. (2012) Hb S‐São Paulo: a 58 Lacy ME, Wellenius GA, Sumner AE, Correa A,
new sickling hemoglobin with stable polymers and Carnethon MR, Liem RI et al. (2017) Association of
decreased oxygen affinity. Arch Biochem Biophys, sickle cell trait with hemoglobin A1c in African
519, 23–31. Americans. JAMA, 317, 507–515.
49 Steinberg MH and Nagel RL. New and recombi- 59 Cooley JC, Peterson WL, Engel CE and Jernigan JP
nant mutant hemoglobins of biological interest. (1954) Clinical triad of massive splenic infarction,
In: Steinberg MH, Forget BG, Higgs DR and Nagel sicklemia trait, and high altitude flying. JAMA, 154,
RL, eds. Disorders of Hemoglobin: Genetics, 111–113.
Pathophysiology, and Clinical Management. 60 Sheikha A (2005) Splenic syndrome in patients at
Cambridge University Press, Cambridge, 2001, high altitude with unrecognized sickle cell trait:
pp. 1195–1211. splenectomy is often unnecessary. Can J Surg, 48,
50 Harteveld CL, Ponjee G, Bakker‐Verweij M, 317–381.
Arkesteijn SG, Phylipsen M and Giordano PC 61 Goldberg NM, Dorman JP, Riley CA and Ambruster
(2012) Hb Haaglanden: a new nonsickling EJ (1985) Altitude‐related specific infarction in
β7Glu>Val variant. Consequences for basic diag- sickle cell trait: Case reports of a father and son.
nostics, screening, and risk assessment when West J Med, 143, 670–672.
dealing with HbS‐like variants. Int J Lab Hematol, 62 Gitlin SD and Thompson CB (1989) Non‐altitude‐
34, 551–555. related splenic infarction in a patient with sickle cell
51 Witkowska HE, Lubin BH, Beuzard Y, Baruchel S, trait. Am J Hematol, 87, 697–698.
Esseltine DW, Vichinsky EP et al. (1991) Sickle cell 63 Tso AC, Roper DR, Wong CL, Bain B and Layton
disease in a patient with sickle cell trait and DM (2011) Splenic infarction in a patient with sickle
compound heterozygosity for hemoglobin S and cell trait and hereditary spherocytosis. Am J
hemoglobin Quebec‐Chori. N Engl J Med, 325, Hematol, 86, 695–696.
1150–1154. 64 Magid D, Fishman EK and Siegelman SS (1984)
52 Wong YS, Tanaka KR, Greenberg LH and Okada T Computed tomography of the spleen and liver
(1969) Hematuria associated with haemoglobin in sickle cell disease. Am J Roentgenol, 143,
CHarlem: a sickling hemoglobin variant. J Urol, 245–249.
102, 762–764. 65 Baron JM, Weinshelbaum EI and Block GE (1968)
53 Cohen‐Solal M, Préhu C, Wajcman H, Poyart C, Splenic rupture associated with bacterial endocar-
Bardakdjian‐Michau J, Kister J et al. (1998) A new ditis and sickle cell trait. JAMA, 205, 112–114.
sickle cell disease phenotype associating Hb S trait, 66 Domingues MC, Domingues LAW, Ostonoff M,
severe pyruvate kinase deficiency (PK Conakry), Matias C, Araujo AS. Florêncio R et al. (2003) Report
and an α2 globin gene variant (Hb Conakry). Br J of a rare case of splenic sequestration in a patient
Haematol, 103, 950–956. with sickle cell trait. Blood, 102, 31b.
54 Nagel RL, Daar S, Romero JR, Suzuka M, Gravell 67 Jones SR, Binder RA and Donowho EM (1970)
D, Bouhassira E et al. (1998) HbS‐Oman heterozy- Sudden death in sickle‐cell trait. N Engl J Med, 28,
gote: a new dominant sickle syndrome. Blood, 92, 323–325.
4375–4382. 68 Koppes GM, Daly JJ, Coltman CA and Butkus DE
55 Naik RP and Haywood C (2015) Sickle cell trait (1977) Exertion‐induced rhabdomyolysis with
diagnosis: clinical and social implications. acute renal failure and disseminated intravascular
Hematology, 2015, 160–167. coagulation in sickle cell trait. Am J Med, 63,
56 Williams TN, Mwangi TW, Wambua S, Peto TE, 313–317.
Weatherall DJ, Gupta S et al. (2005) Negative 69 Ridha A, Khan A, Al‐Abayechi S and Puthenveetil
epistasis between the malaria‐protective effects of V (2014) Acute compartment syndrome secondary
to rhabdomyolysis in a sickle cell trait patient. disease among African American U.S. Army soldiers.
Lancet, 384, 2172. Br J Haematol, 185, 532–540.
70 Drehner D (1999) Death among U.S. Air Force basic 83 Pecker LH and Naik RP (2018) The current state of
trainees, 1956–1996. Mil Med, 164, 841–847. sickle cell trait: implications for reproductive and
71 Thompson AA (2013) Sickle cell trait testing and genetic counseling. Blood, 132, 2331–2338.
athletic participation: a solution in search of a prob- 84 Adsay NV, deRoux SJ, Sakr W and Grignon D
lem? Hematology Am Soc Hematol Educ Program 2013, (1999) Cancer as a marker of genetic medical dis-
2013, 632–637. ease: an unusual case of medullary carcinoma of the
72 Nelson DA, Deuster PA, Carter R, Hill OT, Wolcott kidney. Am J Surg Pathol, 22, 260–264.
VL and Kurina LM (2016) Sickle cell trait, rhabdo- 85 Patel K, Livni N and Macdonald D (2006) Renal
myolysis, and mortality among U.S. Army Soldiers. medullary carcinoma, a rare cause of haematuria in
N Engl J Med, 375, 435–442. sickle cell trait. Br J Haematol, 132, 1.
73 Ajayi AA and Kolawole BA (2004) Sickle cell trait 86 Dimashkieh H, Choe J and Mutema G (2003) Renal
and gender influence type 2 diabetic complications medullary carcinoma: a report of 2 cases and
in African patients. Eur J Intern Med, 15, 312–315. review of the literature. Arch Pathol Lab Med, 127,
74 McAnneny A, Durden F, Pearson GD and Tiwari P e135–e138.
(2012) Intra‐flap thrombosis secondary to acute sickle 87 Srivastava S, Srivastava R and Ghorpade KG (2017)
crisis: a case report. Microsurgery, 32, 585–587. Sickled erythrocytes in urine as a clue to the diag-
75 Ouzzif Z, El Maataoui A, Traore Z, Biaz A, El nosis of sickle cell trait. Saudi J Kidney Dis Transpl,
Machtani S, Dami A et al. (2017) A retinopathy in 28, 909–911.
young patient with co‐inheritance of heterozygous 88 Rimer BA (1975) Sickle‐cell trait and pregnancy: a
alpha +‐thalassemia and sickle trait: a case report. review of a community hospital experience. Am J
BMC Ophthalmol, 17, 6. Obstet Gynec, 123, 6–9.
76 Xu JZ and Thein SL (2019) The carrier state for 89 Larrabee KD and Monga M (1997) Women with
sickle cell disease is not completely harmless. sickle cell trait are at increased risk of preeclampsia.
Haematologica, 104, 1106–1111. Am J Obstet Gynecol, 177, 425–428.
77 Gupta AK, Kirchner KA, Nicholson R, Adams JG, 90 Agrawal R, Berube C, Steinberg G and George TI
Schechter AN, Noguchi CT and Steinberg MH (2013) Moyamoya syndrome with sickle cell trait.
(1991) Effects of alpha‐thalassemia and sickle Int J Lab Hematol, 35, e8–e9.
polymerization tendency on the urine‐concentrat- 91 Caughey MC, Loehr LR, Key NS, Derebail VK,
ing defect of individuals with sickle cell trait. J Clin Gottesman RF, Kshirsagar AV et al. (2014) Sickle cell
Invest, 88, 1963–1968. trait and incident ischemic stroke in the
78 Naik RP, Wilson JG, Ekunwe L, Mwasongwe S, Atherosclerosis Risk in Communities study. Stroke,
Duan Q, Li Y et al. (2016) Elevated D‐dimer levels in 45, 2863–2867.
African Americans with sickle cell trait. Blood, 127, 92 Austin H, Key NS, Benson JM, Lally C, Dowling
2261–2263. NF, Whitsett C and Hooper WC (2007) Sickle cell
79 Bucknor MD, Goo JS and Coppolino ML (2014) The trait and the risk of venous thromboembolism
risk of potential thromboembolic, renal and cardiac among blacks. Blood, 110, 908–912.
complications of sickle cell trait. Hemoglobin, 38, 93 Poehling KA, Light LS, Rhodes M, Snively BM,
28–32. Halasa NB, Mitchel E et al. (2010) Sickle cell trait,
80 Key NS, Connes OP and Derebail VK (2015) hemoglobin C trait, and invasive pneumococcal
Negative health implications of sickle cell trait in disease. Epidemiology, 21, 340–346.
high income countries: from the football field to the 94 Jackson H, Bentley CR, Hingorani M, Atkinson P,
laboratory. Br J Haematol, 170, 5–14. Aclimandos WA and Thompson GM (1995) Sickle
81 Naik RP, Irvin MR, Judd S, Gutiérrez OM, Zakai retinopathy in patients with sickle trait. Eye, 9,
NA, Derebail VK et al. (2017) Sickle cell trait and the 589–593.
risk of ESRD in Blacks. J Am Soc Nephrol, 28, 95 Steinberg MH and Embury SH (1986) β‐thalassae-
2180–2187. mia in blacks: genetic and clinical aspects and inter-
82 Hu J, Nelson DA, Deuster PA, Marks ES, O’Connor actions with sickle hemoglobin gene. Blood, 69,
FG and Kurina LM (2019) Sickle cell trait and renal 985–990.
252 Chapter 4
96 Wambua S, Mwacharo J, Uyoga S, Macharia A cis of beta+‐thalassemia and hemoglobin S. Am J

and Williams TN (2006) Co‐inheritance of α+ tha- Hematol, 26, 237–245.
lassaemia and sickle trait results in specific effects 108 Baklouti F, Ouazana R, Gonnet C, Lapillonne A,
on haematological parameters. Br J Haematol, 133, Delaunay J and Godet J (1989) β+‐Thalassemia in cis
206–109. of a sickle cell gene ‐ occurrence of a promotor muta-
97 Felice AE, Altay CA, Milner PF and Huisman THJ tion on a βS chromosome. Blood, 74, 1817–1822.
(1981) The occurrence and identification of α‐tha- 109 Joly P, Schluth‐Bolard C, Lacan P, Barro C, Pissard S,
lassaemia‐2 among hemoglobin S heterozygotes. Labalme A et al. (2013) HBB loss of heterozygosity in
Am J Clin Pathol, 76, 70–73. the hemopoietic lineage gives rise to an unusual
98 Head CE, Conroy M, Jarvis M, Phelan L and Bain sickle‐cell trait phenotype. Haematologica, 98, e7–e8.
BJ (2004) Some observations on the measurement 110 Shooter C, McKenzie TS, Oakley M, Jacques T,
of haemoglobin A2 and S percentages by high per- Clark B and Thein SL (2015) First reported dupli-
formance liquid chromatography in the presence cation of the entire beta globin gene cluster caus-
and absence of thalassaemia. J Clin Pathol, 57, ing an unusual sickle cell trait phenotype. Br J
276–280. Haematol, 170, 128–131.
99 Steinberg MH (1991) The interactions of β‐thalas- 111 Levere RD, Lichtman HC and Levine J (1964)
saemia with hemoglobinopathies. Hematol/Oncol Effect of iron deficiency anaemia on the metabo-
Clinics North Am, 5, 453–473. lism of the heterogenic haemoglobins in sickle cell
100 Beutler E and West C (2005) Hematologic differ- trait. Nature, 202, 499–501.
ences between African‐Americans and whites; the 112 Heller P, Yakulis VJ, Epstein RB and Friedland S
role of iron deficiency and α‐thalassemia on hemo- (1963) Variation in the amount of hemoglobin S in
globin levels and mean corpuscular volume. a patient with sickle cell trait and megaloblastic
Blood, 106, 740–745. anemia. Blood, 21, 479–483.
101 Mears JG, Lachman HM, Labie D and Nagel RL 113 Steinberg MH. Sickle cell trait. In: Steinberg MH,
(1983) Alpha‐thalassemia is related to prolonged Forget BG, Higgs DR and Nagel RL, eds. Disorders
survival in sickle cell anemia. Blood, 62, 286–290. of Hemoglobin: Genetics, Pathophysiology, and
102 Wilson CI, Hopkins PL, Cabello‐Inchausti B, Clinical Management. Cambridge University Press,
Melnick SJ and Robinson MJ (2000) The peripheral Cambridge, 2001, pp. 811–830.
blood smear in patients with sickle cell trait: a 114 Zurbriggen K, Schmugge M, Schmid M, Durka S,
morphologic observation. Lab Med, 31, 445–447. Kleinert P, Kuster T et al. (2002) Analysis of minor
103 Mowafy N (2005) Hematomorphology. Hematology, hemoglobins by matrix‐assisted laser desorption/
10, Suppl. 1, 182–185. ionization time‐of‐flight mass spectrometry. Clin
104 Thompson GR (1963) Malaria and stress in Chem, 51, 989–996.
relation to haemoglobins S and C. Br Med J, ii, 115 Akinsheye I, Alsultan A, Solovieff N, Ngo D,
976–978. Baldwin CT, Sebastiani P et al. (2011) Fetal hemo-
105 Bain BJ and Phelan L (1996) An evaluation of the globin in sickle cell anemia. Blood, 118, 19–27.
HemoCard Hemoglobin S test and four sickle cell 116 Matthay KK, Mentzer WC, Dozy AM, Kan YW
solubility kits (Ortho Sickledex, Dade Sickle‐Sol, and Bainton DF (1979) Modification of hemo-
Microgen Bioproducts S‐Test and Lorne Sickle‐ globin H disease by sickle trait. J Clin Invest, 64,
Check) for the detection of haemoglobin S and the 1024–1032.
HemoCard Hemoglobin A plus S test for the 117 Steinberg MH. Compound heterozygous and other
detection of haemoglobins A and S. MDA sickle haemoglobinopathies. In: Steinberg MH,
Evaluation Report MDA/96/56. Medical Devices Forget BG, Higgs DR and Nagel RL, eds. Disorders
Agency, London. of Hemoglobin: Genetics, Pathophysiology, and Clinical
106 Higgs DR, Clegg JB, Weatherall DJ, Serjeant BE Management. Cambridge University Press,
and Serjeant GR (1984) The interaction of the ααα Cambridge, 2001, pp. 786–810.
globin gene haplotype and sickle haemoglobin. Br 118 Ustun C, Kutlar F, Holley L, Seigler N, Burgess R
J Haematol, 58, 671–678. and Kutlar A (2003) Interaction of sickle cell trait
107 Baklouti F, Francina A, Dorleac E, Richard G, with hereditary spherocytosis: splenic infarcts and
Rosenberg D, Godet J et al. (1987) Association in sequestration. Acta Haematol, 109, 46–49.
119 Quinn CT, Smith EP, Arbabi S, Khera PK, Lindsell 132 Ataga KI, Moore CG, Jones S, Olajide O, Strayhorn
CJ, Niss O et al. (2016) Biochemical surrogate D, Hinderliter A and Orringer EP (2006)
markers of hemolysis do not correlate with Pulmonary hypertension in patients with sickle
directly measured erythrocyte survival in sickle cell disease: a longitudinal study. Br J Haematol.
cell anemia. Am J Hematol, 91, 1195–1201. 134, 109–115.
120 Piel FB, Steinberg MH and Rees DC (2017) Sickle 133 Niss O, Fleck R, Makue F, Alsaied T, Desai P,
cell disease. N Engl J Med, 376, 1561–1573. Towbin JA et al. (2017) Association between dif-
121 Gardner K, Douiri A, Drasar E, Allman M, Mwirigi fuse myocardial fibrosis and diastolic dysfunction
A, Awogbade M and Thein SL (2016) Survival in in sickle cell anemia. Blood, 130, 205–213.
adults with sickle cell disease in a high‐income 134 Desai PC, Kendel N, Huang Y, Heinlein M, Kraut
setting. Blood, 128, 1436–1438. E and Raman S (2019) Hypoxia in sickle cell dis-
122 DeBaun MR, Ghafuri DL, Rodeghier M, Maitra P, ease due to right to left shunting. Am J Hematol, 94,
Chaturvedi S, Kassim A and Ataga KI (2019) E53–E55.
Decreased median survival of adults with sickle 135 Adekile AD, Owunwanne A, Al‐Za’abi K, Haider
cell disease after adjusting for left truncation bias: MZ, Tuli M and Al‐Mohannadi S (2002) Temporal
a pooled analysis. Blood, 133, 615–617. sequence of splenic dysfunction in sickle cell dis-
123 Brunson A, Keegan THM, Bang H, Mahajan A, ease. Am J Hematol, 69, 23–27.
Paulukonis S and Wun T (2017) Increased risk of 136 Rogers ZR, Wang WC, Luo Z, Iyer RV, Shalaby‐
leukemia among sickle cell disease patients in Rana E, Dertinger SD et al. (2011) Biomarkers of
California. Blood, 130, 1597–1599. splenic function in infants with sickle cell anemia:
124 Topley JM, Rogers DW, Stevens MC and Serjeant baseline data from the BABY HUG Trial. Blood,
GR (1981) Acute splenic sequestration and hyper- 117, 2614–2617.
splenism in the first five years in homozygous 137 Sadarangani M, Makani J, Komba AN, Ajala‐Agbo
sickle cell disease. Arch Dis Child, 56, 765–769. T, Newton CR, Marsh K and Williams TN (2009) An
125 Guth AA, Pachter HL and Jacobowitz GR (1996) observational study of children with sickle cell dis-
Rupture of the pathologic spleen: is there a role for ease in Kilifi, Kenya. Br J Haematol, 146, 675–682.
nonoperative therapy? J Trauma, 41, 214–218. 138 Italia K, Kangne H, Shanmukaiah C, Nadkarni
126 Shoemaker MT, Pitney AC, Harford DJ and Barker AH, Ghosh K and Colah RB (2015) Variable
JA (2004) Acute splenic rupture in an adult with phenotypes of sickle cell disease in India with
homozygous sickle cell anemia treated with chronic the Arab‐Indian haplotype. Br J Haematol, 168,
transfusions. J Pediatr Hematol Oncol, 26, 849–851. 156–158.
127 Sharma D (2009) Sub capsular splenectomy for 139 Issa H and Al‐Salem AH (2010) Hepatobiliary man-
delayed spontaneous splenic rupture in a case of ifestations of sickle cell anemia. Gastroenterology
sickle cell anemia. World J Emerg Surg, 4, 17. Res, 3, 1–8.
128 Ogunsile FJ, Currie KL, Rodeghier M, Kassim A, 140 Karkoska K, Louie J, Appiah‐Kubi AO, Wolfe L,
DeBaun MR and Sharma D (2018) History of par- Rubin L, Rajan S and Aygun B (2018) Transfusion‐
vovirus B19 infection is associated with silent cer- transmitted babesiosis leading to severe hemoly-
ebral infarcts. Pediatr Blood Cancer, 65, e26767. sis in two patients with sickle cell anemia. Pediatr
129 Bao EL, Lareau CA, Brugnara C, Fulcher IR, Barau Blood Cancer, 65, e26734.
C, Moutereau S et al. (2019) Heritability of fetal 141 Ataga KI, Derebail VK and Archer DR (2014) The
hemoglobin, white cell count, and other clinical glomerulopathy of sickle cell disease. Am J
traits from a sickle cell disease family cohort. Am J Hematol, 89, 907–914.
Hematol, 94, 522–527. 142 Vazquez B, Shah B, Zhang X, Lash JP, Gordeuk VR
130 Mian UK, Tang J, Allende APM, Heo M, Bernstein and Saraf SL (2014) Hyperfiltration is associated
N, Vattappally L et al. (2018) Elevated fetal haemo- with the development of microalbuminuria in
globin levels are associated with decreased inci- patients with sickle cell anemia. Am J Hematol, 89,
dence of retinopathy in adults with sickle cell 1156–1157.
disease. Br J Haematol, 183, 807–811. 143 Lebensburger JD, Aban I, Pernell B, Kasztan M,
131 Farber HW and Loscalzo J (2004) Pulmonary arte- Feig DI, Hilliard LM and Askenazi DJ (2019)
rial hypertension. N Engl J Med, 351, 1655–1665. Hyperfiltration during early childhood precedes
254 Chapter 4
albuminuria in pediatric sickle cell nephropathy. 154 Elmariah H, Garrett ME, De Castro LM, Jonassaint
Am J Hematol, 94, 417–423. JC, Ataga KI, Eckman JR et al. (2014) Factors associ-
144 Aban I, Baddam S, Hilliard LM, Howard TH, Feig ated with survival in a contemporary adult sickle
DI and Lebensburger JD (2017) Severe anemia cell disease cohort. Am J Hematol, 89, 530–535.
early in life as a risk factor for sickle‐cell kidney 155 Neonato MG, Guilloud‐Bataille M, Beauvais P,
disease. Blood, 129, 385–387. Bégue P, Belloy M, Benkerrou M et al. (2000)
145 Addae S, Adzaku F, Mohammed S and Annobil S Acute clinical events in 299 homozygous sickle
(1990) Sickle cell disease in permanent residents of cell patients living in France. Eur J Haematol, 65,
mountain and low altitudes in Saudi Arabia. Trop 155–164.
Geogr Med, 42, 342–348. 156 McGann PT, Hernandez AG and Ware RE (2017)
146 Chopra R, Al‐Mulhim AR and Al‐Baharani AT Sickle cell anemia in sub‐Saharan Africa: advanc-
(2005) Fibrocongestive splenomegaly in sickle cell ing the clinical paradigm through partnerships
disease: a distinct clinicopathological entity in the and research. Blood, 129, 155–161.
Eastern province of Saudi Arabia. Am J Hematol, 157 Higgs DR, Aldridge BE, Lamb J, Clegg JB,
79, 180–186. Weatherall DJ, Hayes RJ et al. (1982) The interac-
147 Steinberg MH. Overview of sickle cell anemia tion of alpha‐thalassaemia and homozygous
pathophysiology. In: Costa FF and Conran N, eds. sickle‐cell disease. N Engl J Med, 306, 1441–1446.
Sickle cell Anemia. Springer, Heidelberg, 2016, pp. 158 Nouraie M, Reading NS, Campbell A, Minniti CP,
49–73. Rana SR, Luchtman‐Jones L et al. (2010) Association
148 Purohit P, Patel S, Mohanty PK, Das P and of G6PD with lower haemoglobin concentration
Panigrahi J (2016) Fetal hemoglobin modifies the but not increased haemolysis in patients with
disease manifestation of severe Plasmodium falcipa- sickle cell anaemia. Br J Haematol, 150, 218–225.
rum malaria in adult patients with sickle cell ane- 159 Bailey S, Higgs DR, Morris J and Serjeant GR
mia. Med J Hematol Infect Dis, 8, e2016055. (1991) Is the painful crisis of sickle‐cell disease due
149 Habibi A, Mekontso‐Dessap A, Guillaud C, Michel to sickling? Lancet, 337, 735.
M, Razazi K, Khellaf M et al. (2016) Delayed hemo- 160 Castro O and Haddy TB (1983) Improved survival
lytic transfusion reaction in adult sickle‐cell of iron‐deficient sickle erythrocytes. N Engl J Med,
disease: presentations, outcomes, and treatments 308, 527.
of 99 referral center episodes. Am J Hematol, 91, 161 Ballas SK (1998) Sickle cell disease: clinical man-
989–994. agement. Bailliere’s Clin Haematol, 11, 185–214.
150 King KE, Shirey RS, Lankiewicz MW, Young‐ 162 Ballas SK, Gay RN and Chehab FF (1997) Is Hb A2
Ramsaran J and Ness PM (1997) Delayed hemo- elevated in adults with sickle‐α‐thalassemia
lytic transfusion reactions in sickle cell disease: (βS/βS;‐α/‐α)? Hemoglobin, 21, 405–450.
simultaneous destruction of patient’ red cells. 163 Nolan VG, Wyszynski DF, Farrer LA and Steinberg
Transfusion, 37, 376–381. MH (2005) Hemolysis‐associated priapism in
151 Shafeek S, Jabbar M, Strevens MJ, Oakley S, sickle cell disease. Blood, 106, 3264–3267.
Bareford D and Smith N (2000) Two cases of post 164 Mukherjee MB, Lu CY, Ducrocq R, Gangakhedkar
transfusion hyperhaemolysis in sickle cell disease. RR, Colah RB, Kadam MD et al. (1997) Effect of α‐
Br J Haematol, 108, Suppl. 1, 38. thalassaemia on sickle cell anemia linked to the
152 Platt OS, Brambilla DJ, Rosse WF, Milner PF, Castro Arab‐Indian haplotype in India. Am J Hematol, 55,
O, Steinberg MH and Klug PP (1994) Mortality in 104–109.
sickle cell disease: life expectancy and risk factors 165 Wonkam A, Rumaney MB, Bitoungui VJN, Vorster
for early death. N Engl J Med, 330, 1639–1644. AA, Ramesar R and Ngogang J (2014)
153 Ohene‐Frempong K and Steinberg MH. Clinical Coinheritance of sickle cell anemia and α‐thalas-
aspects of sickle cell anemia in adults and chil- semia delays disease onset and could improve
dren. In: Steinberg MH, Forget BG, Higgs DR and survival in Cameroonian patients (Sub‐Saharan
Nagel RL, eds. Disorders of Hemoglobin: Genetics, Africa). Am J Hematol, 89, 664–665.
Pathophysiology, and Clinical Management. 166 Davis LR (1976) Changing blood picture in sickle‐
Cambridge University Press, Cambridge, 2001, cell anaemia from shortly after birth to adoles-
pp. 611–670. cence. J Clin Pathol, 29, 898–901.
167 Serjeant GR, Grandison Y, Lowrie Y, Mason K, Study Group (1996) Hematologic profile and
Phillips J, Serjeant BE and Vaidya S (1981) The lymphocyte subpopulations in hemoglobin SC
development of haematological changes in disease: comparison with SS and black controls.
homozygous sickle cell disease: a cohort study Am J Hematol, 52, 150–154.
from birth to 6 years. Br J Haematol, 48, 533–543. 178 Moliterno AR, Williams DM, Gutierrez‐Alamillo
168 Meier ER, Wright EC and Miller JL (2014) LI, Salvatori R, Ingersoll RG and Spivak JL (2004)
Reticulocytosis and anemia are associated with an Mpl Baltimore: a thrombopoietin receptor poly-
increased risk of death and stroke in the newborn morphism associated with thrombocytosis. Proc
cohort of the Cooperative Study of Sickle Cell Natl Acad Sci U S A, 101, 11444–11447.
Disease. Am J Hematol, 89, 904–906. 179 Embury SH, Dozy AM, Miller J, Davis JR, Kleman
169 Saraf S, Farooqui M, Infusino G, Oza B, Sidhwani KM, Preisler H et al. (1982) Concurrent sickle‐cell
S, Gowhari M et al. (2011) Standard clinical prac- anemia and α‐thalassemia: effect on severity of
tice underestimates the role and significance of anemia. N Engl J Med, 306, 270–274.
erythropoietin deficiency in sickle cell disease. Br J 180 Rabb LM, Grandison Y, Mason K, Hayes RJ,
Haematol, 153, 386–392. Serjeant B and Serjeant GR (1983) A trial of folate
170 Head CA, Brugnara C, Martinez‐Ruiz R, Kacmarek supplementation in children with homozygous
R M, Bridges K R, Kuter D et al. (1997) Low con- sickle cell disease. Br J Haematol, 54, 589–594.
centrations of nitric oxide increase oxygen affinity 181 Dhar M, Bellevue R and Carmel R (2003)
of sickle erythrocytes in vitro and in vivo. J Clin Pernicious anemia with neuropsychiatric dys-
Invest, 100, 1193–1198. function in a patient with sickle cell anemia
171 Sherwood JB, Goldwasser E, Chilcote R, treated with folate supplementation. N Engl J Med,
Carmichael LD and Nagel RL (1986) Sickle cell 349, 2204–2207.
anemia patients have low erythropoietin levels for 182 Ballas SK and Smith ED (1992) Red blood cell
their degree of anemia. Blood, 67, 46–49. changes during the evolution of the sickle cell
172 Fabry M and Old JM, Laboratory methods for painful crisis. Blood, 29, 2154–2163.
diagnosis and evaluation of hemoglobin disor- 183 Vichinsky E and Styles E (1996) Pulmonary com-
ders. In: Steinberg MH, Forget BG, Higgs DR and plications. Hematol/Oncol Clin North Am, 10,
Weatherall DJ, eds. Disorders of Hemoglobin, 2nd 1275–1287.
edn. Cambridge University Press, Cambridge, 184 Siow W, Matthey F and Bain BJ (2017) The signifi-
2009, pp. 658–686. cance of irregularly contracted cells and hemig-
173 Dover GJ, Chang VT, Boyer SH, Sergeant GR, hosts in sickle cell disease. Am J Hematol, 92.
Antonarakis S and Higgs DR (1987) The cellular 966–967.
basis for different fetal hemoglobin levels among 185 Winchenne A, Cecchini J, Deux JF, De Prost N,
sickle cell individuals with two, three and four Razazi K, Carteaux G et al. (2017) A clinical risk
alpha‐globin genes. Blood, 69, 341–344. score for pulmonary artery thrombosis during
174 Thame M, Grandison Y, Mason K. Thompson M, acute chest syndrome in adult patients with sickle
Higgs D, Morris J et al. (1991) The red cell distribu- cell disease. Br J Haematol, 179, 627–634.
tion width in sickle cell disease – is it of clinical 186 Alvarez O, Montague NS, Marin M, O’Brien R and
value? Clin Lab Haematol, 13, 229–237. Rodriguez MM (2015) Quantification of sickle
175 Castro O, Brambilla DJ, Thorington B, Reindorf cells in the peripheral smear as a marker of dis-
CA, Scott RB, Gillette P et al. and the Cooperative ease severity. Fetal Pediatr Pathol, 34, 149–154.
Study of Sickle Cell Disease (1994) The acute chest 187 Bain BJ (1993) Blood film features of sickle cell‐
syndrome in sickle cell disease: incidence and risk haemoglobin C disease. Br J Haematol, 83,
factors. Blood, 84, 643–649. 516–518.
176 Tassel C, Arnaud C, Kulpa M, Fleurence E, Kandem 188 Weinstein RS, Warth JA, Near K and Marikovsky
A, Madhi F et al. (2011) Leukocytosis is a risk factor Y (1989) Sequestrocytes: a manifestation of trans-
for lung function deterioration in children with cellular cross‐bonding of the red cell membrane in
sickle cell disease. Respir Med, 105, 788–795. sickle cell anemia. J Cell Sci, 94, 593–600.
177 Wong W‐Y, Zhou Y, Operskalski EA, Hassett J, 189 Ward PC, Smith CM and White JG (1979)
Powars DR, Mosley JW and the Transfusion Safety Erythrocytic ecdysis. An unusual morphologic
256 Chapter 4
finding in a case of sickle cell anemia with inter- 202 Smith SGW, Glass UH, Acharya J and Pearson TC
current cold‐agglutinin syndrome. Am J Clin (1989) Pulse oximetry in sickle cell disease. Clin
Pathol, 72, 479–485. Lab Haematol, 11, 185–188.
190 Austin A, Lund K and Bain BJ (2019) Striking dys- 203 Kato GJ, McGowan V, Machado RF, Little JA,
erythropoiesis in sickle cell anemia following an Taylor J, Morris CR et al. (2006) Lactate dehydro-
aplastic crisis. Am J Hematol, 94, 378. genase as a biomarker of hemolysis‐associated
191 Adekile AD, Al‐Zaabi K, Haider MZ and Tuli M nitric oxide resistance, priapism, leg ulceration,
(1997) Molecular and hematological correlates of pulmonary hypertension, and death in patients
splenic function among Arab SS patients. Blood, with sickle cell disease. Blood, 107, 2279–2285.
90, Suppl. 1, 216. 204 Goldberg MA, Brugnara C, Dover GJ, Schapira L,
192 Bunn HF. Sickle hemoglobin and other hemo- Charache S and Bunn HF (1990) Treatment of
globin mutants. In: Stamatoyannopoulos G, sickle cell anemia with hydroxyurea and erythro-
Nienhuis AW, Majerus PW and Varmus H, eds. poietin. N Engl J Med, 323, 366–372.
The Molecular Basis of Blood Diseases, 2nd edn, WB 205 Leiva O, McMahon L, Sloan JM, Lee J and Lerner
Saunders, Philadelphia, 1994, pp. 207–256. A (2019) Recognition of hemophagocytic lympho-
193 Warth JA and Rucknagel DL (1984) Density histiocytosis in sickle cell vaso‐occlusive crises is a
ultracentrifugation during and after pain crisis; potentially lifesaving diagnosis. Haematologica,
increased dense echinocytes in crisis. Blood, 64, 104, e167–e169.
507–515. 206 Bain BJ and Lee L (2010) Pseudo‐Gaucher cells in
194 Griffin PJ, Sebastiani P, Edward H, Baldwin CT, sickle cell anemia. Am J Hematol, 85, 435.
Gladwin MT, Gordeuk VR et al. (2014) The genet- 207 Habara AH, Shaikho EM and Steinberg MH (2017)
ics of hemoglobin A2 regulation in sickle cell ane- Fetal hemoglobin in sickle cell anemia: The Arab‐
mia. Am J Hematol, 89, 1019–1023. Indian haplotype and new therapeutic agents. Am
195 Almeida AM, Henthorn JS and Davies SC (2001) J Hematol, 92, 1233–1242.
Haemoglobin F levels in patients with sickle cell 208 Merghoub T, Perichon B, Maier‐Redelsperger M,
diseases. Blood, 98, 13b. Dibenedetto SP, Samperi P, Ducrocq R et al. (1997)
196 Steinberg MH (1998) Pathophysiology of Dissection of the association status of two poly-
sickle cell disease. Bailliere’s Clin Haematol, 11, morphisms in the β‐globin gene cluster with vari-
163–184. ations in F‐cell number in non‐anemic individuals.
197 Smetanina NS, Gu L‐H and Huisman THJ (1998) Am J Hematol, 56, 239–243.
Comparison of the relative quantities of γ‐messen- 209 Gonçalves I, Ducrocq R, Lavinha J, Nogueira PJ,
ger RNA and fetal hemoglobin in SS patients with Peres MJ, Picanco I et al. (1998) Combined effect of
different haplotypes. Acta Haematol, 100, 4–8. two different polymorphic sequences within the β
198 Melcher K and Hoyer J (2009) Using quantitative globin gene cluster on the level of Hb F. Am J
hemoglobin A2 level and Mcv to discriminate Hematol, 57, 269–276.
between homozygous haemoglobin S and hemo- 210 Thomas JJ, Kutlar A, Scott DF and Lanclos KD
globin S/beta‐zero thalassaemia. Int J Lab Hematol, (1998) Inhibition of gene expression by the Gγ 5′
31, Suppl. 1, 85–86. flanking region of the Bantu βS chromosome. Am J
199 Vasseur C, Domingues‐Hamdi E, Pakdaman S, Hematol, 59, 51–56.
Barau C, Pissard S, Le Corvoisier P et al. (2017) 211 Vathipadiekal V, Farrell JJ, Wang S, Edward HL,
Elevated soluble α‐hemoglobin pool in sickle cell Shappell H, Al‐Rubaish AM et al. (2016) A candi-
anemia. Am J Hematol, 92, E593–E595. date transacting modulator of fetal hemoglobin
200 Bain BJ and Phelan L (1996) An assessment of gene expression in the Arab‐Indian haplotype of
HemoCard Hemoglobin C and HemoCard sickle cell anemia. Am J Hematol, 91, 1118–1122.
Hemoglobin E kits for the detection of haemo- 212 Leonardo FC, Brugnerotto AF, Domingos IF,
globins C and E. MDA Evaluation Report Fertrin KY, de Albuquerque DM, Bezerra AC et al.
MDA/96/57. Medical Devices Agency, London. (2016) Reduced rate of sickle‐related complica-
201 US Department of Health and Human Services tions in Brazilian patients carrying HbF‐promot-
(1993) Guideline: laboratory screening for sickle ing alleles at the BCL11A and HMIP‐2 loci. Br J
cell disease. Lab Med, 24, 515–522. Haematol, 173, 456–460.
213 Marcus SJ, Kinney TR, Schultz WH, O’Branski EE 225 Gangaraju R, May JE, Williams LA, Reddy VB,
and Ware RE (1997) Quantitative analysis of eryth- MacLennan P and Marques MB (2019) Fat embo-
rocytes containing fetal hemoglobin F (F cells) in lism syndrome due to bone marrow necrosis in
children with sickle cell disease. Am J Hematol, 54, patients with hemoglobinopathies: A life‐threaten-
40–46. ing complication mimicking thrombotic thrombo-
214 Chang YP, Maier‐Redelsperger M, Smith KD, cytopenic purpura. Am J Hematol, 94, E64–E66.
Contu L, Ducrocq R, de Montalembert M et al. 226 Gualandro SF, Fonseca GH, Yokomizo IK,
(1997) The relative importance of the X‐linked Gualandro DM and Suganuma LM (2015) Cohort
FCP locus and β‐globin haplotypes in determining study of adult patients with haemoglobin SC dis-
haemoglobin F levels: a study of SS patients ease: clinical characteristics and predictors of mor-
homozygous for βS haplotype. Br J Haematol, 96, tality. Br J Haematol, 171, 631–637.
806–814. 227 Naessens V, Ward R and Kuo KHM (2018) A pro-
215 Milner PF, Garbutt GJ, Nolan‐Davis LB and posed treatment algorithm for adults with haemo-
Wilson JT (1986) The effect of hemoglobin F and α globn SC disease. Br J Haematol, 182, 607–609.
thalassaemia on the red cell indices in sickle cell 228 Ware M, Tyghter D, Staniforth S and Serjeant G
anemia. Am J Hematol, 21, 383–395. (1998) Airline travel in sickle cell disease. Lancet,
216 Rising JA, Sautter RL and Spicer SJ (1974) 352, 652.
Hemoglobin G‐Philadelphia/S. Am J Clin Pathol, 229 Zimmerman SA and Ware RE (2000) Palpable
61, 92–102. splenomegaly in children with haemoglobin SC
217 Dacie J. The Haemolytic Anaemias, Volume 2, The disease: haematological and clinical manifesta-
Hereditary Haemolytic Anaemias, Part 2, 3rd edn. tions. Clin Lab Haematol, 22, 145–150.
Churchill Livingstone, Edinburgh, 1988. 230 Carneiro‐Proietti ABF, Kelly S, Miranda Teixeira
218 Krauss JS, Bures K and Kenimer E (1999) The elu- C, Sabino EC, Alencar CS, Capuani L et al.;
tion of αMontgomery2βS2 hybrid tetramers by the International Component of the NHLBI Recipient
variantTM apparatus. Blood, 94, Suppl. 1, 25b. Epidemiology and Donor Evaluation Study
219 Benkerrou M, Alberti C, Couque N, Haouari Z, Ba (REDS‐III) (2018) Clinical and genetic ancestry
A, Missud F et al. (2013) Impact of glucose‐6‐phos- profile of a large multi‐centre sickle cell disease
phate dehydrogenase deficiency on sickle cell anae- cohort in Brazil. Br J Haematol, 182, 895–908.
mia expression in infancy and early childhood: a 231 Williams S, Maude GH and Serjeant GR (1986)
prospective study. Br J Haematol, 163, 646–654. Clinical presentation of sickle cell‐hemoglobin C
220 Lionnet F, Hammoudi N, Stojanovic KS, Avellino disease. J Pediatr, 109, 586–589.
V, Grateau G, Girot R and Haymann J‐P (2012) 232 Kammeyer R, Devnani R and Mehta R (2016)
Hemoglobin SC disease complications: a clinical Cerebral fat embolism syndrome mimicking throm-
study of 179 cases. Haematologica, 97, 1136–1141. botic thrombocytopenic purpura in a patient with
221 Nagel RL, Fabry ME and Steinberg MH (2003) The hemoglobin SC disease. Am J Hematol, 91, 539–542.
paradox of hemoglobin SC disease. Blood Rev, 17, 233 Pecker LH, Schaefer BA and Luchtman‐Jones L
167–178. (2017) Knowledge insufficient: the management
222 Adams JG and Heller P (1977) Hemoglobin of haemoglobin SC disease. Br J Haematol, 176,
Arlington Park. A new hemoglobin variant with 515–526.
two amino acid substitutions in the beta chain. 234 Lee K, Préhu C. Mérault G, Kéclard L, Roudot‐
Hemoglobin, 1, 419–426. Thoroval F, Bachir D et al. (1998) Genetic and
223 Nisbet‐Brown E, Stehmaier K, Walker L, Waye JS hematological studies in a group of 114 adult
and Chui DHK (2000) Sickle cell disease in a four patients with SC sickle cell disease. Am J Hematol,
year old child with apparent HbS trait. Blood, 59, 15–21.
96, 20b. 235 Serjeant GR and Serjeant BE (1972) A comparison
224 Powars DR, Hiti A, Ramicone E, Johnson C and of erythrocytic characteristics in sickle cell syn-
Chan L (2002) Outcome in hemoglobin SC disease: dromes in Jamaica. Br J Haematol, 23, 205–213.
a four‐decade observational study of clinical, 236 Ballas SK, Larner J, Smith ED, Surrey S, Schwartz
hematologic, and genetic factors. Am J Hematol, 70, E and Rappaport EF (1987) The xerocytosis of SC
206–215. disease. Blood, 69, 124–128.
258 Chapter 4
237 Nagel RL and Lawrence C (1991) The distinct 250 Yang Y‐M, Donnell CA, Farrer JH and Mankad
pathobiology of sickle cell–hemoglobin C disease. VN (1990) Molecular characterization of β‐globin
Hematol Oncol Clin North Am, 5, 433–451. gene mutations in Malay patients with Hb E‐β‐
238 Ballas SK and Kosher W (1988) Erythrocytes in Hb thalassaemia and thalassaemia major. Br J
SC disease are microcytic and hyperchromic. Am J Haematol, 72, 73–80.
Hematol, 28, 37–39. 251 Christakis J, Vavatsi N, Hassapopoulou H,
239 Steinberg MH and Hebbel RP (1983) Clinical Angeloudi M, Papadopoulou M, Loukopoulos D
diversity of sickle cell anemia: genetic and cellular et al. (1991) A comparison of sickle cell syndromes
modulation of disease severity. Am J Hematol, 14, in northern Greece. Br J Haematol, 77, 386–391.
405–416. 252 Kuloziz AE, Bail S, Kar BC, Serjeant BE and
240 Diggs LW and Bell A (1965) Intraerythrocytic Serjeant GR (1991) Sickle cell‐B+ thalassaemia in
hemoglobin C crystals in sickle cell‐hemoglobin C Orissa state, India. Its interactions with the sickle
disease. Blood, 25, 218–223. cell gene. Br J Haematol, 77, 215–220.
241 Bertles JF, Rabinowitz R and Dobler J (1970) 253 Alsultan A, Ngo D, Bae H, Sebastiani P, Baldwin
Hemoglobin interaction: modification of solid CT, Melista E et al. (2013) Genetic studies of fetal
phase composition in the sickling phenomenon. hemoglobin in the Arab‐Indian haplotype sickle
Science, 169, 375–377. cell‐β0 thalassemia. Am J Hematol, 88, 531–532.
242 Lin MJ, Nagel RL and Hirsch RE (1989) 254 Cancio MI, Aygun B, Chui DHK, Rothman JA,
Acceleration of hemoglobin C crystallization by Scott JP, Estepp JH and Hankins JS (2017) The clin-
hemoglobin S. Blood, 74, 1823–1825. ical severity of hemoglobin S/Black (A γδβ)0‐
243 Lawrence C, Fabry ME and Nagel RL (1991) The thalassemia. Pediatr Blood Cancer, 64, e26596.
unique red cell heterogeneity of SC disease. Blood, 255 Serjeant BE, Mason KP and Serjeant GR (1978) The
78, 2104–2112. development of haemoglobin A2 in normal Negro
244 Rucknagel DL and Rising JA (1975) A heterozy- infants and in sickle cell disease. Br J Haematol, 39,
gote for HbβS, HbβC and HbαG Philadelphia in a family 259–263.
presenting with evidence of heterogeneity of 256 Tadmouri GO, Yüksel L and Basak AN (1998)
hemoglobin alpha chain loci. Am J Med, 59, 53–60. HbS/βdel‐thalassemia associated with high levels
245 Lawrence C, Hirsch RE, Fatalieo NA, Patel S, of hemoglobins A2 and F in a Turkish family. Am J
Fabry ME and Nagel RL (1997) Molecular interac- Hematol, 59, 83–86.
tions between α‐G Philadelphia, Hb C, and Hb S: 257 McCurdy PR and Gieschen MM (1960) Clinical
phenotypic implications for SC α‐G Philadelphia and physiologic studies in a Negro with sickle‐cell
disease. Blood, 90, 2819–2825. hemoglobin D disease. N Engl J Med, 262,
246 Serjeant GR, Ashcroft MT, Serjeant BE and Milner 961–964.
PF (1973) The clinical features of sickle cell/β tha- 258 Charache S and Conley CL (1964) Rate of sickling
lassaemia in Jamaica. Br J Haematol, 24, 19–30. of red cells during deoxygenation of blood from
247 Serjeant GR, Sommereaux A‐M, Stevenson M, persons with various sickling disorders. Blood, 24,
Mason K and Serjeant B (1979) Comparison of 25–48.
sickle cell‐β0 thalassaemia with homozygous 259 Schneider RD, Ueda S, Alperin JB, Levin WC,
sickle cell disease. Br J Haematol, 41, 83–93. Jones RT and Brimhall B (1968) Hemoglobin D Los
248 Day ME, Rodeghier M, Driggers J, Bean CJ, Angeles in two Caucasian families: hemoglobin
Volanakis EJ and DeBaun MR (2019) A significant SD disease and hemoglobin D thalassemia. Blood,
proportion of children of African descent with 32, 250–259.
HbSβ0 thalassaemia are inaccurately diagnosed 260 Özsoylu S (1969) Haemoglobin S‐D disease in a
based on phenotypic analyses alone. Br J Haematol, Turkish family. Scand J Haematol, 6, 10–14.
185, 153–156. 261 Huisman THJ (1997) Combinations of β chain
249 Gonzales‐Redondo JM, Kutlar F, Kutlar A, abnormal hemoglobins with each other and with β‐
Stoming TA, de Pablos JM, Kilione Y and Huisman thalassemia determinants with known mutations:
THJ (1988) Hb S(C)‐β+‐thalassaemia: different influence on phenotype. Clin Chem, 43, 1850–1856.
mutations are associated with different levels of 262 Torres LS, Okumura JV, Belini‐Júnior É, Oliveira
normal HbA2. Br J Haematol, 70, 85–89. RG, Nascimento PP, Silva DG et al. (2016)
Phenotypic diversity of sickle cell disease in Haemoglobion E variants: a clinical, haemato-

patients with a double heterozygosity for Hb S logical and biosynthetic study of 4 South African
and Hb D‐Punjab. Hemoglobin, 40, 356–358. families. Acta Haematol, 72, 135–137.
263 Adekile A, Mullah‐Ali A and Akar NA (2010) 274 Gupta R, Jarvis M and Yardumian A (2000)
Does elevated haemoglobin F modulate the phe- Compound heterozygosity for haemoglobin S and
notype in Hb SD‐Los Angeles? Acta Haematol, 123, haemoglobin E. Br J Haematol, 108, 463.
135–139. 275 Masiello D, Heeney MM, Adewoye AH, Eung SH,
264 Milner PF, Miller C, Grey R, Seakins M, Dejong Luo HY, Steinberg MH and Chui DHK (2007)
WW and Went LM (1970) Hemoglobin O Arab in Hemoglobin SE disease‐A concise review. Am J
four negro families and its interaction with hemo- Hematol, 82, 643–649.
globin S and hemoglobin C. N Engl J Med, 283, 276 Knox‐Macaulay HHM, Ahmed MM, Gravell D,
1417–1425. Al‐Kindi S and Ganesh A (2007) Sickle cell‐haemo-
265 Charache S, Zinkham WH, Dickerman JD. globin E (HbSE) compound heterozygosity: a clin-
Birmhall B and Dover GJ (1977) Hemoglobin SC, ical and haematological study. Int J Lab Hematol,
SS/G Philadelphia and SO Arab diseases. 29, 292–301.
Diagnostic importance of an integrative analysis 277 Edison ES, Shaji RV, Chandy M and Srivastava
of clinical, hematologic and electrophoretic find- A (2011) Interaction of hemoglobin E with
ings. Am J Med, 62, 439–446. other abnormal hemoglobins. Acta Haematol, 126,
266 Zimmerman SA, O’Branski EE, Rosse WF and 246–248.
Ware RE (1999) Hemoglobin S/O(Arab): thirteen 278 Eichhorn RF, Buurke EJ, Blok P, Berends MJ and
new cases and review of the literature. Am J Jansen CL (1999) Sickle cell‐like crisis and bone
Hematol, 60, 279–284. marrow necrosis associated with parvovirus B19
267 O’Keeffe EK, Rhodes MM and Woodworth A infection and heterozygosity for haemoglobins S
(2009) A patient with a previous diagnosis of and E. J Intern Med, 245, 103–106.
hemoglobin S/C disease with an unusually severe 279 Gürkan E (2006) Vaso‐occlusive manifestations in
disease course. Clin Chem, 55, 1228–1231. a patient with sickle cell‐hemoglobin E (HbSE)
268 Silvestroni E, Bianco I and Baglioni C (1965) disease. Am J Hematol, 81, 149–156.
Interaction of hemoglobin Lepore with sickle cell 280 Rayburg M, Kalinyak KA, Towbin AJ, Baker PB
trait and microcythemia (thalassemia) in a south- and Joiner CH (2010) Fatal bone marrow embo-
ern Italian family. Blood, 25, 457–469. lism in a child with hemoglobin SE disease. Am J
269 Fairbanks VF, McCormick DJ, Kubik KS, Rezuke Hematol, 85, 182–184.
WN, Black D, Ochaney MS and Schwartz D (1997) 281 Zertal‐Zidani S, Ducrocq R, Weil‐Oliver C, Elion J
Hb S/Hb Lepore with mild sickling symptoms: a and Krishnamoorthy R (2001) A novel δβ
hemoglobin variant with mostly δ‐chain sequences fusion gene expresses hemoglobin A (HbA) not
ameliorates sickle‐cell disease. Am J Hematol, 54, Hb Lepore: Senegalese δ0β+ thalassaemia. Blood,
164–165. 98, 1261–1263.
270 Ngo DA, Aygun B, Akinsheye I, Steinberg MH, 282 Al Jahdhamy R, Makki H, Farrell G and Al Azzawi
Hankins JS and Chui DHK (2010) Characterization AS (2002) A case of compound heterozygosity for
of hemoglobin F decline in compound heterozy- Hb S and Hb S Oman. Br J Haematol, 116, 504.
gotes for HbS and deletional hereditary persis- 283 Luo H‐Y, Heeney M, Wang WC, Eung SH, Ware
tence of fetal hemoglobin. Blood, 116, 683. RE, Steinberg MH and Chui DH (2006)
271 Aksoy M (1960) The hemoglobin E syndromes. Hemoglobinopathies mimicking Hb S/β‐thalas-
II. Sickle‐cell‐hemoglobin E disease. Blood, 15, semia: Hb S/S with α‐thalassemia and Hb S/
610–613. Volga. Am J Hematol, 81, 361–365.
272 Schroeder WA, Powars D, Reynolds RD and 284 Rhoda MD, Arous N, Garel MC, Mazarin M,
Fisher JI (1976) Hb‐E in combination with Hb‐S Monplaisir N, Braconnier F et al. (1986) Interaction
and Hb‐C in a black family. Hemoglobin, 1, of hemoglobin Siriraj with hemoglobin S: a mild
287–289. sickle cell syndrome. Hemoglobin, 10, 21–31.
273 Bird AR, Wood K, Leisegang F, Mathew CG, 285 Kayişli GÖ, Keser I, Özeş ON, Canatan D, Yeşílípek
Ellis P, Hartley PS and Karabus CD (1984) A and Lüleci G (2005) Compound heterozygosity
260 Chapter 4
for two beta chain variants: the mildly unstable 289 Sweeting I, Serjeant BE, Serjeant GR, Kulozik AE
Hb Tyne (codon 5 ProSer) and HbS (codon 6 and Vetter B (1998) Hb S‐Hb Monroe; a sickle cell‐
GluVal). Turk J Haematol, 22, 37–40. beta‐thalassemia syndrome. Hemoglobin, 22,
286 Kendall AG, Ojwang PJ, Schroeder WA and 153–156.
Huisman TH (1973) Hemoglobin Kenya, the prod- 290 Al Balushi HWM, Wali Y, Al Awadi M, Al‐Subhi T,
uct of a gamma‐beta fusion gene: studies of the Rees DC, Brewin JN et al. (2017) The super sickling
family. Am J Hum Genet, 25, 548–563. haemoglobin HbS‐Oman: a study of red cell sick-
287 Wilcox I, Boettger K, Greene L, Malek A, ling, K+ permeability and associations with dis-
Davis L, Steinberg MH et al. (2009) ease severity in patients heterozygous for HbA
Hemoglobin Kenya composed of α‐ and (Aγβ)‐ and HbS‐Oman (HbA/S‐Oman genotype). Br J
fusion‐globin chains, associated with hereditary Haematol, 179, 256–265.
persistence of fetal haemoglobin. Am J Hematol, 291 Swensen JJ, Agarwal AM, Esquilin JM, Swierczek
84, 55–58. S, Perumbeti A, Hussey D et al. (2010) Sickle cell
288 McFarlane A, Warkentin TE and Cartín W (2011) A disease resulting from uniparental disomy in a
novel sickling hemoglobinopathy. N Engl J Med, child who inherited sickle cell trait. Blood, 116,
365, 1548–1549. 2822–2825.
4.1 (a) T 4.3 (a) T 4.5 (a) T 4.7 (a) F 4.9 (a) F 4.11 (a) T
(b) F (b) F (b) T (b) T (b) T (b) T
(c) T (c) F (c) F (c) F (c) T (a) F
(d) T (d) F (d) T (d) T (d) T (a) T
(e) F (e) T (e) F (e) F (e) F (a) T
4.2 (a) F 4.4 (a) F 4.6 (a) F 4.8 (a) T 4.10 (a) T 4.12 (a) T
(b) T (b) T (b) F (b) T (b) F (b) T
(c) T (c) F (c) T (c) T (c) T (c) T
(d) T (d) F (d) T (d) T (d) T (d) T
(e) T (e) T (e) F (e) T (e) T (e) T
5 Other significant haemoglobinopathies
Variant haemoglobins may be clinically significant acetate electrophoresis at alkaline pH. There are
but many are clinically silent. The recognition of also other, less common, variant haemoglobins that
those that are clinically silent can nevertheless be are diagnostically important because they have
important in the diagnostic laboratory since their similar retention times to S, C, E, D‐Punjab or glyco
presence can lead to diagnostic confusion. Problems sylated haemoglobin A on HPLC. If cellulose ace
of two types can arise. Clinically irrelevant variant tate electrophoresis is the primary method, it is
haemoglobins can be confused with clinically sig important to distinguish haemoglobins such as hae
nificant variants because of a similar electrophoretic moglobin G‐Philadelphia or haemoglobin D‐Iran,
mobility or high performance liquid chromatogra which are not clinically important, from haemoglo
phy (HPLC) retention time. In addition, there can bin D‐Punjab/Los Angeles, which is of importance
be coinheritance of two variants, leading to the because of its interaction with haemoglobin S. If
presence of multiple bands or peaks that can be dif HPLC is the primary method the uncommon vari
ficult to interpret. This chapter will therefore deal ant haemoglobins that have a similar retention time
both with clinically relevant haemoglobinopathies to clinically important variants must similarly be
and with other variant haemoglobins that can cause distinguished from each other (e.g. haemoglobin
diagnostic confusion. E from haemoglobin Lepore and haemoglobin
More than 1150 variant haemoglobins have been D‐Punjab from haemoglobin G‐Philadelphia) (see
described. The majority of recognised variant hae Table 2.3).
moglobins are α or β chain variants. A variant α A β chain variant would be expected to comprise
chain leads to variant forms of haemoglobins A, A2 about 50% of total haemoglobin. However if the
and F. A variant β chain leads to a variant of haemo abnormal β chain is synthesised at a reduced rate or
globin A. δ and γ chain variants also occur. There is if there is preferential combination of α chains with
no reason to doubt the existence of variant ε and ζ the normal rather than the variant β chain the pro
chain variants. Functional abnormalities of haemo portion will be less. Among the variant β chains
globin that can result from mutations in globin synthesised at a considerably reduced rate is βE,
genes are shown in Table 5.1. with the result that the proportion of haemoglobin
It should be noted that, although low oxygen E in heterozygotes does not usually exceed 25–30%.
affinity haemoglobins lead to anaemia and some Variant chains with a reduced affinity for α chain, in
times cyanosis, there is no clinically significant comparison with βA, include βS and βC, probably
abnormality as there is normal oxygen delivery to because they are more electropositive than normal β
tissues. The anaemia is the result of a reduced eryth chain [1, 2]; as a result of the reduced affinity, the
ropoietic drive. percentage of the variant is somewhat less than
The variant haemoglobins that are of diagnostic 50%. The converse is seen with variant β chains,
but not clinical significance are mainly haemoglob including βJ‐Baltimore and βJ‐Iran that are more
ins with the mobility of either S or C/E on cellulose electronegative [1, 2] and have a greater affinity

261
262 Chapter 5
Table 5.1 The types of functional abnormality that can occur as a result of mutations in globin genes.
Functional abnormality Example
Polymerisation leading to sickle cell formation Haemoglobin S

Haemoglobin C‐Harlem
Interaction with haemoglobin S, permitting sickling in compound Haemoglobin D‐Punjab/Los Angeles

heterozygotes Haemoglobin C
Haemoglobin O‐Arab
Reduced solubility leading to crystal formation and haemolytic anaemia Haemoglobin C
Increased oxygen affinity leading to polycythaemia Haemoglobin Chesapeake

Haemoglobin Kempsey
Reduced oxygen affinity leading to anaemia Haemoglobin S

Haemoglobin Kansas
Haemoglobin instability leading to a Heinz body haemolytic anaemia Haemoglobin Köln
Extreme instability leading to a thalassaemic phenotype Haemoglobin Terre Haute
Reduced rate of synthesis leading to a thalassaemic phenotype Haemoglobin Lepore

Haemoglobin E
Increased tendency to oxidation leading to methaemoglobin formation Haemoglobin M‐Saskatoon

and cyanosis Haemoglobin M‐Hyde Park
than normal β chain for α chains; the percentage of thalassaemia. As for β globin variants, charge may
the variant is therefore greater than 50%. When influence the affinity of the variant α chain for the
there is coexisting α thalassaemia, a positively normal β chain. For example αM‐Iwate, which is more
charged variant chain such as βS or βC competes less electropositive than normal α chain, combines
well than normal β chain for the reduced number of preferentially with electronegative β chains so that
α chains so that the percentage of the variant is haemoglobin M‐Iwate comprises 22–27% of total
lower than in individuals with a full complement of haemoglobin, even though it is an α1 variant [2].
α genes. The converse is seen with variants such as Haemoglobin G‐Philadelphia (see p. 295) illustrates
J‐Baltimore when the negatively charged variant the complexity of the interaction between an α
chain is more able to compete for the reduced pool chain variant and α thalassaemia. The αG‐Philadelphia
of α chains and the variant is present in an even gene can occur either as one of two α genes on
higher percentage [1]. a chromosome (αGα) or as the only α gene on a
Predicting the percentage of an α chain variant is chromosome (−αG) as a result of the mutation having
more complex since not only are there four α genes occurred in a fusion α2α1 gene on a chromosome
but the α2 gene is transcribed at a higher rate than with a 3.7 kb deletion. The former mutation would
the α1 globin gene. The two allelic α2 genes nor be expected to lead to the variant being about
mally contribute between them about 75% of α 12.5% of total haemoglobin and the latter to the
chains. An α chain variant would therefore be variant being somewhat more than a third of total
expected to comprise either about 37.5% or about haemoglobin. A further complicating factor is that
12.5% of total haemoglobin. Proportions may dif the same ethnic group may have both haemoglobin
fer if: (i) the variant α chain is synthesised at a G‐Philadelphia and a high prevalence of unlinked
reduced rate; (ii) the variant α chain shows a α thalassaemia. There may then be α thalassaemia
greater or lesser affinity for β chain than does in trans to the variant α gene, giving the genotype
the normal α chain; (iii) the variant haemoglobin −αG/−α with haemoglobin G‐Philadelphia being
is also unstable; or (iv) there is coexisting α about 45% of total haemoglobin.
Other significant haemoglobinopathies 263
Haemoglobin C incidence of haemoglobin C in North Africa and a

Haemoglobin C is a variant haemoglobin with a low incidence in southern Europe (Spain and
mutation in the β globin gene at the same site as the Sicily). However it should be noted that some early
mutation in the gene encoding βS. It was first recog reports of the presence of haemoglobin C, based
nised by Itano and Neel in 1950 [3]. Its structure is only on electrophoresis at alkaline pH, may have
α2β26Glu→Lys. It can be present in the heterozygous state been a misidentification of haemoglobin O‐Arab as
(haemoglobin C trait), in the homozygous state (hae haemoglobin C. Haemoglobin C appears to have
moglobin C disease) and in a variety of compound had an independent origin in Oman and Thailand
heterozygous states such as sickle cell/haemoglobin [5, 6]. Two cases have been reported from Iraq, the
C disease and haemoglobin C/β thalassaemia. Sickle identification being based on HPLC only [7].
cell/haemoglobin C disease has been discussed in There are also variant haemoglobins in which the
Chapter 4 (see p. 222). Other haemoglobinopathies haemoglobin C mutation is one of two mutations,
with haemoglobin C will be discussed in this chapter. specifically haemoglobin Arlington Park and
Haemoglobin C is thought to have originated in haemoglobin C‐Rothschild.
West Africa, west of the Niger River (Fig. 5.1) (see Haemoglobin C appears to protect from severe
Table 4.1). In northern Ghana the proportion of falciparum malaria [8], with homozygosity giving
individuals with haemoglobin C is as high as 40% very high protection and heterozygosity giving
and in northern Ivory Coast up to 50%. In Burkina moderate protection [9]. However conflicting results
Faso it is 15–40%. It would appear that haemoglo have been reported from Mali where the incidence
bin C arose in the region spanning Burkina Faso of malaria was found to be higher in people with
and the north of the Ivory Coast and Ghana. It is haemoglobin C trait [10].
found in individuals of African descent in the Oxyhaemoglobin C is prone to crystallisation, but
Caribbean (3.5% prevalence), USA (2% prevalence crystals dissolve on deoxygenation so that obstruc
among Afro‐Americans), Canada and the UK. A tion of capillaries by cells containing crystals is not
high incidence has been noted in a Bedouin tribe in likely. In comparison with haemoglobin A, crystal
northern Israel [4]. There is also a significant lisation is inhibited by haemoglobins F and A2 [5].
0.12%
Morocco
1–6%
Algeria
1–13%
Senegal 1–6% Mali Niger

Gambia <2% 1–8%
Guinea Bissau Guinea

<1.5% Burkina
<1.5% Faso 15–40%
(40–50%) a Nigeria
Ghana 1–9%
1–50%
8–40%
Ivory Coast
Liberia
1–3% Togo Benin
7–17% 7–27%
Fig. 5.1 Distribution of haemoglobin
C in north‐west Africa.
264 Chapter 5
Haemoglobin C trait is conflicting data as to whether this results from

coexisting α thalassaemia trait. In one study, indi
Haemoglobin C trait describes the heterozygous
viduals with C trait and with the normal comple
condition when there is one normal β gene and one
ment of four α globin genes have a mean cell volume
βC gene. It is of no clinical significance but is of
(MCV), on average, around the bottom of the nor
significance in counselling prospective parents.

mal range [12], whereas in another study there was
This is largely because of the possibility of sickle
no difference between C trait and normal [13]. The
cell/haemoglobin C disease if one parent has hae
mean cell haemoglobin concentration (MCHC) is,
moglobin C trait and the other has sickle cell trait.
on average, higher than normal, usually around the
top of the normal range. The red cell distribution
Clinical features width (RDW) is increased.
There are usually no clinical features although an
Blood film. The blood film (Fig. 5.2) is sometimes
increased incidence of invasive pneumococcal
normal but usually shows microcytosis, target cells,
disease has been reported [11].
irregularly contracted cells or a combination of
these features.
Laboratory features
Blood count. The haemoglobin concentration (Hb) is Other investigations. Haemoglobin electrophoresis
usually normal but microcytosis is common. There (Fig. 5.3), HPLC (Fig. 5.4) and capillary electrophoresis
(a)
Fig. 5.2 Blood films of four patients with

haemoglobin C trait showing the range of
features that may be observed: (a) normal
film; (b) one irregularly contracted cell and
(b) one hemi‐ghost;
(c)
Fig. 5.2 Continued. (c) irregularly contracted

cells; and (d) target cells and other poikilocytes.
May–Grünwald–Giemsa (MGG) ×100 objective. (d)
(Fig. 5.5) show haemoglobin A to constitute somewhat

more than 50% of haemoglobin and haemoglobin C
slightly less. The proportion of haemoglobin C is
lower in those with coexisting α thalassaemia trait. In
one study the mean percentage of haemoglobin C
(plus A2) was around 44% in those thought likely to
have four α genes, around 37.5% in those thought
likely to have three α genes and around 32% in those
likely to have two α genes [14]. In another study the
mean levels were 37% and 32% in those with four and
three α genes, respectively [13]. In a single patient with
haemoglobin C trait and haemoglobin H disease the
haemoglobin C was 24% [5]. In subjects with five α
Fig. 5.3 Haemoglobin electrophoresis on cellulose genes, the haemoglobin C percentage tends to be
acetate at alkaline pH showing haemoglobin C trait (lane higher than in those with four α genes [15]. On cel
4) and haemoglobin C/β+ thalassaemia compound lulose acetate at alkaline pH, haemoglobin C has
heterozygosity (lane 5). the same mobility as haemoglobins E, A2 and O‐Arab.
266 Chapter 5
Fig. 5.4 High performance liquid chromatography (HPLC) chromatogram (Bio‐Rad Variant II) in haemoglobin C trait.
Peaks from left to right are haemoglobin F (shaded), two small peaks representing post‐translationally modified
haemoglobin A, haemoglobin A0, haemoglobin A2 (shaded), two small peaks representing post‐translationally modified
haemoglobin C and haemoglobin C0.
On citrate agar or agarose gel at acid pH it can be sepa haemoglobin A2 and haemoglobin Lepore.
rated from haemoglobins E and A2 (same mobility as Haemoglobin C can be detected immunologically [16].
A) and from haemoglobin O‐Arab (similar mobility to Red cell density is increased as a result of
haemoglobin S). On HPLC, haemoglobin C can be increased K+/Cl− co‐transport, loss of intracellular
separated from haemoglobin E and haemoglobin O‐ potassium and resultant cellular dehydration [9].
Arab but, depending on the specific instrument/ Osmotic fragility is decreased. Red cell survival is
reagent system, there may be overlap with normal or slightly reduced.
Fig. 5.5 Capillary electrophoresis (Sebia Capillarys 3) in haemoglobin C trait showing haemoglobins A, A2 and C. Note
that haemoglobins A2 and C overlap.
Diagnosis Clinical features

Diagnosis is dependent on identifying haemoglobin Individuals with haemoglobin C disease either
A and haemoglobin C by at least two independent have a normal Hb or are mildly or moderately
techniques, with haemoglobin C being present in a anaemic. Because of the chronic haemolysis, there is
lower amount than haemoglobin A. an increased incidence of gallstones. The spleen
may be enlarged.
Haemoglobin C disease
Laboratory features
Haemoglobin C disease describes the homozygous
state in which there are two βC genes and no normal Blood count. The Hb ranges from about 80 g/l up to
β gene. As a consequence, about 95% of total hae normal. There is often marked microcytosis. In one
moglobin is haemoglobin C with the remainder study, patients with a normal complement of four
being haemoglobins A2 and F. Homozygosity for α genes and no iron deficiency had, on average, an
haemoglobin C leads to a clinically mild, chronic MCV of 55 fl [12]. The MCHC is, on average,
haemolytic anaemia. around the top of the normal range and the
268 Chapter 5
roportion of hyperdense cells is increased. The

p The percentage of dense cells is markedly
cause of the microcytosis and increased MCHC is increased. Osmotic fragility is markedly reduced.
activation of the K+/Cl− co‐transporter, which Bilirubin concentration is normal or increased.
leads to loss of water from the cell; cells are Oxygen affinity is reduced as a result of reduction
therefore smaller, denser and less deformable than of intracellular pH, rather than any alteration of the
normal [5]. The reticulocyte count is mildly ele oxygen affinity of haemoglobin C [5]. Red cell
vated, usually 2–4%. Thrombocytosis is common survival is reduced to about a third of normal [5].
in children [17]. The bone marrow shows erythroid hyperplasia
and characteristic dyserythropoietic features
Blood film. The blood film (Fig. 5.6) characteristi with an irregular nuclear membrane (Fig. 5.10).
cally shows numerous target cells and numerous On ultrastructural examination, there can also be
irregularly contracted cells [18]. There is also micro duplication of the nuclear membrane (Fig. 5.11).
cytosis. There may be occasional nucleated red
blood cells (NRBC). Occasional cells may contain
Diagnosis
haemoglobin C crystals (Figs 5.6c and Fig 5.7).
Crystals may be tetragonal or hexagonal. Usually Diagnosis depends on identifying haemoglobin
all the haemoglobin in a cell has been incorporated C as the sole variant haemoglobin, in the
into the crystal so that the crystal is in a cell that absence of haemoglobin A, by at least two
otherwise appears empty of haemoglobin. The independent techniques. The differential diagnosis
majority of crystals are 6–10 μm in length and includes compound heterozygosity for haemoglobin
2–3 μm in diameter with pointed ends. Crystals can C and β0 thalassaemia and, if cellulose acetate elec
be present in vivo but can also form in vitro, particu trophoresis is the primary method used, compound
larly if the film dries slowly [19]. heterozygosity for haemoglobin C and either hae
Both crystals and NRBC are more often seen in moglobin E or haemoglobin C‐Harlem. The latter
patients who have been splenectomised. two compound heterozygous states are rare.
Other investigations. Haemoglobin electrophoresis

Haemoglobin C/β thalassaemia
(Fig. 5.8) and HPLC (Fig. 5.9) show that haemo
globin C comprises almost all the haemoglobin. Haemoglobin C may be coinherited with either β0
Haemoglobin F can be slightly elevated but does not or β+ thalassaemia. The latter is more common
usually exceed 3%. because β+ thalassaemia is more common than β0

patients with haemoglobin C
disease showing the range of
features that may be observed:
(a) target cells and irregularly
(a) contracted red cells;
(b)
(c)
Fig. 5.6 Continued.

(b) irregularly contracted cells
and only an occasional target
cell; (c) target cells and
irregularly contracted cells; and
(d) irregularly contracted cells
and one haemoglobin C crystal
(adjacent to the neutrophil).
MGG ×100. (d)
270 Chapter 5
in the ethnic groups that are most likely to inherit consequent on aplastic crises has been observed. If
haemoglobin C. This compound heterozygous state is haemoglobin C is coinherited with mild β+ thalas
observed particularly in those with African ancestry saemia the features are similar to those of homozy
but has also been reported in Italians and Turks. gosity for haemoglobin C. There is a mild to
moderate haemolytic anaemia and some spleno
Clinical features megaly. Fatal spontaneous rupture of the spleen
during pregnancy has been reported in a patient
Compound heterozygosity for haemoglobin C and with previously undiagnosed haemoglobin C/β
β thalassaemia leads to a moderately severe anaemia thalassaemia; the spleen showed extramedullary
with splenomegaly. If haemoglobin C is coinherited haemopoiesis [20].
with β0 thalassaemia or severe β+ thalassaemia, the
clinical picture resembles thalassaemia intermedia
with moderately severe anaemia, splenomegaly
and sometimes hypersplenism. Worsening anaemia

Fig. 5.7 Ultrastructural examination of an erythrocyte acetate at alkaline pH in haemoglobin C disease (lane b);
containing a haemoglobin C crystal. (By courtesy of the other lanes show either haemoglobin A alone or
late Professor Sunitha N. Wickramasinghe.) haemoglobin A plus haemoglobin S.

(Bio‐Rad Variant II) in haemoglobin
C homozygosity; peaks, from left to
right, are haemoglobin F (shaded),
unidentified, haemoglobin A2
(shaded), glycosylated haemoglobin
C and haemoglobin C0; the shoulder
on the left of the haemoglobin C0
peak represents post‐translationally
modified haemoglobin C.
Fig. 5.10 Bone marrow aspirate of a

patient with haemoglobin C disease
showing erythroid hyperplasia and
an irregular nuclear margin. MGG ×100.

examination showing the
characteristic abnormality of the
nuclear membrane in homozygosity
for haemoglobin C. (By courtesy of
the late Professor Sunitha N.
Wickramasinghe.)
Laboratory features irregularly contracted cells. There is more anisocy

tosis and poikilocytosis than in haemoglobin C
Blood count. The Hb varies from 70 to 100 g/l in hae
disease, particularly in cases of haemoglobin

moglobin C/β0 thalassaemia. In haemoglobin C/β+
C/β0 thalassaemia. Haemoglobin C crystals are
the Hb can be reduced or, occasionally, n
ormal. The
sometimes present.
MCV is markedly reduced. The reticulocyte count
is moderately elevated.
Other investigations. The major haemoglobin is
aemoglobin C with haemoglobin F usually being
h
Blood film. The blood film (Fig. 5.12) shows 2–10% (most often greater than 5%). Haemoglobin
hypochromia, microcytosis, target cells and A can be totally absent (haemoglobin C/β0
272 Chapter 5
(a)

patients with haemoglobin C/β
thalassaemia compound
heterozygosity: (a) haemoglobin
C/β0 thalassaemia showing
hypochromia, irregularly
contracted cells, three nucleated
red blood cells and a cell
containing a haemoglobin C
crystal; (b) haemoglobin C/β+
thalassaemia showing irregularly
contracted cells and target cells;
the electrophoretic pattern of this
patient is shown in Fig. 5.3 (lane
(b) 5). MGG ×100.
thalassaemia) or, when haemoglobin C is coinher problematical since it is not infrequent for
ited with a mild β+ thalassaemia, up to 20–30% of patients with h aemoglobin C homozygosity to
total haemoglobin (Figs 5.13 and 5.14). have microcytosis, as a result of cellular dehydra
Osmotic fragility is markedly reduced. tion or coexisting α thalassaemia trait. A distinc
tion can be made by family studies or
deoxyribonucleic acid (DNA) analysis.
Diagnosis
Diagnosis of haemoglobin C/β+ thalassaemia is
Coinheritance of haemoglobin C and other
straightforward, being dependent on identifica
variant haemoglobins or thalassaemias
tion of haemoglobin A and haemoglobin C by two
independent techniques with haemoglobin C Coinheritance of haemoglobin C and either
being more than haemoglobin A. Diagnosis of haemoglobin Lepore or δβ thalassaemia leads

haemoglobin C/β0 thalassaemia can be more to a clinically mild disease resembling
Fig. 5.13 HPLC chromatogram (Bio‐Rad Variant II) in a patient with haemoglobin C/β+ thalassaemia. The peaks from
left to right are: haemoglobin F (shaded), post‐translationally modified haemoglobin A (two peaks), haemoglobin A0,
haemoglobin A2 (shaded), post‐translationally modified haemoglobin C (two peaks) and haemoglobin C0. The red cell
indices were red cell count (RBC) 6.8 × 1012/l, haemoglobin concentration (Hb) 128 g/l, mean cell volume (MCV) 55.4 fl,
mean cell haemoglobin (MCH) 18.7 pg and red cell distribution width (RDW) 19.
coinheritance of haemoglobin C and mild β+ aemoglobin C [5]. In haemoglobin C/δβ thalas

h
thalassaemia. In haemoglobin C/haemoglobin
saemia, haemoglobin C is around 75% and
Lepore compound heterozygosity, haemoglobin haemoglobin F around 25% of total haemoglobin.
C is around 80%, haemoglobin Lepore 10–15% Coinheritance of haemoglobin C and deletional
and haemoglobin F 6–12%. Haemoglobin hereditary persistence of fetal haemoglobin is
Lepore‐Boston inhibits the crystallisation of also clinically mild.
274 Chapter 5
Fig. 5.14 Capillary electrophoresis (Sebia Capillarys 2) in a patient with haemoglobin C/β+ thalassaemia. Note that the
A2 peak overlaps with the haemoglobin C peak. Same patient as Fig. 5.13.
Haemoglobin C crystallisation is accelerated in K‐Woolwich or haemoglobin P‐Galveston does not

haemoglobin C heterozygotes who are also hete differ in severity from haemoglobin C trait [24].
rozygous for haemoglobin Korle‐Bu, leading to a Coinheritance of haemoglobins C and E has
mild haemolytic anaemia with microcytosis and been described [6, 25–28]. It has been noted that
increased numbers of hyperdense cells [21, 22]. the haemoglobin E percentage tends to be higher
Crystals are cubic [21]. Compound heterozygosity than in haemoglobin E trait [6, 26]. For example,
for C and N‐Baltimore also leads to accelerated three children had haemoglobin E plus A2 of
crystallisation of haemoglobin C in comparison 33–37% (cf. 25–30% in haemoglobin E simple
with haemoglobin C trait [22]. The phenotype may heterozygotes), haemoglobin C of 54–56% and
be intermediate in severity between that of haemoglobin F of 2.1–5.8% [6]. In one patient,
heterozygosity and homozygosity for haemoglobin who was clinically well, the haematocrit and
C. If haemoglobin Riyadh is coinherited with reticulocyte count were normal [25]; the red cells
haemoglobin C, crystallisation is retarded and were normocytic and normochromic but showed
microcytosis is usually the only feature [22, 23]. cytoplasmic folding and stomatocyte formation.
Haemoglobin C coinherited with haemoglobin In a second patient there was a mild anaemia
(Hb 99 g/l) with red cell indices suggestive of tha to an atypical form of haemoglobin H disease.
lassaemia trait; haemoglobin C was 60% and hae One reported case had a chronic haemolytic
moglobin E 39% [26]. An adult patient had a anaemia and significant splenomegaly [29]. There
normal Hb, red cell indices suggesting thalassaemia was marked microcytosis. Only haemoglobins
trait, a blood film showing target cells and A and C (20%) were detected on haemoglobin
irregularly contracted cells and HPLC showing electrophoresis and only occasional inclusion‐
38% E plus A2 and 60% haemoglobin C [27]. Two of containing cells were found on a haemoglobin H
the three children mentioned above were anaemic preparation.
(Hb 75 and 95 g/l) and red cells were markedly As is the case with the sickle cell mutation, the hae
microcytic (MCV 52 and 61 fl) in the absence of moglobin C mutation can occur as one of two muta
coexisting α thalassaemia trait (although iron tions on a chromosome. Haemoglobin Arlington Park
deficiency was not excluded) [6]. Figure 5.15 is α2β26Glu→Lys,95Lys→Glu. There is no net change in charge
shows the blood film and other investigations of relative to haemoglobin A so that this variant haemo
an 8‐month‐old baby with haemoglobin C/ globin is electrophoretically silent. Nevertheless it
haemoglobin E compound heterozygosity [28]. interacts in the same manner as haemoglobin C with
Coinheritance of haemoglobin C and the α chain haemoglobin S to give a clinically significant sickling
variant, haemoglobin G‐Philadelphia, does not disorder. Haemoglobin C‐Rothschild has been
differ clinically or haematologically from hae described in a heterozygote; it moves cathodal to hae
moglobin C trait although haemoglobin C crys moglobin C at alkaline pH and between haemoglob
tallisation is accelerated [5]. Electrophoresis on ins S and C on citrate agar at acid pH.
cellulose acetate at alkaline pH shows four bands:
haemoglobin A, haemoglobin G, haemoglobin C
Haemoglobin E
and a slow‐moving haemoglobin G–C hybrid
haemoglobin (Fig. 5.16). On agarose gel at acid pH Haemoglobin E is a β chain variant, α2β26Glu→Lys,
there are only two bands, A plus G‐Philadelphia which is common in South‐East Asia (Table 5.2)
and C plus C‐G hybrid (Fig. 5.17). [30–39]. It is the second most common variant
Coexistence of haemoglobin C heterozygosity haemoglobin in the world, after haemoglobin S.
and the genotype of haemoglobin H disease leads It was first described by Chernoff and colleagues
Fig. 5.15 An 8‐month‐old baby

with compound heterozygosity for
haemoglobins C and E: (a) Blood
film showing target cells and
irregularly contracted cells (MGG
×100); the red cell indices were
RBC 4.45 × 109/l, Hb 92 g/l, MCV
57.8 fl, MCH 20.7 pg and MCHC
358 g/l. (Continued on pp. 276–277) (a)
276 Chapter 5
(b)
Fig. 5.15 Continued. (b) HPLC chromatogram (Bio‐Rad Variant II) showing, from left to right, haemoglobin F, two
small peaks of post‐translationally modified haemoglobin E, haemoglobins E0 plus A2 (32.9%), two small peaks of
modified haemoglobin C and haemoglobin C0 (59.7%).
in 1954 [40] and independently in the same year by Assam), Bangladesh, Pakistan, Nepal, Vietnam,
Itano and colleagues [41]. The highest prevalence is Malaysia, the Philippines, Indonesia and Turkey.
in some parts of Thailand, in Cambodia and in Laos. Although haemoglobin E is prevalent in Sri Lanka, it
Thailand and Myanmar (Burma) have an overall is not prevalent in southern India; it is thought to
prevalence of around 14–15%. Gene frequency in have reached Sri Lanka during migration from
Thailand varies from 8 to 50–70%, being highest in north‐eastern India during the fifth century bc.
north‐eastern Thailand. Haemoglobin E is also Occasional cases have been observed in individuals
found in Sri Lanka, north‐eastern India (Bengal and of apparent Northern European Caucasian descent

acetate at alkaline pH in a patient with heterozygosity for
both haemoglobin C and the α‐chain variant
haemoglobin G‐Philadelphia; four bands are apparent
representing, from left to right, haemoglobin A,
haemoglobin G, haemoglobin C and hybrid C/G.
(c)
The βE chain is synthesised at a reduced rate
in comparison with βA. This is because the muta
tion both leads to slower excision of intervening
sequence 1 and also creates a false splicing site
towards the 3′ end of exon 1 so that there is a pro
portion of abnormally spliced messenger ribonu
cleic acid (mRNA). Post‐transcriptional processing
of the latter is abnormal. The result of the reduced
rate of synthesis of βE chain and therefore of haemo
globin E is that heterozygotes, compound heterozy
gotes and homozygotes show some thalassaemic
features. Haemoglobin E can therefore be regarded
(d)
as a thalassaemic haemoglobinopathy. The α:non‐α
Fig. 5.15 Continued. (c) Cellulose acetate electrophoresis chain synthesis ratio is 1.2–2.1 in heterozygotes [42].
at alkaline pH, showing the patient in lane 6; (d) acid Haemoglobin E also has weakened α1β1 contacts,
agarose electrophoresis showing the patient in lane 6 leading to instability in conditions of increased
(haemoglobin E has the same mobility as haemoglobin A oxidant stress. The most significant clinical conse
at this pH). quences occur if haemoglobin E is coinherited
with β thalassaemia trait, this often leading to tha
lassaemia major or intermedia. Homozygosity for
and a single affected family has been observed in haemoglobin E produces a clinically mild condition
the former Czechoslovakia. There are also variant and is thus of much less significance.
haemoglobins in which the haemoglobin E muta Haemoglobin E heterozygosity may protect
tion is one of two mutations, specifically haemoglo against malaria, although this is uncertain [39, 43],
bin Corbeil and haemoglobin T‐Cambodia. and may protect against severe falciparum malaria
278 Chapter 5
Fig. 5.17 Haemoglobin electrophoresis on agarose gel at acid pH in a patient with heterozygosity for both h
aemoglobin C and
the α‐chain variant haemoglobin G‐Philadelphia; at this pH, since haemoglobin G moves with haemoglobin A and the hybrid
haemoglobin moves with haemoglobin C only two bands are apparent (apart from a faint haemoglobin F band); FASC
indicates a control sample containing haemoglobins F, A, S and C.
[44]. Haemoglobin E homozygosity is protective red cell indices not infrequently resemble those of
against malaria [43]. thalassaemia trait [45]. It is not uncommon for
individuals with haemoglobin E trait to also have a
deletion of one or two α genes. However even those
Haemoglobin E trait
with a full complement of α genes can have
Haemoglobin E trait is an asymptomatic condition microcytosis and mild anaemia. In a report of
with no clinical significance except for the possibil 34 such cases the average Hb was 124 g/l, the
ity of homozygous or compound heterozygous MCV 79.7 fl and the MCH 26.2 pg [46].
states in the children of heterozygotes.
Blood film. The blood film (Fig. 5.18) may be normal
Clinical features or may show hypochromia, microcytosis, target
cells, irregularly contracted cells, basophilic stip
There are usually no clinical features although
pling or any combination of these features.
increased susceptibility to oxidant‐induced hae
molysis has been suspected.
Other investigations. Haemoglobin electrophoresis at
alkaline pH (Fig. 5.19) shows the variant haemoglo
Laboratory features
bin to have the same mobility as haemoglobins C
Blood count. Some patients have a normal blood and A2. On citrate agar or agarose gel at acid pH the
count. Others have an increased red cell count mobility of haemoglobin E is the same as that of hae
(RBC) and reduced mean cell haemoglobin (MCH) moglobin A and A2. Haemoglobin E has a character
and MCV, with or without mild anaemia. The Hb istic mobility on isoelectric focusing, being well
does not usually fall much below 120 g/l. The separated from haemoglobin A and moving close to
MCHC is normal or, occasionally, increased. The haemoglobins C and A2. On HPLC it is easily
Table 5.2 Prevalence of haemoglobin E carriers in various countries. (From references

30–39 and other sources.)
Country Percentage
India 0–3.5*
Pakistan 0.5–1
Bangladesh 4
Bhutan 1.5–6.5
Nepal <1
Sri Lanka 0–13† (overall 0.5%) and 0–3.3% in different
districts reported
Myanmar (Burma) 1–33
Thailand 8–40‡
Laos 20–40‡
Cambodia 15–30‡
Vietnam 2–4§
Southern China 1–2.5
Malaysia 1–40¶
Indonesia 1–13
Philippines ~1
Turkey 0.5–1
Iran 0.02
Jamaica 0.007
* But 22% in Kolkata (Calcutta), 60% in some parts of west Bengal and 50–80% in Assam.
† Common in the Veddah; equally common in Tamils and Sinhalese [39].
‡ Haemoglobin E occurs in 70% of the So people of north‐east Thailand, in 50% of the Khmer people on
the borders of Laos, Thailand and Cambodia and in 50% of the Kachari people in Assam [35].
§ Much higher incidence among Khmer population in Vietnam (20–30%) and in certain other ethnic
groups including the Ede and the Vân Kiêv (5–50%); ranges from 0 to 36% in eight ethnic groups in
southern Vietnam [38].
¶ More frequent in the aboriginal population than in the Malays [32].
separated from haemoglobins A and C but with Above 39% haemoglobin E suggests that the diagno
some instruments co‐elutes with haemoglobin A2 sis is E/β thalassaemia not E trait [42]. A trimodal
(Fig. 5.20). Haemoglobin E can be separated from distribution of the variant haemoglobin is found,
haemoglobin A2 on capillary electrophoresis, per with modal percentages of 29.1, 27.3 and 17.4%,
mitting the latter to be measured (Fig. 5.21); the hae correlating with the presence of four, three or two α
moglobin A2 percentage is then found to be generally globin genes [46]. Individuals with less than 25%
increased in patients with haemoglobin E trait [47– haemoglobin E almost always have coexisting α
50]; in one study there was a mean value of 3.4% in thalassaemia trait [46]. When an individual is het
comparison with a mean normal of 2.6% [49] and in erozygous for haemoglobin E and also has the gen
a second study the mean value was 4.04%, with six otype of haemoglobin H disease the percentage of
of seven results being elevated [50]. Haemoglobin E haemoglobin E is even lower, sometimes as low as
can also be detected by immunoassay [16]. 10%. On the basis of extensive experience in
In haemoglobin E heterozygotes, the variant Thailand, Fucharoen has further reported that het
usually comprises 30% or less of total haemoglobin. erozygotes have 25–30% haemoglobin E when
280 Chapter 5
(a)
(b)
Fig. 5.18 Blood films of two patients with haemoglobin E trait showing: (a) microcytosis, minimal hypochromia and one
target cell and (b) microcytosis (the MCV was 72 fl) and target cells. MGG ×100.
three or four α genes are present, 19–21% when two bin E reported in various haemoglobin E syn
α genes are present and 13–15% when the genotype dromes are likely to represent haemoglobin E plus
is that of haemoglobin H disease [42]. A single haemoglobin A2.
patient has been reported with a β++ thalassaemia Haemoglobin F can be increased, being more
mutation in cis to the haemoglobin E mutation, than 1% (up to 5.7%) in 20 of 74 Thai patients [52];
leading to a haemoglobin E percentage of only this was mainly attributable to coinheritance of a
3.1% [51]. The percentage of haemoglobin E can polymorphism in the BCL11A gene and to a lesser
also be lowered by coexisting iron deficiency. It extent to the XmnI polymorphism of the Gγ gene
should be noted that the percentages of haemoglo (HBG2).
Haemoglobin E is slightly unstable on heat and to levels comparable with those seen in heterozy
isopropanol stability tests. Red cell protoporphy gotes for inherited deficiency [54].
rin, often used as a screening test for iron defi
ciency, can be elevated in haemoglobin E
Diagnosis
heterozygosity [53]. Osmotic fragility of red cells
is reduced. Pyrimidine 5′ nucleotidase is reduced Diagnosis is dependent on demonstrating the pres
ence of haemoglobin E and haemoglobin A by two
independent techniques, with haemoglobin E
being about a third of total haemoglobin. If the
primary diagnostic method is cellulose acetate

electrophoresis at alkaline pH, the differential
diagnosis is mainly with haemoglobin C trait. If
HPLC is the primary diagnostic method, the dif
ferential diagnosis is with heterozygosity for hae
moglobin Lepore or other uncommon haemoglobins
such as haemoglobin D‐Iran (Fig. 5.22) and haemo
globin G‐Coushatta (Fig. 5.23). The percentage of
the variant and the precise retention time are use
ful. Although the latter two variant haemoglobins
elute in the E window the shape of the chromato
gram differs from that of haemoglobin E and the
percentage is much higher.
Screening tests for haemoglobin E are applica
ble in countries with a high prevalence of this
variant haemoglobin. A one‐tube osmotic fragility
and a dichlorophenolindophenol (DCIP) test [55]
acetate at alkaline pH in a patient with haemoglobin E
have been used but the CMU‐E test devised at the
trait (lane d) note that the variant haemoglobin plus Chiang Mai University in Thailand has been
haemoglobin A2 comprises only about 30% of total found to have better sensitivity (100%) and speci
haemoglobin. ficity (99%) [56].

haemoglobin E heterozygosity;
haemoglobin E appears in the A2
window with a retention time of
3.66 minutes, and together with
haemoglobin A2 comprises 30.2% of
total haemoglobin; peaks, from left to
right, are haemoglobin F (shaded),
haemoglobins A0 and haemoglobin E
plus A2 (shaded).
282 Chapter 5
Fig. 5.21 Capillary

electrophoresis (Sebia
Capillarys 2) in a patient
with haemoglobin E trait.
The peaks from left to
right are haemoglobins
A, E and A2. Note
that haemoglobin A2
separates from
haemoglobin E.

haemoglobin D‐Iran heterozygosity;
haemoglobin D‐Iran appears in the A2
window and, together with
haemoglobin A2, comprised 46.6%
with a retention time of 3.53 minutes;
peaks, from left to right, are
haemoglobin F, post‐translationally
modified haemoglobin A (two peaks),
haemoglobin A0 and haemoglobin
D‐Iran plus A2 (shaded).
Haemoglobin E homozygosity Clinical features

(haemoglobin E ‘disease’)
There has been reported to be usually no anaemia
Individuals with the genotype βEβE are often or splenomegaly and rarely any evidence of hae
completely asymptomatic. It has therefore been
molysis [31, 32, 57]. However a study of 84 Indian
suggested that the term ‘haemoglobin E homozygo patients found a mean Hb of 97 g/l, 31% spleno
sity’ is preferable to the term ‘haemoglobin E megaly, 12% hepatomegaly, 21% cholelithiasis and
disease’. jaundice as a frequent presenting feature [58].

haemoglobin G‐Coushatta
heterozygosity; haemoglobin
G‐Coushatta appears in the A2
window and, together with
haemoglobin A2, comprised 51.5%
with a retention time of 3.40 minutes;
peaks, from left to right, are
haemoglobin F, post‐translationally
haemoglobins A0 and haemoglobin
G‐Coushatta plus A2 (shaded).
Fig. 5.24 Blood film of a patient

with haemoglobin E homozygosity
showing microcytosis, target cells
and several irregularly contracted
cells. MGG ×100.
Laboratory features Blood film. The blood film (Fig. 5.24) usually shows
hypochromia and microcytosis with variable num
Blood count. The blood count often resembles that
bers of target cells and irregularly contracted cells.
of β thalassaemia trait with a normal Hb or very mild
anaemia, an increased RBC and reduced MCV and Other investigations. Haemoglobin electrophoresis
MCH. The MCHC is usually normal. The reticulo (Fig. 5.25), HPLC (Fig. 5.26) and capillary electro
cyte count is usually normal but may be slightly ele phoresis (Fig. 5.27) show the major haemoglobin to
vated. The MCV and MCH are significantly lower be haemoglobin E with haemoglobin E plus A2 con
when there is coexisting α0 thalassaemia heterozy stituting 85–99% of total haemoglobin, the rest
gosity but with overlapping values (one study show being haemoglobin F. Except when using capillary
ing MCV mean values of 57.1 and 63.7 fl, respectively, electrophoresis, haemoglobin A2 is difficult to quan
and MCH mean values of 18 and 20.3 pg, respec titate in the presence of haemoglobin E but with
tively) [59]; coexisting homozygosity for α+ thalas appropriate techniques can be shown to be some
saemia can also raise the haemoglobin A2 and was what increased; in one study, a mean value of 4.4%
not excluded in those without α0 thalassaemia. was observed in seven patients in comparison with
284 Chapter 5
a mean of 2.6% in control subjects [49]. Haemoglobin c omprising up to 15% of total haemoglobin [42] but
A2 is significantly higher when there is coexisting α0 is usually less than 10%. The percentage of haemo
thalassaemia heterozygosity, with mean values of globin F is decreased by coexisting α thalassaemia
5.1% and 8.8% in one study; all subjects with coex trait and is increased by the XmnI polymorphism in
isting α0 thalassaemia had a haemoglobin A2 of the Gγ gene, by two polymorphisms in HBS1L‐MYB
above 5.3%, whereas this was observed in 44% of and by a polymorphism of BCL11A [60]. Red cell life
subjects without coexisting α0 thalassaemia [59]. span is normal [31]. Osmotic fragility is reduced.
Haemoglobin F may be normal or elevated, The α:non‐α chain synthesis ratio is increased, with
reported ratios of 2.1 and 2.2 [31] and 1.59 and 1.61
[57]. The oxygen dissociation curve is slightly right‐
shifted [31], probably as a result of increased intra
cellular 2,3‐diphosphoglycerate (2,3‐DPG) since
purified haemoglobin E has a normal oxygen affin
ity. Ultrastructural examination shows precipitated
α chains in a proportion of late erythroblasts and
bone marrow reticulocytes [61]. In one study, hyper
bilirubinaemia was observed in 28% of patients and
was found to correlate for homozygosity for a
UGT1A1 polymorphism [58]. Lactate dehydroge
nase (LDH) was often increased, this together with
the frequent observation of cholelithiasis suggest
ing haemolysis was at least a contributory factor to
hyperbilirubinaemia in this population [58].
Diagnosis
The diagnosis requires identification of haemoglo
Fig. 5.25 Haemoglobin electrophoresis on cellulose bin E as the sole variant haemoglobin, in the absence
acetate at alkaline pH in a patient with haemoglobin E of haemoglobin A, by two independent techniques.
homozygosity (lane c); AFSC, control sample containing The differential diagnosis is with haemoglobin
haemoglobins A, F, S and C (bands are faint). E/β0 thalassaemia compound heterozygosity and

Variant II) in a patient with haemoglobin
E homozygosity; haemoglobin E appears
in the A2 window with a retention time
of 3.65 minutes, and together with
haemoglobin A2 comprised 94.1% of
total haemoglobin. Note that
haemoglobin E that has undergone
post‐translational modifications appears
in the ‘P3 window’ and the A0 windows;
this should not be misinterpreted as
indicating that haemoglobin A is present;
peaks, from left to right, are haemoglobin
F, post‐translationally modified
haemoglobin E (two peaks) and
haemoglobin E plus A2 (shaded).
Fig. 5.27 Capillary electrophoresis (Sebia Capillarys 3) in haemoglobin E homozygosity showing haemoglobins F, E and
A2. Note that haemoglobin A2 is elevated and that, in contrast to compound heterozygosity for haemoglobin E and β0
thalassaemia, there is only a trivial increase in haemoglobin F.
requires assessment of clinical features and of the eterozygous state is quite common in Thailand
h
relative proportions of haemoglobins E, F and, if and surrounding countries but also occurs
possible, A2 (see p. 287), family studies and some throughout a large part of South‐East Asia, stretching
times DNA analysis. from Indonesia to Sri Lanka, northeast India and
Bangladesh, making this the commonest β thalas
Problems and pitfalls. Because of the genetic signifi saemia phenotype in the world.
cance, it is essential to distinguish homozygosity Haemoglobin E‐Saskatoon should be distin
for haemoglobin E from haemoglobin E/β0 thalas guished from haemoglobin E; it does not interact
saemia, and to identify coexisting α0 thalassaemia with β thalassaemia [26].
heterozygosity.
Clinical features
Haemoglobin E/β thalassaemia
The severity of compound heterozygosity for
Haemoglobin E may be coinherited with either β0 haemoglobin E and β thalassaemia is very variable
or β+ thalassaemia [13, 62–65]. The compound with the clinical picture ranging from that of
286 Chapter 5
thalassaemia minor through thalassaemia inter only mildly affected. During pregnancy or inter
media to thalassaemia major. Most patients have a current infection, patients who are not usually
disease that is at least moderately severe. The vari transfusion dependent may become sufficiently
ation in severity is not always readily explicable anaemic to require transfusion.
although it is in part related to the presence or In resource‐poor settings where magnetic reso
absence of haemoglobin A and, if haemoglobin A nance imaging (MRI) is not readily available, serum
is present, to its quantity. A high haemoglobin F or ferritin can be useful for monitoring possible iron
coinheritance of α thalassaemia trait or haemoglo overload and reducing the need for MRI. In a Thai
bin Constant Spring can ameliorate the condition study of transfusion‐dependent and transfusion‐
[64, 66], whereas inheritance of triple α increases independent thalassaemia, predominantly E/β tha
the severity [66]. For non‐deletional hereditary lassaemia, it was found possible to diagnose liver
persistence of fetal haemoglobin, homozygosity is iron overload with cut‐off points of >2500 and
required for the condition to be ameliorated [65]. >1700 μg/l, respectively, in the two groups and to
The severity is also increased if a greater propor exclude cardiac iron overload with cut‐off points of
tion of mRNA is alternatively spliced [42]. <2500 and <3000 μg/l, respectively [69].
Methaemoglobin is increased on average to 2.7%, An analysis of 50 British patients with hae
the level correlating with disease severity and moglobin E/β thalassaemia (half of Bangladeshi
previous splenectomy and possibly being related origin, a quarter Indian/Pakistani and a quarter
to ascorbic acid deficiency [67]. originating in South‐East Asia) gives an idea of
The most severely affected individuals are trans the usual degree of severity of this condition [70].
fusion dependent and have hepatosplenomegaly, Half the patients were regularly transfused and
intermittent jaundice, growth retardation, delayed nearly half had required splenectomy. A smaller
sexual maturation and overexpansion of the bone number of US patients had disease of similar
marrow cavity, leading to facial deformity and severity [70], whereas of 22 Canadian patients,
malpositioned teeth; there is an increased inci 30% were regularly transfused and 17% had
dence of gallstones, leg ulcers, osteoporosis and required splenectomy [71]. In Thailand, patients
osteomalacia. Overexpansion of the bone marrow are generally only transfused when this appears
cavity contributes to high output cardiac failure. essential so that the natural history of the
There is a high rate of thromboembolism and untreated disease is more readily apparent; about
hypoxaemia also occurs in splenectomised half of affected individuals have a thalassaemia
patients, possibly as the result of platelet aggre major phenotype and half a thalassaemia interme
gates reaching the lungs [42]. Non‐splenectomised dia phenotype. Of 76 Thai patients, 14% were
patients can develop pulmonary hypertension and considered to have a ‘mild’ phenotype, 55% a

right heart failure but splenectomy aggravates the moderately severe phenotype and 30% a severe
hyperthrombotic state. Iron overload can cause phenotype [42]. However, it should be noted that
diabetes mellitus and liver fibrosis. Less severely classification as ‘mild’ was possible, with transfu
affected individuals may have splenomegaly and sion independence but with an Hb as low as
facial deformity but do not require regular trans 76 g/l; most of these patients would have to be
fusions to maintain life; however these patients regarded as having thalassaemia intermedia. Of
with the phenotype of non‐transfusion‐dependent 78 Indian patients, 37% had mild disease, 19%
thalassaemia can nevertheless develop extramed moderate (some becoming transfusion dependent
ullary haemopoiesis, gallstones, osteoporosis, pul later in life) and 44% severe (transfusion
monary hypertension and, less often, leg ulcers or dependent) [72].
hypogonadism [68]. Extramedullary haemopoiesis Hydroxycarbamide therapy may ameliorate the
has sometimes led to compression of the spinal disease by increasing haemoglobin F synthesis.
cord or brain by tumour‐like masses of haemopoi Patients who respond either show a reduction in
etic tissue, leading to paraplegia or convulsions. transfusion requirement or become transfusion
Hypersplenism can occur. Occasional patients are independent [73].
Laboratory features body‐like α chain inclusions [76], in addition to the

usual post‐splenectomy changes.
Blood count. The Hb is lower than in haemoglobin E
disease. It varies from 25 to 130 g/l, with the average
Other investigations. Haemoglobin electrophoresis
being around 70–80 g/l in comparison with aver
and HPLC show haemoglobins E/A2 and F in the
ages of around 95–115 g/l in patients with homozy
case of haemoglobin E/β0 thalassaemia and A, E/A2
gosity for haemoglobin E. Because chain synthesis
and F in the case of haemoglobin E/β+ thalassaemia.
is more balanced, the Hb is higher in those who
When haemoglobin A is present, it usually consti
coinherit α thalassaemia; for example, in a study of
tutes around 10% of total haemoglobin. Studies in
non‐transfusion‐dependent Thai patients mainly
Thailand have shown that compound heterozy
with E/β0 thalassaemia, the mean Hb was 76 g/l in
gotes for E and β0 thalassaemia usually have
42 patients with four α genes in comparison
haemoglobin E/A2 representing 30–60% of total
with 119 g/l in a patient with homozygous α+
haemoglobin; there is little overlap with homozy
thalassaemia and a mean of 105 g/l in three patients
gous E, where the percentage is usually ≥65% [77].
with heterozygous α0 thalassaemia [74]. The MCV
Conversely, haemoglobin F is usually ≥15% in E/β0
and MCH are, on average, more reduced than in
thalassaemia, whereas in EE it is less than 15% and
haemoglobin E trait. The MCHC is reduced and the
usually less than 5% [77]. However some patients
RDW is increased. The reticulocyte count is slightly
with E/β0 thalassaemia have a higher haemoglobin
increased (e.g. 4–6%).
E (e.g. >75%) and haemoglobin F <15% [77]. In 56
Hydroxycarbamide therapy is associated with a
non‐transfusion‐dependent Sri Lankan patients, the
slight rise in Hb (usually less than 10 g/l rise), the
mean level of haemoglobin F was 28% with a range
expected rise in the MCV and MCH and an
of 6–51% [78]. The percentage of haemoglobin F is
appreciable fall in the reticulocyte count [75].
influenced by coinheritance of α thalassaemia; in a
Blood film. The blood film (Figs 5.28 and 5.29) shows study of Thai patients, there was a mean haemoglo
anisocytosis, poikilocytosis, hypochromia, microcy bin F of 33% in patients with four α genes in com
tosis, target cells, irregularly contracted cells and parison with mean levels of 25.7% in 20 patients
NRBC. There may be basophilic stippling. If sple with −α3.7/αα [74]. Haemoglobin F percentage is
nectomy has been undertaken there will be Heinz also influenced by polymorphisms affecting
Fig. 5.28 Blood film of a 6‐year‐old

boy with haemoglobin E/β+
thalassaemia showing anisocytosis,
target cells and irregularly contracted
cells. There is one cell with the
haemoglobin condensed into one third
of the cell. The red cells indices were
RBC 3.97 × 1012/l, Hb 60 g/l, MCV
54.7 fl and MCH 15.1 pg.
Haemoglobin A was 10%. MGG ×100.
288 Chapter 5

with haemoglobin E/β thalassaemia
compound heterozygosity severe
enough to have required
splenectomy showing poikilocytosis,
some hypochromic cells,
Pappenheimer bodies, a nucleated
red blood cell and erythrocytes
containing precipitates that are likely
to represent precipitated α chain.
MGG ×100.
aemoglobin F synthesis [74]. The increase in hae

h in children under the age of 2 years than in
moglobin F is also in part attributable to erythropoi adults, since haemoglobin F is higher in young
etin‐driven erythroid hyperplasia and possibly children with E homozygosity than it is in
increased F cell production combined with prefer adults [82]. Nevertheless application of this for
ential survival of F cells [79]. It is not surprising mula considerably reduces the number of cases
that, overall, haemoglobin F is very variable, from needing molecular testing. E/δβ 0 thalassaemia
5 to 87% [63]. is characterised by a higher haemoglobin F
When haemoglobins are quantified by capil (mean 55%) and a low haemoglobin A 2 (mean
lary electrophoresis, permitting haemoglobin A 2 2.4%) [81].
to be distinguished from E, the findings are Because of the variability in haemoglobin F, it is
somewhat different. Haemoglobin E is approxi prudent to perform molecular analysis when
mately 82–95% in E homozygotes and approxi haemoglobin E is more than 75% and haemoglobin
mately 78–80% in E/β 0 thalassaemia [80]. F is between 5 and 15%; in high prevalence areas, if
Haemoglobin A 2 was found to be above 6% in capillary electrophoresis is used and a haemoglobin
E/β 0 thalassaemia and was above this level in A2 percentage is therefore available, the need for
only one of 19 E homozygotes [80]. Surprisingly, molecular analysis can be reduced by use of the
in this study haemoglobin F was higher in EE above formula.
(around 2–12%) than in compound heterozy Tests for haemoglobin instability are weakly posi
gotes (around 2–8%), but the difference was not tive. Osmotic fragility is reduced. Red cell life span
significant [80]. The MCV was above 55 fl in in is reduced. There is a marked increase in 2,3‐DPG
18 of 19 EE and in only one of nine E/β 0 thalas with a resultant marked increase in P50 (despite the
saemias, while the Hb was higher in EE (around increased percentage of haemoglobin F), which
85–150 g/l) than in compound heterozygotes means that quite severe anaemia is well tolerated
(around 62–98 g/l). The formula (7.3 × Hb [78]. Erythropoietin concentration is markedly
A 2) + Hb F has been found useful to screen for increased [78]. The bone marrow shows erythroid
E/β 0 thalassaemia, with a score above 60 having hyperplasia, dyserythropoiesis, increased mac
100% sensitivity and 96.5% specificity in adults rophage activity and increased storage iron. Serum
[81]. Sensitivity appears to be equally good in transferrin receptor concentration is increased,
children but with specificity being much lower reflecting expanded erythropoiesis.
Hydroxycarbamide therapy is associated with a Hb being around 70–80 g/l and the MCV averag
significant rise in haemoglobin F percentage and a ing 67 fl [83]. In another series, the Hb was around
reciprocal fall in haemoglobin E percentage [75]. 70–90 g/l (similar to uncomplicated haemoglobin
There is an increase in Hb, MCV and MCH and a H disease) but the MCV was lower at 52–58 fl [84].
fall in serum ferritin and plasma bilirubin [73]. In addition to the blood film features of haemo
Regular blood transfusion is associated with a globin H disease (microcytosis and marked poikil
greater decrease in haemoglobin F synthesis than in ocytosis), there can also be irregularly contracted
haemoglobin E synthesis [79]. cells and erythrocytes with the haemoglobin
retracted to one or both ends of the cell (Fig. 5.30)
Diagnosis [85]. The haemoglobins present are A, E and
Bart’s, with there being more A than E and more E
The differential diagnosis is as for haemoglobin E
than Bart’s [86]. Only a small percentage of cells
homozygosity (see p. 282).
(1–5%) [83], if any [84], contain haemoglobin H
inclusions and haemoglobin H is usually not
Coinheritance of haemoglobin E and other
detected on electrophoresis [84]. Haemoglobin F
variant haemoglobins, thalassaemias
is increased to 3–13%, whereas in uncomplicated
and haematological disorders
haemoglobin H disease it averages 1–2% [84]. It
Coinheritance of haemoglobins S and E has been appears that the reduced quantity of α chain com
described on p. 238 and the interaction with β tha bines preferentially with the normal β chain rather
lassaemia, α thalassaemia trait and haemoglobin C than with βE [83] with the percentage of haemo
in this chapter. globin E plus A2 being as low as 8–16% [84]. The
Because both are relatively common in South‐ designation EABart’s disease has been used.
East Asia, haemoglobin E heterozygosity or Coinheritance of haemoglobin E heterozygosity
homozygosity may coexist with the genotype of and − −/αCSα or − −/αPSα has a more severe clinical
haemoglobin H disease. In one series of patients, phenotype than coinheritance of haemoglobin E
individuals with heterozygosity for haemoglobin heterozygosity and − −/−α [42], with the Hb being
E and the genotype of haemoglobin H disease had on average about 20 g/l lower [87] and with hae
a condition that was clinically similar to uncom moglobin H inclusions being found in around 5%
plicated haemoglobin H disease, with the average of erythrocytes.
Fig. 5.30 Blood film of a Thai patient

with haemoglobin E heterozygosity
and the genotype of haemoglobin H
disease showing microcytosis, marked
poikilocytosis (including target cells
and schistocytes) and an erythrocyte
with haemoglobin retracted to one
end of the cell. The full blood count
showed RBC 5.34 × 1012/l, Hb 88 g/l,
314 g/l. MGG ×100.
290 Chapter 5
Homozygosity for haemoglobin E coinherited severe phenotype with reticulocytosis and a hae
with the genotype of haemoglobin H disease moglobin concentration of less than 30 g/l [70]. The
gives a severe thalassaemia intermedia pheno mechanism is a marked increase in haemoglobin E
type with an Hb similar to that in AEBart’s dis instability.
ease but on average a lower MCV (mean 61 fl).
Haemoglobins present are usually E, Bart’s and
Haemoglobin D‐Punjab/D‐Los Angeles
F, comprising about 80%, 1–25% (average 10%)
and 1–7%, respectively, of total haemoglobin [83, Haemoglobin D‐Punjab is a β chain variant,
88]. No inclusions are detected in haemoglobin H α2β2121Glu→Gln, initially described under the name of
preparations, suggesting that the βE chains do haemoglobin D‐Los Angeles. The latter is the
not form tetramers. The designation EFBart’s correct designation and is used in North America,
disease has been used but it should be noted that whereas in the UK the designation D‐Punjab tends
some patients have been described with no to be used. This variant haemoglobin has also been
increase of haemoglobin F and with haemoglo designated haemoglobin D‐Cyprus, D‐Conley,
bin Bart’s not being detected [88]. Closely related D‐Chicago, D‐North Carolina, D‐Portugal and
clinical phenotypes result from the coinheritance haemoglobin Oak Ridge [91]. Its only major
of either βEβE or βEβ0 and either − −/−α or − −/αCSα. importance is because of its interaction with hae
Although the αCS gene is present in patients with moglobin S (see p. 233). Its highest incidence is
coinheritance of − −/αCSα and haemoglobin E among Sikhs in the Punjab, who show a preva
homozygosity, no haemoglobin Constant Spring lence of 2–3%. Gujeratis have a prevalence of
is detected [87]. about 1% and a similar prevalence is found in
Coinheritance of haemoglobin E heterozygosity British Pakistanis. There is a low but significant
and − −SEA/αα is of genetic significance. It leads to a prevalence among Afro‐Americans (0.4%) [92]
lower haemoglobin E plus A2 percentage (mean and Afro‐Caribbeans (0.01) [34]. It has also been
22.8%, cf. 27.6%), lower MCV (68.6 fl, cf. 75.7 fl) and reported in individuals with American Indian
lower MCH (22.5 pg, cf. 24.9 pg) [89]. Algorithms ancestry. There is a low prevalence among
have been devised to indicate which patients Caucasian populations in England, France and
require testing for α0 thalassaemia in antenatal Portugal who have had a close contact with India.
screening programmes; they differ according In England the highest prevalence is in Norfolk,
to the Hb [89]. where it has been attributed to the sojourn of
Interaction of haemoglobin E and haemoglobin the IXth foot regiment in India. Haemoglobin
Lepore [26] can cause significant anaemia with the D‐Punjab is also found in Pakistan, Afghanistan,
phenotype of thalassaemia intermedia. Iran, China, Turkey, former Yugoslavia, Greece,
Coinheritance of haemoglobin E and haemoglo Holland, Australia and North Africa. The preva
bin D‐Punjab may be associated with mild anaemia lence in Sicily is 0.5% [93] and in Sri Lanka is
and microcytosis; haemoglobin D is around 60–65% 0–1.3% in different districts [39]. There is also a
[90]. The condition is asymptomatic. variant haemoglobin, haemoglobin Cleveland, in
Coinheritance of haemoglobin E and deletional which the haemoglobin D mutation is one of two
hereditary persistence of fetal haemoglobin leads to mutations.
red cell indices similar to those of thalassaemia trait It is important to distinguish haemoglobin
but no clinical abnormality [24]. D‐Punjab from other α and β chain variants with
Coinheritance of haemoglobin E and the elon similar electrophoretic properties but with less or
gated β chain variant, haemoglobin Tak, has been no clinical significance. Other β chain variants
reported in one patient. A mild polycythaemia was with similar electrophoretic mobility to haemo
observed [86]. globin D‐Punjab on cellulose acetate at alkaline
A single individual has been reported in whom pH include haemoglobin D‐Ibadan, α2β287Thr→Lys,
coinheritance of haemoglobin E disease and pyrim haemoglobin D‐Iran (see Fig. 5.15), α2β222Glu→Gln,
idine 5′ nucleotidase deficiency led to a clinically haemoglobin G‐Coushatta (see Fig. 5.16),
α2β222Glu→Ala, and haemoglobin Korle‐Bu, and haemoglobin Korle‐Bu by HPLC (Fig. 5.31). On
α2β273Asp→Asn, none of which appears to be of clinical capillary electrophoresis, haemoglobin D‐Punjab
significance. Haemoglobin Korle‐Bu, initially separates from haemoglobins A and S and appears
described as haemoglobin G and then as haemo in the ‘D zone’ as do haemoglobins G‐Philadelphia
globin G‐Accra, originated in West Africa (Ghana and Lepore (Fig. 5.32). Although G‐Philadelphia
or Ivory Coast) [94]. It moves slightly on the A elutes in the D‐Punjab window the two can be dis
side of S on electrophoresis at acid pH and has the tinguished on HPLC by the presence of the minor
same retention time as haemoglobin A2 on HPLC. G2 fraction in association with haemoglobin
The commonest α chain variant with the potential G‐Philadelphia. A distinction can similarly be
to be confused with haemoglobin D‐Punjab is hae made on capillary electrophoresis.
moglobin G‐Philadelphia (see later). In heterozygotes, haemoglobin D is somewhat less
Haemoglobin D has a normal stability and a nor than 50% of total haemoglobin. Coexisting α thalas
mal or slightly increased oxygen affinity. saemia reduces the percentage of the variant [96].
Osmotic fragility and red cell survival are
normal.
Haemoglobin D‐Punjab trait
Heterozygosity for haemoglobin D is of genetic but
Diagnosis
no other clinical significance.
Diagnosis requires the identification of haemo
globins A and D‐Punjab by two independent
Clinical features techniques. It should be noted that a reliable identi
fication of haemoglobin D‐Punjab by a combination
Haemoglobin D heterozygotes are asymptomatic.
of cellulose acetate electrophoresis at alkaline pH
and acid agarose electrophoresis is not possible.
Laboratory features Isoelectric focusing and HPLC are more useful.
Confirmation by DNA analysis is possible.
Blood count. The blood count is normal [95], unless Haemoglobin A2 is underestimated by HPLC in
there is coexisting α thalassaemia trait. The reticulo the presence of haemoglobin D, whereas it appears
cyte count is normal. to be accurate by capillary zone electrophoresis
[48, 50].
Blood film. The blood film may be normal or show
some target cells. Haemoglobin D‐Punjab disease
Haemoglobin D‐Punjab homozygosity (haemoglo
Other investigations. On cellulose acetate at alkaline bin D disease) is associated with a clinically mild
pH, haemoglobin D‐Punjab has the same electro phenotype.
phoretic mobility as haemoglobin S and a variety
of other α and β chain variants randomly desig
Clinical features
nated haemoglobin D or G. On citrate agar or aga
rose gel at acid pH haemoglobin variants There may be mild haemolysis and sometimes a
designated D or G separate from S and move with mild haemolytic anaemia [95, 97, 98]. Some homozy
or near to haemoglobin A; D/G variants cannot be gotes have splenomegaly.
separated from each other by either of these
electrophoretic techniques. On isoelectric focusing
Laboratory features
D‐Punjab is easily separated from haemoglobins A
and S and from the other relatively common D/G Blood count. Reported haemoglobin concentration
variants, including G‐Philadelphia. It can be sepa has ranged from 90–100 g/l up to normal levels.
rated from S, A, G‐Ibadan, G‐Coushatta, D‐Iran The red cell indices may be suggestive of
292 Chapter 5
Fig. 5.31 HPLC (Bio‐Rad Variant II) in haemoglobin D‐Punjab heterozygosity. The peaks from left to right are: injection
artefact, haemoglobin F, four low peaks representing post‐translationally modified haemoglobins A and D, haemoglobin
A0, haemoglobin A2 (shaded) and haemoglobin D‐Punjab.
thalassaemia with an elevated RBC and reduced

Diagnosis
MCV and MCH. The reticulocyte count may be nor
mal or elevated to 2–4%. The blood film shows The diagnosis is dependent on identification of
infrequent to numerous target cells and may show haemoglobin D‐Punjab as the sole variant haemo
irregularly contracted cells. Osmotic fragility is globin in the absence of haemoglobin A, by two
reduced. Red cell survival is slightly reduced. reliable independent techniques (Figs 5.33 and

Haemoglobin D constitutes the major part of hae 5.34). If there is microcytosis, the differential
moglobin; haemoglobin A2 and F are not increased. diagnosis is with compound heterozygosity for

Fig. 5.32 Capillary electrophoresis (Sebia Capillarys 3) in haemoglobin D‐Punjab heterozygosity. The peaks from left to
right are: haemoglobin D, haemoglobin A and haemoglobin A2.
haemoglobin D‐Punjab and β0 thalassaemia and more difficult by the apparently normal haemo
molecular analysis is required. globin A2 in some individuals [101].
Haemoglobin D‐Punjab/β thalassaemia Clinical features
Compound heterozygosity for haemoglobin Compound heterozygotes more often resemble β

D‐Punjab and β0 thalassaemia produces a mild thalassaemia trait than β thalassaemia intermedia.

thalassaemic condition [24, 99]. Haemoglobin
There is mild anaemia and sometimes splenomegaly.
D‐Punjab/β+ thalassaemia [91, 100] is less common
than haemoglobin D‐Punjab/β0 thalassaemia.
Laboratory features
Curiously one of the reported cases had a more
severe thalassaemic phenotype than some of the Blood count. Reported Hb has usually ranged from
cases of haemoglobin D‐Punjab/β0 thalassaemia 105 g/l up to normal levels but occasional patients
[100]. In antenatal screening it is important to dis have been more severely anaemic. The RBC may be
tinguish this compound heterozygous state from elevated and there is marked reduction of the MCV
haemoglobin D‐Punjab homozygosity; this is made and MCH. The reticulocyte count is slightly elevated.
294 Chapter 5
Fig. 5.33 HPLC (Bio‐Rad Variant II) in haemoglobin D‐Punjab homozygosity. The peaks from left to right are: injection
artefact, haemoglobin F, four low peaks representing post‐translationally modified haemoglobin D, haemoglobin A2
(shaded) and haemoglobin D‐Punjab.
Blood film. The blood film (Fig. 5.35) shows anisocy has been reported to range from high‐normal
tosis, poikilocytosis, hypochromia, numerous target levels (3%) to levels similar to those seen in β
cells and some irregularly contracted cells. thalassaemia trait (4.6–8%) [24, 26]. Reported

normal levels probably result from underestima
Other investigations. Haemoglobin electrophore tion of haemoglobin A2 by HPLC in the presence
sis and HPLC show almost all the haemoglobin of haemoglobin D‐Punjab. This appears not to be
to be haemoglobin D. Haemoglobin F is elevated a problem when microcolumn chromatography
in some patients. The haemoglobin A2 percentage or capillary electrophoresis is used [48, 50, 101].
Fig. 5.34 Capillary electrophoresis (Sebia Capillarys 3) in haemoglobin D‐Punjab homozygosity. The peaks from left to
right are: haemoglobin F, haemoglobin D and haemoglobin A2.
Coinheritance of haemoglobin D‐Punjab cause diagnostic confusion, either when it occurs

and other variant haemoglobins alone or when it is coinherited with other variant
and thalassaemias haemoglobins. This haemoglobin has also been
designated haemoglobin Stanleyville I, D‐St
Coinheritance of haemoglobin D‐Punjab and haemo
Louis, D‐Baltimore, D‐Washington, G‐Bristol and
globin S has been described on p. 233. Coinheritance
G‐Azakouli. This variant haemoglobin is found
with haemoglobin G‐Philadelphia is discussed later.
in 0.044% of Afro‐Caribbeans [34]. It occurs in a
variety of other ethnic groups including Afro‐
Diagnosis Americans, Algerians, Italians (northern Italy
and Sardinia), Chinese and Melanesians.
The diagnosis and differential diagnosis is as for
Interestingly it can result from two different
haemoglobin D‐Punjab homozygosity.
mutations, AAC→AAG on a chromosome with
the −α3.7 deletion and AAC→AAA in the α2 gene
on a chromosome with the normal complement
Haemoglobin G‐Philadelphia
of two α genes. The former is found in Afro‐
Haemoglobin G‐Philadelphia, α268Asp→Lysβ2, is of Caribbeans and Afro‐Americans and the latter in
no clinical significance but has the potential to Italians,
296 Chapter 5

with haemoglobin D‐Punjab/β0
thalassaemia compound
heterozygosity showing
hypochromia, microcytosis and
poikilocytosis, including target cells.
MGG ×100.
Haemoglobin G‐Philadelphia trait On cellulose acetate electrophoresis and HPLC,

haemoglobin G‐Philadelphia, being an α chain vari
Haemoglobin G‐Philadelphia trait is of no clinical
ant, is associated with a variant haemoglobin A2,
significance. In about 80% of cases this mutated
producing a split haemoglobin A2 band or peak.
gene occurs on an −α3.7 chromosome, whereas the
Similarly, at birth there is a variant haemoglobin F.
other 20% of cases have αGα.
The A2 variant, haemoglobin G2, has been reported
to be present in a lesser amount than haemoglobin
A2, with the total A2 plus G2 percentage being nor
Clinical features
mal or slightly elevated [102].
There are no clinical features. Haemoglobin G‐Philadelphia can be distin
guished from haemoglobin D‐Punjab by isoelectric
focusing and HPLC (Fig. 5.36). On capillary electro
Laboratory features phoresis, G‐Philadelphia appears in the same zone
as D‐Punjab and Lepore but the lower percentage
Blood count. In those with the genotype in −αG/αα
and the presence of haemoglobin G2 indicate that it
there may be mild anaemia or a normal Hb. In one
is an α chain variant and permits distinction from
study the MCV in adults with this genotype was
D‐Punjab (Fig. 5.37).
71–80 fl and the MCH 22.7–26.8 pg [102]. In those
who also have α+ thalassaemia in trans the microcy
tosis is more marked.
Diagnosis
The diagnosis requires identification of haemoglo
Blood film. The blood film is normal.
bin G‐Philadelphia and haemoglobin A by two reli
able independent techniques. It can be difficult to
Other investigations. The variant haemoglobin com detect the split A2/G2 band on electrophoresis so
prises 20–25%, 30–35% or 45–48% of total haemoglo diagnosis by a combination of cellulose acetate
bin, with the trimodal distribution likely to indicate alkaline electrophoresis and agarose gel acid elec
groups of individuals with one αG gene and the num trophoresis is not very reliable. Isoelectric focusing
ber of normal α genes being three, two or one. and HPLC give more dependable information.

haemoglobin G‐Philadelphia
heterozygosity; haemoglobin
G‐Philadelphia appears in the
D window and was 30% with a
retention time of 4.23 minutes; note that
the G2 fraction, which helps to identify
this as an α chain variant, appears in
the S window; peaks, from left to right,
are haemoglobin F (shaded), post‐
translationally modified haemoglobin
A (two peaks) and haemoglobins A0,
A2 (shaded), G and G2.
Haemoglobin G‐Philadelphia homozygosity Haemoglobin O‐Arab

and coinheritance of haemoglobin
Haemoglobin O‐Arab (also reported as haemoglobin
G‐Philadelphia and other variant
Egypt) is a β chain variant, α2β2121Glu→Lys. It has been
haemoglobins or thalassaemias
found in a great variety of ethnic groups but is not
Homozygosity for −αG has been reported. There is common in any of them. Despite its name, it is actu
marked hypochromia and microcytosis. ally quite uncommon among Arabs. It appears to be
Compound heterozygosity for −αG and α0 thalas of African rather than Arab origin, with a distribution
saemia leads to haemoglobin H disease with similar to that of the Ottoman empire [104]. It has
haemoglobin G‐Philadelphia but no haemoglobin A. been suggested that it entered the Turkish domain
Individuals who are heterozygous for both with Sudanese contingents of the Turkish army or,
haemoglobin G‐Philadelphia and haemoglobin
alternatively, that it arose among the Greek Pomaks,
D‐Punjab may have a normal Hb or mild anaemia who have the highest prevalence known [105]. In
[103]. The blood film and reticulocyte count are addition to its occurrence among Arabs (in Israel,
normal. Haemoglobin electrophoresis at alkaline
Yemen, Egypt, Saudi Arabia), it has been reported in
pH shows 30–40% haemoglobin A, approximately Greeks, in Italians, in Eastern Europe (Bulgaria, for
45% haemoglobin with the mobility of S (represent mer Yugoslavia, Hungary, Albania), in gypsies and
ing haemoglobin D‐Punjab and haemoglobin among Africans (Kenyans, Sudanese, Moroccans,
G‐Philadelphia), approximately 15% haemoglobin Tunisians) and those of African descent. In a survey
with the mobility of C/E/A2 (representing haemo in Jamaica the prevalence was 0.016% [34]. Its main
globin A2 and the hybrid G‐Philadelphia /D‐Punjab importance is because of possible interaction with
haemoglobin) and about 2% of haemoglobin G2. haemoglobin S (see p. 233) and β thalassaemia.
Haemoglobin electrophoresis at acid pH is normal
since both variant haemoglobins move with hae Haemoglobin O‐Arab trait
moglobin A.
Coinheritance of haemoglobin G‐Philadelphia Haemoglobin O‐Arab trait is of genetic but no other
with haemoglobin S heterozygosity and homozygo clinical significance.
sity has been discussed on pp. 200 and 221, and
with haemoglobin C and haemoglobin S plus C on Clinical features. Haemoglobin O‐Arab trait causes
pp. 275 and 229. no clinical abnormality.
298 Chapter 5
Fig. 5.37 Capillary electrophoresis (Sebia Capillarys 3) in haemoglobin G‐Philadelphia heterozygosity. The peaks from
left to right are: unidentified, haemoglobin A, haemoglobin F, haemoglobin G‐Philadelphia, haemoglobin A2 and
haemoglobin G2.
Laboratory features ahead, whereas on agarose gel at acid pH it moves

Blood count. The Hb is normal. The MCV may be close to, but slightly behind, haemoglobin S
normal or borderline low. (Fig. 5.39). On HPLC, haemoglobin O‐Arab elutes in
the C window but has a retention time between
Blood film. The blood films can show slight aniso those of haemoglobins S and C (Fig. 5.40) (see also
cytosis, poikilocytosis, hypochromia and a few Fig. 2.19o). In addition, there is a minor peak in the
target cells [24]. haemoglobin D window [106] and another peak also
representing post‐translationally modified O‐Arab
Other investigations. The percentage of dense red close to the main peak (see Fig. 5.40). On capillary
cells is increased [9]. Haemoglobin O‐Arab is 38–43% electrophoresis, the O‐Arab peak appears in the
of total haemoglobin. It has a mobility similar to haemoglobin A2 zone. The relatively low percentage
haemoglobin C at alkaline pH (Fig. 5.38). On citrate may result, as with haemoglobins S and C, from the
agar at acid pH it moves close to S but slightly positive charge of the βO‐Arab chain. Individuals with
diagnosis is with haemoglobins C, E and C‐Harlem.

Physicochemical characteristics are very similar to
those of haemoglobin C‐Harlem but in the simple
heterozygotes the distinction can be made easily
because C‐Harlem has a positive sickle solubility
test. If the primary diagnostic method is HPLC,
haemoglobin C‐Harlem must again be considered.
Haemoglobin O‐Arab disease

Homozygosity for haemoglobin O‐Arab has been
reported in Bulgaria, former Yugoslavia (a gypsy
family), Morocco, Tunisia, the Sudan and Kenya
Fig. 5.38 Cellulose acetate electrophoresis at alkaline pH [24, 104, 107–109].
in a patient with haemoglobin O‐Arab heterozygosity;
haemoglobin O‐Arab has the same mobility as
Clinical features
haemoglobins C and E but is present as a higher
percentage than haemoglobin E; AFSC indicates a control Homozygotes may be asymptomatic with compen
sample containing haemoglobins A, F, S and C. sated haemolysis or there may be recurrent jaundice
and anaemia. The spleen may be enlarged.
Laboratory features
Blood count. The Hb can be normal or reduced
C and the MCV is reduced. The reticulocyte count is
mildly increased.
S
A Blood film. The blood film shows numerous target
F cells and sometimes NRBC.
a b c d e f g h
Other investigations. Haemoglobin O‐Arab comprises
Patient
almost all the haemoglobin with a small amount of
Fig. 5.39 Agarose gel electrophoresis at acid pH in a haemoglobin A2. Haemoglobin F may be increased, at
patient with haemoglobin O‐Arab heterozygosity: lanes least in children [110]. The percentage of dense cells is
show, from left to right, (a) F plus faint S; (b) A + C; (c) A
increased, the abnormality being comparable to that
plus O‐Arab; (d) faint F and A plus S; (e, f and g) A plus C;
seen in homozygosity for haemoglobin C [9].
(h) A, F, S and C (control sample); note that the
haemoglobin O‐Arab is slightly slower than haemoglobin
S (i.e. slightly closer to C). Haemoglobin O‐Arab/β thalassaemia
Haemoglobin O‐Arab/β0 thalassaemia has been
described in Bulgaria. Hungary and Italy and hae
coexisting α thalassaemia trait have a somewhat moglobin O‐Arab/β+ thalassaemia in Saudi Arabia,
lower percentage of the variant haemoglobin. Turkey and in an Afro‐Caribbean [24, 26].
Diagnosis Clinical features

If the primary diagnostic method is cellulose ace This compound heterozygous state causes slight to
tate electrophoresis at alkaline pH the differential moderate anaemia with jaundice and splenomegaly.
300 Chapter 5

haemoglobin O‐Arab heterozygosity;
haemoglobin O‐Arab has a longer
retention time than haemoglobin S, 4.84
minutes on this sample; the major
peaks, from left to right, are post‐
translationally modified haemoglobin
A (two peaks), haemoglobins A0 and A2
(shaded), two small peaks representing
glycated O‐Arab and other
post‐translationally modified O‐Arab,
and O‐Arab.

compound heterozygosity for
haemoglobin O‐Arab and β0
thalassaemia showing target cells and
irregularly contracted cells. The
patient was of Turkish Cypriot,
Egyptian and Jamaican ancestry. Red
cell indices were RBC 6.78 × 109/l, Hb
121 g/l, MCV 55.4 fl, MCH 17.8 pg
and MCHC 320 g/l. MGG ×100.
The haemolytic anaemia may be episodic, being Other investigations. Haemoglobins present are
precipitated by intercurrent infection. haemoglobins O and A2 with or without haemo
globin A. Haemoglobin F may be mildly
increased,
Laboratory features
Blood count. The reported Hb has ranged from
60 g/l to normal. The MCV is reduced and the retic
Coinheritance of haemoglobin O‐Arab
ulocyte count is mildly increased.
and other variant haemoglobins or
thalassaemias
Blood film. The blood film (Fig. 5.41) shows anisocy
tosis, poikilocytosis, hypochromia, microcytosis Coinheritance of haemoglobin S and haemoglobin
and target cells, which may be infrequent or O‐Arab causes significant disease. This condition
numerous. has been described on p. 233.
• replacement of an internal non‐polar amino acid

Unstable haemoglobins
with a polar amino acid, which has to be oriented
The term ‘unstable haemoglobin’ is best restricted to outwards and thus disrupts the molecule (inter
those variant haemoglobins that cause clinically rec ference with tertiary structure);
ognisable haemolysis. Other haemoglobins, for • interference of the binding of α and β subunits to
example haemoglobin E, are unstable in vitro but this each other; specifically impairment of the α1β1
probably does not contribute significantly to the asso dimeric bonds (interference with quaternary struc
ciated clinical features. Both α and β chain variant ture); this can result in dissociation into monomers,
haemoglobins can be unstable. A small number of γ which favours methaemoglobin formation;
chain variants have been described that lead to hae • elongation of the β chain;
moglobin instability and haemolysis in neonates, • probably interference of binding of α chain to
including haemoglobin F‐Poole [111] and haemoglo alpha haemoglobin stabilising protein.
bin F‐Bonheiden [112]. In addition, there are two γ The first unstable haemoglobins identified were
chain variants that are methaemoglobins and also two β chain variants, haemoglobin Zurich and
unstable (haemoglobin F‐M‐Fort Ripley and haemo haemoglobin Köln, both identified in the early
globin F‐M‐Osaka), which cause neonatal pseudocy 1960s. Subsequently, a case of ‘congenital non‐
anosis [113]. Occasionally the fetal form of an α chain spherocytic haemolytic anaemia’ reported in the
variant is more unstable than the adult form; this is 1950s was found to be attributable to haemoglo
the case with haemoglobin Hasharon, with α2Hasharonγ2 bin Bristol. Since these early reports, more than
being more unstable than α2Hasharonβ2 [114] and being 135 unstable haemoglobins have been described,
associated with haemolysis in neonates. Occasionally of which haemoglobin Köln appears to be the
a β chain variant is associated with neonatal jaundice most common. Many more unstable β variants
[115]. δ chain variants can also be unstable but because have been reported than unstable α variants,
of their low concentration this is of no clinical signifi perhaps because α variants, being a lower propor
cance. Several patients have been described with two tion of total haemoglobin, are more likely to go
unstable haemoglobins, together with haemoglobin unrecognised.
A or a β0‐thalassaemia determinant. This results not Mutations leading to an unstable haemoglobin
from three β genes in the genome but from post‐trans can be:
lational modification of an unstable haemoglobin. • a point mutation leading to replacement of one
Both haemoglobin Sydney and haemoglobin Atlanta amino acid by another or replacement of an amino
can be modified in this manner, producing haemo acid by the imino acid proline [24];
globin Sydney‐Coventry and haemoglobin Atlanta‐ • a point mutation followed by post‐translational
Coventry, respectively. Haemoglobin instability can modification of the haemoglobin encoded; leucine
result from an unstable globin chain or an unstable in an abnormal haem pocket is modified to hydrox
haemoglobin molecule. The causative primary abnor yleucine [116, 117];
mality can affect the secondary structure (α helix and • deletions of one to eight codons leading to dele
intervening turns), the tertiary three‐dimensional tion of a small number of amino acids; for example,
structure of the monomer or the quaternary structure haemoglobin Gun Hill has a β chain that lacks five
(relationship between monomers). amino acids including the haem‐binding site [118]
Haemoglobin instability may be consequent on: and haemoglobin J‐Biakra (unstable only in vitro)
• an abnormality of the haem pocket so that haem has an α chain that lacks eight amino acids [119];
is not firmly bound and water can enter the nor • tandem duplication of codons leading to duplica
mally hydrophobic haem pocket; haem depleted tion of a small number of amino acids; for example,
dimers and tetramers are present; haemoglobin Fairfax has five extra amino acids
• interference with the α helical structure, often [120], the same amino acids that are deleted in
because an amino acid is replaced by the imino acid haemoglobin Gun Hill;
proline, or interference with the interhelical bends • frameshift or STOP codon mutations leading to
(abnormality of secondary structure); synthesis of an elongated β chain.
302 Chapter 5
Of these various mechanisms, by far the that Heinz body haemolytic anaemia is less likely in
commonest is a single amino acid substitution. smokers with haemoglobin Zurich than in non‐
Unstable haemoglobins show a greater or lesser smokers [1, 121]. Some unstable haemoglobins are
tendency to Heinz body formation. Heinz bodies synthesised at a reduced rate. For example, reticulo
are composed of hemichromes, which are deriva cytes of a heterozygote for haemoglobin Köln syn
tives of ferric haemoglobin (methaemoglobin) in thesised haemoglobins A and Köln in a ratio of
which the haem has been lost from the haem 80:20 [125].
pocket and has bound elsewhere to denatured The proportion of abnormal haemoglobin for
globin [121]. Heinz bodies bind to the inner sur unstable β chain variants is very variable, ranging
face of the red cell membrane, probably by hydro from 35–40% in the case of haemoglobin Ham
phobic interactions rather than covalent bonds mersmith, to 10–15% in the case of haemoglobin
[121]. Binding is preferentially to band 3. The Köln, to almost undetectable in the case of very
membrane can also be damaged by free haem and unstable haemoglobins such as haemoglobin
free iron, leading to lipid peroxidation. The cell Cagliari. The proportion of an unstable β chain vari
becomes less deformable and may be trapped in ant is determined by:
the spleen; membrane loss and removal of Heinz • the rate of synthesis of the variant chain;
bodies occur in the spleen. • the rate of breakdown of the βvariant chain before
Most unstable haemoglobins cause haematologi association with α chain can occur;
cal abnormalities in heterozygotes and homozygo • the rate of association of α and βvariant chains in
sity has not been described. An exception to this comparison with the rate of association of α chains
generalisation is haemoglobin Bushwick, which is and normal βA chains;
only slightly unstable; it causes no significant • the rate of breakdown of αβvariant dimers before
abnormality in heterozygotes but caused significant assembly of dimers to tetramers can occur;
haemolytic anaemia in the one homozygote • the rate of denaturation of unstable haemoglobin
described [122]. Another exception is haemoglobin (with consequent removal as Heinz bodies).
Taybe, an α chain variant, which causes mild anae In the case of unstable α chain variants the pro
mia in heterozygotes and a more severe haemolytic portion of variant haemoglobin is usually less than
anaemia with hypochromia and microcytosis in 15% but may be 1–2% or even less. The explanation
homozygotes or compound heterozygotes with α for the variable proportion of the variant haemo
thalassaemia [123]; homozygous triplets with globin is as for the β chain variants.
hydrops fetalis have been described [124]. Unstable In some instances the proportion of an unstable
haemoglobins have been described in a great vari haemoglobin is also affected by the fact that the
ety of different ethnic groups and sometimes an mutation occurs in cis to an α thalassaemia determi
identical mutation has been found in a small num nant (haemoglobin Suan‐Dok, haemoglobin Petah
ber of individuals in very different parts of the Tikva) or in trans to a β thalassaemia determinant
world, indicating that independent mutations have (haemoglobin Leiden) [121].
occurred. A significant proportion of unstable Depending on the degree of instability of a
haemoglobins, probably about a third, are new variant globin chain or of a variant haemoglobin,
mutations, both parents being normal. various degrees of abnormality are possible: (i) a very
Unstable haemoglobins may, in addition to being unstable α or β chain is destroyed so rapidly that no
unstable, show either increased or decreased oxy variant globin chain or haemoglobin is detectable
gen affinity. Interaction with 2,3‐DPG can be and the phenotype is that of thalassaemia; (ii) an
impaired. They can also be particularly prone to unstable globin binds haem but cannot bind to
oxidation to methaemoglobin. Haemoglobin Zurich other globin chains so that it precipitates as Heinz
has an unusual abnormality of the haem pocket, bodies in erythroid precursors, leading to dyseryth
which is associated with an increased affinity for ropoietic and ineffective erythropoiesis, as in domi
carbon monoxide and an increase in carboxyhae nant β thalassaemia intermedia resulting from
moglobin; paradoxically, this reduces instability so heterozygosity for a hyperunstable β chain; (iii) a
lesser degree of instability permits a variant be intermittent or may intermittently worsen. Such
haemoglobin to be synthesised, leading to a haemo deterioration is often attributable to intercurrent
lytic anaemia and the clinical features recognised in infection or exposure to oxidant drugs. In the case
association with an unstable haemoglobin. Highly of haemoglobin Zurich, haemolysis may occur only
unstable α chain variants are usually clinically silent on exposure to oxidants, and smokers, for the rea
but if they interact with an α thalassaemia determi sons described on p. 302, exhibit less haemolysis
nant the phenotype may either be that of haemoglo than non‐smokers. Acute deterioration caused by
bin H disease or a haemolytic anaemia or may folic acid deficiency and aplastic crises due to par
simulate β thalassaemia intermedia; in the latter vovirus B19 infection have also been recognised.
instance there is no haemoglobin H detectable, When an unstable haemoglobin is abnormally
there is dyserythropoiesis and globin chain synthe prone to oxidation to methaemoglobin the patient
sis studies show a paradoxically increased α:β ratio may be cyanosed. Low oxygen affinity of an unsta
[126]. A phenotype with features of both thalassae ble haemoglobin can also cause cyanosis.
mia and a Heinz body haemolytic anaemia can be Some variant haemoglobins are only slightly
produced either by a reduced rate of synthesis of an unstable so that although in vitro tests for instability
unstable variant or by marked instability, leading to are positive there is usually no associated clinical
destruction of the variant globin chain or of αβ abnormality.
dimers before assembly of haemoglobin tetramers
can occur.
Laboratory features
Patients with a very unstable haemoglobin are
sometimes treated by splenectomy. Although this Blood count. The Hb may be normal, except during
can lead to a reduction in haemolysis, there may episodic haemolysis, or may be mildly, moderately
also be an increase in the platelet count, leading to or severely reduced. On average, the Hb is higher
thrombotic complications. They are also sometimes when the unstable haemoglobin also has increased
treated with hydroxycarbamide, leading to a bene oxygen affinity. The MCV can be elevated and the
ficial increase in haemoglobin F percentage and a MCH and MCHC reduced. The reduction of MCH
fall in the percentage of the unstable variant. and MCHC may be attributable to loss of haem or
removal of Heinz bodies by the spleen. Loss of
haem from the haem pocket can lead to a low
Clinical features
MCHC despite normocytic cells in the blood film
An unstable haemoglobin can cause severe, since the staining of the red cells in a blood film is
moderate or mild anaemia. Depending on the attributable to the globin content whereas the
severity of the abnormality and the chain that is chemical measurement of haemoglobin is depend
abnormal, presentation may be in infancy, childhood ent on the haem molecule [118]. The reticulocyte
or adult life. One α chain variant, haemoglobin count may be elevated constantly or intermittently.
Hasharon, causes significant haemolysis in neonates The degree of elevation of the reticulocyte count is
but not in adult life, indicating that α2Hasharonγ2 is not necessarily proportional to the reduction of the
more unstable than α2Hasharonβ2 [1]. There may be Hb because of the effect of altered oxygen affinity. If
intermittent or constant jaundice and passage of an unstable haemoglobin has a high oxygen affinity
dark brown or almost black urine, the latter the reticulocyte count will be higher in relation to
attributable to the excretion of dipyrroles, which are the Hb than if there is an unstable haemoglobin
abnormal breakdown products of haem. The with a low oxygen affinity. There can be thrombocy
incidence of gallstones is increased and they can topenia, which is sometimes disproportionate to the
occur in children [127]. Leg ulcers can occur. The degree of splenomegaly and not readily explicable.
spleen may be enlarged and hypersplenism Occasionally, when an unstable haemoglobin has
sometimes occurs. Rarely moya moya has been increased oxygen affinity, there is polycythaemia
described. When haemolysis is severe and chronic, rather than anaemia. Polycythaemia has also occasion
pulmonary hypertension can occur. Anaemia may ally been observed to develop following splenectomy.
304 Chapter 5
During haemolytic crises there is a fall in the Hb give normal results and DNA analysis or mass spec
and a rise in the reticulocyte count. The white cell trometry is required for demonstration of the
count and the neutrophil count usually also rise. variant haemoglobin.
When the unstable haemoglobin is a β chain
Blood film. The blood film (Fig. 5.42) can show anae variant, the concentration of haemoglobin A2 may
mia, macrocytosis, polychromatic macrocytes, mild be increased, as a consequence of selective dena
hypochromia, basophilic stippling and the presence turation and removal of the unstable variant.
of irregularly contracted cells and keratocytes (‘bite Haemoglobin F concentration can be moderately
cells’). Sometimes there is hypochromia and micro increased (e.g. to 10–12%). Methaemoglobin may be
cytosis [123]. Although it is often considered that present in fresh blood or, more often, appears
Heinz bodies cannot be detected on a Romanowsky‐ abnormally rapidly as the specimen ages.
stained film, it has been pointed out that they can in Heat and, usually, isopropanol tests for haemo
fact be identified in patients who have required globin instability are positive. The instability of
splenectomy because of the presence of an unstable some haemoglobins can also be demonstrated by
haemoglobin (and also in other patients with severe their response to mechanical shaking. The heat test
oxidant‐induced haemolytic anaemia) [24]. is more sensitive than the isopropanol test so will
During haemolytic crises there is an increase in detect variant haemoglobins showing a lesser
polychromasia and in the number of irregularly degree of instability. For example, the instability of
contracted cells and keratocytes. Functional hypo haemoglobin Olmsted is detected only by a heat
splenism, consequent on reticuloendothelial over test [129]. For this reason it has been suggested that
load, can lead to the appearance of Howell–Jolly both tests be employed. Interestingly the colour of
bodies. the precipitate varies, depending on the tendency of
the unstable haemoglobin to lose haem. Thus hae
Other investigations. Haemoglobin electrophoresis moglobin Hammersmith gives a red‐brown precipi
may be abnormal (Fig. 5.43) but is basically normal tate representing haemoglobin Hammersmith with
in more than half of the reported unstable haemo most of its haem groups still attached, whereas hae
globins, specifically in those with no change in moglobin Santa Ana loses haem group so readily
charge [1]. If the variant haemoglobin has a normal that the precipitate is almost white [24].
electrophoretic mobility there may nevertheless be A Heinz body test may be positive or may become
a tail behind the main haemoglobin band repre positive on incubation of the blood at 37°C for
senting either denatured haemoglobin or haemo 24 hours. During acute haemolytic episodes, includ
globin with a variable degree of haem depletion ing those induced by drugs, a previously negative
[128]; such bands are more likely if the specimen is Heinz body test may become positive. This is both
not fresh. Disappearance of an abnormal band because of an increased rate of formation of Heinz
when haemin is added demonstrates that it results bodies and because reticuloendothelial overload
from haem depletion. In the case of unstable β chain means that, once formed, Heinz bodies are not
variants, free α chains may form a discrete band being pitted from red cells by hepatic and splenic
near the origin, just behind haemoglobin A2. macrophages. If a splenectomy has been necessary,
Occasionally, two abnormal bands have been a Heinz body test is likely to be positive, with Heinz
reported, for example in the case of haemoglobin bodies sometimes being present in almost every
Rush, apparently representing migration of asym red cell.
metric hybrids (α2ββRush) as well as haemoglobin A Oxygen affinity may be increased (e.g. haemo
and haemoglobin Rush [1]. globin Köln) or reduced (e.g. haemoglobin
Some unstable haemoglobins appear as a discrete Hammersmith). Of the reported unstable haemo
peak on HPLC; in others there is a minor compo globins, oxygen affinity has been found to be
nent apparent, representing degraded haemoglobin reduced in 30%, normal in 20% and increased in
(Fig. 5.44). The shape and number of minor peaks 50% [1]. Since these conditions are generally
varies according to the age of the sample. Sometimes heterozygous, the oxygen dissociation curve is

standard electrophoretic and HPLC techniques all generally biphasic, representing the mixture of

(a)
(b)
Fig. 5.42 Blood films of three

patients with unstable
haemoglobins showing:
(a) irregularly contracted cells,
macrocytosis and
thrombocytopenia in a patient
with haemoglobin Köln; (b)
several stomatocytes and several
irregularly contracted cells in a
patient with haemoglobin Siriraj;
and (c) hypochromia and
poikilocytosis in a patient
with haemoglobin St Mary’s.
MGG ×100. (c)
306 Chapter 5
ormal and variant haemoglobins. Red cell life

n There may be an increase in bilirubin concentra
span is reduced but measurements may be inaccu tion (mainly unconjugated bilirubin) and LDH
rate for the following reasons: 51Cr may bind prefer together with evidence of intravascular haemolysis
entially to the unstable haemoglobin; 51Cr may lead such as a reduced serum haptoglobin concentration,
to denaturation of the unstable haemoglobin; loss of reduced serum haemopexin and the presence of
51
Cr may indicate removal of a Heinz body by the serum methaemalbumin and urinary haemosiderin.
spleen rather than death of a red cell [24]. It should be noted that inaccurate (low) pulse oxi
Bone marrow examination (Fig. 5.45) may meter measurements of oxygen saturation have
show erythroid hyperplasia and dyserythropoietic been reported as a result of the abnormal absorp
features such as bi‐ and multinuclearity and
tion spectrum of the variant haemoglobin, with a
nuclear irregularity or lobulation [130]; ring number of unstable haemoglobins including hae
sideroblasts are sometimes present. moglobin Köln, haemoglobin Hammersmith and
haemoglobin Cheverly [131, 132].
Diagnosis
Haemoglobin electrophoresis, HPLC or both should
be performed. However, since other tests may all
give normal results, diagnosis requires a specific
test for an unstable haemoglobin followed by DNA
analysis or mass spectrometry. A blood film
examination is particularly useful.
Coinheritance of an unstable haemoglobin

and other variant haemoglobins or
thalassaemia
The severity of disease caused by an unstable β
Fig. 5.43 Haemoglobin electrophoresis on cellulose chain variant may be reduced by coinheritance of α
acetate at alkaline pH in a patient with haemoglobin thalassaemia [133]. Coinheritance of an unstable
Köln; a faint band representing denatured haemoglobin haemoglobins (e.g. haemoglobin Acharnes, haemo
Köln is apparent behind haemoglobin A and between S globin Arta or haemoglobin Lulu Island) with β
and C positions; AFSC, control sample. thalassaemia has been associated with the clinical

haemoglobin Köln heterozygosity
showing minor components that
represent denatured haemoglobin
Köln (far right); other peaks, from left
to right, are haemoglobin F (shaded),
haemoglobins A0 plus unaltered
haemoglobin Köln and haemoglobin
A2 (shaded).
(a)
Fig. 5.45 Bone marrow aspirate

in haemoglobin St Mary’s:
(a) MGG ×100 showing erythroid
hyperplasia and
dyserythropoiesis (lobulated
nuclei and basophilic stippling);
(b) Perls’ stain ×100 showing
ring sideroblasts. (b)
picture of β thalassaemia intermedia (see Table as lavender‐blue, slate grey or brownish‐slate.

3.10). Coinheritance of an unstable haemoglobin However, with haemoglobins M‐Boston and
and β0 thalassaemia in trans can lead to a severe M‐Iwate, which have a low oxygen affinity, there is
haemolytic anaemia [134]. also an element of true cyanosis [121]. In Japan,
terms meaning ‘black mouth’ or ‘black blood’ or
‘black child’ were used for carriers of haemoglobin
Haemoglobin M
M‐Iwate. The first haemoglobin M was recognised
The designation haemoglobin M is given to a vari in 1948, a year before the first description of haemo
ety of α, β and γ chain haemoglobin variants that globin S [135].
show an increased tendency to oxidation to methae The molecular abnormality is usually replace
moglobin, with resultant cyanosis or, more correctly, ment of a histidine residue in the haem pocket by
pseudocyanosis (Fig. 5.46). The abnormal colour of tyrosine, so that the iron of the haem molecule is
the patient’s skin differs from the purplish‐blue of stabilised in the ferric (Fe3+) form. Either the proxi
true cyanosis and has been variously described mal or the d istal histidine may be involved
308 Chapter 5
(Fig. 5.47). The exception is haemoglobin

M‐Milwaukee, in which the longer side chain of
glutamic acid, substituted for valine at β67, reaches
the haem molecule and the ferric state is stabilised
[121]. Once the variant globin chain has been
oxidised it becomes non‐functional from the point
of view of oxygen transport. In the case of the
three α chain variants M‐Boston, M‐Iwate and
M‐Milwaukee, only the β chains can bind oxygen;
in the case of the two β chain variants M‐Saskatoon
and M‐Hyde Park, only the α chains can bind
oxygen. Conversion to methaemoglobin may be
accelerated by exogenous oxidants. Some M hae
moglobins have a reduced oxygen affinity. Some
are also unstable and instability may be aggra
vated by oxidant stress. Haemoglobin M‐Hyde
Park is an example of an unstable haemoglobin M
with partial haem loss leading to haemoglobin
instability and haemolysis. They all have a marked
reduction in cooperativity indicated by a low n
number [136] (see later). These variant haemoglob
ins are summarised in Table 5.3 [2, 136–141].
The α chain M haemoglobins investigated to
date have all been α1 variants [2]. It has been pos
tulated that this is because with an α2 mutation
the preferential combination of the variant α
Fig. 5.46 Clinical photograph of a patient with chain with normal non‐α chains could lead to a
haemoglobin M‐Saskatoon showing pseudocyanosis. haemoglobin M percentage incompatible with
(With thanks to Dr Thomas Erblich and the patient.) fetal viability.
CH
H
CH C CH3
H3C CH CH2
N N O N NH
CH Fe CH
Distal
N N histidine
NH N of globin
H3C CH3 chain
Proximal CH2 C CH2 Fig. 5.47 Diagram showing the

histidine H relationship of haem to the proximal
of globin CH2COOH CH2COOH and distal histidines of the haem
chain
pocket; mutation of either of the
haem
corresponding codons can lead to a
haemoglobin M.
Table 5.3 M haemoglobins [2, 136–141].
Haemoglobin Oxygen Usual percentage n number* Other features

affinity of variant
β chain variants
Haemoglobin M‐Saskatoon (α2β263(E7)His→Tyr) Normal 35% 1.2 Mild or no haemolysis
Haemoglobin M‐Hyde Park (α2β2 )†

92(F8)His→Tyr
Normal NA 1.3 Mild haemolysis
Haemoglobin M‐Milwaukee (α β 2 2
67(E11)Val→Glu
) Reduced 50% 1.2 No haemolysis
Haemoglobin Chile (α2β2 28Leu→Met

) Cyanosis
α chain variants
Haemoglobin M‐Iwate (α287(F8)His→Tyrβ2)‡ Reduced 19% 1.1 Not anaemic
Haemoglobin M‐Boston (α2 58(E7)His→Tyr

β2) Reduced NA 1.2 Sometimes anaemic
γ chain variants
Haemoglobin F M‐Osaka (α2Gγ263(E7)His→Tyr) 5‐8%, 12%, 17.5% Neonatal cyanosis [137]
Haemoglobin F M‐Fort Ripley (α2Gγ292(F8)His→Tyr) Slightly 10%, 9% Neonatal cyanosis, also

increased unstable [138]
Haemoglobin F‐Circleville (α2Gγ263(E7)His→Leu)§ Normal 14% Neonatal cyanosis [139]
Haemoglobin F M‐Viseu (Hb Shady Grove) 16%, 7–15% Neonatal cyanosis [140, 141]
(α2Gγ228Leu→Met)
NA, not available.

* The n number indicates the degree of cooperativity between haemoglobin subunits; normal haemoglobins have an n value of 2.7–3.0
whereas a haemoglobin with no cooperativity has an n value of 1 [136].
† Sometimes called haemoglobin Milwaukee II.
‡ Also known as haemoglobin Kankakee, haemoglobin M‐Oldenberg and haemoglobin M‐Sendai [2].
§ Presumptive identification as a haemoglobin M.
Clinical features haemoglobin M‐Saskatoon, which may be associ

ated with a compensated haemolytic state.
There is cyanosis from birth in the case of α chain
variants and from around 3–6 months of age in the
Blood film. The blood film can show poikilocytosis.
case of β chain variants. Babies with γ chain variants
(haemoglobin F‐M‐Osaka, haemoglobin F‐M‐Fort Other investigations. Diagnosis is usually by spec
Ripley and haemoglobin F‐Circleville) are mildly trometry since the absorbance spectrum of methae
cyanosed at birth but cyanosis lessens as β chain moglobin differs from that of haemoglobin A. The
production takes over from γ [113]; such babies absorption spectra of different haemoglobin Ms
may require supplemental oxygen in the first few differ from each other but not sufficiently to permit
months of life. When there is a significant amount definitive diagnosis. However, if haemoglobin is
of methaemoglobin present, a blood sample appears first oxidised, the spectra are characteristic (Fig. 5.48).
macroscopically brownish. Haemoglobin electrophoresis is normal at alkaline
pH but electrophoresis at a pH of 7 in a phosphate
buffer shows abnormal mobility of haemoglobin
Laboratory features
Ms since the mutation involved histidine, which
Blood count. The Hb may be normal or high. ionises at neutral pH [142]. Other congenital and
Occasional patients are anaemic. The reticulocyte acquired causes of methaemoglobinaemia need to
count is sometimes elevated, particularly in the be excluded (see p. 336). In contrast to low a ffinity
case of carriers of haemoglobin M‐Hyde Park and haemoglobins, oxygenation of the sample in vitro
310 Chapter 5
0.5 0.5
Boston
0.4 0.4
Hyde Park
Absorbance
Absorbance
0.3 0.3
Saskatoon Iwate
0.2 0.2
0.1 0.1
Fig. 5.48 Absorption spectra of
haemoglobins M‐Iwate, M‐Boston,
0 0 M‐Hyde Park and M‐Saskatoon in
500 600 500 600
Wavelength (μm) Wavelength (μm) comparison with methaemoglobin A.
(Modified from reference 1.)
does not restore the normal colour. Haemoglobin which is associated with marked microcytosis and
electrophoresis may show an abnormal band but unbalanced chain synthesis) [144]. The first high
the normal and variant haemoglobins are more affinity haemoglobin recognised was an α chain
readily separated if they are both converted to variant, haemoglobin Chesapeake, described by
methaemoglobin prior to electrophoresis. The oxy Charache, Weatherall and Clegg in 1966 [145]. The
gen dissociation curve can be abnormal in shape diagnosis of a high affinity haemoglobin should
and either right or left shifted. A test for haemoglo be considered when there is unexplained poly
bin instability should be performed since some cythaemia, particularly in a young person or
unstable haemoglobins also show accelerated for when a patient with a high Hb has a family his
mation of methaemoglobin. Heinz bodies may be tory of polycythaemia. It is important that patients
present, particularly in the case of haemoglobin with polycythaemia resulting from a high affinity
M‐Saskatoon. haemoglobin are not misdiagnosed as polycythae
Pulse oximetry is inaccurate in the presence of mia vera. In the past, such misdiagnosis led to
methaemoglobin since it has similar light absorp exposure to 32P, for up to 10 years, and even to 32P‐
tion characteristics to oxyhaemoglobin, giving a induced myelodysplastic syndrome or acute mye
falsely elevated estimate of haemoglobin saturation loid leukaemia [1, 146]. Onset of a haematological
[143]; co‐oximetry is required. abnormality is from birth in the case of an α chain
variant and during the first year of life in the case
of a β chain variant.
High affinity haemoglobins
More than 120 high affinity variant haemoglob
The presence of a variant haemoglobin with a ins have been described. In most cases there is
high oxygen affinity usually leads to polycythae only a single copy of the mutant gene but in the
mia. The exception is when a high affinity haemo case of the α variants haemoglobin Tarrant and
globin is also very unstable, as is the case with haemoglobin Longview, individuals with two
haemoglobin Köln. About a third of unstable hae copies of the gene have been described and com
moglobins show an increased oxygen affinity but pound heterozygosity for haemoglobin Tarrant
since the dominant feature is haemolysis they are and haemoglobin Jackson has also been recog
categorised primarily as unstable haemoglobins nised [147]. Homozygosity for the β chain variant
rather than as high affinity haemoglobins. haemoglobin Abruzzo has likewise been described
Some high affinity haemoglobins also produce a [147]. Compound heterozygosity for a β chain var
thalassaemic phenotype (e.g. haemoglobin Crete, iant and β0 thalassaemia can occur, and leads to a
more severe phenotype. The molecular abnormal hyperviscosity – headache, vertigo, tinnitus and
ity may result in: paraesthesia – can occur [150]. Serum erythropoi
• interference with the α1β2 contacts, which nor etin is normal when the patient is in a stable state
mally allow movement of the haemoglobin subu but is increased by venesection. Theoretical con
nits in relation to each other on oxygenation, with siderations would suggest that transfer of oxygen
stabilisation of the R (oxy) conformation or destabi to the fetus might be impaired when a woman has
lisation of the T (deoxy) conformation; a high affinity haemoglobin but in practice preg
• interference with α1β1 contacts with major disrup nancy is generally uneventful.
tion of the quaternary structure and favouring of
the R (oxy) conformation;
Laboratory features
• limited polymerisation, restraining the quater
nary conformation and favouring the R (oxy) Blood count. The RBC, Hb and haematocrit are high
conformation; in the normal range or elevated but other red cell
• reduced binding to 2,3‐DPG; indices are normal. Whether the Hb is above the
• abnormality of the amino end of the α chain or of normal range is determined by:
either the amino or carboxy end of the β chain, lead • the degree of shift of the oxygen dissociation
ing to disruption of the quaternary structure and curve (i.e. the P50o2);
favouring of the R (oxy) conformation; • the percentage of the variant haemoglobin (deter
• alteration of the haem pocket. mined in turn by whether the variant haemoglobin
Occasional high affinity haemoglobins (e.g. Hb is an α or a β variant and whether it is also unstable
Headington) have a mutation leading to alteration and whether there is coexisting thalassaemia).
of a surface amino acid residue; the molecular Somewhat surprisingly, high affinity haemoglo
mechanism of the altered affinity is not clear [148]. bins have been associated not only with a true
Haemoglobin Olympia, which also has a substitu polycythaemia but also with a relative polycythae
tion affecting a surface amino acid, has now been mia (i.e. with a reduced plasma volume but normal
shown to be characterised by self‐aggregation of red cell mass) [151]. The white cell count and the
haemoglobin tetramers [121]. platelet count are usually normal but some indi
High affinity haemoglobins often have reduced viduals have been reported with an elevated white
cooperativity and a reduced Bohr effect and some cell count [152].
are unstable.
The molecular abnormalities responsible include Blood film. The blood film appears ‘packed’ because
not only single point mutations but also double the high haematocrit and increased whole blood
point mutations (haemoglobin Poissy, only one of viscosity leads to a thick film.
the mutations is relevant), deletions, insertions and
frameshift mutations (haemoglobin Tak) [121]. Other laboratory features. The total red cell mass, as
A number of high affinity haemoglobins form measured by radio‐isotope dilution techniques, is
stable hybrid tetramers (with both a normal and a raised in relation to what is normal for an individ
variant β chain) leading to a broad peak on HPLC ual of the same gender, height and weight. The
and two variant bands on electrophoresis [149]. plasma volume can be normal or reduced. The par
tial pressure of oxygen in arterial blood is normal.
Haemoglobin electrophoresis on cellulose acetate
Clinical features
at alkaline pH sometimes shows a haemoglobin
The individual with a high affinity haemoglobin variant with abnormal electrophoretic mobility but
may be plethoric but otherwise well. In later life it this is not necessarily so. Electrophoresis at acid pH
is likely that the risk of thrombosis is increased. should be carried out even if electrophoresis at
In contrast to polycythaemia vera, there is no alkaline pH is normal as sometimes a variant
hepatomegaly or splenomegaly. Symptoms of haemoglobin is revealed only at acid pH. About
312 Chapter 5
three quarters of unstable haemoglobins can be

detected by either electrophoresis or isoelectric 100
focusing. HPLC or mass spectrometry permits
detection of other high affinity haemoglobins [153].
Determining the P50o2 on a blood gas analyser can
O2 saturation (%)
be used for screening for a high affinity haemoglo
bin and should be done whenever there is a suspi
50
cion of a high affinity haemoglobin, even if no
electrophoretically abnormal variant or HPLC
abnormality has been detected. If an abnormality is
detected, an oxygen dissociation curve can be per
formed. If a relevant abnormality is found and no
variant haemoglobin has been detected by other 0
techniques, DNA analysis is indicated. The variant 0 20 40 60 80 100 120
haemoglobin is usually around 50% in the case of a PO2 (mmHg)
β chain variant and 12–40% in the case of an α chain Haemoglobin Bethesda
variant. The percentage of β chain variants is higher Haemoglobin Luton
if there is coexisting β thalassaemia in trans, being Haemoglobin A
almost 100% in the case of coexisting β0 thalassae Haemoglobin Kansas
mia. The percentage of α chain variants is similarly
greater if there is coexisting α thalassaemia. Fig. 5.49 Oxygen dissociation curves showing two high
The oxygen dissociation curve is left shifted and affinity haemoglobins and one low affinity haemoglobin
the P50o2 is decreased (Fig. 5.49). Reported P50o2 has in comparison with haemoglobin A; the vertical dotted
ranged from as low as 9.5 mmHg (haemoglobin line shows the normal partial pressure of oxygen in
Heathrow), 10 mmHg (haemoglobin McKees venous blood.
Rocks) or 11 mmHg (haemoglobin Syracuse) to
slightly below normal. Since heterozygosity is lower the Hb a marked rise of erythropoietin
usual, the oxygen dissociation curve is often bipha concentration occurs.
sic, representing the mixture of normal and variant
haemoglobins. In other cases the entire dissociation
Diagnosis
curve is shifted, possibly because of the presence of
hybrid haemoglobin molecules (i.e. α2βAβX) [154]. At least two standard electrophoretic or HPLC tech
The normal sigmoid shape of the oxygen dissocia niques are recommended, in addition to an oxygen
tion curve is usually lost to a variable extent, this dissociation curve and measurement of P50o2. A test for
being quantified by the n value. This is calculated an unstable haemoglobin should also be carried out in
according to Hill’s equation: KPn = [Y/1−Y], where patients with unexplained polycythaemia since some
K is a constant, P is the partial pressure of oxygen high affinity haemoglobins are unstable.
and Y is the fractional oxygen saturation. The log of
Y/1−Y plotted against the log of Po2 normally gives
Coinheritance of a high affinity
a straight line but, when the oxygen dissociation
haemoglobin and other variant
curve is biphasic, the line is bent. The value of n
haemoglobins or thalassaemia
when cooperativity is normal is about 2.6–2.7,
whereas n is approximately 1 when there is no Coinheritance of a high affinity β chain variant hae
cooperativity. High affinity haemoglobins typically moglobin and δβ0 thalassaemia has been observed
have n values of 1.1–1.9. to aggravate the polycythaemia, a result that would
Concentration of 2,3‐DPG is increased. The be predicted since the only haemoglobins present
serum erythropoietin concentration is normal are the variant (haemoglobin Headington or hae
or high but if the individual is venesected to moglobin Tak) and haemoglobin F, which also has
a high oxygen affinity [148, 155]. Coinheritance of a (whole blood P50o2 greater than 50 mmHg) are
high affinity haemoglobin and β0 thalassaemia can characterised by cyanosis. Cyanosis has been
similarly result in more marked polycythaemia observed, for example, with haemoglobins Kansas,
than in the simple heterozygoous state for the vari Beth Israel, St. Mande [1], Titusville [164] and
ant haemoglobin [156]. Coinheritance of β thalas Bassett [165]. In the case of an α chain variant, the
saemia and the high affinity β chain variants cyanosis is present from birth whereas with a β
haemoglobin Crete and haemoglobin San Diego chain variant it appears during the first year of life.
did not prevent the development of polycythaemia O2 saturation is reduced. If the variant haemoglobin
[24]. Coinheritance of haemoglobin Tak with is a low percentage, the Hb may fall within the
haemoglobin E was associated with a mild poly normal range [164] and cyanosis may be absent.
cythaemia [86]. Tissue extraction of O2 does not have a linear
relationship to P50o2; initially it increases as P50
increases, but then decreases so that by a whole
Low affinity haemoglobins blood P50 of 80 mmHg, oxygen extraction is again
normal [164].
At least 70 low affinity haemoglobins have been The mechanism of reduced oxygen affinity [121,
described. Some variant haemoglobins that are 166] may be:
more important for other characteristics also have • alteration of the α1β2 interface that either stabi
a low oxygen affinity so that a low haemoglobin lises the T (deoxy) conformation (e.g. haemoglobin
concentration in homozygous states is partly con Titusville) or destabilizes the R (oxy) conformation
sequent on better oxygen delivery to tissues with a (e.g. haemoglobin Kansas);
lessening of erythropoietic drive. This is true of • alteration of the α1β1 contacts that disrupt the
sickle cell anaemia. Some unstable haemoglobins molecule, favouring the T (deoxy) conformation
also have reduced oxygen affinity so that a low Hb (e.g. haemoglobin Presbyterian);
is well tolerated. If the instability is mild, as in the • increased affinity for 2,3‐DPG (e.g. haemoglobin
case of haemoglobin Kansas, the only clinical fea Aalborg);
tures may be a well‐tolerated anaemia; when affin • stearic hindrance in the haem environment (e.g.
ity is markedly reduced, there can be cyanosis. An haemoglobin Bologna);
unexplained association of low affinity haemo • alteration of the N‐terminal residue of the α chain
globins with pulmonary hypertension has been with resultant stabilisation of the deoxy form (e.g.
reported in three families [157–159]. More low haemoglobin Thionville) or abnormal interaction
affinity β chain than α chain variants have been with chloride ions, favouring the deoxy form
described, probably because the lower percentage (haemoglobin Lyon‐Bron) [166].
of an α chain variant means that its effects are less An increased concentration of 2,3‐DPG can con
likely to be noticed. In the neonatal period, low tribute to the low affinity, since desaturation leads
affinity γ chain variants, such as haemoglobin F‐ to a rise in pH which increases the synthesis of
Heuried, which is also unstable [160], and haemo 2,3‐DPG [121]. Low affinity haemoglobins some
globin F‐Cincinnati [161], can lead to cyanosis. times have reduced cooperativity or a reduced
Incidental diagnosis, particularly in neonates, can Bohr effect.
result from detection of a low O2 saturation despite
normal arterial blood gases [162]. The rate of
detection of low affinity haemoglobins is likely to
Haemoglobin F variants
have increased in the USA with the introduction of
screening for cyanotic congenital heart disease by At least 69 haemoglobin F variants are known,
pulse oximetry [163]. somewhat more being Gγ variants than Aγ variants
Variant haemoglobins with a moderate decrease [113]. In the neonatal period most Gγ variants consti
in oxygen affinity are characterised by anaemia, tute around 31% of total haemoglobin while Aγ vari
whereas those with a marked decrease in affinity ants are around 13% [113]. Several haemoglobin F
(Bio‐Rad Variant II) showing a double
peak resulting from the presence of a
haemoglobin A2 variant; both the
variant and the normal A2 appeared
within the A2 window so that
quantification was accurate, whereas
the most frequently observed A2
variant, A2′ (see Fig. 5.51), is
quantified as haemoglobin S; peaks,
from left to right, are haemoglobin F
(shaded), post‐translationally
haemoglobin A0 and haemoglobins A2
plus A2 variant (double peak, shaded).
Fig. 5.51 HPLC chromatogram (Bio‐Rad Variant II) in a patient with both haemoglobin A2′ and β thalassaemia
heterozygosity. The peaks from left to right are haemoglobin F (shaded), post‐translationally modified haemoglobin A
(two peaks), haemoglobin A2 (shaded) and haemoglobin A2′ (in S window). Note that the normal A2 peak is 3.5% but if
the A2′ peak is added, the total is 5.1% and the diagnosis of β thalassaemia heterozygosity becomes evident.
Fig. 5.52 Capillary electrophoresis (Sebia Capillarys 3) in a patient with both haemoglobin A2′ and β thalassaemia
heterozygosity. Note that the normal A2 peak is 3.5% but if the A2′ peak (labelled 4) is added, the total is 5.1%. Same
patient as Fig. 5.51.
variants have two amino acid substitutions. Many in Africa and occurs in between 1 and 2% of Afro‐
of these are not of any clinical significance but there Americans (1.6% overall of 14 321 individuals in
are several unstable fetal methaemoglobins, which nine studies [168]). Its highest frequency is among
can cause cyanosis in the neonatal period. One GγAγ the Herero people of Namibia where heterozy
fusion gene leads to γ thalassaemia. Several haemo gotes are more than 18% of the population [169].
globin F variants are M haemoglobins (see earlier). It also occurs with polymorphic frequency in the
Dogon region of Mali, being found overall in 2.5%
of the population but in up to 12% in one caste
Haemoglobin A2 variants
group [170]. Other A2 variants are common in very
More than 20 haemoglobin A2 variants (Fig. 5.50) specific ethnic groups. For example, approaching
are known [167]. They are not of any clinical sig 5% of Babinga pygmies in the Central African
nificance but can complicate the diagnosis of β tha Republic carry haemoglobin A2‐Flatbush and a
lassaemia trait since their presence can lead to a similar percentage of Sumatrans carry haemoglo
failure to measure the total of haemoglobin A2 plus bin A2‐Indonesia [157]. Most A2 variants are syn
haemoglobin A2 variant (Figs 5.51 and 5.52). The thesised at an approximately normal rate. Several
most common of these variants is haemoglobin A2′ very rare variants are unstable and therefore pre
(initially known as haemoglobin B2). It originated sent at a very low percentage [157].
316 Chapter 5
Check your knowledge 5.5 Haemoglobin O‐Arab

(a) is an α chain variant
One to five answers may be correct. Answers to (b) is the commonest variant haemoglobin in
almost all questions can either be found in this Arabs
chapter or can be deduced from information given. (c) can interact with haemoglobin S to cause
Correct answers are given on p. 324. a clinically severe disease
(d) is electrophoretically silent
5.1 Haemoglobin C homozygosity is usually
(e) has a significantly increased oxygen
associated with
affinity
(a) an increased reticulocyte count
(b) an increased incidence of gallstones
(c) a need for an increased iron intake to 5.6 A high oxygen affinity haemoglobin
prevent anaemia (a) usually causes anaemia
(d) a blood film showing target cells and (b) is associated with a reduced serum
irregularly contracted cells erythropoietin
(e) approximately equal proportions of (c) usually compromises the outcome of
haemoglobins A and C pregnancy because of reduced oxygen
delivery to the fetus
5.2 Haemoglobin E (d) is associated with a left‐shifted oxygen
(a) has its highest prevalence in South‐East dissociation curve
Asia (e) may also be unstable
(b) can interact adversely with β thalassaemia
(c) is synthesised at a reduced rate because a 5.7 Heterozygotes for haemoglobin E
splicing defect leads to impaired produc (a) often have 25–30% of the variant
tion of mRNA haemoglobin
(d) is unstable in vitro (b) occasionally have a normal blood film
(e) can be confused with haemoglobin C on and red cell indices
agarose gel electrophoresis at pH 6.5 (c) usually have West African ancestry
(d) can have red cell indices similar to those
5.3 Haemoglobin G‐Philadelphia of thalassaemia trait
(a) is of major clinical significance because of (e) on cellulose acetate electrophoresis at
its interaction with haemoglobin S alkaline pH can be readily confused with
(b) has arisen by at least two independent β thalassaemia trait
mutations
(c) is associated with a variant form of
5.8 Irregularly contracted cells can be a feature of
haemoglobin A2
(a) β thalassaemia trait
(d) has a high oxygen affinity
(b) homozygosity for haemoglobin C
(e) moves with haemoglobin S on electropho
(c) homozygosity for haemoglobin E
resis at alkaline pH and with haemoglo
(d) pulmonary crisis in sickle cell anaemia
bin A on electrophoresis at acid pH
(e) an unstable haemoglobin
5.4 An unstable haemoglobin
(a) may have increased oxygen affinity 5.9 Haemoglobin D‐Punjab (D‐Los Angeles)
(b) may have decreased oxygen affinity (a) occurs in a small proportion of Africans
(c) may be electrophoretically silent and Afro‐Caribbeans
(d) may be prone to conversion to (b) is of no clinical or genetic significance
methaemoglobin (c) is an α chain variant and is therefore
(e) usually only causes symptoms in associated with a split haemoglobin A2
homozygotes band on cellulose acetate electrophoresis
(d) has its highest prevalence in southern Pathophysiology, and Clinical Management, 2nd edn.
India Cambridge University Press, Cambridge, 2009.
(e) on cellulose acetate electrophoresis at The Globin Gene Server, hosted by Pennsylvania State
alkaline pH can easily be confused with University, USA and McMaster University, Canada.
haemoglobin S http://globin.cse.psu.edu/
5.10 Reduced red cell survival is characteristic of References

(a) Homozygosity for haemoglobin C
1 Bunn HF. Sickle hemoglobin and other hemoglobin
(b) Heterozygosity for haemoglobin Köln
mutants. In: Stamatoyannopoulos G, Nienhuis AW,
(c) Homozygosity for haemoglobin E
Majerus PW and Varmus H, eds. The Molecular Basis
(d) Heterozygosity for haemoglobin O‐Arab
of Blood Diseases, 2nd edn. WB Saunders,
(e) Haemoglobin H disease
Philadelphia, 1994, pp. 207–256.
2 Ameri A, Fairbanks VF, Yanik GA, Mahdi F,
5.11 An unstable haemoglobin may be associated
Thibodeau SN, McCormick DJ et al. (1999)
with Identification of the molecular genetic defect of
(a) Heinz body haemolytic anaemia patients with methemoglobin M Kanakee (M‐Iwate),
(b) a positive isopropanol test α87 (F8) His → Tyr: evidence for electrostatic model of
(c) oxidant‐induced haemolysis αM hemoglobin assembly. Blood, 94, 1825–1826.
(d) an abnormal band on haemoglobin 3 Itano HA and Neel JV (1950) A new inherited
electrophoresis representing haem‐ abnormality of human hemoglobin. Proc Natl Acad
depleted haemoglobin Sci U S A, 36, 613–617.
(e) splenomegaly 4 Eliakim R and Rachmilewitz EA (1983) Hemog
lobinopathies in Israel. Hemoglobin, 7, 479–485.
5.12 Cyanosis may be caused by heterozygosity for 5 Nagel RL and Steinberg MH. Hemoglobin SC dis
(a) haemoglobin C ease and HbC disorders. In: Steinberg MH, Forget
(b) haemoglobin D BG, Higgs DR and Nagel RL, eds. Disorders of
(c) haemoglobin E Hemoglobin: Genetics, Pathophysiology, and Clinical
(d) haemoglobin M Management. Cambridge University Press,
(e) low oxygen affinity haemoglobins Cambridge, 2001, pp. 756–785.
6 Sanchaisuriya K, Fucharoen G, Sae‐ung N,
5.13 The percentage of a variant haemoglobin is Siriratmanawong N, Surapot S and Fucharoen S
(2001) Molecular characterization of hemoglobin C
determined by
in Thailand. Am J Hematol, 67, 189–193.
(a) whether it is an α or β chain variant
7 Hassan MK, Taha JY, Al‐Naama LM, Widad NM
(b) the oxygen affinity of the variant
and Jasim SN (2003) Frequency of haemoglobi
haemoglobin
nopathies and glucose‐6‐phosphate dehydrogenase
(c) whether there is coexisting thalassaemia
deficiency in Basra. East Mediterr Health J, 9, 45–54.
(d) the affinity of the variant chain for
8 Agarwal A, Guindo A, Cissoko Y, Taylor JG,
normal α or β chain Coulibaly D, Koné A, Kayentao K et al. (2000)
(e) whether the variant chain is synthesised Hemoglobin C associated with protection from
at a normal or reduced rate severe malaria in the Dogon of Mali, a West African
population with a low prevalence of hemoglobin S.
Blood, 96, 2358–2363.
Further reading
9 Nagel RL, Fabry ME and Steinberg MH (2003)
Bain BJ, Wild BJ, Stephens AD and Phelan LA. Variant The paradox of hemoglobin SC disease. Blood Rev,
Haemoglobins: a Guide to Identification. Wiley‐Blackwell, 17, 167–178.
Oxford, 2010. 10 Lopera‐Mesa TM, Doumbia S, Konaté D, Anderson
Steinberg MH, Forget BG, Higgs DR and Weatherall JM, Doumbouya M, Keita AS et al. (2015) Effect of
DJ, eds. Disorders of Hemoglobin: Genetics, red blood cell variants on childhood malaria in
318 Chapter 5
Mali: a prospective cohort study. Lancet Haematol, 2, 23 Hirsch RE, Witkowska HE, Shafer F, Lin MJ, Balazs
e140–e149. TC, Bookchin RM and Nagel RM (1997) Hb C com
11 Poehling KA, Light LS, Rhodes M, Snively BM, pound heterozygotes [HbC/Hb Riyadh and HbC/
Halasa NB, Mitchel E et al. (2010) Sickle cell trait, Hb N‐Baltimore] with opposing effects upon haemo
hemoglobin C trait, and invasive pneumococcal globin C crystallization. Br J Haematol, 97, 259–265.
disease. Epidemiology, 21, 340–346. 24 Dacie J. The Haemolytic Anaemias, Volume 2, The
12 Ballas SK, Larner J, Smith ED, Surrey S, Schwartz E Hereditary Haemolytic Anaemias, Part 2, 3rd edn.
and Rappaport EF (1987) The xerocytosis of SC dis Churchill Livingstone, Edinburgh, 1988.
ease. Blood, 69, 124–128. 25 Schroeder WA, Powars D, Reynolds RD and Fisher
13 Couto FD, De Albuquerque ABL, Adorno EV, De JI (1976) Hb‐E in combination with Hb‐S and Hb‐C
Moura Neto JP, De Freitas Abbehusen L, De Oliveira in a black family. Hemoglobin, 1, 287–289.
JLB et al. (2003) α‐thalassaemia 2, 3.7 kB deletion 26 Huisman THJ (1997) Combinations of β chain
and haemoglobin AC heterozygosity in pregnancy: abnormal hemoglobins with each other and with
a molecular and haematological analysis. Clin Lab β‐thalassemia determinants with known mutations:
Haematol, 235, 29–34. influence on phenotype. Clin Chem, 43, 1850–1856.
14 Huisman THJ (1977) Trimodality in the percentages 27 Swart L and du Pisani A (2012) Images 307:
of β chain variants in heterozygotes: the effect of the Haemoglobin C/E compound heterozygosity in a
number of active Hbα structural loci. Hemoglobin, 1, 47‐year‐old female. www.bloodmed.com (accessed
349–382. March 2016).
15 Higgs DR, Clegg JB, Weatherall DJ, Serjeant BE and 28 Spencer Chapman M, Kiritkumar K, Lund K and
Serjeant GR (1984) The interaction of the ααα globin Bain BJ (2019) An unusual hemoglobinopathy:
gene haplotype and sickle haemoglobin. Br J compound heterozygosity for hemoglobins C and
Haematol, 58, 671–678. E. Am J Hematol, 94, 144.
16 Bain BJ and Phelan L (1996) An assessment of 29 Giordano PC, Harteveld CL, Michiels JJ, Terpstra
HemoCard Hemoglobin C and HemoCard W, Batelaan D, van Delft P et al. (1997) Atypical hae
Hemoglobin E kits for the detection of haemoglob moglobin H disease in a Surinamese patient result
ins C and E. MDA Evaluation Report MDA/96/57. ing from the combination of − −SEA and −α3.7
Medical Devices Agency, London. deletions with haemoglobin C heterozygosity. Br J
17 Cook CM, Smeltzer MP, Mortier NA, Kirk SE, Haematol, 96, 801–805.
Despotovic JM, Ware RE and Hankins JS (2013) The 30 Na‐Nakorn S and Minnich V (1957) Studies on
clinical and laboratory spectrum of Hb C [β6(A3) hemoglobin E. III. Homozygous hemoglobin E and
Glu→Lys, GAG>AAG] disease. Hemoglobin, 37, 16–25. variants of thalassaemia and hemoglobin E. A fam
18 Bain BJ (1993) Blood film features of sickle cell‐ ily study. Blood, 12, 529–538.
haemoglobin C disease. Br J Haematol, 83, 516–518. 31 Fairbanks VF, Oliveros R, Brandabur JH, Willis RR
19 Kraus AP and Diggs LW (1956) In vitro crystalliza and Fiester RF (1980) Homozygous hemoglobin E
tion of haemoglobin occurring in citrated blood mimics β‐thalassaemia minor without anemia or
from patients with haemoglobin C. J Lab Clin Med, hemolysis: hematologic, functional, and biosyn
47, 700–705. thetic studies of first North American cases. Am J
20 Boldorini R, Bozzola C, Gallarotti E and Ribaldone Hematol, 8, 109–121.
R (2006) Fatal splenic rupture in a pregnant woman 32 Lachant NA (1987) Hemoglobin E: an emerging
with hemoglobin C/β‐thalassemia and myeloid hemoglobinopathy in the United States. Am J
metaplasia. Arch Pathol Lab Med, 130, 1231–1232. Hematol, 25, 449–462.
21 Nagel RL, Lin MJ, Witkowska HE, Fabry ME, 33 Flint J, Harding RM, Boyce AJ and Clegg JB (1993)
Bestak M and Hirsch RE (1993) Compound hete The population genetics of the haemoglobinopa
rozygosity for hemoglobin C and Korle‐Bu: moder thies. Bailliere’s Clin Haematol, 6, 215–262.
ate microcytic anemia and acceleration of crystal 34 Serjeant GR, Serjeant BE, Forbes M, Hayes RJ, Higgs
formation. Blood, 82, 1907–1912. DR and Lehmann H (1986) Haemoglobin gene
22 Lawrence C, Hirsch RE, Fatalieo NA, Patel S, Fabry frequencies in the Jamaican population: a study in
ME and Nagel RL (1997) Molecular interactions 100,000 newborns. Br J Haematol, 64, 253–262.
between α‐G Philadelphia, Hb C, and HbS: pheno 35 Flint J, Harding RM, Boyce AJ and Clegg JB (1998)
typic implications for SC α‐G Philadelphia disease. The population genetics of the haemoglobinopa
Blood, 90, 2819–2825. thies. Bailliere’s Clin Haematol, 11, 1–51.
36 de Silva S, Fisher CA, Premawardhena A, capillary electrophoresis with the Primus high‐
Lamabadusuriya SP, Peto TEA, Perera G et al. (2000) pressure liquid chromatography in the evaluation
Thalassaemia in Sri Lanka: implications for the of hemoglobinopathies. Am J Clin Pathol, 130,
future health burden of Asian populations. Lancet, 824–831.
355, 786–791. 48 Higgins TN, Khajuria A and Mack M (2009)
37 Ashtiani MT, Monajemzadeh M, Sina AH, Berenji F, Quantification of HbA2 in patients with and without
Abdollahi M, Said MG and Alam M (2009) β‐thalassemia and in the presence of HbS, HbC, HbE,
Prevalence of haemoglobinopathies in 34,030 and HbD Punjab hemoglobin variants: comparison
healthy adults in Tehran, Iran. J Clin Pathol, 62, of two systems. Am J Clin Pathol, 131, 357–362.
924–925. 49 Mais DD, Gulbranson RD and Keren DF (2009) The
38 O’Riordan S, Hien TT, Miles K, Allen A, Quyen NN, range of hemoglobin A2 in hemoglobin E heterozy
Hung NQ et al. (2010) Large scale screening for hae gotes as determined by capillary electrophoresis.
moglobin disorders in southern Vietnam: implica Am J Clin Pathol, 132, 34–38.
tions for avoidance and management. Br J Haematol, 50 Borbely N, Phelan L, Szydlo R and Bain B (2013)
150, 359–364. Capillary zone electrophoresis for haemoglobinop
39 Premawardhena A, Allen A, Piel F, Fisher C, Perera athy diagnosis. J Clin Pathol, 66, 29–39.
L, Rodrigo R et al. (2017) The evolutionary and clini 51 Singha K, Fucharoen G and Fucharoen S (2019)
cal implications of the uneven distribution of the Molecular characterisation of haemoglobin E‐Udon
frequency of the inherited haemoglobin variants Thani (HBB:c.[79G>A;92+7A>G]): a novel form of Hb
over short geographical distances. Br J Haematol, E‐β‐thalassaemia syndrome. J Clin Pathol, 72, 322–324.
176, 475–484. 52 Kesornsit A, Jeenduang N, Horpet D, Plyduang T
40 Chernoff AL, Minnich V and Chongchareonsuk S and Nuinoon M (2018) Quantitative trait loci influ
(1954) Hemoglobin E, a hereditary abnormality of encing Hb F levels in Southern Thai Hb E (HBB:
human hemoglobin. Science, 120, 605–606. c.79G>A) heterozygotes. Hemoglobin, 42, 23–29.
41 Itano HA, Bergren WR and Sturgeon P (1954) 53 Graham EA, Felgenhauer J, Detter JC and Labbe RF
Identification of a fourth abnormal hemoglobin. (1996) Elevated zinc protoporphyrin associated
J Am Chem Soc, 76, 2278. with thalassemia trait and hemoglobin E. J Pediatr,
42 Fucharoen S. Hemoglobin E disorders. In: Steinberg 129, 105–110.
MH, Forget BG, Higgs DR and Nagel RL, eds. 54 Rees DC, Duley DA and Marinaki AM (2003)
Disorders of Hemoglobin: Genetics, Pathophysiology, Pyrimidine 5′ nucleotidase deficiency. Br J Haematol,
and Clinical Management. Cambridge University 120, 375–383.
Press, Cambridge, 2001, pp. 1139–1154. 55 Chapple L, Harris A, Phelan L and Bain BJ (2006)
43 Goheen MM, Campino S and Cerami C (2017) The Reassessment of a simple chemical method using
role of the red blood cell in host defence against fal DCIP for screening for haemoglobin E. J Clin Pathol,
ciparum malaria: an expanding repertoire of evolu 59, 74–76.
tionary alterations. Br J Haematol, 179, 543–556. 56 Wanapirak C, Sirichotiyakul S, Luewan S and
44 Hutagalung R,Wilairatana P, Looareesuwan S, Srisupundit K (2009) Comparison of the accuracy of
Brittenham GM, Aikawa M and Gordeuk VR (1999) dichlorophenolindophenol (DCIP), modified DCIP,
Influence of hemoglobin E on the severity of and hemoglobin E tests to screen for the HbE trait
Falciparum malaria. J Infect Dis, 179, 283–286. in pregnant women. Int J Gynaecol Obstet, 107,
45 Fairbanks VF, Gilchrist GS, Brimhall B, Jereb JA and 59–60.
Goldston EC (1979) Hemoglobin E trait reexam 57 Wong SC and Ali MAM (1982) Hemoglobin E dis
ined: a cause of microcytosis and erythrocytosis. ease. Am J Hematol, 13, 15–21.
Blood, 53, 109–115. 58 Jayasree D, Shaji RV, George B, Mathews V,
46 Sanchaisuriya K, Fucharoen G, Sae‐ung N, Sae‐ue Srivastava A and Edison ES (2016) Clinical, hemato
N, Baisungneon R, Jetsrisuparb A and Fucharoen S logical and molecular analysis of homozygous Hb
(1997) Molecular and hematological characteriza E (HBB: c.79G > A) in the Indian population.
tion of HbE heterozygotes with α‐thalassemia Hemoglobin, 40, 16–19.
determinant. Southeast Asian J Trop Med Public 59 Pornprasert S, Thichak S, Kongthai K and
Health, 28, Suppl 3, 100–103. Wangchauy C (2018) Comparison of HbA2, E, F and
47 Keren DF, Hedstrom D, Gulbranson R, Ou CN red cell parameters in homozygous HbE with and
and Bak R (2008) Comparison of Sebia Capillarys without α0‐thalassemia trait. Lab Med, 49, 118–122.
320 Chapter 5
60 Pakdee N, Yamsri S, Fucharoen G, Sanchaisuriya K, 72 Italia K, Dabke P, Sawant P, Nadkarni A, Ghosh K

Pissard S and Fucharoen S (2014) Variability of and Colah RB (2016) Hb E‐β‐thalassemia in five
hemoglobin F expression in hemoglobin EE dis Indian states. Hemoglobin, 40, 310–315.
ease: hematological and molecular analysis. Blood 73 Italia KY, Jijina FF, Merchant R, Panjwani S, Nadkami
Cells Mol Dis, 53, 11–15. AH, Sawant PM et al. (2010) Effect of hydroxyurea
61 Wickramasinghe SN, Hughes M, Wasi P, Fucharoen on the transfusion requirements in patients with
S and Litwinczuk R (1984) Ultrastructure and cell severe HbE‐β‐thalassemia: a genotypic and pheno
cycle distribution of erythropoietic cells in hete typic study. Acta Haematol, 63, 147–150.
rozygotes and homozygotes for haemoglobin E. Br 74 Yamsri S, Pakdee N, Fucharoen G, Sanchaisuriya K
J Haematol, 57, 685–694. and Fucharoen S (2016) Molecular understanding
62 Mehta BC, Agarawal MB, Varandani DG, Joshi RH of non‐transfusion‐dependent thalassemia associ
and Bhargava AB (1980) Hemoglobin E‐thalassae ated with hemoglobin E‐β‐thalassemia in Northeast
mia: a study of 16 cases. Acta Haematol, 64, 201–204. Thailand. Acta Haematol, 136, 233–239.
63 Weatherall D and Clegg JB. The Thalassaemia 75 Fucharoen S, Siritanaratkul N, Winichagoon P,
Syndromes. Blackwell Science, Oxford, 1981. Chowthaworn J, Siriboon W. Muangsup W et al.
64 Winichagoon P, Thonglairoam V, Fucharoen S, (1996) Hydroxyurea increases hemoglobin F levels
Wilairat P, Fukumaki Y and Wasi P (1993) Severity and improves the effectiveness of erythropoiesis in
differences in β‐thalassaemia/haemoglobin E syn β‐thalassaemia/haemoglobin E disease. Blood, 87,
dromes: implication of genetic factors. Br J Haematol, 887–892.
83, 633–639. 76 Wickramasinghe SN, Hughes M, Wasi P, Fucharoen
65 Issaragrisil S, Siritanaratkul N and Fucharoen S. S and Modell B (1981) Ineffective erythropoiesis in
Diagnosis and Management of Thalassemia: Thailand as haemoglobinEβ ‐thalassaemia: an electron micro
a Model. American Society of Hematology, scope study. Br J Haematol, 48, 451–457.
Education Program Book, American Society of 77 Pornprasert S, Tookjai M, Punyamung M and
Hematology, Washington DC, 2001, pp. 483–488. Koggthai K (2019) A formula to identify potential
66 Sharma V, Kumar B, Kumar G and Saxena R (2009) cases of β‐thalassemia/HbE disease among patients
Alpha globin gene numbers: an important modifier with absent HbA, HbE >75% and HbF between
of HbE/beta thalassemia. Hematology, 14, 297–300. 5 and 15%. Lab Med, 50, 158–162.
67 Allen A, Fisher C, Premawardhena A, Bandara D, 78 Allen A, Fisher C, Premawardhena A, Peto T, Allen
Perera A, Allen S et al. (2012) Methemoglobinemia S, Arambepola M, Thayalsutha V et al. (2010)
and ascorbate deficiency in hemoglobin E β thalas Adaptation to anemia in hemoglobin E‐ß thalas
semia: metabolic and clinical implications. Blood, semia. Blood, 116, 5368–5370.
120, 2939–2944. 79 Rees DC, Porter JB, Clegg JB and Weatherall DJ
68 Ekwattanakit S, Siritanaratkul N and Viprakasit V (1999) Why are haemoglobin F levels increased in
(2018) A prospective analysis for prevalence of HbE/β thalassaemia. Blood, 94, 3199–3204.
complications in Thai nontransfusion‐dependent 80 Prasing W and Pornprasert S (2014) Measurement
Hb E/β‐thalassemia and α‐thalassemia (Hb H dis of HbA2 by capillary electrophoresis for diagnosing
ease). Am J Hematol, 93, 623–629. β‐thalassemia/HbE disease in patients with low
69 Krittayaphong R, Viprakasit V, Saiviroonporn P, HbF. Lab Med, 45, 226–230.
Wangworatrakul W and Wood JC (2018) Serum fer 81 Singha K, Fucharoen G, Sanchaisuriya K and
ritin in the diagnosis of cardiac and liver iron over Fucharoen S (2018) EE score: an index for simple
load in thalassaemia patients real‐world practice: a differentiation of homozygous hemoglobin E and
multicentre study. Br J Haematol, 182, 301–305. hemoglobin E‐β0‐thalassemia. Clin Chem Lab Med,
70 Rees DC, Styles L, Vichinsky EP, Clegg JB and 56, 1507–1513.
Weatherall DJ (1998) The hemoglobin E syndromes. 82 Arong A, Wangwok G, Singha K, Sanchaisuriya K,
Ann NY Acad Sci, 850, 334–343. Fucharoen G and Fucharoen S (2019) Differentiation of
71 Fouladi M, Macmillan ML, Nisbet‐Brown E, Klein homozygous hemoglobin E and hemoglobin E‐β0 ‐
N, Barlas J, Waye JS and Olivieri NF (1998) thalassemia in children. Int J Lab Hematol, 41, e70–e72.
Hemoglobin E/β thalassemia: the Canadian experi 83 Fucharoen S, Winichagoon P, Siritanaratkul N,
ence. Ann NY Acad Sci, 850, 410–411. Chowthaworn J and Pootrakul P (1998) α‐ and
β‐ thalassemia in Thailand. Ann NY Acad Sci, 850, Punjab. Interaction with α thalassaemia and
412–414. diagnosis by gene mapping. Scand J Haematol, 32,
84 Boonsa S, Sanchaisuriya K, Fucharoen G, Wiangnon 275–282.
S, Jetrsrisuparb A and Fucharoen S (2004) The 97 Stout C, Holland CK and Bird RM (1964)
diverse molecular basis and haematological fea Hemoglobin D in an Oklahoma family. Arch Intern
tures of Hb H and AEBart’s disease in Northeast Med, 114, 296–300.
Thailand. Acta Haematologica, 111, 149–154. 98 Dutta RN, Grover J and Lai M (1972) Haemoglobin
85 Bain BJ (2009) Hypochromic microcytic anemia D disease. J Indian Med Assoc, 58, 42–47.
with a variant haemoglobin. Am J Hematol, 84, 59. 99 Marengo‐Rowe AJ, McCraken AW and Flanagan
86 Hoyer JD, Wick MJ, Thibodeau SN, Viker KA, P (1968) Complete suppression of haemoglobin A
Conner R and Fairbanks VF (1998) Hb Tak con synthesis in haemoglobin D Los Angeles‐beta tha
firmed by DNA analysis: not expressed as thalas lassaemia. J Clin Pathol, 21, 508–510.
semia in a Hb Tak/Hb E compound heterozygote. 100 Schneider RG, Ueda S, Alperin JB, Levin WC,
Hemoglobin, 22, 45–52. Jones RT and Brimhall B (1968) Hemoglobin D Los
87 Sae‐ung N, Fucharoen G, Sanchaisuriya K and Angeles in two Caucasian families: hemoglobin
Fucharoen S (2007) α0‐ thalassemia and related dis SD disease and hemoglobin D thalassemia. Blood,
orders in northeast Thailand: a molecular and 32, 250–259.
hematological characterization. Acta Haematol, 117, 101 Das S and Mashon RS (2015) Coinheritance of Hb
78–82. D‐Punjab and β‐thalassemia: diagnosis and impli
88 Viprakasit V and Tanphaichitr VS (2003) An unu cations in prenatal diagnosis. Hemoglobin, 39,
sual phenotype for homozygous HbE with a single 138–140.
α globin gene: a pitfall in clinical diagnosis and 102 Knox‐Macauley HMH, Fleming AF and Lamb J
genetic counselling. Blood, 102, 36b. (1984) Haemoglobin GPhiladelphia and its inter
89 Leckngam P, Limweeraprajak E, Kiewkarnkha T action with haemoglobin S and alpha‐thalassae
and Tatu T (2017) The Hb E (HBB: c.79G>A), mean mia in Nigerians. Clin Lab Haematol, 6, 113–121.
corpuscular volume, mean corpuscular hemo 103 Rothman MC and Ranney HM (1971) Double het
globin cutoff points in double heterozygous Hb erozygote for hemoglobin G (α268Lysβ2A) and hemo
E/− −SEA α‐thalassemia‐1 carriers are dependent on globin D (α2Aβ2121Gln). Blood, 37, 177–183.
hemoglobin levels. Hemoglobin, 41, 38–43. 104 Nagel RL, Krishnamoorthy R. Fattoum S, Elion J,
90 Edison ES, Shaji RV, Chandy M and Srivastava A Genard N, Romero J and Fabry ME (1999) The
(2011) Interaction of hemoglobin E with other abnor erythrocyte effects of haemoglobin OARAB. Br J
mal hemoglobins. Acta Haematol, 126, 246–248. Haematol, 107, 516–521.
91 Wong SC and Ali MAM (1980) Haemoglobin D Los 105 Papadopoulos V, Dermitzakis E, Konstantinidou
Angeles, D‐β+‐thalassaemia, and D‐β0‐thalassaemia. D, Petridis D, Xanthopoulidis G and Loukopoulos
A report of two Canadian families. Acta Haematol, D (2005) HbO‐Arab mutation originated in the
63, 151–155. Pomak population of Greek Thrace. Haematologica,
92 Chernoff AI (1956) On the prevalence of hemoglobin 90, 255–257.
D in the American Negro. Blood, 11, 907–909. 106 Joutovsky A and Nardi M (2004) Hemoglobin
93 Giambona A, Vinciguerra M, Cannata M, Cassarà F, C and hemoglobin O‐Arab variants can be
Fiorentino G, Leto F et al. (2011) The genetic hetero diagnosed using the Bio‐Rad Variant II high‐

geneity of β‐globin gene defects in Sicily reflects the performance liquid chromatography system
historic population migrations of the island. Blood without further confirmatory tests. Arch Pathol Lab
Cells Mol Dis, 46, 282–287. Med, 128, 435–439.
94 Hoyer JD and Kroft SH, eds. Color Atlas of 107 Heard SE, Westwood NB, Pearson TC and Stephens
Hemoglobin Disorders. College of American AD (1991) Homozygous haemoglobin O‐Arab in
Pathologists, Northfield, 2003. pregnancy. Clin Lab Haematol, 13, 319–320.
95 Chernoff AI (1958) The hemoglobin D syndromes. 108 Ibrahim SA, Mustafa D, Mohamed AO and
Blood, 13, 116–127. Mohed MB (1992) Homozygous haemoglobin O
96 Trent RJ, Harris MG, Fleming PJ, Wyatt K, Hughes disease and conjugated hyperbilirubinaemia in a
WG and Kronenberg H (1984) Haemoglobin D Sudanese family. Br Med J, 304, 26–27.
322 Chapter 5
109 Kendall AG and Barr RD (1973) Haemoglo tandem duplication of β90‐94 inserted between
binopathies in Kenya. Trans Soc Trop Med Hyg, 67, B94‐B95 (proved by DNA sequence analysis); a
770–781. previously undescribed unstable hemoglobin
110 Dror S (2013) Clinical and hematological features variant associated with severe hemolytic anemia
of homozygous hemoglobin O‐Arab [Beta 121 Glu and stroke during early childhood. Blood, 90,
→ Lys]. Paediatr Blood Cancer, 60, 506–507. Suppl 1, 27b.
111 Lee‐Potter JP, Deacon‐Smith RA, Simkiss MJ, 121 Nagel RL. Disorders of hemoglobin function and
Kamuzora H and Lehmann H (1975) A new cause stability. In: Steinberg MH, Forget BG, Higgs DR
of haemolytic anaemia in the newborn: a descrip and Nagel RL, eds. Disorders of Hemoglobin:
tion of an unstable fetal haemoglobin: F Poole, Genetics, Pathophysiology, and Clinical Management.
α2Gγ2130 tryptophan → glycine. J Clin Pathol, 28, Cambridge University Press, Cambridge, 2001,
317–320. pp. 1155–1194.
112 Van den Driessch M, Moerman J, Moens M, Van 122 Srivastava P, Kaeda JS, Roper D, Vulliamy TJ,
Eldere S, Derclaye I and Philippe M (2005) Severe Buckley M and Luzzatto L (1995) Severe hemo
hereditary haemolytic anaemia in a Caucasian lytic anemia associated with the homozygous
newborn: a new fetal haemoglobin variant Hb F‐ state for an unstable haemoglobin variant (Hb
Bonheiden (Gγ 38(C4) Thr→Pro), Eur J Pediatr, Bushwick). Blood, 86, 1977–1982.
164, 261–262. 123 Koren A, Levin C, Zalman L, Palmor H, Filon D,
113 Huisman THJ (1997) Gamma chain abnormal Chubar E et al. (2016) Hb TAYBE: clinical and mor
human fetal hemoglobin variants. Am J Hematol, phological findings in 43 patients. Eur J Haematol,
55, 159–163. 97, 137–144.
114 Brugnara C and Platt OS, The neonatal erythro 124 Arnon S, Tamary H, Dgany O, Litmanovitz I,
cyte and its disorders. In: Nathan DG, Orkin SH, Regev R, Bauer S et al. (2016) Hydrops fetalis asso
Ginsburg D and Look AT, eds. Nathan and Oski’s ciated with homozygosity for hemoglobin Taybe
Hematology of Infancy and Childhood, 6th edn. (a 38/39 THR deletion) in newborn triplets. Am J
Saunders, Philadelphia, 2003, pp. 19–55. Hematol, 76, 263–266.
115 Jiang H, Zhou JY, Li J and Li DZ (2019) Unstable 125 White JM and Brain MC (1970) Defective synthe
hemoglobin variants: the need for clinical vigi sis of an unstable haemoglobin: haemoglobin
lance in infants with congenital jaundice. Koln (β98Val–Met). Br J Haematol, 18, 195–209.
Hemoglobin, 43, 60–62. 126 Traeger‐Synodinos J, Papassotiriou I, Mataxotou‐
116 Casey P, Kynoch PAM, Lang A, Lehmann H, Mavrommati A, Vrettou C, Stamoulakatou A and
Nozari G and Shinton HK (1978) Double heterozy Kanavakis E (2000) Distinct phenotypic expres
gosity for two unstable haemoglobins: Hb Sydney sion associated with a new hyperunstable alpha
(β67 [E11] Val→ Ala) and Hb Coventry (β141 [H19] globin variant (Hb Heraklion,α1cd37(C2)Pro>0):
Leu deleted). Br J Haematol, 38, 195–209. comparison to other α‐thalassemic hemoglobi
117 Brennan SO, Shaw J, Allen J and George PM (1992) nopathies. Blood Cells Mol Dis, 26, 276–284.
β141 Leu is not deleted in the unstable haemoglo 127 Hayakawa I, Hataya H and Kaneko T (2016)
bin Atlanta‐Coventry but is replaced by a novel Choledocholithiases in a child with hemoglobin
amino acid of mass 128 daltons. Br J Haematol, Evans [alpha2 62(E11) Val→Met]. Clin Case Rep, 4,
81, 99–103. 661–663.
118 Rieder RF and Bradley TB (1968) Hemoglobin 128 de Furia FG and Miller DR (1972) Oxygen affin
Gun Hill: an unstable protein associated with ity in haemoglobin Köln disease. Blood, 39,
chronic hemolysis. Blood, 32, 355–369. 398–406.
119 Wajcman H, Dahmane M, Préhu C, Costes B, 129 Phyliky RL and Fairbanks VF (1997) Throm
Promé D, Arous N et al. (1998) Haemoglobin boembolic complication of splenectomy in unsta
J‐Biakra: a new mildly unstable α1 gene variant ble hemoglobin disorders: Hb Olmsted, Hb Köln.
with a deletion of eight residues (α50‐57,α51‐58 or Am J Hematol, 55, 53.
α52‐59) including the distal histidine. Br J 130 Frisch B, Lewis SM, Sherman D, White JM and
Haematol, 100, 401–406. Gordon‐Smith EC (1974) The ultrastructure of
120 Greenberg J, Fairbanks VF, Hoyer J, Thibodeau SN haemopoiesis in two haemoglobinopathies. Br J
and Fairfax VA (1998) Hemoglobin Fairfax, a Haematol, 28, 109–117.
131 Hohl RJ, Sherburne AR, Feely JE, Huisman THJ 144 Papassotiriou I, Traeger‐Synodinos J, Marden MC,
and Burns CP (1998) Low pulse oximeter‐meas Kister J, Liapi D, Prome D et al. (2005) The homozy
ured hemoglobin oxygen saturations with hemo gous state for Hb Crete [β129 (H7) Ala→Pro] is
globin Cheverly. Am J Hematol, 59, 181–184. associated with a complex phenotype including
132 Hurford WE and Kratz A (2003) Case 23‐2004: a erythrocytosis and functional anemia. Blood Cells
50‐year‐old woman with low oxygen saturation. Mol Dis, 34, 229–234.
N Engl J Med, 351, 380–387. 145 Charache S, Weatherall DJ and Clegg JB (1966)
133 Fay KC, Brennan SO, Costello JM, Potter HC, Polycythemia associated with a hemoglobinopa
Williamson DA, Trent RJ et al. (1993) Haemoglobin thy. J Clin Invest, 45, 813–822.
Manukau β67[E11] Val→Gly: transfusion‐depend 146 Bagby GC, Richert‐Boe K and Koler RD (1978) 32P
ent haemolytic anaemia ameliorated by coexisting and acute leukemia: development of leukemia in a
alpha thalassaemia. Br J Haematol, 85, 352–355. patient with hemoglobin Yakima. Blood, 52,
134 Lie‐Injo LE, Ganesan J, Randhawa ZI, Peterson D 350–354.
and Kane JP (1977) Hb Leiden‐beta (0) thalassemia 147 Oliveira JL, Coon LM, Frederick LA, Hein M,
in a Chinese with severe hemolytic anemia. Am J Swanson KC, Savedra ME et al. (2018) Genotype‐
Hematol, 2, 335–342. phenotype correlation of hereditary erythrocyto
135 Horlein H and Weber G (1948) Uber chronisch sis mutations, a single center experience. Am J
familiare methamoglobinanue and eine neue Hematol, 93, 1029–1041.
modification des Methamoglobins Dtsch Med 148 Rochette J, Barnetson R, Kiger L, Kister J,
Wochenschr, 73, 476, cited by Nagel, 2001 [121]. Littlewood TJ, Webster R et al. (1994) Association of
136 Sherwood LM and Parris EE (1970) Clinically novel high oxygen affinity haemoglobin variant
important variants of human hemoglobin. N Engl with δβ thalassaemia. Br J Haematol, 86, 118–124.
J Med, 282, 142–152. 149 Mais DD, Boxer LA, Gulbranson RD and Keren
137 Glader BE, Zwerdling D, Kutlar F, Kutlar A, DF (2007) Hemoglobin Ypsilanti: a high‐oxygen‐
Wilson JB and Huisman TH (1989) Hb F‐M‐Osaka affinity hemoglobin demonstrated by two auto
or alpha 2G gamma 2(63)(E7)His‐‐‐‐Tyr in a mated high‐pressure liquid chromatography
Caucasian male infant. Hemoglobin, 13, 769–773. systems. Am J Clin Pathol, 128, 850–853.
138 Molchanova TP, Wilson JB, Gu LH, Hain RD, 150 Yudin J and Verhovsek M (2019) How we diag
Chang LS, Poon AO and Huisman TH (1992) A nose and manage altered oxygen affinity hemo
second observation of the fetal methemoglobin globin variants. Am J Hematol, 94, 597–603.
variant Hb F‐M‐Fort Ripley or alpha 2G gamma 151 Beard MEJ, Potter HC, Spearing RL and Brennan
2(92)(F8)His‐‐‐‐Tyr. Hemoglobin, 16, 389–398. SO (2001) Haemoglobin Pierre‐Benite – a high
139 Dainer E, Shell R, Miller R, Atkin JF, Pastore M, affinity variant associated with relative poly
Kutlar A et al. (2008) Neonatal cyanosis due cythaemia. Clin Lab Haematol, 23, 407–409.
to a novel fetal hemoglobin: Hb F‐Circleville 152 Stephens A (1977) Polycythaemia and high
[Ggamma63(E7)His‐‐>Leu, CAT>CTT]. Hemoglobin, affinity haemoglobins. Br J Haematol, 36,
32, 596–600. 153–159.
140 Bento C, Magalhães Maia T, Carvalhais I, Moita F, 153 Williamson D, Langdown JV, Myles T, Mason C,
Abreu G, Relvas L et al. (2013) Transient neonatal Henthorn J and Davies SC (1992) Polycythaemia
cyanosis associated with a new Hb F variant: Hb F and microcytosis arising from the combination of
Viseu. J Pediatr Hematol Oncol, 35, e77–e80. a new high oxygen affinity haemoglobin (Hb
141 Marks A, Luo H‐y, Chui DHK and Greenberg J Luton α89 His→Leu) and α thalassaemia trait. Br
(2013) Hemoglobin Shady Grove: a novel fetal J Haematol, 82, 621–622.
methemoglobin variant, Pediatr Blood Cancer, 60, 154 David O, Ivaldi G, Rabino‐Massa E and Ricco G
E55–E66. (2002) Functional studies on nine different haemo
142 Wild BJ and Bain BJ (2004) Detection and quantifi globins with high oxygen affinity. Acta Haematol,
cation of normal and variant haemoglobins: an 108, 132–138.
analytical review. Ann Clin Biochem, 41, 355–369. 155 Prakobkaew N, Singsanan S, Fucharoen G, Surapot
143 Janssen WJ, Collard HR and Dhaliwal G (2005) S and Fucharoen S (2010) Secondary erythrocytosis
Why “why” matters. N Engl J Med, 352, caused by hemoglobin Tak/(δβ)0‐thalassemia syn
1270–1271. drome. Acta Haematol, 124, 115–119.
324 Chapter 5
156 Ropero P, Villegas A, Gonzalez AF, Anguita E, 163 Ostrander RJ and Green NC (2011) Neonatal cya
Sanchez J, Carreno DL et al. (2000) Hb Johnstown nosis from a haemoglobin variant. N Engl J Med,
[beta 109 (G11) Val‐‐>Leu]: second case described 365, 378.
and associated for the first time with beta0‐ 164 Luo H‐y, Irving A, Prior J, Lim E, Eung SH, Skelton
thalassemia in two Spanish families. Am J Hematol, TP et al. (2004) Hemoglobin Titusville, a low oxy
65, 298–301. gen affinity variant hemoglobin, in a family of
157 Rich S and Hart K (1991) Familial pulmonary Northern European background. Am J Hematol, 77,
hypertension in association with an abnormal
384–386.
hemoglobin. Insights into the pathogenesis of
165 Abdulmalik O, Safo MK, Lerner NB, Ochotorena J,
primary pulmonary hypertension. Chest, 99,
Daikhin E, Lakka V et al. (2004) Characterization of
1208–1210. hemoglobin Bassett (α94Asp→Ala), a variant with
158 Wille RT, Krishnan K, Cooney KA, Bach DS and very low oxygen affinity. Am J Hematol, 77, 268–276.
Martinez F (1996) Familial association of primary 166 Lacan P, Souillet G, Aubry M, Prome D, Richelme‐
pulmonary hypertension and a new low‐oxygen David S, Kister J et al. (2002) New alpha 2 globin
affinity beta‐chain hemoglobinopathy, Hb chain variant with low oxygen affinity affecting the
Washtenaw. Chest, 109, 848–850. N‐terminal residue and leading to N‐acetylation
159 Wu Y, Ramani GV, Gai Q, Lemon LC and Baer MR [Hb Lyon‐Bron α 1(NA1)Val ‐‐> Ac‐Ala]. Am J
(2010) Rare hemoglobinopathy presenting as pro Hematol, 69, 214–218.
gressive dyspnea. Am J Hematol, 85, 355–357. 167 Steinberg MH and Adams JG (1991) Haemoglobin A2:
160 Crowley MA, Mollan TL, Abdulmalik OY, Butler origin, evolution and aftermath. Blood, 78, 2165–2177.
AD, Goodwin EF, Sarkar A et al. (2011) A hemo 168 Vella F (1977) Variation in hemoglobin A2.
globin variant associated with neonatal cyanosis Hemoglobin, 1, 619–650.
and anemia. N Engl J Med, 364, 1837–1843. 169 Spurdle AB, Krause A, Ramsay M and Jenkins T
161 Kohli‐Kumar M, Zwerdling T and Rucknagel DL (1994) The high frequency of the Hb B2 variant
(1995) Hemoglobin F‐Cincinnati, α2Gγ2 41(C7) in the Herero population: a founder effect?
Phe→Ser in a newborn with cyanosis. Am J Hemoglobin, 18, 317–323.
Hematol, 49, 43–47. 170 Bennani M, Mombo L‐E, Chaventre A, Barakat A,
162 Mina M, James R and Gandhi S (2018) Haemoglobin Ducrocq R, Nagel RL and Krishnamoorthy R
Titusville: a case study and review of the literature. (2003) Origin of Hb A2′ (Hb B2) [δ16(A13)
J Paediatr Child Health, 54, 449–452. Gly→Arg(GGC→CGC)]. Hemoglobin, 27, 105–110.
5.1 (a) T 5.4 (a) T 5.7 (a) T 5.10 (a) T 5.13 (a) T
(b) T (b) T (b) T (b) T (b) F
(c) F (c) T (c) F (c) F (c) T
(d) T (d) T (d) T (d) F (d) T
(e) F (e) F (e) F (e) T (e) T
5.2 (a) T 5.5 (a) F 5.8 (a) T 5.11 (a) T

(b) T (b) F (b) T (b) T
(c) T (c) T (c) T (c) T
(d) T (d) F (d) T (d) T
(e) F (e) F (e) T (e) T
5.3 (a) F 5.6 (a) F 5.9 (a) T 5.12 (a) F

(b) T (b) F (b) F (b) F
(c) T (c) F (c) F (c) F
(d) F (d) T (d) F (d) T
(e) T (e) T (e) T (e) T
6 Acquired abnormalities of globin chain synthesis or
haemoglobin structure
Acquired disorders of globin chain synthesis can also been reported. It should be noted that cases
result from: (i) mutation or deletion of a globin described as acquired β or δβ thalassaemia are not
gene; (ii) altered methylation status of a globin gene as well characterised as acquired haemoglobin H
leading to altered expression; or (iii) the influence of disease and it is less clear that this is a distinct entity.
other genes on the expression of globin genes. The molecular mechanism – deletion of the α gene
Acquired somatic mutations of globin genes are cluster from one chromosome 16 – was identified in
very rare. Altered expression is much more a single patient with acquired α thalassaemia trait
common. as a feature of an MDS [2].
Post‐translational modification of haemoglobin More than 80 cases of acquired haemoglobin H
structure can also occur as a result of inherited disease have been described [7–21]. This syndrome
abnormalities, other than those of globin genes, or has been associated mainly with MDS but also with
as a result of exposure to toxic drugs or chemicals. AML, primary myelofibrosis, atypical chronic mye
loid leukaemia and various difficult‐to‐classify
myelodysplastic/myeloproliferative neoplasms. It
Acquired thalassaemia
has been described in a 10‐year‐old boy with
Alterations in the rates of globin chain synthesis, Down’s syndrome and it was postulated that this
with α:β ratios similar to those observed in thalas represented the onset of MDS [22]. The type of MDS
saemia, are quite common in myeloid malignancies. most strongly associated with acquired haemoglo
This could be regarded as a mild form of acquired bin H disease is refractory anaemia with ring side
thalassaemia. Peters et al. [1] observed an increased roblasts (Fig. 6.1) (myelodysplastic syndrome with
α:β ratio in six of 11 patients with myelodysplastic ring sideroblasts and unilineage dysplasia in the
syndromes (MDS) (α:β chain synthesis ratios of 2016 World Health Organization classification) but
1.28–2.43, normal range 0.97–1.19) and in two of some cases have had refractory anaemia or refrac
four patients with acute myeloid leukaemia (AML) tory anaemia with excess of blasts. Cases of AML
(both evolved from MDS, α:β chain synthesis ratios have typically been erythroleukaemia (Fig. 6.2),
of 1.8 and 2.1). In this study, red cell hypochromia sometimes with sideroblastic erythropoiesis.
and microcytosis were not confined to patients with Atypical myelodysplastic/myeloproliferative neo
abnormalities in the α:β chain synthesis ratio. plasms have included several cases of refractory
The phenotype of α or β thalassaemia is much less anaemia, with or without ring sideroblasts, with
common among cases of leukaemia, MDS and coexisting thrombocytosis.
related disorders than an alteration in the ratio of α One case of acquired haemoglobin H disease has
and β globin chain synthesis. When acquired thalas been reported which was associated with what
saemia occurs, the phenotype is most often that of appeared to be acute lymphoblastic leukaemia [17],
haemoglobin H disease, although acquired α thalas an unexpected association. Although no immu
saemia trait [2] and β or δβ thalassaemia [3–6] have nophenotyping was available, there was strong

325
326 Chapter 6
Fig. 6.1 Blood film of a

60‐year‐old man with acquired
haemoglobin H disease
associated with refractory
(a)
anaemia with ring sideroblasts.
The full blood count (FBC)
showed white cell count (WBC)
9.2 × 109/l, haemoglobin
concentration (Hb) 100 g/l, mean
cell volume (MCV) 66 fl and
platelet count 53 × 109/l. The
haemoglobin H percentage was
9% and the α:β chain synthesis
was 0.16: (a) dimorphic red cells
with one of the hypochromic
cells containing several
Pappenheimer bodies;
(b) dimorphic red cells with
normocytic normochromic cells
and microcytic target cells. (With
thanks to Dr A. Hendrick.)
May–Grünwald–Giemsa (MGG)
(b) ×100 objective.
circumstantial evidence of lymphoid lineage: the intact and show a normal pattern of methylation, α
cells showed periodic acid–Schiff (PAS)‐block posi chain messenger ribonucleic acid (mRNA) is greatly
tivity and were negative for Sudan black B and reduced. In many patients the defect in α chain syn
myeloperoxidase and the disease remitted on two thesis is so severe that it is clear that all four α genes
occasions with vincristine and prednisone. The dis are downregulated. Interestingly, one patient with
appearance of the haemoglobin H with disease acquired haemoglobin H disease had five α genes,
remission, on two occasions, provided evidence as a result of a constitutional triple α, and despite
that the haemoglobin H‐containing red cells and this had an α:β globin chain synthesis ratio of 0.09
the leukaemic blasts belonged to the same clone. If [18]. In this patient all five α genes must have been
the leukaemia was indeed lymphoid, a leukaemo downregulated. Clearly a trans‐acting factor was
genic mutation in a pluripotent lymphoid–myeloid implicated. This was supported by the observation
stem cell could be postulated. that the ratio of α2 and α1 gene transcripts was nor
It has been demonstrated that in acquired haemo mal [19]. The great majority of patients with
globin H disease, although all four α genes are acquired haemoglobin H disease, more than 90%,
Acquired abnormalities of globin chain synthesis 327
(a)
Fig. 6.2 Acquired haemoglobin

H disease in a patient with
erythroleukaemia: (a) blood film,
MGG ×100, and (b) haemoglobin
H preparation. Haemoglobin H
was 15% of total haemoglobin. (b)
have been male. The explanation of all these obser disease with sideroblastic erythropoiesis; (ii) the pos
vations was revealed when it was demonstrated sibility that there is coexistence of abnormal clonal
that the cause of acquired haemoglobin H disease is cells, showing a defect in α chain synthesis, and sur
an acquired somatic mutation of the ATRX gene [21, viving normal cells with normal haemoglobin synthe
23, 24]. A unique case involved chromosome 16 sis; and (iii) the possibility that the erythroid precursors
rather than the ATRX gene on the X chromosome. with a defect in synthesis of α chain and haemoglobin
Acquired haemoglobin H disease is associated with represent a subclone of the neoplastic population. In
variable anaemia and reticulocytosis. Microcytosis is occasional cases the α chain synthetic defect leading to
common but not invariable, whereas inherited hae haemoglobin H synthesis has either appeared during
moglobin H disease is invariably associated with the course of the illness [11] or has decreased with dis
microcytosis. The red cells usually show marked ease progression [18], supporting the third possibility.
anisocytosis, poikilocytosis and hypochromia. The The percentage of haemoglobin H varies considerably
blood film is often dimorphic, a feature that is not seen between cases, from a few per cent to as high as
in the inherited form of the disease. There are three 60–65%. Cells containing haemoglobin H inclusions
possible explanations for the dimorphism: (i) the vary from a few per cent to more than 90%. Cases with
frequency of association of acquired haemoglobin H an ATRX mutation have more severe haemoglobin H
328 Chapter 6
disease than is seen in the inherited ATRX syndrome, Acquired β, δβ and γδβ thalassaemias have been
even when the mutation is the same, with a median of described in MDS [4, 5], AML (erythroleukaemia)
15% haemoglobin H and a median of 30% of cells hav [3], juvenile myelomonocytic leukaemia [6] and
ing H inclusions [21, 24, 25]. Haemoglobin H levels are common variable immunodeficiency [27]. The
also often higher than in inherited haemoglobin H dis phenotype has usually resembled the heterozy
ease and the α:β chain synthesis ratio is often much gous state for these thalassaemic disorders [3–5,
lower, varying from 0.05 to 0.66. Some cases have 27]. However, a child with juvenile myelomono
traces of haemoglobin Bart’s. When haemoglobin H cytic leukaemia has been described with a
constitutes a large percentage of total haemoglobin, phenotype resembling β (or more correctly δβ)
the oxygen dissociation curve becomes very abnormal thalassaemia major [6]. Although it is common for
and, because of the hyperbolic dissociation curve of children with this type of leukaemia to have an
haemoglobin H, symptoms are much more severe increased synthesis of haemoglobin F with rever
than would be expected for the degree of anaemia. sion to the usual fetal Gγ:Aγ ratio, this child dif
The percentage of haemoglobin A2 is reduced. fered in also having marked microcytosis and
Uncommonly, haemoglobin F is increased [26]. very unbalanced chain synthesis. The phenotype
At least seven patients have been reported with of some cases of acquired β, δβ and γδβ thalassae
acquired thalassaemic conditions resembling β tha mia has differed from that of the inherited disor
lassaemia (Table 6.1) [3–6, 27, 28]. Two cases with a ders. Two patients had a significant macrocytosis
low‐normal or reduced haemoglobin A2 percentage rather than microcytosis [3, 4] and one had marked
and an increased haemoglobin F percentage were reticulocytosis [5].
reported as ‘acquired δβ thalassaemia’ [3, 4]. The The leukaemias and MDS can be viewed as
remaining cases were reported as ‘acquired β tha acquired genetic disorders affecting a haemopoietic
lassaemia’ [5, 6, 27, 28], although none of them had stem cell. It is therefore not surprising that there can
an increased haemoglobin A2 percentage. Two of be mutations affecting globin chain synthesis.
these cases had a significantly increased haemoglo However, although the relevant mRNA is greatly
bin F percentage [5, 6] and are therefore better reduced, mutations in globin genes have rarely
regarded as acquired δβ thalassaemia. The other been detected [5, 6, 19, 25].
two had microcytosis and unbalanced chain syn Two cases of acquired β thalassaemia and one of
thesis, but with no increase in either haemoglobin acquired α thalassaemia have been reported in
A2 or haemoglobin F [27]; a further patient had patients with common variable immune deficiency
deletion of the β gene cluster [28] (and could [27], another unexpected association. In one patient
therefore be regarded as having acquired γδβ
there was significant anaemia and microcytosis
thalassaemia). with normal haemoglobin electrophoresis and an
Table 6.1 Acquired β, δβ and γδβ thalassaemia. (From references 3–6, 27, 28.)
Sex/age (years) Hb (g/l) MCV (fl) F (%) A2 (%) α:β* α:β+γ* G

γ:Aγ Disease Reference
M, 27 89 ↑ 39 0.85 2.2 1.2–1.4 99:1 M6 AML† 3

M, 73 65 107 16 1.9 1.85 1.18 0.85:1 RARS 4
M, 26 103 80 3 1.3 1.7 1.7 NS RAEB 5
M, 30 78 67 <1 2.4 1.4 NS NS CVID 27
M, 22 138 75 <1 2.9 1.15 NS NS CVID 27
F, 3 105 67 77 0.4 2.24 0.45 2.2:1 JMML 6
F, 65 97 66 3.1 2.4 MDS 28
AML, acute myeloid leukaemia; CVID, common variable immunodeficiency; JMML, juvenile myelomonocytic leukaemia; MDS,
myelodysplastic syndrome; RAEB, refractory anaemia with excess of blasts; RARS, refractory anaemia with ring sideroblasts.
* Globin chain synthesis ratios.
† M6 AML (erythroleukaemia) according to the French–American–British classification with preceding aplastic anaemia.
α:β chain synthesis ratio of 0.76. X‐linked polymor thalassaemia and with unstable β chain variants.
phism analysis suggested clonal haemopoiesis and However, in acquired conditions increased synthesis
lymphopoiesis. It was postulated that there had is the only known mechanism for a higher haemoglobin F.
been damage to a common lymphoid–myeloid The main determinant of haemoglobin F in the neo
stem cell with subsequent clonal expansion. natal period is post‐conceptual age, so that premature
It is possible for a somatic mutation to interact babies have a higher proportion; the levels are, how
with a germline mutation, as occurred in a patient ever, appropriate for the gestational age of the neo
with heterozygosity for β0 thalassaemia who had nate. Post‐mature babies have a lower haemoglobin F
lost the second β allele in a proportion of cells, lead than babies born at term. Intrauterine hypoxia or
ing to the phenotype of β thalassaemia intermedia growth retardation leads to a higher haemoglobin F in
[29]. This mutation occurred during embryogene the neonate. On multiple regression analysis, neonatal
sis, rather than as a result of a haematological neo haemoglobin F level correlates, in addition, with male
plasm. However, a number of instances have been sex, twin births and maternal cigarette smoking [35].
reported of late onset β thalassaemia intermedia or An association with maternal cigarette smoking was
major as a result of acquired uniparental disomy in confirmed in a second study [36]. Maternal diabetes
patients who had previously manifest β thalassae mellitus is also associated with a higher level. It should
mia heterozygosity [30, 31]. be noted that although these conditions cause a higher
than normal percentage of haemoglobin F at birth,
and this is therefore correctly regarded as congenital,
Increased or decreased haemoglobin F
this is nevertheless ‘acquired’ in the sense that it is not
Increased synthesis of haemoglobin F is relatively genetic but is due to adverse conditions operating
common. Decreased synthesis, other than in γ tha during intrauterine life.
lassaemia, is rare (or at least is difficult to demon A study of sudden infant death syndrome sug
strate beyond the neonatal period, since the gested that haemoglobin F percentage was increased
percentage of haemoglobin F is normally low). A in comparison with levels seen in infants matched
reduced percentage has been reported in Down’s for post‐conceptual age [37], but this was not con
syndrome [32] and in a single neonate with a chro firmed in two subsequent studies [38, 39].
mosome group C/D translocation [33]. Decreased Acquired causes of an increased percentage of
synthesis of haemoglobin F in the fetal or neonatal haemoglobin F are shown in Table 6.2 [34–36,
period can be viewed as premature switching from 40–48]. (For inherited causes see Table 3.13.)
γ chain to β chain synthesis, since any deficit in hae A rise in haemoglobin F level is seen in 15–20% of
moglobin F synthesis is compensated for by haemo pregnant women, with levels up to 5% [34]. This is
globin A synthesis. related to an increasing red cell mass and a burst of
An alteration of haemoglobin F level can result not F cell production. An increased rate of synthesis of
only from an alteration in the rate of synthesis but haemoglobin F is also a feature of ‘stress erythro
also from longer or shorter survival of haemoglobin poiesis’ when there is rapid erythroid expansion.
F‐containing cells. For example, in the neonatal There is an associated increased in mean cell vol
period a reduced level of haemoglobin F can result ume (MCV) and increased expression of i antigen,
from a haemolytic anaemia that leads to rapid normally expressed at high levels only in the fetal
destruction of haemoglobin F‐containing cells, with and neonatal periods. The ratio of Gγ:Aγ synthesis in
new cells that are formed containing a higher propor stress erythropoiesis may be that characteristic of
tion of haemoglobin A. There is no alteration in the fetal life [42]. Stress erythropoiesis may be seen fol
relative rates of synthesis of β and γ globin chains [34]. lowing blood loss or acute haemolysis, during
The converse is seen later in life when an increased recovery from bone marrow or red cell aplasia and
haemoglobin F level in some inherited disorders of shortly after erythropoietin administration.
globin chain synthesis results from preferential Haemoglobin F synthesis is increased in certain
survival of cells that contain more haemoglobin F. leukaemias. This is characteristic of juvenile myelo
This can occur in sickle cell anaemia, β+ and β0 monocytic leukaemia (Figs 6.3 and 6.4) and is often
330 Chapter 6
Table 6.2 Some acquired causes of an increased percentage of haemoglobin F. (From references 34–36, 40–48 and other
sources.)
Neonatal period (acquired in utero)

Premature babies
Small for gestational age babies
Chronic intrauterine hypoxia
Infants of diabetic mothers [43, 44]
Maternal cigarette smoking [35, 36]
Infancy and childhood

Juvenile myelomonocytic leukaemia
Recovery from transient erythroblastopenia of childhood
Adults or any age

Pregnancy (second trimester when rapid erythroid expansion is occurring)
‘Stress erythropoiesis’
Blood loss or phlebotomy
Acute or severe haemolysis
Recovery from bone marrow failure, e.g. following treatment of acute leukaemia, recovery from severe infection
Recovery from pure red cell aplasia
Recovery from iron deficiency anaemia
Erythropoietin administration.
During regeneration following bone marrow transplantation
Aplastic anaemia, particularly during androgen administration
Pernicious anaemia
Some cases of acute myeloid leukaemia (particularly erythroleukaemia), myelodysplastic syndromes, paroxysmal
nocturnal haemoglobinuria and polycythaemia vera
Tumours including hydatidiform mole, carcinoma of the bronchus and breast, ovarian and testicular tumours,
gastrointestinal tumours, hepatoma and multiple myeloma [45]
Thyrotoxicosis [46]
Diabetes mellitus (minor increase) [48]
Starvation ketoacidosis [34]
Kala azar [42]
Drug‐induced
Hydroxycarbamide administration in sickle cell anaemia or β thalassaemia intermedia
Butyrate administration in sickle cell anaemia
Acylating agents, such as phenylacetate and phenylbutyrate, in sickle cell anaemia
Azacitidine administration in sickle cell anaemia
Erythropoietin
Sodium valproate
used as one of the features defining this disorder. increased percentage of haemoglobin F has been
The mechanism is unknown but the disease is char associated with a worse prognosis in children with
acterised by reversion to other features of fetal hae this type of leukaemia [49].
mopoiesis, such as decreased haemoglobin A2, Haemoglobin F levels can be raised pharmaco
decreased expression of erythrocyte I antigen, logically (e.g. by administration of butyrate, acylat
increased expression of erythrocyte i antigen and ing agents or cytotoxic agents, which alter gene
reduced expression of carbonic anhydrase. When expression). Higher levels observed in patients
the haemoglobin F percentage is increased, the Gγ:Aγ with human immunodeficiency virus (HIV) infec
ratio is that usually seen in fetal life rather than that tion may be related to the administration of zido
usually seen beyond the first few months of life. An vudine [50].
(a)
(b)
Fig. 6.3 Juvenile myelomonocytic

leukaemia showing: (a) blood film
MGG ×100; (b) Kleihauer test
showing increased F‐containing
cells (×40); and (c) negative control
for Kleihauer test (×40). (With
thanks to Dr Bridget Wilkins.) (c)
332 Chapter 6
Fig. 6.4 High performance liquid chromatography (HPLC) chromatogram (Bio‐Rad Variant II) in juvenile myelomono
cytic leukaemia showing, from left to right, post‐translationally modified haemoglobin F, haemoglobin F0 and haemo
globin A. Note the acquired absence of haemoglobin A2.
noted that an acquired condition leading to decreased

Increased or decreased
synthesis of haemoglobin A2 can lower the percent
haemoglobin A2
age in a patient with β thalassaemia trait and, in a
A low level of haemoglobin A2 synthesis is physio mild case, could obscure the diagnosis. A low haemo
logical in the fetus and neonate. globin A2 percentage, with or without an increased
At other stages of life, a low rate of synthesis can be percentage of haemoglobin F, may be predictive of
inherited or acquired. Reduced synthesis of haemo leukaemic transformation in aplastic anaemia [51].
globin A2 is relatively common as an acquired disor An increased haemoglobin A2 percentage is an
der, as a result of iron deficiency or impaired delivery uncommon acquired abnormality except in the
of iron to developing erythroid cells. It should be context of HIV infection and its treatment.
Table 6.3 Some acquired causes of an increased or impaired [62]. Not only haemoglobin A but also
decreased percentage of haemoglobin A2 [46, 51–61]. other normal and variant haemoglobins are glyco
sylated. Glycosylated haemoglobins can sometimes
Increased percentage
lead to diagnostic confusion. For example, on high
Hyperthyroidism [46]
Megaloblastic anaemia consequent on deficiency of
performance liquid chromatography (HPLC), a gly
vitamin B12 or folic acid (some cases)* [52, 53] cosylated haemoglobin may have the same
Human immunodeficiency virus (HIV) infection and its characteristics as another normal or variant haemo
treatment with zidovudine [54–57] globin; with some systems, glycosylated haemoglo
Associated with hypertrophic osteoarthropathy [58] bin S elutes very close to haemoglobin A.
Malaria [59] Haemoglobin A1c is lower than normal in patients
Decreased percentage with haemolytic anaemia since glycosylation
Iron deficiency increases with the average age of the red cell. It is
Anaemia of chronic disease [60] higher than normal in transient erythroblastopenia
Sideroblastic anaemia [53] of childhood [63], indicating the older red cell pop
Lead poisoning [53]
ulation. It is significantly increased in iron defi
Juvenile myelomonocytic leukaemia
Some myelodysplastic syndromes, including acquired ciency anaemia [64]. Other real and artefactual
haemoglobin H disease causes of increased or decreased haemoglobin A1c
Some cases of acute myeloid leukaemia, particularly percentage are shown in Table 6.4 [63–66].
erythroleukaemia
Some cases of aplastic anaemia [51]
Hypothyroidism [53] Carboxyhaemoglobinaemia
Associated with chemotherapy‐induced increased
Carboxyhaemoglobin is formed when carbon
haemoglobin F synthesis [61]
monoxide combines with haem iron. Carbon
* But note that folic acid deficiency has been reported to lower the monoxide is a product of incomplete combustion
haemoglobin A2 percentage in individuals with β thalassaemia [52]. of hydrocarbons. Endogenous production of car
bon monoxide as a result of catabolism of haemo
globin (specifically haem), and to a lesser extent
Acquired abnormalities leading to an increased myoglobin, is a physiological process so that a
or reduced percentage of haemoglobin A2 are low level of carboxyhaemoglobin is detectable in
summarised in Table 6.3 [46, 51–61]. (For inherited everyone [42, 67]. Healthy non‐smokers have
causes of an abnormal haemoglobin A2 percentage around 1% of carboxyhaemoglobin with higher
see Tables 3.8 and 3.12.) levels (around 5%) being observed in pregnancy
and in the presence of haemolytic anaemia [68] or
a haematoma [69]. Fetuses and neonates have
Increased or decreased glycosylated
higher levels than adults [69]. A much higher
haemoglobin
level of carboxyhaemoglobin in the blood occurs
Addition of glucose to the N‐terminus of the β if there is exposure to exogenous carbon monox
chain, forming haemoglobin A1c, is a normal post‐ ide. Increased levels of carboxyhaemoglobin can
translational modification of haemoglobin. The result from exposure to:
concentration is usually 3–4%. Other glycosylated • car exhaust fumes (deliberate or accidental,
fractions are present at a lower concentration; these including snow‐blocked car exhaust pipe [70] – the
include haemoglobins A1a and A1b. A higher concen risk is less when a catalytic converter is fitted);
tration of blood glucose in diabetes mellitus leads to • coal, gas, peat and charcoal fires, water heaters,
an increased percentage of glycosylated haemoglo central heating boilers, heating appliances, power
bin (Fig. 6.5). Glycosylation is irreversible but does generators, other combustion engines – particularly
not have any major effect on haemoglobin function. when ventilation is poor and if a fire intended for
The oxygen affinity is mildly increased as inter outdoor use is used indoors or in a sealed tent;
action with 2,3‐diphosphoglycerate (2,3‐DPG) is • inhaled smoke in house fires;
334 Chapter 6
(a) (b)
Fig. 6.5 HPLC chromatogram showing: (a) increased haemoglobin A1c in a diabetic patient and (b) normal haemoglobin
A1c in a control sample.
Table 6.4 Causes of real and apparent increased and decreased percentages of haemoglobin A1c [63–66].
Increased percentage Decreased percentage
Factitious Presence of a variant haemoglobin with a Presence of various variant haemoglobins including
retention time overlapping with haemoglobins S, C, Takamatsu, G‐Szuhu, Himeyi, O‐
haemoglobin A1c Padova, Camden, Riyadh, J‐Meerut, Sherwood Forest,
Manitoba and G‐Coushatta [65]
True Diabetes mellitus Shortened red cell life span

Aged population of red cells, e.g. recent onset
of pure red cell aplasia or transient
erythroblastopenia of childhood [63]
Iron deficiency anaemia [64]
HIV infection [65]
HIV, human immunodeficiency virus.
• industrial fumes, including both propane and are seen in suicide attempts using car exhaust fumes
methane used as fuels and methylene chloride and with accidental exposure to industrial chemicals
(a common component of paint remover and other or the combustion products of poorly ventilated
solvents, which is metabolised by the liver to car domestic heaters. Carbon monoxide poisoning is
bon monoxide) [67]; caused by suicide attempts in about a half of
• cigarette smoke and hookah smoking. instances, is associated with burns in about a quarter
Cigarette smokers often have carboxyhaemo of cases and results from other unintentional expo
globin levels of 5–10% but sometimes up to 20%. sure to carbon monoxide in another quarter [72].
Higher levels (e.g. 38%) can be seen with hookah A high concentration of carboxyhaemoglobin gives
smoking [71]. Part of the risk to the fetus of cigarette the blood a cherry‐red colour. Symptoms include
smoking in pregnancy is attributable to carbon mon headache, lethargy, nausea and vomiting followed
oxide [69]. Much higher levels, which may be fatal, by drowsiness, convulsions, coma and death.
The affinity of haemoglobin for carbon m onoxide the percentage of carboxyhaemoglobin (assuming
is 200–250 times as great as its affinity for oxygen. that the individual has been removed from the
The process of formation of carboxyhaemoglobin is source of carbon monoxide and is breathing
slowly reversible. Production of carboxyhaemo inspired air with a low content of carbon
globin moves the oxygen dissociation curve to the monoxide).
left and makes it more hyperbolic. This is because, If a pregnant woman is exposed to carbon mon
once two carbon monoxide molecules are bound to oxide, effects in the fetus are even more severe
haem groups, the molecule changes to an oxy con [67]. Fetal haemoglobin has a greater affinity for
formation, increasing the affinity for oxygen. This carbon monoxide than haemoglobin A so that the
means that, in individuals with an increased per percentage of carboxyhaemoglobin is 10–15%
centage of carboxyhaemoglobin, the degree of higher in the fetus. In addition, the half‐life of fetal
tissue hypoxia is greater than would be expected carboxyhaemoglobin F is about twice as long as
from the percentage of carboxyhaemoglobin pre that of maternal carboxyhaemoglobin, so that
sent. The increased oxygen affinity caused by bind recovery takes longer [75]. There is an exaggerated
ing to carbon monoxide is more important than the leftwards shift of the oxygen dissociation curve
decreased affinity attributable to 2,3‐DPG in with resultant severe tissue hypoxia. There is a
explaining the varying P50o2 seen in normal indi similar increased vulnerability to carbon monox
viduals [42]. Carboxyhaemoglobin does not func ide poisoning during the first few months of life,
tion in oxygen transport and, in addition, oxygen when fetal haemoglobin levels remain high, and
delivery to tissues is impaired as is shown by the later in life in individuals with a persistent eleva
altered shape of the dissociation curve. The effects tion of fetal haemoglobin [73].
of tissue hypoxia are further aggravated by the Haemoglobin Zurich is an interesting example of
binding of carbon monoxide to myoglobin and to the interaction between inherited and acquired
mitochondrial cytochromes [68, 73]. The impaired abnormalities of the haemoglobin molecule. It has a
tissue oxygen delivery in individuals with a chronic greater affinity for carbon monoxide than does hae
increase in the percentage of carboxyhaemoglobin moglobin A and, paradoxically, this protects ciga
means that there is an erythropoietin‐driven rette smokers from haemolysis.
increase in the rate of haemoglobin synthesis and Carboxyhaemoglobin is detected and measured
the haemoglobin concentration is increased. This by spectroscopy (Fig. 6.6). However, it should be
may mean that increased blood viscosity com noted that the severity of carbon monoxide poison
pounds the effects of reduced delivery of oxygen to ing correlates poorly with the carboxyhaemoglobin
tissues. percentage and this should not be used for judging
When the concentration of oxygen in the inspired the necessity for hyperbaric oxygen therapy [76].
air is lower, the effects of any influences likely to Pulse oximetry is inaccurate in measuring oxyhae
raise the carboxyhaemoglobin concentration are moglobin when carboxyhaemoglobin is present
greater; for example, poor combustion of a stove in since the technique does not distinguish between
a sealed tent would have a greater effect at altitude. carboxyhaemoglobin and oxyhaemoglobin [67].
Conversion of carboxyhaemoglobin to oxyhaemo This is because the wavelengths employed by most
globin is accelerated by removal from the source of pulse oximeters are selected to distinguish between
carbon monoxide and by ventilation with oxygen, oxygenated and non‐oxygenated haemoglobin and
or hyperbaric oxygen treatment. The half‐life of not to distinguish oxyhaemoglobin from other
carboxyhaemoglobin is 240–320 minutes breathing forms of haemoglobin. A pulse CO‐oximeter (or co‐
air, 80–100 minutes breathing 100% oxygen and oximeter), however, measures the percentages of
about 20 minutes with hyperbaric oxygen [73]. carboxyhaemoglobin, oxyhaemoglobin and deoxy
Hyperbaric oxygen therapy has been demonstrated haemoglobin. Carboxyhaemoglobin may be seri
to reduce neurological damage [74]. An increased ously underestimated by some spectrophotometers
rate of ventilation, for example as a consequence of if the patient has been administered hydroxocobal
vigorous exercise or artificial ventilation, lowers amin as part of emergency management [77].
336 Chapter 6
22
20
18
16 COHb
Hb
Absorbance
14 O2Hb
12
10
8
6
4
2
Fig. 6.6 Spectroscopy showing
0
carboxyhaemoglobin (COHb),
510 550 600 650
Wavelength in nm oxyhaemoglobin (O2Hb) and
deoxyhaemoglobin (Hb).
Symptoms are considerably greater when an acute

Methaemoglobinaemia rise in the percentage of methaemoglobin occurs
Methaemoglobin is formed by oxidation of haem than when there is chronic methaemoglobinaemia.
iron (i.e. by conversion from the ferrous (Fe2+) to A concentration of 0.15–0.2 g/l causes cyanosis (in
the ferric (Fe3+) form). Auto‐oxidation of haemo comparison with the 0.5 g/l of deoxyhaemoglobin
globin occurs, the α chain being oxidised more that is needed to produce cyanosis). Methaemoglobin
rapidly than the β chain in an intact molecule. is detected by spectroscopy (Fig. 6.11) or by co‐
Auto‐oxidation is increased by a rise in tempera oximetry, which permits the detection of carboxy‐,
ture, increased 2,3‐DPG and reduced pH [69]. The met‐ and sulphaemoglobins. It should be noted
rate of production of methaemoglobin is increased that, as for patients with carboxyhaemoglobinae
by sepsis. Oxidation of haemoglobin is increased mia, pulse oximetry and conventional blood gas
by exposure to exogenous oxidants. Various red analysis are misleading in patients with methaemo
cell enzymes convert methaemoglobin back to globinaemia [81, 82], since methaemoglobin has
haemoglobin so that the level is normally less similar light absorption characteristics to oxyhae
than 1%. moglobin. Pulse oximetry overestimates the oxygen
Methaemoglobin does not function in oxygen saturation. Congenital and acquired causes of
transport and, in addition, the presence of methae methaemoglobinaemia are shown in Table 6.5 [42,
moglobin leads to a left shift of the oxygen dissocia 69, 78, 83–116]. It is also possible that methaemoglo
tion curve, which further impairs oxygen delivery binaemia can be the result of endogenous oxidant
to tissues; this is because oxidation of some haems production by gut bacteria since two patients have
in a partly oxidised tetramer favours the oxy con been reported in whom combined met‐ and
formation of the tetramer. High levels of methae sulphaemoglobinaemia responded to oral adminis
moglobin are associated with chocolate‐coloured tration of antibiotics [117] and neonates with
blood, cyanosis (or pseudocyanosis) (Figs 6.7–6.9), diarrhoea and methaemoglobinaemia have been

headache, tachycardia, dyspnoea, tachypnoea and reported [118].
finally coma and death. In patients with glucose‐6‐ Infants are more susceptible to methaemoglobi
phosphate dehydrogenase deficiency, intravascular naemia than adults. This is only in part because
haemolysis can lead to plasma and serum being haemoglobin F is more susceptible to oxidation. It is
brown [78, 79]. The blood film may show evidence related more to the reduced level of methaemo
of oxidative damage to red cells [80] (Fig. 6.10). globin reductase at this age.
Fig. 6.7 Hands of a patient

with anaemia and mild
methaemoglobinaemia caused by
dapsone, in comparison of the hand
of a healthy subject. (With thanks to
Professor Lawrence Hirst.)
The congenital methaemoglobinaemias caused

by methaemoglobin reductase (NADH‐cytochrome
b5 reductase) deficiency enter into the differential
diagnosis of methaemoglobinaemia. They usually
have 10–20% of methaemoglobin. The reduced
oxygen‐carrying capacity of the blood leads to sec
ondary polycythaemia. In type I deficiency, meth
aemoglobinaemia is the only clinical m anifestation,
whereas in type II deficiency, in which tissues other
than red cells are affected, there is also neurological
impairment and growth retardation. Both have an
autosomal recessive inheritance. Individuals with
an inherited methaemoglobin reductase deficiency
have an increased susceptibility to oxidant drugs
and chemicals.
Patients with severe symptoms from methaemo
globinaemia are treated with intravenous methylene
blue. In life‐threatening situations, exchange transfu
sion can be employed. Methylene blue is, however,
contraindicated in patients with methaemoglobi
naemia associated with haemolysis in glucose‐6‐
phosphate dehydrogenase deficiency [87, 88].
Sulphaemoglobinaemia
Sulphaemoglobin is produced by irreversible
oxidation and chemical alteration of haemoglo
bin. The mechanism is probably production of
Fig. 6.8 The tongue of a patient with methaemoglobinaemia. a
ferrohaemoglobin–peroxide complex that is
338 Chapter 6
Fig. 6.9 A bag of blood removed

during exchange transfusion (bottom)
from a patient with severe
methaemoglobinaemia induced by an
oxidant chemical, showing the
chocolate colour conveyed by
methaemoglobin. The upper bag, in
contrast, is a bag of normal donor
blood.

with dapsone‐induced
methaemoglobinaemia showing
irregularly contracted cells and
keratocytes (‘bite cells’). One cell
contains a Howell–Jolly body,
indicative of functional hyposplenism
due to splenic overload during an
episode of acute haemolysis. MGG ×100.
s ulphated in the presence of hydrogen sulphide responded to oral administration of antibiotics

[117]. The causes of sulphaemoglobinaemia are [117]. Abnormal gut flora as a result of constipa
summarised in Table 6.6 [42, 119–121]. As already tion has also been suspected as a cause [62]. The
mentioned, it is also likely that sulphaemoglobi drugs that can cause methaemoglobinaemia can
naemia can result from an abnormal population also cause sulphaemoglobinaemia. The common
of gut bacteria since three patients have been est cause is exposure to drugs such as dapsone
reported in whom sulphaemoglobinaemia, or and previously phenacetin (sometimes now pre
combined met‐ and sulphaemoglobinaemia, sent as an adulterant in cocaine). Occupational
22
20
18
16
Absorbance
14 Hb
O2Hb
12
10
8
6
MetHb
4
2
0
510 550 600 650
Wavelength in nm
Fig. 6.11 Spectroscopy showing methaemoglobin (MetHb), oxyhaemoglobin (O2Hb) and deoxyhaemoglobin (Hb).
Table 6.5 Some causes of an increased concentration of methaemoglobin. (Derived from references 42, 69, 78, 83–116
and other sources.)
Inherited
Haemoglobin M and haemoglobin F‐M
As a feature of an unstable haemoglobin
As a feature of haemoglobin E/β thalassaemia correlating with disease severity and previous splenectomy [83]
Deficiency of NADH‐cytochrome b5 reductase
Type I (deficiency of red cells only)
Type II (generalised tissue deficiency)
Deficiency of cytochrome b5 (single patient) [84]
Acquired
Residence at high altitude [85]
Exposure to oxidant drugs and chemicals
Nitrates, nitrites and related compounds

Nitrites such as sodium nitrite*, amyl nitrite† [86], butyl nitrite†, isobutyl nitrite† (therapeutic doses of nitrites may produce
up to 5% methaemoglobin; ‘poppers’, used recreationally, are volatile amyl, butyl or isobutyl nitrite)
Nitric oxide (NO) (by inhalation), including nitric oxide administered to neonates with pulmonary hypertension
Nitrates (converted to nitrites by hepatic metabolism), e.g. nitrate‐rich diet (particularly in young babies); excessive
amount of sodium nitrate used as preservative for sausage meat; Chinese dumplings with a high content of nitrates or
containing nitrite‐cured meat [87]; nitrate‐contaminated well water‡ including when it is used for home dialysis or for
reconstituting powdered milk for infants)
Nitroglycerine (intravenously)
Nitroprusside
Nitrobenzenes (shoe and floor polish, paint solvents, agricultural products) [86]
Nitroethane (nail polish remover, propellant, fuel additive) [86]
Nitrofurantoin [86]
Other
3‐Aminopyridine‐2‐carboxaldehyde thiosemicarbazone (Triapine) [88]
Aniline (purchased as a recreational drug) [89] or contained in petrol octane booster
Celecoxib [90]
Clofazimine
Copper sulphate (ingested with suicidal intent)
Dapsone (rarely topical dapsone) [91]
Diarylsulphonylureas [86]
Doxorubicin [86]
Doxycycline (when used in high doses for sclerotherapy) [92]
Favism in G6PD‐deficient subjects [93]
Flutamide [94]
Henna
Hypochlorous acid (contained in bleach) [95]
Local or topical anaesthetics§
Topical benzocaine [96] including when used as an oral gel for post‐chemotherapy stomatitis [97] and when used for
transoesophageal echocardiography [98]
Prilocaine [99]
Procaine
Lidocaine
Lysol (50% cresol in linseed oil, potassium hydroxide and water) [100]
Mephedrone (in ‘snow’ – a recreational drug) [101]
Metoclopramide [86]
Methylene blue (in G6PD‐deficient subjects or in high dose) [102]
Naphthalene
Paracetamol (acetominophen)
Phenacetin (now withdrawn from legal market)
Phenazopyridine [103]
Potassium permanganate [104]
Primaquine, chloroquine [86]
Propanil (a herbicide)
Rasburicase [105]
Riluzole (overdose) [106]
Sulphonamides [96] including sulphamethoxazole (constituent of cotrimoxazole)
Unidentified chemicals used by goldsmiths [107]
G6PD, glucose‐6‐phosphate dehydrogenase.

* Used for curing meat, as a food preservative, as an insecticide and to inhibit corrosion (may contaminate water in pipes and tanks when
it has been used as a corrosion‐inhibiting solution) [99]; used in treatment of cyanide poisoning [108]; present in some antifreeze [109].
† Including ‘recreational’ use.
‡ The risks from well water relate to the application of nitrogenous fertilizers to surrounding farmland [42].
§ Benzocaine and novocaine can be present as adulterants in street heroin [78].
Table 6.6 Causes of an increased

Inherited concentration of sulphaemoglobin
Autosomal dominant inherited sulphaemoglobinaemia [42] [42, 119–121.]
As a feature of some unstable haemoglobins, e.g. haemoglobin Olmstead [42]
Acquired
Exposure to drugs and chemicals (as for methaemoglobinaemia), but
occasionally drug‐induced sulphaemoglobinaemia occurs in the absence of
methaemoglobinaemia, e.g. with sumatriptan therapy [119]
Sulphur‐containing ointments
Occupational exposure to hydrogen sulphide (factory workers, sewage
workers, livestock farmers, workers in sulphurous thermal baths)
Escherichia coli septicaemia (in a neonate) [120]
Intestinal Morganella morganii (in a neonate) [121]
22
SulfHb
20
18
16
Hb
Absorbance
14 O2Hb
12
10
8
6
4
2
Fig. 6.12 Spectroscopy showing
0
sulphaemoglobin (sulfHb),
510 550 600 650
oxyhaemoglobin (O2Hb) and
Wavelength in nm
deoxyhaemoglobin (Hb).
exposure to sulphur‐containing compounds can Check your knowledge

also be responsible. Flutamide, an anti‐androgen
used for treating c arcinoma of the prostate, has One to five answers may be correct. Answers to
been reported to cause sulphaemoglobinaemia most questions can be either found in this chapter
and methaemoglobinaemia [94]. A single case of or deduced from information given. Answers are
sulph‐ and methaemoglobinaemia apparently given on p. 347.
caused by intravenous all‐trans‐retinoic acid has
also been reported [115]. Sulphaemoglobin lacks 6.1 Cyanosis or pseudocyanosis can be
cooperativity [62] and does not function in caused by
oxygen transport. However, it leads to reduced (a) methaemoglobinaemia
oxygen affinity of non‐sulphinated monomers, (b) sulphaemoglobinaemia
ameliorating the effect on tissue oxygen delivery. (c) carboxyhaemoglobinaemia
It has been postulated that sulphaemoglobinae (d) increased haemoglobin A1c
mia would be much more deleterious in patients (e) a low affinity haemoglobin
with sickle cell disease because favouring of the
deoxy conformation would also favour mainte 6.2 A reduced percentage of haemoglobin A2 can
nance of polymerisation [69]. A concentration of be caused by
5–16 g/l or above can cause cyanosis (or pseudo (a) haemochromatosis
cyanosis). At very high levels the beneficial effect (b) diabetes mellitus
on oxygen affinity is negated by an extremely (c) anaemia of chronic disease
right‐shifted oxygen dissociation curve. In severe (d) iron deficiency
acute sulphaemoglobinaemia there are neurolog (e) sideroblastic anaemia
ical effects, pulmonary oedema and death [69].
Sulphaemoglobin can be detected by spectros 6.3 An increased percentage of haemoglobin F in
copy (Fig. 6.12) and co‐oximetry. Pulse oximetry a neonate can be caused by
is inaccurate, the oxygen saturation being over (a) intrauterine hypoxia
estimated [82]. (b) post‐mature baby
Methylene blue is ineffective in the treatment of (c) maternal diabetes
sulphaemoglobinaemia but, as most patients are (d) maternal alcohol intake
asymptomatic, specific treatment is not indicated. (e) maternal cigarette smoking
342 Chapter 6
6.4 Oxygen affinity of whole blood is increased 6.10 Methaemoglobinaemia can result from
by (a) recreational use of nitrites
(a) a high percentage of haemoglobin F (b) domestic heaters in poorly ventilated
(b) methaemoglobinaemia rooms
(c) sulphaemoglobinaemia (c) occupational exposure to nitrates
(d) a high percentage of haemoglobin A1c (d) cigarette smoking
(e) carboxyhaemoglobinaemia (e) an inherited abnormality of haemoglobin
structure
6.5 An increased percentage of haemoglobin F
can be caused by 6.11 The increased percentage of carboxyhaemo
(a) pregnancy globin in cigarette smokers
(b) the menopause (a) is greater at increased altitude
(c) recovery from aplastic anaemia (b) can lead to an increased haemoglobin
(d) juvenile myelomonocytic leukaemia concentration
(e) hydatidiform mole (c) can reach 50% of total haemoglobin
(d) is reduced by a period of vigorous exercise
6.6 Acquired haemoglobin H disease is a rare but (e) does not have any disadvantageous
recognised feature of effects
(a) chronic lymphocytic leukaemia
(b) acute myeloid leukaemia 6.12 Sulphaemoglobinaemia
(c) myelodysplastic syndromes (a) may occur together with
(d) non‐Hodgkin’s lymphoma methaemoglobinaemia
(e) congenital sideroblastic anaemia (b) usually causes severe symptoms
(c) can be corrected by intravenous injection
6.7 Carbon monoxide can be derived from of methylene blue
(a) metabolism of haemoglobin (d) can be caused by dapsone therapy
(b) metabolism of glucose (e) alters the absorption spectrum of
(c) poorly ventilated domestic heaters haemoglobin
(d) exhaust fumes of motor vehicles
(e) cigarette fumes
References
6.8 An increased percentage of haemoglobin A1c
1 Peters RE, May A and Jacobs A (1986) Increased
can result from
α:non‐α globin chain synthesis ratios in myelodys
(a) haemolytic anaemia
plastic syndromes and myeloid leukaemias. J Clin
(b) exposure to carbon monoxide
Pathol, 39, 1233–1235.
(c) recent onset of pure red cell aplasia 2 Steensma DP, Viprakasit V, Hendrick A, Goff DK,
(d) poorly controlled diabetes mellitus Leach J, Gibbons RJ and Higgs DR (2004) Deletion
(e) β thalassaemia trait of the α‐globin gene cluster as a cause of acquired
α‐thalassaemia in myelodysplastic syndrome.
6.9 Carboxyhaemoglobin
Blood, 103, 1518–1520.
(a) is formed during the transport of carbon
3 Markham RE, Butler F, Goh K and Rowley P (1983)
dioxide from tissues to the lungs Erythroleukemia manifesting as δβ thalassemia.
(b) is produced by the catabolism of Hemoglobin, 7, 71–78.
myoglobin 4 Aoyagi Y, Akimoto H, Yamaoka K, Koike A,
(c) is not present in healthy individuals Kanayama M, Miyaji T and Yamamoto K (1989)
(d) causes the blood to be chocolate‐brown Refractory anemia with ringed sideroblasts compli
in colour cated by δβ‐thalassemia‐like hemoglobinopathy.
(e) cannot be reconverted to oxyhaemoglobin Rinsho Ketsueki, 30, 674–679.
5 Hoyle C, Kaeda J, Leslie J and Luzzatto L (1991) acquired hemoglobin H disease. Am J Med, 75,
Case report: acquired β thalassaemia trait in MDS. 181–191.
Br J Haematol, 79, 116–117. 19 Anognou NP, Ley TJ. Chesbro B, Wright G, Kitchens
6 Honig GR, Suarez CR, Vida LN, Lu S‐J and Liu ET C, Liebhaber S et al. (1983) Acquired α‐thalassemia
(1998) Juvenile myelomonocytic leukemia (JMML) in preleukemia is due to decreased expression of all
with the hematologic phenotype of severe β thalas four α‐globin genes Proc Natl Acad Sci U S A, 80,
semia. Am J Hematol, 58, 67–71. 6051–6055.
7 White JC, Ellis M, Coleman PN, Beaven GH, Gratzer 20 Mercieca J, Bain B, Barlow G and Catovsky D (1996)
WB, Shooter EM and Skinner ER (1960) An unstable Teaching cases from the Royal Marsden and St
haemoglobin associated with some cases of leukae Mary’s Hospitals, Case 10: microcytic anaemia and
mia. Br J Haematol, 6, 171–177. thrombocytosis. Leuk Lymphoma, 18, 185–186.
8 Beaven GH, Stevens BL, Dance N and White JC 21 Higgs DR, Buckle VJ, Gibbons R and Steensma D.
(1963) Occurrence of haemoglobin H in leukaemia. Unusual types of α thalassemia. In: Steinberg MH,
Nature, 199, 1297–1298. Forget BG, Higgs DR and Nagel RL, eds. Disorders
9 Rosenweig AI, Heywood JD, Motulsky AG and of Hemoglobin: Genetics, Pathophysiology, and Clinical
Finch CA (1968) Hemoglobin H as an acquired Management, 2nd edn. Cambridge University Press,
defect in alpha‐chain synthesis. Acta Haematol, 8, Cambridge, 2009, pp. 296–320.
91–101. 22 Yin XL, He YY, Zhang XH, Zhou TH and Zhang TL
10 Hamilton RW, Schwartz E, Atwater J and Erslev AJ (2010) Acquired hemoglobin H disease associated
(1971) Acquired Hb H disease. N Engl J Med, 285, with Down syndrome. J Pediatr Hematol Oncol, 32,
1217–1221. e244–e245.
11 Boehme WM, Piira TA, Kurnik JE, and Bethlenfalvay 23 Gibbons RJ, Pellagatti A, Garrick P, Wood WG,
NC (1978) Acquired hemoglobin H disease in Malik N, Ayyub H et al. (1995) Identification of
refractory sideroblastic anemia. A preleukemic acquired somatic mutations in the gene encoding
marker. Arch Intern Med, 138, 603–606. chromatin‐remodeling factor ATRX in the alpha‐
12 Lindsey RJ, Jackson JM and Raven JL (1978) thalassaemia myelodysplasia syndrome (ATMDS).
Acquired haemoglobin H disease complicating a Nat Genet, 34, 446–449.
myeloproliferative syndrome: a case report. 24 Steensma DP, Higgs DR, Fisher CA and Gibbons RJ
Pathology, 10, 329–334. (2004) Acquired somatic ATRX mutations in myelo
13 Weatherall DJ, Old J, Longley J, Wood WH, Clegg dysplastic syndrome associated with α thalassemia
JB, Pollock A and Lewis MJ (1978) Acquired (ATMDS) convey a more severe hematologic phe
haemoglobin H disease in leukaemia: pathophysi notype than germline ATRX mutations. Blood, 103,
ology and molecular basis. Br J Haematol, 38, 2019–2026.
305–322. 25 Helder J and Deisseroth A (1987) SI nuclease analy
14 Veer A, Kosciolek BA, Bauman AW, Rowley PT sis of α‐globin gene expression in preleukemic
(1979) Acquired hemoglobin H disease in idiopathic patients with acquired hemoglobin H disease after
myelofibrosis. Am J Hematol, 6, 199–206. transfer to mouse erythroleukemia cells. Proc Natl
15 Dash S and Dash RJ (1980) Idiopathic myelofibrosis Acad Sci U S A, 84, 2387–2390.
without splenomegaly and with an acquired hae 26 Rinaldi CR, Rinaldi P, Pane F, Camera A and Rinaldi
moglobin disorder. Indian J Cancer, 17, 193–195. C (2010) Acquired Hb H disease associated with
16 Yoo D, Schechter GP, Amigable AN and Nienhuis elevated Hb F level in patient affected by primary
AW (1980) Myeloproliferative syndrome with side myelofibrosis. Ann Hematol, 89, 827–828.
roblastic anemia and acquired hemoglobin H disease. 27 Belickova M, Schroeder HW, Guan YL, Brierre J,
Cancer, 45, 78–83. Berney S, Cooper MD and Prchal JF (1994) Clonal
17 Kueh YK (1982) Acute lymphoblastic leukemia with hematopoiesis and acquired thalassaemia in com
brilliant cresyl blue erythrocytic inclusions‐acquired mon variable immunodeficiency. Mol Med, 6, 56–61.
hemoglobin H disease? N Engl J Med, 307, 193–194. 28 Brunner AM and Steensma DP (2016) Myelodysplastic
18 Higgs DR, Wood WG, Barton C and Weatherall DJ syndrome associated with acquired beta thalassemia:
(1983) Clinical features and molecular analysis of “BTMDS”. Am J Hematol, 91, E325–E327.
344 Chapter 6
29 Badens C, Mattei MG, Imbert AM, Lapoumeroulie 43 Perrine SP, Greene MF and Faller DV (1985) Delay
C, Martini N, Michel G and Lena‐Russo D (2002) A in fetal globin switch in infants of diabetic mothers.
novel mechanism for thalassaemia intermedia, N Engl J Med, 312, 334–338.
Lancet, 359, 132–133. 44 Peters F, Rohloff D, Kohlmann T, Renner F,
30 Chang JG, Tsai WC, Chong IW, Chang CS, Lin CC Jantschek G, Kerner W and Fehm HL (1998) Fetal
and Liu TC (2008) β‐thalassemia major evolution hemoglobin in starvation ketosis in young women.
from β‐thalassemia minor is associated with pater Blood, 91, 691–694.
nal uniparental isodisomy of chromosome 11p15. 45 Kellen JA, Bush RS and Malkin A (1980) Alkali‐
Haematologica, 93, 913–916. resistant hemoglobin (Hb F) in cancer patients.
31 Bento C, Maia TM, Milosevic JD, Carreira IM, Cancer, 45, 1448–1450.
Kralovics R and Ribeiro ML (2013) β thalassemia 46 Kendall AG and Bastomsky CH (1981) Haemoglobin
major due to acquired uniparental disomy in a previ A2 in hyperthyroidism. Hemoglobin, 5, 571–577.
ously healthy adolescent. Haematologica, 98, e4–e6. 47 Little JA, Dempsey NJ, Tuchman M and Ginder GD
32 Wilson MG, Schroeder WA and Graves DA (1968) (1995) Metabolic persistence of fetal hemoglobin.
Postnatal change of haemoglobin F and A2 in Blood, 85, 1712–1718.
infants with Down’s syndrome (G trisomy). 48 Lee MH, Kim JH, Im MW and Choi JW (2009)
Pediatrics, 42, 349–353. Association between enhanced fetal haemoglobin
33 Weller SDV, Apley J and Raper AB (1966) levels and ineffective reticulocyte production in
Malformations associated with precocious synthe diabetics. Acta Haematol, 122, 247–251.
sis of adult haemoglobin. Lancet, i, 777–779. 49 Luna‐Fineman S, Shannon KM, Atwater SK, Davis
34 Nienhuis AW and Benz EJ (1977) Regulation of J, Masterson M, Ortega J et al. (1999) Myelodysplastic
hemoglobin synthesis during the development of and myeloproliferative disorders of childhood: a
the red cell (third of three parts). N Engl J Med, 297, study of 167 patients. Blood, 93, 459–466.
1430–1436. 50 Poli‐Neto A, Nonoyama K, Oshiro M, Ebner‐Filho
35 Fagan DG, Lancashire RJ, Walker A and Sorohan T W, Miguita K, Watanabe CI and Barretto OC de O
(1995) Determinants of fetal haemoglobin in new (1998) Increased fetal hemoglobin levels in H.I.V.
born infants. Arch Dis Child, 72, F111–F114. patients. Br J Haematol, 102, 96.
36 Cochran DL, Conrad ME and Matney J (1997) 51 Aksoy M and Erdem S (1978) Followup study of the
Hemoglobin F and risk factors for sudden infant mortality and the development of leukemia in 444
death syndrome. Lab Med, 28, 53–57. pancytopenic patients with chronic exposure to
37 Giulian GG, Gilbert EF and Moss RL (1987) Elevated benzene. Blood, 52, 285–292.
fetal hemoglobin levels in sudden infant death syn 52 Alperin JB, Dow PA and Petteway MB (1977)
drome. N Engl J Med, 316, 1122–1126. Haemoglobin A2 levels in health and various hema
38 Cheron G, Bachoux I, Maier M, Massonneau M, tologic disorders. Am J Clin Pathol, 67, 219–226.
Peltier JY and Girot R (1989) Fetal hemoglobin in 53 Steinberg MH and Adams JG (1991) Hemoglobin
sudden infant death syndrome. N Engl J Med, 320, A2: origin, evolution, and aftermath. Blood, 78,
1011–1012. 2165–2177.
39 Zielke HR, Meny RG, O’Brien MJ, Smialek JE, 54 Routy JP, Monte M, Beaulieu R, Toma E, St‐Pierre L
Kutlar F, Huisman THJ and Dover GJ (1989) Normal and Dumont M (1993) Increase of hemoglobin A2 in
fetal hemoglobin levels in the sudden infant death human immunodeficiency virus‐I‐infected patients
syndrome. N Engl J Med, 321, 1359–1364. treated with zidovudine. Am J Hematol, 43, 86–90.
40 Papayannopoulou T, Vichinsky E and 55 Howard J, Henthorn JS, Murphy S and Davies SC
Stamatoyannopoulos G (1980) Fetal Hb production (2005) Implications of increased haemoglobin A2
during acute erythroid expansion I. Observation in levels in HIV‐positive women in the antenatal
patients with transient erythroblastopenia and clinic. J Clin Pathol, 58, 556–558.
post‐phlebotomy. Br J Haematol, 44, 535–546. 56 Wilkinson MJ, Bain BJ, Phelan L and Benzie A (2007)
41 Weatherall D and Clegg JB. The Thalassaemia Increased haemoglobin A2 percentage in HIV infec
Syndromes. Blackwell Science, Oxford, 1981. tion: disease or treatment? AIDS, 14, 37–42.
42 Bunn HF and Forget BG. Hemoglobin: Molecular, Genetic 57 Galactéros F, Amaudric F, Préhu C, Feingold N,
and Clinical Aspects. WB Saunders, Philadelphia, 1986. Doucet‐Populaire F, Sobel A and Rosa J (1993)
Acquired unbalanced hemoglobin chain synthesis MH, Forget BG, Higgs DR and Nagel RL, eds. Disorders
during HIV infection. C R Acad Sci III, 316, of Hemoglobin: Genetics, Pathophysiology, and Clinical
437–440. Management. Cambridge University Press, Cambridge,
58 Stephens AD, Angastiniotis M, Baysal E, Chan V, 2001, pp. 1214–1233.
Fucharoen S, Giordano PC et al. on behalf of the 70 Choo V (1996) Snow‐blocked car exhaust causes
International Council for the Standardisation of CO poisoning. Lancet, 347, 384.
Haematology (ICSH) (2012) ICSH recommenda 71 Maalem R, Alali A and Alqahtani S (2019) Tobacco
tions for the measurement of haemoglobin A2. Int J hookah smoking‐induced carbon monoxide poi
Lab Haematol, 34, 1–13. soning: a case report of non‐ambient exposure. Clin
59 Eng L‐I L, Lopez CG and Lopes M (1971) Case Rep, 7, 1178–1180.
Haemoglobin A2 in malaria patients. Trans R Soc 72 Pullinger R (1996) Something in the air: survival
Trop Med Hyg, 65, 480–483. after dramatic, unsuspected case of accidental car
60 Ali MAM and Schwertner E (1974) Hemoglobin A2 bon monoxide poisoning. Lancet, 312, 897–898.
level: a proposed test for confirming the diagnosis 73 Leach RM, Rees PJ and Wilmshurst P (1998)
of iron deficiency. Am J Clin Pathol, 63, 549–553. Hyperbaric oxygen therapy. Br Med J, 317, 1140–1143.
61 Cech P, Testa U, Dubart A, Schneider P, Bachmann F 74 Weaver LK, Hopkins RO. Chan KJ, Churchill S,
and, Guerrasio A (1982) Lasting Hb F reactivation Elliot CG. Clemmer TP et al. (2002) Hyperbaric oxy
and Hb A2 reduction induced by the treatment of gen for acute carbon monoxide poisoning. N Engl J
Hodgkin’s disease in a woman heterozygous for Med, 347, 1057–1067.
beta‐thalassemia and the Swiss type of the hetero 75 Longo LD (1977) The biological effects of carbon
cellular hereditary persistence of fetal haemoglo monoxide on the pregnant woman, fetus, and new
bin. Acta Haematol, 67, 275–284. born infant. Am J Obstet Gynecol, 129, 69–103.
62 Nagel RL and Steinberg MH. Hemoglobins of the 76 Pitkin AD, Broome JR and Salmon J (1999) Hyperbaric
embryo and fetus and minor hemoglobins of adults. oxygen therapy in intensive care. Part II: carbon mon
In: Steinberg MH, Forget BG, Higgs DR and Nagel oxide poisoning. Br J Intens Care, 9, 189–195.
RL, eds. Disorders of Hemoglobin: Genetics, 77 Livshits Z, Lugassy DM, Shawn LK and Hoffman
Pathophysiology, and Clinical Management. Cambridge RS (2012) Falsely low carboxyhemoglobin level
University Press, Cambridge, 2001, pp. 197–230. after hydroxocobalamin therapy. N Engl J Med, 367,
63 Karsten J, Anker AP and Odink RJ (1996) 1270–1271.
Glycosylated haemoglobin and transient erythro 78 Baeza‐Trinidad R and de la Cuesta‐López MR
blastopenia of childhood. Lancet, 347, 273. (2015) Chocolate coloured serum in methaemoglo
64 Coban E, Ozdogan M and Timuragaoglu A (2004) binaemia. Lancet, 386, e6.
Effect of iron deficiency anemia on levels of hemo 79 Sherwood GB, Paschal RD and Adamski J (2016)
globin A1c in nondiabetic patients. Acta Haematol, Rasburicase‐induced methemoglobinemia: case
112, 126–128. report, literature review, and proposed treatment
65 Moriwaki Y, Yamamoto T, Shibutani Y, Harano T, algorithm. Clin Case Rep, 4, 315–319.
Takahashi S and Hada T (2000) Abnormal haemo 80 Morris A, Bain BJ, Atta M and Layton DM (2017) A
globins, Hb Takamatsu and Hb G‐Szuhu, detected puzzling case of methemoglobinemia. Am J Hematol,
during the analysis of glycated haemoglobin 92, 1103–1104.
(HbA1C) by high performance liquid chromatogra 81 Sin DD and Shafran SD (1996) Dapsone‐ and pri
phy. J Clin Pathol, 53, 854–857. maquine‐induced methemoglobinemia in HIV‐
66 Kabadi UM, Gopal V, Hood L, Kabadi MU and Platt infected individuals. J Acquir Immune Defic Syndr
K (1992) Elevated glycosylated hemoglobin concen Hum Retrovirol, 12, 477–481.
trations in AIDS. AIDS, 6, 236–238. 82 Hohl RJ, Sherburne AR, Feely JE, Huisman THJ and
67 Ernst A and Zibiak JD (1998) Carbon monoxide poi Burns CP (1998) Low pulse oximeter‐measured
soning. N Engl J Med, 339, 1603–1608. hemoglobin oxygen saturations with hemoglobin
68 Blumenthal I (2001) Carbon monoxide poisoning. Cheverly. Am J Hematol, 59, 181–184.
J R Soc Med, 94, 270–272. 83 Allen A, Fisher C, Premawardhena A, Bandara D,
69 Nagel RL and Jaffé ER. CO‐, NO‐, met‐, and sulf‐hemo Perera A, Allen S et al. (2012) Methemoglobinemia
globinemias: the dyshemoglobinemias. In: Steinberg and ascorbate deficiency in hemoglobin E β
346 Chapter 6
thalassemia: metabolic and clinical implications. 97 Kaczorowska‐Hac B, Stefanowicz J, Stachowicz‐

Blood, 120, 2939–2944. Stencel T, Kozlowska M, Adamkiewicz‐Drozynska
84 Hurford WE and Kratz A (2003) Case 23‐2004: a 50‐ E and Balcerska A (2012) Methemoglobinemia in
year‐old woman with low oxygen saturation. N postchemotherapy stomatitis topical treatment: 2
Engl J Med, 351, 380–387. pediatric cases. J Pediatr Hematol Oncol, 34, 137–139.
85 Gourdin D, Vergnes H and Gutierez N (1975) 98 Aryal MR, Gupta S, Giri S and Fraga JD (2013)
Methaemoglobin in man living at high altitude. Br J Benzocaine‐induced methaemoglobinaemia: a
Haematol, 29, 243–246. life‐threatening complication after a transoesoph
86 AgarwalN,NagelRLandPrchalJ.Dyshemoglobinemias. ageal echocardiogram (TEE). BMJ Case Rep,
In: Steinberg MH, Forget BG, Higgs DR and Weatherall doi:10.1136/bcr‐2013‐009398.
DJ, eds. Disorders of Hemoglobin: Genetics, Pathophysiology 99 Finan A, Keenan P, O’Donovan FO, Mayne P and
and Clinical Management. Cambridge University Press, Murphy J (1998) Methaemoglobinaemia associ
Cambridge, 2009, pp. 607–622. ated with sodium nitrite in three siblings. Br Med J,
87 Janssen WJ, Dhaliwal G, Collard HR and Saint S 317, 1138–1139.
(2005) Clinical problem‐solving. Why “why” 100 Chan TK, Mak LW and Ng RP (1971)
matters. N Engl J Med, 351, 2429–2434. Methemoglobinemia, Heinz bodies and acute
88 Foltz LM, Dalal BI, Wadsworth LD, Broady R, Chi massive intravascular hemolysis in Lysol poison
K, Eisenhuer E et al. (2006) Recognition and man ing. Blood, 38, 739–744.
agement of methemoglobinemia and hemolysis in a 101 Ahmed N, Hoy BPS and McInerney J (2010)
G6PD‐deficient patient on experimental anticancer Methaemoglobinaemia due to mephedrone (‘snow’).
drug Triapine. Am J Hematol, 81, 210–211. BMJ Case Rep, doi:10.1136/bcr.04.2010.2879.
89 Centers for Disease Control and Prevention (CDC) 102 Goluboff N and Wheaton R (1961) Methylene blue
(2012) Severe methemoglobinemia and hemolytic induced cyanosis and acute hemolytic anemia
anemia from aniline purchased as 2C‐E (4‐ethyl‐2,5‐ complicating the treatment of methemoglobine
dimethoxyphenethylamine), a recreational drug, mia. J Pediat, 58, 86–89.
on the Internet – Oregon, 2011. MMWR Morb Mortal 103 Williamson K, Htet N and Nanini S (2016)
Wkly Rep, 61, 85–88. Methemoglobinemia from prolonged therapeutic
90 Umbreit J (2007) Methemoglobin – it’s not just blue: use of phenazopyridine. Ann Clin Case Rep, 1, 1085.
a concise review. Am J Hematol, 82, 134–144. 104 Mahomedy MC, Mahomedy YH, Canham PA,
91 Swartzentruber GS, Yanta JH and Pizon AF (2015) DowningJWandJealDE(1975)Methaemoglobinaemia
Methemoglobinemia as a complication of topical following treatment dispensed by witch doctors.
dapsone. N Engl J Med, 372, 491–492. Two cases of potassium permanganate poisoning.
92 Coughlin K, Flibotte J, Cahill AM, Osterhoudt K, Anaesthesia, 30, 190–193.
Hedrick H and Vrecenak J (2019) Methemoglobinemia 105 Kizer N, Martinez E and Powell M (2006) Report
in an infant after sclerotherapy with high‐dose dox of two cases of rasburicase‐induced methemo
ycycline. Pediatrics, 143, e20181642. globinemia. Leuk Lymphoma, 47, 2648–2650.
93 Leunbach TL, Pedersen JF, Trydal T, Thorgaard P, 106 Viallon A, Page Y and Bertrand JC (2000)
Helgestad J and Rosthøj S (2014) Acute favism: Methemoglobinemia due to riluzole. N Engl J Med,
methemoglobinemia may cause cyanosis and low 343, 665–666.
pulse oximetry readings. Pediatr Hematol Oncol, 31, 107 Jayaprakash K (2003) Acquired methaemoglobinemia
104–106. (MetHb) in goldsmiths – a hitherto unobserved occu
94 Kouides PA, Abboud CN and Fairbanks VF (1996) pational hazard. Indian J Occup Environ Med, 7, 16–18.
Flutamide‐induced cyanosis refractory to methyl 108 Baskin SI, Horowitz AM and Nealley EW (1992)
ene blue therapy. Br J Haematol, 94, 73–75. The antidotal action of sodium nitrite and sodium
95 Mitzumani T and Hojo M (2011) Severe hypoxaemia thiosulfate against cyanide poisoning. J Clin
due to methaemoglobinaemia and aspiration pneu Pharmacol, 32, 368–375.
monia. BMJ Case Rep, doi:10.1136/bcr.03.2011.3937. 109 Farkas AN, Scoccimarro A and Pizon AF (2017)
96 Donnelly GB and Randlett D (2000) Methemo Methemoglobinemia due to antifreeze ingestion.
globinemia. N Engl J Med, 343, 337. N Engl J Med, 377, 1993–1994.
110 Cummings MH, Day S, Norton S, Cook G and treatment with chronic graft‐versus‐host disease.
Alexander WD (1994) Petrol octane booster and Blood, 92, 4872–4873.
methemoglobinemia. J Intern Med, 235, 279–280. 117 Levine D, Brunton AT, Kruger A and Hersant M
111 Zinkham WH and Oski FA (1996) Henna: a poten (2000) Recurrent sulphaemoglobinaemia treated
tial cause of oxidative hemolysis and neonatal with neomycin. J R Soc Med, 93, 428.
hyperbilirubinemia. Pediatrics, 97, 707–709. 118 Brugnara C and Platt OS, The neonatal erythro
112 Nakajima W, Ishida A, Arai H and Takada G cyte and its disorders. In: Nathan DG, Orkin SH,
(1997) Methaemoglobinaemia after inhalation of Ginsburg D and Look AT, eds. Nathan and Oski’s
nitric oxide in infant with pulmonary hyperten Hematology of Infancy and Childhood, 6th edn.
sion. Lancet, 350, 1002–1003. Saunders, Philadelphia, 2003, pp. 19–55.
113 Sanders P and Faunt J (1997) An unusual cause of 119 Flexman AM, Del Vicario G and Schwarz KW
cyanosis (isosorbide dinitrate induced methaemo (2007) Dark green blood in the operating theatre,
globinaemia), Aust NZ J Med, 27, 596. Lancet, 369, 1972.
114 Bacon R (1997) Nitrate preserved sausage meat 120 Tangerman A, Bongaerts G, Agbeko R,
causes an unusual food poisoning incident. Semmekrot B and Severijnen R (2002) The origin
Commun Dis Rep, 77, R45–47. of hydrogen sulfide in a newborn with sulfhae
115 Hogan CJ, Wiley JS and Billington T (1997) moglobin induced cyanosis. J Clin Pathol, 55,
Intravascular haemolysis complicating treatment 631–633.
of acute promyelocytic leukaemia. Aust NZ J Med, 121 Murphy K, Ryan C, Dempsey EM, O’Toole PW,
27, 450–451. Ross RP, Stanton C and Ryan CA (2015) Neonatal
116 Moreira V de A, de Medeiros BC, Bonfim CMS, sulfhemoglobinemia and hemolytic anemia asso
Pasquini R and de Medeiros CR (1998) ciated with intestinal Morganella morganii.
Methemoglobinemia secondary to clofazimine Pediatrics, 136, e1641–e1645.
6.1 (a) T 6.3 (a) T 6.5 (a) T 6.7 (a) T 6.9 (a) F 6.11 (a) T
(b) T (b) F (b) F (b) F (b) T (b) T
(c) F (c) T (c) T (c) T (c) F (c) F
(d) F (d) F (d) T (d) T (d) F (d) T
(e) T (e) T (e) T (e) T (e) F (e) F
6.2 (a) F 6.4 (a) T 6.6 (a) F 6.8 (a) F 6.10 (a) T 6.12 (a) T
(b) F (b) T (b) T (b) F (b) F (b) F
(c) T (c) F (c) T (c) T (c) T (c) F
(d) T (d) T (d) F (d) T (d) F (d) T
(e) T (e) T (e) F (e) F (e) T (e) T
7 Organisation of a haemoglobinopathy
diagnostic service
documenting ethnic origin, as in France [1], can

General principles make provision of a rational service more difficult.
The organisation of a haemoglobinopathy service It is essential to know if the patient has been trans-
depends on the ethnic mix of the population served fused in the recent past. Only with this information
and also on whether the service is hospital based or will the testing be clinically relevant and appropri-
a regional scheme, whether antenatal and neonatal ately interpreted. This chapter will discuss mainly
screening is required, and whether children are the provision of a hospital‐based service in a multi‐
included in those tested. It is essential that laborato- ethnic community in which the population to be
ries have clearly defined written protocols that are tested includes neonates and children and also
followed. These will differ, depending on the pregnant women and their partners, but testing in
patient population, the nature of the service that is a resource‐poor setting and point‐of‐care testing
required and the technology employed. Protocols will also be considered. In laboratories serving
will differ according to the primary laboratory populations of a single ethnic origin, the laboratory
technique employed, whether this be haemoglobin procedures will be adapted to local requirements.
electrophoresis, high performance liquid chroma- All abnormal results of haemoglobinopathy test-
tography (HPLC), capillary electrophoresis or isoe- ing should be conveyed to the patient, or in the case
lectric focusing (IEF). It is important to remember of a child to the patient’s parents, in written form
that in a diagnostic laboratory the identification of a together with an appropriate explanation. To avoid
variant haemoglobin is often presumptive rather unnecessary repeat testing, it can also be useful to
than definitive and that it is not possible, or indeed give the patient written confirmation of significant
clinically relevant, to identify every variant haemo- negative results, for example that haemoglobin S is
globin detected. The protocols that are followed not present. Haemoglobinopathy cards are a useful
should ensure that the great majority of clinically way to give patients documentation of results.
relevant disorders of haemoglobin synthesis are These should give the patient’s name, the date of
detected. It is not realistic to hope that all relevant testing, the test results and, if necessary, an interpre-
abnormalities will be detected. For example, in tation. It can also useful to include the haemoglobin
antenatal screening, silent β thalassaemia will concentration (Hb) and the mean cell haemoglobin
generally be missed. Laboratory tests for haemo- (MCH) or the mean cell volume (MCV). In the case
globinopathies should not be performed in a of normal results in patients with microcytosis from
knowledge vacuum. It is essential to know the age ethnic groups in which α thalassaemia trait occurs,
of the patient, the ethnic origin, the red cell indices it is useful to add a statement “This test does not
and the clinical features, if any. Legal restraints on exclude α thalassaemia trait”.

348
Organisation of a haemoglobinopathy diagnostic service 349
Antenatal and preconceptual pregnant. Tests must therefore be performed rap-

haemoglobinopathy/thalassaemia idly in order to permit testing of the partner and
screening and fetal diagnosis offering of fetal diagnosis and termination of
pregnancy, when appropriate, while the pregnancy
Antenatal and preconceptual testing is still in its early stages. The UK aim is to complete
screening of the mother by 10 weeks gestation,
In most Western countries, the aim of antenatal to be able to offer prenatal diagnosis to couples
haemoglobinopathy screening is to detect inherited by 12 weeks and to carry out prenatal diagnosis by
abnormalities in the parents that might lead to such 12 weeks + 6 days. Screening should be repeated in
severe disease in the fetus that termination of preg- subsequent pregnancies. Fetal diagnosis carries a
nancy would be justified and might be requested by small risk of inducing miscarriage of a fetus, possi-
the parents. In some instances, specifically when bly a normal fetus. For this reason, fetal diagnosis
haemoglobin Bart’s hydrops fetalis is predicted in should generally only be undertaken when the
the fetus, continuing the pregnancy may also put potential parents wish to consider termination of
the mother’s own health and even life at risk. pregnancy. Counselling of parents should be based
In countries where termination of pregnancy is on giving very full information on the likely
unacceptable or restricted, different approaches to outcome of pregnancy and the likely severity of any
reducing the prevalence of serious haemoglobi- fetal disease. It may be useful for potential parents
nopathies are taken. to be referred to relevant patient support groups.
In some countries with a high prevalence of Counselling should be non‐directive.
disorders of globin chain synthesis, screening is
The conditions that should be predicted in a fetus
usually performed before pregnancy is undertaken. are shown in Table 7.1 and the abnormalities that
In countries with adequate resources, this can be should therefore be detected in the mother in
achieved by population screening, as is carried out, Table 7.2. It follows that when one of these abnor-
for example, in Cyprus and, more recently, in Saudi malities is detected in the mother, appropriate test-
Arabia and the United Arab Emirates [2], where ing of the potential father will follow. Table 7.3
screening must be carried out before a marriage cer- shows conditions that are generally mild and for
tificate is issued. In high prevalence areas, screening which prediction is not generally considered neces-
of adolescents can also be effective [3]. In some sary. Tables 7.1 and 7.3 represent the consensus
countries, marriage is strongly discouraged and view in the UK [7] but there are some areas of con-
marriage licences are refused if both partners are troversy. Severe haemoglobin H disease is one such
found to have β thalassaemia heterozygosity [4]. In
others, most affected couples marry but prenatal
diagnosis and termination of an affected pregnancy Table 7.1 Haemoglobinopathies of such severity that
is often chosen [5]. The percentage of couples who the likelihood of their occurrence in a fetus should
separate when both are found to be carriers varies be predicted.
greatly between countries, from less than 5% to Haemoglobin Bart’s hydrops fetalis
approaching 60%.
β thalassaemia major and intermedia including that
In countries with fewer economic resources it may
resulting from β thalassaemia/haemoglobin E compound
be possible to make testing cost effective by target- heterozygosity
ing individuals most in need of testing. For example,
Sickle cell disease
in countries such as Pakistan where consanguineous
Sickle cell anaemia
marriages are common and specific disorders of Sickle cell/haemoglobin C disease
globin chain synthesis such as β thalassaemia tend Sickle cell/β thalassaemia
to segregate in families, screening can be targeted on Sickle cell/δβ thalassaemia
extended families with an index case [6]. Sickle cell/haemoglobin Lepore
In many countries screening is commonly under- Sickle cell/ haemoglobin D‐Punjab
taken only when the potential mother is already Sickle cell/ haemoglobin O‐Arab
350 Chapter 7
Table 7.2 Disorders of globin chain synthesis that should testing causes anxiety, particularly if a woman is
be detected in prospective parents in order to predict already pregnant. In addition, the cost of genetic
the occurrence of severe disease in offspring.* testing may be considerable and no health service
has unlimited resources. This must always be borne
α0 thalassaemia trait or haemoglobin H disease
β thalassaemia trait in mind when drawing up protocols. It may not be
δβ thalassaemia trait justifiable to test a large number of couples for rare
Haemoglobin Lepore disorders in order to attempt to identify a very
Haemoglobin E small proportion of patients with a significant
Haemoglobin S abnormality. Screening for α0 thalassaemia provides
Haemoglobin C
an example of where zeal should be tempered by
Haemoglobin D‐Punjab
consideration of what is reasonable. This condition
Haemoglobin O‐Arab
Unstable haemoglobins does occur in native British, Afro‐Caribbeans and
Indians but it is very uncommon in all these ethnic
* Most cases of thalassaemia intermedia and sickle cell disease groups. Both Afro‐Caribbeans and Indians have a
will already have been diagnosed but occasionally those with high percentage of α+ thalassaemia and screening
quite mild disease will be detected only during pregnancy. UK
the large number of individuals with microcytosis
guidelines suggest hereditary persistence of fetal haemoglobin
should also be detected. that is likely to be attributable to heterozygosity
or homozygosity for α+ thalassaemia in the hope of
identifying rare individuals with α0 thalassaemia
Table 7.3 Less severe haemoglobinopathies for which
trait is not usually considered justifiable. Individual
the prediction of the condition in a fetus is not usually
circumstances may dictate that certain couples are
considered essential.
tested; for example, if there is consanguinity, if they
Haemoglobin H disease are native British originating in Lancashire or
Mild sickling conditions Cheshire or if they are of Caribbean origin and may
Sickle cell/haemoglobin E compound heterozygosity have Chinese ancestry.
Sickle cell/deletional hereditary persistence of fetal
In antenatal screening, if one partner is found to
haemoglobin compound heterozygosity
have β thalassaemia heterozygosity and the other
Haemoglobin E homozygosity
Haemoglobin C homozygosity to have α0 heterozygosity (or haemoglobin H
disease) the partner with β thalassaemia heterozy-
gosity should be specifically tested to exclude coex-
area. In countries or ethnic groups where there is a isting α0 thalassaemia. The coexistence of the two
high prevalence of both α0 and αTα or other α+ tha- has been found in 8–9% of Hong Kong Chinese, in
lassaemia heterozygosity, prediction of haemoglo- 4% of Chinese in Guangong province of China and
bin H disease can be attempted. However, it should in 8% of Chinese in Malaysia [8]. The same applies
be noted that haemoglobin H disease is often mild to haemoglobin E heterozygotes; the possibility of
and this diagnosis would not generally be consid- coexisting α0 thalassaemia must be considered (and
ered an indication for termination of pregnancy. will be particularly suspected if the haemoglobin E
The prediction of thalassaemia intermedia is also a percentage is lower than otherwise expected).
difficult area since this condition varies greatly in It should also be noted that red cell indices can be
severity and the severity of the phenotype associ- misleading in homozygosity for haemoglobin
ated with a specific genotype is not always predict- Constant Spring. In a series of 20 patients, the MCH
able. This uncertainty must be conveyed to was not below 26 pg in five subjects and in three it
prospective parents. Tables 7.1 and 7.3 are not was not below 25 pg [9]; the MCV was never below
exhaustive but cover the great majority of likely 80 fl. If screening for haemoglobin Quong Sze is to
clinical situations. Parents with rare abnormalities, be done (for prediction of non‐deletional haemo-
for example an unstable haemoglobin, need to be globin H disease), an MCH of less than 27 pg is an
assessed individually. In any antenatal screening effective threshold, whereas screening by means of
programme it should be remembered that genetic the MCV is ineffective [10].
Assess family origin

Perform FBC and HPLC or CE
Hb F ≥5%, Hb A2 ≥3.5% Hb A2 normal, Variant haemoglobin

Hb A2 not and MCH MCH <25 pg present
increased <27 pg
MCH <25 pg MCH ≥25 pg

Diagnose β
thalassaemia trait
If family origin
Consider
appropriate, assess for Perform sickle solubility
MCH and
α0 thalassaemia trait test or electrophoresis, as
assess for
possible δβ appropriate, to identify
thalassaemia haemoglobin S, C,
of HPFH* D-Punjab, E, Lepore or
O-Arab
Fig. 7.1 Flow chart for universal antenatal screening for variant haemoglobins and α, β and δβ thalassaemias in a high
prevalence area or country, using high performance liquid chromatography (HPLC) or capillary electrophoresis (CE) as
the primary method. FBC, full blood count; Hb A2, haemoglobin A2; Hb F, haemoglobin F; HPFH, hereditary persistence
of fetal haemoglobin; MCH, mean cell haemoglobin. *Hb F ≥5% and <10% more likely to be δβ thalassaemia, HbF ≥10%
more likely to be HPFH.
Uptake of antenatal diagnosis is dependent on A protocol based on HPLC or capillary electro-

cultural factors and on the severity of the condition phoresis for the identification of variant haemo-
that is predicted. In an audit of prenatal diagnosis globins and β thalassaemia trait, applicable both to
for haemoglobinopathies in the UK in 1998 the universal antenatal screening and in other circum-
uptake was 90% among Cypriots, 41 rising to 65% stances, is shown in Fig. 7.1 and a similar protocol,
among Indians, 26% among Pakistanis and 9% which is applicable when cellulose acetate electro-
among Bangladeshis [11]. Uptake is also higher in phoresis is the primary diagnostic tool, in Fig. 7.2.
high prevalence areas of the UK, possibly because In England, universal screening is applied in high
of the availability of well‐informed doctors and prevalence areas (screening tests positive in ≥2% of
counsellors. In general, uptake is higher when β antenatal population) with a selective policy being
thalassaemia major is predicted than when sickle applied in low prevalence areas (screening tests
cell disease is predicted. Uptake is almost universal positive in <1%); some discretion is applicable
when haemoglobin Bart’s hydrops fetalis is pre- between these cut‐off points. Figure 7.3 shows an
dicted, although it should be noted that occasional algorithm applicable in a low prevalence area. The
parents are choosing diagnosis followed by intrau- action limits suggested for the diagnosis of β thalas-
terine transfusion rather than termination of saemia trait have been validated [12]. In the UK,
pregnancy. deoxyribonucleic acid (DNA) sequencing of both α
Perform FBC and electrophoresis
on cellulose acetate at alkaline pH
Hb F ≥5%,* MCH <27 pg, MCH <25 pg Variant haemoglobin

Hb A2 not Hb F normal or present
increased only slightly
increased
Measure If family origin

appropriate, assess Perform sickle solubility
Hb A2‡
Consider MCH for α0 thalassaemia test or electrophoresis at
and assess for trait regardless of acid pH, as appropriate,
possible δβ other findings to identify variant
thalassaemia or Hb A2 ≥3.5% haemoglobins S,C,D-
HPFH† Punjab, E, Lepore and
O-Arab
Diagnose β
thalassaemia trait
Fig. 7.2 Flow chart for universal antenatal screening for variant haemoglobins and α, β and δβ thalassaemias in a high
prevalence area or country, using cellulose acetate electrophoresis as the primary method. *An apparent increase can be
confirmed by a 2‐minute alkali denaturation test. †Hb F ≥5% but < 10% is more likely to be δβ thalassaemia; Hb F ≥10%
is more likely to be HPFH. In the UK antenatal screening programme, the action value for haemoglobin F is 10% if the
MCH is 27 pg or higher and 5% if the MCH is less than 27 pg. ‡Haemoglobin A2 is quantified by elution.

Perform FBC
MCH <25 pg MCH <27 pg MCH ≥27 pg
Perform HPLC
Consider family or CE to
origin and if quantify Hb A2
(and detect If family origin
appropriate,
appropriate, perform
investigate for α0 any variant
haemoglobin) HPLC or CE to
thalassaemia,
investigare for
whether or not Hb
haemoglobin S,C,
A2 is increased or a Hb A2 D-Punjab, E, Lepore
variant ≥3.5% or O-Arab
haemoglobin is
present Fig. 7.3 Flow chart for
Diagnose β selective antenatal testing
thalassaemia in a low prevalence area or
country.
and β globin genes is now employed for definitive to the antenatal clinic; (v) failure to diagnose β
antenatal screening. The cut‐off point adopted in thalassaemia heterozygosity with a normal haemo-
the UK for α0 thalassaemia screening is an MCH of globin A2 because of coexisting δ thalassaemia or a δ
less than 25 pg. A study in Hong Kong indicated chain variant, in the latter instance due to failure to
that about 1.8% of individuals with − −SEA/ (2/110) sum the percentages of A2 and the A2 variant; (vi)
would be missed by this strategy [13]; a cut‐off failure to consider the possibility of α0 thalassaemia
point of 27 pg for the MCH or 80 fl for the MCV is in individuals with β thalassaemia heterozygosity or
therefore preferred in Hong Kong. In countries with haemoglobin E heterozygosity or homozygosity;
a high prevalence and heterogeneity of thalassae- (vii) failure to predict β thalassaemia intermedia
mias and haemoglobinopathies and with limited because of silent or near silent thalassaemia; (viii)
resources, other screening policies are appropriate failure to test for α0 thalassaemia when the red cell
(see later). indices are normal because of coinheritance of α0
When testing of a pregnant woman reveals a thalassaemia and quadruple α [14]. An extra prob-
potentially significant condition it is useful to issue, lem that must be avoided in prenatal screening is
with the test result, a proforma that can be followed stigmatisation of those who are found to be carriers
by antenatal clinic staff during the further manage- of a significant haemoglobinopathy.
ment of the patient. This helps to ensure correct
patient management and can also be used for audit
Fetal diagnosis
purposes.
In antenatal screening and testing, it is essential Fetal diagnosis can be carried out by chorionic villus
to know if a conception is the result of use of a sampling (from 11 weeks, usually at 11–14 weeks) or
donor egg or a donor sperm. Thalassaemia major, on cells obtained by amniocentesis (from 15 weeks,
haemoglobin S/β thalassaemia and haemoglobin H usually at 15–20 weeks). When amniocentesis is
disease have all resulted from the use of donor eggs used, back‐up cells are cultured in case not enough
or sperm. It is also important to know if either part- DNA is obtained; their growth takes 10–14 days so
ner is of unknown ethnic origin, for example, that there may be delay in diagnosis. The aim of the
because of adoption. In women who become preg- UK scheme is to achieve fetal diagnosis by 12 weeks
nant despite having had an allogeneic stem cell plus 6 days of gestation.
transplant, it is the genetic characteristics that are It may be possible to avoid an invasive test to
important; the usual haemoglobinopathy screening exclude haemoglobin Bart’s hydrops fetalis by use
result will be misleading. of ultrasound assessment of the fetus. If the middle
In countries that have regulations governing cerebral artery shows a normal peak systolic veloc-
DNA analysis these should be followed if testing is ity then the fetus is not significantly anaemic and
DNA based. It is particularly important that very hydrops can be excluded.
full information about the implications of testing is Assisted reproduction technology with pre‐
given when family studies are to be undertaken implantation diagnosis of an embryo is also possi-
since these sometimes reveal non‐paternity. This is ble when termination of pregnancy would not be
necessary for protein‐based testing as well as for acceptable.
DNA‐based testing.
Problems and pitfalls fetal diagnosis
Problems and pitfalls in antenatal diagnosis
Problems in fetal diagnosis include: (i) those arising
There are numerous problems and pitfalls in antena- from late booking at an antenatal clinic and delayed
tal diagnosis. These include: (i) late booking for diagnosis; (ii) miscarriage as a result of the diagnos-
antenatal care; (ii) non‐availability of partner; (iii) tic procedure; and (iii) illegality of termination of
lack of knowledge of ethnic origin; (iv) failure to pregnancy in some jurisdictions, sometimes result-
assess sperm and ovum donors for significant car- ing in pregnant women travelling to another
rier states or non‐disclosure of assisted reproduction country for a termination.
354 Chapter 7
programme [16]. Neonatal screening can be per-

Neonatal screening
formed on a cord blood sample (taken by syringe
In most countries the purpose of neonatal screen- and needle or evacuated container and needle from
ing is principally the detection of sickle cell disease, an umbilical cord vessel after carefully wiping any
since early diagnosis has been shown to reduce maternal blood from the surface of the cord), an
mortality during infancy and early childhood. anticoagulated capillary blood sample, a dried spot
However, β thalassaemia major can also be recog- of capillary blood (e.g. on a Guthrie card) or, in sick
nised or suspected; for example, between 2010 and hospitalised babies, a venous sample (Table 7.4).
2016 the screening programme in England led to Cord blood should not be squeezed from the end of
the detection of 113 cases of thalassaemia in addi- the cut cord because of the risk of contamination
tion to 1315 cases of sickle cell disease [15]; the with maternal blood. A capillary sample obtained
policy is for follow‐up to be offered if haemoglobin by heel prick, blotted on to filter paper and permit-
A is 1.5% or less. Different aims may be appropriate ted to dry, is usually most convenient. Often one
in countries without a significant incidence of blood spot on a Guthrie card (used for diagnosis of
sickle cell disease; in Thailand a neonatal screening congenital metabolic diseases and, in some African
programme aiming to detect β0β0 thalassaemia, β0βE countries, for screening for human immunodefi-
and haemoglobin H disease was considered to ciency virus infection) is used for this purpose.
have potential to not only enhance the care of When such a dried blood spot is used, a circle of
babies but also to improve the antenatal screening blood‐impregnated filter paper is punched out and
Table 7.4 Methods applicable to neonatal haemoglobinopathy diagnosis.
Advantages and disadvantages
Method of obtaining blood

By needle and syringe from cord vessel Contamination with maternal blood can occur if technique is not
meticulous; inadvertently obtaining a sample after a baby has been
transfused is generally avoided; there may be more certainty of testing
every neonate if a cord blood sample is used
Heel prick sample into a heparinised Transport and labelling of the sample is more difficult but the sample
capillary tube does not suffer the dilution that occurs when a dried blood spot is eluted;
contamination by maternal blood is avoided but it is important to avoid
inadvertently taking a sample after a blood transfusion has been given
Heel prick sample blotted onto filter paper Transport and labelling is easy but dried samples are more likely to be
and dried denatured giving blurred bands or peaks; contamination by maternal
blood is avoided but it is important to avoid inadvertently taking a
sample after a blood transfusion has been given; the blood sample can
be obtained at the same time as sampling for metabolic testing, e.g. for
phenylketonuria, i.e. at day 5 (or earlier if the baby is about to be
transfused)
Method of detecting variant haemoglobins
High performance liquid chromatography Very sensitive technique; variant haemoglobins are quantified
Isoelectric focusing Very sensitive technique
Capillary electrophoresis Sensitive technique
Cellulose acetate electrophoresis An eluate from a Guthrie spot may be too dilute for this technique; less
sensitive to low concentrations of a normal or a variant haemoglobin
Mass spectrometry Applicable only in specialised centres

the haemoglobin is eluted. A Guthrie spot can also is delay in testing [18]. In selecting a method for
be used for DNA analysis. This is useful if a baby neonatal screening, consideration must be given to
has been transfused and greatly reduces the num- the workload and to whether it is convenient to use
ber of babies that have to be recalled for repeat the same technique both for neonatal screening and
blood sampling. for other screening. A consideration of both the
Universal neonatal screening has been recom- requirement for sensitivity and the need to use the
mended when more than 15% of neonates are born same instrumentation for other purposes means
to ethnic minority mothers [17]. In the UK there is that HPLC is often the technique chosen. When
now universal neonatal screening for sickle cell dis- mass spectrometry is already in use for screening
ease, using a dried blood spot taken ideally at day 5 for metabolic disorders, it may be chosen. In the
and no later than day 8, the day of birth being day 0 UK, the selected methods detect haemoglobins S, C,
[18–20]. UK screening aims to detect the various D‐Punjab, O‐Arab and E; probable β thalassaemia
forms of sickle cell disease and also haemoglobin S/ major is also reported and further investigated [24].
unknown variant and suspected S/hereditary per- The use of sensitive techniques is particularly
sistence of fetal haemoglobin (HPFH). In addition, important in screening premature babies in whom
results are reported if there is haemoglobin C only, the percentage of haemoglobin A or of any variant
D only, E only or CD, CE or DE. Heterozygotes for haemoglobin is likely to be low. Screening of these
these variants and for O‐Arab can be reported. babies should be done soon after birth to avoid the
Suspected β thalassaemia is reported and is further possibility of a sample being taken after a blood
investigated. Universal screening is also the transfusion. The presence of more haemoglobin A
practice in the Netherlands and Spain; Belgium has than F or a prominent haemoglobin A2 band in a
universal screening in two regional centres and neonatal sample should raise the possibility that
France has targeted screening in metropolitan there has been contamination with maternal blood
France and universal screening in overseas territo- or that a post‐transfusion sample has been sent to
ries [20]. Sickle cell disease is also the target of the laboratory.
newborn screening in Greece. Universal screening All haemoglobin variants detected by the initial
is US policy. In some circumstances, considerations screening method should be further investigated by
of cost may dictate selective screening. Screening a supplementary alternative method to make their
can be performed before the baby leaves hospital in presumptive identification more reliable. It should
a hospital‐based scheme or together with screening be noted that a sickle solubility test should not be
for inherited metabolic defects in a community‐ used in neonates because of the high probability of
based scheme. The most suitable techniques are false negative results. Potentially significant hae-
those that are sensitive and can be performed with moglobinopathies should be confirmed by a second
small blood samples (e.g. HPLC, capillary electro- sample, conveniently around the age of 6–7 weeks
phoresis, IEF and tandem mass spectrometry) (testing to be completed by 8 weeks so there is no
[15, 21–24]. A suitable protocol is shown in Fig. 7.4. delay in commencing prophylactic penicillin).
Cellulose acetate electrophoresis can also be used Repeat testing should also be performed on all
but the eluate of a dried blood spot may be too babies whose initial sample showed no haemoglo-
dilute for this technique and, in addition, it is less bin A. In addition, it is prudent to repeat tests if the
sensitive than either HPLC or IEF for the detection predominant haemoglobin present is haemoglobin
of the low percentages of haemoglobin A that may F with very small amounts of haemoglobins A and
be present in premature neonates. It is also less sen- S, since it can be difficult to d istinguish sickle cell
sitive for the detection of small amounts of variant trait from sickle cell/β+ thalassaemia in this circum-
haemoglobins, levels down to 4% being detected in stance. The initial report on such a sample should
one study, whereas HPLC and IEF detected haemo- be circumspect. The detection of only haemoglobins
globin variants down to levels of less than 2% [22]. S and F in a neonate is most often attributable to
Haemoglobin Bart’s is unstable on storage and may sickle cell anaemia. However, compound heterozy-
therefore be missed, regardless of technique, if there gosity for haemoglobin S and either β0 thalassaemia
356 Chapter 7
Obtain capillary sample by heel prick on to

Guthrie card and allow to dry for transport
In the laboratory, punch a circle of blood-impregnated filter

paper from the Guthrie card and prepare an eluate
Apply primary method, either HPLC, IEF, CE or MS
No Abnormality
abnormality present
No further Punch a further circle from the same

action card and apply a second
confirmatory method
Abnormal result
Normal result or diagnosis not
straightforward
Report result Carry out appropriate

confirmatory testing
at 8–12 weeks of age
Fig. 7.4 A protocol for neonatal screening. IEF, isoelectric focusing; MS, mass spectrometry. If the baby has been
transfused in utero or before a blood sample has been taken, follow an alternative policy of DNA analysis for the βS gene.
or deletional HPFH also produces this pattern. In An essential part of any neonatal screening pro-
addition, it has been noted that some babies with gramme is a well‐organised system of follow‐up
sickle cell/β+ thalassaemia compound heterozygo- and appropriate management of babies found to
sity also have only S and F detectable at birth, have a significant abnormality. Information and an
particularly if
cellulose acetate electrophoresis is appropriate explanation must also be given to the
the detection method employed [25]. Studies on parents of babies found to have sickle cell trait or
parental samples can help distinguish S/HPFH other heterozygous conditions. It is important for
from clinically significant conditions and avoid all those involved in neonatal screening schemes to
unnecessary follow‐up and further testing of these remember that β thalassaemia trait will not be
babies [26]. detected by testing in neonates or young infants.
If it is known that one or both parents has β thalas- are planned, so prudent practice is to test all patients
saemia trait then this should be borne in mind when of an appropriate ethnic group.
issuing a report. In testing for haemoglobin S it is necessary to
bear in mind the very wide range of ethnic groups
in which this variant haemoglobin can occur
Problems and pitfalls in neonatal
(see Chapter 4). If routine surgery is being planned,
diagnosis
all patients of an appropriate ethnic group should
Problems and pitfalls in neonatal diagnosis include: have an FBC and HPLC, capillary electrophoresis
(i) baby not tested because of arrival in a country as or haemoglobin electrophoresis, supplemented,
a neonate or in early infancy; (ii) diagnosis missed when a relevant abnormality is detected, by a sickle
when testing is selective rather than universal; (ii) solubility test. If emergency anaesthesia is required,
intrauterine or neonatal transfusion prior to test; the patient should be assessed for clinical features
(iii) follow‐up testing not done; and (iv) parents not suggestive of an undiagnosed sickling disorder. If
informed of carrier state. there are no such features, an FBC and sickle solu-
Although it is desirable to inform parents when a bility test should be performed; if the Hb is reduced,
carrier state is detected, in some European coun- a blood film should be examined. If the patient has
tries, including Germany and Switzerland, this is clinical features compatible with a sickling disor-
not currently (2018) permitted [20]. der, an FBC, blood film and sickle solubility test
should be performed. The purpose of the blood
film in these circumstances is to facilitate the diag-
Pre‐anaesthetic testing
nosis of patients with sickle cell disease with a nor-
It is important to detect all patients with sickle cell mal Hb who might otherwise be assumed to have
disease before anaesthesia in order to ensure that sickle cell trait. When resources permit, a definitive
the necessity for preoperative blood transfusion is diagnosis of sickle cell disease should be made
considered and that the patient is kept warm, well prior to surgery. This is more likely to be feasible in
hydrated and well oxygenated both during surgery laboratories using HPLC or capillary electrophore-
and in the post‐operative period. Although patients sis as the primary diagnostic method rather than
with sickle cell anaemia and other severe sickling haemoglobin electrophoresis. If a definitive diag-
disorders will usually already have been diagnosed nosis cannot be made rapidly, but sickle cell dis-
prior to presenting with a condition requiring ease is suspected, surgery should proceed on the
immediate surgery and anaesthesia, this is not nec- assumption that the patient does have sickle cell
essarily so in patients with sickle cell/haemoglobin disease and appropriate attention should be paid
C disease and sickle cell/β+ thalassaemia. Such to oxygenation and hydration. When full testing is
patients may have a normal Hb so that diagnosis, in not possible in an emergency situation it is impor-
an emergency situation, cannot rest on a sickle tant to ensure that an adequate pre‐ transfusion
solubility test and full blood count (FBC) alone. sample is available for full testing on the next
Supplementing these tests with a blood film makes working day. A laboratory protocol for pre‐anaes-
provisional identification of these compound thetic testing is shown in Fig. 7.5. These procedures
heterozygous states more accurate. will mean that the provisional diagnosis is correct
It is also conventional to test any patient at risk of in the majority of patients. Some patients with
having sickle cell trait prior to anaesthesia, in order sickle cell disease with a high percentage of haemo-
to ensure that hypoxia does not occur during sur- globin F or sickle cell/β+ thalassaemia with a high
gery. It could be argued that no patient should be percentage of haemoglobin A may be missed, but
permitted to become hypoxic and that sickle screen- they are the patients most likely to have mild dis-
ing is therefore unnecessary if sickle cell disease can ease and least likely to suffer complications in rela-
be excluded. However, the detection of sickle cell tion to surgery and anaesthesia. It should also be
trait can also be relevant to the prolonged applica- noted that false negative results with a sickle solu-
tion of a tourniquet and if cell salvage techniques bility test can be seen in young infants. However, a
358 Chapter 7
Assess if surgery and anaesthesia are urgent or non-urgent
Non-urgent Urgent
Assess patient
Assess patient and

perform: FBC; Clinical Clinical
blood film; HPLC, features of features of
CE or CAE; sickle SCD absent SCD present
solubility test if
relevant; proceed
to surgery when
investigations
Perform sickle Perform sickle
completed
solubility test solubility test
and FBC; and FBC and
examine blood examine blood
film if anaemic film
Normal FBC
and sickle
solubility Sickle Relevant
test negative solubility clinical features
test positive, or blood film
FBC and film suggestive of
do not SCD, sickle
suggest SCD solubility test
positive
Regard as
normal
Regard as sickle trait Regard as SCD, consult

and avoid hypoxia haematologist, avoid hypoxia,
hypotension, hypothermia
and dehydration
Fig. 7.5 A protocol for pre‐anaesthetic testing to detect patients with sickle cell trait or sickle cell disease prior to
routine or emergency surgery. CAE, cellulose acetate electrophoresis; SCD, sickle cell disease.
false negative test is only expected with quite a low

Problems and pitfalls in pre‐anaesthetic
haemoglobin S percentage when anaesthetic com-
testing
plications would be unlikely. Definitive testing by
HPLC, capillary electrophoresis or haemoglobin Problems and pitfalls in pre‐anaesthetic testing
electrophoresis is required, but emergency surgery include: (i) failure to test at a pre‐assessment clinic
can proceed. before planned surgery; (ii) failure to test all patients
of appropriate ethnic origin; and (iii) failure to aemoglobins are also unstable. Relevant investiga-
h
recognise that a patient has sickle cell disease
tions include:
(e.g. sickle cell/haemoglobin C compound hete- • FBC, blood film and reticulocyte count;
rozygosity) rather than sickle trait. • HPLC, capillary electrophoresis or haemoglobin
electrophoresis;
• isopropanol and heat instability tests;
Other haemoglobinopathy
• oxygen dissociation curve and estimation of P50o2;
investigations
• family studies;
Investigation of haemolytic anaemia • investigation to exclude other causes of
polycythaemia.
Haemolytic anaemia can be consequent on the
presence of an unstable haemoglobin. When this is
suspected, the following tests should be performed: Identification of an unknown variant
• FBC, blood film and reticulocyte count; haemoglobin
• HPLC, capillary electrophoresis or haemoglobin
Uncommon haemoglobins that are not readily
electrophoresis;
identifiable are occasionally detected in screening
• isopropanol and heat instability tests;
programmes. Variant haemoglobins may also be
• haemoglobin A2 estimation;
detected because of an aberrant result when
• a test for Heinz bodies, repeated after incubation
measuring haemoglobin A1c for monitoring of
for 24 hours at 37°C, if initially negative;
diabetes. It is not always necessary, or indeed
• family studies;
possible, to identify a variant haemoglobin but it
• tests to exclude other causes of a haemolytic
is necessary to determine whether or not it is of
anaemia.
potential clinical significance. Routine tests
should be applied, as shown in Figs 7.1 and 7.2 in
Investigation of unexplained cyanosis order to identify the common variant haemoglob-
Unexplained cyanosis can be caused by a low ins likely to be clinically important, including
oxygen affinity haemoglobin, methaemoglobinae- haemoglobins S, C, D‐Punjab, E, O‐Arab, Lepore
mia or sulphaemoglobinaemia. Some inherited and Constant Spring. If the variant haemoglobin
methaemoglobins are also unstable. Relevant
remains unidentified after following one of
investigations include: these protocols, the following factors should be
• FBC, blood film and reticulocyte count; considered:
• HPLC, capillary electrophoresis or haemoglobin • whether is there any personal or family history
electrophoresis; suggestive of a clinically significant variant haemo-
• haemoglobin electrophoresis after conversion of globin (e.g. anaemia, jaundice, polycythaemia);
all haemoglobin to methaemoglobin; • whether there is an elevated reticulocyte count
• isopropanol and heat instability tests; or any blood film abnormality suggestive of
• spectrophotometry for the detection and quanti- haemolysis;
fication of sulph‐ and methaemoglobin; • whether there are red cell indices suggestive of a
• oxygen dissociation curve and estimation of P50o2; thalassaemic condition;
• partial pressure of oxygen in arterial blood to • whether the patient is in the reproductive age
exclude hypoxia as a cause of cyanosis; range or has close relatives who might be consider-
• family studies. ing pregnancy.
Further tests that might be indicated include a
sickle solubility test (since not all sickling haemo-
Investigation of unexplained polycythaemia
globins have the elution time or electrophoretic
Polycythaemia can be consequent on a high affinity mobility of haemoglobin S) and a test for an unsta-
haemoglobin or, rarely, on homozygosity for ble haemoglobin. Other tests that might be selec-
deletional HPFH. Some high oxygen affinity
tively applied if the variant haemoglobin has not
360 Chapter 7
been identified by HPLC, capillary electrophoresis screening for haemoglobin E heterozygosity, sensi-
or haemoglobin electrophoresis include: tivity increased from 89 to 98% when an MCV <80 fl
• oxygen dissociation curve and determination of or MCH <27 pg was supplemented by a red cell dis-
P50o2; tribution width (RDW) <14.45% [30]. Point‐of‐care
• haemoglobin absorption spectrum; tests applicable to some types of α0 thalassaemia in
• electron spray mass spectrometry; a resource‐poor setting have also been developed
• DNA sequencing or other molecular techniques. (see p. 102).
With the increasing availability of mass spec- In countries where sickle cell disease is the main
trometry and DNA sequencing, there is no longer a clinical problem, a sickle solubility test and elec-
significant role for globin chain electrophoresis, trophoresis at alkaline pH [31] or a point‐of‐care
tryptic digestion and peptide fingerprinting, or glo- test (see later) may suffice for basic diagnosis.
bin chain synthesis studies. When β thalassaemia is a significant problem,
HPLC or capillary electrophoresis is likely to be
preferred to electrophoresis since it permits meas-
Haemoglobinopathy investigations
urement of haemoglobin A2 as well as provisional
in a resource‐poor setting
identification of haemoglobin S and other variant
Appropriate investigations in a resource‐poor haemoglobins.
setting will depend on the prevalence of various
haemoglobinopathies and on financial and other
Point‐of‐care testing for haemoglobin S
constraints. In all such settings an FBC, including
MCV or MCH, is highly desirable. When an auto- Point‐of‐care testing for detection of haemoglob-
mated counter is not available, a modified osmotic ins S and possibly also haemoglobin C can be use-
fragility test (Naked Eye Single Tube Red cell ful in a resource‐poor setting, sometimes including
Osmotic Fragility Test, NESTROFT), can be used to use in outreach clinics. An ideal test will be fast,
screen for α and β thalassaemia trait. Abnormal inexpensive, not dependent on an electricity sup-
results are found also in iron deficiency and sickle ply, refrigeration or expensive equipment, and
cell trait. If an automated counter is available, use of with a high sensitivity. A high specificity is also
this screening test may not be advised, since false desirable since this reduces the need for follow‐up
negative results are obtained in a small percentage testing.
of cases of β thalassaemia heterozygosity. However, A novel point‐of‐care test for the detection of
a study from Thailand found that the osmotic fra- sickle cell disease depends on the ability of nor-
gility test showed 95% sensitivity for the detection mal cells to move through a piece of modified
of α0 or β thalassaemia, while the MCV, with a cut‐ chromatography paper while sickled cells cannot.
off point of 78.1 fl, showed 93% sensitivity [27]. The A drop of blood is mixed with sickle solubility
dichlorophenolindophenol (DCIP) test can be used reagents and a drop of the mixture is then applied
to screen for haemoglobin E heterozygosity, haemo- to the paper. Normal, sickle cell trait (AS) and
globin E homozygosity and haemoglobin E/β tha- sickle cell anaemia (SS) can be distinguished visu-
lassaemia compound heterozygosity; positive ally, with sickle cell/haemoglobin C disease (SC)
results are also obtained in haemoglobin H disease results falling between AS and SS [32]. A modifi-
and with unstable haemoglobins. Modifications of cation of the method makes it suitable for neona-
the DCIP tests improve sensitivity and specificity tal screening, a study in Angola finding 100%
[28]. In Thailand, initial antenatal screening for α0 sensitivity and 83% specificity for detection of
and β thalassaemia and haemoglobin E is by means sickle cell anaemia [33].
of either a modified osmotic fragility test or red cell Another novel method uses density separation of
indices plus a DCIP test; testing for haemoglobin H red cells, with the presence of dense cells permitting
inclusions and immunochromatographic detection sickle cell anaemia and sickle cell/haemoglobin C
of ζ chain or haemoglobin Bart’s can have a supple- disease to be distinguished from each other and
mentary role [29]. In a Sri Lankan study, when from sickle cell trait or normal with a sensitivity
of 90–91% and a specificity of 88–91%; sickle cell Check your knowledge

trait and normal samples are not distinguished
from each other [34]. One to five answers may be correct. Answers to
A rapid point‐of‐care immunoassay, SickleSCAN, most questions can be either found in this chapter
distinguishes between haemoglobins S, C and or deduced from information given. Answers are
‘haemoglobin other than S or C’ (usually haemoglo- given on p. 364.
bin A) [35]. It is applicable to dried blood spots and
7.1 An antenatal haemoglobinopathy screening
may thus be suitable not only for screening for
programme should detect
carriers but also for testing of neonates in resource‐
(a) α0 thalassaemia
poor settings. No power source is needed. A
(b) α+ thalassaemia
preliminary study in Nigeria showed 100% sensi-
(c) β0 thalassaemia
tivity in detection of sickle cell disease (one SS, one
(d) β+ thalassaemia
SC) from the age of 9 months upwards (younger
(e) δβ thalassaemia
infants not tested); SS, SC and AS were generally
distinguished [36].
7.2 Sickle cell disease, which could complicate
Another rapid point‐of‐care test, HemoTypeSC,
surgery and anaesthesia, could be observed in
uses monoclonal antibodies and a competitive
(a) Arabs
lateral flow immunochromatographic assay to
(b) Greeks
recognise the presence of haemoglobins S, C and
(c) Cypriots
‘not S or C’ (mainly haemoglobin A) [37, 38]. It
(d) North Africans
requires only 15 μl of blood and will distinguish
(e) Afro‐Caribbeans
between: haemoglobin A plus S (either sickle cell
trait of haemoglobin S/β+ thalassaemia); haemo- 7.3 Haemoglobin electrophoresis plus HPLC at 6
globin A plus C (either haemoglobin C trait of weeks of age could provide a definitive
haemoglobin C/β+ thalassaemia); haemoglobin S diagnosis of
plus C; haemoglobin S only (sickle cell anaemia (a) β thalassaemia trait
or haemoglobin S/β0 thalassaemia); haemoglobin (b) sickle cell/hereditary persistence of fetal
C only (either haemoglobin C homozygosity or haemoglobin
haemoglobin C/β0 thalassaemia); and ‘haemoglo- (c) sickle cell/β0 thalassaemia
bin other than S or C only’ (mainly haemoglobin (d) sickle cell/haemoglobin C disease
A only) [37]. Testing is reliable irrespective of the (e) sickle cell trait
presence of haemoglobin F, even at high levels
[37]. It requires no instrumentation and no power 7.4 It is important that an antenatal haemoglobi-
source so is useful in resource‐poor settings nopathy screening programme detects
where the only common variant haemoglobins (a) haemoglobin C trait
are S and C. In a large study carried out largely in (b) haemoglobin D‐Punjab trait
Ghana, the sensitivity and specificity for SS/Sβ0 (c) haemoglobin E trait
thalassaemia and CC were both 100%, while for (d) haemoglobin G‐Philadelphia trait
SC the sensitivity was 97% and the specificity (e) haemoglobin Lepore trait
100% [38]. For A + S, A + C and A only the sensi-
tivity and specificity were both above 99% [38]. 7.5 In an antenatal haemoglobinopathy screening
A Ugandan study similarly found a very high programme, investigation for α0 thalassaemia
degree of accuracy in the detection of haemoglob- trait should be carried out in
ins A and S in patients from the age of 1 month (a) Chinese
upwards [39]. (b) Laotians
Microfluidic electrophoretic techniques, such as (c) Cambodians
HemeChip, which requires a power source, are (d) Japanese
under development. (e) Vietnamese
362 Chapter 7
7.6 A patient with a high oxygen affinity Internet resources

haemoglobin could have normal results for
(a) cellulose acetate electrophoresis at EMQN Best Practice Guidelines for molecular and
alkaline pH haematology methods for carrier identification and
(b) oxygen dissociation curve prenatal diagnosis of the haemoglobinopathies.
(c) agarose gel electrophoresis at acid pH https://www.ncbi.nlm.nih.gov/pmc/articles/
(d) measurement of P50o2 PMC4666573/
(e) isopropanol test Public Health England. NHS Sickle Cell and
Thalassaemia Screening Programme. Handbook for
7.7 A neonatal haemoglobinopathy screening Antenatal Laboratories, 4th edn, October 2017. https://
programme should aim to detect assets.publishing.service.gov.uk/government/
cases of uploads/system/uploads/attachment_data/
(a) α thalassaemia trait file/656094/Antenatal_Laboratory_Handbook.pdf
(b) β thalassaemia trait Public Health England. NHS Newborn Blood Spot
(c) sickle cell anaemia (NBS) Screening Programme, 2018. https://www.
(d) sickle cell/β0 thalassaemia compound gov.uk/topic/population‐screening‐programmes/
heterozygosity newborn‐blood‐spot
(e) hereditary persistence of fetal Public Health England. NHS Sickle Cell and
haemoglobin Thalassaemia (SCT) Screening Programme, 2019.
https://www.gov.uk/topic/population‐screening‐
7.8 In an antenatal haemoglobinopathy screen- programmes/sickle‐cell‐thalassaemia
ing programme, investigation for α0 thalas-
saemia trait should be carried out in
References
(a) Greeks
(b) Cypriots 1 Guerrier G, Piot MA, Resiere D, Guerrier M,
(c) Spaniards Baillard C, Moro MR et al. (2017) Adequate clinical
(d) Portuguese practice limited by the ethnic French taboo. Lancet,
(e) Austrians 389, 2189–2190.
2 Baysal E, Yousef I, Zeinali M, Abdullah H, Essa M,
7.9 Suitable primary methods for Taleb M and Redha A (2007) Molecular basis of
neonatal haemoglobinopathy screening haemoglobinopathies in the UAE: implications for
include prenatal diagnosis and prevention programmes.
Emirates Med J, 25, 7–21.
(a) cellulose acetate electrophoresis at
3 Cao A, Congiu R, Sollaino MC, Desogus MF,
alkaline pH
Demartis FR, Loi D et al. (2008) Thalassaemia and
(b) citrate agar or agarose gel electrophore-
glucose‐6‐phosphate dehydrogenase screening in
sis at acid pH
13‐ to 14‐year‐old students of the Sardinian popula-
(c) isoelectric focusing
tion: preliminary findings. Commun Genet, 11,
(d) high performance liquid
121–128.
chromatography 4 Younis K (2006) Premarital testing for β‐thalassemia
(e) sickle solubility test trait: Palestine experience 2000 – 2005. Pan Arab
Med J, Suppl, 89.
7.10 Emergency pre‐anaesthetic screening should 5 Al‐Allawi NA, Jalal SD, Ahmed NH, Faraj AH, Shalli
detect patients with A and Hamamy H (2013) The first five years of a
(a) sickle cell trait preventive programme for haemoglobinopathies in
(b) β thalassaemia trait Northeastern Iraq. J Med Screen, 20, 171–176.
(c) sickle cell anaemia 6 Ahmed S, Saleem M, Modell B and Petrou M (2002)
(d) sickle/cell haemoglobin C disease Screening extended families for genetic hemoglobin
(e) haemoglobin C disease disorders in Pakistan. N Engl J Med, 347, 1162–1168.
7 Ryan K, Bain BJ, Worthington D, James J, Plews D, 18 Henthorn JS, Almeida AM and Davies SC (2004)
Mason A et al.; British Committee for Standards in Neonatal screening for sickle cell disorders. Br J
Haematology (2010) Significant haemoglobinopa- Haematol, 124, 259–263.
thies: guidelines for screening and diagnosis. Br J 19 Public Health England. NHS Sickle Cell and
Haematol, 149, 35–49. Thalassaemia Screening Programme. Handbook for
8 Wee YC, Tan KL, Kuldip K, Tai KS, George E, Tan Newborn Laboratories, 4th edn, 2017. https://assets.
PC et al. (2008) Alpha‐thalassaemia in association publishing.service.gov.uk/government/uploads/
with beta‐thalassaemia patients in Malaysia: a system/uploads/attachment_data/file/585126/
study of the co‐inheritance of both disorders. NHS_SCT_Handbook_for_Newborn_Laboratories.
Commun Genet, 11, 129–134. pdf
9 Pootrakul P, Winichagoon P, Fucharoen S, 20 Lobitz S, Telfer P, Cela E, Allaf B, Angastiniotis M,
Pravatmuang P, Piankijagum A and Wasi P (1981) Backman Johansson C et al.; with the endorsement
Homozygous haemoglobin Constant Spring: a need of EuroBloodNet, the European Reference Network
for revision of concept. Hum Genet, 59, 250–255. in Rare Haematological Diseases (2018) Newborn
10 Yang Y, Lou J‐W, Liu Y‐H, He Y and Li D‐Z (2014) screening for sickle cell disease in Europe: recom-
Screening and Diagnosis of Hb Quong Sze [HBA2: mendations from a Pan‐European Consensus
c.377T > C (or HBA1)] in a prenatal control program Conference. Br J Haematol, 183, 648–660.
for thalassemia. Hemoglobin, 38, 158–160. 21 International Committee for Standardization in
11 Modell B, Petrou M, Layton M, Varnavides L, Haematology (1988) Recommendations for neona-
Moisely C, Ward RHT et al. (1998) Audit of prenatal tal screening for haemoglobinopathies. Clin Lab
diagnosis for haemoglobin disorders in the United Haematol, 10, 335–345.
Kingdom: the first twenty years. Ann NY Acad Sci, 22 Chapman C and Chambers K (1997) Neonatal
850, 420–422. haemoglobinopathy screening methods. MDA

12 Daniel Y, Cartwright R, Rennie K, Streetly A and Evaluation Report MDA/97/64. Medical Devices
Howard J (2014) Evaluation of the validity of Hb A2 Agency, London.
and mean corpuscular haemoglobin action values 23 Campbell M, Henthorn JS and Davies SC (1999)
in antenatal screening for beta thalassaemia carriers Evaluation of cation‐exchange HPLC compared
in England. Br J Haematol, 166, 607–611. with isoelectric focusing for neonatal hemoglobi-
13 Ma ESK, Chan AYY, Ha SY, Lau YL and Chan LC nopathy screening. Clin Chem, 45, 9251–9275.
(2001) Thalassemia screening based on red cell indi- 24 Daniel Y and Henthorn J (2015) Tandem mass spec-
ces in the Chinese. Haematologica, 86, 1310–1311. trometry for sickle cell and thalassaemia newborn
14 Liu S, Huang LY, Jiang F, Sun X‐F and Li D‐Z (2018) screening pilot study. Public Health England.
Complex interactions between thalassemia defec- https://assets.publishing.service.gov.uk/government/
tive alleles compromise screening and cause severe uploads/system/uploads/attachment_data/
anemia in a Chinese family. Int J Lab Hematol, 40, file/488858/Tandem_Mass_Spectrometry_for_
e55–e58. Sickle_Cell_and_Thalassaemia_Newborn_
15 Streetly A, Sisodia R, Dick M, Latinovic R, Hounsell Screening_Pilot_Study_2015.pdf
K and Dormandy E (2018) Evaluation of newborn 25 US Department of Health and Human Services
sickle cell screening programme in England: 2010– (1993) Guideline: laboratory screening for sickle cell
2016. Arch Dis Child, 103, 648–653. disease. Lab Med, 24, 515–522.
16 Jindatanmanusan P, Riolueang S, Glomglao W, 26 Serjeant GR, Serjeant BE, Hambleton IR, Oakley M,
Sukontharangsri Y, Chamnanvanakij S, Torcharus Thein SL and Clark B (2017) A plea for the newborn
K and Viprakasit V (2013) Diagnostic applications diagnosis of Hb S‐hereditary persistence of fetal
of newborn screening for α‐thalassaemias, haemo- hemoglobin. Hemoglobin, 41, 216–217.
globins E and H disorders using isoelectric focusing 27 Sirichotiyakul S, Wanapirak C, Srisupundit K,
on dry blood spots. Ann Clin Biochem, 51, 237–247. Luewan S and Tongsong T (2009) A comparison of
17 Report of a Working Party of the Standing the accuracy of the corpuscular fragility and mean
Medical Advisory Committee on Sickle Cell, corpuscular volume tests for the alpha‐thalassemia
Thalassaemia and other Haemoglobinopathies. 1 and beta‐thalassemia traits. Int J Gynaecol Obstet,
HMSO, London, 1993. 107, 26–29.
364 Chapter 7
28 Wanapirak C, Sirichotiyakul S, Luewan S, 34 Kumar AA, Patton MR, Hennek JW, Lee SY,
Srisupundit K and Tongsong T (2009) Comparison D’Alesio‐Spina G, Yang X et al. (2014) Density‐
of the accuracy of dichlorophenolindophenol based separation in multiphase systems provides a
(DCIP), modified DCIP, and haemoglobin E tests to simple method to identify sickle cell disease. Proc
screen for the HbE trait in pregnant women. Int J Natl Acad Sci U S A, 111, 14864–14869.
Gynaecol Obstet, 107, 59–60. 35 McGann PT, Schaefer BA, Paniagua M, Howard TA
29 Jomoui W, Fucharoen G, Sanchaisuriya K and and Ware RE (2016) Characteristics of a rapid,
Fucharoen S (2017) Screening of (–SEA) α‐thalassae- point‐of‐care lateral flow immunoassay for the
mia using an immunochromatographic strip assay diagnosis of sickle cell disease. Am J Hematol, 91,
for the ζ‐globin chain in a population with a high 205–210.
prevalence and heterogeneity of haemoglobinopa- 36 Nwegbu MM, Isa HA, Nwankwo BB, Okeke CC,
thies. J Clin Pathol, 70, 63–68. Edet‐Offong UJ, Akinola NO et al. (2017) Preliminary
30 Nishad AA, de Silva IS, Perera HL, Pathmeswaran evaluation of a point‐of‐care testing device
A, Kastutiratne KT and Premawardhena AP (2014) (SickleSCAN™) in screening for sickle cell disease.
Role of red cell distribution width in screening for Hemoglobin, 41, 77–82.
Hb E trait in population screening for hemoglobin 37 Quinn CT, Paniagua MC, DiNello RK, Panchal A
disorders. J Pediatr Hematol Oncol, 36, e490–e492. and Geisberg M (2016) A rapid, inexpensive and
31 Heimlich JB, Chipoka G, Kamthunzi P, Krysiak R, disposable point‐of‐care blood test for sickle cell
Majawa Y, Mafunga P et al. (2016) Establishing disease using novel, highly specific monoclonal
sickle cell diagnostics and characterizing a antibodies. Br J Haematol, 175, 724–732.
paediatric sickle cell disease cohort in Malawi. 38 Steele C, Sinski A, Asibey J, Hardy‐Dessources MD,
Br J Haematol, 174, 325–329. Elana G, Brennan C et al. (2019) Point‐of‐care screen-
32 Yang X, Kanter J, Piety NZ, Benton MS, Vignes SM ing for sickle cell disease in low‐resource settings: a
and Shevkoplyas SS (2012) A simple, rapid, low‐ multi‐center evaluation of HemoTypeSC, a novel
cost test for the diagnosis of sickle cell disease using rapid test. Am J Hematol, 94, 39–45.
a paper‐based hemoglobin solubility assay. Blood, 39 Nankanja R, Kadhumbula S, Tagoola A, Geisberg
120, 245. M, Serrao E and Balyegyusa S (2019) HemoTypeSC
33 Piety NZ, George A, Serrano S, Lanzi MR, Patel PR, demonstrates >99% field accuracy in a sickle cell
Noli MP et al. (2017) A paper‐based test for screen- disease screening initiative in children of southeast-
ing newborns for sickle cell disease. Sci Rep, 7, 45488. ern Uganda. Am J Hematol, 94, E164–E166.
7.1 (a) T 7.3 (a) F 7.5 (a) T 7.7 (a) F 7.9 (a) T
(b) F (b) F (b) T (b) F (b) F
(c) T (c) F (c) T (c) T (c) T
(d) T (d) T (d) F (d) T (d) T
(e) T (e) T (e) T (e) F (e) F
7.2 (a) T 7.4 (a) T 7.6 (a) T 7.8 (a) T 7.10 (a) T
(b) T (b) T (b) F (b) T (b) F
(c) T (c) T (c) T (c) F (c) T
(d) T (d) F (d) F (d) F (d) T
(e) T (e) T (e) T (e) F (e) F
8 Self‐assessment: test cases
All the case studies are based on real patients pre most likely explanation or explanations for each
senting real diagnostic problems. The reader is case. No patient had been transfused and all were
advised that not all are straightforward. Careful adults. (For patient 3, note the quantity of the vari
thought and, if necessary, reference back to earlier ant haemoglobin.)
chapters is advised before looking at the answers Patient 1 ……………………………………………
given in the second half of this chapter. Patient 2 ……………………………………………
Patient 3 ……………………………………………
Patient 4 ……………………………………………
Exercise 8.1
Patient 5 ……………………………………………
You are provided with a diagrammatic representa Patient 6 ……………………………………………
tion of results of haemoglobin electrophoresis on Patient 7 ……………………………………………
cellulose acetate at alkaline pH and haemoglobin Patient 8 ……………………………………………
electrophoresis on agarose gel at acid pH and with Patient 9 ……………………………………………
the results of a sickle solubility test on a control Patient 10 …………………………………………..
sample and samples from patients 1–11. Give the Patient 11 …………………………………………..
Cellulose acetate Agarose gel

electrophoresis – alkaline electrophoresis – acid
AF S C FA S C Sickle solubility test
Control
1 Positive
2 Negative
3 Negative
4 Positive
5 Positive
6 Positive
7 Negative
8 Negative
9 Negative
10 Positive
11 Negative

365
366 Chapter 8
Exercise 8.2 Afro‐Caribbean patients, patient 1 and patient 2.

Give the most likely diagnosis in each.
You are provided with haemoglobin electrophore
Patient 1 ……………………………………………
sis at alkaline pH, a sickle solubility test and a high
Patient 2 ……………………………………………
performance liquid chromatography (HPLC)
chromatogram (Bio‐Rad Variant) on two

Self‐assessment: test cases 367
Exercise 8.3 pg and mean cell haemoglobin concentration

(MCHC) 332 g/l.
You are provided with a blood film and the results
What is the most likely diagnosis? …………………
of electrophoresis at alkaline pH (lane d) on a
What test would you perform to confirm the
33‐year‐old pregnant Bangladeshi woman with
diagnosis? ………………………………………………
the following red cell indices: red cell count (RBC)
What abnormality in the partner would be
4.39 × 1012/l, haemoglobin concentration (Hb) 110
most likely to lead to serious disease in the fetus?
g/l, haematocrit (Hct) 0.32, mean cell volume
…………………………………………………………
(MCV) 74 fl, mean cell haemoglobin (MCH) 25.1
368 Chapter 8
Exercise 8.4 serum ferritin 28 μmol/l (normal range 10–300). You

are provided with a photograph of the blood film
A young woman was referred to a haematologist
and the electrophoretic pattern on cellulose acetate
because of borderline anaemia (Hb between 100 and
at alkaline pH (lane b). The haemoglobin A2 was
115 g/l) and microcytosis that had not responded to
estimated at 1.8 and 2.5% on two occasions (normal
oral iron. The patient stated that as far as she knew
range 2.3–3.5%). Haemoglobin F was 11.4% by alkali
her ancestry was totally English, although her father
denaturation with a heterogeneous distribution on
had once been told that he had ‘Spanish blood’. The
a Kleihauer test.
haematologist stopped the iron therapy and a week
later performed various tests with the following
What advice should be given to the patient?
results: RBC 5.37 × 1012/l, Hb 116 g/l, MCV 69 fl,
…………………………………………………………
MCH 21.7 pg, MCHC 312 g/l, red cell distribution
width (RDW) 19.2% (normal range 11.1–14.9) and
Exercise 8.5 0.35, MCV 72 fl, MCH 21.7 pg and MCHC 301 g/l.
Electrophoresis at acid pH was normal.
You are provided with the blood film, the electro
phoretic pattern at alkaline pH (lane d) and the
What is the clinical significance of this result?
HPLC chromatogram (Bio‐Rad Variant) of a
…………………………………………………………
34‐year‐old woman from Gibraltar with the following
red cell indices: RBC 4.83 × 1012/l, Hb 105 g/l, Hct
370 Chapter 8
Exercise 8.6 What is the most likely diagnosis in each family

member?
You are provided with blood films, red cell indices
Father ………………………………………………
and the results of haemoglobin electrophoresis on a
17‐year‐old daughter ………………………………
middle‐aged man, who was complaining of fatigue,
15‐year‐old daughter ………………………………
and from three of his children.
10‐year‐old son ……………………………………
(With thanks to Dr Francis Toolis.)
Family member RBC (× 1012/l) Hb (g/l) MCV (fl) MCHC (g/l) Haemoglobin electrophoresis
Father (a) 5.47 107 65 302 A + A2 (A2 5.2%)

Daughter, aged 17 (b) 3.96 117 87 343 A + A2 (A2 2.5%)
Daughter, aged 15 (c) 5.43 109 64 314 A + A2 (A2 5.1%)
Son, aged 10 (d) 4.61 129 88 339 A + A2 (A2 2.3%)
(a)
(b)
(c)
(d)
372 Chapter 8
Exercise 8.7 What explanations of these results would you

consider and what further investigations would
A 74‐year‐old white American woman was hospi
you perform? …………………………………………
talised because of back pain caused by osteoporosis.
A blood film showed hypochromia and a full blood
count (FBC) showed: RBC 2.9 × 1012/l, Hb 69 g/l, Exercise 8.8
Hct 0.21, MCV 74 fl, MCH 23 pg and RDW 18.4%
You are provided with photographs of a blood film,
(normal range 11.1–14.9). She was transfused
haemoglobin electrophoresis on cellulose acetate at
2 units of red cells and was discharged. Several
alkaline pH (lane b), haemoglobin electrophoresis
weeks later she was readmitted with an Hb of 59
on agarose gel at acid pH (fourth lane from left)
g/l and was found to have a bleeding gastric ulcer.
and the results of globin chain synthesis studies
She was transfused 5 units of blood over 3 days.
(with thanks to Professor Lucio Luzzatto) on an
Subsequently haemoglobin electrophoresis showed
Afro‐Caribbean patient with the following red cell
a band with the mobility of haemoglobin S and a
indices: RBC 5.43 × 1012/l, Hb 107 g/l, Hct 0.34,
sickle solubility test was positive. Haemoglobins
MCV 63 fl, MCH 19.7 pg and MCHC 312 g/l.
were quantified by HPLC as follows: haemoglobin
What is the diagnosis? ……………………………
A 89.7%, haemoglobin A2 2.9%, haemoglobin F 0.4%
and haemoglobin S 7.0%.
374 Chapter 8
Exercise 8.9 What is the likely diagnosis in each family

member?
You are provided with red cell indices and photo
Mother ……………………………………………
graphs of blood films, cellulose acetate electrophoresis
Father ………………………………………………
and results of sickle solubility tests on a child and his
Child ………………………………………………
parents. In the cellulose acetate electrophoresis the
mother is lane a, the father lane b and the child lane f.
RBC (× 1012/l) Hb (g/l) MCV (fl) MCH (pg) Sickle solubility test
Mother (a) 4.39 117 80 26.6 Negative

Father (b) 4.96 146 88 29.4 Positive
Child (c) 2.05 72 98 35.2 Positive
(a)
(b)
(c)
376 Chapter 8
Exercise 8.10 Patient 1 …………………………………………….

Patient 2 …………………………………………….
You are provided with a diagrammatic representa
Patient 3 …………………………………………….
tion of results of haemoglobin electrophoresis on
Patient 4 …………………………………………….
cellulose acetate at alkaline pH and on agarose gel
Patient 5 …………………………………………….
at acid pH together with the results of a sickle solu
Patient 6 …………………………………………….
bility test on a control sample and samples from
Patient 7 …………………………………………….
patients 1–11. Give the most likely explanation or
Patient 8 …………………………………………….
explanations for each case. No patient had been
Patient 9 …………………………………………….
transfused and all were adults. For the purposes of
Patient 10 ……………………………………………
this exercise, ignore the thickness of the bands.
Patient 11 ……………………………………………
Cellulose acetate Agarose gel

electrophoresis – alkaline electrophoresis – acid
AF S C FA S C Sickle solubility test
Control
1 Positive
2 Positive
3 Positive
4 Positive
5 Negative
6 Negative
7 Positive
8 Negative
9 Negative
10 Negative
11 Positive
Exercise 8.11 MCHC 307 g/l. The haemoglobin A2 was 5.5%.

You are provided with a photograph of the patient’s
A young man had a routine blood count performed.
blood film and of haemoglobin electrophoresis on
This was noted to show abnormal red cell indices,
cellulose acetate at alkaline pH (lane c).
leading to haemoglobin electrophoresis being per
Suggest two possible explanations of the abnor
formed. Three of the patient’s grandparents were of
mality observed. ………………………………………
Indian ethnic origin. The other was half‐Indian and
What is the likely clinical significance?
half‐Chinese. There was no history of consanguin
…………………………………………………………
ity. Red cell indices were: RBC 5.56 × 1012/l, Hb
110 g/l, Hct 0.36, MCV 65 fl, MCH 19.8 pg and
378 Chapter 8
Exercise 8.12 with the photographs of a blood film and an electro

phoretic strip at alkaline pH (lane e).
A 23‐year‐old Afro‐Caribbean woman had a positive
What is the most likely explanation of the abnor
sickle solubility test and the following red indices:
mal red cell indices and electrophoretic abnormality?
RBC 4.42 × 1012/l, Hb 115 g/l, Hct 0.35, MCV 78 fl,
…………………………………………………………
MCH 26 pg and MCHC 318 g/l. You are provided
Exercise 8.13 ifferential count showed 91% erythroid cells; 40%

d
of remaining cells were blast cells. Erythroblasts
A 67‐year‐old man of Northern European Caucasian
showed increased siderotic granulation, including a
origin presented with a history of breathlessness and
low percentage of ring sideroblasts. Haemoglobin
ankle oedema. On physical examination he was pale
electrophoresis showed 15% haemoglobin H and
but no other abnormality was detected. His FBC
a haemoglobin H preparation showed typical
showed white cell count (WBC) 5.7 × 109/l, RBC 5.3 ×
inclusions. The α:β chain synthesis ratio was 0.49.
1012/l, Hb 94 g/l, Hct 0.36, MCV 68.6 fl, MCH 17.7
What abnormalities are shown by the blood and
pg, MCHC 261 g/l and platelet count 651 × 109/l.
bone marrow films? ……………………………………
You are provided with photographs of his peripheral
Explain the nature of the patient’s condition.
blood film and his bone marrow aspirate.
…………………………………………………………
The bone marrow was hypercellular with
(With thanks to Dr Jane Mercieca.)
increased number of megakaryocytes, some of
which were hypolobated. The bone marrow
380 Chapter 8
Exercise 8.14 from a young African man with sickle cell anaemia.
The blood film also showed Howell–Jolly bodies.
You are provided with photographs of a blood film
Explain the blood film and CT scan features in
and a computed tomography (CT) scan of the abdo
relation to each other. ………………………………
men, performed without any contrast medium,
Exercise 8.15 on agarose gel at acid pH (lanes 6 and 8) from a

young Afro‐Caribbean woman with anaemia and
You are provided with photographs of a blood film,
recurrent limb pains.
haemoglobin electrophoresis on cellulose acetate at
What is the most likely diagnosis? ………………
alkaline pH (lane e) and haemoglobin electrophoresis
382 Chapter 8
Exercise 8.16 Haemoglobin electrophoresis on agarose gel at acid

pH showed two major bands with the mobility of A
You are provided with photographs of a blood film
and S.
and haemoglobin electrophoresis on cellulose ace
tate at alkaline pH (sixth lane) from an African
What is the significance of this diagnosis?
woman who has been hospitalised with a breast
…………………………………………………………
abscess. Red cell indices were: RBC 4.23 × 1012/l, Hb
What are the possible explanations of the micro
107 g/l, MCV 75 fl, MCH 25.3 pg, MCHC 335 g/l
cytosis? …………………………………………………
and RDW 15.3%. A sickle solubility test was positive.
Exercise 8.17 there were two bands with the mobilities of A and S.
The sickle solubility test was positive.
You are provided with photographs of the blood
film and haemoglobin electrophoresis at alkaline pH
What is the significance to the patient?
(lane 5) of an Afro‐Caribbean woman in the first tri
…………………………………………………………
mester of pregnancy. On agarose gel at acid pH
384 Chapter 8
Exercise 8.18 1012/l, Hb 115 g/l, Hct 0.38, MCV 66 fl, MCH 20.1
pg and MCHC 304 g/l.
A 32‐year‐old pregnant Chinese woman had normal
What diagnoses are likely? …………………………
haemoglobin electrophoresis (haemoglobin A2
What tests should be performed in her partner
2.1%) and the following red cell indices: RBC 5.71 ×
and why? ………………………………………………
Exercise 8.19 screening showed haemoglobins A, F and A2 with

haemoglobin F being 23% of total haemoglobin and
You are provided with a photograph of the blood
haemoglobin A2 1.6%.
film of an Afro‐Caribbean woman in the first trimester
of pregnancy. Her red cell indices were: RBC 4.2 ×
What are the implications for management of
1012/l, Hb 120 g/l, MCV 88 fl, MCH 28.6 pg and
the patient? …………………………………………
MCHC 324 g/l. She was found to be Rh D negative
with no atypical antibodies. Haemoglobinopathy
Exercise 8.20 occasional café‐au‐lait spots. Results of FBC were:

WBC 28.3 × 109/l, Hb 84 g/l, MCV 78 fl and platelet
You are provided with a photograph of the blood
count 96 × 109/l. Monocytes and neutrophils were
film of a 2‐year‐old Northern European Caucasian
both increased. Haemoglobin electrophoresis
boy noted by his parents to be miserable and fretful.
showed 12% haemoglobin F and 1.0% haemoglobin
On examination, he had hepatosplenomegaly,
A2. The carbonic anhydrase band on the stained
moderate enlargement of lymph nodes and an
electrophoretic strip appeared to be reduced.
eczematous rash. There was a family history of
What is the likely diagnosis? ……………………
neurofibromatosis and the child was noted to have
386 Chapter 8
Exercise 8.21 Suggest reasons that might explain why the girl
is more anaemic than her father. ……………………
You are provided with a photograph of the blood film
(With thanks to Dr Michael Makris.)
of a 17‐year‐old girl with a white English mother and
a Pakistani father. She had been found to be anaemic
when she required extraction of wisdom teeth. There
was no hepatomegaly or splenomegaly. The red cell
indices and results of haemoglobin A2 quantification
in the girl and her parents are tabulated.
Propositus Father Mother
RBC (× 1012/l) 4.04 5.99 4.22

Hb (g/l) 85 120 136
MCV (fl) 68.8 62.9 93.1
MCH (pg) 21.0 20.2 32.2
Haemoglobin A2 4.7 4.4 2.3
Haemoglobin F 3.2 <1.0 <1.0
Exercise 8.22 Haemoglobin F was 18.3% and haemoglobin A2 was

2.6% by HPLC and 1.8% by microcolumn
You are provided with an HPLC chromatogram
chromatography.
(Bio‐Rad Variant) of a patient with the following
What is the likely diagnosis? ………………………
red cell indices: RBC 4.56 × 1012/l, Hb 113 g/l, Hct
0.34, MCV 75 fl, MCH 24.7 pg and MCHC 332 g/l.
Retention time (minutes)

388 Chapter 8
Exercise 8.23 What is the most likely diagnosis in the mother?

…………………………………………………………
You are provided with photographs of blood films
What is the most likely diagnosis in the daughter?
of a girl aged 18 (a) and her mother (b). The girl’s
…………………………………………………………
father had a normal blood film. Other details are
tabulated.
Propositus Mother Father
Ethnic origin Italian‐African Italian West African

RBC (× 1012/l) 3.65 5.78 5.1
Hb (g/l) 70 105 153
MCV (fl) 60 56 90
MCH (pg) 19.2 18.2 30
MCHC (g/l) 319 323 326
Ferritin (μmol/l) 430 33 50
(a)
(b)
Exercise 8.24 MCH 23.4 pg and MCHC 340 g/l. Haemoglobin

electrophoresis showed haemoglobin A 23%, haemo
You are provided with a photograph of the blood globin S 73% and haemoglobin A2 4%.
film of a patient with the following red cell indices: What is the most likely diagnosis? ………………
RBC 5.39 × 1012/l, Hb 126 g/l, Hct 0.37, MCV 69 fl,
Exercise 8.25 on auscultation of the lungs. Arterial blood gas

analysis showed a normal partial pressure of oxygen
A 3‐year‐old white English boy was brought to the
and reduced partial pressure of carbon dioxide. An
paediatric accident and emergency department by
FBC showed WBC 14.3 × 109/l, Hb 110 g/l, MCV 85
his father who had found him playing in the garden
fl, MCH 27 pg and platelet count 360 × 109/l. The
shed several hours previously. The child had com
laboratory scientist performing the blood count and
plained of abdominal pain and had vomited.
blood gas analysis noted that the blood was choco
Thereafter he was appeared restless and lethargic.
late‐brown in colour.
On examination in the emergency department he
was noted to have central cyanosis, tachycardia,
What test should be done? …………………………
tachypnoea and hypotension. He was afebrile. Heart
What treatment should be given? …………………
sounds were normal and there was no abnormality
390 Chapter 8
Exercise 8.26 between an AC sample (to the left) and an FASC

control sample (to the right). Her WBC and platelet
You are provided with results of cellulose acetate
count were normal. Red cell indices were: RBC 4.71
electrophoresis at alkaline pH, agarose gel electro
× 1012/l, Hb 161 g/l, Hct 0.46, MCV 96.8 fl, MCH
phoresis at acid pH and an HPLC chromatogram
34.1 pg, MCHC 352 g/l and RDW 12.2%.
(Bio‐Rad Variant II) on an Irish woman known to
What is the likely diagnosis? ………………………
have had a high haemoglobin concentration for
(With thanks to Mairead O’Reilly.)
some time. On the agarose gel electrophoresis, the
patient’s blood sample is second from the right
Cellulose acetate
AFSC
electrophoresis
at alkaline pH
Propositus
S
A
F
Agarose gel electrophoresis Propositus

at acid pH
Exercise 8.27 1.35 minutes. On cellulose acetate electrophoresis

at alkaline pH, the A band was broadened by a
slightly faster component. Agarose gel electrophoresis
(Bio‐Rad Variant II) on an anaemic pregnant African
at acid pH was normal.
woman (Hb 98 g/l, MCV 79 fl). The ‘P2 fraction’ on
Is this likely to indicate pregnancy‐related
Bio‐Rad Variant II HPLC analysis was 41.4% with
diabetes or is there an alternative explanation?

the abnormal peak having a retention time of
…………………………………………………………
Exercise 8.28 following intravenous injection of methylene

blue.
Three previously healthy children whose father
What is the likely cause of the cyanosis?
was a butcher became profoundly cyanosed, with
…………………………………………………………
a deteriorating level of consciousness, within a
What is the probable underlying cause?
short time of each other. Rapid recovery occurred
…………………………………………………………
392 Chapter 8
Exercise 8.29 Make a provisional diagnosis. ……………………

After reaching a provisional diagnosis, exam
The blood of a 34‐year‐old pregnant Caucasian
ine the HPLC chromatogram (Bio‐Rad Variant II)
woman was sent for antenatal thalassaemia and
provided and assess whether the provisional
haemoglobinopathy screening in a hospital with a
diagnosis is confirmed. State the significance of
policy of universal screening. Her red cell indices
your findings. The variant haemoglobin was
were: RBC 4.23 × 1012/l, Hb 134 g/l, Hct 0.39, MCV
41.9% and had a retention time of 4.38 minutes.
92 fl, MCH 31.6 pg and MCHC 346 g/l. You are pro
…………………………………………………………
vided with cellulose acetate electrophoresis at alka
line pH and agarose gel electrophoresis at acid pH.
Cellulose acetate
electrophoresis
at alkaline pH
Propositus
AFSC
Agarose gel
electrophoresis
at acid pH
A
F
Propositus
Exercise 8.30 basophilic stippling and some irregularly con

tracted cells. You are provided with haemoglobin
Haemoglobinopathy investigation was requested
electrophoresis on cellulose acetate at alkaline pH
on a 15‐year‐old boy hospitalised with vomiting
and an HPLC chromatogram (Bio‐Rad Variant II)
and diarrhoea. His ethnic origin was not totally
(done on washed cells). The smaller peak on the
clear; he was stated to be ‘black’ and had a name
right hand side of the chromatogram was in the S
that suggested he might be Arab or from a Muslim
window and was 20%; a sickle solubility test was
country. Red cell indices were: RBC 5.18 × 1012/l, Hb
positive. Haemoglobin F was 3% and haemoglobin
86 g/l, Hct 0.286, MCV 55 fl, MCH 16.6 pg, MCHC
A2 was quantified as 4%.
300 g/l and RDW 24%. His blood film showed
Explain your findings and give a provisional
hypochromia, microcytosis, numerous target cells,
diagnosis. ………………………………………………
AC
AC
AC
Patient
Patient
AFSC
394 Chapter 8
Exercise 8.31 F of 2.5%. Her red cell indices were: RBC 4.64 × 1012/l,
Hb 129 g/l, Hct 0.41, MCV 89 fl, MCH 27.7 pg and
Preoperative haemoglobinopathy screening was
MCHC 312 g/l.
requested in a 26‐year‐old woman. She was found
Does she have β thalassaemia trait? ………………
to have a haemoglobin A2 of 5.3% and haemoglobin
How would you proceed? ………………………
Exercise 8.32 are provided with a photograph of his blood film.

HPLC showed haemoglobin F 96.7% and haemo
A 13‐year‐old Ghanaian boy had splenomegaly and
globin A2 3.3%.
tired easily. One brother was similarly affected
What abnormalities are shown in the blood film?
while another had a milder anaemia. His FBC
…………………………………………………………
showed: RBC 5.32 × 1012/l, Hb 85 g/l, Hct 0.30,
MCV 57 fl, MCH 16.1 pg and MCHC 281 g/l. You
Exercise 8.33 saturation 8%. The infant’s mother was of Thai eth
nic origin and her father was Afro‐Caribbean. You
A 21‐month‐old infant was being followed in paedi
are provided with photographs of a blood film and
atric haematology outpatients after detection of an
an HPLC chromatogram (Bio‐Rad Variant II).
abnormality on neonatal haemoglobinopathy
What abnormalities are shown by the blood film?
screening. She was clinically well. Her FBC (with
…………………………………………………………
age‐appropriate reference ranges in brackets)
Considering the ethnic origin, what are the most
showed Hb 92 g/l (105–135 g/l), MCV 57.8 fl (70–85
likely diagnoses? ……………………………………
fl), MCH 20.7 pg (25–35 pg), MCHC 358 g/l (310–
Comment on the proportions of the variant haemo
350 g/l) and RDW 16.3% (12.5–15.5%). Serum iron
globins. …………………………………………………
was 5 μmol/l, transferrin 2.6 g/l and transferrin
396 Chapter 8
Exercise 8.34 been disclosed but her name suggested that she was
not Northern European. You are provided with her
A blood sample from a 31‐year‐old woman com
HPLC chromatogram (Bio‐Rad Variant II).
plaining of tiredness was sent for an FBC and hae
moglobinopathy investigations by her general
What is the most likely ethnic origin?
practitioner. This showed RBC 5.0 × 1012/l, Hb
…………………………………………………………
99 g/l, Hct 0.33, MCV 66 fl, MCH 20.1 pg, MCHC
300 g/l and RDW 20%. Her ethnic origin had not
Exercise 8.35 normal blood count and a positive sickle solubility

test.
What diagnoses can be made in this patient?
(Bio‐Rad Variant II) on an adult patient with a
…………………………………………………………
398 Chapter 8
Exercise 8.36 stove, the pipe of which is stated to have fitted badly.
It is believed that doctors will be able to save the life
A pregnant Afro‐Caribbean woman was known to
of Madame Zola, who was also affected by the nox-
have β thalassaemia trait and this was confirmed on
ious vapour.
her presentation to the antenatal clinic in her second
From such details as have hitherto been obtained
pregnancy. Her partner was tested and was found to
it seems that M. and Madame Zola returned yester-
have an Hb of 130 g/l, MCV 58 fl and MCH 17.8 pg.
day from the country where they had been staying
HLPC (Bio‐Rad Variant II) showed haemoglobin A2
for about three months. Their house in Rue de
3% and a peak in the haemoglobin S window of 2.5%.
Brielle’s was very cold, not having been inhabited
No further investigations are done.
for so long, and as there was a considerable fall in
At 5 months of age the baby was found to have
the temperature, M. Zola ordered the fire to be
anaemia (Hb 80 g/l), hepatosplenomegaly and
lighted in the grate of the bedroom, which is a vast
growth retardation. HPLC showed haemoglobin A
apartment. The footman set about lighting the fire,
51.5%, haemoglobin F 46% and haemoglobin A2 2.5%.
but it did not draw at all well. After dinner, which
What has gone wrong? ……………………………
M. and Madame Zola ate with good appetite they
retired to rest. That was about ten o’clock.
Exercise 8.37 This morning at 9.30 a.m. some workmen went to
The following is an extract from the Manchester the house to execute certain repairs which had been
Guardian of 30th September, 1902. ordered to be carried out in M. Zola’s room. The
TRAGIC DEATH OF E. ZOLA servants, who were already a little alarmed at hav-
ACCIDENTALLY ASPHYXIATED ing heard no sound in the bedroom, knocked loudly
MADAME ZOLA NEARLY SHARES HIS FATE at the door, which, as they received no response, they
We regret to announce the death, under circum broke in …. M. Zola was found lying half out of bed
stances of a most tragic character, of the …. He was quite dead. … Madame Zola was found
renowned French novelist, M. Emile Zola. The lying in bed showing no signs of life …. The fright-
following telegram explains how M. Zola died: ened servants instantly threw open the windows
Paris and gave the alarm.
M. Emile Zola was this morning found dead in his What is the ‘noxious vapour’ to which death is
house from accidental asphyxiation. Madame Zola attributed? ……………………………………………
is seriously ill. Explain the mechanism of death. …………………
2.00 p.m. The death of M. Zola appears to have Did the servants take the right action?
been caused by poisonous gases emitted from a …………………………………………………………
Answers Patient 5: SD (e.g. S plus D‐Punjab since the propor

tions are equal and a β chain variant is favoured).
Exercise 8.1 Patient 6: AS plus G‐Philadelphia.
Patient 7: AD plus G‐Philadelphia (i.e. heterozygo
Patient 1: AS. sity for a β chain variant designated D or G, e.g.
Patient 2: AD or AG (heterozygosity for a number of D‐Punjab, and an α chain variant designated D or
haemoglobins designated D or G, both α and β G, e.g. G‐Philadelphia).
chain variants, for example D‐Punjab, D‐Iran, D‐ Patient 8: AC.
Copenhagen, D‐Norfolk or G‐Philadelphia). Patient 9: AE (note that the quantity of the variant
Note: the quantity of the variant and haemoglo haemoglobin is less than in C trait and, in addi
bin A are similar so a β chain variant is favoured. tion, the mobility at acid pH is different).
Patient 3: A plus Lepore. Patient 10: A plus C‐Harlem.
Patient 4: SS or Sβ0 thalassaemia. Patient 11: A plus O‐Arab.
Exercise 8.2 Exercise 8.5

Patient 1: AS (sickle cell trait). There is a band with the mobility of S at alkaline
Patient 2: A plus D‐Punjab (the HPLC chromato pH, which we can deduce had the same mobility as
gram is typical of D‐Punjab heterozygosity). A at acid pH. On the HPLC chromatogram it is
seen that the variant haemoglobin is eluting in the
haemoglobin A2 window and, together with haemo
Exercise 8.3 globin A2, comprises 14.3% of total haemoglobin.
Haemoglobin E trait is most likely because of the Clearly it is not haemoglobin A2 on the grounds of
ethnic origin, the thalassaemic indices, the mobil the proportion of the variant and its electrophoretic
ity of the variant haemoglobin at alkaline pH mobility. These findings are indicative of haemo
and the relatively low percentage of the vari globin Lepore trait.
ant haemoglobin (the variant plus haemoglobin The significance is similar to that of β thalassaemia
A 2 appears to be around 30%). Haemoglobin C trait.
and haemoglobin O‐Arab are much less likely. For a bonus point you might have noted that
The provisional diagnosis could be confirmed the P2 fraction eluting after haemoglobin F, repre
by haemoglobin electrophoresis at acid pH or by senting glycosylated haemoglobin, is also quite
HPLC. high.
The condition of most potential significance in
the partner would be β thalassaemia trait since the
Exercise 8.6
compound heterozygous state for haemoglobin E
and β thalassaemia often leads to the clinical Father: β thalassaemia trait plus hereditary
picture of thalassaemia intermedia or even thalas elliptocytosis.
saemia major. Daughter aged 17: hereditary elliptocytosis.
Daughter aged 15: β thalassaemia trait.
Son aged 10: normal.
Exercise 8.4 The father’s blood film shows anisocytosis,
The findings are indicative of heterozygosity for δβ poikilocytosis including elliptocytes and ovalo
thalassaemia since there are thalassaemic indices cytes, microcytosis and hypochromia. There is a
with elevation of haemoglobin F but not haemoglo teardrop poikilocyte showing basophilic stippling.
bin A2. It would be sensible to test the patient’s His haemoglobin A2 is elevated. The film is more
father and, if he were normal, the patient’s mother, strikingly elliptocytic than is usual in β thalassaemia
to help confirm the diagnosis. trait and consideration of the conditions he has
The patient’s father was tested and was found to transmitted to two of his children indicates that he
have the same haematological abnormality as the is likely to have both β thalassaemia trait and
patient. The clinical significance is very similar to hereditary elliptocytosis.
the clinical significance of β thalassaemia hete
rozygosity. The patient should not take iron unless
she is demonstrated to be iron deficient and should
Exercise 8.7
be warned of the possibility of thalassaemia inter The clinical history and initial laboratory find
media or major in the fetus if her partner should ings suggest a diagnosis of iron deficiency anae
happen to be heterozygous for β (or δβ) thalassaemia. mia. The subsequent demonstration of a variant
If she were to become pregnant it should be noted haemoglobin, identified as haemoglobin S is
that her high fetal haemoglobin percentage would likely to be correct since it was based on electro
be expected to result in a positive Kleihauer test. phoresis at acid and alkaline pH and HPLC
If she were Rh D negative she would need an (although a positive sickle solubility test would
alternative test for detection of feto‐maternal not be expected). An explanation needs to be
haemorrhage. found for haemoglobin S being present in such
400 Chapter 8
a low concentration. There are two possible present, in addition to the major C band. A labora
explanations: tory error might be suspected as the explanation of
• She has been transfused with blood from a donor this discrepancy but in fact agarose gel electropho
with sickle cell trait. resis is more sensitive than cellulose acetate electro
• She herself has sickle cell trait and the percentage phoresis for the detection of a low concentration of
of the variant haemoglobin has been lowered by the normal and variant haemoglobins. Globin chain
recent transfusion. synthesis studies confirmed that the patient was a
It is unlikely that the patient has sickle cell compound heterozygote for βC and β+ thalassaemia
disease since there is nothing in the history to sug with greatly reduced synthesis of βA globin chain.
gest this. The fact that she is identified as ‘white’
does not exclude her having haemoglobin S.
However, haemoglobin electrophoresis on a resid Exercise 8.9
ual blood sample from each of the donor bags con
The mother has D or G trait. The father has sickle cell
firmed that the patient had been transfused with
trait. The child obviously has a sickling disorder and
blood from a donor with sickle cell trait.
must have inherited D from the mother and S from the
It is not rare for clinical staff to request investiga
father. The blood film of the child indicates that there
tion for a haemoglobinopathy when there is no
has been interaction between haemoglobin S and the
clear clinical indication for the test nor is it rare for
other variant haemoglobin, producing a clinically sig
a test to be requested, inappropriately, on a blood
nificant disorder. This is likely to be haemoglobin S/
sample taken after transfusion. The laboratory
haemoglobin D‐Punjab compound heterozygosity.
should be alert for this explanation of anomalous
Other haemoglobins designated D and G do not
results. Transfusion of blood from a donor with hae
interact adversely with haemoglobin S. Haemoglobin
moglobin C has also been reported.
D‐Punjab was confirmed on further testing.
This case shows the importance of performing
Ahmad E and Sykes E (1999) Clinical pathology
either HPLC or electrophoresis at acid pH in
rounds: low level of hemoglobin S in a white
patients who, on cellulose acetate electrophoresis,
woman. Lab Med, 30, 572–575.
Suarez AA, Polski JM, Grossman BJ and Johnston M have a single band with the mobility of haemoglo
(1999) Blood transfusion–acquired hemoglobin C. bin S. Misdiagnosis of compound heterozygous
Arch Pathol Lab Med, 123, 642–643. states as sickle cell anaemia may lead to paternity
being questioned, in cases in which the father does
not have haemoglobin S, with serious social and
Exercise 8.8
possibly legal consequences.
The blood film shows target cells, irregularly The situation is more straightforward if HPLC is
contracted cells and a haemoglobin C crystal.
the initial test, rather than cellulose acetate electro
Electrophoresis on cellulose acetate at alkaline pH phoresis or capillary electrophoresis, since S and
shows only haemoglobin C but on agarose gel D‐Punjab have different retention times (see facing
faint bands with the mobilities of F and A are also page).
HPLC chromatogram (Bio‐Rad Variant II) in another patient with haemoglobins S and D‐Punjab. The peaks from left
to right are altered haemoglobin F, haemoglobin F0 (shaded), post‐translationally modified haemoglobin S in the A0
window, haemoglobin A2, haemoglobin S and haemoglobin D‐Punjab.
402 Chapter 8
Exercise 8.10 thalassaemia. The genotype −α/−α is found in 1–2%

of Afro‐Caribbeans. The genotype − −/αα is very
Patient 1: SC.
rare in this ethnic group so is a much less likely
Patient 2: SE.
explanation. The S band is so faint that, except for
Patient 3: S plus C‐Harlem or SS plus G‐Philadelphia
the positive sickle solubility test, haemoglobin
(or other D/G α chain variant).
Lepore might have been suspected.
Patient 4: S plus O‐Arab.
Patient 5: CC plus G‐Philadelphia.
Patient 6: AC plus G‐Philadelphia. Exercise 8.13
Patient 7: SC plus G‐Philadelphia.
The blood film is dimorphic and shows target cells,
Patient 8: CC or Cβ0 thalassaemia.
Pappenheimer bodies and giant platelets. The bone
Patient 9: EE or Eβ0 thalassaemia.
marrow shows marked erythroid hyperplasia.
Patient 10: O‐Arab homozygote or O‐Arab/β0
Inherited haemoglobin H disease is unlikely in
thalassaemia.
view of the ethnic origin and the presence of hae
Patient 11: C plus C‐Harlem.
matological abnormalities indicative of a haemato
logical neoplasm. Because blasts are a high
Exercise 8.11 percentage of non‐erythroid cells the disease was
classified, according to the French–American–
The electrophoretic strip shows only haemoglobin
British (FAB) classification, as M6 acute myeloid
A2 and a variant haemoglobin that is slightly faster
leukaemia (AML) rather than as a myelodysplastic
than haemoglobin A. There is no common variant
syndrome (MDS); however, as blast cells were only
haemoglobin with this mobility. The two possible
3.6% of all nucleated cells the classification, accord
explanations are:
ing to the 2016 WHO Classification of Tumours
• compound heterozygosity for β0 thalassaemia
of Haemopoietic and Lymphoid Tissues, would
and an uncommon variant haemoglobin;
be as MDS. It is of interest that the patient has
• homozygosity for an uncommon haemoglobin.
both myelodysplastic features (anaemia and some
In the absence of consanguinity it is unlikely that
ring sideroblasts) and myeloproliferative features
the patient would be homozygous for an uncom
(thrombocytosis with giant platelets). This repre
mon variant haemoglobin. DNA analysis confirmed
sents an ‘overlap syndrome’, i.e. a condition with
that he was heterozygous for β0 thalassaemia and
overlapping myelodysplastic/myeloproliferative
electron spray mass spectrometry (with thanks to
features. This patient illustrates the particular asso
Dr Barbara Wild) identified the variant haemoglo
ciation of acquired haemoglobin H disease with a
bin as haemoglobin Tacoma.
myeloid neoplasm with prominent erythroid
As these investigations are said to have resulted
involvement.
from a ‘routine blood count’ it appears that the
patient is asymptomatic (this was confirmed) and
Bain BJ (1999) The relationship between the myelopro
the variant haemoglobin is therefore not likely to be liferative syndromes and the myeloproliferative
of any clinical significance. The β thalassaemia trait, disorders. Leuk Lymphoma, 34, 443–449.
however, could have genetic significance. Mercieca J, Bain B, Barbour G and Catovsky D (1996)
Teaching cases from the Royal Marsden and St Mary’s
Hospitals Case 10 Microcytic anaemia and throm
Exercise 8.12 bocytosis. Leuk Lymphoma, 21, 185–186.
The patient has sickle cell trait since she has a positive
sickle solubility test and a variant haemoglobin with
Exercise 8.14
the mobility of haemoglobin S. However, the haemo
globin S percentage is unusually low. This, together The blood film shows that the patient has the features
with the thalassaemic indices, suggests that as well expected in sickle cell anaemia, including hyposplenic
as having sickle cell trait she is homozygous for α+ features (a target cell and a Howell–Jolly body).
The CT scan shows that rather than having splenic haemoglobin D‐Punjab, which does interact.
atrophy, the patient’s spleen is of normal size and Although the precise variant was not identified in
abnormally dense. This unusual appearance sug this case, it was shown by HPLC not to be haemo
gests that there is deposition of calcium in the globin D‐Punjab.
spleen, as a result of recurrent splenic infarction. Only the haemoglobin S heterozygosity is likely
Despite the normal‐sized spleen the patient has to be clinically significant. If she requires a general
functional hyposplenism. anaesthetic for drainage of the breast abscess the
anaesthetist will wish to know that she has sickle
cell trait. This would also be of potential genetic
Exercise 8.15
significance.
It is clear from the history and the blood film that The anaemia and microcytosis could be caused
the patient has some type of sickle cell disease. by the effects of the infection, if it has been going
Haemoglobin electrophoresis at alkaline pH sug on for some time, leading to anaemia of chronic
gests possible compound heterozygosity for hae disease. Alternatively, the patient could have a
moglobins S and C. However at acid pH it is clear coincidental iron deficiency anaemia. α thalassae
that there is no haemoglobin C present. Compound mia trait is also quite likely in this ethnic group.
heterozygosity for haemoglobin S and C‐Harlem
should be suspected (and was the answer given in
Exercise 8.17
the first edition of this book)*. However, further
investigation including family studies, citrate agar The patient is heterozygous for both βS and
electrophoresis and mass spectrometry led to a αG‐Philadelphia, hence the three bands on electrophoresis
diagnosis of compound heterozygosity for haemo at alkaline pH. The haemoglobin G‐Philadelphia is
globin S and haemoglobin O‐Arab. The variable very unlikely to be of any clinical significance.
mobility of haemoglobin O‐Arab on electrophore However if the patient’s partner also has sickle cell
sis at acid pH can cause problems in diagnosis of trait there is a one in four risk of the fetus having
compound heterozygous states. These problems do sickle cell anaemia. There would also be significant
not arise in the simple heterozygous state since genetic implications if the patient’s partner had β
haemoglobin C‐Harlem has a positive sickle solu thalassaemia trait, haemoglobin C, haemoglobin
bility test and haemoglobin O‐Arab does not. D‐Punjab or haemoglobin O‐Arab. The couple
*Haemoglobin O‐Arab has been similarly misiden concerned might wish to consider termination of
tified as haemoglobin C‐Harlem by others. HPLC pregnancy if significant fetal disease were predicted.
can be useful. Red cell indices and haemoglobin electrophoresis or
HPLC should be performed on her partner, fol
Joutovsky A and Nardi M (2004) Hemoglobin C and lowed, if indicated, by consideration of antenatal
hemoglobin O‐Arab variants can be diagnosed diagnosis of any significant abnormality in the fetus.
using the Bio‐Rad Variant II high‐performance
liquid chromatography system without further
confirmatory tests. Arch Pathol Lab Med, 128, Exercise 8.18
435–439. As the patient has marked microcytosis but a nor
mal Hb she is unlikely to have iron deficiency. The
red cell indices are suggestive of thalassaemia and
Exercise 8.16
as she has a normal haemoglobin A2 percentage it is
The patient is a compound heterozygote for hae likely that she has α thalassaemia. Since she is
moglobin S and a β chain variant, haemoglobin D Chinese she could be heterozygous for α0 thalassae
or G. Since she is asymptomatic and the blood film mia (− −/αα) or homozygous for α+ thalassaemia
does not show any features of sickle cell disease, the (−α/−α). The microcytosis is too marked for hete
second haemoglobin is likely to be a variant that rozygosity for α+ thalassaemia to be a likely diagnosis.
does not interact with haemoglobin S, rather than The implications are as shown in the table:
404 Chapter 8
Abnormality present in mother Findings in partner Possible abnormality in fetus
−α/−α αα/αα −α/αα (α thalassaemia trait)

−α/αα −α/αα or −α/−α (α thalassaemia trait)
−α/−α −α/−α (α thalassaemia trait)
− −/αα −α/αα (α thalassaemia trait) or −α/− − (haemoglobin H disease)
− −/αα αα/αα αα/αα (normal) or − −/αα (α thalassaemia trait)
−α/αα αα/αα (normal), −α/αα (α thalassaemia trait), − −/αα
(α thalassaemia trait) or − −/−α (haemoglobin H disease)
−α/−α −α/αα (α thalassaemia trait) or − −/−α (haemoglobin H disease)
− −/αα αα/αα (normal) or − −/αα (α thalassaemia trait) or − −/− −
(haemoglobin Bart’s hydrops fetalis)
β thalassaemia trait Beware: a diagnosis of β thalassaemia trait in the partner does
not exclude his also having α thalassaemia trait; molecular
analysis to exclude − −/αα is indicated
Exercise 8.19 for triple α. This condition was harmless in the

mother but in the daughter it aggravated the chain
The high haemoglobin F with normal red cell indices
imbalance attributable to the β thalassaemia trait
is likely to be caused by deletional hereditary persis
and led to a more severe phenotype.
tence of fetal haemoglobin. This has no clinical signifi
cance except that the Kleihauer test will be positive.
Bain BJ, Swirsky D, Bhavnani M, Layton M, Parker
The patient is Rh D negative and an alternative tech N, Makris M et al. (2001) British Society for Haema
nique will have to be used after delivery to detect and tology Slide Session, Annual Scientific Meeting,
quantitate any fetal cells in the maternal circulation. Bournemouth, 2000, Clin Lab Haematol, 23, 265–269.
Exercise 8.20
Exercise 8.22
The findings are those of juvenile myelomonocytic
leukaemia. An increased haemoglobin F percentage The findings are those of heterozygosity for δβ
for age is one of the criteria that can be used for thalassaemia. Hereditary persistence of fetal hae
making this diagnosis. The low haemoglobin A2 moglobin is excluded by the ‘thalassaemic’ red cell
and low carbonic anhydrase show reversion to fetal indices. Note that in addition to the increased hae
type erythropoiesis. moglobin F (shaded) shown on HPLC, there are
complex early peaks (to the left), which represent
post‐translationally modified haemoglobin F.
Exercise 8.21
It appears that both the daughter and the father are
Exercise 8.23
heterozygous for β thalassaemia since they both
have microcytosis and an increased haemoglobin The most likely diagnosis in the mother is β thalas
A2 percentage. This was confirmed on molecular saemia trait and in the daughter is compound het
analysis, both having the IVS1 5 G→C mutation. erozygosity for haemoglobin S and β thalassaemia
Possible explanations of the more severe phenotype (note the nucleated red blood cell and, on the right
in the daughter include: of the image, one sickle cell). In fact she had S/β0
• coinheritance of a ‘silent’ β thalassaemia allele thalassaemia.
from the mother;
• coinheritance of homozygosity or heterozygosity
Exercise 8.24
for triple α.
The latter explanation was found to be correct; Haemoglobin S/β+ thalassaemia compound
the mother and the daughter were heterozygous heterozygosity.
Exercise 8.25 Exercise 8.29

The most likely diagnosis is methaemoglobinaemia The variant haemoglobin has electrophoretic charac
caused by exposure to a toxic substance found in teristics suggestive of haemoglobin E but it is
the garden shed. odd that the patient is Caucasian. In addition,
Methaemoglobin should be tested for by spec the proportions of haemoglobin A and the variant
trometry or by co‐oximetry. A co‐oximeter is an haemoglobin appears to be similar, a very unlikely
instrument that passes monochromatic light at four finding if this were haemoglobin E. The results of
wavelengths through the test sample and is thus HPLC analysis exclude the possibility of haemoglobin
able to quantitate carboxyhaemoglobin, methaemo E since there is a variant haemoglobin in the
globin, oxyhaemoglobin and haemoglobin. ‘S window’. These are the features of haemoglobin
The correct treatment is intravenous methylene blue. E‐Saskatoon, which does not have the genetic impli
cations of haemoglobin E. This case shows that
Wentworth P, Roy M, Wilson B, Padusenko J, Smeaton A
even with two independent methods, a provisional
and Burchell N (1999) Clinical pathology rounds:
identification can be wrong.
toxic methemoglobinemia in a 2‐year‐old child.
Lab Med, 30, 311–315.
Exercise 8.30
Exercise 8.26
The HPLC chromatogram and haemoglobin electro
A variant haemoglobin is present, suggesting that
phoresis show the presence of haemoglobin S at an
the polycythaemia is the result of a high affinity
unusually low level of 20%. Haemoglobin A2 appears
haemoglobin. This was haemoglobin Kempsey and
slightly elevated but its quantification can be inac
although the variant haemoglobin appears in the ‘D
curate in the presence of haemoglobin S as post‐
window’ of the HPLC chromatogram its curious
translationally modified haemoglobin S appears in the
shape is characteristic of haemoglobin Kempsey.
A2 window. There is a slight increase in haemoglobin
F. There is an abnormal fraction eluting early, which
Exercise 8.27
has the form expected of haemoglobin Bart’s (although
The grossly increased ‘P2 fraction’ has nothing to no haemoglobin Bart’s was visible on haemoglobin
do with diabetes, although it would give a factitious electrophoresis). A haemoglobin H preparation was
result on haemoglobin A1c quantification. It should negative. These findings are the consequence of sickle
be recognised as a variant haemoglobin. It was cell trait plus the genotype of haemoglobin H disease.
identified by mass spectrometry as haemoglobin The reported levels of haemoglobin S in this condi
Hope (with thanks to Dr Barbara Wild). tion vary between 10 and 25%. Haemoglobin H is
present in only very trivial amounts.
Exercise 8.28
The recovery with methylene blue suggests Exercise 8.31
methaemoglobinaemia.
Given the father’s occupation, the likely underly The possibility of elevation of haemoglobin A2 for
ing cause is accidental exposure to sodium nitrite, another reason (e.g. treatment of retroviral infection)
used for curing meat. should be considered. Alternatively, could the
In the family described, the sodium nitrite had indices be atypical because of coexisting liver
been introduced into the domestic environment disease, megaloblastic anaemia (unlikely as the
for use as an insecticide and had been emptied haemoglobin concentration is normal) or hydrox
into a sugar bowl by one of the children. ycarbamide therapy (not very likely in a young
pre‐operative patient)?
Finan A, Keenan P, O’Donovan FO’, Mayne P and The explanation was found, on DNA analysis, to
Murphy J (1998) Methaemoglobinaemia associated be coexisting α and β thalassaemia trait, known
with sodium nitrite in three siblings. Br Med J, 317, to normalise the red cell indices but not the ele
1138–1139. vated haemoglobin A2 that would be expected in
406 Chapter 8
β thalassaemia trait. Specifically she had −α/αα and which is typical of this condition (although in this
the β+ thalassaemia mutation −29A→G. instance it is partly artefactually lowered as the
baseline is high).
The most likely ethnic origin is South‐East Asian
Exercise 8.32
(Thai, Cambodian, Laotian, Malaysian or Filipino)
The blood film shows hypochromia, microcytosis or Chinese.
and poikilocytosis. The nucleated red blood cell is a
micronormoblast with defective haemoglobinisa
Exercise 8.35
tion. The clinical features, in the light of the total
absence of haemoglobin A, are indicative of β tha The patient has sickle cell trait and diabetes
lassaemia intermedia. mellitus. The peaks from left to right are: injection
This was found to be due to compound heterozy artefact, haemoglobin F, glycated haemoglobin A
gosity for β0 thalassaemia (mutation IVS2 849 A→G) (8.3%), other post‐translationally modified hae
and the Ghanaian deletional hereditary persistence moglobin A, haemoglobin A0 with the shoulder
of fetal haemoglobin (HPFH‐2). The phenotype of on the left of the main peak representing glycated
this compound heterozygous state is often milder haemoglobin S, haemoglobin A2 (plus some post‐
than in this patient. translationally modified haemoglobin S) and
haemoglobin S0.
It is important to recognise the evidence of diabetes
Exercise 8.33
mellitus and draw this to the attention of clinical
The blood film shows target cells and irregularly staff if this diagnosis is not already known.
contracted cells.
The infant has biochemical evidence of iron defi
Exercise 8.36
ciency. Considering the ethnic origin, the variant
haemoglobins are likely to be haemoglobin E (from The results of HPLC were misinterpreted in the
her Thai mother) and haemoglobin C (from her Afro‐ father. A small peak in the haemoglobin S window
Caribbean father). This provisional diagnosis was could represent carry over from a previous speci
confirmed by cellulose acetate electrophoresis at men but in fact he had haemoglobin A2′ and this
alkaline pH and acid agarose electrophoresis. Note should have been added to haemoglobin A2; the
that haemoglobin E (plus A2) is only 30.4% while sum of haemoglobin A2 and the variant of 5.5%
haemoglobin C is 57.8%, reflecting the fact that hae would have revealed that he, as well as the mother,
moglobin E is a thalassaemic haemoglobinopathy. had β thalassaemia trait. The baby has β thalassae
Not also the two peaks of post‐translationally modi mia major.
fied haemoglobin E and two of post‐translationally
modified haemoglobin C.
Exercise 8.37
Spencer‐Chapman M, Kiritkumar K, Lund K and Bain BJ The noxious vapour was carbon monoxide.
(2019) An unusual hemoglobinopathy: compound Carbon monoxide poisoning causes death by
heterozygosity for hemoglobins C and E. Am J asphyxiation. Carboxyhaemoglobin has no oxygen‐
Haematol, 94, 144. combining activity and, in addition, increases the
oxygen affinity of the remaining haemoglobin, fur
ther impairing oxygen deliver to tissues.
Exercise 8.34
The servants’ instinctive action in throwing open
The HPLC chromatogram shows the double peak of the windows has a sound physiological basis since
haemoglobin H that is characteristic of this instru carboxyhaemoglobin is slowly converted to oxy
ment. This finding, together with the blood count, haemoglobin on breathing room air. Removing
confirms a diagnosis of haemoglobin H disease. Madame Zola to another room may have been even
Note the low haemoglobin A2 percentage (1.4%) more effective.
Appendix: electronic resources
Sickle cell disease (specific) Sickle Cell Society. Standards for the Clinical Care of
Adults with Sickle Cell Disease in the UK, 2018.
Guidelines https://www.sicklecellsociety.org/wp‐content/
uploads/2018/04/Web‐version‐FINAL‐SCS‐
British Society for Haematology. Management of Standards‐GSM‐6.4.18.pdf
Acute Chest Syndrome in Sickle Cell Disease, 2015.
https://b‐s‐h.org.uk/guidelines/guidelines/
management‐of‐acute‐chest‐syndrome‐in‐sickle‐ Websites
cell‐disease
American Society of Hematology. (Sickle cell a dvocacy.)
British Society for Haematology. Red Cell Transfusion in
h t t p : / / w w w. h e m a t o l o g y. o r g / A d v o c a c y /
Sickle Cell Disease, Parts I and II, 2016. https://b‐s‐h.org.
4329.aspx
uk/guidelines/guidelines/red‐cell‐transfusion‐in‐
European Network for Rare and Congenital Anaemias
sickle‐cell‐disease‐part‐l and https://b‐s‐h.org.uk/
(ENERCA) (2012) (Diagnosis and management of
guidelines/guidelines/red‐cell‐transfusion‐
sickle cell disease.) https://www.enerca.org/
in‐sickle‐cell‐disease‐part‐ii
activities/training/enerca‐recommendations‐
British Society for Haematology. Guidelines for the Use of
2013.html
Hydroxycarbamide in Children and Adults with Sickle
Sickle Cell Disease Association of America.
Cell Disease, 2018. https://b‐s‐h.org.uk/guidelines/
(Patient support group.) www.sicklecelldisease.org
guidelines/guidelines‐for‐the‐use‐of‐
Sickle Cell Information Centre, Atlanta, Georgia.
hydroxycarbamide‐in‐children‐and‐adults‐with‐
http://www.emory.edu/PEDS/SICKLE
sickle‐cell‐disease
Sickle Cell Society. (UK patient support group.)
National Institute for Health and Care Excellence
https://www.sicklecellsociety.org
(NICE). Sickle Cell Disease: Managing Acute Painful
Episodes in Hospital. Clinical guideline CG143, 2012.
https://www.nice.org.uk/guidance/cg143
Thalassaemia (specific)
National Institute for Health and Care Excellence
(NICE). Sickle Cell Disease. Quality standard QS58,
Guidelines
2014. https://www.nice.org.uk/guidance/qs58
National Institute for Health and Care Excellence Stephens AD, Angastiniotis M, Baysal E, Chan V,
(NICE). Sickle Cell Disease: Acute Painful Episode Fucharoen S, Giordano PC et al.; International
Overview, 2017. https://pathways.nice.org.uk/ Council for the Standardisation of Haematology
pathways/sickle‐cell‐disease‐acute‐painful‐episode (ICSH) (2012) ICSH recommendations for the
Public Health England. Sickle Cell Disease in Children: measurement of haemoglobin A2. Int J Lab Hematol,
Standards for Clinical Care, 2010. https://www.gov. 34, 1–13. doi: 10.1111/j.1751‐553X.2011.01368.x.
uk/government/publications/sickle‐ https://onlinelibrary.wiley.com/doi/full/10.1111/
cell‐disease‐in‐children‐standards‐for‐clinical‐care j.1751‐553X.2011.01368.x

407
408 Appendix: electronic resources
Stephens AD, Angastiniotis M, Baysal E, Chan V, Davis https://www.gov.uk/topic/population‐screening‐

B, Fucharoen S et al.; International Council for The programmes/sickle‐cell‐thalassaemia
Standardisation of Haematology (ICSH) (2012) Public Health England. Sickle Cell and Thalassaemia
ICSH recommendations for the measurement Screening: Handbook for Laboratories, 2017. https://
of haemoglobin F. Int J Lab Hematol, 34, 14–20. www.gov.uk/government/publications/sickle‐
doi: 10.1111/j.1751‐553X.2011.01367.x. https:// cell‐and‐thalassaemia‐screening‐handbook‐for‐
onlinelibrary.wiley.com/doi/full/10.1111/ laboratories
j.1751‐553X.2011.01367.x Public Health England. NHS Sickle Cell and Thalassaemia
Screening Programme. Handbook for Antenatal
Laboratories, 4th edn, 2017. https://assets.
Websites publishing.service.gov.uk/government/uploads/
system/uploads/attachment_data/file/656094/
Bluebird bio Inc. Challenge TDT. (Educational website
Antenatal_Laboratory_Handbook.pdf
on transfusion‐dependent β thalassaemia.) https://
Public Health England. NHS Sickle Cell and Thalassaemia
www.challengetdt.com
Screening Programme. Handbook for Newborn Laboratories,
Cooley’s Anemia Foundation. (Patient support group.)
4th edn, 2017. https://assets.publishing.service.gov.
https://www.thalassemia.org
uk/government/uploads/system/uploads/
Thalassaemia International Federation. http://
attachment_data/file/585126/NHS_SCT_Handbook_
thalassaemia.org.cy
for_Newborn_Laboratories.pdf
Thalassaemia Society of NSW. (Patient support group.)
Public Health England. Sickle Cell and Thalassaemia
https://thalnsw.org.au
Screening Handbook, 2018. https://www.gov.uk/
UKTS, previously UK Thalassaemia Society. (Patient
government/publications/handbook‐for‐sickle‐cell‐
support group.) https://www.ukts.org
and‐thalassaemia‐screening
General (both sickle cell Websites

and thalassaemia, or other
Bio‐Rad Laboratories. Bio‐Rad Library of Variants.
haemoglobinopathy)
https://hemoglobins.bio‐rad.com/Pages/
StartPage.aspx
Guidelines
Ithanet. (An international thalassaemia and haemoglo-
Accessible Publication of Genetic Information (APoGI), binopathy genetic information resource.) http://
University College, London. http://2000.apogi. www.ithanet.eu
info/data/html/hb/menu.htm Public Health England. NHS Newborn Blood Spot (NBS)
British Society for Haematology. Significant Screening Programme. https://www.gov.uk/topic/
Haemoglobinopathies: Guidelines for Screening and population‐screening‐programmes/newborn‐
Diagnosis, 2010. https://b‐s‐h.org.uk/guidelines/ blood‐spot
guidelines/significant‐haemoglobinopathies‐ South Thames Sickle Cell and Thalassaemia Network.
guidelines‐for‐screening‐and‐diagnosis http://www.ststn.co.uk
Daniel Y and Henthorn J. Tandem Mass Spectrometry for Thalassaemia and Sickle Cell Australia. (Patient sup-
Sickle Cell and Thalassaemia Newborn Screening Pilot port group.) https://www.tasca.org.au
Study, 2015. https://assets.publishing.service.gov. The Globin Gene Server, hosted by Pennsylvania State
uk/government/uploads/system/uploads/ University, USA and McMaster University, Canada.
attachment_data/file/488858/Tandem_Mass_ http://globin.cse.psu.edu
Spectrometry_for_Sickle_Cell_and_Thalassaemia_ UK Forum on Haemoglobin Disorders. (A multidisci-
Newborn_Screening_Pilot_Study_2015.pdf plinary group of health care professionals involved
Public Health England. NHS Sickle Cell and Thalassaemia in thalassaemia and sickle cell disease care.)
(SCT) Screening Programme, 2013 with later updates. https://www.haemoglobin.org.uk
Appendix: electronic resources 409
Genes Ithanet. (Resource provided by an international coop-

erative group giving genetic information on hae-
Accessible Publication of Genetic Information (APoGI), moglobinopathies.) http://www.ithanet.eu/db/
University College, London. http://2000.apogi. ithagenes
info/data/html/hb/menu.htm The Globin Gene Server. http://globin.cse.psu.edu
Index
Notes: Page numbers in italics refer at acid pH 35–37, 36 deletions 87, 88, 88, 96
to figures; those in bold to tables. at alkaline pH 34–35, 37 mutations 90
This index is in word‐by‐word order. case studies 365, 376, 398, 402 α thalassaemia 86, 86–116
Greek characters are indexed as if all‐trans‐retinoic acid 341 antenatal screening 351,
spelt out in full, i.e. α = alpha. almost silent β thalassaemia 351–353, 352
Prefixes to chemicals, globin genes trait 130–131, 132 β thalassaemia coinheritance 133
and thalassaemias are not ignored α chain case study 384, 403–404
in the alphabetical order; thus free 144, 148, 215, 304 coinheritance 116
δ‐aminolaevulinic acid starts with ‘d’ precipitates deletional 87–88, 88
and β thalassaemia starts with ‘b’. β thalassaemia major 144, DNA analysis 77, 77–78, 78, 79
147, 148, 149 ethnic‐specific mutations 91
acetylation β thalassaemia trait 123, 126 globin chain synthesis ratio 75
co‐translational 12 haemoglobin E haemoglobin E/β thalassaemia
post‐translational 4, 20 homozygosity 284 coinheritance 287
acetyltransferases 12 haemoglobin H disease 105 haemoglobin G‐Philadelphia
acquired abnormalities 325–347 production, α thalassaemia 87 coinheritance 262, 297
acquired thalassaemia 325, synthesis 3, 3, 12 major see haemoglobin Bart’s
328, 328 α chain variants 22, 261 hydrops fetalis
case study 379, 402 % of total haemoglobin 22–24, non‐deletional 87, 88–90, 89
acute chest syndrome 203, 209 262 prevalence 92–95
acute hepatic crisis 204 cellulose acetate electrophoresis 33 sickle cell/β thalassaemia
acute intrahepatic cholestasis 204 high affinity 312 coinheritance 230
acute lymphoblastic leukaemia low affinity 313 unstable haemoglobin
194, 325–326 M haemoglobins 308, 309, 309 coinheritance 306
adrenal insufficiency 145 sickle cell anaemia α thalassaemia trait
adults coinheritance 221–222 acquired 325
haemoglobin F 15 sickle cell trait coinher- β thalassaemia coinheritance 133
haemoglobins 2, 3, 4 itance 200–201, 201 β thalassaemia trait coinheritance
African haplotypes, βS gene unstable 89, 90, 301, 302 133, 394, 405–406
haemoglobin F levels 218, 219 α globin genes 9 haemoglobin C
origin and spread 185, 186, acquired thalassaemia 326 coinheritance 265
186–187 cluster 8–9, 9 sickle cell anaemia coinheritance
A
γ globin chain 2 deletions 87, 87–88, 88 207, 207, 208, 210, 221
polymorphisms 2, 219–220 duplication 86, 86 sickle cell/haemoglobin C disease
variants 313–315 mutations 22, 85, 87–90, 89, 91 coinheritance 224, 229
A
γδβ thalassaemias 86, 149–153 see also α1 gene; α2 gene sickle cell trait coinheritance
agarose gel electrophoresis α‐locus control region (LCRA) 9, 11 194, 194

411
412 Index
α0 thalassaemia 87 ααα see triple α ATRX gene mutations

antenatal screening 103, 350, ααT thalassaemia 95 acquired somatic 105, 327–328
351, 352, 353 α:β globin chain synthesis ratio 75 inherited 20, 90
deletional 87, 88, 88, 100 acquired thalassaemias 325, 326 ATRX syndrome 88, 89, 90, 105
DNA analysis 75, 76, 103 inherited thalassaemias 75 clinical features 106
ethnic groups 91, 100 αTα thalassaemia 95 laboratory features 113
heterozygosity see α0 amniocentesis 353 auto‐oxidation, haemoglobin 1, 336
thalassaemia trait anaemia
homozygosity see haemoglobin α0 thalassaemia 100 babesiosis, transfusion‐transmitted
Bart’s hydrops fetalis α+ thalassaemia 96 205, 229
immunological detection 70 β thalassaemia major 144 Bantu haplotype, βS gene 185, 186
non‐deletional 89 haemoglobin Bart’s hydrops haemoglobin F levels 218,
osmotic fragility test 72 fetalis 114, 115–116 219, 220
α0 thalassaemia trait 100–103 haemoglobin C/β sickle cell/haemoglobin C
coinheritance 102–103 thalassaemia 270 disease 226
diagnosis 100–102 haemoglobin Constant basophilic stippling
haemoglobin E coinheritance Spring 98 β thalassaemia trait 123, 124
102–103, 283, 284, 290 haemoglobin H disease 103, dominant β thalassaemia 136, 136
laboratory features 100, 101–102 105–106, 107 haemoglobin Constant
prevalence 92–95 sickle cell anaemia 204–205, Spring 98, 98
α1 (HBA1) gene 9, 9 207–208 BCL11A 11
deletions 87–88, 88 sickle cell/β thalassaemia 230 BCL11A gene
mutations 87–90, 89, 95 unstable haemoglobins 303 haploinsufficiency 20
splice site mutation 88–89 see also haemoglobin polymorphism 15, 16, 221
α2 (HBA2) gene 9, 9 concentration reduced expression 15
deletions 87–88, 88, 95 anaesthesia, testing prior to Benin haplotype, βS gene 185, 186
mutations 87–90, 89, 95 357–359, 358 haemoglobin F levels 218, 219
α+–α0 thalassaemia 89, 89 anisochromasia 123, 124 sickle cell/haemoglobin C
α+ thalassaemia 87, 95–100 antenatal screening/ disease 226
clinical significance 96 testing 349–353 β chain
compound heterozygosity 95–100 conditions predicted 349, 350 precipitates 110, 110
deletional 87, 87, 88, 95 haemoglobin Bart’s hydrops synthesis 3, 3, 10, 10–12
diagnosis 99–100 fetalis 103, 349, 351 β chain variants 22, 261
ethnic groups 91, 95 problems and pitfalls 353 % of total haemoglobin 24,
haemoglobin E selective 351, 352 261–262
coinheritance 103, 283 universal 351, 351, 352 α thalassaemia
heterozygosity 95–100 ANTXR1 gene 16, 220 coinheritance 116
coinheritance 100 aplastic anaemia β thalassaemia 117, 119–120
diagnosis 99–100 congenital 164 haemoglobin D‐Punjab
laboratory features 96–99 low haemoglobin A2 332 vs 290–291
prevalence 92–95 see also red cell aplasia high affinity 310–311, 312
homozygosity 95–100 Arab‐Indian haplotype, βS low affinity 313
coinheritance 100 gene 186, 186 M haemoglobins 309, 309
diagnosis 99–100 haemoglobin F levels 218–220 not interacting with
laboratory features 96, sickle cell/α thalassaemia haemoglobin S 243
96–99, 97 trait 207 unstable 117, 119, 301, 302
neonates 98–99 sickle cell/β thalassaemia 230 β globin gene 9, 9
non‐deletional 89, 89 ASH1L gene mutations 20, 121 cluster 8–9, 9
sickle cell anaemia coinheritance assisted reproduction mutations 22, 24
see sickle cell/α+ thalassaemia technology 353 consequences 23, 117, 119–120
Index 413
dominant β sickle cell coinheritance see sickle osmotic fragility test 72, 128
thalassaemia 136–137 cell/β thalassaemia prevalence 92–95
sites 120, 121 unstable haemoglobin quadruple α coinheritance
somatic 141, 142 coinheritance 306–307 134, 142
β‐locus control region (LCRB) 9, 11 β thalassaemia intermedia 121, sickle cell trait coinher-
deletions 153 139–142 itance 196, 201
polymorphisms, haemoglobin F acquired 329 silent see silent β thalassaemia trait
levels and 15, 16, 160 clinical features 139–141, 140 somatic mutations with 329
β thalassaemia 86, 116–149 genotypes 134, 135, 141 triple α coinheritance see triple
acquired 325, 328, 328 laboratory features 141–142, α/β thalassaemia trait
α thalassaemia 142–144 βS gene see sickle cell gene
coinheritance 133 prenatal diagnosis 141 BGLT3 long non‐encoding
antenatal testing 350 unstable haemoglobins 302 RNA 15
β0 116, 120, 120 β thalassaemia major 121, 142–149 bilirubin
β+ 116, 119–120, 121 case study 398, 406 haemoglobin E
compound heterozygosity clinical features 144–146, homozygosity 284
116–117, 121 145–146 haemoglobin H disease 110
deletional 117, 119 dominant β thalassaemias 136 HPLC interpretation 44–46, 46
DNA analysis 76, 76–77 genotypes 142–144 sickle cell anaemia 216
dominant 134–137, 136–137 globin chain synthesis ratio 75 bite cells 304, 338
genetic defects 116–121, haemoglobin F 148, 150, 164 blood count 30–31
119–120, 120–121 laboratory features 146–149, α0 thalassaemia 100
haemoglobin C coinheritance see 147–150 α+ thalassaemia 96, 97
haemoglobin C/β β thalassaemia minor 121, 134 β thalassaemia trait 122
thalassaemia β thalassaemia trait 121–134 haemoglobin C/β
haemoglobin D‐Punjab almost silent 130–131, 132 thalassaemia 271
compound heterozygosity α thalassaemia haemoglobin C disease 267–268
293–294, 296 coinheritance 133 haemoglobin C/haemoglobin
haemoglobin E coinheritance see α thalassaemia trait coinher- E 274–275
haemoglobin E/β itance 133, 394, 405–406 haemoglobin C trait 264
thalassaemia antenatal screening 103, 350, haemoglobin D‐Punjab/β
haemoglobin Lepore compound 351, 351–353, 352 thalassaemia 293
heterozygosity 139 case study 398, 406 haemoglobin D‐Punjab
haemoglobin O‐Arab clinical features 121–122 disease 291–292
coinheritance 299–300, 300 coinheritance 133–134, 135 haemoglobin E/β
haemoglobin Tacoma δ thalassaemia/chain variant thalassaemia 287
coinheritance 377, 402 coinheritance 133 haemoglobin E/haemoglobin H
hereditary elliptocytosis diagnosis 124–133 disease 289
coinheritance 134, globin chain synthesis ratio 75 haemoglobin E
370–371, 399 haemoglobin A2 see haemoglobin homozygosity 283
heterozygosity see β thalassaemia A2, β thalassaemia trait haemoglobin E trait 279
trait haemoglobin A2’ 127–128, 129, haemoglobin H disease 106
high affinity haemoglobin 314, 315, 315 haemoglobin G‐Philadelphia
coinheritance 313 haemoglobin F 128, 129–130, 163 trait 296
homozygosity 116, 121 haemoglobin H disease haemoglobin M 309
HPFH coinheritance 134, coinheritance 113, 133, 134 high affinity haemoglobins 311
135, 160 HPLC 127, 129 pre‐anaesthesia 357
case study 394, 406 laboratory features 122–124 resource‐poor settings 360
inclusion body 136 neonatal period 123–124, 130, sickle cell anaemia 207–210, 209
non‐deletional 116–117, 119–120 130, 356–357 sickle cell/β thalassaemia 230
414 Index
blood count (cont’d) methaemoglobinaemia 336, 338 sickle cell/haemoglobin D‐

sickle cell/haemoglobin C pre‐anaesthesia 357 Punjab disease 233, 237
disease 223–224 sickle cell anaemia 210–214, unstable haemoglobins 306, 307
sickle cell trait 194 211–215, 224 bone marrow infarction 203, 222
unstable haemoglobins 303–304 sickle cell/β thalassaemia 230, bony deformity
see also red cell indices 231, 232 β thalassaemia major 144, 145
blood films 30–31 sickle cell/haemoglobin C haemoglobin E/β
acquired haemoglobin H disease 224, 224–226, thalassaemia 286
disease 326, 327, 327 225–226 haemoglobin H disease 106
α0 thalassaemia 100, 101 sickle cell/haemoglobin D‐ sickle cell anaemia 204, 205
α+ thalassaemia 96, 98 Punjab disease 233, 237 breast abscess, case study 382, 403
β thalassaemia intermedia 141– sickle cell/haemoglobin E buffering action, haemoglobin 1
142, 142–143 disease 239, 242
β thalassaemia major 146–148, sickle cell/haemoglobin O‐Arab Cameroon haplotype, βS gene
147–148 disease 233, 239 185, 218
β thalassaemia trait 123, sickle cell/HPFH 237, 242 capillary electrophoresis 38, 39, 40
124–125 sickle cell trait 194, 195 antenatal screening 351, 351
δβ thalassaemia 150, 150, 152 unstable haemoglobins 304, 305 β thalassaemia trait 127
dominant β thalassaemia 136, blood gas analysis 312 haemoglobin A2
136 blood samples 30 quantification 66–67
εγδβ thalassaemia 153, 154 neonatal screening 354, haemoglobin C/β
haemoglobin Bart’s hydrops 354–355 thalassaemia 274
fetalis 116, 116 blood transfusions haemoglobin C trait 264–265, 267
haemoglobin C/β thalassaemia β thalassaemia minor 139 haemoglobin D‐Punjab
224, 271, 272 donor with sickle cell trait 372, disease 295
haemoglobin C disease 224, 399–400 haemoglobin D‐Punjab
268, 268–269 haemoglobin E/β trait 291, 293
haemoglobin C/haemoglobin thalassaemia 286 haemoglobin E/β
E 275, 275 sickle cell anaemia 214 thalassaemia 288
haemoglobin C trait 264, boat‐shaped cells haemoglobin E homozygosity
264–265 sickle cell anaemia 210, 213 283, 285
haemoglobin D‐Punjab/β sickle cell/haemoglobin C haemoglobin E trait 279, 282
thalassaemia 294, 296 disease 224, 224, 225 haemoglobin F
haemoglobin E/β thalassae- body temperature, oxygen quantification 69
mia 287, 287–288 dissociation curve 6, 7 haemoglobin G‐
haemoglobin E/haemoglobin H Bohr effect 5 Philadelphia 296, 298
disease 289, 289 bone infarction 192, 202, 203, 204 neonatal screening 38, 41–43,
haemoglobin E homozygosity bone marrow examination 354
283, 283 β thalassaemia intermedia sickle cell anaemia 214–215,
haemoglobin E trait 279, 280 142, 144 216, 217
haemoglobin H disease 108, β thalassaemia major 148, 149 sickle cell/β thalassaemia 235
108–109, 111, 112 β thalassaemia trait 123, 126 sickle cell/haemoglobin C 228
haemoglobin Lepore trait 138, dominant β thalassaemia sickle cell/haemoglobin E
138 136, 137 239–240
haemoglobin O‐Arab/β haemoglobin C disease 268, 271 sickle cell trait 194, 199
thalassaemia 300, 300 haemoglobin H disease 108–110, split A2 band 65, 65
haemoglobin S‐Oman compound 109–110 capillary isoelectric focusing 63
heterozygotes 246 sickle cell anaemia 217, 219–220 car exhaust fumes 333, 334
high affinity haemoglobins 311 sickle cell/β thalassaemia carbamated haemoglobin 1
HPFH 158, 159 231, 236 carbon dioxide (CO2) transport 1
Index 415
carbon monoxide 4, 333 countries see geography laboratory features 150, 150–152
haemoglobin Zurich affin- crossover during meiosis, Sardinian 133, 153
ity 302, 335 abnormal 18 δβ+ thalassaemia 153
poisoning 334–335, 398, 406 non‐homologous 21, 21–22 δβ0 thalassaemia 149–150, 153
carboxyhaemoglobin 4, 333–335 unequal 85, 86, 90 A
γ 149–150, 157
spectroscopy 335, 336 CSDA (KBX3) gene 15 high affinity haemoglobin
carboxyhaemoglobinaemia cyanosis coinheritance 312–313
333–335 case studies 389, 391, 405 HPFH vs 153, 156–157
case study 398, 406 haemoglobin M carriers 307, 308 sickle cell compound heterozy-
cardiovascular benefits, β investigation of unexplained 359 gosity (S/δβ0) 218, 236
thalassaemia 122 low affinity haemoglobins 313 δ β thalassaemia, sickle cell
0 +
cardiovascular complications methaemoglobinaemia 336, 337 coinheritance 240

β thalassaemia 141, 145 neonatal 309, 309, 313 densitometry, scanning see
sickle cell anaemia 203 unstable haemoglobins 303 scanning densitometry
sickle cell trait 193 deoxyhaemoglobin
CEBPE gene polymorphism 141 dactylitis 202, 203 measurement 335, 336
cellulose acetate electrophore- dapsone 337, 338, 338 nitric oxide binding 1–2
sis 32, 33–34 deletions, gene 18–19, 21 deoxyribonucleic acid see DNA
antenatal screening 351, 352 α thalassaemias 87, 87–88, 88 diabetes mellitus 4
case studies 365, 376, 398, 402 β thalassaemia 117, 119 capillary electrophoresis 38
control samples 32, 33 HPFH 155–156, 157–160, 163 case study 397, 406
globin chains 74, 74 somatic 141, 142 diagnosis of sickle cell trait 192
isoelectric focusing vs 62 thalassaemia 85–86, 86 haemoglobin A1c 44, 45, 333, 334
neonatal screening 354, 355 unstable haemoglobins 301 dichlorophenolindophenol
scanning densitometry 33–34, 34 δ‐aminolaevulinate dehydrase 7, 8 test 72, 72, 360
split A2 band 33 δaminolaevulinate synthase 7, 8 haemoglobin E 281
supplementary alternative δ‐aminolaevulinic acid 7, 8 2,3 diphosphoglycerate
procedures 33 δ globin chain (2,3‐DPG) 5–7, 6
typical mobilities 34, 35, 36 gene 9, 9 low affinity haemoglobins 313
Central African Republic haplo- gene mutations 22, 85 sickle cell anaemia 208
type, βS gene see Bantu synthesis rate 3, 13 DNA analysis 75–78, 76
haplotype, βS gene unstable variants 301 antenatal diagnosis 351–353
cerebral haemorrhage/ variants 13–14, 22, 133 dominant β thalassaemia 134–137,
infarction 202 δ thalassaemia 13–14, 86, 154 136–137
chorionic villus sampling 353 β thalassaemia coinheritance 133 Down’s syndrome 155, 325, 329
chromosome translocations 19 δβ fusion gene 85 drug‐induced abnormalities
citrate agar electrophoresis 35–37, δβ thalassaemia 153 high haemoglobin F 330, 330
37, 38 haemoglobin Lepore trait 137–138 methaemoglobinaemia 339–340
CMU‐E test 281 see also haemoglobin Lepore sulphaemoglobinaemia 338–341
co‐oximetry 335, 336, 405 δβ thalassaemia 86, 149–153 duplications, gene 18, 21–22
co‐translational modification 12 acquired 325, 328 thalassaemias 85–86
cognitive impairment 141, 144 antenatal screening 351, 352 unstable haemoglobins 301
common variable immunodefi- β thalassaemia trait vs 133 dyserythropoiesis, sickle cell
ciency (CVID) 328, 328–329 case studies 368, 387, 399, 404 anaemia 210, 214
congenital dyserythropoietic Corfu 153, 154 dyserythropoietic anaemia,
anaemia 75, 157, 164 DNA analysis 76 congenital 75, 157, 164
cooperativity 5, 6 A
γ 86, 149–153
coproporphyrinogen III 7, 8 genetic defects 153 EABart’s disease 289
Corfu δβ thalassaemia 153, 154 haemoglobin C EFBart’s disease 290
coronary artery disease 122 coinheritance 272–273 eggs, donor 353
416 Index
EKLF see KLF1 F cells 15, 155 γ chain variants 22, 313–315
electrophoresis leukaemia 331 low affinity 313
globin chain 74, 74 quantification 69 M haemoglobins 309, 309
haemoglobin see haemoglobin sickle cell anaemia 221 unstable 301
electrophoresis sickle cell/haemoglobin C γ thalassaemia 86, 155
electrospray ionisation mass disease 226 γδβ thalassaemia see εγδβ
spectrometry 78–79 fat embolism 223, 226 thalassaemia
elliptocytes 123, 124 FCP1 gene 15, 20 gap polymerase chain reaction
elliptocytosis, hereditary 134, ferritin 8 (PCR) 76
370–371, 399 β thalassaemia intermedia 142 GATA1 11
embryo β thalassaemia trait 122, 130 GATA1 gene mutation 20, 120, 121
globin chain synthesis 3, 3–4 haemoglobin E/β gene conversion 17
haemoglobins 2–4, 3, 4 thalassaemia 286 gene deletions see deletions, gene
enhancers haemoglobin H disease 105, 106 gene duplications see duplications,
globin chain synthesis 11 ferrochetalase 7, 8 gene
mutations 17 fetal diagnosis 349, 353 gene fusions see fusion genes
ε globin chain 3, 3 β thalassaemia intermedia 141 gene insertions 19, 20
gene 9, 9 DNA analysis 76 gene inversions 19
εγδβ (γδβ) thalassaemia 86, 153 problems and pitfalls 353 genetics, haemoglobin
acquired 328 fetal haemoglobin see haemoglobin F synthesis 7–12
β thalassaemia trait vs 133 fetus geography
laboratory features 153, 154 globin chain synthesis 3, 3–4 haemoglobin C 188–191, 263, 263
triple α coinheritance 153 haemoglobins 2, 3, 4 haemoglobin E 276–278, 279
ERCC2 (XPD) gene mutation 20, fever 6, 7 haemoglobin S 185–187, 186,
120, 121 FKLF/FKLF2 11 188–191
erythroblastosis fetalis 116, 116 flutamide 341 thalassaemias 92–95
erythroleukaemia 325, 327, 328 folic acid G
γ globin chain 2
erythrophagocytosis, haemoglobin deficiency 131 polymorphisms 219
H disease 103 supplements 96, 122, 208–209 variants 313–315
erythropoietin, serum 142, 208, 312 frameshift mutations 19, 21 γ: γ ratio 15, 159
G A
ethnic groups α thalassaemias 89, 89 globin chain synthesis 8–12, 10

α thalassaemia mutations 91 dominant β thalassaemia 136 abnormalities 16–22
α0 thalassaemia 91, 100 frontal bossing 144, 145, 204 acquired abnormalities 325–347
α+ thalassaemia mutations 91, 95 full blood count (FBC) see blood analysis of rate 74–75, 75, 75
antenatal screening 349, 350 count early life 3, 3–4
δβ thalassaemia 153 fusion genes 18, 21, 21 haem function 7, 12
haemoglobin S and C α thalassaemias 95 thalassaemias 22, 85–86, 86
prevalence 188–191 thalassaemias 85, 86, 90 globin chains
HPFH 157, 162–163 see also δβ fusion gene cooperativity 5, 6
sickle cell disease 186–187 electrophoresis 74, 74
thalassaemia prevalence 90, 92–95 gallstones interactions between 5, 6
exercise, vigorous 192–193 β thalassaemia 121, 141 globin genes 8–11
extramedullary haemopoiesis haemoglobin H disease 105 clusters 8–9, 9
β thalassaemia intermedia sickle cell anaemia 204, 205 DNA sequences influencing
140, 140 γ chain expression 16, 20
β thalassaemia major 144 gene mutations 16, 22, 85 evolutionary origin 9
haemoglobin Bart’s hydrops genes 9, 9 locus control regions see locus
fetalis 114 HPFH 157 control regions, globin gene
haemoglobin E/β synthesis 3, 3 mutations 16–24, 17–19, 23
thalassaemia 286 two species 2 promoters 9, 11
Index 417
regulation of expression 11 normal 12–16 genetic control of synthesis 13

structure 9, 10 post‐translational modifications haemoglobin E/β
transcription 10, 10–11, 12 see post‐translational thalassaemia 288
globin mRNA 10, 11–12 modifications haemoglobin E
mutations 17 structure see haemoglobin homozygosity 283–284
globin pseudogenes 8, 9, 9 structure haemoglobin G‐Philadelphia
duplications 17 variants see variant trait 296
glucose‐6‐phosphate dehydrogenase haemoglobins haemoglobin H disease 111
(G6PD) haemoglobin A 2, 3 haemoglobin Lepore trait 138
β thalassaemia trait 123 β thalassaemia major 148 haemoglobin S interaction 185
haemoglobin H disease 110 capillary electrophoresis 39, 40 HPFH 158, 159
glucose‐6‐phosphate dehydrogenase glycosylation 4 HPLC 47, 65–66, 66
(G6PD) deficiency haemoglobin S interaction 185 increased
β thalassaemia trait oxygen dissociation curve 6 acquired causes 332–333, 333
coinheritance 122 sickle cell/β thalassaemia case study 394, 405–406
β thalassaemia trait vs 123 230, 235 inherited causes 124–127,
methaemoglobinaemia 336 sickle cell trait 194 128
osmotic fragility test 72 synthesis 7 microcolumn chromatography
sickle cell anaemia haemoglobin A0 4 67, 67–68
coinheritance 222 haemoglobin A1c 4 quantification 64–68
glutathionylation, haemoglobin 4 HPLC 44, 45, 47, 73, 73 sickle cell anaemia 215, 218
glycosylated (glycated) incidental diagnosis of sickle cell sickle cell/β thalassaemia 230
haemoglobins 4 trait 192 sickle cell/haemoglobin C
HPLC 44, 45, 46, 47 increased or decreased 333, disease 226
increased or decreased 333, 334 334, 334 sickle cell trait 196
see also haemoglobin A1c quantification 72–73 split A2 band 33, 65, 65
growth retardation haemoglobin A2 12–14 structure 2, 3
β thalassaemia major 144 β thalassaemia major 148, 150 synthesis rates 12–13, 13
haemoglobin H disease 105, 106 β thalassaemia trait 124–129, variants 13–14, 14, 127–128,
gut bacteria 336, 338 127 154, 315
Guthrie spots 30, 354–355 borderline high A2 haemoglobin A2’ 13–14, 14, 315
levels 133–134 β thalassaemia trait 127–128,
haem 1, 2 coinherited conditions 133, 134 129, 314, 315, 315
functions 7, 12 neonatal period 130, 130 case study 398, 406
modifications 1 normal A2 percentages laboratory features 314, 315
synthesis 7, 8 130–133, 132 quantification 64–65, 65
haem‐regulated inhibitor (HRI) 7 with normal red cell haemoglobin A2‐Flatbush 315
haematin 1 indices 134 haemoglobin A2‐Indonesia 315
haemochromatosis, quantification 128, 129 haemoglobin AIII 4
hereditary 122 silent and almost silent haemoglobin Aalborg 313
haemoglobin 1–29 disease 130–131 haemoglobin Abruzzo 310
adult 2, 3, 4 capillary electrophoresis 38, 39, haemoglobin Adana 90
auto‐oxidation 1, 336 40, 66–67 haemoglobin Agrinio 90
carbamated 1 cellulose acetate electrophoresis haemoglobin anti‐Kenya 21
electrophoresis see haemoglobin 33–34, 36 haemoglobin anti‐Lepore 21
electrophoresis decreased haemoglobin Arlington Park 222,
embryonic 2–4, 3, 4 acquired causes 332–333, 333 263, 275
fetal 2, 3, 4 inherited causes 154, 155 haemoglobin
functions 1–2 δ thalassaemia 154, 154 Atlanta‐Coventry 301
genetics of synthesis 7–12 δβ thalassaemia 150, 151 haemoglobin B2 13–14
418 Index
haemoglobin Bart’s 87, 114 haemoglobin H disease haemoglobin C/β

α0 thalassaemia 100, 101, 102 coinheritance 265, 275 thalassaemia 271
α+ thalassaemia 98–99 heterozygosity see haemoglobin haemoglobin Constant
detection 116, 117 C trait Spring 98
haemoglobin Bart’s hydrops homozygosity see haemoglobin C haemoglobin E/β
fetalis 114 disease thalassaemia 287
haemoglobin H disease 111, 113 sickle cell coinheritance see haemoglobin H disease
oxygen affinity 6, 114 sickle cell/haemoglobin C 103–105, 106
haemoglobin Bart’s hydrops disease HPFH 158
fetalis 114–116 haemoglobin C/β sickle cell anaemia 207–208,
antenatal prediction 103, 349, 351 thalassaemia 268–272 209, 218
clinical features 114, 115 case study 372–373, 400 sickle cell/β thalassaemia 230
diagnosis 116 clinical features 270 sickle cell/haemoglobin C
fetal diagnosis 353 diagnosis 272 disease 223
genetic defects 86, 114–115 laboratory features 224, sickle cell/haemoglobin D‐
laboratory features 115–116, 271–272, 272–274 Punjab disease 233
116–117 haemoglobin C disease 267–268 sickle cell/haemoglobin E
haemoglobin Bassett 313 blood films 224, 268, 268–269 disease 239
haemoglobin Beth Israel 313 laboratory features 267–268, sickle cell/haemoglobin O‐Arab
haemoglobin Bethesda 312 268–271 disease 233
haemoglobin Bologna 313 haemoglobin C‐Harlem unstable haemoglobins 303
haemoglobin Boston‐Kuwait 136 cellulose acetate haemoglobin Constant Spring 22,
haemoglobin Bristol 20, 301 electrophoresis 36 90, 95
haemoglobin Bushwick 31, 302 haemoglobin C compound antenatal screening 350
haemoglobin C 263–275 heterozygosity 240–241 diagnosis 38, 99–100
% of total haemoglobin 261, haemoglobin O‐Arab vs 234, haemoglobin H disease 111, 113
262, 265 240, 299, 403 hydrops fetalis 98
α thalassaemia trait second mutations 187, 191 laboratory findings 98, 98
coinheritance 265 sickle cell compound red cell indices 98, 350
β thalassaemia coinheritance see heterozygosity 234–235, 240 haemoglobin Costa Rica 22
haemoglobin C/β haemoglobin C‐Ndjamena 191 haemoglobin Crete 313
thalassaemia haemoglobin C‐New Cross 191 haemoglobin D
case study 394–395, 406 haemoglobin C‐Rothschild 263, 275 haemoglobin A2
coinheritance 263, 268–275, haemoglobin C trait 264–267 quantification 67
272–278 clinical features 264 haemoglobin S vs 194
crystals 268, 269, 270 diagnosis 267 sickle cell/haemoglobin D‐
δβ thalassaemia laboratory features 264–266, Punjab disease 233, 238
coinheritance 272–273 264–267 haemoglobin D‐Baltimore see
DNA analysis 76 haemoglobin C‐Ziguinchor 191 haemoglobin
electrophoresis 34, 36, 40, haemoglobin Chesapeake 310 G‐Philadelphia
264–266, 265 haemoglobin Chile 309 haemoglobin D‐Chicago see
geographic distribution 188–191, haemoglobin Cleveland 290 haemoglobin D‐Punjab
263, 263 haemoglobin concentration (Hb) haemoglobin D‐Conley see
haemoglobin A2 quantification α0 thalassaemia 100 haemoglobin D‐Punjab
65, 66 α+ thalassaemia 96, 98 haemoglobin D‐Cyprus see
haemoglobin C‐Harlem β thalassaemia intermedia 141 haemoglobin D‐Punjab
compound β thalassaemia major 146 haemoglobin D disease see
heterozygosity 240–241 β thalassaemia trait 122 haemoglobin D‐Punjab
haemoglobin D‐Punjab haemoglobin Bart’s hydrops disease
coinheritance 187 fetalis 115–116 haemoglobin D‐Ibadan
Index 419
haemoglobin D‐Punjab sickle cell crisis 209 α+ thalassaemia coinheritance

vs 290–291 sickle cell/haemoglobin C 103, 283
sickle cell compound disease 224 clinical features 282
heterozygotes 233 haemoglobin E 275–290 diagnosis 284–285
haemoglobin D‐Iran % of total haemoglobin 279–280, laboratory features 283–284,
diagnostic relevance 36, 261 287, 288 283–285
haemoglobin D‐Punjab α0 thalassaemia trait haemoglobin E‐Saskatoon 36, 285
vs 290–291 coinheritance 102–103, 283, case study 392, 405
haemoglobin E trait vs 281, 282 284, 290 haemoglobin E trait 278–281
sickle cell compound β chain synthesis rate 261, 278 antenatal screening 350
heterozygotes 233 β thalassaemia coinheritance case study 367, 399
haemoglobin D‐Los Angeles see see haemoglobin E/β diagnosis 281, 282, 283
haemoglobin D‐Punjab thalassaemia laboratory features 279–281,
haemoglobin D‐North Carolina see case study 394–395, 406 280–282
haemoglobin D‐Punjab coinheritance 285–290 haemoglobin Egypt see haemoglobin
haemoglobin D‐Portugal see dichlorophenolindophenol O‐Arab
haemoglobin D‐Punjab test 72, 72 haemoglobin
haemoglobin D‐Punjab 290–295 disease see haemoglobin E electrophoresis 31–38
β thalassaemia homozygosity α0 thalassaemia 100, 101
coinheritance 293–294, 296 DNA analysis 76 α+ thalassaemia 99
coinheritance 293–295 electrophoresis 34, 36, 239, β thalassaemia trait 127
detection 36, 290–291 279–280, 281 δβ thalassaemia 151
DNA analysis 75–76, 76 geography 276–278, 279 haemoglobin Bart’s hydrops
haemoglobin C haemoglobin A2 quantification fetalis 116, 117
coinheritance 187 66, 67 haemoglobin C disease 268, 270
haemoglobin E haemoglobin C coinheritance haemoglobin C/haemoglobin E
coinheritance 290 274–275, 275–277 277
haemoglobin G‐Philadelphia haemoglobin H disease haemoglobin C/haemoglobin
coinheritance 297 coinheritance 280, 289, G‐Philadelphia 275, 277
haemoglobin G‐Philadelphia 289–290 haemoglobin C trait 264–265, 265
vs 291, 296 haemoglobin Tak haemoglobin D‐Punjab
haemoglobin S coinheritance coinheritance 313 290–291, 293
233, 237–238 heterozygosity see haemoglobin haemoglobin E/β
case study 374–375, 400 E trait thalassaemia 287–288
HPLC 65, 291, 292 resource‐poor settings 360 haemoglobin E homozygosity
isoelectric focusing 61, 61 sickle cell compound 283, 284
haemoglobin D‐Punjab dis- heterozygosity 238–240, haemoglobin E trait 279–280, 281
ease 291–293, 294, 295 242–244 haemoglobin H disease 110, 110
haemoglobin D‐Punjab trait 291, thalassaemic features 22, 117, 278 haemoglobin Lepore trait 138,
292, 293 haemoglobin E/β 139
case study 365, 399 thalassaemia 285–289 haemoglobin M 309–310
haemoglobin D‐St Louis see α thalassaemia coinheritance 287 haemoglobin O‐Arab trait
haemoglobin case study 367, 399 298, 299
G‐Philadelphia clinical features 285–286 high affinity
haemoglobin D‐Washington see laboratory features 279–280, haemoglobins 311–312
haemoglobin 287–288, 287–289 sickle cell anaemia 214–215,
G‐Philadelphia haemoglobin E homozygosity 215, 221
haemoglobin Dhohar 117 (disease) 282–285 sickle cell/β thalassaemia 233
haemoglobin distribution width α0 thalassaemia coinheritance sickle cell/haemoglobin C 227,
(HDW) 102–103, 283, 284 228–229
420 Index
haemoglobin electrophoresis (cont’d) laboratory techniques 68–69 sickle cell anaemia

sickle cell/haemoglobin neonates 329 coinheritance 221
D‐Punjab 233 post‐translational sickle cell/haemoglobin C
sickle cell/haemoglobin E 239, modification 4 disease coinheritance 229
243 quantification 68–69, 130 sickle cell trait coinheritance
sickle cell/haemoglobin sickle cell anaemia 163, 200–201, 201, 383, 403
Lepore 236 217–220, 218 trait (heterozygosity) 296,
sickle cell/haemoglobin sickle cell/β thalassaemia 230 297, 298
O‐Arab 233, 239 sickle cell/haemoglobin C haemoglobin G2 296
sickle cell/HPFH 237 disease 226, 227 haemoglobin Gower 1: 3, 3
sickle cell trait 194, 196 sickle cell/haemoglobin D‐ haemoglobin Gower 2: 3, 3, 4
unstable haemoglobins 304, 306 Punjab disease 233, 238 haemoglobin Gun Hill 301
haemoglobin F 14–16 structure 2, 3 haemoglobin H 87, 103
age‐related changes 14, variants see γ chain variants acquired haemoglobin H
14–15, 217 haemoglobin F‐Bonheiden 301 disease 327, 328
β thalassaemia intermedia 142 haemoglobin F‐Cincinnati 313 detection and quantification 43,
β thalassaemia major 148, haemoglobin F‐Circleville 309, 309 110, 110–111
150, 164 haemoglobin F‐Heuried 313 haemoglobin Bart’s hydrops
β thalassaemia trait 128, haemoglobin F‐M‐Fort fetalis 115
129–130, 163 Ripley 301, 309, 309 haemoglobin E/haemoglobin H
capillary electrophoresis 38, 40 haemoglobin F‐M‐Osaka 301, disease 289
carbon monoxide affinity 335 309, 309 inclusions
cells containing see F cells haemoglobin F‐M‐Viseu 309 acquired haemoglobin H
cellulose acetate haemoglobin F‐Poole 301 disease 327, 327
electrophoresis 33 haemoglobin F quantitative trait α0 thalassaemia 102, 102
decreased 155, 329 loci (HBFQTL) 15 detection 70, 70, 71
δβ thalassaemia 150, 151, 152 haemoglobin F‐Sardinia 2 haemoglobin H disease 108,
distribution 69, 157, 157 haemoglobin Fairfax 301 111, 111–112
Down’s syndrome 155, 329 haemoglobin G 194 oxygen dissociation curve 6,
genetic control of adult haemoglobin G‐Azakouli see 103, 111–113
levels 15–16 haemoglobin G‐Philadelphia red cell lysates 31
γ: γ ratio 15
G A
haemoglobin G‐Bristol see haemoglobin H/Constant Spring
haemoglobin E/β haemoglobin G‐Philadelphia disease 105, 111, 113
thalassaemia 287–288 haemoglobin G‐Coushatta 36, haemoglobin H disease 96, 103–114
haemoglobin E 281, 283, 290–291 acquired 105, 325–328, 326, 327
homozygosity 284 haemoglobin case study 379, 402
haemoglobin Lepore trait G‐Philadelphia 295–297 antenatal screening 349–350
138–139, 163 α thalassaemia β thalassaemia trait coinher-
hereditary persistence see coinheritance 262, 297 itance 113, 133, 134
hereditary persistence of coinheritance 297 case study 396, 406
fetal haemoglobin diagnostic relevance 36, 261 clinical features 105–106
increased 16 haemoglobin A2 quantification coinheritance 113–114
acquired causes 329–330, 330, 65, 127 genetic abnormalities 103,
331–332 haemoglobin C coinheritance 104–105, 106
case studies 384–385, 387, 404 275, 277 haemoglobin C trait
haemoglobin E trait 280 haemoglobin D‐Punjab coinheritance 265, 275
inherited causes 155–164, 156 coinheritance 297 haemoglobin E coinheritance
sickle cell anaemia see sickle haemoglobin D‐Punjab vs 291, 280, 289, 289–290
cell anaemia, with high 296 haemoglobin Paksé
haemoglobin F homozygosity 297 coinheritance 111, 113
Index 421
HPFH coinheritance 160, 161 compound heterozygosity 139 haemoglobin Memphis 221
laboratory features 106–113 as δβ+ thalassaemia 153 haemoglobin Monroe/
sickle cell trait coinheritance see DNA analysis 76 haemoglobin S compound
sickle cell trait/haemoglobin electrophoresis 33–34, 36, 139, heterozygosity 243
H disease 139, 236 haemoglobin
haemoglobin Haaglanden 187 genetic defect 21, 85, 137–138 Montgomery 221–222
haemoglobin Hammersmith haemoglobin C haemoglobin N‐Baltimore/
302, 304 coinheritance 272–273 haemoglobin C 274
haemoglobin Hasharon 36, 301, 303 haemoglobin E haemoglobin New York
haemoglobin Headington 311, coinheritance 290 haemoglobin H disease
312–313 heterozygosity see haemoglobin coinheritance 114
haemoglobin Heathrow 74, 312 Lepore trait haemoglobin S compound
haemoglobin Hofu/haemoglobin S homozygosity 139 heterozygosity 243
compound HPLC 139, 139, 236, 241 haemoglobin North Shore 117, 243
heterozygosity 242 sickle cell compound haemoglobin O‐Arab 297–300
haemoglobin Hope heterozygosity 235–236, 241 β thalassaemia coinher-
case study 391, 405 haemoglobin Lepore itance 299–300, 300
haemoglobin S compound Baltimore 138–139 coinheritance 299–300
heterozygosity 241–242, 243 haemoglobin Lepore Boston/ DNA analysis 76
haemoglobin Hopkins II 221 Washington 138–139, 273 electrophoresis 36, 298, 299
haemoglobin I 187 haemoglobin Lepore haemoglobin C‐Harlem vs 234,
haemoglobin I‐Toulouse 243 Hollandia 138 240, 299, 403
haemoglobin Icaria 90, 98 haemoglobin Lepore Leiden 138 haemoglobin C vs 265–266
haemoglobin J‐Baltimore 261–262 haemoglobin Lepore trait 137–139 sickle cell coinheritance see sickle
haemoglobin J‐Biakra 301 case study 369, 399 cell/haemoglobin O‐Arab
haemoglobin J‐Iran 261–262 high haemoglobin F 138–139, 163 disease
haemoglobin J‐Sardegna 20 laboratory features 138–139, haemoglobin O‐Arab disease 299
haemoglobin Jackson 310 138–139 haemoglobin O‐Arab trait
haemoglobin Jamaica Plain 187, haemoglobin Longview 310 297–299, 299, 300
191, 243 haemoglobin Lufkin 243 haemoglobin Oak Ridge see
haemoglobin K‐Woolwich/ haemoglobin Luton 312 haemoglobin D‐Punjab
haemoglobin C 274 haemoglobin Lyon‐Bron 313 haemoglobin Olmsted 304
haemoglobin Kansas 312, 313 haemoglobin M 307–310, 309 haemoglobin Olympia 311
haemoglobin Kempsey 390, 405 clinical features 307, 308, 309 haemoglobin P‐Galveston/
haemoglobin Kenya laboratory features 309–310, 310 haemoglobin C 274
deletional HPFH 159, 160 molecular abnormality haemoglobin Paksé 90, 95
genetic defect 21 307–308, 308 diagnosis 99–100
haemoglobin S compound see also methaemoglobin haemoglobin H disease 111, 113
heterozygosity 242 haemoglobin M‐Boston 307, 308, haemoglobin Palencia 138
haemoglobin Knossus 133 309, 310 haemoglobin Parchman 138
haemoglobin Köln 301, 302, haemoglobin M‐Hyde Park 308, haemoglobin Poissy 311
305, 306 309, 309, 310 haemoglobin Portland
haemoglobin Korle‐Bu 22 haemoglobin M‐Iwate 262, 307, (or Portland 1) 3, 3
electrophoresis 36, 291 308, 309, 310 haemoglobin Bart’s hydrops
haemoglobin C haemoglobin M‐Milwaukee 308, fetalis 114
coinheritance 274 309 hereditary persistence 157
sickle cell anaemia haemoglobin M‐Saskatoon 308, haemoglobin Presbyterian 313
coinheritance 221 308, 309, 309, 310 haemoglobin Q‐India 66, 66
haemoglobin Koya Dora 90 haemoglobin Malay 76, 117 haemoglobin Quebec‐Chori/
haemoglobin Lepore 86, 138 haemoglobin McKees Rocks 312 haemoglobin S 241
422 Index
haemoglobin Quong Sze 90, 350 haemoglobin Seal Rock 90 β thalassaemia major 144
haemoglobin Raleigh 20 haemoglobin Setif 187 sickle cell anaemia 202, 204
haemoglobin Riyadh/haemoglobin haemoglobin Shelby 243 sickle cell/β thalassaemia 229
C 274 haemoglobin Siriraj sickle cell/haemoglobin C
haemoglobin Rush 304 blood film 305 disease 222
haemoglobin S 185–244 haemoglobin S compound unstable haemoglobins 301
% of total haemoglobin 261, 262 heterozygosity 241–242, haemolytic anaemia
countries and ethnic 245–246 α+ thalassaemia 96
groups 188–191 haemoglobin St Mande 313 haemoglobin A1c 333
detection 63–64, 194–196 haemoglobin St Mary’s 305, 307 haemoglobin H disease 103, 105
DNA analysis 76 haemoglobin Stanleyville I see investigations 359
electrophoresis 33, 34, 36, 40 haemoglobin unstable haemoglobins 302, 303
gene see sickle cell gene G‐Philadelphia haemopexin, serum 149
glycated, HPLC 44, 46 haemoglobin Stanleyville II 221 haemophagocytosis 217, 228
haemoglobin A2 quantification haemoglobin structure 1–7, 2 hair‐on‐end appearance 144,
64–65, 66, 67, 68 acquired abnormalities 325–347 145, 204
haemoglobin S/haemoglobin primary 5 hand‐foot syndrome 202, 203
D‐Punjab compound quarternary 2, 5 haptoglobin, serum 149, 216
heterozygosity, case secondary 5 HBA1 gene see α1 gene
study 382, 403 tertiary 2, 5, 5 HBA2 gene see α2 gene
heterozygosity see sickle cell trait see also post‐translational HBB gene see β globin gene
homozygosity see sickle cell modifications HBBP1 pseudogene 15
anaemia haemoglobin HBD (δ) gene 9, 9
immunological detection 69–70 Sydney‐Coventry 301 HBE1 (ε) gene 9, 9
point‐of‐care testing 360–361 haemoglobin Syracuse 312 HBFQTL2 see HBSIL‐MYB intergenic
polymerisation of deoxygenated haemoglobin Tacoma 377, 402 polymorphism block 2
185, 186 haemoglobin Tak 290, 312–313 HBFQTL3 15
quantification 68, 196 haemoglobin Tarrant 310 HBFQTL4 15
sickle cell/α thalassaemia haemoglobin Taybe 302 HBFQTL5 15
trait 194 haemoglobin Thionville 313 HBFQTL6 15
sickle cell anaemia 214–215, 215 haemoglobin Titusville 313 HBG1(Aγ) gene 9, 9
sickle cell/β thalassaemia 230 haemoglobin Tyne/haemoglobin HBG2 (Gγ) gene 9, 9
sickle cell trait 194, 194, 196, 196 S 242 polymorphism 15, 219
haemoglobin S‐Antilles 187, 191 haemoglobin Vicksburg 117 HBS1L gene 13
compound heterozygotes 241 haemoglobin Volga/haemoglobin HBSIL‐MYB intergenic
heterozygotes 243 S 241 polymorphism block 2
haemoglobin S‐Cameroon 191 haemoglobin Westmead/α0 (HMIP‐2) 15, 16, 20, 221
haemoglobin S‐Clichy 191 thalassaemia 103 HBZ gene 9
haemoglobin S‐Oman 187, 191 haemoglobin Zunyi 119 headaches 145
compound heterozygotes haemoglobin Zurich 301, 302, 303 heat stability test 70–71, 304
241, 246 carbon monoxide affinity heel prick blood samples 30, 354,
heterozygotes 243 302, 335 354–355
haemoglobin S‐Providence 191 electrophoresis 36 Heinz bodies 302, 304
haemoglobin S‐San Martin 191 haemoglobinopathies detection tests 71, 72, 304
haemoglobin S‐São Paulo 191 diagnostic service see service, haemoglobin H disease 111, 112
haemoglobin S‐South End 191 haemoglobinopathy sickle solubility test 64
haemoglobin S‐Travis 191 diagnostic hemi‐ghosts
haemoglobin S‐Wake 191 other significant 261–324 sickle cell anaemia 209, 210, 213
haemoglobin San Diego 243, 313 terminology 22 sickle cell/haemoglobin C
haemoglobin Santa Ana 304 haemolysis disease 225, 226
Index 423
HemoCards 69–70 α0 thalassaemia 100, 101, 102 sickle cell/haemoglobin

HemoTypeSC test 361 α+ thalassaemia 99 O‐Arab 233–234, 240
hepatic steatosis 142 antenatal screening 351 sickle cell trait 194, 197–198
hepatitis C 122 β thalassaemia major 148, 150 unstable haemoglobins 304, 306
hereditary elliptocytosis 134, β thalassaemia trait 127, 129 high resolution melting analysis
370–371, 399 δβ thalassaemia 151, 152 (HRMA) 78
hereditary persistence of fetal elution patterns 47, 47–59 Hill’s equation 312
haemoglobin haemoglobin A1c quantification histiocytes, sea‐blue 217, 219
(HPFH) 155–161 73, 73 histone acetylases 11
β thalassaemia coinheritance haemoglobin A2 quantification histone deacetylase 1 (HDAC1) 11
134, 135, 160 65–66, 66 HMIP‐2 see HBSIL‐MYB intergenic
case study 394, 406 haemoglobin Bart’s hydrops polymorphism block 2
coinheritance 160, 161 fetalis 116, 117 Howell–Jolly bodies 210, 214,
deletional 155–156, 157–160 haemoglobin C/β 224, 338
blood films 158, 159 thalassaemia 273 HPFH see hereditary persistence of
case study 384, 404 haemoglobin C disease 268, 270 fetal haemoglobin
coinheritance 160 haemoglobin C/ HPLC see high performance liquid
gene defects 157, 158 haemoglobin E 276 chromatography
haematological features haemoglobin C trait 264–265, human immunodeficiency virus
158, 159 266, 266 (HIV) infection 330, 332
δβ thalassaemia vs 153, 156–157 haemoglobin D‐Punjab hydrops fetalis
δβ0 type 157–158, 159 disease 294 haemoglobin Bart’s see
DNA analysis 76 haemoglobin D‐Punjab haemoglobin Bart’s
A
γ 160–161, 162–163 trait 291, 292 hydrops fetalis
G
γ 160–161, 162, 219 haemoglobin E/β haemoglobin Constant
γ γ 163
G A
thalassaemia 287–288 Spring 98
haemoglobin E coinheritance 290 haemoglobin E homozygosity haemoglobin H disease 106
haemoglobin H coinheritance 283, 284 hydroxycarbamide
160, 161 haemoglobin E trait 279, 281 β thalassaemia 142
non‐deletional 156, 160–161, haemoglobin F haemoglobin E/β thalassaemia
162–163 quantification 68–69 286, 287, 289
sickle cell compound haemoglobin G‐ sickle cell anaemia 202,
heterozygosity 160, 218, Philadelphia 296, 297 209–210, 217, 221
219–221, 236–237, 242 haemoglobin H disease 110, sickle cell/haemoglobin C
hereditary spherocytosis 192, 110–111 disease 224
201, 229 haemoglobin Lepore trait 139, 139 unstable haemoglobins 303
Herrick, James 185 haemoglobin O‐Arab trait hydroxyurea see
high affinity 298, 300 hydroxycarbamide
haemoglobins 310–313 interpretation 44–46 hyperhaemolysis, sickle cell
case study 390, 405 neonatal screening 354, 355 anaemia 214
clinical features 311 retention times 39–40, 47, 60 hypersplenism
coinheritance 312–313 sickle cell anaemia 214–215, 218 β thalassaemia 141, 144, 148
diagnosis 312 sickle cell/β thalassaemia 234 sickle cell anaemia 202, 212
laboratory features 73–74, 74, sickle cell/haemoglobin C 227 hyphaemia, post‐traumatic 193
311–312, 312 sickle cell/haemoglobin hyposplenism
unstable 303, 310 D‐Punjab 233, 238 case study 380, 402–403
high altitude 192, 205 sickle cell/haemoglobin E 239, sickle cell anaemia 210, 214
high performance liquid 244 sickle cell/β thalassaemia 230
chromatography sickle cell/haemoglobin sickle cell/haemoglobin C
(HPLC) 38–47 Lepore 236, 241 disease 222–223, 225
424 Index
hypothermia 6 haemoglobin C disease 268, liver disease

hypoxia 268–269 β thalassaemia 122, 140, 140,
carboxyhaemoglobinaemia 335 haemoglobin C trait 264, 264–265 145–146
sickle cell formation 192, 357 sickle cell anaemia 209, 210, 213 haemoglobin H disease 106
sickle cell/haemoglobin C sickle cell anaemia 204
immunoassays 69–70 disease 224, 225, 226 locus control regions, globin
inclusion body β unstable haemoglobins 305 gene 11
thalassaemia 136 isoelectric focusing (IEF) 57–63, 61 deletions 19, 87
industrial fumes 334 applications 61–62, 62 DNA sequences competing
ineffective erythropoiesis capillary 63 with 19
β thalassaemia major 144 densitometric scanning 61, 61 see also α‐locus control region;
dominant β neonatal screening 354, 355 β‐locus control region
thalassaemia 134–136 isoelectric point (pI) 61 low affinity haemoglobins 313
haemoglobin H disease 103, 105 isopropanol test 71, 71, 304 oxygen dissociation
infants curve 73–74, 312
globin chain synthesis 3 jaundice 204, 205, 303 LUC7L gene 90
haemoglobin F 14, 14–15 juvenile myelomonocytic leukaemia lymphocyte count, sickle cell
haemoglobins 4 (JMML) anaemia 208
methaemoglobinaemia 336 acquired thalassaemia 328, 328
sickle cell anaemia 202, 210, 212 case study 385, 404 macrophages, sickle cell
sickle solubility test 64 increased haemoglobin F anaemia 217
see also neonates 329–330, 331–332 malaria
initiation codon 12 haemoglobin C 263
mutations 17, 89, 89, 119 KBX3 gene 15 haemoglobin E 278
insertions, gene 19, 20 kidney, medullary carcinoma 193 sickle cell disease 187, 192, 194
inversions, gene 19 Kleihauer test 69, 69, 331 thalassaemia 96, 122
iron KLF1 11 MCS‐R2 enhancer, deletion 88, 88
haem molecule 1 KLF1 gene mean cell haemoglobin (MCH)
haem synthesis 7 control of haemoglobin F 15, 16 α0 thalassaemia 100
uptake by erythroid cells 7, 8 haploinsufficiency 157 α+ thalassaemia 96, 96, 98
iron deficiency mutations 11, 13, 20, 130, 157 β thalassaemia major 146
α0 thalassaemia vs 100–102 β thalassaemia trait 122, 123,
β thalassaemia trait vs 123, laboratory techniques 30–84 128, 131, 132
130, 131 more specialised tests 73–79 haemoglobin Constant
case studies 372, 394–395, quality assurance 79 Spring 98
399–400, 406 resource‐poor situations 72 haemoglobin E
globin chain synthesis 12 sample collection 30 homozygosity 283
haem synthesis 7 lactate dehydrogenase (LDH) 216, haemoglobin H disease 106
haemoglobin A1c 333 228, 284 HPFH 158
low haemoglobin A2 332 LCRA see α‐locus control region sickle cell anaemia 208
osmotic fragility test 72 LCRB see β‐locus control region sickle cell/β thalassaemia 230
sickle cell anaemia 208, 216 leg ulceration 204 sickle cell/haemoglobin C
iron overload leptocytes 148 disease 224
β thalassaemia intermedia 140 leukaemia sickle cell trait 194
β thalassaemia major 145 acquired thalassaemias 325–326, unstable haemoglobins 303
β thalassaemia trait 122 327, 328, 328 mean cell haemoglobin concentra-
haemoglobin E/β case studies 379, 385, 402, 404 tion (MCHC)
thalassaemia 286 increased haemoglobin F β thalassaemia trait 122
haemoglobin H disease 106 329–330, 331–332 haemoglobin C disease 267–268
irregularly contracted red cells lipid profile, β thalassaemia trait 122 haemoglobin C trait 264
Index 425
haemoglobin Constant clinical features 336, 337–338 unstable haemoglobins 301, 303
Spring 98 laboratory features 336, 338–339 see also infants
haemoglobin H disease 106 methionine aminopeptidase 12 neural tube defects 96, 122
sickle cell anaemia 208 methylene blue 337 neurological complications
sickle cell/β thalassaemia 230 microcolumn chromatography 67, sickle cell anaemia 202
sickle cell/haemoglobin C 67–68 sickle cell/haemoglobin C
disease 224 mis‐sense mutations 16, 17 disease 223
unstable haemoglobins 303 mosaic isodisomy 243 neutrophil count, sickle cell
mean cell volume (MCV) MPL‐Baltimore 208 anaemia 208, 209
α0 thalassaemia 100 multiple ligation‐dependent probe new‐sense mutations 17
α+ thalassaemia 96, 96, 98 amplification (MPLA) 78 NFE4p22 enhancer 11
β thalassaemia major 146 MYB gene 13, 15, 221 nitrate/nitrite exposure 339, 405
β thalassaemia trait 122, 123, myelodysplastic/myeloproliferative nitric oxide (NO) 1–2, 202
128, 131, 132 neoplasms, atypical 325 S‐nitrosohaemoglobin 2
haemoglobin C disease 267 case study 379, 402 nonsense mutations 17, 20
haemoglobin C trait 264 myelodysplastic syndromes nucleated red blood cells (NRBC)
haemoglobin Constant (MDS) 325, 326, 328, 328 case study 394, 406
Spring 98 myelolipoma 146 sickle cell anaemia 210, 211,
haemoglobin E myoglobin 6 212, 213
homozygosity 283
haemoglobin H disease 106 NADH‐cytochrome organisation, haemoglobinopathy
HPFH 158 b5‐methaemoglobin service see service, haemo-
sickle cell anaemia 208, 218 reductase 1 globinopathy diagnostic
sickle cell/β thalassaemia 230 deficiency 337 osmotic fragility test 72, 360
sickle cell/haemoglobin C Napoleon hat sickle cells 243, 246 β thalassaemia trait 72, 128
disease 223–224 neonatal screening 354–357 haemoglobin E trait 281
sickle cell/haemoglobin D‐ capillary electrophoresis 38, osteomyelitis 203
Punjab disease 233 41–43, 354 osteonecrosis
sickle cell/haemoglobin E haemoglobin H disease 113 sickle cell anaemia 203, 204
disease 239 methods 354, 354–355 sickle cell/haemoglobin C
sickle cell/haemoglobin O‐Arab problems and pitfalls 357 disease 222, 223
disease 233 protocol 356 oxygen, transport 1
sickle cell trait 194, 194 universal vs targeted 355 oxygen affinity
unstable haemoglobins 303 neonates embryonic haemoglobins 4
megaloblastic anaemia α0 thalassaemia 100, 102 haemoglobin Bart’s 6, 114
β thalassaemia trait vs 131–133 α+ thalassaemia 98–99 haemoglobin H 103
sickle cell anaemia 204–205, 212 β thalassaemia trait 123–124, unstable haemoglobins 302,
sickle cell/β thalassaemia 230 130, 130, 356–357 303, 304–306
methaemoglobin 1, 4, 336–337 cyanosis 309, 309, 313 oxygen dissociation curve 5–7, 6
haemoglobin E/β εγδβ thalassaemia 153, 154 carboxyhaemoglobinaemia 335
thalassaemia 286 haemoglobin Bart’s hydrops haemoglobin E homozygosity 284
haemoglobin M carriers 308, 309 fetalis 114, 115 haemoglobin H 103
laboratory features 309–310, 310 haemoglobin F 329 high and low affinity
unstable haemoglobins 303, 304 haemoglobin H disease 111, 113 haemoglobins 73–74, 74,
see also haemoglobin M sample collection 30, 354, 312, 312
methaemoglobin reductase 1 354–355 methaemoglobinaemia 310, 336
deficiency 337 sickle cell anaemia 202, 210, sickle cell anaemia 216
methaemoglobinaemia 336–337 211, 215–216, 218 sickle cell/haemoglobin C
case studies 389, 391, 405 sickle cell/β thalassaemia 231 disease 228
causes 336, 339–340 sickle cell trait 194–197, 198 unstable haemoglobins 304–306
426 Index
oxyhaemoglobin unstable haemoglobins 301 red cell aplasia, parvovirus‐

measurement 335, 336 variant haemoglobins 20–21 induced 204, 210, 214
nitric oxide interactions 1, 2 post‐zygotic mitotic recombina- red cell distribution width (RDW)
tion 243, 244 β thalassaemia trait 122
P50o2 310, 313 PPARGC1A 15 sickle cell anaemia 208, 209
Pappenheimer bodies 224 pre‐anaesthetic testing 357–359, 358 sickle cell/β thalassaemia 230
acquired haemoglobin H preconceptual screening 349 sickle cell/haemoglobin C
disease 326 pregnancy disease 224
β thalassaemia major 148 α+ thalassaemia 96 red cell indices
haemoglobin E/β antenatal testing 349–353 α0 thalassaemia 100, 101
thalassaemia 288 β thalassaemia trait 131 α+ thalassaemia 96, 96, 97, 98
sickle cell/β thalassaemia 230 carbon monoxide exposure 335 β thalassaemia trait 122, 123, 128
paraproteins haemoglobin Bart’s hydrops δβ thalassaemia 150–153
haemoglobin electrophoresis fetalis 115 haemoglobin Constant
32, 32 haemoglobin F 329 Spring 98, 350
sickle solubility test 63, 64 termination 349, 353 haemoglobin E trait 279
parvovirus B19 infection preimplantation diagnosis 353 haemoglobin H disease 106, 107
sickle cell anaemia 204, 210, premature infants 329, 355 HPFH 158
211–212 prenatal diagnosis see fetal diagnosis sickle cell anaemia 208
sickle cell/haemoglobin E 239 priapism 141, 192, 223 sickle cell/β thalassaemia 230
Pauling, Linus 2, 185 promoters, globin gene 9, 11 sickle cell/haemoglobin C
pH, oxygen dissociation curve 5, 6 protocols, laboratory 348 disease 223–224
phenacetin 338 protoporphyrin IX 7, 8 sickle cell trait 194, 194
platelet count, sickle cell anae- pseudo‐Gaucher cells 217, 220 see also blood count
mia 208, 209 pseudocyanosis 307, 308, 336, 337 red cell lysates 31
poikilocytes, SC 224, 224–225, 225 pseudogenes, globin see globin red cell protoporphyrin
point mutations 16–21, 17 pseudogenes α+ thalassaemia 98
α thalassaemia 89 ψα1 pseudogene 9, 9 β thalassaemia trait 130
HPFH 160, 161, 162–163 ψα2 pseudogene 9, 9 haemoglobin E trait 281
unstable haemoglobins 301 ψβ pseudogene 9, 9 red cell survival
point‐of‐care testing, haemoglobin ψζ pseudogene 9, 9 haemoglobin H disease 103
S 360–361 pulmonary hypertension 141, sickle cell anaemia 202, 216–217
polyadenylation, RNA 12 144, 203 sickle cell/β thalassaemia 231
α thalassaemia mutations 89, 89 pulmonary infarction 203, 204 sickle cell/haemoglobin C
β thalassaemia mutations 119 pulse CO‐oximetry see co‐oximetry disease 222
polychromasia 210, 224, 225, 230 pulse oximetry unstable haemoglobins 306
polycythaemia carboxyhaemoglobinaemia 335 refractory anaemia with excess of
case study 390, 405 methaemoglobinaemia 310, 336 blasts (RAEB) 325, 328
high affinity haemoglobins 310, unstable haemoglobins 306 refractory anaemia with ring
311, 312, 313 pyridoxal 5’‐phosphate 7, 8 sideroblasts (RARS) 325,
investigation of unexplained 359 pyrimidine 5’ nucleotidase 326, 328
unstable haemoglobins 303 deficiency 290 renal disease
polycythaemia vera 123, 130 pyruvate kinase deficiency 192, β thalassaemia 141, 146
polymerase chain reaction 201, 244 sickle cell disease 193, 205
(PCR) 76, 76–78, 78 resource‐poor settings 72, 360–361
porphobilinogen 7, 8 quadruple α (αααα) 18, 86 restriction enzyme polymerase
post‐translational modifications 4 β thalassaemia trait chain reaction (PCR) 78
capillary electrophoresis 38 coinheritance 134, 142 restriction fragment length
HPLC interpretation 44, 45, quality assurance 79 polymorphism (RFLP)
46, 47 quintuple α (ααααα) 18, 86 analysis 77, 185
Index 427
reticulocyte count 31, 31 sickle cell/α thalassaemia trait sickle cell disease 185–260
β thalassaemia major 148 clinical features 207, 207 βS heterozygotes see sickle cell trait
β thalassaemia trait 122 haemoglobin F 215, 221 βS homozygotes see sickle cell
sickle cell anaemia 208, 218 laboratory features 210 anaemia
sickle cell/β thalassaemia 230 sickle cell anaemia 201–222 causes 192
sickle cell/haemoglobin C aggravating factors 207 compound heterozygotes 187,
disease 224 ameliorating factors 206–207, 207 192, 233–243
sickle cell/haemoglobin O‐Arab blood count 207–210, 209 detection tests 63–64
disease 233 blood films 210–214, 211–215, 224 geographic origins and
unstable haemoglobins 303 bone marrow examination 217, spread 185–187, 186
reticulocyte indices, β thalassaemia 219–220 heterozygotes with 192,
trait 122 capillary electrophoresis 214–215, 243–244
retinal disease 193, 203, 222, 229 216, 217 neonatal screening 354, 355
reverse dot blot analysis 78, 79 case study 380, 402–403 point‐of‐care testing 360–361
rhabdomyolysis, exertional 192–193 clinical features 202–206, 203, pre‐anaesthetic testing 357–359,
RNA splice site mutations 88–89, 204, 204, 205 358
89, 119 coinheritance 221–222 terminology 187
diagnosis 221 uniparental disomy 244
S‐nitrosylation 2 haemoglobin electrophoresis sickle cell gene (βS)
same‐sense (neutral) mutations 214–215, 215 geographic origins and
16, 17 haemoglobin F 163, 217–221 spread 185–187, 186
sample collection 30 with high haemoglobin F 207, 218 heterozygosity (ββS) see sickle cell
SAR1A gene 15 blood films 210, 215 trait
Sardinian δβ thalassaemia 133, 153 clinical features 205–206, 207 homozygosity (βSβS) see sickle cell
Saudi Arabian haplotype, βS gene HPLC 214–215, 218 anaemia
see Arab‐Indian haplotype, laboratory features 207–217 second mutations 187, 191, 192
βS gene neonatal screening 355–356 sickle cell haemoglobin see
scanning densitometry terminology 187 haemoglobin S
cellulose acetate electrophoresis sickle cell/β thalassaemia 187, sickle cell/haemoglobin C
33–34, 34 229–232 disease 222–229
haemoglobin A2 64, 128, 129 case studies 388–389, 404 clinical features 222–223, 223
haemoglobin F 68 clinical features 229 coinheritance 224, 229
isoelectric focusing 61, 61 coexisting α thalassaemia 230 diagnosis 228–229
Senegal haplotype, βS gene 185, 186 diagnosis 68, 231–232 HPLC 56–58
haemoglobin F levels 218, laboratory features 230–231, laboratory features 223–228
219, 220 231–236, 235 point‐of‐care testing 360–361
sickle cell/haemoglobin C neonatal screening 355–356 pre‐anaesthetic testing 357
disease 226 pre‐anaesthetic testing 357 sickle cell/haemoglobin C‐Harlem
service, haemoglobinopathy S/β0 218, 229, 232 compound heterozygosity
diagnostic 348–364 S/β+ 68, 196, 229, 231–232 234–235, 240
antenatal and preconceptual sickle cell crises sickle cell/haemoglobin D‐Punjab
testing 349–353 blood count 209 compound heterozygosity
fetal diagnosis 353 blood films 210–211, 213 case study 382, 403
neonatal screening 354–357 sickle cell anaemia 203 sickle cell/haemoglobin D‐Punjab/
other investigations 359–360 sickle cell/β thalassaemia 229 Los Angeles disease 233,
pre‐anaesthetic testing 357–359 sickle cell/haemoglobin C 237–238
resource‐poor settings 360–361 disease 222, 223 case study 374–375, 400
sickle cell/α+ thalassaemia sickle cell/δ0β+ thalassaemia 240 sickle cell/haemoglobin E compound
case study 378, 402 sickle cell/δβ0 thalassaemia heterozygosity 238–240,
laboratory features 208, 210, 212 218, 236 242–244
428 Index
sickle cell/haemoglobin sickle cell trait 187–201 prenatal diagnosis 76, 141
G‐Philadelphia compound α thalassaemia trait skin prick blood samples 30
heterozygotes 200–201, 201 coinheritance 194, 194 skull deformity 144, 145, 204, 205
case study 383, 403 β thalassaemia in cis with 196, 201 smoking, cigarette 302, 329, 334
sickle cell/haemoglobin Hofu blood count 194, 194 somatic mosaicism 22
compound blood films 194, 195 somatic mutations
heterozygosity 242 capillary electrophoresis 194, 199 acquired haemoglobin H
sickle cell/haemoglobin Hope case studies 365, 382, 399, 403 disease 105, 327–328
compound heterozygosity clinical features 192–193 β thalassaemia 141, 142
241–242, 243 coinheritance 198–201 South‐East Asian ovalocytosis
sickle cell/haemoglobin Kenya diabetes with 192, 397, 406 72, 134
compound diagnosis 68, 197 Southern blot analysis 77, 77–78
heterozygosity 242 haemoglobin electrophoresis spectroscopy
sickle cell/haemoglobin 194, 196 carboxyhaemoglobin 335, 336
Lepore 235–236, 241 hereditary spherocytosis methaemoglobin 309, 310,
sickle cell/haemoglobin Monroe with 192, 201 336, 339
compound HPLC 194, 197–198 sulphaemoglobin 341
heterozygosity 243 laboratory features 194–197 sperm, donor 353
sickle cell/haemoglobin New York neonatal screening 355 spherocytosis, hereditary 192,
compound pre‐anaesthetic testing 357–359, 201, 229
heterozygosity 243 358 splenectomy
sickle cell/haemoglobin O‐Arab pyruvate kinase deficiency β thalassaemia major 148
disease 233–234 with 192, 201, 244 β thalassaemia trait 123, 125
case study 381, 403 transfusion from donor haemoglobin E/β thalassaemia
laboratory features 233–234, with 372, 399–400 286, 287, 288
239–240, 240 sickle cell trait/haemoglobin H haemoglobin H disease 105, 106
sickle cell/haemoglobin disease 198–200, 199–200 unstable haemoglobins 303
Quebec‐Chori compound case study 393, 405 splenic infarction
heterozygosity 241 haemoglobin S levels 196, 198, 200 case study 380, 402–403
sickle cell/haemoglobin S‐Antilles sickle cells 185, 186 sickle cell anaemia 203
compound haemoglobin S‐Oman 243, 246 sickle cell/β thalassaemia 229
heterozygosity 241 sickle cell anaemia 202, 210, sickle cell/haemoglobin C
sickle cell/haemoglobin S‐Oman 211–214, 217, 219, 220 disease 222
compound heterozygosity sickle cell/haemoglobin C sickle cell trait 192
241, 246 disease 224, 224 splenic rupture, spontaneous 192,
sickle cell/haemoglobin Siriraj sickle cell/haemoglobin 202, 270
compound heterozygosity D‐Punjab disease 233, 237 splenic sequestration
241–242, 245–246 sickle cell trait 192–193, 194 sickle cell anaemia 202, 209, 212
sickle cell/haemoglobin Tyne sickle solubility test 33, 63, 63–64 sickle cell/haemoglobin C
compound case studies 365, 376, 398, 402 disease 222, 224
heterozygosity 242 false positives 63, 63–64 sickle cell trait 192
sickle cell/haemoglobin Volga pre‐anaesthesia 357–358 splenomegaly
compound sickle cell anaemia 215 β thalassaemia major 144
heterozygosity 241 sickle cell trait 194–196 β thalassaemia trait 122
sickle cell/hereditary persistence of sickle test 63 haemoglobin H disease 105
fetal haemoglobin (S/HPFH) SickleSCAN test 361 sickle cell anaemia 203–204
160, 218, 236–237 silent β thalassaemia trait 129, sickle cell/β thalassaemia 229
βS haplotypes and haemoglobin F 130–131, 132 sickle cell/haemoglobin C
levels 219–221 almost 130–131, 132 disease 222, 223
blood film 237, 242 β0 thalassaemia sickle cell/haemoglobin
neonatal screening 356 coinheritance 143 D‐Punjab disease 233
Index 429
split A2 band 33, 65, 65 haemoglobin E/β M haemoglobins 308

SSP (stage selector protein) 11 thalassaemia 286 mutations 301–302
STOP codons see termination codons thalassaemia minor see β phenotypes 302–303
stress erythropoiesis 210, 214, thalassaemia minor post‐translational
329, 330 thalassaemic haemoglobinopathies modification 301
sudden infant death syndrome 329 117, 120 uroporphyrinogen III 7, 8
sulphaemoglobin 4, 337–341 TOX gene 15
sulphaemoglobinaemia 337–341, trans‐acting factors, variant haemoglobins 16–24
340 erythroid‐specific 11 % of total haemoglobin 22–24,
transcription, globin gene 10, 261–262
target cells 10–11, 12 abnormalities of globin chain
haemoglobin C disease 268, transcription factors 11 synthesis 16–22
268–269 transfer RNA (tRNA) 12 diagnostic vs clinical
haemoglobin C trait 264, 264–265 transferrin 7, 8 importance 261
sickle cell anaemia 210, 214 transferrin receptors (TfR) 7, 8 electrophoresis 31–38, 32,
sickle cell/β thalassaemia 230 β thalassaemia major 149 36, 38
sickle cell/haemoglobin C transfusion reactions, sickle cell functional abnormalities 262
disease 224, 225, 225, 226 anaemia 214 HPLC 38–47, 47–59, 60
sickle cell trait 194, 195 transient erythroblastopenia of identification of
termination (STOP) codons 12 childhood 333 unknown 359–360
α thalassaemias 89, 89–90 translation, globin chain 10, 11–12 immunoassays 69–70
β thalassaemias 119, 120 mutations affecting 89–90 incidental detection 73
mutations affecting 17, 19, 20, 21 translocations, chromosomal 19 isoelectric focusing 61–62, 62
termination of pregnancy 349, 353 triple α (ααα) 18, 86 other significant 261–324
thalassaemia 22, 85–184 εγδβ thalassaemia specialised diagnostic tests 73–79
acquired 325–329, 328 coinheritance 153 tetramer stability 22, 32–33
analysis of globin chain synthesis mechanism of generation 86, 90 thalassaemias and 22
rate 74–75, 75 prenatal testing 141 vascular occlusion
antenatal screening 349–353 sickle cell trait with 196 sickle cell anaemia 202–204, 204
classification 85, 86 triple α/ β thalassaemia trait 134 sickle cell trait 192, 193
DNA analysis 76, 76–78 β thalassaemia intermedia 141 venous thromboembolism
neonatal screening 354 β thalassaemia major 142 β thalassaemia 141, 144
osmotic fragility screening test 72 case study 386, 404 sickle cell/haemoglobin C
prevalences 90, 92–95 disease 223
unstable haemoglobin UGT1A1 polymorphism 284 sickle cell trait 193
coinheritance 306–307 umbilical cord blood samples 30,
unstable haemoglobins 90, 120, 354, 354 white cell count (WBC)
302–303 uniparental disomy 142, 244 β thalassaemia major 148
see also specific thalassaemias acquired 329 sickle cell anaemia 208, 209
thalassaemia intermedia unstable haemoglobins 301–307
antenatal screening 350 % of total haemoglobin 302 XPD (ERCC2) gene mutation 20,
β thalassaemia see β thalassaemia α thalassaemias 90, 303 120, 121
intermedia β thalassaemias 120, 302–303
δβ thalassaemia 150 clinical features 303 ZBTB7A gene 15
haemoglobin E/β coinheritance 306–307 ζ globin chain 3, 3
thalassaemia 286 detection 70–71, 71 α0 thalassaemia 102
haemoglobin E/haemoglobin H diagnosis 306 gene 9, 9
disease 290 haemoglobin H 103 zinc protoporphyrin
thalassaemia major high affinity 303, 310 α0 thalassaemia 102
β thalassaemia see β thalassaemia laboratory features 303–306 β thalassaemia trait 123
major low affinity 313 Zola, Emile 398, 406

Barbara J Bain - Haemoglobinopathy Diagnosis

Uploaded by

Copyright:

Available Formats

Barbara J Bain - Haemoglobinopathy Diagnosis

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Barbara J Bain - Haemoglobinopathy Diagnosis

Uploaded by

Copyright:

Available Formats

Haemoglobinopathy Diagnosis

Honorary Consultant Haematologist

Limit of Liability/Disclaimer of Warranty

Library of Congress Cataloging‐in‐Publication Data

Names: Bain, Barbara J., author.

Set in 9/12pt Palatino by SPi Global, Pondicherry, India

α the Greek letter, alpha γ thalassaemia a thalassaemic condition resulting

haemoglobin J a group of variant haemoglobins HIV human immunodeficiency virus

Haemoglobins and their structure

Haemoglobinopathy Diagnosis, Third Edition. Barbara J. Bain.

Yolk sac Liver Spleen Bone marrow

Globin synthesis (%)

Haemoglobin Globin chains Period when normally present

A α2β2* Major haemoglobin in adult life

Gower 1 ζ2ε2 Significant haemoglobin during early intrauterine life

Gower 2 α2ε2 Significant haemoglobin during early intrauterine life

Portland or ζ2γ2 Significant haemoglobin during early intrauterine life

Oxygen saturation (%)

1.3 2.5 5 7.5 10 12.5 13 kPa

Fig. 1.8 Diagram of chromosomes 11

Fig. 1.9 Diagrammatic representation of the LCRA ζ ψζ ψα2 ψα1 α2 α1 θ

Table 1.2 The sequences

ε (HBE1) CCAAT AATAAAG

β (HBB) CACCC CCAAT CATAAAA

δ (HBD) CCAAC CATAAAA

Fig. 1.11 Diagram showing the rate of 1

Fig. 1.12 High performance liquid

Type of mutation Possible consequence Examples

Gene conversion* Conversion of Gγ gene to Aγ gene, giving

Conversion of Aγ gene to Gγ gene, giving

Conversion of ψζ1 to a gene that

Within non‐coding sequence, Production of a new splice site Some β thalassaemias

Mutation 5′ or 3′ to the gene Mutation of an enhancer Some β thalassaemias

Type of mutation Possible consequence Examples

Deletion or duplication of one or more genes or pseudogenes

Duplication of α gene Triple, quadruple or quintuple ααα†/αα, ααα/ααα, αααα‡/αα, αααα/αααα

Triplication of entire α globin gene Six α genes on a single chromosome αα:αα:αα/αα

Duplication of Gγ gene Double, triple or quadruple Gγ gene γ γ γ, GγGγGγAγ (homozygotes have

Duplication of the ζ or ψζ gene Double, triple or quadruple ζ2ψζ1ψζ1/ζ2ψζ1 or ζ2ψζ1ψζ1/ζ2ψζ1ψζ1

Abnormal crossover during meiosis leading to gene fusion

δβδ fusion – double crossover Reduced rate of synthesis of Haemoglobin Parchman

βδ fusion (with preservation of Anti‐Lepore haemoglobins, e.g.

βAγ fusion (with preservation of Haemoglobin anti‐Kenya

γ γ fusion (designated ‐GγAγ‐)

Deletion of DNA sequences but without a frameshift in coding sequence

Type of mutation Possible consequence Examples

Deletion plus inversion

Deletion plus insertion

Insertion within a coding sequence but without a frameshift

Insertion of a duplicated sequence including the start codon

Loss of an α gene α thalassaemia; may be part of a

Deletion of a locus control region

Creation of a sequence that competes for the locus control region

Mutation in FCP1 at Xp22.2 Hereditary persistence of fetal haemoglobin

An HBSIL‐MYB intergenic polymorphism at 6q22.3–23.1 Hereditary persistence of fetal haemoglobin

BCL11A haploinsufficiency (2p16.1) Increased haemoglobin F associated with cognitive

KLF1 at 19p13.13 Increased haemoglobin A2 [42, 43]

ASH1L at 1q22 β thalassaemia [76]

Anti-Lepore βδ fusion gene

α genes on one ­chromosome. Duplicated α genes with a structurally abnormal haemoglobin,

α genes on one chromosome. Duplicated α genes with a structurally abnormal haemoglobin,

Capillary isoelectric focusing A sickle solubility test can be performed by pur