Orthopedic Trauma (2016)

Orthopedic Trauma
Diagnosis, Operative
Techniques and Management
Newman Wagner
Orthopedic Trauma: Diagnosis, Operative Techniques and
Management
Orthopedic Trauma: Diagnosis,
Operative Techniques and
Management
Edited by Newman Wagner
Published by Academic Pages,
5 Penn Plaza,
19th Floor,
New York, NY 10001, USA
Orthopedic Trauma: Diagnosis, Operative Techniques and Management

Edited by Newman Wagner
© 2017 Academic Pages
International Standard Book Number: 978-1-9789-2749-0
This book contains information obtained from authentic and highly regarded sources. Copyright for all individual chapters
remain with the respective authors as indicated. All chapters are published with permission under the Creative Commons
Attribution License or equivalent. A wide variety of references are listed. Permission and sources are indicated; for detailed
attributions, please refer to the permissions page and list of contributors. Reasonable efforts have been made to publish reliable
data and information, but the authors, editors and publisher cannot assume any responsibility for the validity of all materials
or the consequences of their use.
Copyright of this ebook is with Academic Pages, rights acquired from the original print publisher, Callisto Reference.
The publisher’s policy is to use permanent paper from mills that operate a sustainable forestry policy. Furthermore, the
publisher ensures that the text paper and cover boards used have met acceptable environmental accreditation standards.
Trademark Notice: Registered trademark of products or corporate names are used only for explanation and identification
without intent to infringe.
Contents

Preface IX
Chapter 1 Are work return and leaves of absence predictable after an unstable pelvic
ring injury? 1
Alessandro Aprato, Alexander Joeris, Ferdinando Tosto, Vasiliki Kalampoki,
Elke Rometsch, Marco Favuto, Alessandro Stucchi, Matheus Azi
and Alessandro Massè
Chapter 2 Adjacent segment infection after surgical treatment of spondylodiscitis 6

Ahmed Ezzat Siam, Hesham El Saghir and Heinrich Boehm
Chapter 3 Single intra-articular injection of high molecular weight hyaluronic

acid for hip osteoarthritis 17
Fabrizio Rivera
Chapter 4 Long-term clinical results and MRI changes after autologous chondrocyte
implantation in the knee of young and active middle aged patients 23
Donato Rosa, Giovanni Balato, Giovanni Ciaramella, Ernesto Soscia,
Giovanni Improta and Maria Triassi
Chapter 5 Outcomes of intramedullary nailing for acute proximal humerus fractures:

a systematic review 31
Jason Wong, Jared M. Newman and Konrad I. Gruson
Chapter 6 Mid-term results of Miller-Galante unicompartmental knee replacement

for medial compartment knee osteoarthritis 41
Hemanth Kumar Venkatesh and S. S. Maheswaran
Chapter 7 An unusual case of neurothekeoma of the arm in an adult 49

Federica Bergamin, Ezio Nicola Gangemi, Claudia Cerato, Alessandra Clemente,
Marco Borsetti, Adolfo Suriani and Stefano Taraglio
Chapter 8 Achieving hip fracture surgery within 36 hours: an investigation of risk factors
to surgical delay and recommendations for practice 53
Adeel Aqil, Fahad Hossain, Hassaan Sheikh, Joseph Aderinto, George Whitwell
and Harish Kapoor
Chapter 9 Effectiveness of intra-articular injections of sodium hyaluronate-chondroitin

sulfate in knee osteoarthritis: a multicenter prospective study 60
Fabrizio Rivera, Luca Bertignone, Giancarlo Grandi, Roberto Camisassa,
Guido Comaschi, Diego Trentini, Marco Zanone, Giuseppe Teppex,
Gabriele Vasario and Giorgio Fortina
__________________________ WORLD TECHNOLOGIES __________________________

9I &RQWHQWV
Chapter 10 Congenital idiopathic talipes equinovarus before and after walking age:
observations and strategy of treatment from a series of 88 cases 67
Cesare Faldini, Francesco Traina, Matteo Nanni, Ilaria Sanzarello, Raffaele Borghi
and Fabrizio Perna
Chapter 11 A modified Austin/chevron osteotomy for treatment of hallux valgus

and hallux rigidus 74
Michele Vasso, Chiara Del Regno, Antonio D’Amelio and Alfredo Schiavone Panni
Chapter 12 The use of a dual mobility cup in the management of recurrent dislocations
of hip hemiarthroplasty 79
Christian Carulli, Armando Macera, Fabrizio Matassi, Roberto Civinini
and Massimo Innocenti
Chapter 13 Biocompatibility of magnesium implants in primary human reaming

debris-derived cells stem cells in vitro 85
Olga Charyeva, Olga Dakischew, Ursula Sommer, Christian Heiss,
Reinhard Schnettler and Katrin Susanne Lips
Chapter 14 Correlating preoperative imaging with intraoperative fluoroscopy in iliosacral

screw placement 96
Jonathan G. Eastman and Milton L. Chip Routt Jr.
Chapter 15 Cross-cultural adaptation and validation of the International Knee

Documentation Committee Subjective Knee Form in Greek 104
George A. Koumantakis, Konstantinos Tsoligkas, Antonios Papoutsidakis,
Athanasios Ververidis and Georgios I. Drosos
Chapter 16 Effectiveness of extracorporeal shockwave therapy in three major

tendon diseases 111
Christian Carulli, Filippo Tonelli, Matteo Innocenti, Bonaventura Gambardella,
Francesco Muncibì and Massimo Innocenti
Chapter 17 Arthrodesis of proximal inter-phalangeal joint for hammertoe: intramedullary

device options 117
Matteo Guelfi, Andrea Pantalone, Janos Cambiaso Daniel, Daniele Vanni,
Marco G. B. Guelfi and Vincenzo Salini
Chapter 18 FiberWire tension band for patellar fractures 122

Lawrence Camarda, Alessandra La Gattuta, Marcello Butera, Francesco Siragusa
and Michele D’Arienzo
Chapter 19 The role of femoral offset and abductor lever arm in total hip arthroplasty 128
Filip Bjørdal and Kristian Bjørgul
Chapter 20 Supplemental S1 fixation for type C pelvic ring injuries: biomechanical study
of a long iliosacral versus a transsacral screw 134
Pooria Salari, Berton R. Moed and J. Gary Bledsoe
Chapter 21 Effectiveness of psychological support in patients undergoing primary total hip

or knee arthroplasty: a controlled cohort study 142
V. Tristaino, F. Lantieri, S. Tornago, M. Gramazio, E. Carriere and A. Camera

&RQWHQWV 9,,
Chapter 22 Blood metal ions after hybrid metal-on-polyethylene Exeter—Trident total

hip replacement 153
Rohit Singh, Gopikanthan Manoharan, Pete Craig, Simon Collier, Phillip Shaylor
and Ashok Sinha
Chapter 23 Intraosseous concentration and inhibitory effect of different intravenous

cefazolin doses used in preoperative prophylaxis of total knee arthroplasty 158
Chayanin Angthong, Pongpaibool Krajubngern, Warawut Tiyapongpattana,
Boonchana Pongcharoen, Piya Pinsornsak, Nattapol Tammachote
and Wanna Kittisupaluck
Chapter 24 External versus internal fixation for bicondylar tibial plateau fractures:
systematic review and meta-analysis 162
David Metcalfe, Craig J. Hickson, Lesley McKee and Xavier L. Griffin
Chapter 25 Deep peroneal nerve palsy with isolated lateral compartment syndrome
secondary to peroneus longus tear: a report of two cases and a review
of the literature 173
Kunihiko Hiramatsu, Yasukazu Yonetani, Kazutaka Kinugasa,
Norimasa Nakamura, Koji Yamamoto, Hideki Yoshikawa and Masayuki Hamada
Chapter 26 Isolated reconstruction of the medial patellofemoral ligament with autologous

quadriceps tendon 178
Giovanni Vavalle and Michele Capozzi
Chapter 27 Platelet-rich plasma versus autologous blood versus steroid injection in lateral
epicondylitis: systematic review and network meta-analysis 186
Alisara Arirachakaran, Amnat Sukthuayat, Thaworn Sisayanarane,
Sorawut Laoratanavoraphong, Wichan Kanchanatawan
and Jatupon Kongtharvonskul
Chapter 28 Is there a link between the neutrophil-to-lymphocyte ratio and venous

thromboembolic events after knee arthroplasty? A pilot study 198
Tyler Barker, Victoria E. Rogers, Vanessa T. Henriksen, Kimberly B. Brown,
Roy H. Trawick, Nathan G. Momberger and G. Lynn Rasmussen
Chapter 29 Calcific tendinitis of the rotator cuff: state of the art in diagnosis and treatment 204
Giovanni Merolla, Sanjay Singh, Paolo Paladini and Giuseppe Porcellini
Chapter 30 Is end-stage lateral osteoarthritic knee always valgus? Mechanical alignment

analysis and radiographic severity assessment 212
Su Chan Lee, Viral Gondalia, Byoung Yoon Hwang, Hye Sun Ahn, Choon Key Lee,
David J. Hunter and Kwang Am Jung
Chapter 31 Investigation of bone quality of the first and second sacral segments amongst
trauma patients: concerns about iliosacral screw fixation 218
Dane Salazar, Sean Lannon, Olga Pasternak, Adam Schiff, Laurie Lomasney,
Erika Mitchell and Michael Stover
Chapter 32 Aseptic lysis L2–L3 as complication of abdominal aortic aneurysm repair 226
Federico Mancini, Andrea Ascoli-Marchetti, Luca Garro and Roberto Caterini
Chapter 33 Malunited extra-articular distal radius fractures: corrective osteotomies using

volar locking plate 230
Luigi Tarallo, Raffaele Mugnai, Roberto Adani and Fabio Catani

9,II&RQWHQWV
Chapter 34 Complication rates and reduction potential of palmar versus dorsal locking
plate osteosynthesis for the treatment of distal radius fractures 236
F. Wichlas, N. P. Haas, A. Disch, D. Machó and S. Tsitsilonis
Chapter 35 Non-operative treatment versus percutaneous fixation for minimally displaced

scaphoid waist fractures in high demand young manual workers 242
Haroon Majeed
Chapter 36 Variable-angle locking plate with or without double-tiered subchondral support

procedure in the treatment of intra-articular distal radius fracture 248
Keikichi Kawasaki, Tetsuya Nemoto, Katsunori Inagaki, Kazunari Tomita
and Yukio Ueno
Chapter 37 Predictive factors of hospital length of stay in patients with operatively treated
ankle fractures 252
Matthew R. McDonald, Vasanth Sathiyakumar, Jordan C. Apfeld, Benjamin Hooe,
Jesse Ehrenfeld, William T. Obremskey and Manish K. Sethi

Permissions

List of Contributors

Preface
This book has been an outcome of determined endeavour from a group of educationists in the field. The primary
objective was to involve a broad spectrum of professionals from diverse cultural background involved in the field
for developing new researches. The book not only targets students but also scholars pursuing higher research for
further enhancement of the theoretical and practical applications of the subject.
This book provides comprehensive insights into the field of orthopedic trauma. It involves diagnosis and treatment of
diseases related to the musculoskeletal system. Orthopedics was usually restricted to treating structural deformities
but now it has forayed into wider areas such as sports injuries, degenerative diseases, infections, tumors, etc. These
problems can affect the physical movement of a person and may persist for a longer period of time. As this field
is emerging at a rapid pace, the contents of this book will help the readers understand the modern concepts and
applications of the subject. It presents researches and studies performed by experts across the globe. This book
will serve as a valuable source of reference to a broad spectrum of readers such as orthopedicians, surgeons,
traumatologists, professionals and students.
It was an honour to edit such a profound book and also a challenging task to compile and examine all the relevant
data for accuracy and originality. I wish to acknowledge the efforts of the contributors for submitting such brilliant
and diverse chapters in the field and for endlessly working for the completion of the book. Last, but not the least;
I thank my family for being a constant source of support in all my research endeavours.
Editor

1
Are work return and leaves of absence
predictable after an unstable pelvic ring injury?
Alessandro Aprato1 • Alexander Joeris2 • Ferdinando Tosto1 • Vasiliki Kalampoki2 •
Elke Rometsch2 • Marco Favuto1 • Alessandro Stucchi1 • Matheus Azi3 •

Alessandro Massè1
Abstract not associated with any of the factors investigated. Poly-

Background Resuming work after surgical treatment of trauma, ICU admission, and time from trauma to definitive
an unstable pelvic ring injury is often impeded because of surgery were associated with longer leaves of absence.
residual disability. The aim of this study was to test which Conclusions Work reintegration after pelvic ring injuries
factors influence return to work, ability to return to the is a major issue for patients and health care systems: 58 %
same job function as before the injury, leaves of absence, of patients were not able to return to or lost their job.
and incapacitation after sustaining a pelvic fracture. Factors correlated with leaves of absence were injury
Materials and methods We performed a retrospective severity, delayed definitive fixation, and ICU admission.
study on patients with surgically treated pelvic fractures. Level of evidence IV (case series).
Medical records were reviewed to document patients’
demographic data, the extent of follow-up care, diagnosis Keywords Pelvic fracture Pelvic ring injuries
of the injury (according to the Tile system of classifica- Morbidity Productivity loss Leave of absence
tion), type of surgical treatment, injury severity, and the
time from trauma to definitive surgery. We also recorded
the classification of patients’ physical status according to Introduction
the American Society of Anesthesiologists (ASA) and
details about admission to the intensive care unit (ICU). Morbidity and complications are frequent after an unsta-
Patients were interviewed to note the number of days ble pelvic injury [1]. Patients usually undergo a long
before returning to work and their ability to maintain their rehabilitation and injury often creates a chronic disabling
previously held jobs. condition, which frequently requires long-lasting analgesic
Results Fifty patients were included in the study, and their therapies and impairs the ability to work [2]. The latter
mean age was 46.3 ± 12.6 years. The median time to return generates a huge societal impact through loss of employ-
to work was 195 days. Twelve patients (24 %) lost their ment or the need for professional retraining. Many studies
jobs and 17 (34 %) resumed their previous job with a [3–18] have shown low rates of job reintegration after a
change of tasks. ICU admission and time from trauma to pelvic fracture but, to our knowledge, the association
definitive surgery were negatively correlated with return to between days of work absence and pre-trauma health sta-
the previously held job. Returning to the same job tasks was tus, the severity of injury, and job characteristics has not
yet been investigated.
& Alessandro Aprato
We performed a retrospective study on patients with
[email protected] pelvic ring injuries treated surgically in our referral center.
The main aim of this study was to evaluate the type of
1
Medical School, University of Turin, Turin, Italy fracture, pre-trauma health status, time from trauma to
2
Clinical Investigation and Documentation (C.I.D.) definitive surgery, severity of injury, and job characteris-
Department, AO Foundation, Dübendorf, Switzerland tics, and their influence on work resumption, the ability to
3
Manoel Victorino Hospital, Salvador, Brazil maintain the previously held job, and leaves of absence.

2 Orthopedic Trauma: Diagnosis, Operative Techniques and Management
Materials and methods Table 1 Demographic, clinical, and job-related characteristics of the
study participants
Between 2010 and 2012, we surgically treated 93 pelvic Variables
ring injuries in our referral center. All fractures were
Gender
operated upon by at least two surgeons of our pelvic sur-
Female 18 (36.0)
gery team, which consists of three specialized surgeons.
Male 32 (64.0)
The study was approved by the local ethics committee and
was conducted in accordance with the ethical standards laid Age at surgery (years) 46.3 ± 12.6
down in the 1964 Declaration of Helsinki and its later Time from trauma to definitive surgery (days) 11 (6; 15)
amendments. Hospital charts were retrospectively Follow-up period (months) 38.5 (18; 60)
reviewed after patients had given informed consent to the Type of fracture
use of their data. B 12 (24.0)
Patients were considered ineligible if they did not work C 38 (76.0)
before the trauma (e.g., students and retirees), if they had Polytrauma (ISS [15)
been operated on less than 9 months previously so that no No 36 (72.0)
follow-up of 9 months or longer was possible, or if no Yes 14 (28.0)
phone contact information was available in the medical ASA classification score
records. Data on demographics, diagnoses (according to the 0–1 40 (80.0)
Tile system of classification [19]), type of surgical treat- 2–3 10 (20.0)
ment, Injury Severity Score (ISS) [20] on arrival, American ICU admission
Society of Anesthesiologists’ (ASA) physical status clas- No 18 (36.0)
sification [21], intensive care unit (ICU) admission, time Yes 32 (64.0)
from trauma to definitive surgery, and follow-up were Sedentary worker
retrieved from medical records and recorded in a custom No 25 (50.0)
database. Yes 25 (50.0)
Fractures were grouped into B and C types, according to Job sector
the Tile classification system. The pre-trauma health status Private 40 (80.0)
was classified as morbid for ASA scores higher than one. Public 10 (20.0)
Polytrauma was defined as an ISS score of greater than 15 Resumption of the previous job
points on hospital admission [22]. No 12 (24.0)
Patients were interviewed by phone about their type of Yes 38 (76.0)
work, their ability to return to work, number of days before Maintenance of the same job tasks a
returning to work, and their ability to return to the same job No 17 (46.0)
tasks. Yes 20 (54.0)
Time to get back to work (days)a 195 (150; 300)
Statistical methodology
Data are presented as N (%), mean ± standard deviation, or median
(p25; p75)
All data were analyzed with standard descriptive statistics.
ASA American Society of Anesthesiologists physical status classifi-
Univariate analysis was performed with regard to (1)
cation; ICU intensive care unit
readmission to the former job (yes or no) and (2) mainte- a
Of the 38 patients who resumed their previous job, information
nance of the same job tasks (yes or no). This was done with regarding maintenance of the same job tasks was not available for one
the chi-squared test or Fisher’s exact test for categorical patient
outcomes and Student’s t test or the Mann–Whitney test for
continuous outcomes. The Kolmogorov–Smirnov test was considered statistically significant. All analyses were per-
used to determine whether data were normally distributed. formed using Stata version 12 (Stata Corporation, College
The relationship between leaves of absence and study Station, TX, USA).
characteristics was assessed with univariate linear regres-
sion models. Since the values for days off work were
skewed, they had to be log-transformed to use them in the Results
regression models. As a consequence, regression coeffi-
cients and 95 % confidence intervals were converted into a Of the 88 patients surgically treated for a pelvic fracture in
percent increase in the respective variable using the for- our referral center, 12 patients were lost to follow-up, 15
mula [exp(b)-1] 9 100. P values lower than 0.05 were were excluded because their surgery had taken place less

Are work return and leaves of absence predictable after an unstable pelvic ring injury? 3
Table 2 Demographic, clinical, and job-related characteristics of the study participants according to (a) resumption of the previous job and
(b) maintenance of the same job tasks along with univariate association tests
Variables Resumption of the former job P value Maintenance of the same job tasks P value
No (N = 12) Yes (N = 38) No (N = 17) Yes (N = 20)
Gender 0.639* 0.717*

Female 5 (41.7) 13 (34.2) 5 (29.4) 7 (35.0)
Male 7 (58.3) 25 (65.8) 12 (70.6) 13 (65.0)
Age at surgery (years) 46.1 ± 9.8 46.4 ± 13.5 0.937 46.2 ± 12.4 47.5 ± 14.4 0.768
Time from trauma to definitive surgery (days) 14.5 (9; 21.5) 10 (5; 15) 0.047à 12 (8; 15) 7.5 (5; 13) 0.069à
§
Type of fracture 0.705 0.288§
B 2 (16.7) 10 (26.3) 3 (17.6) 7 (35.0)
C 10 (83.3) 28 (73.7) 14 (82.4) 13 (65.0)
Polytrauma (ISS [15) 0.718§ 0.763*
No 8 (66.7) 28 (73.7) 12 (70.6) 15 (75.0)
Yes 4 (33.3) 10 (26.3) 5 (29.4) 5 (25.0)
ASA classification score 0.686§ 0.999§
0–1 9 (75.0) 31 (81.6) 14 (82.4) 17 (85.0)
2–3 3 (25.0) 7 (18.4) 3 (17.6) 3 (15.0)
ICU admission 0.036§ 0.272*
No 1 (8.3) 17 (44.7) 9 (52.9) 7 (35.0)
Yes 11 (91.7) 21 (55.3) 8 (47.1) 13 (65.0)
Sedentary worker 0.508* 0.072*
No 7 (58.3) 18 (47.4) 11 (64.7) 7 (35.0)
Yes 5 (41.7) 20 (52.6) 6 (35.3) 13 (65.0)
Job sector 0.416§ 0.137§
Private 11 (91.7) 29 (76.3) 15 (88.2) 13 (65.0)
Public 1 (8.3) 9 (23.7) 2 (11.8) 7 (35.0)
Data are presented as N (%), mean ± standard deviation, or median (p25; p75)
ASA American Society of Anesthesiologists physical status classification; ICU intensive care unit
à
P value derived from *Chi-squared test; Student’s t test; Mann–Whitney test; §
Fisher’s exact test
than 9 months before the start of the study, 6 patients were (gender, age at surgery, type of fracture, ISS, pre-trauma
students, and 5 had retired before the trauma occurred. health status, sedentary work, and job sector) were not
Thus, 50 patients were included in the analysis. Their mean associated with return to work (all p values [0.05).
age was 46.3 years (range 18–83) and men represented a With regard to maintaining the same job tasks, there was
higher proportion (64 %). Demographic, clinical, and job- a trend of patients with a shorter time from trauma to
related data are presented in Table 1. Almost 75 % of the definitive surgery (p = 0.069) and sedentary work
study patients suffered a type C fracture. Of the 38 (76 %) (p = 0.072) to be more likely to maintain the same tasks.
patients who returned to their previous job, 20 (54 %) None of the remaining characteristics was associated with
managed to maintain the same job tasks. The median time maintenance of the same job tasks upon job resumption.
to return to work was 195 days (range 150–300). The univariate regression models investigating the
The analysis of potential predictors for returning relationship of patient characteristics with leaves of
(N = 38) or not returning (N = 12) to the previously held absence are shown in Table 3. We found highly significant
job revealed that job resumption was significantly associ- evidence that polytrauma patients have more than twice as
ated with ICU admission (p = 0.036); of the 12 patients many days off work (representing an increase of 120 % of
who lost their jobs, 11 had been in an ICU (Table 2). There the average days of work absence; p \ 0.0001) compared
was some evidence that job resumption was also associated with non-polytrauma patients. In addition, there was evi-
with the time elapsing between trauma and definitive sur- dence for an association of time to get back to work with
gery; a longer time appeared to have a negative effect on both the time elapsing between trauma and definitive sur-
return to work (p = 0.047). All remaining characteristics gery and with ICU admission. The average increase in days

Table 3 Univariate linear

Variables Category/increment Percent change (95 % CI) P value
regression models evaluating
the relationship between leave Gender Female Baseline 0.154
of absence (in days) and clinical
and job-related characteristics Male 27.2 (-53.3, 13.2)
of the study participants: Age at surgery 10 years more -0.2 (-15.2, 17.4) 0.979
percent change, 95 % Time from trauma to definitive surgery 10 days more 41.3 (6.5, 87.4) 0.018
confidence intervals (CIs), and
Type of fracture B Baseline 0.995
P values
C 0.2 (-38.7, 63.6)
Polytrauma No Baseline \0.0001
Yes 120.1 (45.9, 232.3)
ASA classification score 0–1 Baseline 0.997
2–3 0.1 (-42.7, 74.9)
ICU admission No Baseline 0.019
Yes 62.9 (9.0, 143.5)
Sedentary worker No Baseline 0.620
Yes 11.3 (-27.7, 71.3)
Job sector Private Baseline 0.096
Public 51.0 (-7.4, 146.2)
The models predicted log10-transformed leaves of absence. All parameter estimates have been
exponentiated
ASA American Society of Anesthesiologists physical status classification; ICU intensive care unit
off work was 41.3 % (p = 0.018) for each additional productivity loss) if the ability to return to work is used as
10 days from trauma to definitive surgery and 62.9 % the sole outcome.
(p = 0.019) for patients who had been admitted to ICU Our study has several limitations. Patients in an unsta-
compared to patients with no ICU admission. None of the ble clinical condition could not be transferred to our hos-
remaining study variables were significantly related to pital and were treated by our team of surgeons on-site.
leaves of absence. They were not included in the study due to the difficulties
of retrieving their charts from a hospital other than ours.
This ultimately means that our study population is subject
Discussion to selection bias because the most severe cases are not
contained. It is conceivable that the work resumption rates
Pelvic fractures severely affect the post-trauma work proof these patients would probably range at the lower end.
ductivity of the patient. In this study, we evaluated how Therefore, our results may underestimate the true impact of
factors, such as complexity of the fracture, pre-trauma pelvic fracture on societal costs.
health status, time from trauma to definitive surgery, Limitations of this analysis also include selection and
severity of injury, and job characteristics influenced information biases as well as general limitations of work-
patients’ return to work, the ability to maintain the former ing with phone interview data. With regard to limitations of
job tasks, and the extent of leave of absence. the study design, the retrospective collection of data and
Many studies [3–18] have described the productivity patient interview limit the conclusions drawn from this
loss subsequent to these fractures but time to return to work study. Furthermore, in our study, only employees from a
has not been described. We found that the mean time to single European country were included; thus, findings may
return to work is approximately 195 days in patients who not be generalizable to other geographical regions.
underwent surgery for a pelvic ring injury. Our results show also similarity to Gabbe et al.’s study
A systematic review [1] showed work return rates [23]; severity of injury (evaluated by ISS, ICU admission,
ranging from 57 to 84 % for operatively treated, unstable, or both) and not fracture type was identified as an impor-
and open book pelvic fractures; our population showed tant predictor for work return in both studies. Additionally,
similar results (24 % chance of losing the job). Further- we found that the time elapsed between trauma and
more, our study showed that only 46 % of patients who definitive surgery had a significant correlation with time to
resumed their job were able to maintain their job tasks, return to work, although this parameter may be interpreted
which emphasizes the risk of underestimating the true as linked to the severity of injury. Time from trauma to
impact (e.g., through the need for retraining or because of definitive surgery has been considered a good indicator of

Are work return and leaves of absence predictable after an unstable pelvic ring injury? 5
clinical outcomes for pelvic fracture [24]; according to our 3. Browner BD, Cole JD, Graham JM, Bondurant FJ, Nunchuck-
results, it can also be used to predict productivity loss and, Burns SK, Colter HB (1987) Delayed posterior internal fixation
of unstable pelvic fractures. J Trauma 27(9):998–1006
thus, further related costs. 4. Cole JD, Blum DA, Ansel LJ (1996) Outcome after fixation of
Surprisingly, it appears that work return and leaves of unstable posterior pelvic ring injuries. Clin Orthop Relat Res
absence are not merely a signal of general decline in pre- 329:160–179
trauma health status but that the sudden exogenous trauma 5. Gruen GS, Leit ME, Gruen RJ, Garrison HG, Auble TE, Peitzman
AB (1995) Functional outcome of patients with unstable pelvic
generates productivity loss independently of pre-trauma ring fractures stabilized with open reduction and internal fixation.
health conditions. J Trauma 39(5):838–844; discussion 44–45
Based on our study, we found that work characteristics 6. Henderson RC (1989) The long-term results of nonoperatively
(i.e., private or public and sedentary or physically active treated major pelvic disruptions. J Orthop Trauma 3(1):41–47
7. Kabak S, Halici M, Tuncel M, Avsarogullari L, Baktir A, Basturk
employment) do not influence work resumption, the ability M (2003) Functional outcome of open reduction and internal
to maintain the same job tasks, or the number of days of fixation for completely unstable pelvic ring fractures (type C): a
sick leave. However, further studies with larger cohorts are report of 40 cases. J Orthop Trauma 17(8):555–562
needed to confirm these findings. 8. Keating JF, Werier J, Blachut P, Broekhuyse H, Meek RN,
O’Brien PJ (1999) Early fixation of the vertically unstable pelvis:
Our data did not show correlations between the variables the role of iliosacral screw fixation of the posterior lesion.
analyzed and return to the same job tasks; we think that the J Orthop Trauma 13(2):107–113
complexity of job types may act as a relevant confounding 9. Korovessis P, Baikousis A, Stamatakis M, Katonis P (2000)
factor. Medium- and long-term results of open reduction and internal
fixation for unstable pelvic ring fractures. Orthopedics
Work reintegration after pelvic fracture is a major issue 23(11):1165–1171
for the patient, health care facilities, and social systems: 10. Leung KS, Chien P, Shen WY, So WS (1992) Operative treat-
58 % of patients were not able to return to work or lost ment of unstable pelvic fractures. Injury 23(1):31–37
their jobs. Factors correlated with leaves of absence were 11. Miranda MA, Riemer BL, Butterfield SL, Burke CJ III (1996)
Pelvic ring injuries. A long term functional outcome study. Clin
injury severity, ICU admission, and the time elapsed Orthop Relat Res 329:152–159
between trauma and definitive fixation. 12. Monahan PR, Taylor RG (1975) Dislocation and fracture-dislo-
cation of the pelvis. Injury 6(4):325–333
Acknowledgments The authors thank all the AOCID (AO Clinical 13. Ragnarsson B, Olerud C, Olerud S (1993) Anterior square-plate
Investigation and Documentation) staff for their help in preparing the fixation of sacroiliac disruption. 2–8 years follow-up of 23 con-
manuscript. The corresponding author was supported by the AO secutive cases. Acta Orthop Scand 64(2):138–142
Foundation via an AO Trauma fellowship at AOCID. 14. Slatis P, Karaharju EO (1980) External fixation of unstable pelvic
fractures: experiences in 22 patients treated with a trapezoid
Compliance with ethical standards compression frame. Clin Orthop Relat Res 151:73–80
15. Tornetta P III, Matta JM (1996) Outcome of operatively treated
Conflict of interest The authors declare that they have no conflict unstable posterior pelvic ring disruptions. Clin Orthop Relat Res
of interest. 329:186–193
16. Van den Bosch EW, Van der Kleyn R, Hogervorst M, Van Vugt
Ethical standards The study conforms to the 1964 Helsinki dec- AB (1999) Functional outcome of internal fixation for pelvic ring
laration and its later amendments. The study was approved by the fractures. J Trauma 47(2):365–371
responsible Ethical Committee, and all the patients provided informed 17. Madhu TS, Raman R, Giannoudis PV (2007) Long-term outcome
consent before being enrolled. in patients with combined spinal and pelvic fractures. Injury
38(5):598–606. doi:10.1016/j.injury.2006.11.005
18. Suzuki T, Shindo M, Soma K, Minehara H, Nakamura K, Uchino
M et al (2007) Long-term functional outcome after unsta-
ble pelvic ring fracture. J Trauma 63(4):884–888. doi:10.1097/01.
ta.0000235888.90489.fc
19. Tile M (1988) Pelvic ring fractures: should they be fixed? J Bone
Jt Surg Br 70(1):1–12
20. Baker SP, O’Neill B, Haddon W Jr, Long WB (1974) The injury
severity score: a method for describing patients with multiple
injuries and evaluating emergency care. J Trauma 14(3):187–196
21. Saklad M (1941) Grading of patients for surgical procedures.
References Anesthesiology 2(3):281–284
22. Butcher NE, Balogh ZJ (2009) The definition of polytrauma: the
1. Papakostidis C, Kanakaris NK, Kontakis G, Giannoudis PV need for international consensus. Injury 40(Suppl. 4):S12–S22
(2009) Pelvic ring disruptions: treatment modalities and analysis 23. Gabbe BJ, Hofstee DJ, Esser M, Bucknill A, Russ MK, Cameron
of outcomes. Int Orthop 33(2):329–338. doi:10.1007/s00264-008- PA et al (2014) Functional and return to work outcomes fol-
0555-6 lowing major trauma involving severe pelvic ring fracture. ANZ J
2. MacKenzie EJ, Morris JA, Jurkovich GJ, Yasui Y, Cushing BM, Surg. doi:10.1111/ans.12700
Burgess AR et al (1998) Return to work following injury: the role 24. Katsoulis E, Giannoudis PV (2006) Impact of timing of pelvic
of economic, social, and job-related factors. Am J Public Health fixation on functional outcome. Injury 37(12):1133–1142. doi:10.
88(11):1630–1637 1016/j.injury.2006.07.017

2
Adjacent segment infection after surgical treatment
of spondylodiscitis
Ahmed Ezzat Siam2,3 • Hesham El Saghir1,2 • Heinrich Boehm1
Abstract Conclusions Adjacent segment infection after surgical

Background This is the first case series to describe treatment of spondylodiscitis is a rare complication
adjacent segment infection (ASI) after surgical treatment of (1.94 %). It is associated with multimorbidity and shows a
spondylodiscitis (SD). high mortality rate and a high neurological affection rate.
Materials and methods Patients with SD, spondylitis who Possible explanations are: haematomas of repeated micro-
were surgically treated between 1994 and 2012 were fractures around screw loosening, haematogenous spread,
included. Out of 1187 cases, 23 (1.94 %) returned to our direct inoculation or a combination of these factors. ASI
institution (Zentralklinik Bad Berka) with ASI: 10 males, may also lead to proximal junctional kyphosis, as found in
13 females, with a mean age of 65.1 years and a mean this series. We suggest early surgical intervention with
follow-up of 69 months. anterior debridement, reconstruction and fusion with pos-
Results ASI most commonly involved L3–4 (seven terior instrumentation, followed by antimicrobial therapy
patients), T12–L1 (five) and L2–3 (four). The mean inter- for 12 weeks.
val between operations of primary infection and ASI was Level of evidence Level IV retrospective uncontrolled
36.9 months. All cases needed surgical intervention, case series.
debridement, reconstruction and fusion with longer
instrumentation, with culture and sensitivity-based post- Keywords Adjacent segment infection
operative antimicrobial therapy. At last follow-up, six Spondylodiscitis Spondylitis Spinal infection Adjacent
patients (26.1 %) were mobilized in a wheelchair with a segment disease
varying degree of paraplegia (three had pre-existing
paralysis). Three patients died within 2 months after the
ASI operation (13 %). Excellent outcomes were achieved Introduction
in five patients, and good in eight.
Spondylodiscitis (SD) is a rare disease with incidence
varying globally from one per 100,000 to one per 250,000/
Parts of this study have been presented as an abstract in the EuroSpine year [1, 2]. In many patients, clinical and imaging findings
Congress 2013 in Liverpool, United Kingdom from October 2nd–4th. suggest the diagnosis before microbiological confirmation
is obtained, and a causative organism remains unknown in
& Ahmed Ezzat Siam up to 40 % of patients [2–4], causing greater difficulty for
[email protected]
physicians in selecting the most appropriate antimicrobial
1
Department of Spinal Surgery, Zentralklinik Bad Berka, treatment [5].
Bad Berka, Germany Although an elevation in C-reactive protein (CRP) and/
2
Spine Unit, El Hadara University Hospital, Alexandria, Egypt or erythrocytic sedimentation rate (ESR) should not be
3 taken as pathognomonic for an infection, both serve as
Department of Spinal Surgery with Scoliosis Centre, Schön
Klinik Vogtareuth, Krankenhausstrasse 20, screening and surveillance tests in the diagnosis and
83569 Vogtareuth, Germany treatment of spinal infections [6]. The high sensitivity,

Adjacent segment infection after surgical treatment of spondylodiscitis 7
specificity and accuracy of magnetic resonance imaging (ASA) score [10]. The ASA score is a subjective assess-
(MRI) make it the main imaging diagnostic tool in spinal ment of a patient’s overall health that is based on five
infection [7]. A clinical diagnosis of spondylitis can be classes:
made in patients with positive blood cultures and com-
1. Patient is completely healthy and fit.
patible clinical history in combination with corresponding
2. Patient has mild systemic disease.
changes on laboratory and imaging studies. A definitive
3. Patient has severe systemic disease that is not
diagnosis of spondylitis can only be made on microscopic
incapacitating.
or bacteriological examination and culture of infected tis-
4. Patient has incapacitating disease that is a constant
sues [8].
threat to life.
Pyogenic infection in the postoperative period is a well
5. A moribund patient who is not expected to live for
documented complication of spinal surgery. In this case,
24 h with or without surgery.
the infection occurs mainly in the operated spinal segment.
Adjacent segment infection (ASI) is a very uncommon Data collection, assessment of the radiological findings
complication [9]. and statistical analysis were performed by the first author
To the best of the authors’ knowledge, no previously (AES) and critically revised by the others. The adjacent
published study has described ASI after surgical treatment segment lordosis angle was measured between the end-
of SD. The current study aims to report and discuss this plates above and below on the lateral views of postopera-
rare phenomenon. tive radiographs and compared to those at the time of
presentation with ASI (Table 2).
A diagnosis of spondylodiscitis was made on clinical,
Materials and methods radiological and microbiological grounds, with patients
fulfilling the following criteria:
Study design
1. Clinical symptoms suggestive of spondylodiscitis
(back pain unrelieved by rest; radiating pain ± neuro-
Single-centre, multi-surgeon, retrospective study of clinical
logical deficits ± fever) with laboratory abnormalities:
and radiological outcome measures.
WBC, ESR and CRP levels.
2. Abnormal MRI (and other imaging modalities) fea-
Patients
tures compatible with infection of the spine.
3. Isolation of the causative microorganism or typical
The medical database of our institution (Zentralklinik Bad
histological pattern from percutaneous disk or epidural
Berka) was reviewed for patients with spinal infection who
abscess puncture or biopsy.
were surgically treated from 1994 to 2012. Patients with
ASI were included. Patients with same level recurrent ASI was defined similarly, with infection of the adjacent
infection were excluded, as well as patients with ASI after segment (vertebra or intervertebral body) after surgical
surgery for spinal pathologies other than SD. Data were treatment of the primarily treated segment(s).
collected regarding demographics, presenting signs and
symptoms, and predisposing and risk factors (Table 1). We Treatment
also collected information regarding the level(s) of spinal
involvement, perioperative inflammatory markers [white All patients underwent operative treatment primarily and
blood cell count (WBC), erythrocyte sedimentation rate secondarily. The surgical approach was either posterior,
(ESR), C-reactive protein (CRP)], microbiological exami- anterior or combined anterior and posterior, with debride-
nation (blood cultures, intra-operative biopsy) and imaging ment, fusion and longer instrumentation. Autologous bone
modalities. Routinely, plain radiographs in anteroposterior graft was harvested via a separate incision from the iliac
and lateral views, and magnetic resonance imaging (MRI) crest.
of the whole spine routinely T1- and T2-weighted with and Unless general health condition or intra-operative com-
without contrast medium were performed. Additionally, plications precluded it, all patients were mobilized with
computed tomography (CT) imaging was done in cases of assistance on the first postoperative day. Postoperative
marked bone destruction. This review also included the treatment included a culture-based antimicrobial therapy, or
management of this phenomenon as regards antimicrobial a broad-spectrum antimicrobial therapy when no organism
treatment and surgical intervention, as well as surgical was isolated. This was given for a mean of 12 weeks and was
data, complications and outcomes (Figs. 1, 2, 3, 4). stopped according to clinical, laboratory and radiological
The general condition of the patient was categorized findings of recovery. After ASI surgery, antimicrobial ther-
according to the American Society of Anesthesiologists apy was continued for at least 12 weeks in all patients.

8
Table 1 Demography, comorbidities, primary presentation, neurological function and FU outcomes

Patient Age Risk factors, previous ASA Main Duration of WBC ESR CRP Frankel Frankel FU FU after Final outcome
(years)/sex operations, co-morbidities score presentation symptoms (/mm3) (mm/h) (mg/ grade grade Frankel primary (Odom’s criteria)
and co-infections (months) dL) (primary) (ASI) grade operation
(months)
1 69.4/M MO, DM, BPH, HT, AA, renal 4 Fever 1.5 18.4 94 226.7 E E E 59.4 Good
stones, hypothyroidism. ASI:
fever, UTI, sepsis
2 68.8/F DM, HT, PNP, heel ulcers, UTI, 3 Weakness 2 weeks 8.5 65 42.2 C C C 45.7 Good
TKR
3 40.8/F MO 1 BP 1 9.8 90 94.8 E E E 182 Excellent
4 55.1/M Alcoholism, paraplegia sub-L2 2 Weakness 5 12.3 68 139.1 C C C 67.3 Fair
5 50.6/M DM, HT 3 Paraplegia 1.5 7.8 93 103.1 B B B 13.1 Died MOF, septic shock
(40 days)
6 70.8/F HT, breast cancer, OP, IHD 4 BP 5 9.7 84 229.7 E C B 16.5 Died MOF, pneumonia,
septicemia (16 days)
7 75.6/F Poliomyelitis, Parkinsonism, MO, 4 BP 5 9.1 17 5 C C C 32.7 Good
HT, DM, IHD, hypothyroidism
8 77.8/F Sigmoid colon cancer, depression, 4 BP 1.5 4.8 48 45.9 E E E 30.3 Died GIT bleeding
hypothyroidism, MO, OP, HT,
UTI
9 81.1/M PLS, re-infection 1 BP 1 10.8 100 172 E E E 93.8 Excellent
10 56/F Alcoholism, alcoholic liver 3 BP 1 week 10.3 48 67.4 E E E 4.8 Good
cirrhosis, chronic anaemia
11 47.6/M HT 2 Weakness 1 day 2.6 91 45.9 D C C 73.2 Died MOF, sepsis
MRSA (29 days)
12 75.9/F DM, IHD, OP 3 BP 3 weeks 10 110 267 E E E 95.8 Excellent
13 69.8/F Septicaemia, shoulder infection, 4 BP 2.5 20.9 135 155.8 E E E 38 Excellent
fever
14 57.7/F Hypothyroidism, hip abscess 2 Paraparesis 2 weeks 10.5 66 7.9 C E E 130.5 Excellent
15 73/M HT 2 BP 3.5 7.2 70 72.2 E D C 30 Fair
16 65.7/M Parkinsonism, UTI 2 BP 1 7 4 5 E E C 70.3 Fair
17 73.1/M Chronic respiratory insufficiency, 4 BP 1 5.8 140 121.8 C C B 10.2 Poor
HT, paraplegia, OP, IHD,
hemiparesis, vertebral fracture
18 55.7/F Bronchial asthma, re-infection 2 BP 4 8.1 31 11.2 E E E 87.5 Good
19 69.9/M DM, HT, cortison, RI, HI. ASI: 4 Weakness 6 days 6.5 69 74.8 C C D 46.4 Good
septicaemia, fever, toe
infection
20 80.2/F OP, HT, MVR, UTI, HBV, 3 BP 1 7.1 104 113 E D D 36.2 Good
cholecystectomy

Follow-up
syndrome, RI renal insufficiency, RA rheumatoid arthritis, BP back pain, WBC white blood cell count, ESR erythrocyte sedimentation rate, CRP C-reactive protein, FU follow-up, MOF
M male, F female, ASA American Society of Anaesthesiology (general condition), MO morbid obesity, DM diabetes mellitus, BPH benign prostatic hyperplesia, HT hypertension, AA aortic
aneurysm, ASI adjacent segment infection, UTI urinary tract infection, PNP polyneuropathy, TKR total knee replacement, OP osteoporosis, IHD ischaemic heart disease, PLS postlaminectomy
Died MOF (after long
intensive care and
Preoperative, postoperative and last follow-up (FU) neu-
multiple wound
(Odom’s criteria)
rological findings were assessed according to Frankel’s
Final outcome
infetions)
classification: [11].
Good
1. ‘Complete’ (A). Paralysis, both motor and sensory,
Fair
below the level marked.
2. ‘Sensory only’ (B). Some sensation present below the
operation
(months)
FU after
primary
level of the lesion but motor paralysis complete below

210.3
156
58
that level.
3. ‘Motor Useless’ (C). Some motor power present below
Frankel
grade
the lesion but of no practical use to the patient.

FU
E
D
4. ‘Motor Useful’ (D). Useful motor power below the

level of the lesion.
Frankel
(ASI)
grade
5. ‘Recovery’ (E). Free of neurological symptoms.

D
E
The final functional outcome was completed by ques-

(primary)
Frankel
tionnaires including Odom’s criteria [12] which catego-

grade
rized patients’ satisfaction into four grades: excellent,

E
good, fair and poor.

10.1
(mg/
CRP
138
31
dL)
– Excellent: all preoperative symptoms relieved, abnor-

mal findings unchanged or improved.
(mm/h)
– Good: minimum residual of preoperative symptoms not

ESR
140
88
35
requiring medication or limiting activity, and abnormal

findings unchanged or improved.
(/mm3)
WBC
10.9
21.7
6.3
– Fair: definite relief of some preoperative symptoms

with others remaining unchanged or only slightly
Duration of
improved.
symptoms
(months)
– Poor: symptoms and signs unchanged from preopera-

1 week
tive status or worse.

4
6
presentation
Statistical analysis
Main
multiorgan failure, MRSA methicillin-resistant Staphylococcus aureus

BP
BP
BP
Descriptive statistics were used. Quantitative variables

(e.g. age, laboratory values, operative data, interval
score
ASA
between infections) were summarized by mean value and

3
the standard deviation if appropriate. Qualitative demo-

cholecystectomy, cortisone. ASI:
graphic variables (e.g. gender and disease characteristics as

DM, RI, HT, bilateral TKR,
well as potential prognostic factors) were summarized by

ovarian cancer operation,
operations, co-morbidities
IHD, MO, HT, RI. ASI:
counts and percentages. Analytical statistics were used to

Risk factors, previous
compare the preoperative and postoperative values as

septicaemia, UTI
RA, foot infection,
and co-infections
regards the laboratory findings. Because of the small

chemotherapy
septicaemia
number of cases, non-parametric tests were used, in this

case the Wilcoxon signed-rank test. The same test was used
to analyse the difference between the lordosis angle of the
adjacent segment after primary surgery and at the time of
presentation with ASI. To analyse the possible correlation
(years)/sex
Table 1 continued
between different variables and the outcomes according to

57.7/M
65.1/F
59.6/F
Age
Odom’s criteria, the Kruskal–Wallis test was used. Statis-

tical significance was defined as p \ 0.05. The statistical
Patient
analysis was performed using SPSS version 13.0 (SPSS

21
22
23
Inc., Chicago, IL, USA).

Fig. 1 Patient 4: sagittal MRI cuts T2- and T1-weighted, and lateral radiographs; a preoperatively, b after primary operation, c adjacent segment
infection, screw loosening and marked adjacent segment kyphosis and d last FU after 5.5 years
Results in six (26.1 %), ASA 3 in eight (34.8 %) and ASA 4 in

seven (30.4 %). This distribution reflects the generally bad
Demography condition of these patients (Table 1).
Between 1994 and 2012, 1187 patients were surgically Clinical presentation
treated in Zentralklinik Bad Berka because of SD. Out of
these, 23 (10 males, 13 females) returned with ASI At the time of primary infection, the main symptoms were
(1.94 %), with a mean age 65.1 ± 10.9 years. The primary back pain in 16 patients (69.6 %) and neurological dete-
infection was lumbar in 13 (56.5 %), thoracolumbar in four rioration in six (26.1 %). The average period of conser-
(17.4 %), thoracic in three (13 %), cervical in one (4.3 %) vative treatment was 2.17 months before surgery
and combined thoracic and lumbar in two cases (8.7 %). (Table 2).
Single-level infection was found in 16 patients (65.6 %), At the time of treatment of ASI, patients presented most
double-level in four (17.6 %) and three levels in three commonly with recurrence of severe back pain (15 cases,
(13 %). Comorbidities were found in 19 patients (82.6 %); 65.2 %). The mean interval between the operation of pri-
most commonly hypertension (HT) (12 patients, 52.2 %), mary infection and the operation of ASI was 36.88 months.
diabetes mellitus (DM) (7, 30.4 %), osteoporosis (5, Neurologically, one patient had Frankel grade B para-
21.7 %) and ischemic heart disease (IHD) (5, 21.7 %) plegia (4.3 %), six patients had paraparesis grade C
(Table 1). The general condition of the patients before the (26.1 %), one had grade D (4.3 %), and 15 patients
primary surgery was ASA 1 in two patients (8.7 %), ASA 2 (65.2 %) were neurologically free (grade E).

Fig. 2 Patient 18: a preoperative MRI and radiographs, b postoperative, c ASI in MRI and radiographs and d after reoperation
Fig. 3 Patient 8: sagittal MRI cuts T2- and T1-weighted, and lateral radiographs; a preoperatively, b after primary operation, c adjacent segment
infection, no screw loosening or marked adjacent segment kyphosis, and d last FU after 2.5 years

Fig. 4 Patient 12: sagittal MRI cuts T2- and T1-weighted, and lateral radiographs; a preoperatively, b after primary operation, c adjacent
segment infection, and d after ASI surgery
Laboratory findings (82.6 %), bi-segmental in seven cases (30.4 %) and multi-
segmental in one case (4.3 %).
The mean preoperative laboratory values were WBC Multifocal non-contiguous spinal infection was diag-
9830 ± 4743/mm3, ESR 77.8 ± 36.7 mm/h and CRP nosed in four patients (17.4 %); two cervical and two
94.8 ± 77.2 mg/dL. The difference in relation to the thoracic spinal infections coincided with lumbar infection.
immediate postoperative values was not statistically sig- An epidural abscess was found in four patients (17.4 %)
nificant (p = 0.813, 0.465 and 0.594, respectively). At and psoas abscess in seven (30.4 %).
readmission with ASI, the mean values were WBC
10,496 ± 6697/mm3, ESR 79.8 ± 37 mm/h and CRP Primary surgery
94 ± 83 mg/dL. Postoperative values did not differ sig-
nificantly after ASI operation (p = 0.859, 0.345 and 0.889, The mean operative time was 217 ± 69.5 min with a mean
respectively). blood loss of 1223 ± 710 ml. Mono- and bi-segmental
spinal fusions were done in seven (30.4 %) and eight
Diagnostic imaging (34.8 %) patients, respectively. Three-, four- and five-
segment fusions were performed in five patients (21.7 %),
The most common primarily involved levels were L3–4 two (8.7 %) and one (4.3 %), respectively. Interbody
(seven, 30.4 %), L4–5 (seven, 30.4 %) and L2–3 (five, fusion was done in 16 patients (69.6 %), while corpectomy
21.7 %). ASI most commonly involved L3–4 (seven, was done in seven patients (30.4 %). Eight patients had
30.4 %), T12–L1 (five, 21.7 %) and L2–3 (four, 17.4 %). bone graft only (34.8 %), and 15 had bone graft and cage
ASI involved cranial segment in ten patients (43.5 %), (65.2 %). Minimally invasive techniques (e.g. video-as-
caudal segment in ten (43.5 %), floating segment in two sisted thoracoscopic surgery and percutaneous instrumen-
(8.7 %) and adjacent segments cranially and caudally in tation) were used in six patients (26.1 %), while an open
one case (4.3 %), mono-segmental affection in 19 cases technique was used in 17 patients (73.9 %) (Table 3).

Table 2 Patient data at admission and possible explanations of adjacent segment infection
Patient Interval ASI Presentation WBC ESR CRP Septicaemia Screw Change
before level(s) (ASI) (/mm3) (mm/h) (mg/dL) loosening in ASK
ASI surgery
(months)
1 5.3 L3–4 Fever 12.7 54 86.2 Yes (Escherichia coli) No –

2 19.6 L3–4, L5–S1 BP 11.7 132 93.7 No Yes 15°
(?cage)
3 11.9 L3–4 BP 10 83 99 No No –
4 14.4 L2–3 Decubitus ulcer 7.5 57 38.2 No Yes 28°
5 6 T11–12 BP 10.2 73 240.7 Yes (MRSA) Yes 26°
6 8.6 T12–L1 BP 7.1 94 71.4 Yes (Pseudomonas aeruginosa Yes 17°
MDR)
7 16.8 L3–4 BP 5.1 17 3 No No –
8 12.3 L2–3 BP 9.5 61 55.4 No No –
9 61.1 T12–L1 BP 10.3 84 203 No Yes –
10 14 days L3–4 BP 14.3 85 84.4 No No –
11 72.3 L1–L2 Weakness 2.4 81 62.7 Yes (Staphylococcus aureus) Yes –
12 86.2 L4–5 BP 5.7 97 38.9 No No –
13 27.4 T12–L1 BP 7 84 45.5 Yes (Staphylococcus aureus) No –
14 6 L1–2 BP 14.3 140 180 No Yes –
15 17.4 L1–2 Cauda equina 8.9 90 44.8 No Yes –
syndrome
16 39 T11–L1 BP 6 5 1.9 No No –
17 4.5 L3–4 Wound infection 10.2 140 274.4 No Yes 10°
18 46 L2–3 BP 5.2 37 7.7 No No –
19 25.9 T1–2 Fever 23.4 105 270.2 Yes (Staphylococcus aureus) No –
20 9.9 L2–3 BP 34 47 63.5 No Yes –
21 4.4 T2–3 Fever 5.1 140 102.3 Yes (Pseudomonas aeruginosa Yes –
MDR)
22 205.6 L3–4 GIT infection 11 84 89 Yes (Escherichia coli) Yes 15°
23 147.3 L5–S1 BP 9.8 45 6.4 No Yes –
ASI adjacent segment infection, BP back pain, WBC white blood cell count, ESR erythrocyte sedimentation rate, CRP C-reactive protein, MRSA
methicillin-resistant Staphylococcus aureus, MDR multi-drug resistant, GIT gastrointestinal tract, ASK adjacent segment kyphosis
Postoperative treatment fusions were done in six patients (26.1 %), while the
majority of patients had long-segment fusions; five seg-
A broad-spectrum antimicrobial was started on the same ments in five patients, and more than five segments in five
day after surgery (or continued), and was shifted according patients (21.7 %). Interbody fusion was done in 21 patients
to culture and sensitivity tests. The most common causative (91.3 %), while corpectomy was done in only two patients
organism identified (in primary SD) was Staphylococcus (8.7 %). Ten patients had bone graft and cages (43.5 %),
aureus, in five patients (21.7 %). In 8 patients, no organism and 13 had bone graft only (56.5 %). Minimally invasive
could be isolated (34.8 %) (Table 3). techniques were used in eight patients (34.8 %) and open
technique in 15 patients (65.2 %) (Table 3). Of 11 patients
Surgery for adjacent segment infection with positive microbiological findings, eight (72.7 %) had
a recurrence of the same micro-organism with multiple
The mean operative time was 215 ± 106.8 min with a antimicrobial drug resistance and three (27.3 %) had a
mean blood loss of 1241 ± 587.1 ml. Bi-segmental spinal superadded infection with another organism.

Table 3 Operative data of

Primary surgery Surgery for ASI
spondylodiscitis and adjacent
segment infection surgical Operative time (min) 217 ± 69.5 215 ± 106.8
treatment
Blood loss (ml) 1223 ± 710 1241 ± 587.1
Mono-segmental operation 7 (30.4 %) 0
Bi-segmental 8 (34.8 %) 6 (26.1 %)
Three segments 5 (21.7 %) 4 (17.4 %)
Four segments 2 (8.7 %) 3 (13 %)
Five segments 1 (4.3 %) 5 (21.7 %)
[Five segments 0 5 (21.7 %)
One setting operation 22 (95.7 %) 20 (87 %)
Two settings 1 (4.3 %) 3 (13 %)
Posterior approach only 1 (4.3 %) 5 (21.7 %)
Anterior approach only 1 (cervical) (4.3 %) 0
Anterior and posterior approaches 21 (91.3 %) 18 (78.3 %)
Bone graft only 8 (34.8 %) 13 (56.5 %)
Bone graft and cage 15 (65.2 %) 10 (43.5 %)
Interbody fusion 16 (69.6 %) 21 (91.3 %)
Corpectomy and fusion 7 (30.4 %) 2 (8.7 %)
Open technique 17 (73.9 %) 15 (65.2 %)
Minimally invasive technique 6 (26.1 %) 8 (34.8 %)
Microorganism
Staphylococcus aureus 5 (21.7 %) 4 (17.4 %)
Staphylococcus epidermidis 3 (13 %) 3 (13 %)
Pseudomonas aeruginosa 4 (17.4 %) 3 (13 %)
Enterococcus faecalis 2 (8.7 %) 0
Escherechia coli 1 (4.3 %) 1 (4.3 %)
No organism 8 (34.8 %) 12 (52.2 %)
Radiological results neurological status or other variables, as statistically anal-

ysed in Table 4.
Marked increase in adjacent segment kyphosis ([10°)
occurred in six patients (26.1 %) and screw loosening was
identified in 13 patients (56.5 %) at the time of presenta- Discussion
tion with ASI (Table 2).
To the best of the authors’ knowledge, no previously
Functional outcome published study has described the prevalence of ASI after
surgical fusion in SD. We conducted a PubMed/Medline
At the later presentation with ASI, two patients had dete- search and review of the available literature up to
riorated to grade C (8.7 %) and three had weakness grade December 2014. This phenomenon has been described only
D (13 %). The mean FU period was 69 ± 55.13 months in three case reports: one lumbar in Germany [13] and two
after primary surgery. At last FU, six patients (26.1 %) cervical in India [9, 14]. This study presents the first case
were mobilized in a wheelchair with a varying degree of series of ASI in the literature. The main limitations of this
paraplegia (three had pre-existing paralysis). The others did study are the variable treatment options, the wide variation
not have neurological changes during the FU period in FU period and the retrospective design of the study with
(Table 1). Three patients died within 2 months after ASI a low level of evidence.
operation because of sepsis and/or multi-organ failure Because of long conservative treatment of SD, no causa-
(13 %). Subjectively, out of 18 surviving patients at the tive organism could be isolated in many patients previ-
time of this study, an excellent outcome was achieved in ously confirmed to have this disease. The surgical approach
five (27.8 %), good in eight (44.4 %), fair in four (22.2 %) with radical debridement, posterior stabilization and recon-
and poor in one patient (5.6 %). This outcome was not struction of anterior column using expandable titanium cages
significantly related to age, sex, region affected, is a widespread and accepted method [13]. Korovessis et al.

Table 4 Statistical analysis of

Outcome (Odom’s criteria) Excellent Good Fair Poor Died Total p value
different factors as regards the
final functional outcome Sex
according to Odom’s criteria
Male 1 3 3 1 2 10 (43 %) 0.406
Female 4 5 1 0 3 13 (57 %)
Involvement of segments
Monosegmental 5 5 4 1 2 17 (74 %) 0.171
Bisegmental 0 2 0 0 3 5 (22 %)
Multisegmental 0 1 0 0 0 1 (4 %)
General condition (ASA score)
ASA 1 (best) 2 0 0 0 0 2 (9 %) 0.133
ASA 2 1 1 3 0 1 6 (26 %)
ASA 3 1 4 1 0 2 8 (35 %)
ASA 4 (worst) 1 3 0 1 2 7 (30 %)
Preoperative neurology (Frankel grade)
B (paralysis) 0 0 0 0 1 1 (4 %) 0.66
C 1 3 1 1 0 6 (26 %)
D 0 0 0 0 1 1 (4 %)
E (normal) 4 5 3 0 3 15 (65 %)
Region of primary infection
Cervical 0 1 0 0 0 1 (4 %) 0.793
Thoracic 1 1 0 0 1 3 (13 %)
Thoracolumbar 1 1 1 1 0 4 (17 %)
Lumbar 3 5 3 0 4 15 (65 %)
Other factors
Septicemia 1 3 0 0 4 8 (35 %) 0.119
Screw loosening 2 3 3 1 4 13 (57 %) 0.406
ASK ([10°) 0 1 1 1 3 6 (26 %) 0.092
None of these factors significantly affected the outcome of the disease
ASA American Society of Anesthesiologists, ASK adjacent segment kyphosis
have even shown that the use of titanium mesh cages may further 12 weeks, depending on the causative organism and
have a beneficial influence on eradication of infection and culture and sensitivity examinations.
fusion [15]. In addition, Ruf et al. did not find any association From the authors’ research into an explanation of this
between titanium cages and persistence or recurrence of phenomenon, the following hypotheses are presented:
infection [16]. The goals of surgical intervention are to pre-
1. Haematogenous infection route: prolonged preopera-
serve neurological function and to facilitate stable bony
tive and postoperative antimicrobial treatment should
fusion without severe kyphosis. Procedures range from
have minimized the risk of re-infection via this route
decompression, debridement and drainage to interbody fusion
[13]. In this study, eight patients (34.8 %) had positive
and grafting, and are decided on a case-by-case basis [2].
blood cultures within the FU time.
Based on the current study, we suggest early surgical
2. Direct infection of adjacent segment by intra-opera-
intervention because of the higher incidence of multi-drug-
tively contaminated screws. This was suspected by
resistant micro-organisms and before extension of bone
Lange et al., using cannulated screws [13]. No cannu-
destruction and expected deterioration of general condition
lated screws were used in our series, which opposes the
and neurological functions of the patient. The usual sur-
hypothesis that bacteria are being shielded from antibi-
gical treatment consists of anterior debridement, recon-
otic treatment within the cannulation of screws. We still
struction and fusion combined with open or percutaneous
suggest that direct contamination during surgery by
posterior instrumentation. This allows adequate eradication
faulty drilling or by cranially located screws may have a
of the septic focus, resistance-adjusted antimicrobial ther-
role in ASI, as also suggested by Kulkarni and Hee [14].
apy and early mobilization. Postoperative antimicrobial
Seven patients had ASI with the same infecting
therapy should be immediately started (or continued) for a

organism as primary SD; four of these organisms had 2. Butler JS, Shelly MJ, Timlin M, Powderly WG, O’Byrne JM
acquired more antimicrobial resistance. (2006) Non-tuberculous pyogenic spinal infection in adults: a
12-year experience from a tertiary referral centre. Spine
3. Screw loosening is a very important finding in 13 31:2695–2700
patients (56.5 %) in this series. We assume that slowly 3. Carragee EJ (1997) Pyogenic vertebral osteomyelitis. J Bone
progressing loosening of screws causes repeated Joint Surg Am 79:874–880
micro-fractures in pedicles and endplates. These 4. Hadjipavlou AG, Mader JT, Necessary JT, Muffoletto AJ (2000)
Haematogenous pyogenic spinal infections and their surgical
micro-fractures lead to small haematomas in bone management. Spine 25:1668–1679
tissue in pedicles, endplates and most importantly in 5. D’Agostino C, Scorzolini L, Massetti AP, Carnevalini M, d’Et-
the endplate–disc attachment. Subsequent infection of torre G, Venditti M, Vullo V, Orsi GB (2010) A seven-year
these haematomas in previously infected and operated prospective study on spondylodiscitis: epidemiological and
microbiological features. Infection 38:102–107
spinal region is likely to occur, especially in multi- 6. Schinkel C, Gottwald M, Andress HJ (2003) Surgical treatment
morbid patients with poor general condition. of spondylodiscitis. Surg Inf 4(4):387–391
4. Proximal junctional kyphosis (PJK) is a well-known 7. An HS, Seldomridge A, Spinal infections (2006) Diagnostic tests
complication of spinal instrumentation, especially with and imaging studies. Clin Orthop Relat Res 444:27–33
8. Cheung WY, Luk KD (2012) Pyogenic spondylitis. Int Orthop
long-segment fusions [17, 18]. We assume that the 36(2):397–404
adjacent segment infection is a direct cause of many 9. Basu S, Sreeramalingam R (2012) Adjacent level spondy-
cases of PJK. In the current study, six patients had lodiscitis after anterior cervical decompression and fusion. Indian
marked increase in kyphosis prior to ASI. From the J Orthop 46(3):360–363
10. Dripps RD (1963) New classification of physical status. Anes-
authors’ point of view, in cases of junctional kyphosis, thesiol 24:111
ASI should be suspected and intra-operative biopsies 11. Mannion AF, Elfering A, Staerkle R et al (2005) Outcome
should be sent to histopathology and microbiology for assessment in low back pain: how low can you go? Eur Spine J
exclusion of low-grade or subclinical infection. 14(10):1014–1026
12. Odom GL, Finney W, Woodhall B (1958) Cervical disk lesions.
J Am Med Assoc 166:23–28
Acknowledgments Special thanks go to the team of the hospital 13. Lange T, Schulte TL, Bullmann V (2010) Two recurrences of
library in Zentralklinik Bad Berka, Mrs. Zuellinger and Mrs. Stahl, adjacent spondylodiscitis after initial surgical intervention with
for their invaluable efforts in collection of related scientific articles. posterior stabilization, debridement, and reconstruction of the
anterior column in a patient with spondylodiscitis: a case report.
Compliance with ethical standards Spine 35(16):E804–E810
14. Kulkarni AG, Hee HT (2006) Adjacent level discitis after anterior
Conflict of interest All authors declare that they do not have any cervical discectomy and fusion (ACDF): a case report. Eur Spine
potential conflict of interest. J 15(5):559–563
15. Korovessis P, Repantis T, Iliopoulos P et al (2008) Beneficial
Ethical standards This study conforms to the 1964 Helsinki dec- influence of titanium mesh cage on infection healing and spinal
laration and its later amendments. It was approved by the responsible reconstruction in hematogenous septic spondylitis: a retrospective
Ethics Committee and because of its retrospective design an informed analysis of surgical outcome of twenty-five consecutive cases and
consent was not required. review of literature. Spine 33:E759–E767
16. Ruf M, Stoltze D, Merk HR, Ames M, Harms J (2007) Treatment
of vertebral osteomyelitis by radical debridement and stabiliza-
tion using titanium mesh cages. Spine 32(9):E275–E280
17. Ahn DK, Park HS, Choi DJ, Kim KS, Yang SJ (2010) Survival
and prognostic analysis of adjacent segments after spinal fusion.
Clin Orthop Surg 2(3):140–147
18. Kim HJ, Lenke LG, Shaffrey CI, Van Alstyne EM, Skelly AC
(2012) Proximal junctional kyphosis as a distinct form of adjacent
segment pathology after spinal deformity surgery: a systematic
review. Spine 37(22 Suppl):S144–S164
References
1. Acosta FL Jr, Chin CT, Quiñones H, Ames CP, Weinstein PR,

Chou D (2004) Diagnosis and management of adult pyogenic
osteomyelitis of the cervical spine. Neurosurg Focus 17:E2

3
Single intra-articular injection of high molecular weight
hyaluronic acid for hip osteoarthritis
Fabrizio Rivera1
Abstract T0 ? 12 months); after a significant change between T0

Background Intra-articular (IA) injection of hyaluronic and T0 ? 3 months, the score remained stable. The evo-
acid (HA) into the hip joint appears to be safe and well lution of the pain VAS showed a statistically highly sig-
tolerated but only a small number of randomized clinical nificant improvement (p \ 0.001) between T0 and
trials in humans has been published. The objective of this T0 ? 3 months; thereafter it remained stable from the first
prospective study was to evaluate the efficacy and safety of quarter post-intervention. No serious adverse event was
a single IA injection of high-molecular-weight (2800 kDa) noted; 12 cases (0.5 %) of pain associated with transient
HA (Coxarthrum) for hip osteoarthritis. synovitis are noteworthy.
Materials and methods All patients received a single IA Conclusion This study shows that a single IA injection of
administration of 2.5 % sodium hyaluronate (75 mg/3 mL) Coxarthrum is effective from the third month and that the
of high molecular weight. Fluoroscopy requires an iodized results are stable or continue to improve up to 1 year.
contrast medium (iopamidol, 1 ml) which highlights the Level of evidence IV.
capsule before administering HA. Patients were evaluated
before IA injection (T0), after 3 months, after 6 months Keywords Hip Viscosupplementation Hyaluronic acid
and after 1 year from injection. Results were evaluated by
the Brief Pain Inventory (BPI II), Harris Hip Score and a
visual analog scale of pain (pain VAS). All treated patients Introduction
were considered for statistical analysis.
Results Two hundred seven patients were included at T0. Osteoarthritis has a very high prevalence globally. It is a
The mean age was 67 years (range 46–81). Regarding BPI source of pain and deterioration of joint function, with
severity score, changes in pain between T0 and the three important socioeconomic consequences. The related pain is
following visits were statistically highly significant poorly evaluated by doctors, who underestimate its inten-
(p \ 0.001). Changes in pain score compared to the pre- sity when the pain is reported by the patient as being high,
vious visit were statistically significant for the worst pain in and overestimate it when it is reported as being weak [1].
the second quarter post-intervention (p = 0.037) and for The incidence of hip arthritis is increasing with age and is
mean pain in the second semester post-intervention estimated at between 47.3 (95 % confidence interval [CI]
(p = 0.043) The evolution of the Harris Hip Score was 27.8–66.8) [1] and 88/100 000 patient-years (95 % CI
statistically highly significant (p \ 0.001) between T0 and 65–101) [2]. Hip pain is reported by 19.2 % (95 % CI
the following visits (T0 ? 3 months, T0 ? 6 months and 17.9–20.6) of people aged 65 years and older. Less than
half (48 %) of the symptomatic respondents had unilateral
problems affecting one hip or knee joint only [1].
& Fabrizio Rivera Optimal management of osteoarthritis requires a com-
[email protected]
bination of non-pharmacological and pharmacological
1
Department of Orthopedic Trauma, SS Annunziata Hospital, modalities. Among the latter are injections of intra-articu-
Via Ospedali 14, 12038 Savigliano (CN), Italy lar (IA) hyaluronic acid (HA), first isolated in 1934 by Karl

Meyer in the vitreous humor; the first human clinical use of good or full joint mobility, and hip disease persisting for at
IA HA in the treatment of knee arthritis was made in 1975, least 3 months. Patients were excluded from the study
and the first trials date back to 1980 [1]. HA is a where they had severe arthritis for which it was no longer
polysaccharide macromolecule, a glycosaminoglycan of possible to recognize radiographic joint space (Kellgren–
high molecular weight (MW) composed of repetitions of Lawrence stage 4), had inflammatory, autoimmune and
disaccharides of glucuronic acid and N-acetylglucosamine; septic disease (rheumatoid arthritis, connective tissue dis-
it is a constituent of synovial fluid in normal and osteoar- ease, osteomyelitis), or had surgical indication for hip
thritic joints and is synthesized by chondrocytes and syn- arthroplasty. All patients received a single injection of a
oviocytes [2]. HA has complex biological properties that single administration of 2.5 % sodium hyaluronate (75 mg/
could explain its analgesic effects (anti-inflammatory by 3 mL) of high MW (2800 kDa) (Coxarthrum). This is a
inhibiting the formation and release of prostaglandin, sterile, viscoelastic, transparent, homogeneous preparation
immunomodulatory in situ), irrespective of its mechanical composed of purified HA, without any avian protein; it is
action on the joint fluid. The concentration of HA in an not cross-linked by a chemical agent, which limits as much
arthritic joint has been found to decrease to 50–33 % of as possible the risks of allergic and cytotoxic reactions.
normal levels, and includes a reduction in molecular size. Injections were performed by fluoroscopic guidance. Flu-
Molecular interaction has also been observed, with a con- oroscopy requires an iodized contrast medium (iopamidol,
sequent decrease in elasticity and viscosity of the synovial 1 ml) which highlights the capsule before administering
fluid [3]. HA. Patients were evaluated before IA injection (T0), after
HA may be useful in patients with knee or hip 3 months, after 6 months and after 1 year from injection.
osteoarthritis. The symptomatic benefit is delayed in The first endpoint was the score on the Modified Brief Pain
comparison with that of intra-articular injections of corti- Inventory (BPI II) comprising (1) a score of pain severity
costeroids, but it is prolonged. The IA injections of HA are (BPI severity score) rated between zero and 10 and mea-
widely used and recommended in existing guidelines as a suring the pain which the subjects felt before the present
useful therapeutic modality to treat patients with knee visit (the worst pain, the lightest pain, the mean pain) and
osteoarthritis; there is less experimental evidence of effi- the pain now, that is to say the pain felt during the visit; (2)
cacy for hip arthritis than for knee arthritis [4]. an impact score (BPI interference score) rated between
IA injection of HA into the hip joint appears to be safe zero and 10, describing disturbances of social life (work,
and well tolerated [5] but only a small number of ran- sleep and mood); (3) an overall impact score adding the
domized clinical trials in humans has been published [6–9]. previous score and four other items (activities in general,
Data from a meta-analysis in knee arthritis suggested ability to walk, relationships with others, the enjoyment of
that the heterogeneity between trials might be due to the life). Another criterion of evaluation was the Harris Hip
higher MW products having greater efficacy. Indeed, HA Score whose range is from zero to 100 with points dis-
preparations may broadly be classified according to their tributed within four areas: ‘‘pain’’ domain, maximum 44
MW and formulation type: solutions of low MW points; ‘‘function’’ domain, maximum 47 points; ‘‘range of
(500–1200 kDa), solutions of high MW (6000 kDa), cross- motion’’ domain, maximum 5 points; ‘‘no deformity’’
linked HA and solutions of non-animal stabilized HA domain, maximum 4 points. Finally, a visual analog scale
(NASHA) [10–12]. (VAS) of pain (pain VAS), scored from zero to 10, also
As a consequence, the objective of this prospective allowed judging the effectiveness of IA HA.
study was to evaluate the efficacy and safety of IA injection All treated patients were considered for statistical anal-
of a single dose of high MW (2800 kDa) HA (75 mg/ ysis, which was performed in SASÒ software (version 9.2).
3 mL) (Coxarthrum, LCA Pharmaceutical, Chartres, At each study time (T0, T0 ? 3 months, T0 ? 6 months
France) for hip osteoarthritis. and T0 ? 12 months), mean, standard deviation and median
endpoints were calculated. For the same endpoint, compar-
isons were made at different study times using Student’s
Materials and methods t test for paired samples. Results were considered statisti-
cally significant for values of p \ 0.05.
The study protocol was approved according to the modal-
ities planned by Ethical Committee. We conducted a sin-
gle-center, prospective, unblinded study. After baseline Results
(T0), patients were to be reviewed at 3, 6 and 12 months.
Inclusion criteria were age more than 40 years, mono- or Two hundred seven patients were included at T0. One
bilateral hip arthritis with X-ray proof of at least partially hundred twenty-six were women (61 %) and eighty-one
preserved joint space (Kellgren–Lawrence stage 2–3 [10]), were men (49 %). The mean age was 67 years (range

Single intra-articular injection of high molecular weight hyaluronic acid for hip osteoarthritis 19
Table 1 Number of patients completing the questionnaires at the various visits

No. of patients available at
T0 T0 ? 3 months T0 ? 6 months T0 ? 12 months
BPI II questionnaire (severity and interference) 207 207 150 121

Harris Hip Score 207 207 150 121
VAS of pain 165 176 128 104
Table 2 Evolution of patients’ pain (BPI II severity score)

Mean pain rating (out of 10) ± SD
Before and Between T0 and Between T0 ? 3 months Between T0 ? 6 months and
at T0 T0 ? 3 months and T0 ? 6 months T0 ? 12 months
Worst pain 6.03 (1.51) 4.78 (1.95)* 4.90 (2.16)*, 4.80 (2.00)*
Slightest pain 3.80 (1.92) 2.91 (1.68)* 2.52 (1.61)* 2.42 (1.43)*
Mean pain 4.93 (1.49) 3.78 (1.64)* 3.42 (1.68)* 3.22 (1.57)*,
Pain during visit 4.07 (2.04) 3.00 (1.94)* 2.73 (1.98)* 2.55 (1.63)*
* Statistically highly significant (p \ 0.001) compared with T0

Statistically significant (p \ 0.05) compared with the previous visit
46–81). Mean body mass index (kg/m2) was 22.8 (range

18.8–29.9). Radiological evaluation of osteoarthritis
showed a Kellgren–Lawrence stage 2 in 83 (40.1 %)
patients and a Kellgren–Lawrence stage 3 in 124 (59.9 %)
patients. The number of included patients who completed
the questionnaires in the various planned visits gradually
decreased over time; however, data from three-quarters of
patients were still available after 6 months, and data from
over half the patients after 1 year. The data of pain VAS
were less available than those of the BPI II questionnaires
and the Harris Hip Score (Table 1).
Regarding the BPI severity score, changes in pain Fig. 1 Evolution of BPI Severity Score during follow-up
between T0 and the three following visits were statistically
highly significant (p \ 0.001). Changes in pain score
compared to the previous visit were statistically significant semester post-intervention for mood. It was significant
for the worst pain in the second quarter post-intervention (p \ 0.05) concerning professional activities for the second
(p = 0.037) and for mean pain in the second semester post- semester post-intervention (Table 3).
intervention (p = 0.043) (Table 2). Changes in pain The evolution of the BPI interference overall score
severity (BPI severity score) are shown in Fig. 1. Note the between T0 and the three following visits was statistically
parallelism of the curves, although the intensity of the highly significant (p \ 0.001). The evolution of the BPI
worst pain is virtually unchanged from T0 ? 3 months interference overall score measured against the previous
onwards. visit was statistically significant in the second quarter post-
The evolution of the BPI interference score, describing intervention (p \ 0.01) and statistically highly significant
disturbances of social life, measured between T0 and the during the second semester post-intervention (p \ 0.001)
three following visits, was statistically highly significant (Table 4).
(p \ 0.001) for the three items describing disturbances of The evolution of the Harris Hip Score was statistically
social life (work, sleep and mood). The evolution of the highly significant (p \ 0.001) between T0 and the fol-
BPI interference score measured against the previous visit lowing visits (T0 ? 3 months, T0 ? 6 months and
was also highly significant (p \ 0.001) from the second T0 ? 12 months); after a significant change between T0
quarter post-intervention for sleep and for the second and T0 ? 3 months, the score remained stable (Fig. 2).

Table 3 Evolution of the BPI interference score describing disturbances of social life
Mean rating (out of 10) ± SD
Before and Between T0 and Between T0 ? 3 months Between T0 ? 6 months
at T0 T0 ? 3 months and T0 ? 6 months and T0 ? 12 months
Professional activities 4.44 (2.15) 3.42 (1.94)* 3.26 (2.45)* 2.59 (1.97)*,
,H
Sleep 3.80 (2.23) 2.59 (1.76)* 2.01 (1.78)* 1.31 (1.17)*,H
Mood 4.11 (2.18) 3.14 (2.12)* 2.79 (2.45)* 1.83 (1.61)*,H

Statistically significant (p \ 0.05) compared with previous visit
H
Statistically highly significant (p \ 0.001) compared with previous time
Table 4 Evolution of the BPI interference overall score at the different visits
Before and Between T0 and Between T0 ? 3 months Between T0 ? 6 months
at T0 T0 ? 3 months and T0 ? 6 months and T0 ? 12 months
Mean (±SD) 30.40 (13.65) 22.81 (11.92)* 19.83 (13.72)*, 14.17 (9.78)*,H
Median 30 20 20.5 10

Statistically significant (p \ 0.01) compared with previous time
H
Statistically highly significant (p \ 0.001) compared with previous visit
prospective study, changes in all outcome measures were

significantly in favor of a single IA injection, whether it
was the BPI severity score, the BPI interference score, the
Harris Hip Score or the pain VAS. This trend was very
clear from T0 ? 3 months; then the results remained
stable or continued to improve on all these criteria.
Only a small number of scientific papers containing
statistically significant results about hip IA injection of HA
are available in the literature despite the hip being the
second most common site of arthritis.
Fig. 2 Evolution of Harris Hip Score during follow-up Conrozier and colleagues [11] retrospectively evaluated
a group of 56 patients with severe or moderate hip
As for the evolution of the Harris Hip Score, the evo- osteoarthritis after one or two IA administrations of high
lution of the pain VAS showed a statistically highly sig- MW HA. At 90 days follow-up, 58.9 % of the patients
nificant improvement (p \ 0.001) between T0 and reported a benefit of the infiltration treatment.
T0 ? 3 months; thereafter it remained stable from the first A prospective double-blind study compared the effect of
quarter post-intervention. high MW and low MW HA, together with a placebo. In this
No serious adverse event was noted; 12 cases (0.5 %) of study 59 patients were evaluated at time intervals of 1, 3
pain associated with transient synovitis (during 24 h) are and 6 months after the first infiltration. Similarly to our
noteworthy. study, improvement of scores was noted at 1 month and
remained significant up to 6 months in both groups com-
pared to the placebo group (p \ 0.001). No significant
Discussion differences were observed between the results obtained in
the two study groups treated with the different HA mole-
To evaluate the efficacy and safety of a single IA of HA cules [12].
(75 mg/3 mL) of high MW (2800 kDa) (Coxarthrum) for Berg and Olsson [13] studied a group of 31 patients with
hip osteoarthritis pain management, we included in our hip osteoarthritis at 2 weeks and 3 months follow-up after
study two hundred seven patients. In this unblinded a single administration of non-animal, stabilized HA

Single intra-articular injection of high molecular weight hyaluronic acid for hip osteoarthritis 21
(NASHA). Three months after administration, there was a narrow IA space, performing a ‘‘blind’’ hip IA injection is
statistically significant 68 % improvement in symptoms not recommended [20]. Ultrasound guidance does not need
(p \ 0.007). contrast media and can also be repeated without causing
Colen and colleagues [5] evaluated the efficacy of hip problems of radiation load for the patient or operator, but
viscosupplementation by analysing the results of 16 trials exposure to radiation during fluoroscopy is minimal and
with a total of 509 patients, with evidence levels varying there is no difference in the speed of IA injection between
from I to IV, and using various types of preparations. the two techniques when performed by experts. The choice
Notwithstanding the relatively low level of evidence in the between ultrasound or fluoroscopy is based on the experi-
trials, the authors concluded that viscosupplementation ence of the operator in using both methods [20]. However,
may be an alternative therapy for treating coxarthrosis. when IA injection is performed under fluoroscopy, the
Intra-articular infiltration has proved to be safe and well- amount of radiopaque contrast agent must be as low as
tolerated therapy. However, the authors state the need for possible to avoid viscosupplement dilution [19].
trials on a larger number of people in order to avoid having In our experience, we observed 12 cases of pain asso-
to consider HA infiltration in the hip as an extremely ciated with transient synovitis after IA injection. As
selective choice that depends on the experience of the reported in the literature [21–24], transient synovitis cor-
operator. The same group of authors [14] reported that related with the reaction to a foreign body is a minor
51 % of the patients had not undergone surgery 3 years complication with an incidence of between 5 and 10 %.
after viscosupplementation, after evaluation of a group of This adverse reaction normally resolves in 24–48 h fol-
120 patients who were candidates for surgical treatment lowing infiltration. without long-term clinical effects.
with a total hip arthroplasty. Further studies are needed on unresolved hip viscosup-
To overcome the problem of selectivity and operator- plementation issues including cost–effectiveness of ther-
dependent approach to hip viscosupplementation, a clear apy, relation between molecular weight and effectiveness,
identification of whether or not the patient is suitable for and how to best incorporate viscosupplementation into an
HA infiltration treatment in the hip is mandatory. Although arthritis therapy algorithm. Our experience proved the
studies of variability in the efficacy of HA infiltration efficacy of IA HA injection for hip arthritis treatment.
therapy in the hip compared to the gravity of the hip A single dose of HA (75 mg/3 mL) of high MW
arthritis have not yet been carried out, it can reasonably be (2800 kDa) is proving to be safe and effective for pain
assumed, as reported by knee viscosupplementation stud- control in patients with hip arthritis (Kellgren–Lawrence
ies, that the possibilities of efficacy and duration of the stages 2 and 3) before indications for hip arthroplasty. Vis-
beneficial effects of the treatment are inversely propor- cosupplementation is effective from the third month and the
tional to the gravity of the disease [15]. For this reason results are stable or continue to improve up to 1 year.
selection criteria for the candidate patient are vital to obtain
pain relief in cases of hip arthritis. These selection criteria Compliance with ethical standards
consist of hip pain for at least 3 weeks, X-ray proof of at Conflict of interest This research received no specific grant from
least partially-preserved joint space, and good or full joint any funding agency in the public, commercial, or not-for-profit
mobility. Hip viscosupplementation can be used as an sectors.
alternative to or in combination with drugs for pain control.
Ethical standards This study was authorized by the local ethical
This type of approach to viscosupplementation therapy
committee and was performed in accordance with the ethical stan-
does not correspond to the inclusion criteria reported in dards of the 1964 Declaration of Helsinki as revised in 2000. The
Van den Bekerom and colleagues’ study [14]. Considering ethical committee waived the need for informed consent since the
hyaluronic acid as a pain therapy, using its beneficial rights and interests of the patients would not be violated and their
privacy and anonymity would be assured by this study design.
effects on cartilage due to both the pharmacological and
the physical properties of the molecule [16–18], then the
use of injection in cases of low or medium degrees of hip
arthritis is mandatory. For this reason, patient candidates
for hip arthroplasty were excluded for our study.
Recently, to clarify some aspects of viscosupplementa-
tion treatment, a review of the literature confirmed that IA
HA is an effective treatment for mild to moderate
osteoarthritis but it is not an alternative to surgery in References
advanced cartilage degeneration [19].
Fluoroscopic guidance is one of the possible radiologi- 1. Tait RC, Chibnall JT (1997) Physician judgments of chronic pain
cal guidances when performing IA hip injection. Due to the patients. Soc Sci Med 45:1199–1205

2. Wilson MG, Michet CJ Jr, Ilstrup DM, Melton LJ 3rd (1990) efficacy of different molecular weight hyaluronan solutions in the
Idiopathic symptomatic osteoarthritis of the hip and knee: a treatment of knee osteoarthritis. Rheumatol Int 26:325–330
population-based incidence study. Mayo Clin Proc 13. Berg P, Olsson U (2004) Intra-articular injection of non-animal
65(9):1214–1221 stabilised hyaluronic acid (NASHA) for osteoarthritis of the hip:
3. Balazs EA, Denlinger JL (1993) Viscosupplementation: a new a pilot study. Clin Exp Rheumatol 2(3):300–306
concept in the treatment of osteoarthritis. J Rheumatol 39:3–9 14. Van den Bekerom MP, Rys B, Mulier M (2008) Viscosupple-
4. Zhang W, Moskowitz RW, Nuki G, Abramson S, Altman RD, mentation in the hip: evaluation of hyaluronic acid formulations.
Arden N, Bierma-Zeinstra S, Brandt KD, Croft P, Doherty M, Arch Orthop Trauma Surg 18(3):275–280
Dougados M, Hochberg M, Hunter DJ, Kwoh K, Lohmander LS, 15. Evanich DJ, Evanich CJ, Wright MB, Rydlewicz JA (2001)
Tugwell P (2008) OARSI recommendations for the management Efficacy of intrarticular hyaluronic acid injections in knee
of hip and knee osteoarthritis, part II: OARSI evidence-based, osteoarthritis. Clin Orthop 390:173–181
expert consensus guidelines. Osteoarthritis Cartilage 2:137–162 16. Vitanzo PC Jr, Sennett BJ (2006) Hyaluronans: is clinical
5. Van den Bekerom MP, Lamme B, Sermon A, Mulier M (2008) effectiveness dependent on molecular weight? Am J Orthop
What is the evidence for viscosupplementation in the treatment of 35(9):421–428
patients with hip osteoarthritis? Systematic review of the litera- 17. Lindholm A, Ronéus B, Lindblad G, Jones B (1996) Hyaluronan
ture. Arch Orthop Trauma Surg 128:815–823 turnover in the synovial fluid in metacarpophalangeal–and middle
6. Colen S, van den Bekerom MP, Bellemans J, Mulier M (2010) carpal joints in standardbred horses. Acta Vet Scand
Comparison of intra-articular injections of hyaluronic acid and 37(2):147–151
corticosteroid in the treatment of osteoarthritis of the hip in 18. Lùrati A, Laria A, Mazzocchi D, Re KA, Marrazza M, Scarpellini
comparison with intra-articular injections of bupivacaine. Design M (2015) Effects of hyaluronic acid (HA) viscosupplementation
of a prospective, randomized, controlled study with blinding of on peripheral Th cells in knee and hip osteoarthritis.
the patients and outcome assessors. BMC Musculoskelet Disord Osteoarthritis Cartilage 23(1):88–93
16(11):264–269 19. Henrotin Y, Raman R, Richette P, Bard H, Jerosch J, Conrozier
7. Atchia I, Kane D, Reed MR, Isaacs JD, Birrell F (2011) Efficacy T, Chevalier X, Migliore A (2015) Consensus statement on vis-
of a single ultrasound-guided injection for the treatment of hip cosupplementation with hyaluronic acid for the management of
osteoarthritis. Ann Rheum Dis 70(1):110–116 osteoarthritis. Semin Arthritis Rheum 30 2015.04.011
8. Migliore A, Massafra U, Bizzi E, Vacca F, Martin-Martin S, 20. Rivera F (2014) Can viscosupplementation be used in the hip? An
Granata M, Alimonti A, Tormenta S (2009) Comparative, dou- Italian perspective. Orthopedics 37(1):48–55
ble-blind, controlled study of intra-articular hyaluronic acid 21. Goomer RS, Leslie K, Maris T, Amie ID (2005) Native
(Hyalubrix) injections versus local anesthetic in osteoarthritis of hyaluronan produces less hypersensitivity than cross-linked
the hip. Arthritis Res 11(6):183–186 hyaluronan. Clin Orthop 434:239–245
9. Richette P, Ravaud P, Conrozier T, Euller-Ziegler L, Mazières B, 22. Morton AH, Shannon P (2003) Increased frequency of acute local
Maugars Y, Mulleman D, Clerson P, Chevalier X (2009) Effect reaction to intra-articular Hylan G-F 20 (Synvisc) in patients
of hyaluronic acid in symptomatic hip osteoarthritis: a multi- receiving more than one course of treatment. J Bone Joint Surg A
center, randomized, placebo-controlled trial. Arthritis Rheum 85:2050–2051
60(3):824–830 23. Goldberg VM, Coutts RD (2004) Pseudoseptic reactions to hylan
10. Ahlbäck S (1968) Osteoarthrosis of the knee: a radiographic viscosupplementation: diagnosis and treatment. Clin Orthop
investigation. Acta Radiol Stockholm Suppl 277:70–72 419:130–137
11. Conrozier T, Bertin P, Bailleul F, Mathieu P, Charlot J, Vignon 24. Juni P, Reichenbach S, Trelle S, for the Swiss Viscosupple-
E, Treves R, Chevalier X (2006) Clinical response to intra-ar- mentation Trial Group et al (2007) Efficacy and safety of
ticular injections of hylan G-F 20 in symptomatic hip intraarticular hylan or hyaluronic acids for osteoarthritis of the
osteoarthritis: the OMERACT-OARSI criteria applied to the knee: a randomized controlled trial. Arthritis Rheum
results of a pilot study. Joint Bone Spine 73(6):705–709 56:3610–3619
12. Kotevoglu N, Lyibozkurt PC, Hiz O, Toktas H, Kuran B (2006) A
prospective randomised controlled clinical trial comparing the

4
Long-term clinical results and MRI changes after autologous
chondrocyte implantation in the knee of young and active middle
aged patients
Donato Rosa1 • Giovanni Balato1 • Giovanni Ciaramella1 • Ernesto Soscia2 •
Giovanni Improta1 • Maria Triassi1
Abstract improvements in all clinical outcome parameters, even as

Background Autologous chondrocyte implantation (ACI) much as 12 years after implantation. A significant decrease
represents a valid surgical option for symptomatic full- in the MOCART score was recorded at final measurement.
thickness chondral lesions of the knee. Here we report Reoperation was required in 2 patients; failure was caused
long-term clinical and MRI results of first-generation ACI. by partial detachment of the graft in both cases.
Materials and methods Fifteen patients (mean age Conclusion Autologous chondrocyte implantation is an
21.3 years) underwent first-generation ACI for symp- effective and durable solution for the treatment of large,
tomatic chondral defects of the knee between 1997 and full-thickness cartilage and osteochondral lesions, even in
2001. The mean size of the lesions was 5.08 cm2 (range young and active middle-aged patients. High-resolution
2–9 cm2). Patients were evaluated using the International MRI is a useful and noninvasive method for evaluating the
Knee Documentation Committee (IKDC) Knee Examina- repaired tissue.
tion Form, the Tegner Activity Scale, and the Knee Injury Level of evidence IV.
and Osteoarthritis Outcome Score (KOOS). High-resolu-
tion MRI was used to analyze the repair tissue with nine Keywords Autologous chondrocyte implantation
variables (the MOCART scoring system). Chondral lesion Magnetic resonance imaging Knee
Results The mean follow-up period was 148 months Osteochondritis dissecans
(range 125–177 months). ACI resulted in substantial
Introduction
& Giovanni Balato
[email protected]
Cartilage lesions of the knee in orthopedic patients are an
Donato Rosa underestimated problem. Despite the advances made in
[email protected]
scientific knowledge and technology, treatment of these
Giovanni Ciaramella lesions remains troublesome. Autologous chondrocyte
[email protected]
implantation (ACI), first reported in 1994 by Brittberg
Ernesto Soscia et al., was introduced as an alternative means of treating
[email protected]
symptomatic full-thickness chondral lesions of the knee
Giovanni Improta [1]. After ACI, cartilage repair tissue consists mainly of
[email protected]
cartilage-like tissue that mimics the macroscopic, micro-
Maria Triassi scopic, and biomechanical features of healthy hyaline
[email protected]
cartilage [2, 3]. Magnetic resonance imaging (MRI) is the
1
Department of Public Health, School of Medicine, Federico II most reproducible and least aggressive technique for
University, Via S. Pansini 5, Bl. 12, 80131 Naples, Italy assessing cartilage regeneration after ACI. One validated
2
Institute of Biostructure and Bioimaging, National Research scoring system for the morphologic MRI evaluation of
Council, Via S. Pansini 5, 80131 Naples, Italy cartilage repair sites is the Magnetic Resonance

Observation of Cartilage Repair Tissue (MOCART) system Table 1 Patient demographic characteristics as well as prior and
[4, 5]. Although satisfactory results in mid-term pain relief concomitant procedures
have been reported [6], only a limited number of studies Patient characteristic N = 15
have examined the long-term results of ACI in terms of Age, years
clinical assessment, patient satisfaction, and magnetic Mean ± SD 21,33 ± 8,92
resonance imaging (MRI) results [7–13]. Range 13–45
The purposes of the study reported in the present paper Gender
were therefore to: Male 9/15
• Evaluate the overall long-term results of ACI in terms Female 6/15
of clinical assessment, patient satisfaction, and mag- Previous procedures
netic resonance imaging (MRI) results Debridement/lavage 1/15
• Compare the long-term with the short-term clinical Procedures performed concurrently with cartilage harvest
results Medial meniscectomy 2/15
• Evaluate the correlation between the subjective clinical Lateral meniscectomy 1/15
outcome and the radiological MOCART scoring system Procedures performed with implantation
and its variables. ACL reconstruction 1/15
Patellar alignment 1/15
SD standard deviation
Materials and methods
Table 2 Characteristics of chondral lesions
Between 1997 and 2001, 15 patients (nine men and six Defect characteristic
women), with a mean age of 21 years (range 13–45),
underwent autologous chondrocyte implantation using the Acute traumatic injury 8/15
original periosteum-cover technique. All patients had knee Osteochondritis dissecans 7/15
pain and had decreased their physical activity due to the Total surface area, cm2
presence of a chondral defect in the concerned knee. A Mean ± SD 5.08 ± 2.01
symptomatic full-thickness cartilage lesion (Outerbridge Range 2–9
grade III or IV) or an osteochondral lesion (2–12 cm2) was Defect location
considered an indication for ACI. Exclusion criteria were Medial femoral condyle 10/15
age [45 years, prolonged osteoarthritis (Kellgren–Lawr- Lateral femoral condyle 2/15
ence grade 2 or more), obesity (BMI [ 35 kg/m2), a kiss- Patella 2/15
ing lesion, active inflammatory arthritis or infection, Tibial plate 1/15
varus/valgus alignment [5, and/or untreated knee insta- SD standard deviation
bility. No patient had undergone any previous surgical
attempts to treat the chondral defect, except for one case in
which a meniscectomy was performed together with a and written informed consent for the publication of their
chondral debridement for a patellar lesion. A trauma was individual clinical details in this paper; this was approved
the cause of the chondral defect in eight cases, whilst by the institutional review board of our department and is
osteochondritis dissecans was the underlying cause in compliant with the Declaration of Helsinki.
seven cases. Patients with osteochondritis dissecans were
rated International Cartilage Repair Society (ICRS) stage 2 Surgical technique
(partial discontinuity, stable on probing) or 3 (having an
unstable but not dislocated fragment). The mean size of the The ACI technique consisted of a two-step procedure, as
lesion surface was 5.08 ± 2.01 cm2 (range 2–9 cm2). originally described by Brittberg [1]. First, an arthroscopy
Lesions were localized on the medial femoral condyle in 10 was performed, where small pieces of full-thickness car-
cases, on the lateral femoral condyle in two cases, on the tilage were harvested from a low-weight-bearing area of
patella in two cases, and on the tibial plateau in one case. the trochlea or from the upper area of the medial condyle;
During the first arthroscopic step, three partial meniscec- these pieces weighed approximately 200–300 mg. The
tomies (two medial and one lateral) were performed, while biopsy material was placed in a nutrient medium and
an anterior cruciate ligament (ACL) reconstruction and a transported within 24 h to a chosen laboratory. Chondro-
patellar alignment (Tables 1, 2) were performed during the cytes were isolated from the cartilage by enzymatic treat-
implantation of chondrocytes. All individuals provided oral ment, and the number of chondrocytes was increased via

Long-term clinical results and MRI changes after autologous chondrocyte implantation in the knee of young... 25
monolayer culture, as described previously [1]. After functional evaluation at baseline. Before the surgical pro-
3–4 weeks, an autologous pool of chondrocytes was ready cedure/surgery, all of the patients underwent a physical
to be implanted. The surgical approach used in the examination, and weight-bearing standing radiographs as
implantation step of the procedure depended on the size well as magnetic resonance images (MRIs) of the affected
and location of the defect: a medial or lateral parapatellar knees were recorded. Follow-up was accomplished in all
arthrotomy was performed. Defect edges were marked and patients for a mean period of 148.1 ± 15.76 months (range
then cut using a surgical blade, creating a contained lesion 125–177 months). Each year, a clinical evaluation was
with surrounding healthy cartilage. In all cases with performed. At the final follow-up, functional evaluation
osteochondritis dissecans, the lesion was identified and the was performed with the IKDC Knee Examination Form,
fragment was removed along with fibrous tissue and the Tegner Activity Level Score, and the Knee Injury and
degenerated bone until healthy, bleeding bone was reached. Osteoarthritis Outcome Score (KOOS) (Italian version LK
A periosteal flap was harvested from the proximal medial 1.0) [17, 18]. Cases in which further surgery was per-
subcutaneous border of the tibia. An incision was made formed after ACI were defined as treatment failures.
about 3 cm below the insertion of the pes anserinus. With
the inner cambium layer facing the lesion, the periosteal Radiological evaluation
flap was sutured to the surrounding cartilage using inter-
rupted absorbable sutures. The periosteal rim was sealed Thirteen of the patients periodically underwent magnetic
with fibrin glue except for one corner, where the suspen- resonance imaging (1.5 T, Siemens Symphony) according
sion of cultured chondrocytes (Carticel) was injected into to the following acquisition protocol:
the defect. The implant was completed by closing the
• Axial TSE PD FS 2D 1-7/180 (TE 37, TR 2500, matrix
corner with a final suture and the fibrin glue.
192 9 256, FOV 180 9 180).
• Sagittal TSE PD FS 2D 1-7/180 (TE 38, TR 2000,
Postoperative rehabilitation protocol
matrix 240 9 320, FOV 180 9 180).
• Coronal TSE PD FS 2D 1-7/180 (TE 37, TR 2000,
The goal of rehabilitation was to protect the graft while
matrix 256 9 256, FOV 180 9 180).
promoting maturation of the newly implanted chondrocytes
• Sagittal TSE PD 2D 1-7/180 (TE 38, TR 2000, matrix
by implementing a program that focused on regaining full
240 9 320, FOV 180 9 180).
range of motion (ROM), progressive weight bearing, lower
• Sagittal 3D spoiled GRE T1 (Fi 3D 1/40) (TE 8, TR 34,
extremity strengthening, flexibility, and proprioceptive
matrix 192 9 256, FOV 180 9 180).
training. In particular, when at least 24 h had passed fol-
lowing surgery, the knee was mobilized with the help of a The mean time of the first MRI after the implantation
continuous passive motion (CPM) machine. Weight-bear- was 12 months (range 6–30 months). At final follow-up
ing activity was typically barred until after the first 2 weeks after a mean of 148.1 months (range 122–175 months), 11
of implantation in order to preserve the physical properties patients were studied using a high-field MRI instrument
of the graft. Partial weight bearing was then permitted until (3 T, Siemens Magnetom Trio) that was available at the
4 weeks after surgery. From 4 to 6 weeks after surgery, the time in the radiologic department of our institution. The
patient could progress to the use of one crutch, with the examination was performed with a dedicated knee coil.
load gradually increased over the subsequent 6 weeks so The acquisition protocol was as follows:
that full weight bearing had occurred by week 12. By
• Axial TSE PD FS (TE 11, TR5890, matrix 256 9 256,
3 months after surgery, the patient had recovered their full
thickness 3 mm, FOV 160 9 160).
active range of motion with a normalized gait pattern. At
• Sagittal TSE PD FS (TE 11, TR4660, matrix
6–9 months after surgery, the patient continued progressive
320 9 320, thickness 3 mm, FOV 160 9 160).
strength training and transitioned to more functional
• Coronal FFE 3D T1 hi-res VIBE (Te5, Tr14,2, FA 25,
activities. From 9 to 18 months after surgery, the goal of
slice thickness 0.6, matrix 512 9 512, FOV
the rehabilitation was to implement sports-specific activity
150 9 150).
and eventually facilitate the return of the patient to
• Sagittal FFE DP 3D hi-res (Te232, Tr2200, matrix
competition.
230 9 250, FA 120, slice thickness 0.8, FOV
162 9 181).
Clinical evaluation
• Sagittal T2 3D hi-res (Te4.9, Tr11, matrix 480 9 512,
slice thickness 0.6, FA 40, FOV 140 9 150).
The International Knee Documentation Committee (IKDC)
Knee Examination Form [14, 15] and the Tegner Activity The images were evaluated by an expert radiologist
Level Score [16] were used to perform clinical and according to the MOCART scoring system.

Statistical analysis both short-term and long-term follow-ups; the paired t-test
showed a significant decrease in the MOCART score from
Paired sample t-tests were used to determine whether fol- 59 ± 29.13 to 43.5 ± 32.91 (P = 0.0226) (Fig. 3).
low-up data were significantly increased or decreased from Table 3 shows the results for each variable of the
the baseline clinical scores. In this context, the MRI score MOCART score at both short-term and long-term follow-
after short-term follow-up was compared with the long- ups. The correlation coefficients and the results of the t-test
term follow-up MOCART score by paired t-test. To for subjective outcomes and the different variables of the
determine the correlation between clinical outcome and MRI classification system indicated that there were statis-
MRI score, the KOOS and the IKDC scores were corre- tically significant correlations between degree of defect
lated with the MOCART score and with the nine variables repair and pain KOOS as well as between effusion and the
of the MRI scoring system. For the statistical analysis, pain and symptoms KOOS (p \ 0.05).
Spearman’s correlation coefficient (rs) and Student’s t-test
were calculated. To evaluate the relationships of the
MOCART score and some of its variables (degree of defect Discussion
filling, integration of border zone, surface of the repair
tissue, structure of the repair tissue, signal intensity of the In this study, we evaluated the durability of ACI in patients
repair tissue) with the KOOS variables and those of the treated for full-thickness cartilage defects. The most
IKDC, Spearman’s correlation coefficient was calculated important finding of our study was confirmation of the
considering the ranks of the variables, not their numerical long-term effectiveness of ACI, even up to 14 years after
values. An independent samples t-test was used for the the first implantation performed in our center. Clinical and
remaining MOCART variables (subchondral lamina, sub- functional improvements, with significantly increased
chondral bone, adhesions, effusion). All tests were per- mean IKDC and Tegner Activity Level scores, were
formed using the statistical software package R (R observed in 86.6 % of cases. A functional evaluation was
Development Core Team, 2005). In all instances, P \ 0.05 also performed using the Knee Injury and Osteoarthritis
was regarded as statistically significant. Outcome Score (KOOS) at final follow-up. The KOOS
results were compared with the age-specific KOOS scores
for the general population, as obtained in the epidemio-
Results logical study of Paradowski et al. [19]. The mean KOOS
scores for the 18–34 year-old age group were 92.2 men/
All 15 patients were retrospectively followed up after ACI 92.1 women for pain, 87.2 men/89.1 women for symptoms,
for a mean period of 148.1 ± 15.76 months (range 94.2 men/95.2 women for ADL, 85.1 men/86.4 women for
125–177 months). Two patients (13.3 %) needed an oper- sports, and 85.3 men/83.6 women for quality of life. At the
ation after ACI, entailing removal of the graft and treat-
ment of the defect with microfractures. Failure was in both
cases due to partial detachment of the graft and degener-
ation of the graft area. The graft site was also filled with
fibrous tissue that was partially lifted at its medial aspect,
exposing the subchondral bone. The defect area was deb-
rided from the fibrous tissue and the chondral lesion was
exposed. The microfracture technique was then performed
as a treatment for the lesion. At the final follow-up, sig-
nificant increases in all scores were recorded. Compared
with the pre-procedure findings, the mean IKDC score
improved significantly, increasing from 37.20 ± 19.54 to
76.32 ± 32.36 (P = 0.000314) (Fig. 1). The Tegner
Activity Level Score showed significant improvement after
surgery, increasing from 2.33 ± 1.34 to 4.93 ± 2.43
(P = 0.0011) (Fig. 2). The KOOS scores were as follows:
pain 79.63 ± 33.33; symptoms 76.42 ± 32.47; ADL
85.09 ± 34.62; sport 70.33 ± 31.13; knee-related quality
of life 74.17 ± 32.72. The mean MOCART score at the
first follow-up was 55 ± 26.53, whereas that at the last Fig. 1 IKDC score: improvement from pre-operative levels to final
follow-up was 45 ± 31.62. Ten patients underwent MRI at follow-up

With only 2 failures (13.3 %), the results of the pre-

sent study compare favorably with other such reports in
the literature [2, 9, 11]. Our treatment failure rate was
rather low compared with the reported rates of 16–24 %
observed in comparable treatment settings [8, 17]. The
two failures occurred in our study due to early deterio-
ration of the graft site; this complication always leads to
clinical failure and a new intervention always becomes
necessary. In both cases, it became necessary to remove
the graft and treat the osteochondral damage with
microfractures [20]. Four of our patients (26.6 %)
showed hypertrophy of the graft on MRI at first radio-
logic follow-up, but in none of those cases was it
symptomatic, so we did not perform a second-look
arthroscopy in any of these cases. Transplant hypertrophy
is a complication associated with the use of periosteum
[20–23]. Thus, several modifications of the initial tech-
Fig. 2 Tegner score: improvement from pre-operative levels to final nique, such as periosteal flap peeling or flap substitution
follow-up with synthetic membranes or fibrin matrix, were pro-
posed to minimize its incidence and attain satisfactory
results [21, 23].
In addition to providing data on long-term clinical out-
comes, our study also contributes information on MRI
assessment. MRI is a noninvasive method for assessing
structural repair outcomes, and is considered the most
effective tool for evaluating the internal structures of the
knee joint. A second look via arthroscopy would enable
better evaluation of the obtained repair tissue, but the
invasive nature of this procedure would not allow it to be
performed daily in a clinical setting. Moreover, the risks
associated with such an invasive approach are not accept-
able for ethical reasons, in particular for patients with
satisfactory outcomes. It may only be justifiable in cases of
failed treatment when the patient needs further cartilage
treatment. In our study, second-look arthroscopy was only
performed in two patients for whom ACI failed.
The cartilage and ACI graft were assessed with 3D
Fig. 3 MOCART score: comparison of MOCART score at sequences, which provided superior spatial resolution,
12 months with that at final follow-up aiding definition of the defect filling, the integration of the
graft with the underlying bone and adjacent native carti-
final follow-up evaluation, our patients had a mean age of lage, and the status of the subchondral bone and bone
33.6 years and, if the two failed implantations are exclu- marrow. To describe the repair tissue, we used the previ-
ded, the average KOOS scores were 95.1 men/86.6 women ously published MOCART classification [8]. We used the
for pain, 93.3 men/80 women for symptoms, 97.8 men/98.8 MOCART score to evaluate the results after a mean fol-
women for ADL, 88.7 men/69 women for sport, and 89.1 low-up period of 148.1 months. We compared the MRI
men/80 women for quality of life. The KOOS results findings with the clinical outcomes. Initially, the MRI
obtained in our study are comparable with the results of variables were correlated with the subjective patient eval-
Paradowski et al. [19]. The increases observed in all clin- uation using the KOOS and IKDC scoring systems. Sta-
ical and functional scores at the last follow-up might be tistically significant correlations between the clinical
related to the young age of the study population at the time outcome and some of the radiological variables were
of ACI. Indeed, the mean age of the participating patients found. A statistically significant correlation of filling of the
was 21.33 years (range 13–45), 10 years less than the defect with KOOS pain was observed (P \ 0.05). Effusion
average age of patients in other studies [10–12]. was statistically significantly correlated with KOOS pain

Table 3 MRI evaluation of

Variables First follow-up Last follow-up
repair tissue 1–2 years and
10 years after ACI implantation 1. Degree of defect repair and filling of the defect
Complete 2/10 (20 %) 2/10 (20 %)
Hypertrophy 4/10 (40 %) 4/10 (40 %)
Incomplete
[50 % of the adjacent cartilage 3/10 (30 %) 0/10 (0 %)
\50 % of the adjacent cartilage 0/10 (0 %) 1/10 (10 %)
Subchondral bone exposed 1/10 (10 %) 3/10 (30 %)
2. Integration to border zone
Complete 7/10 (70 %) 4/10 (40 %)
Incomplete
Demarcating border visible (split-like) 0/10 (0 %) 1/10 (10 %)
Defect visible
\50 % of the length of the repair tissue 1/10 (10 %) 2/10 (20 %)
[50 % of the length of the repair tissue 2/10 (20 %) 3/10 (30 %)
3. Surface of the repair tissue
Surface intact 5/10 (50 %) 5/10 (50 %)
Surface damaged
\50 % of repair tissue depth 2/10 (20 %) 2/10 (20 %)
[50 % of repair tissue depth or total degeneration 3/10 (30 %) 3/10 (30 %)
4. Structure of the repair tissue
Homogeneous 3/10 (30 %) 1/10 (10 %)
Inhomogeneous or cleft formation 7/10 (70 %) 9/10 (90 %)
5. Signal intensity of the repair tissue
Normal (identical to adjacent cartilage) 3/10 (30 %) 2/10 (20 %)
Nearly normal 5/10 (50 %) 5/10 (50 %)
Abnormal 2/10 (20 %) 3/10 (30 %)
6. Subchondral lamina
Intact 2/10 20 %) 1/10 (10 %)
Not intact 8/10 (80 %) 9/10 (90 %)
7. Subchondral bone
Intact 4/10 (40 %) 1/10 (10 %)
Not intact 6/10 (60 %) 9/10 (90 %)
8. Adhesions
No 8/10 (80 %) 2/10 (20 %)
Yes 2/10 (20 %) 8/10 (80 %)
9. Effusions
No 8/10 (80 %) 6/10 (60 %)
Yes 2/10 (20 %) 4/10 (40 %)
and symptoms (P \ 0.05). No statistically significant cor- scoring system, only one remained unchanged over time:
relation was found for the other variables. the surface of the repair tissue. Patients who had complete
Marlovits et al. compared clinical scores with MRI or hypertrophic filling at the first follow-up presented a
variables and found statistically significant correlations stable degree of defect repair over time, whilst the others
with only four of the nine MOCART variables (filling of who had incomplete filling showed a deterioration in the
the defect, structure of the repair tissue, subchondral bone, score for this variable at last follow-up (Table 3). The
signal intensity of the repair tissue) [5]. Our study evalu- observed reductions in MOCART variable scores can be
ated MRI images at long-term follow-up for only ten knees explained by graft aging and the alteration of the whole
and compared these data with images taken at the first joint. The MOCART variables that were more likely to
follow-up. Among the nine variables of the MOCART show reduced scores were those linked to the underlying

bone alteration and to the presence of adhesions. The References

health and integration of the patch were seen to be com-
promised in a few cases. Those variations are probably 1. Brittberg M, Lindahl A, Nilsson A, Ohlsson C, Isaksson O,
Peterson L (1994) Treatment of deep cartilage defects in the knee
linked to inflammation, which can be present in a joint that
with autologous chondrocyte transplantation. N Engl J Med
does not work optimally; moreover, the new cartilage is 331:889–895
probably less strong than that surrounding it. Perhaps a 2. Peterson L, Brittberg M, Kiviranta I, Akerlund EL, Lindahl A
longer follow-up period, although difficult to implement, (2002) Autologous chondrocyte transplantation. Biomechanics
and long-term durability. Am J Sports Med 30:2–12
may reveal if the MOCART score plateaus after decreasing
3. Peterson L, Minas T, Brittberg M, Lindahl A (2003) Treatment of
or if it continuously decreases. osteochondritis dissecans of the knee with autologous chondro-
In conclusion, first-generation ACI seems to be an cyte transplantation: results at two to ten years. J Bone J Surg Am
effective and durable treatment for large, full-thickness 85:17–24
4. Marlovits S, Singer P, Zeller P, Mandl I, Haller J, Trattnig S
chondral and osteochondral defects of the knee. ACI pro-
(2006) Magnetic resonance observation of cartilage repair tissue
vides satisfactory results in terms of both pain relief and (MOCART) for the evaluation of autologous chondrocyte trans-
knee function rehabilitation, which appear to be sustained plantation: determination of interobserver variability and corre-
in the majority of patients according to long-term follow- lation to clinical outcome after 2 years. Eur J Radiol 57:16–23
5. Marlovits S, Striessnig G, Resinger CT, Aldrian SM, Vecsei V,
up results. Magnetic resonance imaging plays an important
Imhof H et al (2004) Definition of pertinent parameters for the
role during the post-procedure follow-up of cartilage repair evaluation of articular cartilage repair tissue with high-resolution
procedures, as it permits accurate and noninvasive assess- magnetic resonance imaging. Eur J Radiol 52(3):310–319
ment of the status of cartilage repair, even though there is 6. Browne JE, Anderson AF, Arciero R, Mandelbaum B, Moseley
JB Jr, Micheli LJ, et al (2005) Clinical outcome of autologous
no significant linear correlation between the overall MRI
chondrocyte implantation at 5 years in US subjects. Clin Orthop
score and the subjective and objective knee scores. Relat Res 436:237–254
Some limitations of our study need to be acknowledged. 7. Peterson L, Vasiliadis HS, Brittberg M, Lindahl A (2010)
Treatment effects may have been overestimated or under- Autologous chondrocyte implantation: a long-term follow-up.
Am J Sports Med 38:1117–1124
estimated because of the lack of a control group. Com-
8. Moseley JB, Anderson AF, Browne JE, Mandelbaum BR,
parison to a control group would aid accurate Micheli LJ, Fu F et al (2010) Long-term durability of autologous
interpretation, as it would allow the spontaneous evolution chondrocyte implantation: a multicenter, observational study in
of untreated lesions of a similar size to be evaluated. The US patients. Am J Sports Med 38:238–246
9. Brittberg M (2008) Autologous chondrocyte implantation—
literature provides only very limited data on patients with
technique and long-term follow-up. Injury 39(Suppl 1):s40–s49
untreated cartilage lesions [24]. Cicuttini et al. suggested 10. Minas T (2001) Autologous chondrocyte implantation for focal
that full-thickness cartilage lesions in young patients may chondral defects of the knee. Clin Orthop Relat Res 391:S349–
provoke early osteoarthritis over time [25]. To obtain more S361
11. Pelissier A, Boyer P, Boussetta Y, Bierry G, Van Hille W, Hamon P
reliable data, a second study arm of patients with healthy
et al (2014) Satisfactory long-term MRI after autologous chon-
knees or untreated chondral lesions would be of special drocyte implantation at the knee. Knee Surg Sports Traumatol
interest. However, it should be remembered that such a Arthrosc 22(9):2007–2012. doi:10.1007/s00167-013-2428-9
control group would be difficult to create due to ethical 12. Moradi B, Schonit E, Nierhoff C, Hagmann S, Oberle D, Got-
terbarm T et al (2012) First-generation autologous chondrocyte
considerations, and this remains a limitation of our analysis
implantation in patients with cartilage defects of the knee: 7 to
and other analyses of the long-term outcomes of ACI and 14 years’ clinical and magnetic resonance imaging follow-up
other treatment options [2, 8, 7]. evaluation. Arthroscopy 28(12):1851–1861
13. Mithöfer K, Minas T, Peterson L, Yeon H, Micheli LJ (2005)
Functional outcome of knee articular cartilage repair in adoles-
Compliance with ethical standards cent athletes. Am J Sports Med 33:1147–1153
14. Irrgang JJ, Anderson AF, Boland AL, Harner CD, Kurosaka M,
Conflict of interest None. Neyret P et al (2001) Development and validation of the inter-
national knee documentation committee subjective knee form.
Ethical standards The study conforms to the 1964 Helsinki dec- Am J Sport Med 29:600–613
laration and its later amendments; it was approved by the responsible 15. Anderson AF, Irrgang JJ, Kocher MS, Mann BJ, Harrast JJ,
Ethics Committee and all the patients provided informed consent International Knee Documentation Committee (2006) The Inter-
before being enrolled. national Knee Documentation Committee Subjective Knee
Evaluation Form: normative data. Am J Sports Med 34:128–135
16. Tegner Y, Lysholm J (1985) Rating system in the evaluation of
knee ligament injuries. Clin Orthop Relat Res 198:43–49
17. Roos EM, Roos HP, Ekdahl C, Lohmander LS (1998) Knee
injury and Osteoarthritis Outcome Score (KOOS)—validation of
a Swedish version. Scand J Med Sci Sports 8:439–448
18. Roos EM, Roos HP, Lohmander LS, Ekdahl C, Bynnon BD
(1998) Knee injury and Osteoarthritis Outcome Score (KOOS)—

development of a self-administered outcome measure. J Orthop chondrocyte implantation—a systematic review. Osteoarthr Car-
Sports Phys Ther 28:88–96 til 19:779–791
19. Paradowski PT, Bergman S, Sunden-Lundius A, Lohmander LS, 23. Kreuz PC, Steinwachs M, Erggelet C, Krause SJ, Ossendorf C,
Roos EM (2006) Knee complaints vary with age and gender in Maier D et al (2007) Classification of graft hypertrophy after
the adult population. Population-based reference data for the autologous chondrocyte implantation of full-thickness chondral
Knee injury and Osteoarthritis Outcome Score (KOOS). BMC defect in the knee. Osteoarthr Cartil 15:1139–1147
Musculoskelet Disord 7:38 24. Shelbourne KD, Jari S, Gray T (2003) Outcome of untreated
20. Vasiliadis HS, Doukas M, Batistatou A, Georgoulis A, Beris AE traumatic articular cartilage defects of the knee: a natural history
(2008) Chondromatosis of the knee joint 8 months after autolo- study. J Bone J Surg Am 85A(Suppl 2):8–16
gous chondrocyte implantation. Am J Sports Med 10:562–567 25. Cicuttini FM, Wluka AE, Stuckey SL (2001) Tibial and femoral
21. Beris AE, Lykissas MG, Kostas-Agnantis I, Manoudis GN (2012) cartilage changes in knee osteoarthritis. Ann Rheum Dis
Treatment of full-thickness chondral defect of the knee with 60:977–980
autologous chondrocyte implantation: a functional evaluation
with long-term follow-up. Am J Sports Med 40:562–567
22. Harris JD, Siston RA, Brophy RH, Lattermann C, Flaningan DC
(2011) Failure, re-operations, and complications after autologous

5
Outcomes of intramedullary nailing for acute proximal humerus
fractures: a systematic review
Jason Wong1 • Jared M. Newman1 • Konrad I. Gruson1
Abstract four-part fractures was 13.6, 17.4, and 63.2 %,

Background While proximal humerus fractures remain respectively.
common within the elderly population, the optimal treat- Conclusions Intramedullary nailing of acute, displaced
ment method remains controversial. Intramedullary nailing two- and three-part proximal humerus fractures yields
has been advocated as an effective and less invasive sur- satisfactory clinical outcomes, although reoperation and
gical technique. The purpose of this study is to elucidate complication rates remain high. Use of this implant for
the demographics, outcomes, and complications of intra- four-part fractures cannot be recommended until further
medullary nailing for acute, displaced proximal humerus clinical studies with larger patient numbers are available.
fractures. Level of evidence Level IV, Systematic review.
Materials and methods Multiple computerized literature
databases were used to perform a systematic review of Keywords Systematic review Intramedullary nail
English-language literature. Studies that met our stated Proximal humerus fracture Outcomes
criteria were further assessed for the requisite data, and
when possible, similar outcome data were combined to
generate frequency-weighted means. Introduction
Results Fourteen studies with 448 patients met our
inclusion criteria. The frequency-weighted mean age was Proximal humerus fractures account for 4–5 % of all
64.3 years, and mean follow-up was 22.6 months. Females fractures and occur most frequently in elderly female
accounted for 71 % of the included patients. Three-part patients [1–4]. From 1999 to 2005 there was a 25 % rela-
fractures (51 %) were most commonly treated. The overall tive increase of proximal humerus fractures treated surgi-
frequency-weighted mean Constant score was 72.8, and cally [2]. Commonly utilized techniques include
American Shoulder and Elbow Surgeons (ASES) score was percutaneous fixation [5], open reduction with locking
84.3. Frequency-weighted mean forward elevation, plate fixation (ORIF) [6], intramedullary nailing (IMN) [7,
abduction, extension, and external rotation were 137.3°, 8], hemiarthroplasty (HA) [9], and reverse shoulder
138.4°, 33.8°, and 43.1°, respectively. The Constant score arthroplasty (RSA) [10]. Continued debate exists as to
for two- and three-part fractures was significantly higher which of these represents the ‘‘gold standard’’ to manage
than for four-part fractures (p = 0.007 and p = 0.0009, acute, displaced proximal humerus fractures. Amongst
respectively). The reoperation rate for two-, three-, and nonarthroplasty techniques, some studies have reported on
successful use of locked plating in treating the more
complex, three- and four-part fractures [6]. Others have
& Konrad I. Gruson shown similarly good clinical outcomes with the use of a
[email protected] locked, antegrade intramedullary nail [11, 12]. The pur-
1 ported advantages of IMN include decreased soft tissue
Department of Orthopaedic Surgery, Albert Einstein College
of Medicine, 1250 Waters Place, 11th Floor, Bronx, disruption, preservation of blood supply, and shorter
NY 10461, USA operative time.

The purpose of this study is to critically evaluate the from the articles: (1) number of acute, displaced proximal
outcomes following locked, antegrade IMN of acute, dis- humerus fractures treated with IMN, including the number
placed proximal humerus fractures reported in the literature of two-, three-, and four-part fractures; (2) mean patient
and present a concise systematic review. Specifically, we age; (3) mean, minimum, and range of follow-up; (4)
attempt to determine: (1) the demographics of patients who patient gender; (5) use of a surgical or nonoperative control
undergo IMN for two-, three-, and four-part proximal group; (6) mean time from injury to surgery; (7) functional
humerus fractures; (2) the outcomes following IMN for outcome scores; and (5) complications and reoperation
acute proximal humerus fractures, including functional rates.
scores and range-of-motion (ROM) data; (3) the rate and We identified 661 initial manuscripts using our search
types of complications following IMN for displaced terms (Fig. 1). Four hundred and eighty-four articles were
proximal humerus fractures; and (4) any difference in excluded following review of the article title because of
outcomes between two-, three-, and four-part proximal irrelevance to the study question. One hundred and thirty-
humerus fractures. three were then excluded following a review of the
abstract. Of these, 102 were either not written in English or
were review articles or case reports, and 31 failed the
Methods and materials above study criteria. This left 44 articles which required a
full-text review. Of these, 30 were excluded following a
We used the PubMed, EMBASE, ScienceDirect, and Web full-text review because they did not meet our study
of Science computerized literature databases to search all inclusion criteria. In particular, two studies reported on
years from the beginning of the database through April treatment for both acute and subacute fractures ([6 weeks
2014. Articles were retrieved by using the following key- from injury) [12, 13]. One of these did not provide out-
words: ‘‘intramedullary nailing proximal humerus,’’ ‘‘in- comes data specifically for acute fractures and was exclu-
tramedullary nailing proximal humerus fracture,’’ and ded [12]. In the other study, outcomes for acute fractures
‘‘proximal humeral nailing.’’ In addition to these keywords, could be determined, and it was included [13]. Two sets of
we utilized the medical subject heading (MeSH) ‘‘shoulder two articles [14–17] within the 44 full-text-reviewed
fractures’’ combined with ‘‘fracture fixation, intramedul- papers were from the same group of authors reporting on
lary’’ to maximize search specificity and sensitivity in the the same or similar cohorts of patients at a later time point.
PubMed database. In both cases, the more recent article was included [15, 17].
Inclusion criteria for studies in this systematic review Three studies additionally reported on outcomes fol-
included published studies that: (1) were written in the lowing IMN of fracture dislocations [18–20], and three
English language; (2) had a minimum clinical follow-up of studies additionally reported on outcomes following IMN
12 months; (3) reported on the use of antegrade IMN for of combined proximal humerus and shaft fractures [11, 21,
acute two-, three-, and four-part proximal humerus frac- 22]. Only two of these studies were included, as explicit
tures; (4) utilized at least one validated outcome measure; outcomes data for acute proximal humerus fractures could
and (5) had C10 patients for review. Exclusion criteria be elicited [11, 19]. Finally, one study reported on acute
included studies that: (1) were review articles, case reports, fractures, fracture dislocations, and combined proximal
or technical papers; (2) provided combined outcomes data humerus and shaft fractures [23]. Demographic and out-
for fracture-dislocations and/or proximal humeral fractures comes data were extracted for the acute fracture subgroup
with diaphyseal extension without individual data for acute only. A total of 14 articles were ultimately included [7, 8,
fractures; (3) involved the use of flexible intramedullary 11, 13, 15, 17, 19, 23–29].
devices; (4) involved fractures resulting from bony Functional outcomes were measured using a variety of
metastasis; and (6) did not explicitly report a minimum scoring systems: American Shoulder and Elbow Surgeons
12-month follow-up. Finally, the reference lists of all the (ASES) [8, 27, 29], Constant score (CS) [2, 7, 8, 17, 26,
full-text papers were reviewed to identify any additional 29], modified Constant score (mCS) [7, 17], relative Con-
studies that met the stated inclusion criteria. stant score (rCS) [19, 20, 26], Neer score [11, 13, 23, 24,
The search strategy was independently implemented by 27], relative Neer score (rNeer) [26], Simple Shoulder Test
two of the authors to select references from the above- (SST) [8], Oxford Shoulder Score (OSS) [17], Japanese
mentioned databases. Disagreement between two inde- Orthopaedic Association shoulder score (JOA) [25], and
pendent reviewers was resolved by consensus and arbitra- Shoulder Pain and Disability Index (SPADI) [28]. The JOA
tion of the senior author. The article titles and abstracts score is categorized as excellent (85–95 points), satisfac-
were screened according to the eligibility criteria. The full tory (75–84 points), unsatisfactory (65–74 points), and
texts of the articles that met the inclusion criteria were poor (\65 points) [25]. The SPADI score is categorized as
thoroughly reviewed. The following data were extracted excellent (0–25), good (26–50), fair (51–75), and poor

Outcomes of intramedullary nailing for acute proximal humerus fractures: a systematic review 33
Fig. 1 Flow diagram outlining

Unique references identified by title, abstract, keyword (n=661)
the systematic review process
used in this study Combined Pubmed, EMBASE, ScienceDirect, and Webscience (n=661)
Excluded based on title (n=484)
Failed criteria (n=484)
Excluded based on abstract (n= 133)
Not English, letter, editorial, review article (n=

102)
Excluded based on full text (n=30)
Letter, editorial, review article (n=0)

Duplicate patient series (n=2)
Total included references (n=14)
Added from manual reference search (n=0)
(76–100) [28]. Relative scores refer to the comparison of operated on between 1993 and 2007, although one study
the operative shoulder with the contralateral, noninjured did not provide this information [28]. All of the studies
shoulder, expressed as a percentage. utilized either inclusion or exclusion criteria. Four studies
Ten of the studies reported the use of statistical analysis were performed across multiple institutions [7, 15, 19, 26].
[7, 8, 15, 17, 19, 23, 24, 26, 27, 29]. Each of the selected Nine studies reported on the number of surgeons involved,
studies contributed data to the patient demographics. In si- with a mean of 2.4 amongst those studies [7, 8, 11, 17, 23,
tuations where more than one study provided data for any 26–29].
of the outcome measures, the data were pooled and fre-
quency-weighted means were calculated. The frequency- Demographics
weighted mean represents the mean from each individual
study weighted by the number of patients in that study. A Demographic data are provided in Table 1. There were a
standard Student t test was used to compare the frequency- total of 529–563 patients in the 14 studies at baseline. The
weighted means for the demographic and outcomes data. p- final patient total was 448. Using the Neer classification,
Value \0.05 was considered statistically significant. 185 (41 %) had two-part fractures, 230 (51 %) had three-
part fractures, and 33 (13 %) had four-part fractures. The
frequency-weighted mean age was 64.3 years. All studies
Results except for one reported on patient gender [11], with 338
females (71 %) and 139 males (29 %). Two studies pro-
Nine studies were published as level IV evidence [7, 8, 11, vided only baseline age and gender demographics; there-
13, 23, 28], three as level III evidence [17, 19, 27], and one fore, the gender tally adds up to more than the final study
each as level II [15] and level I evidence [29]. Three patient number [15, 19]. Only four studies reported the
studies included a comparative operative group using a dominance of the operated extremity [8, 15, 28, 29], with
locked proximal humeral plate [15, 17, 29]. No study 50 % involving the dominant shoulder. The frequency-
included a nonoperative control group. The patients were weighted mean follow-up was 22.6 months (range

Table 1 Demographic and operative details of included studies

Authors Publication Type of No. of patients Neer fracture type Mean Female/ Mean Minimum
date study age male follow-up follow-up
Baseline Final Two- Three- Four- (years) (months) (months)
part part part
Hatzidakis et al. [7] 2011 Retrospective 48 38 38 0 0 65 28/10 20 12

Nolan et al. [8] 2011 Retrospective 18 13 9 4 0 71 10/3 42 24
Adedapo et al. [11] 2001 Retrospective 16 16 0 10 6 67.1 NR 12 12
Lin et al. [13] 1998 Retrospective 16–18 16 6 0 0 67.6 10/6 19.6 14
Konrad et al. [15] 2012 Prospective 58a 47 0 47 0 64.8 447/11 12 12
Trepat et al. [17] 2012 Retrospective 15 13 3 0 0 64.5 7/6 12 12
Gradl et al. [19] 2007 Prospective 69–96b 69 7 35 17 67.2 67/36 12 12
Lin et al. [23] 2006 Prospective 22–27 22 0 22 0 53.3 10/12 23.9 19
Kazakos et al. [24] 2007 Retrospective 31 27 16 11 0 65.9 17/10 12 12
Koike et al. [25] 2008 Retrospective 54 54 29 22 3 66 44/10 18 13
Linhart et al. [26] 2007 Retrospective 97 51 5 31 5 68.4 39/12 12 12
Park et al. [27] 2012 Retrospective 43 43 0 43 0 60.2 34/9 65 35
Sforzo et al. [28] 2009 Retrospective 14 14 7 5 2 56 9/5 40 12
Zhu et al. [29] 2011 Prospective 28 25 5 0 0 54.8 16/9 36 36
Totals 529–563 448 185 230 33 338/139c
Frequency-weighted 64.3 22.6 16.4
mean
NR, not reported
a
Only baseline demographic data provided
b
Four four-part fracture dislocations and 12 fractures with shaft extension excluded. Mean age and follow-up reported for entire baseline cohort
of 112 patients, but weighted for 69 patients
c
Total does not add up to total patient number given (b)
12–65 months), and the mean minimum follow-up was Functional outcomes
16.4 months. Six studies explicitly reported on the mean
time from injury to surgery, with a frequency-weighted Functional outcomes are reported in Table 2. Three studies
mean of 4.9 days (range 1.1–9.4 days) [13, 23, 25–27, 29]. (n = 81) reported ASES scores, with a frequency-weighted
mean of 84.3 [8, 27, 29]. Seven studies (n = 225) reported
Surgical technique the CS, with a frequency-weighted mean of 72.8 [7, 8, 11,
17, 19, 26, 29]. The rCS was reported in three studies
Nine studies reported performing the surgery in the beach- (n = 167), with a frequency-weighted mean of 81.4 [15,
chair position [8, 13, 17, 19, 23–26, 28], one used the 19, 26]. Two studies (n = 51) reported the mCS, with a
lateral position [15], two used the supine position [11, 29], frequency-weighted mean of 91.9 [7, 17]. Five studies
and one used either the beach-chair or lazy lateral position (n = 124) reported a Neer score, with a frequency-
[7]. One study did not report patient positioning [27]. weighted mean of 84.5 [11, 13, 23, 24, 27]. One study
While all studies (406 patients) utilized a deltoid-splitting (n = 51) reported a rNeer score of 84.7 [26]. One study
approach, one predominately used the deltopectoral (n = 13) reported a mean OSS score of 19.8 [17]. One
approach (42 patients) [27]. The most commonly used study (n = 54) reported a mean JOA score of 81.0 [25].
intramedullary device was the Polarus nail (Acumed, One study (n = 13) reported a mean SST score of 6.8 [8].
Beaverton, OR, USA) (36 %). Other intramedullary nails One study (n = 14) reported a mean SPADI score of 30
included the Targon PH nail (Aesculap, Tuttlingen, Ger- [28].
many) (27 %), Synthes PHN (Synthes GmbH, Oberdorf,
Switzerland) (16 %), Stryker T2 nail (Stryker Orthopae- Pain outcomes
dics, Mahwah, NJ, USA) (9 %), Humeral Locked Nail
(United, Taipei, Taiwan) (8 %), Synthes EX spiral blade Pain scores were reported as a component of either the CS
(Synthes, West Chester, PA, USA) (2 %), and the Uniflex [7] or the Neer score [23], visual analogue scale (VAS) [8,
humeral nail (Biomet, Warsaw, IN, USA) (1 %). 29], or subjectively as mild, moderate or severe [11, 17].

Table 2 Postoperative outcomes
Authors Postoperative scores (mean) Pain scores Range of motion
ASES CS rCS mCS Neer CS JOA VAS Neer None Mild Mod Severe FE Abd Ext ER IR (°)
(0–15) (0–30) (0–35) (°) (°) (°) (°)
Hatzidakis et al. [7] — 71.0 — 97.0 — 13.0 — — — — — — — 132.0 — — — —

Nolan et al. [8] 67.8 60.5 — — — — — 2.3 — — — — — 120 — — 38.8 T4 (1), T9 (2), T10 (1),
T12 (1), L1 (2), L2
(2), L4 (3), L5 (1)
Adedapo et al. [11] — 80.4 — — 75.7 — — — — 8.0 5.0 1.0 2.0 143.8 141.9 — 48.8 75.0
Lin et al. [13] — — — — 87.2 — — — — — — — — 153.8 154.7 36.3 48.1 65.3
Konrad et al. [15] — — 89.0 — — — — — — — — — — — — — — —
Trepat et al. [17] — 63.8 — 76.9 — — — — — 8 3 1 1 131 121.5 — 38.8 —
Gradl et al. [19] — 69.9 75.6 — — — — — — — — — — — — — — —
Lin et al. [23] — — — — 85.3 — — — 32.7 — — — — 137.3 133.9 32 39.3 56.6
Kazakos et al. [24] — — — — 83.7 — — — — — — — — — — — — —
Koike et al. [25] — — — — — — 26 — — — — — — — — — — —
Linhart et al. [26] — 71.2 82.1 — — — — — — — — — — — — — — —
Park et al. [27] 86.0 — — — 87 — — — — — — — — — — — — —
Sforzo et al. [28] — — — — — — — — — — — — — 105.6 — — 36.5 —
Zhu et al. [29] 90.0 93.3 — — — — — 0 — — — — — 160.8 — — 47.8 T8
Outcomes of intramedullary nailing for acute proximal humerus fractures: a systematic review
Total (no. of 16.0 8.0 2.0 3.0

patients)
Frequency-weighted 84.3 72.8 81.4 91.9 84.5 13.0 26.0 0.8 32.7 137.3 138.4 33.8 43.1
mean
Standard deviation 7.4 8.2 5.5 8.8 3.7 15.6 11.3 2.1 4.9
ASES American Shoulder and Elbow Surgeons, CS Constant score, rCS relative Constant score, mCS modified Constant score, JOA Japanese Orthopaedic Association, VAS visual analogue
scale, FE forward elevation, Abd abduction, Ext extension, ER external rotation, IR internal rotation

35
One study (n = 38) reported a mean Constant pain score of Complications

13.0 [7]. This score is given on a scale of 0–15, where a
score of 15 represents no pain. One study (n = 22) All studies reported both radiographic outcomes and com-
reported a mean Neer pain score of 32.7 on a scale of 0–35, plications following IMN of acute, displaced proximal
where a score of 35 represents no pain [23]. Two studies humerus fractures. The overall radiographic healing rate in
reported a mean VAS score using a scale of 0–10, with a this series was 99.3 % (445/448). Only five studies explicitly
frequency-weighted mean of 0.8 [8, 29]. Two studies reported radiographic parameters for fracture malunion [8,
(n = 29) reported pain as either mild, moderate or severe 17, 25, 27, 28], four studies explicitly reported a definition of
[11, 17]. Sixteen patients reported no pain, eight had mild nonunion [8, 17, 25, 28], and eight studies formally mea-
pain, two had moderate pain, and three had severe pain. sured the final radiographic neck–shaft angle (NSA) [7, 8,
15, 17, 23, 25, 27, 28]. Five studies formally reported the
Range of motion loss of NSA during the postoperative period [7, 8, 17, 27],
and one study only reported patients with[10° loss of NSA
Range-of-motion outcomes were reported in eight studies [29]. One study (n = 16) reported on complications of a
[7, 8, 11, 13, 17, 23, 28, 29]. These included active forward patient group which included proximal humeral fractures
elevation [7, 8, 11, 13, 17, 23, 28, 29], abduction [11, 13, with shaft extension [11]. Furthermore, one study (n = 47)
17, 23], external rotation at the side [8, 11, 13, 17, 23, 28, included patients with \12 months follow-up in their report
29], extension [13, 23], and hand-in-back internal rotation on complications [15]. These patients were thus excluded
[8, 11, 13, 23, 29]. Eight studies (n = 141) reported a from analysis of the complications for acute proximal
frequency-weighted mean forward elevation of 137.3°. humeral fractures, leaving a total of 385 patients.
Four studies (n = 54) reported a frequency-weighted mean There were 61 (15.8 %) secondary surgeries, mostly for
active abduction of 138.4°. Two studies (n = 38) reported removal of migrated proximal screws. There were 160
a mean extension of 33.8°. Seven studies (n = 119) (41.5 %) reported complications, with 38 instances of sec-
reported active external rotation with a frequency-weighted ondary loss of reduction (10 %), 34 instances of screw
mean of 43.1°. Three studies (n = 54) evaluated internal migration or perforation into the joint (9 %), 33 instances of
rotation based on arc of motion with a frequency-weighted malunion (9 %), 14 instances of avascular necrosis (4 %),
mean of 64.6° [11, 13, 23], whereas two (n = 38) reported and 13 instances of subacromial impingement (4 %). The
values for maximum hand-in-back internal rotation, with a remaining complications are depicted in Fig. 2. No case of
range of T2 through buttock [8, 29]. nerve or vascular injury was reported in this patient series.
Fig. 2 Summary of
complications

Table 3 Subgroup outcomes analysis

Outcome Two-part fracture Three-part fracture Four-part fracture p-Value
(mean ± SD) N (mean ± SD) N (mean ± SD) N Two versus Two versus Three versus
three four four
Forward elevation 140.4 ± 29.2° 95 143.0 ± 35.8° 38 96.0 ± 33.1° 8 0.7 \0.0001 0.001
Abduction 154.7 ± 30.6° 16 140.5 ± 31.8° 32 120.0 ± 20.0° 6 0.2 0.04 0.1
External rotation 45.4 ± 17.7° 57 44.0 ± 16.7° 38 33.8 ± 18.3° 8 0.7 0.1 0.09
Constant score 74.3 ± 17.7 104 75.5 ± 15.6 127 63.6 ± 19.9 27 0.6 0.007 0.0009
Neer score 87.2 ± 7.7 16 85.3 ± 9.9 122 62.5 ± 11.5 6 0.5 0.0001 0.0001
ASES score 85.5 ± 17.2 34 83.5 ± 13.5 47 na na 0.6 na na
ASES American Shoulder and Elbow Surgeons, na not applicable
Outcomes and complications by fracture pattern 63.2 % was found. The reoperation rate for both two- and
three-part fractures was significantly less than for four-part
Further analysis was performed to stratify the functional fractures [p \ 0.0002, 95 % CI (22, 64 %) and (26, 68 %),
outcomes and complications by fracture pattern. The respectively]. There were 29 reported complications
results are given in Table 3. The frequency-weighted amongst the four-part fracture cohort.
ASES score for two-part fractures was found not to be
statistically different from the score for three-part fractures
[85.5 versus 83.5, p = 0.6, 95 % CI (-4.8, 8.8)]. The Discussion
frequency-weighted mean CS for two- and three-part
fractures were statistically greater than for four-part frac- Proximal humerus fractures remain common amongst the
tures [p = 0.007, 95 % CI (2.9, 18.5) and p = 0.0009, elderly, behind fractures of the hip and distal radius [4].
95 % CI (5.0, 18.8), respectively], but were not different Complex three- and four-part fractures account for [50 %
from each other. The frequency-weighted mean forward of cases in patients older than 60 years [3, 4]. Despite their
elevation for two- and three-part fractures were greater prevalence, optimal treatment remains controversial. Zyto
than for four-part fractures [p \ 0.0001, 95 % CI (22.9, et al. found no functional difference between tension band
65.9) and p = 0.001, 95 % CI (19.3,74.7), respectively], fixation versus nonsurgical management of displaced three-
but were not statistically different from each other and four-part proximal humerus fractures at 1 and 3–5 year
[p = 0.7, 95 % CI (-14.4, 9.4)]. The frequency-weighted follow-up [30]. Others have shown comparable clinical
mean abduction was greater amongst two-part fractures results for four-part fractures treated with hemiarthroplasty
versus four-part fractures; however, no significant differ- or nonoperatively [31]. Reverse shoulder arthroplasty may
ences were found in abduction and external rotation play a role in elderly patients with complex three- and four-
between three- and four-part fractures. part proximal humerus fractures [10, 32]. Percutaneous
Complication data and reoperation rates for seven studies fixation has demonstrated good midterm results, although
(n = 125) involving two-part fractures [7, 8, 13, 17, 19, 28, avascular necrosis (AVN) rates are high [5]. While locked
29] and for five studies (n = 109) for three-part fractures [8, proximal humerus plating has become popular [6, 33, 34],
19, 23, 27, 28] were analyzed. Only two studies (n = 19) the complication and reoperation rates remain high [34]. As
explicitly reported complications for four-part fractures [19, such, multiple operative techniques have been recom-
28]. The complication rate for two-part fractures was mended for treatment of displaced fractures, including
33.6 %, and the reoperation rate was 13.6 %. For three-part percutaneous fixation [5], ORIF [6, 34], IMN [7, 8],
fractures, the complication rate was 57.8 % with a reoper- hemiarthroplasty [9], and RSA [10]. Intramedullary nailing
ation rate of 17.4 %. There was no significant difference in has been reported to provide clinical outcomes comparable
reoperation rate (p = 0.5) between two- and three-part to locking plates [15, 19, 29], with less soft tissue dissec-
fractures, though two-part fractures had a significantly lower tion and a possibly improved complication profile.
complication rate [p = 0.0002, 95 % CI (11, 36 %)]. In In the current systematic review on the outcomes of
both instances, secondary fracture displacement/fracture IMN for displaced proximal humerus fractures, we found a
malunion accounted for the majority of complications, fol- frequency-weighted mean patient age of 64.3 years. The
lowed by screw migration/glenohumeral joint penetration. majority of patients were female, corresponding to the
Despite the small number of four-part fractures for which gender that most frequently sustains this fracture. The most
complications could be assessed, a reoperation rate of common fracture pattern treated was three-part fractures,

followed by two-part fractures. Few studies included four- occur beyond 12 months, and our mean follow-up was
part fractures, likely reflecting the technical difficulty in 22.6 months. Loss of reduction can be associated with
utilizing IMN for this pattern, in addition to the number of proximal screw migration or screw penetration into the
alternative implants available. A systematic review of glenohumeral joint. Malunion has been associated with
locking plates for proximal humerus fractures similarly poor clinical outcomes [8, 18]. Despite its correlation with
found that three-part fractures were most commonly trea- outcome, only five studies reported the loss of NSA and
ted, followed by two-part fractures [34]. Their mean age only eight of the studies reported a final NSA. Park et al.
was 62 years, with a majority of patients being female. In demonstrated the importance of restoring and supporting
comparison, patients undergoing RSA for displaced prox- the medial calcar with a screw [27]. The authors also uti-
imal humerus fractures tend to be older with the majority of lized tension band sutures and noted improved radio-
cases involving four-part fractures [32]. In a comparative graphic and clinical outcomes. The benefits of fracture
study of HA versus RSA, the mean age for patients augmentation using calcium sulfate cement and placement
undergoing HA was 74.1 and 74.8 years for RSA [35]. In of inferomedial screws have been demonstrated with
younger, more active patients, operative fixation of dis- locking plates and may play a role with intramedullary
placed proximal humerus fractures has generally been nailing [36, 37]. In our study, the overall AVN incidence
favored over the use of shoulder arthroplasty. was 4 %, lower than that reported by Sproul et al. for
We determined a frequency-weighted mean CS of 72.8 locked plating (10.8 %) [34] and Harrison et al. for per-
and ASES score of 84.3. The CS is comparable to the score cutaneous pinning (26 %) [5]. This finding may be
of 73.6 reported in a systematic review of locked plating expected given the less invasive insertion of IMN, though it
for proximal humerus fractures [34]. Additionally, the may not present radiographically for years [5]. Kloub and
authors found the score for four-part fractures (67.7) to be colleagues found that reduction quality influenced AVN
statistically worse than for two-part fractures (77.4). The development, with 2 % complete necrosis following
CS for two-part fractures (74.3) in the current study was excellent reduction compared with 60 % complete necrosis
statistically better than for four-part fractures (63.6), but it following poor reduction [20]. AVN was associated with
was not different from three-part fractures (75.5). Similar worse clinical outcome. Perhaps not unexpectedly, com-
findings were made with respect to the Neer score, sug- plications amongst two-part fractures (33.6 %) were lower
gesting that the outcomes for two- and three-part fractures in our study than amongst three-part fractures (57.8 %).
treated with IMN are comparable. Our overall frequency- The majority of these complications were related to loss of
weighted mean forward elevation, abduction, and external reduction and malunion, similar to that reported for locked
rotation were 137.3°, 138.4°, and 43.1°, respectively. plating [6, 33, 34]. Future studies need to better stratify
Forward elevation for both two- and three-part fractures complications by injury pattern to better understand which
(140.4° and 143.0°, respectively) were significantly greater fractures would benefit from the use of IMN. Furthermore,
than for four-part fractures (96°), but no statistical differ- fracture augmentation strategies may potentially reduce
ence was noted for abduction and external rotation for postoperative loss of reduction.
three- versus four-part fractures. Forward elevation and Our overall reoperation rate of 15.8 % compares well
abduction reported by Sproul et al. were 98° and 103°, with that reported for locked plating. Sproul and colleagues
though based on only two studies [34]. Solberg et al. reported a reoperation rate of 13.8 %, mostly for screw
reported a higher CS following locked plating of three-part penetration [34]. Most of the reoperations in our study were
proximal humerus fractures compared with hemiarthro- related to proximal screw migration. Proximal screw
plasty, though this difference was not significant for four- migration may be decreased through the use of an end-cap
part fractures [6]. Harrison and colleagues reported a mean as well as threaded bushings within the nail to minimize
ASES score of 82 at a minimum of 3 years following screw back-out [7, 29]. The nail should be inserted at least
closed reduction percutaneous pinning of predominately 5 mm below the subchondral bone of the humeral head and
three- and four-part fractures. The mean forward elevation proximal screw lengths fluoroscopically verified [8, 18]. A
was 140° and external rotation was 41°, though no analysis more medial articular entry point may cause less damage to
was performed by fracture pattern [5]. Our findings of fair the rotator cuff as the more medial aspect of rotator cuff
clinical outcomes and ROM following IMN of four-part has more vascularity [7, 8, 18], though rotator cuff symp-
fractures emphasize the need for further study into the use toms can persist despite meticulous repair [8]. The reop-
of fracture fixation versus arthroplasty in this difficult eration rate for two-part and three-part fractures in our
patient group. study was 13.6 and 17.4 %, respectively. As these result
The most common complication associated with IMN in primarily from screw back-out and fracture displacement,
this review was secondary loss of reduction (24 %) fol- meticulous surgical technique and newer implant designs
lowed by fracture malunion (21 %). Many complications may reduce their incidence. Given the high reoperation rate

amongst four-part fractures (63.2 %), caution should be Compliance with ethical standards
exercised in selecting IMN for this fracture pattern and
Conflict of interest The authors declare that they have no conflicts
preoperative discussions should be held with the patient of interest.
regarding this risk.
There are several limitations of this systematic review, Informed consent This article does not contain any studies with
primarily related to the inherent limitations of the studies human participants or animals performed by any of the authors.
Informed consent was not applicable to the preparation of this
on which this review was based. There were 10 retro- manuscript.
spective studies and 4 prospective studies with the majority
(64 %) published as level IV evidence. Additionally, a
variety of outcome measures and nonuniform assignment
of complications to each individual patient were used.
Future studies with standardized use of outcomes measures
and strict definitions for complications are needed. Despite
this, we were able to pool the data for the Constant, Neer,
and ASES scores to generate frequency-weighted means.
Two studies did not report demographic data for acute
fractures only, and their age and gender distribution were References
weighted only for the cohort of interest. This may have
1. Aaron D, Shatsky J, Paredes JC, Jiang C, Parsons BO, Flatow EL
affected the reported mean age, although the gender dis- (2012) Proximal humeral fractures: internal fixation. J Bone Joint
tribution was similar to the remaining studies. Complica- Surg Am 94(24):2280–2288
tion data were often reported for the entire cohort of 2. Bell JE, Leung BC, Spratt KF, Koval KJ, Weinstein JD, Good-
man DC et al (2011) Trends and variation in incidence, surgical
patients, including instances other than for acute fracture.
treatment, and repeat surgery of proximal humeral fractures in the
We therefore excluded complication data unless explicitly elderly. J Bone Joint Surg Am 93(2):121–131
reported for the acute fractures, and this may have affected 3. Court-Brown CM, Garg A, McQueen MM (2001) The epidemi-
the calculated complication profile following IMN. Finally, ology of proximal humeral fractures. Acta Orthop Scand
72(4):365–371
multiple intramedullary nail designs with differing proxi-
4. Kannus P, Palvanen M, Niemi S, Parkkari J, Jarvinen M, Vuori I
mal screw insertion configurations and the variable use of (1996) Increasing number and incidence of osteoporotic fractures
an endcap were included, which may have overestimated of the proximal humerus in elderly people. BMJ 313(7064):
the complications. Our review was strengthened by the 1051–1052
5. Harrison AK, Gruson KI, Zmitzkowski B, Keener J, Galatz L,
final patient number and the weighted mean follow-up of
Williams G et al (2012) Intermediate outcomes following per-
22.6 months with a minimum of 16.4 months, which is cutaneous fixation of proximal humerus fractures. J Bone Joint
adequate for a study involving fractures. Additionally, we Surg Am 94(13):1223–1228
were able to perform a subgroup analysis amongst the 6. Solberg BD, Moon CN, Franco DP, Paiement GD (2009) Surgical
treatment of three and four-part proximal humeral fractures.
fracture patterns for ROM and functional outcomes. This
J Bone Joint Surg Am 91(7):1689–1697
study represents the first systematic review that the authors 7. Hatzidakis AM, Shevlin MJ, Fenton DL, Curran-Everett D,
are aware of assessing the use of intramedullary nailing for Nowinski RJ, Fehringer EV (2011) Angular-stable locked intra-
acute proximal humerus fractures. medullary nailing of two-part surgical neck fractures of the
proximal part of the humerus. J Bone Joint Surg Am 93(23):
Intramedullary nailing of acute two- and three-part
2171–2179
proximal humerus fractures yields satisfactory clinical and 8. Nolan BM, Kippe MA, Wiater JM, Nowinski GP (2011) Surgical
functional results. Only fair clinical and functional results treatment of displaced proximal humerus fractures with a short
were reported for four-part fractures, suggesting that fur- intramedullary nail. J Shoulder Elbow Surg 20:1241–1247
9. Olerud P, Ahrengart L, Ponzer S, Saving J, Tidermark J (2011)
ther studies with larger patient numbers are needed to
Hemiarthroplasty versus nonoperative treatment of displaced
determine the role of intramedullary nailing for four-part 4-part proximal humeral fractures in elderly patients: a random-
fractures. Further, the complication profile and reoperation ized controlled trial. J Shoulder Elbow Surg 20(7):1025–1033
rate, in particular loss of reduction, remain high regardless 10. Bufquin T, Hersan A, Hubert L, Massin P (2007) Reverse
shoulder arthroplasty for the treatment of three- and four-part
of fracture pattern. Newer implant designs and use of
fractures of the proximal humerus in the elderly: a prospective
augmentation techniques (i.e., rotator cuff sutures, trical- review of 43 cases with a short-term follow-up. J Bone Joint Surg
cium phosphate cement) may reduce its incidence. Br 89(4):516–520
11. Adedapo AO, Ikpeme JO (2001) The results of internal fixation
Acknowledgment None of the authors have a financial or propri- of three- and four-part proximal humeral fractures with the
etary interest in the subject matter or materials discussed in the Polarus nail. Injury 32(2):115–121
manuscript, including (but not limited to) employment, consultancies, 12. Giannoudis PV, Xypnitos FN, Dimitriou R, Manidakis N,
stock ownership, honoraria, and paid expert testimony. Hackney R (2012) Internal fixation of proximal humeral fractures

using the Polarus intramedullary nail: our institutional experience 26. Linhart W, Ueblacker P, Grossterlinden L, Kschowak P, Briem
and review of the literature. J Orthop Surg Res 7:39 D, Janssen A et al (2007) Antegrade nailing of humeral head
13. Lin J, Hou SM, Hang YS (1998) Locked nailing for displaced fractures with captured interlocking screws. J Orthop Trauma
surgical neck fractures of the humerus. J Trauma 45(6):1051–1057 21(5):285–294
14. Blum J, Hansen M, Muller M, Rommens PM, Matuschka H, 27. Park JY, Kim JH, Lhee SH, Lee SJ (2012) The importance of
Olmeda A et al (2009) Proximal humeral fractures and intrame- inferomedial support in the hot air balloon technique for treat-
dullary nailing: experience with a new nail system. Eur J Trauma ment of 3-part proximal humeral fractures. J Shoulder Elbow
Emerg Surg 35:489–498 Surg 21:1152–1159
15. Konrad G, Audige L, Lambert S, Hertel R, Sudkamp NP (2012) 28. Sforzo CR, Wright TW (2009) Treatment of acute proximal humerus
Similar outcomes for nail versus plate fixation of three-part fractures with a Polarus nail. J Surg Orthop Adv 18(1):28–34
proximal humeral fractures. Clin Orthop Relat Res 470:602–609 29. Zhu Y, Lu Y, Shen J, Zhang J, Jiang C (2011) Locking intra-
16. Popescu D, Fernandez-Valencia JA, Rios M, Cune J, Domingo A, medullary nails and locking plates in the treatment of two-part
Prat S (2009) Internal fixation of proximal humerus fractures proximal humeral surgical neck fractures: a prospective ran-
using the T2-proximal humeral nail. Arch Orthop Trauma Surg domized trial with a minimum of 3 years of follow-up. J Bone
129:1239–1244 Joint Surg Am 93:159–168
17. Trepat AD, Popescu D, Fernandez-Valencia JA, Cune J, Rios M, 30. Zyto K, Ahrengart L, Sperber A (1997) Treatment of displaced
Prat S (2012) Comparative study between locking plates versus proximal humeral fractures in elderly patients. J Bone Joint Surg
proximal humeral nail for the treatment of 2-part proximal Br 79(3):412–417
humeral fractures. Eur J Orthop Traumatol 22:373–379 31. Boons HW, Goosen JH, van Grinsven S, van Susante JL, van
18. Cuny C, Scarlat MM, Irrazi M, Beau P, Wenger V, Ionescu N Loon CJ (2012) Hemiarthroplasty for humeral four-part fractures
et al (2008) The telegraph nail for proximal humeral fractures: a for patients 65 years or older: a randomized controlled trial. Clin
prospective 4 year study. J Shoulder Elbow Surg 17:539–545 Orthop Relat Res 470:3483–3491
19. Gradl G, Dietze A, Arndt D, Beck M, Gierer P, Borsch T et al 32. Anakwenze OA, Zoller S, Ahmad CS, Levine WN (2014) Rev-
(2007) Angular and sliding stable antegrade nailing (Targon PH) erse shoulder arthroplasty for acute proximal humerus fractures: a
for the treatment of proximal humeral fractures. Arch Orthop systematic review. J Shoulder Elbow Surg 23:e73–e80
Trauma Surg 127:937–944 33. Solberg BD, Moon CN, Franco DP, Paiement GD (2009) Locked
20. Kloub M, Holub K, Polakova S (2014) Nailing of three- and four- plating of 3- and 4-part proximal humerus fractures in older
part fractures of the humeral head—long-term results. Injury patients: the effect of initial fracture pattern on outcome. J Orthop
455:S28–S37 Trauma 23(2):113–119
21. Georgousis M, Kontogeorgakos V, Kourkouvelas S, Badras S, 34. Sproul RC, Iyengar JJ, Devcic Z, Feeley BT (2011) A systematic
Georgaklis V, Badras L (2010) Internal fixation of proximal review of locking plate fixation of proximal humerus fractures.
humerus fractures with the Polarus intramedullary nail. Acta Injury 42(4):308–413
Orthop Belg 76:462–467 35. Cuff DJ, Pupello DR (2013) Comparison of hemiarthroplasty and
22. Russo R, Cautiero F, Lombardi LV, Visconti V (2009) Telegraph reverse shoulder arthroplasty for the treatment of proximal
antegrade nailing in the treatment of humeral fractures with humeral fractures in elderly patients. J Bone Joint Surg Am
rotator interval split technique. Musculoskelet Surg 93:S7–S14 95(22):2050–2055
23. Lin J (2006) Effectiveness of locked nailing for displaced three- 36. Egol KA, Sugi MT, Ong CC, Montero N, Davidovitch R,
part proximal humeral fractures. J Trauma 61(2):363–374 Zuckerman JD (2012) Fracture site augmentation with calcium
24. Kazakos K, Lyras DN, Galanis V, Verettas D, Psillakis I, phosphate cement reduces screw penetration after open reduc-
Chatzipappas Ch et al (2007) Internal fixation of proximal tion-internal fixation of proximal humeral fractures. J Shoulder
humerus fractures using the Polarus intramedullary nail. Arch Elbow Surg 21(6):741–748
Orthop Trauma Surg 127:503–508 37. Zhang L, Zheng J, Wang W, Lin G, Huang Y, Zheng J et al
25. Koike Y, Komatsuda T, Sato K (2008) Internal fixation of (2011) The clinical benefit of medial support screws in locking
proximal humeral fractures with a Polarus humeral nail. J Orthop plating of proximal humerus fractures: a prospective randomized
Traumatol 9:135–139 study. Int Orthop 35(11):1655–1661

6
Mid-term results of Miller2Galante unicompartmental knee
replacement for medial compartment knee osteoarthritis
Hemanth Kumar Venkatesh1 • S. S. Maheswaran2
Abstract Conclusion This study confirms that fixed-bearing UKR

Background The purpose of this study is to analyse and gives excellent results in patients with medial compartment
report the mid-term results of 175 unicompartmental knee knee arthritis who comply with the inclusion criteria. Age
replacement (UKR) procedures performed for medial and BMI were not considered to influence the clinical and
compartment knee arthritis from January 2001 to January functional outcomes.
2010. Level of evidence-III.
Materials and methods The cohort participants were
selected after stringent inclusion criteria and the average Keywords Arthritis Medial compartment knee Fixed
follow-up was 5.6 years (range 2–10 years). The fixed- bearing UKR
bearing UKR procedure was carried out on all patients.
Results The pre-operative mean knee range of movement
improved from 100° ± 11.3° to 118.3° ± 12° (p value Introduction
\0.001). The pre-operative mean Knee Society (KS) knee
and functional score improved from 47 ± 5.5 and The degenerative changes in idiopathic osteoarthritis of the
55.1 ± 4.6 to 91.8 ± 9.2 and 92 ± 10.1 (p value \0.001), knee begin in the medial compartment in 80–90 % of
respectively. The revision rate of the cohort was 4 % patients [1–3]. This has given rise to the rationale for the
(seven knees) and implant survival rate was 96 % at the treatment of only one compartment, either with a high
end of 10 years; 87 % of the cohort were satisfied with the tibial osteotomy (HTO) or a unicompartmental knee
procedure and had a normal gait pattern. In this study, there replacement (UKR). Improved mid- and long-term results
was no statistical difference between groups with a body of UKR, comparable with the excellent and well-known
mass index (BMI) B30 kg/m2 and those with a results after total knee replacement (TKR), have con-
BMI C30 kg/m2, and between groups aged B55 years and tributed to the use of UKR on younger, active, and obese
those aged C55 years, in clinical and functional outcome populations [4].
following UKR. For the past 20 years, the overall results of UKR have
been promising, and this procedure is especially appro-
priate for anteromedial osteoarthritis of the knee [5–7].
& Hemanth Kumar Venkatesh UKR is less invasive, causes less blood loss, and preserves
[email protected] the bone stock and almost normal knee kinematics in
S. S. Maheswaran comparison with TKR [8–11].
[email protected] The purpose of this study was to evaluate the mid-term
1 results of the fixed-bearing metal-backed Miller-Galante
Department of Orthopedics, Basildon and Thurrock
University Hospital, Nethermayne, Basildon, prosthesis implanted in 148 patients with medial com-
Essex SS165NL, UK partmental osteoarthritis and also to evaluate the functional
2
Department of Orthopedics, University Hospital of North outcome in groups aged B55 years and C55 years, and in
Tees & Hartlepool, Stockton On Tees TS19 8PE, UK groups with a body mass index (BMI) B30 kg/m2 and

C30 kg/m2. Our hypothesis was that UKR surgery

improves the clinical and functional outcome in patients
with medial compartment arthritis of the knee joint. The
criteria for selecting patients for UKR were thoroughly
analysed since much controversy exists about the correct
indications for this procedure.
The research population comprised 148 patients with 175

primary UKRs at the University Hospital of North Tees
and Hartlepool Trust, UK between January 2001 and Jan-
uary 2010.
Patients were selected after thorough clinical and radi-
ological evaluation, and only those with medial compart-
ment disease were selected after fulfilment of inclusion and
exclusion criteria. Inclusion criteria were patients aged Fig. 1 Post UKR, alignment angle AP view
C40 years, BMI \40 kg/m2, no pain at rest, medial com-
partment osteoarthritis (Ahlbäck radiological grades 3 or
4), intact anterior and posterior cruciate ligament (ACL and
PCL, respectively), flexion deformity B10°, cor-
rectable varus deformity B15°, and minimum 90° of knee
flexion (Figs. 1, 2).
Patients with active or suspected infection in the knee,
inflammatory arthritis (rheumatoid arthritis, gout, psoriatic
arthritis), a previous history of HTO, post-traumatic
arthritis, advanced osteoarthritis involving the lateral
compartment and lateral facet arthritis of the patel-
lafemoral (PF) joint were excluded from the study.
Complete radiological assessment was carried out
before surgery to assess the degree of deformity, and
severity of arthritis [3]. Standard weight-bearing antero-
posterior (AP) and lateral radiographs of the knee joint
were used in all patients. Varus and valgus stress views of
the knee joint were taken to confirm the presence of full-
thickness articular cartilage in the lateral compartment. A
skyline view of the patella was used to assess the PF joint
status. Mechanical axis and degree of varus deformity were
estimated by orthogonal views.
All the surgical procedures were performed by the
same senior surgeon (SSM), with metal-backed cemented- Fig. 2 Orthogonal view post UKR
fixed Miller-Galante (Zimmer, Warsaw, IN, USA) UKRs
in 175 knees. Of 175 primary UKRs, osteoarthritis of the
medial compartment knee was common and involved 147 meniscectomy, four patients had undergone open menis-
patients with 174 knee joints. Arthritis secondary to cectomy and one patient had undergone arthroscopic ACL
osteonecrosis of the medial femoral condyle was noted in reconstruction in the past prior to index surgery.
one patient. Bilateral staged UKR was performed in 27 The surgical findings and the status of the PF joint
patients. cartilage and lateral compartment cartilage were recorded
Of 175 patients, 72 (41 %) had undergone previous knee from surgical notes. Intra-operative and post-operative
surgeries—67 patients had undergone arthroscopic complications such as fracture, infection, bleeding and re-
debridement of the joint including partial medial surgery were also recorded.

Mid-term results of Miller–Galante unicompartmental knee replacement for medial compartment knee... 43
Surgical procedure was \5 % (p B 0.05). Paired data were analysed using a

paired t test. Levene’s test was used to statistically test for
All the procedures were performed under regional anaes- equality of variances.
thesia. Standard medial parapatellar approach from the Implant survivorship was calculated by constructing
upper pole of the patella to 1–2 cm distal to the joint line, survivorship tables and Kaplan–Meier survivorship analy-
proximal (1–2 cm) to the tibial tuberosity was performed. sis with 95 % confidence intervals. Revision of UKR to
The patella was never dislocated, only lateralized. The TKR is the final determinant of survivorship.
tibial resection was performed with an extra medullary
guide with a slope of 5°. Based on the tibial resection, the
distal femur was resected through a ‘spacer block’ that Results
allowed a cut parallel (in extension) to the tibial cut. The
femoral cutting block was then positioned for the posterior Of the 148 patients (86; 57.1 % male and 62; 42.1 %
femoral and oblique resections. The trial components were female), 89 knees were replaced on the right side and 86 on
positioned, and dynamic tests were performed to choose the left side and 27 patients underwent bilateral UKR
the polyethylene thickness. The flexion and extension gaps surgery. The average age at the time of index surgery was
were assessed, trying to obtain approximately 2 mm of 61.7 years (range 44–80 years), with a mean age of 62.7
laxity in both positions. Soft-tissue release was not neces- years for males and 60.3 years for females. The mean BMI
sary in any of the cases. Finally, the definitive components of the cohort at the time of index surgery was 29.2 kg/m2
were fixed with PalacosÒ cement (Zimmer). The thickness (range 21–38 kg/m2); the mean BMI for males was 29 and
of the polyethylene ranged from 8-12 mm. A periarticular 29.4 kg/m2 for females. The average follow-up of the
injection with local anaesthetic was given before implan- cohort was 5.6 years (range 2–10 years). The mean length
tation. An immediate full weight-bearing rehabilitation of stay in hospital following index surgery was 2.5 days
protocol was used for all the patients. The patients also (range 2–4 days) (Table 1).
received routine thromboprophylaxis with low-molecular- The mean pre-operative knee range of movement
weight heparin for 2 weeks post-operatively. (ROM) of the 175 UKRs improved from 110.5° (range
Clinical and functional evaluations were performed 80°–135°) to 118.3° (range 60–135°) at the final follow-up.
during post-operative follow-up at regular intervals of The mean difference was -7.9 with 95 % CI (-10.14,
3 months, 6 months, and 1 year and Knee Society Scores -5.4), p value \0.001 (Fig. 3).
(KSS) were used to compare the overall functional and The mean pre-operative KS knee score of the cohort
clinical results. Post-operative radiographs were assessed improved from 47 (range 34–62) to 91.8 (range 51–100) at
for alignment of the components, correction of deformity, the final follow-up. The mean difference was -44.83 with
and signs of loosening of the components. Clinical and 95 % CI (-46.44, -43.23), p value\0.001. The mean pre-
radiological assessments were performed at the final fol- operative KS functional score of the cohort improved from
low-up. A patient satisfaction survey was included at the 55.1 (range 45–65) to 92 (range 55–100) at the final follow-
final follow-up based on the ability to perform daily living up. The mean difference was -36.90 with 95 % CI (-38.5,
activities and no standard scoring systems were used for -35), p value \0.001 (Fig. 4).
the assessment. The results were rated as satisfactory, The independent t test found no statistical difference in
excellent and not satisfactory. knee clinical and functional scores between males and
Failure of the surgery was defined as the revision of females.
UKR to TKR due to any cause such as loosening of The mean BMI of the cohort was 29.2 kg/m2 (range
components, infection, pain or any other indications. 21–38 kg/m2). The sample size with BMI B30 kg/m2 was
117 (72 male, 55 female), and the sample size with BMI
Statistical analysis C30 kg/m2 group was 58 (28 male, 30 female).
Mean pre-operative KS knee scores were 47.4 for BMI
Data were entered into Microsoft Excel spread sheets and B30 kg/m2 and 46.2 for BMI [30 kg/m2. The mean dif-
statistical analysis performed using SPSS software (SPSS ference was 1.2 with 95 % CI (-0.55, 2.96),
Inc. version 18.0). All the scale variables were tested for p value = 0.832. The mean pre-operative knee functional
normality using Kolmogorov-Smirnov test. Patient scores were 55.3 for BMI B30 kg/m2 and 54.9 for BMI
demographics were described using means, standard devi- [30 kg/m2. The mean difference was 0.35 with 95 % CI
ations, and ranges. The independent t-test was used to (-1.12, 1.82), p value = 0.620.
compare KSS clinical and functional outcomes. The level Mean post-operative KS knee scores at recent follow-
for statistical significance was \5 %, i.e., the probability up for BMI B30 kg/m2 was 91.6 and 92.4 for BMI
that the difference measured could have been due to chance [30 kg/m2. The mean difference was -0.85 with 95 %

Table 1 Demographic
N Minimum Maximum Mean Standard deviation
statistics
Statistic Statistic Statistic Statistic Standard error Statistic
Age 175 44 80 61.71 0.651 8.617

Follow-up (months) 175 22 129 63.61 1.992 26.270
BMI 175 21 38 29.19 0.286 3.781
LOS (days) 175 2 4 2.59 0.148 1.723
Pre-operative ROM 175 80 135 110.98 0.956 11.630
Post-operative ROM 175 60 135 117.34 0.997 12.287
BMI Body mass index, LOS length of stay, ROM range of movement
CI (-3.77, 2.06), p value = 0.539. Mean post-operative

knee functional scores at recent follow-up for BMI
B30 kg/m2 were 91.6 and 92.7 for BMI [30 kg/m2. The
mean difference was -1.12 with 95 % CI (-4.48, 2.23),
p value = 0.703.
In the BMI B30 and C30 kg/m2 groups, there was no
statistically significant difference in KS clinical scores,
functional scores and knee ROM scores, (p value [0.05)
(Table 2).
The sample size for the group aged B55 years was 38
(17 male, 21 female). The sample size for the group aged
C55 years was 137 (83 male, 54 female).
Mean pre-operative KS knee score was 46.4 for the
group aged B55 years and 47.2 for the group aged
C55 years. The mean difference was -0.77 with 95 % CI
(-2.70, 1.17), p value = 0.809. Mean post-operative KS
Fig. 3 Error bar graph illustrating the knee range of movement of the knee score was 92.2 for the group aged B55 years and 91.7
cohort for the group aged C55 years. The mean difference was
0.42 with 95 % CI (-2.79, 3.63), p value = 0.539.
Mean pre-operative knee function score was 54.4 for the
group aged B55 years and 55.4 for the group aged
C55 years. The mean difference was -0.94 with 95 % CI
(-2.55, 0.66) p value = 0.285. Mean post-operative knee
function score was 91.3 for the group aged B55 years and
92.1 for the group aged C55 years. The mean difference
was -0.83 with 95 % CI (-4.53, 2.88), p value = 0.455.
This study infers no statistical significant difference in
KS clinical and functional outcomes between two age
groups (p value [0.05) (Table 3).
Patient satisfaction
At the latest follow-up, 45 % of the patients were enthu-

siastic regarding the procedure and 42 % patients were
satisfied with the results. Twelve patients underwent TKR
for the opposite side. They were very satisfied with the
UKR knee outcome and preferred UKR over TKR. Ten
percent of the patients were not satisfied with the procedure
Fig. 4 Error bar chart illustrating the knee scores of the cohort and 3 % of the patients did not reply.

Table 2 Statistics for BMI

BMI-kg/m2 N Mean Standard deviation Standard error mean
C30 and B30 kg/m2 groups
Pre-operative ROM B30 117 111.89 12.089 1.191
C30 58 108.89 10.329 1.540
Post-operative ROM B30 117 117.88 12.149 1.180
C30 58 116.00 12.774 1.904
Pre-operative KSS B30 117 49.00 5.894 0.672
C30 58 49.27 5.601 1.023
Post-operative KSS B30 117 91.94 9.830 0.955
C30 58 92.96 7.992 1.178
Pre-operative KFS B30 117 55.58 5.754 0.651
C30 58 55.00 4.355 0.795
Post-operative KFS B30 117 91.21 12.006 1.166
C30 58 91.98 10.012 1.476
BMI Body mass index, ROM range of movement, KSS knee society score, KFS knee functional score
Table 3 Statistics for groups

Age N Mean Standard deviation Standard error mean
aged C55 years/B55 years
Pre-operative KSS B55 38 49.37 5.718 1.312
C55 137 49.01 5.834 0.622
Post-operative KSS B55 38 91.38 8.988 1.541
C55 137 92.50 9.407 0.866
Pre-operative KFS B55 38 54.21 4.791 1.099
C55 137 55.67 5.497 0.583
Post-operative KFS B55 38 90.15 12.522 2.147
C55 137 91.81 11.101 1.022
Pre-operative ROM B55 38 112.34 12.047 2.130
C55 137 110.60 11.537 1.071
Post-operative ROM B55 38 115.16 14.783 2.613
C55 137 117.90 11.581 1.062
KSS Knee society score, KFS knee functional score, ROM range of movement
Radiographic results deep infection, and progression of arthritis in the opposite

compartment. Six patients developed superficial infections
The average pre-operative varus deformity was 7° (range 2°– post-operatively and were managed with oral antibiotics.
14°) measured on orthogonal X-ray. The average post-op- One patient developed chronic regional pain syndrome post-
erative alignment was 3° (range neutral to 5°). The average operatively and was managed with medical treatment. None
alignment at recent follow-up was 4° (range 2°–8°), and there of the patient had any significant blood loss during the pro-
were no signs of progression of arthritis in the lateral com- cedure or required blood transfusion post-operatively.
partment in the cohort at the last follow-up X-ray. Medial PF
joint arthritis was noted in 60 % (105 patients) of the cohort Revisions
and there was no progression in PF joint arthritis at the recent
follow-up X-ray. The average pre-operative grade of PF joint Four patients (2.28 %) underwent revision surgery to TKR
arthritis was Grade 2, which involved the medial facet more because of unexplained pain. The clinical, biochemical and
commonly than the trochlear groove (range grade 1–3). This radiological investigation including computed tomography
was consistent with surgical findings. (CT) scan failed to identify the source of the pain. The average
period for revision surgery was 31.7 months (range
Complications 19–54 months) from the time of index surgery. There was no
marked improvement in the KS knee score and functional score
There were no significant complications per-operatively or in these patients. The average KS knee and functional scores
post-operatively such as fractures, deep vein thrombosis, were 59.5 (range 55–63) and 66.25 (range 60–70), respectively.

Two patients (1.14 %) underwent revision surgery to that UKR continues to give a result as good as TKR for
TKR at 5 and 7 years following index surgery due to tibial 10–14 years [5]. The present study concluded early
component loosening, which was demonstrated by CT scan recovery of patient following fixed-bearing UKR, reduced
without significant osteolysis. hospital stay (mean 2.5 days), excellent knee ROM
One patient (0.57 %) had revision surgery to TKR because (118.3° ± 12°), and very good patient satisfaction (87 %).
of polyethylene wear, 9.6 years following index surgery. The blood loss was minimal and no patient required blood
In the cohort of BMI B30 kg/m2, the failure rate was transfusion following the UKR procedure.
4.27 % (five knees) and the main factor for failure was Anterior knee pain and medial facet PF arthritis will
unexplained pain in 1.70 % (two knees), loosening of improve following UKR [16–19]. In our series, 105
component 1.70 % (two knees), and polyethylene wear patients (60 %) with medial facet PF arthritis had no pro-
0.85 % (one knee). In the cohort of BMI C30 kg/m2, the gression of PF arthritis or had anterior knee joint pain
failure rate was 3.44 % (two knees) and the factor for during the follow-up. Patients with lateral facet PF arthritis
failure was unexplained pain. and mal-tracking were not included in the study.
Loosening of a component, progression of arthritis to the
Survival analysis retained compartments, and polyethylene wear, were three
major causes of failure in UKR. In the Swedish Knee
Implant survival was calculated by constructing life Arthroplasty Register (SKAR) 2011 [20], the most common
tables and Kaplan–Meier survival analysis plot. The cases cause of failure of UKR was loosening of a component and
were grouped into 1-year intervals, with failure defined as approximately 45 % of revisions were attributed to this
revision to TKR or need for revision. The mean survivor- cause from 2000-2009. Lewold et al. [21] reported risk of
ship of the implant was 96 % at 10.9 years with 95 % CI revision following Oxford UKR was 2.1 % and mean time of
(10.6, 11.4 years) (Fig. 5). revision was 26 months (range 6–74 months). In our series,
two knees (1.14 %) were revised to TKR for tibial compo-
nent loosening and mean time of revision was 6 years (range
Discussion 5–7 years). The KSS knee and functional scores following
the revision surgery were satisfactory and average knee
The long-term success of UKR depends on stringent patient ROM was 110° (range 100°–120°).
selection criteria and surgical technique. The benefits of Polyethylene wear has often been cited as a cause of
UKR over TKR are better ROM, faster recovery and failure after UKR, more so in association with flat articu-
greater patient satisfaction [11–13]. Reports have shown lating surfaces than with congruent mobile bearings [19, 21–
23]. Berger et al. [24] reported no revisions for polyethylene
wear in a series of 51 knees that were treated with flat
articulating surface (fixed-bearing Miller-Galante). In the
SKAR Annual Report 2011 [20], polyethylene wear was the
second most common cause of failure of UKR and 15 % of
revision cases were attributed to this cause from 2000-2009.
In our series, one knee (0.5 %) was revised to TKR for
polyethylene wear at 9.6 years following primary UKR. No
significant osteolysis or implant loosening was noted dur-
ing the surgery.
The other main cause of failure was unexplained pain
which persists even after UKR. The possible explanations
include tibial condyle overload, overhang of the tibial
component, overstretching of the medial collateral liga-
ment (bearing is too thick) and pes anserine bursitis.
Revision arthroplasty does not cure pain in every case, and
in SKAR 2004 [34] unexplained pain following UKRs was
reported to be 5–6 %. Such procedures are often not only
unnecessary but also ineffective [23]. In our study, four
patients (2.28 %) were revised to TKR at an average of
31.7 months (range 19–54 months) for persistent pain.
There was no improvement in function or pain following
Fig. 5 Illustrating Kaplan-Meier implant survival revision surgery and average KS knee and functional scores

were 59.5 and 66.2, respectively. Patients were not satisfied in the other knee preferred UKR over TKR. Most patients
following the revision procedure and continued to com- had near normal knee kinematics and were happy with the
plain of pain. gait pattern following UKR.
One patient developed chronic regional pain syndrome Implant survival was calculated by constructing life
following UKR. TKR was not offered as the next thera- tables and Kaplan–Meir survival analysis plot. The cases
peutic step because it worsens the condition which was were grouped into 1-year intervals, with failure defined as
managed symptomatically. revision to TKR or need for revision. The mean survivor-
Most authors reported that overcorrection of the varus ship of the implant was 96 % at 10.9 years with 95 % CI
deformity into valgus deformity is the usual cause for (10.6, 11.4 years) in our study group, which was compa-
progression of arthritis in the contralateral compartment rable to other published studies [6, 22, 41].
and recommend leaving the UKR knee in a few degrees of The main limitation of our study was that it was a non-
varus or neutral to avoid this [21–24, 40]. In the SKAR randomized case series study (single surgeon) and the
2004 Report [34], approximately 25 % of the UKR revi- results were not compared with a controlled group. Patients
sions were for progression of arthritis. Progression of and their respective clinical and functional results were not
arthritis in other compartments, either PF or lateral, was not matched based on age, BMI and pre-operative limb
encountered in our study. This was attributed to slight alignment. The average follow-up of the study was short
undercorrection or neutral correction of the deformity and when compared with the most series in the literature.
the mean polyethylene thickness used was 9 mm (range The study concludes that fixed-bearing Miller-Galante
8–12 mm).The mean post-operative alignment of the knees UKR is a valid alternative for patients with idiopathic
was 3° (range neutral to 5°) and at recent follow-up was 4° medial compartment knee arthritis and the learning curve is
(range 2°–8°). steep. Proper patient selection is the key for excellent
Studies have reported that TKR patients with a high functional outcome and high rate of survivorship of
BMI have inferior results compared to patients with a implant following UKR.
lower BMI [24–30]. Tabor et al. [31] reported in a mean
follow-up of 20 years in 82 patients that obese patients had Compliance with ethical standards
higher survival than those who were not obese. In another Conflict of interest None of the authors have received any grants or
study of patients with Oxford Phase III UKR, Kuipers et al. funding and do not have any financial and personal relationships with
[32] reported no early difference between obese and non- other people or organisations that could inappropriately influence
obese patients. In our series, there was no statistical sig- (bias) their work.
nificant difference in the clinical and functional outcome
Ethical standards Obtaining the informed consent from involved
following UKRs in the cohort of BMI C30 and B30 kg/m2 patients was waived by the Research Ethics Committee. All proce-
(p [ 0.05). dures involving human participants were in accordance with the 1964
Lidgren et al. [34] reported age at the time of surgery to Helsinki declaration and its later amendments. The study was
approved by the Research Ethics Committee.
be a recognised risk factor for implant survival both in
UKR and TKR. Reliable function and good survival have
been reported for TKR in younger patients, and this form of
treatment has also been advocated for unicompartmental
osteoarthritis [31, 33–38]. The advantages of UKR over
TKR include retention of the cruciate ligaments, preser-
vation of bone stock in the uninvolved compartments and
better functional results [34, 39, 40]. In our study cohort,
39 (21.7 %) were aged B55 years and 137 (78.3 %) were
aged C55 years. There was no statistically significant dif-
ference in KS knee and functional scores between the References
group aged B55 years and those aged C55 years.
1. Peat G, Mc carney R, Croftt P (2001) Knee pain and osteoarthritis
Voss et al. [26] also reported that most patients with
in older adults; a review of community burden and current use of
UKR walked with a normal gait pattern as assessed in the health care. Ann Rheum Dis 60:91–97
gait laboratory. Rourgraff et al. [14] reported better clinical 2. Hootman JM, Helmick CG (2006) Projections of US prevalence
results and prosthetic survivorship for UKR over TKR and of arthritis and associated activities limitations. Arthritis Rheum
54(1):226–229
also reported that more people preferred UKR. In our study
3. American College of Rheumatology Subcommittee on
cohort, 87 % of the patients showed good to excellent Osteoarthritis Guidelines (2000) Recommendations for the
functional and clinical outcome, and preferred UKR for the medical management of osteoarthritis of the hip and knee.
opposite knee. Twelve patients in our study who had TKR Arthritis Rheum 43:1905–1915

4. Ahlback S (1968) Osteoarthrosis of the knee: a radiographic 24. Berger RA, Nedeff DD, Barden RM, Sheinkop MM, Jacobs JJ,
investigation. Acta Radiol Diagn (Stockh) Suppl 277:7–72 Rosenberg AG, Galante JO (1999) Unicompartmental total knee
5. Fisher N, Agarwal M, Reuben SF, Johnson DS, Turner PG (2006) arthroplasty. Clinical experience at 6- to10-year follow up. Clin
Sporting and physical activity following Oxford medial uni- Orthop 367:50–60
compartmental knee arthroplasty. Knee 13(4):296–300 25. Squire MW, Callaghan JJ, Goetz DD, Sullivan PM, Johnston RC
6. Price AJ, Waite JC, Svard UC (2005) Long term clinical results (1999) Unicompartmental knee replacement. A minimum 15 year
of the medial oxford unicompartmental knee arthroplasty. Clin follow up study. Clin Orthop 367:61–72
Orthop 435:171–180 26. Voss F, Sheinkop MB, Galante JO, Barden RM, Rosenberg AG
7. Bedson J, Jordon K, Croft P (2004) The prevalence and history of (1995) Miller-Galante unicompartmental knee arthroplasty at 2-
knee osteoarthritis in general practice: a case-control study. Fam to 5-year follow-up evaluation. J Arthroplasty 10(6):764–770
Pract 21:1–6 27. Foran JR, Mont MA, Rajadhyaksha AD (2004) Total knee
8. Murray DV, Goodfellow JW, O’Connor JJ (1998) The Oxford arthroplasty in obese patients: a comparison with a matched
medial unicompartmental arthroplasty: a 10-year survival study. control group. J Arthroplasty 19:817
J Bone Joint Surg Br 80-B:983–989 28. Winiarsky R, Barth P, Lotke P (1998) Total knee arthroplasty in
9. Bert JM (2005) Unicompartmental knee replacement. Orthop morbidly obese patients. J Bone Joint Surg Am 80:1770
Clin N Am 36(4):513–522 29. Berend KR, Lombardi AV Jr, Mallory TH (2005) Early failure of
10. Laurencin CT, Zelicof SB, Scott RD (1991) Unicompartmental minimally invasive unicompartmental knee arthroplasty is asso-
versus total knee arthroplasty in the same patient. A comparative ciated with obesity. Clin Orthop Relat Res 440:60
study. Clin Orthop 273:151–156 30. Heck DA, Marmor L, Gibson A (1993) Unicompartmental knee
11. Svard UCG, Price AJ (2001) Oxford medial unicompartmental arthroplasty. A multicentre investigation with long term follow
knee arthroplasty. A survival analysis of an independent series. up evaluation. Clin Orthop Relat Res 286:154
J Bone Joint Surg Br 83(2):191–194 31. Tabor OB Jr, Tabor OB, Bernard M (2005) Unicompartmental
12. Tanavalee A, Choi YJ, Tria AJ Jr (2005) Unicondylar knee knee arthroplasty: long term success in middle age and obese
arthroplasty: past and present. Orthopedics 28(12):1423–1433 patients. J Surg Orthop Adv 14:59
13. Ackroyd CE, Whitehouse SL, Newman JH, Joslin CC (2002) A 32. Kuipers BM, Kollen BJ, Kaijser Bots PC (2010) Factors associ-
comparative study of the medial St Georg sled and kinematic ated with reduced early survival in the Oxford phase III medial
total knee arthroplasties. Ten-year survivorship. J Bone Joint unicompartmental knee replacement. Knee 17:48
Surg Br 84-B:667–672 33. Peter MB, Maria SG, Micael GZ, Harpal SK, Aaron JJ, Michael
14. Berger RA, Meneghini RM, Jacobs JJ et al (2005) Results of AM (2011) Outcomes of unicompartmental knee arthroplasty
unicompartmental knee arthroplasty at a minimum of 10 years of stratified by body mass index. J Arthroplasty 26(8):1149–1153
follow-up. J Bone Joint Surg Am 87:999–1006 34. Lidgren L, Knutson K, Robertson O. Swedish Knee arthroplasty
15. BrennerSmith At, Ewings P, Weale AE (2004) Knee scores in a register 2004 annual report. www.myknee.se/en
normal elderly population. Knee 11:279–282 35. Argenson JN, Chevrol-Bankeddache Y, Aubaniac JM (2002)
16. Rougraff BT, Heck DA, Gibson AE (1991) A comparison of Modern unicompartmental knee arthroplasty with cement: a three
tricompartmental and unicompartmental arthroplasty for the to 10-year follow-up study. J Bone Joint Surg Am
treatment of gonarthrosis. Clin Orthop 273:157–164 84-A:2235–2239
17. Jonsson GT (1981) Compartmental arthroplasty for gonarthrosis. 36. Price AJ, Dodd CA, Svard UCG, Murray DW (2005) Oxford
Acta Orthop Scand Suppl 193 medial unicompartmental knee arthroplasty in patients younger
18. Goodfellow Hungerford DS, Zindel M (1976) Patella-femoral and older than 60 years of age. J Bone Joint Surg Br
joint mechanics and pathology. 1. Functional anatomy of the 87(11):1488–1492
patella-femoral joint. J Bone Joint Surg Br 58B:287–290 37. Pennington DW, Swienchowski JJ, Lutes WB, Drake GN (2003)
19. Beard DJ, Pandit H, Ostlere S, Jenkins C, Dodd CAF, Murray Unicompartmental knee arthroplasty in patients 60 years of age
DW (2007) Pre-operative clinical and radiological assessment of or younger. J Bone Joint Surg Am 85:1968–1973
the patellofemoral joint in unicompartmental knee replacement 38. Parratte S, Argenson JNA, Pearce O, Pauly V, Auquier P,
and its influence on outcome. J Bone Joint Surg Br Aubaniac JM (2009) Medial unicompartmental knee replacement
89-B:1602–1607 in the under-50 s. J Bone Joint Surg Br 91 B:351–356
20 Sundberg M, Lidgren L, Knutson K, Robertson O. Swedish Knee 39. Hanssen AD, Stuart MJ, Scott RD, Scuderi GR (2001) Surgical
Arthroplasty Register 2011 annual report. www.myknee.se/en options for the middle aged patient with osteoarthritis of the knee
21. Lewold S, Goodman S, Knutson K, Robertson O, Lidgren I joint. Instr Course Lect 50:499–511
(1995) Oxford meniscal bearing knee versus the Marmor knee in 40. Berger RA, Meneghini RM, Sheinkop MB, Della Valle CJ,
unicompartmental arthroplasty for arthrosis. A Swedish multi- Jacobs JJ, Rosenberg AG, Galante JO (2004) The progression of
centre survival study. J Arthroplasty 10:722–731 patellofemoral arthrosis after medial unicompartmental replace-
22. Argenson Jn, O’Connor JJ (1992) Polyethylene wear in meniscal ment: results at 11 to 15 years. Clin Orthop Rel Res 428:92–99
knee replacement. A one to nine-year retrieval analysis of the 41. Weale AE, Newman JH (1994) Unicompartmental arthroplasty
Oxford knee. J Bone Joint Surg Br 74:228–232 and high tibia osteotomy for osteoarthrosis of the knee: a com-
23. Psychoylos V, Crawford RW, O’Connor JJ, Murray DW (1998) parative study with a 12- to 17-year follow-up period. Clin
Wear of congruent meniscal bearings in unicompartmental knee Orthop 302:134–137
replacement: a retrieval study of 16 specimens. J Bone Joint Surg
Br 80:976–982

7
An unusual case of neurothekeoma of the arm in an adult
Federica Bergamin1 • Ezio Nicola Gangemi1 • Claudia Cerato1 • Alessandra Clemente1 •
Marco Borsetti1 • Adolfo Suriani2 • Stefano Taraglio2
Abstract Neurothekeomas are uncommon benign neo- mixed or hypercellular type, clinicians must consider the
plasms with a peripheral nerve sheath origin. This tumor possibility of an early local recurrence, suggesting a close
usually involves dermis and is described as a small, soli- clinical and radiological follow-up.
tary, slow growing and reddish to flesh-colored nodule or
papule. Neurothekeoma preferentially affects the central Keywords Soft tissue tumors Myxoid tumors Nerve
aspect of the face, the arms or shoulders of women in the sheath tumors Hand surgery Rare tumors
second and third decades of life. This is the first case report
of neurothekeoma involving the wrist developing from
synovial tissue and with uncertain clinical behavior in an Introduction
adult female. The tumor was completely excised under
brachial plexus block. Histopathologically, the examina- Neurothekeomas, or nerve sheath myxomas, are uncom-
tion of the microscopic slides revealed the presence of a mon benign neoplasms with a peripheral nerve sheath
20-mm diameter, well-circumscribed and multilobulated origin [1, 2]. This tumor usually involves dermis or, less
tumor composed of abundant myxoid stroma with cellular frequently, mucosal or submucosal tissue and, typically, is
elements; with immunohistochemistry there was positivity described as a small, solitary, slow growing and reddish to
to vimentin but S100-protein, epithelial membrane antigen, flesh-colored nodule or papule [3]. Neurothekeoma pref-
cytokeratin AE1-3, CD99 and CD34 were all negative. erentially affects the central aspect of the face, the arms or
This pattern suggested a myxoid tumor form of neu- shoulders of women in the second and third decades of life
rothekeoma, mixed subtype. The patient had an atypical [4, 5]. To the best of our knowledge, this is the first case
local recurrence and was re-operated after 3 months. After report of neurothekeoma involving the wrist developing
12 months there was no evidence of clinical recurrences from synovial tissue and with uncertain clinical behavior in
confirmed by magnetic resonance evaluation. Basically, an adult female.
our case report adds an important element in the correct
clinical management of neurotecheomas: faced with a
histological diagnosis with an unusual localization and Case report
A 30-year-old woman presented with a large dorso-lateral

& Federica Bergamin synovial cyst of the right wrist previously ultrasono-
[email protected] graphically checked (2 months before); at the first exami-
1
nation the wrist was painful and swollen without a specific
Department of Surgical Activities, Division of Plastic
localization and the overlying skin appeared without local
Surgery and Hand Surgery, Maria Vittoria Hospital, ASL
TO2, Via Cibrario 72, 10144 Turin, Italy erythema. There was no reported fever, palpitations, irri-
2 tability, dysphagia, dyspnea, weight loss, or other signifi-
Department of Laboratory Diagnostics, Division of
Pathology, Maria Vittoria Hospital, ASL TO2, Via Cibrario cant family medical history. The patient did not have a
72, 10144 Turin, Italy history of smoking but reported a history of recent trauma.

The patient was treated with a functional bandage, rest and

anti-inflammatory drugs for a week. At the next examina-
tion, there was an appreciable reduction of reactive syn-
ovitis of the extensor compartment, with evidence of the
synovial cyst previously described. Therefore, an excision
of the entire cyst was proposed.
Under brachial plexus block an S italic skin incision was
performed on the dorso-ulnar surface of the wrist; the deep
dissection showed a circumscribed and well-cleavable
lesion like a tenosynovitis of the extensor carpi ulnaris
(ECU) without a pedicle joint that was removed en bloc.
Surprisingly, the pathologist found a 20-mm diameter,
well-circumscribed and multilobulated tumor: each lobule
was composed of abundant myxoid stroma with cellular
elements in part fusal, in part epithelial and in part giant;
all lobules showed a tendency to confluence but there was
no cellular polymorphism or nuclear hyperchromia. Mar-
gins were tumor free. With immunohistochemistry there
was positivity to vimentin but S100-protein, epithelial
membrane antigen (EMA), cytokeratin AE1-3, CD99 and
CD34 were all negative. There was a mitotic activity of
7 % (Ki67). This pattern suggested a myxoid tumor form
of neurothekeoma, mixed subtype.
A close follow-up was performed; at 3 months a new
subcutaneous lesion in the dorsal compartment was
detected (Fig. 1a, b) and scanned with ultrasonography.
The patient underwent revision surgery with dorsal and
proximal extension of the previous scar; the subcutaneous
lesion appeared surrounded by scar tissue and adherent to Fig. 1 Clinical features of recurrence. a Dorsal view with evidence
the tendon sheath (Fig. 1c). Histologically, a recurrence of of 4-cm diameter subcutaneous non-solid mass. b Radial view with
neurothekeoma, with discrete myxoid stroma and cellular painless swelling proximal to the ulnar styloid. c View after complete
excision of the mass with evidence of extensor ulnaris carpi and
elements in part epithelial and in part giant, was found extensors digitorum. d Follow-up at 12 months showed no evidence
(Fig. 2). With immunohistochemistry there was positivity of further recurrence
only to vimentin; S100-protein, EMA and CD10 while
cytokeratin AE1-3, CD99, CD34, actina and NSE were all
negative. There was a mitotic activity of 4 % (Ki67). All compartment. While most have been located in the dermis,
margins were negative. the literature describes a few cases where it has arisen in
There was no recurrence by the 12-month follow-up, the mouth, the nose, the cranium, the cauda equine, the
according to magnetic resonance evaluation (Fig. 1d). spinal canal, within the peripheral nerves, the cerebello-
pontine angle, the mediastinal, the hypopharynx and the
external auditory canal [8–14]. To the best of our knowl-
Discussion edge, this is the first case report of neurothekeoma of the
wrist developing in the synovial tissue in an adult female.
Neurothekeoma is a rare and benign tumor of the nerve Usually nerve sheath myxoma does not have aggressive
sheath first described as being like nerve sheath myxoma local growth, does not have a tendency to metastases, and
by Harkin and Reed in 1969 [6] and subsequently termed recurrences are very uncommon [8]. Our group described a
neurothekeoma by Gallagher and Helwig in 1980 [7]. recurrence after 3 months probably due to the atypical site
It typically presents as a benign, asymptomatic, solitary, of localization or to the histologic and immunohisto-
slow growing and dome-shaped lesion that only rarely chemical characteristics.
ulcerates the skin. Neurothekeoma preferentially affects the In fact several studies have classified nerve sheath
dermis of the cervicofacial areas and shoulders of young myxomas into three groups based on cellularity, mucin
women. In our case the lesion was located on the wrist in content and growth pattern: a hypocellular or myxoid type,
the subcutaneous tissue near to the sixth extensor a hypercellular type and a mixed type [8, 15].

An unusual case of neurothekeoma of the arm in an adult 51
S100 and positive to vimentin and melanoma specific

antigen (NK1/C-3). According to Laskin et al. [16] cellular
and mixed type neurothekeoma failed to show convincing
evidence of neurosustentacular differentiation and, thus,
warranted a separate classification. Rarely the tumor may
demonstrate an infiltrative growth pattern, brisk mitotic
activity and cytological pleomorphism. However, these
atypical features do not influence the prognosis of the
tumor [17].
Based on these conisdertaions, mixed type and hyper-
cellular type could show unusual local recurrence because
of the presence of cells with mitotic activity; nevertheless a
complete surgical excision with the tumor margins free, the
clinician has to consider this eventuality. Moreover, Laskin
et al. [16] noticed that myxoid/hypocellular neurothekeo-
mas occurred more commonly in male patients (male:fe-
male ratio 6:5) in the fourth decade and were found on both
the upper and lower limbs. In contrast, the cellular/mixed
group affected more females (male:female ratio 4:7) in the
second decade with an upper body distribution [16].
Generally, neurothekeomas are difficult to diagnose
prior to performing a biopsy, due to the lack of specific
clinical manifestations or imaging characteristics. In fact in
our case the diagnosis of wrist mass was oriented towards
synovial cyst prior to surgical excision. The differential
diagnosis of neurothekeoma should include other neural
entities, such as schwannoma, true neuroma and myxoid
neurofibroma. The effective treatment of choice is com-
plete surgical excision with clear margins. No malignant
transformation or metastases have been reported and local
recurrence is extremely uncommon, but possible, espe-
cially for the cellular/mixed group [16, 17]. Although this
is a rare type of tumor, the clinician should consider this
entity in differential diagnosis, as it is imperative to dis-
tinguish it from malignant lesions, in order to avoid
Fig. 2 Histology features. a En bloc excision of the mass after unnecessary aggressive treatment. Faced with a suspected
recurrence. b Mixed form of neurothekeoma with lobules composed diagnosis is therefore advisable to perform a punch biopsy
by cells in part fusal, in part epithelial and in part giant in a of the lesion to determine with certainty the correct path of
background of abundant myxoid stroma (hematoxylin and eosin, the therapeutic procedure.
magnification 940). c Higher magnification section of mixed
neurothekeoma (hematoxylin and eosin, magnification 9200) Basically, our case report adds an important element in
the correct clinical management of neurothekeomas: faced
with a histological diagnosis with an unusual localization
The hypocellular group consists of well-circumscribed and mixed or hypercellular type, clinicians must consider
multilobulated tumors usually located in the reticular der- the possibility of an early local recurrence, suggesting a
mis and often extending into the subcutaneous fat with a close clinical and radiological follow-up.
prominent myxoid stroma and positive to vimentin and
S100-protein. According to Laskin et al. [16] the myxoid Compliance with ethical standards
type of neurothekeoma showed neurosustentacular differ-
entiation (glial cell, Schwann cell and melanocyte) and is Conflict of interest The authors declare that they have no conflict
of interest.
the bona fide nerve sheath tumor.
The cellular types are composed of ill-defined nests and Ethical standards The patient provided informed consent to the
fascicles of cells, rather scant mucin and are negative to publication of the case report.

7. Gallager RL, Helwig EB (1980) Neurothekeoma—a benign

cutaneous tumor of neural origin. Am J Clin Pathol 74:759–764
8. Allen PW (2000) Myxoma is not a single entity: a review of the
concept of myxoma. Ann Diagn Pathol 4:99–123. doi:10.1053/
adpa.2000.0099
9. Bhat A, Narasimha A, C V, Vk S (2015) Nerve sheath myxoma:
report of a rare case. J Clin Diagn Res 9:ED07–ED09. doi:10.
7860/JCDR/2015/10911.5810
10. Alghamdi KM, Al-Mubarak LA (2012) A myxoid neu-
rothekeoma of the nose. J Drugs Dermatol 11:252–255
References 11. Malkoc M, Ormeci T, Keskinbora M et al (2014) Nerve sheath
myxoma of the dorsal paravertebral space. Int J Surg Case Rep
1. Barnhill RL, Mihm MC (1990) Cellular neurothekeoma. A dis- 5:858–860. doi:10.1016/j.ijscr.2014.10.003
tinctive variant of neurothekeoma mimicking nevomelanocytic 12. Sayan M, Celik A, Ertunc O et al (2015) Mediastinal neu-
tumors. Am J Surg Pathol 14:113–120 rothekeoma: a rare tumor. Asian Cardiovasc Thorac Ann
2. Hornick JL, Fletcher CDM (2007) Cellular neurothekeoma: 23:593–595. doi:10.1177/0218492314552857
detailed characterization in a series of 133 cases. Am J Surg 13. Li J, Shi Z, Jing J (2014) Neurothekeomas of the thoracic and
Pathol 31:329–340. doi:10.1097/01.pas.0000213360.03133.89 lumbar area in an adult man: a case report. Mol Clin Oncol
3. Fetsch JF, Laskin WB, Miettinen M (2005) Nerve sheath myx- 2:156–158. doi:10.3892/mco.2013.211
oma: a clinicopathologic and immunohistochemical analysis of 14. Vij M, Jaiswal S, Agrawal V et al (2013) Nerve sheath myxoma
57 morphologically distinctive, S-100 protein- and GFAP-posi- (neurothekeoma) of cerebellopontine angle: case report of a rare
tive, myxoid peripheral nerve sheath tumors with a predilection tumor with brief review of literature. Turk Neurosurg
for the extremities and a high local recurrence rate. Am J Surg 23:113–116. doi:10.5137/1019-5149.JTN.4255-11.1
Pathol 29:1615–1624 15. Stratton J, Billings SD (2014) Cellular neurothekeoma: analysis
4. Rozza-de-Menezes RE, Andrade RM, Israel MS, Gonçalves of 37 cases emphasizing atypical histologic features. Mod Pathol
Cunha KS (2013) Intraoral nerve sheath myxoma: case report and 27:701–710. doi:10.1038/modpathol.2013.190
systematic review of the literature. Head Neck 35:E397–E404. 16. Laskin WB, Fetsch JF, Miettinen M (2000) The ‘‘neu-
doi:10.1002/hed.23361 rothekeoma’’: immunohistochemical analysis distinguishes the
5. Safadi RA, Hellstein JW, Diab MM, Hammad HM (2010) Nerve true nerve sheath myxoma from its mimics. Hum Pathol
sheath myxoma (neurothekeoma) of the gingiva, a case report and 31:1230–1241. doi:10.1053/hupa.2000.18474
review of the literature. Head Neck Pathol 4:242–245. doi:10. 17. Argenyi ZB, Kutzner H, Seaba MM (1995) Ultrastructural
1007/s12105-010-0183-5 spectrum of cutaneous nerve sheath myxoma/cellular neu-
6. Harkin J, Reed R (1969) Tumors of the peripheral nervous sys- rothekeoma. J Cutan Pathol 22:137–145
tem. In: Harkin JC, Reed RJ (eds) Atlas of Tumor Pathology.
Armed Forces Institute of Pathology, Washington, DC

8
Achieving hip fracture surgery within 36 hours: an investigation
of risk factors to surgical delay and recommendations for practice
Adeel Aqil1 • Fahad Hossain1 • Hassaan Sheikh1 • Joseph Aderinto1 •
George Whitwell1 • Harish Kapoor1
Abstract following a validated propensity score matching process,

Background The UK hip fracture best practice tariff a Pearson chi-square test failed to demonstrate a statistical
(BPT) aims to deliver hip fracture surgery within 36 h of difference in mortality incidence between the hypo-
admission. Ensuring that delays are reserved for conditions and normonatraemic patients [v2(1, N = 512) = 0.10,
which compromise survival, but are responsive to medical p = 0.757].
optimisation, would help to achieve this target. We aimed Conclusions Hip fracture surgery should not be delayed
to identify medical risk factors of surgical delay, and assess in the presence of non-severe and isolated hyponatraemia.
their impact on mortality. Instead, surgical delay may only be warranted in the
Materials and methods Prospectively collected patient presence of medical conditions which contribute to mor-
data was obtained from the National Hip Fracture Database tality and are optimisable.
(NHFD). Medical determinants of surgical delay were Level of evidence III
identified and analysed using a multivariate regression
analysis. The mortality risk associated with each factor Keywords Hip fractures Time to treatment 36 h
contributing to surgical delay was then calculated.
Results A total 1361 patients underwent hip fracture
surgery, of which 537 patients (39.5 %) received surgery Introduction
within 36 h of admission. Following multivariate analyses,
only hyponatraemia was deduced to be a significant risk A fracture of the hip is the commonest cause of injury-
factor for delay RR = 1.24 (95 % CI 1.06–1.44). However, related death in the UK [1]. Prompt surgery has been
associated with higher rates of independent living and
lower 30-day and 1-year mortality rates [2–5]. Earlier
& Adeel Aqil surgery has also been shown to improve patient outcomes
[email protected]
by reducing pain scores, and lowering of the risk of decu-
Fahad Hossain bitus ulcer formation and length of inpatient stay [2, 6, 7].
[email protected]
The inception of best practice tariffs (BPTs), which
Hassaan Sheikh aimed to improve these patient outcomes, stemmed from
[email protected]
the ‘Equity and Excellence: liberating the NHS’ govern-
Joseph Aderinto ment white paper [8]. BPTs are incentivised targets, which
[email protected]
financially compensate organisations for delivering high
George Whitwell quality care. In the context of hip fracture management, the
[email protected]
BPT consists of an initial base tariff, with additional pay-
Harish Kapoor ments if further criteria of best practice have been met. One
[email protected]
of these criteria is delivering hip fracture surgery within
Level 1 Trauma Centre, Leeds General Infirmary, Great 36 h of presentation to a health care institution. This
George St, Leeds LS1 3EX, UK government target is also in accordance with clinical

guidelines set by the British Orthopaedic Association each patient as a separate variable [12]. Biochemical
(BOA) and National Institute of Clinical Excellence parameters collected included admission haemoglobin
(NICE), which state that hip fracture surgery should be levels (Hb), white cell count (WCC), coagulation profile,
performed on the day of, or the day after admission and urea and electrolyte levels. Time to surgery from presen-
within normal working hours [9, 10]. However, the tation was also collected.
National Hip Fracture Database (NHFD) has reported that The primary outcome of interest was a delay to surgery
this specific BPT target was met in only 71.4 % of hip over 36 h from initial hospital presentation. The secondary
fracture patients, equating to £15.9 million in ‘lost’ mon- outcome examined was the occurrence and causes of
etary incentives [1]. mortality within 30 days of admission. Primary and sec-
Clearly, resources must be made available to allow such ondary causes of death were noted from death certificates
a level of service provision and to qualify for the maximum and hospital death records. A total of 1674 patients were
financial reward the BPT has to offer. Optimal clinical initially identified, but following exclusions of incomplete
decision-making could therefore augment and streamline data sets and incorrect or duplicate entries, a total of 1361
management in order to facilitate early surgery. As survival patients were included in the study.
is perhaps the most desirable outcome following a fracture Statistical analysis was undertaken in a two-stage pro-
of the neck of femur (FNOF), and delay to surgery in itself cess. We initially categorised patients into two groups:
carries an increased risk to mortality, then it certainly group1 = time to surgery \36 h; group 2 = time to sur-
follows that delays for medical optimisation would only be gery [36 h). All variables collected were then compared
justified for conditions which also carry a mortality risk [3– between these two groups on initial univariate analysis
7]. Therefore, identifying medical risk factors for surgical using the chi-square or Fisher’s exact test for categorical
delay and their associated mortality risk would assist data and the independent t/Mann–Whitney test for contin-
organisations to rationalise clinical decision-making, and uous variables. A subsequent backward stepwise Cox
thus enhance compliance with the BPT target. regression model was undertaken to identify the most
The primary aim of this study was therefore to identify significant determinants of surgical delay beyond 36 h. Our
medical conditions associated with patients failing to criteria for inclusion of variables into the model included a
achieve the 36-h cut-off for surgery following a hip frac- p value \0.15 on univariate analysis, in accordance with
ture. We subsequently evaluated whether these factors published statistical methods [13]. Results were displayed
were justifiable in risking surgical delay by gauging whe- as relative risks rather than odds ratios, in order to aid
ther they were also associated with an increased risk to clinical interpretation [14].
mortality. The decision to delay hip fracture surgery on medical
grounds is undertaken to avoid significant complications
which may result from precipitous surgery. Death is per-
Materials and methods haps the most important complication to avoid. Therefore,
it is logical to validate variables that risk a delay to surgery
We obtained prospectively collected hip fracture patient beyond 36 h in terms of their impact on mortality. We
information from the UK NHFD from before April 2010 undertook a second-stage analysis to assess mortality
and prior to the inception of the 36-h BPT guideline. Data likelihood at 30 days following surgery for each individual
was subsequently cross-referenced with our institution’s variable which had been found to delay surgery. To limit
patient records. The use of data after the introduction of the potential for selection bias, when assessing one variable’s
BPT guidelines may have risked missing patients with association with mortality we had to control for all other
legitimate causes of delay, who may have had their surgery variables. We therefore derived a single scalar propensity
expedited to meet the 36-h target. Hence analysis of delays score from the regression of all remaining covariates in
was performed on data pre-dating the BPT introduction, Tables 1 and 2. Between-group propensity score matching
allowing all medical causes of delays to be identified and was performed using a ‘‘nearest neighbour’’ matching
an assessment of their risk to mortality to be performed. strategy [15]. An assessment of the matching process
We collected patient-level information including consisted of an evaluation of between-group standardised
demographic data, American Society of Anesthesiologists mean differences and variance ratios according to pub-
(ASA) grade at the time of surgery, fracture type, source of lished standards [16]. Between-group mortality analysis
admission and walking ability [11]. All patient co-mor- used a chi-square test for each variable influencing surgical
bidity data was identified using the International Classifi- delay only after the matching process had been verified as
cation of Disease 10th revision (ICD-10) codes, and these being successful in balancing covariates between the two
were used to calculate the Charlson co-morbidity index for groups.

Achieving hip fracture surgery within 36 hours: an investigation of risk factors to surgical delay... 55
Table 1 Comparison of
Variables Time to surgery \36 h Time to surgery [36 h p value
demographic data between
subjects who did and those who Number 537 (39.5 %) 824 (60.5 %) 0.01
did not have surgery within 36 h
of admission Age in years 84 (24–103) 83 (31–104) 0.279
Gender – – 0.674
Male 143 (26.6 %) 228 (27.6 %) –
Female 394 (73.4 %) 596 (72.4 %) –
Fracture type – – 0.228
Intracapsular undisplaced 197 (36.7 %) 263 (31.9 %) –
Intracapsular displaced 156 (29.1 %) 277 (33.6 %) –
Intertrochanteric 140 (26.1 %) 217 (26.3 %) –
Subtrochanteric 44 (8.2 %) 67 (8.1 %) –
Admission source – – 0.031a
Own home 390 (72.6 %) 650 (78.8 %) –
Residential/nursing home 118 (22.0 %) 128 (15.5 %) –
Already inpatient 11 (2.0 %) 24 (2.9 %) –
Other hospital 2 (0.4 %) 2 (0.2 %) –
Unknown/other 16 (3.0 %) 20 (2.4 %) –
Pre-injury walking ability – – 0.664
Independent 285 (55.1 %) 439 (53.2 %) –
1 stick 122 (22.7 %) 203 (24.6 %) –
2 sticks or frame 99 (18.4 %) 148 (17.9 %) –
Wheelchair/scooter 12 (2.2 %) 13 (1.6 %) –
Unknown 19(3.5 %) 21(%) –
Results are displayed as median (range) for continuous data, and as n (%) of population for discrete data
Continuous data were analysed using an independent t-test, categorical data using chi-square/Fisher’s test
and ordinal data using the Mann–Whitney–Wilcoxon test
a
Included in the multivariate analysis
Results existing anticoagulation therapy (p = 0.303). Furthermore,

the calculated Charlson’s co-morbidity index was also
A total 1361 patients underwent hip fracture surgery, of similar between the two groups (p = 0.835). There was no
which 537 patients (39.5 %) received surgery within 36 h statistical difference in haematological and serum bio-
of admission. The overall median time to surgery from chemical parameters between the two groups (Table 2).
presentation was 23 h (3–36) in group 1 and 72 h (36–774) Following univariate analysis, seven variables, includ-
in group 2. The demographics were similar between ing admission source, history of dementia, ischaemic heart
patients who did (group 1) and those who did not (group 2) disease, MI, cerebrovascular accidents (CVA), urinary tract
receive timely surgery (Table 1). There was no difference infections and hyponatraemia met criteria for inclusion into
between the two groups with respect to age, gender, the Cox regression model. The model thereafter inferred
walking ability, fracture pattern and ASA grade. However, only hyponatraemia to be a significant risk factor for delay
with regards to admission source, there was a higher pro- to surgery beyond 36 h with a covariate adjusted relative
portion of patients presenting from a community care risk (RR) 1.24 (95 % CI 1.06–1.44, p = 0.006).
institution in group 1, whilst a higher proportion of patients The overall 30-day mortality in our cohort of hip frac-
were from their own home in group 2 (p = 0.013). ture patients was 9.0 %. The commonest cause of death
The distribution of the different co-morbidities between was from pneumonia (37 %). Following propensity score
the two groups are summarised in Table 2. There was a analysis, 256 patients with hyponatraemia were matched to
higher proportion of patients with cardiac co-morbidities in 256 patients with normal sodium levels. The absolute
group 1, while a higher proportion of patients in group 2 acceptable propensity score caliper width was 0.01. A near
presented with hyponatraemia (sodium \135 mmol/l) perfect standardised mean difference of 0.0003 and a
(p = 0.00). There was no difference between the two variance ratio of 1 (0.01:0.01) confirmed between-group
groups with respect to a number of factors, including pre- homogeneity to be well within acceptable limits [16]. Thus,

Table 2 Comparison of clinical

Variable Time to surgery \36 h Time to surgery [36 h p value
data between subjects who did
and those who did not have Co-morbidities
surgery within 36 h of hospital
admission Dementia 46 (8.6 %) 52 (6.3 %) 0.116a
Hypertension 18 (3.4 %) 25 (3.0 %) 0.743
Diabetes mellitus 50 (9.3 %) 89 (10.8 %) 0.375
Ischaemic heart dis. 161 (30 %) 195 (23.6 %) 0.010a
COPD/asthma 75 (14 %) 126 (15.3 %) 0.501
Neurological dis. 12 (2.2 %) 17 (2.1 %) 0.830
Stroke 25 (4.7 %) 24 (2.9 %) 0.092a
Thyroid dis. 33 (6.1 %) 43 (5.2 %) 0.467
Malignancy 50 (9.3 %) 79 (9.6 %) 0.865
Alcoholism 18 (3.4 %) 25 (3.0 %) 0.743
Chest infection 66 (12.3 %) 97 (11.8 %) 0.773
Urinary tract infection 99 (18.4 %) 123 (14.9 %) 0.087a
Myocardial infarction 22 (4.1 %) 17 (2.1 %) 0.028a
Cardiac failure 16 (3.0 %) 23 (2.8 %) 0.839
Peripheral vascular dis. 6 (1.1 %) 13 (1.6 %) 0.638b
Peptic ulcer dis. 4 (0.7 %) 5 (0.6 %) 0.745b
Liver disease 4 (0.7 %) 3 (0.4 %) 0.444b
Connective tissue dis. 0 (0 %) 1 (0.1 %) 1.0b
Leukaemia 1 (0.2 %) 3 (0.4 %) 1.0b
Anaemia 64 (11.9 %) 105 (12.7 %) 0.652
Chronic renal failure 41 (7.6 %) 79 (9.6 %) 0.214
Hyponatraemia 96 (17.9 %) 233 (28.2 %) 0.000a
Anticoagulation therapy 7 (1.3 %) 18 (2.2 %) 0.303b
Blood results on admission
HB 12.1 (6–17) 12.0 (7–19) 0.563
Platelet count 264 (43–843) 264 (43–938) 0.313
White cell count 10.3 (4–78) 10.3 (1–67) 0.754
Urea 7.4 (1–34) 7.2 (1–36) 0.950
Creatinine 93 (50–817) 92 (42–512) 0.949
Potassium (K?) 4.4 (2–7) 4.3 (3–7) 0.805
INR 1 (0.8–5.6) 1 (0.8–6.3) 0.540
APTT 29 (20–190) 29 (19–195) 0.450
ASA 0.685
1 52 (9.7 %) 68 (8.3 %)
2 135 (25.1 %) 226 (27.4 %)
3 282 (52.5 %) 420 (51.0 %)
4 68 (12.7 %) 109 (13.2 %)
5 0 (0 %) 1 (0.1 %)
Charlson score (median, range) 4 (0–8) 4 (0–9) 0.835
Results are displayed as median (range) for continuous data, and as n (%) of population for discrete data
Continuous data were analysed using an independent t-test, categorical data using chi-square/Fisher’s test
and ordinal data using the Mann–Whitney–Wilcoxon test
Displayed blood results are serum values in mmol/l
INR international normalised ratio, APTT activated partial thromboplastin time
a
Included in multivariate analysis
b
Fisher’s exact test used

the matching process controlled for all collected variables, 30-day re-admission [21]. Equally, it follows that patients
including time to surgery. The 30-day mortality rates for in group 2 with comparatively normal INRs were delayed
hyponatraemic patients was 10 % (24/256) and 9 % (22/ for other reasons.
256) for normonatraemic patients. This was not statistically We found hyponatraemia to be comparatively more
significant (p = 0.757). common in the delayed surgery cohort. While we did not
formally explore the underlying reasons for this, anecdotally
we believe that hyponatraemia was perceived to be associ-
Discussion ated with peri-operative mortality and morbidity. The asso-
ciation between hyponatraemia and mortality has been
Our study has shown that 60.5 % of patients had surgery demonstrated previously [22]. However, it has also been
delayed beyond 36 h. Furthermore, hyponatraemia was suggested that severe underlying disease is the cause of death
identified as a pre-operative risk factor for this. Interest- while hyponatraemia is merely another complication of this
ingly, the impact of hyponatraemia on 30-day mortality underlying disease. Hence, while it shows an association, it
was not significant. does not necessarily prove causality. Chawla et al. in their
Nationally, the reason for 37.9 % of patients failing to study of just over 45,000 patients found that mortality rates
meet the UK hip fracture BPT target was because of a tended to increase as sodium levels changed from normal to
perceived need for medical optimisation [17]. At first mild hyponatraemia. Surprisingly, as hyponatraemia
glance the rates of delay in our study may seem high. This became more severe (sodium\120 mmol/l) mortality trends
was because data collection predated the NICE guidelines reversed [23]. Furthermore, over the 12 years of their study,
for time to surgery. We realised that the guidelines could only three deaths were directly attributable to adverse
have modified clinical practice owing to the need for hyponatraemia sequelae. Our study also found no difference
expediting surgery within 36 h. Hence, potential medical in mortality incidence between hypo- and normonatraemic
causes for delay that would have otherwise been apparent patients after matching groups for confounders, such as liver
prior to the guidelines would potentially be missed fol- and renal failure, which may have contributed to both mor-
lowing its introduction. Thus, pre-guideline data were used tality and hyponatraemia. This supports the notion that
in an effort to prevent this potential bias. hyponatraemia may not necessarily be singularly causal to
The median age in both groups of our cohort was above mortality. Subgroup analysis of those with severe hypona-
80 years with a higher proportion of females. This is in traemia (sodium\120 mmol/l) was unfortunately precluded
agreement with demographic information published by the because there were only three patients that fell into this
NHFD [18]. With respect to admission source, we found category. We cannot therefore draw conclusions as to whe-
that a comparatively higher proportion of patients admitted ther severe hyponatraemia is a risk factor to mortality and
from their own home with a hip fracture were delayed whether it is reasonable to delay surgery in its presence.
beyond 36 h. Conversely, a higher proportion of patients Interestingly, we found that patients with a history of
admitted from a community care institution were seen in ischaemic heart disease were significantly less likely to have
the timely surgery group. It is entirely possible that clinical their surgery delayed beyond 36 h (23.6 versus 30 %,
practice may have inherently favoured expedited treatment p = 0.01). Patients with a history of myocardial infarction
owing to fears of complications of delayed surgery in also were significantly less likely to have delayed surgery
patients who were perceived to be frailer. Such patients are (4.1 versus 2.1 %, p = 0.03). These patients may have been
more likely to present from a community care institution prioritised as these risk factors are non-modifiable and
than their own home [19, 20]. This is also reflected by the clinical opinion may have been not to increase their risk
comparatively higher proportion of cardiac related co- further by also having delayed surgery. Similarly, there is an
morbidities in the early surgery group. increased awareness of the need to avoid unnecessary delays
This study found that the mean international normalised in order to gain financial compensation for services used in
ratio (INR) and ranges between the two groups were sim- treating such patients, and to avoid increased costs associ-
ilar. This is because patients with comparatively higher ated with longer hospital stays in these patients [24, 25].
INRs in group 1 had been aggressively treated to correct The main weakness of this study lies in the fact that we
the values within the 36-h time frame by using, according present data pertaining to only one major trauma unit. One
to our institution’s formalised protocol, vitamin k therapy. may argue that patient data from other units may yield
This practice is supported by Gleeson et al. who demon- differing results. However, our findings may be more
strated in their cohort of 1080 patients, that an active widely generalisable as our patient population demo-
management strategy for the reversal of warfarin antico- graphics and mortality rates of 9.0 % (n = 123/1361) at
agulation facilitated earlier surgery without increasing 30 days were comparable to other published studies and
complications of thromboembolic events, mortality or NHFD reports [18, 26, 27]. Although retrospective by

design, we cross-referenced prospectively collected data References

from multiple sources, including a national hip fracture
registry and our own hospital-coding database, ensuring 1. Royal College of Physicians (2013) National Hip Fracture
Database national report. http://www.nhfd.co.uk/20/hipfractureR.
that the final dataset was reliable. Non-medical risk factors
nsf/0/CA920122A244F2ED802579C900553993/$file/NHFD
for delay are not available in the NHFD or medical notes Report2013.pdf. Accessed 20 Mar 2015
and hence our regression model is limited by their absence. 2. Al-Ani AN, Samuelsson B, Tidermark J, Norling A, Ekstrom W,
We have, however, made a comprehensive assessment of Cederholm T, Hedstrom M (2008) Early operation on patients
with a hip fracture improved the ability to return to independent
38 medical and demographic variables. These variables are
living. A prospective study of 850 patients. J Bone Jt Surg Am
readily available on initial presentation and are thus easily 90(7):1436–1442. doi:10.2106/JBJS.G.00890
collectable by other units who also wish to make similar 3. McGuire KJ, Bernstein J, Polsky D, Silber JH (2004) The 2004
assessments of their services. Marshall Urist award: delays until surgery after hip fracture
increases mortality. Clin Orthop Relat Res 428:294–301
This type of study is relevant in the current NHS culture
4. Shiga T, Wajima Z, Ohe Y (2008) Is operative delay associated
of target-driven quality health care delivery. Verifying and with increased mortality of hip fracture patients? Systematic
investigating the legitimacy of medical causes of surgical review, meta-analysis, and meta-regression. Can J Anaesth
delay is therefore not only pertinent, but has also been 55(3):146–154. doi:10.1007/BF03016088
5. Weller I, Wai EK, Jaglal S, Kreder HJ (2005) The effect of
specifically highlighted as a vital area for future research
hospital type and surgical delay on mortality after surgery for hip
by the NHFD Scientific Committee [9]. To our knowledge fracture. J Bone Jt Surg Br 87(3):361–366
this study is unique in assessing the risk factors to delay in 6. Grimes JP, Gregory PM, Noveck H, Butler MS, Carson JL (2002)
achieving the 36-h BPT target in these patients. Nationally, The effects of time-to-surgery on mortality and morbidity in
patients following hip fracture. Am J Med 112(9):702–709
delays are also due to a lack of theatre time, equipment or
7. Orosz GM, Magaziner J, Hannan EL, Morrison RS, Koval K,
high dependency beds (43 % of the time) [17]. Therefore, Gilbert M, McLaughlin M, Halm EA, Wang JJ, Litke A, Sil-
whilst streamlining medical decision-making may help berzweig SB, Siu AL (2004) Association of timing of surgery for
improve the likelihood of attaining the BPT, availability of hip fracture and patient outcomes. JAMA, J Am Med Assoc
291(14):1738–1743. doi:10.1001/jama.291.14.1738
clinical resources plays an important part.
8. Department of Health (2010) Equity and excellence: liberating
In conclusion, surgical delays can result when one aims the NHS. http://www.gov.uk/government/uploads/system/uploads/
to avoid medical complications associated with hastened attachment_data/file/213823/dh_117794.pdf. Accessed 20 Mar 2015
hip fracture surgery. However, delay is not justifiable in the 9. NICE clinical guideline 124 (June 2011). The management of hip
fractures in adults. Modified March 2014. http://www.nice.org.
presence of non-severe and isolated hyponatraemia.
uk/guidance/cg124. Accessed 22 Nov 2014
Instead, surgical delay should only be warranted in the 10. British Orthopaedic Association (2012) BOAST 1 guideline
presence of medical conditions which contribute to mor- version 2—patients sustaining a fragility hip fracture. http://
tality and are optimisable. www.boa.ac.uk/publications/boa-standards-for-trauma-boasts/
#toggle-id-1. Accessed 22 Nov 2014
11. Association of Anaesthetists of Great Britain and Ireland safety
Acknowledgments The authors would like to thank the Orthopae-
guideline (2010) Management of proximal femoral fractures.
dic Network, UK for their help in this study.
http://www.aagbi.org/sites/default/files/femoralfractures2012_0pdf.
Accessed 20 Aug 2014
Compliance with ethical standards
12. Charlson ME, Pompei P, Ales KL, MacKenzie CR (1987) A new
method of classifying prognostic comorbidity in longitudinal
Ethical standards Obtaining informed consent from involved
studies: development and validation. J Chronic Dis 40(5):373–383
patients was waived by our Institutional Review Board for this ret-
13. Bursac Z, Gauss CH, Williams DK, Hosmer DW (2008) Pur-
rospective study. All procedures involving human participants were in
poseful selection of variables in logistic regression. Source Code
accordance with the 1964 Helsinki Declaration and its later amend-
Biol Med 3:17. doi:10.1186/1751-0473-3-17
ments. Ethical approval was waived by our institutional review board
14. McNutt LA, Wu C, Xue X, Hafner JP (2003) Estimating the
as it involved retrospective patient data which evaluated service
relative risk in cohort studies and clinical trials of common out-
provision. This study was therefore in accordance with regulations
comes. Am J Epidemiol 157(10):940–943
from the National Patient Safety Agency.
15. Stuart EA (2010) Matching methods for causal inference: a
review and a look forward. Statist Sci: Rev J Inst Math Statist
Conflict of interest No benefits in any form have been received or
25(1):1–21. doi:10.1214/09-STS313
will be received from a commercial party related directly or indirectly
16. Rubin DB (2007) The design versus the analysis of observational
to the subject of this article.
studies for causal effects: parallels with the design of randomized
trials. Stat Med 26(1):20–36. doi:10.1002/sim.2739
17. Royal College of Physicians (2011) National Hip Fracture
Database annual report. http://www.nhfd.co.uk/003/hipfracturer.
nsf/NHFDNationalReport2011_Final.pdf. Accessed 20 Mar 2015
18. Royal College of Physicians (2014) National Hip Fracture
Database annual report. http://www.nhfd.co.uk/20/hipfractureR.
nsf/welcome?readform. Accessed 20 Mar 2015

19. Beaupre LA, Cinats JG, Jones CA, Scharfenberger AV, William 24. Kosy JD, Blackshaw R, Swart M, Fordyce A, Lofthouse RA
CJD, Senthilselvan A, Saunders LD (2007) Does functional (2013) Fractured neck of femur patient care improved by simu-
recovery in elderly hip fracture patients differ between patients lated fast-track system. J Orthop Traumatol 14(3):165–170.
admitted from long-term care and the community? J Gerontol Ser doi:10.1007/s10195-013-0240-4
A Biol Sci Med Sci 62(10):1127–1133 25. Johnson DJ, Greenberg SE, Sathiyakumar V, Thakore R,
20. Eastwood EA, Magaziner J, Wang J, Silberzweig SB, Hannan Ehrenfeld JM, Obremskey WT, Sethi MK (2015) Relationship
EL, Strauss E, Siu AL (2002) Patients with hip fracture: sub- between the Charlson comorbidity index and cost of treating hip
groups and their outcomes. J Am Geriatr Soc 50(7):1240–1249 fractures: implications for bundled payment. J Orthop Traumatol
21. Gleason LJ, Mendelson DA, Kates SL, Friedman SM (2014) 16(3):209–213. doi:10.1007/s10195-015-0337-z
Anticoagulation management in individuals with hip fracture. 26. Tarrant SM, Hardy BM, Byth PL, Brown TL, Attia J, Balogh ZJ
J Am Geriatr Soc 62(1):159–164. doi:10.1111/jgs.12591 (2014) Preventable mortality in geriatric hip fracture inpatients.
22. Mc Causland FR, Wright J, Waikar SS (2014) Association of Bone Jt J 96-B(9):1178–1184. doi:10.1302/0301-620X.96B9.32814
serum sodium with morbidity and mortality in hospitalized 27. Haleem S, Lutchman L, Mayahi R, Grice JE, Parker MJ (2008)
patients undergoing major orthopedic surgery. J Hosp Med Mortality following hip fracture: trends and geographical varia-
9(5):297–302. doi:10.1002/jhm.2168 tions over the last 40 years. Injury 39(10):1157–1163. doi:10.
23. Chawla A, Sterns RH, Nigwekar SU, Cappuccio JD (2011) 1016/j.injury.2008.03.022
Mortality and serum sodium: do patients die from or with
hyponatremia? Clin J Am Soc Nephrol 6(5):960–965. doi:10.
2215/CJN.10101110

9
Effectiveness of intra-articular injections of sodium hyaluronate-
chondroitin sulfate in knee osteoarthritis: a multicenter
prospective study
Fabrizio Rivera1,9 • Luca Bertignone2,3 • Giancarlo Grandi4 • Roberto Camisassa4 •
Guido Comaschi5 • Diego Trentini6 • Marco Zanone7 • Giuseppe Teppex7 •

Gabriele Vasario7 • Giorgio Fortina8
Abstract month 6 (P \ 0.0001). The mean subscore was already

Background Intra-articular injection of hyaluronic acid is significantly decreased 1 month after the last injection at
a well-established therapy for the treatment of knee 25.7 (P \ 0.0001). Pain relief and consumption of anal-
osteoarthritis. The aim of the study was to assess the gesic drugs, both assessed with visual analogic scale
effectiveness and safety of the use of Arthrum HCSÒ (VAS), consistently decreased. The investigators were
(40 mg hyaluronic acid and 40 mg chondroitin sulfate in satisfied/very satisfied as regards the therapeutic effec-
2 mL). tiveness of sodium hyaluronate-chondroitin sulfate in
Materials and methods This was an open, multicenter, reducing pain (77 %), improving mobility (78 %) and
prospective study. Men or women over 40 years of age reducing the consumption of analgesics (74 %). Only one
with documented knee osteoarthritis and WOMAC sub- adverse effect was reported by one patient (knee
score A (severity of pain) C25 were enrolled. They tumefaction).
received three weekly intra-articular injections of sodium Conclusion These results suggest that intra-articular
hyaluronate 2 % and chondroitin sulfate 2 % in combina- injections of Arthrum HCSÒ (sodium hyaluronate plus
tion. WOMAC subscore A was assessed at 1, 3 and chondroitin sulfate) in patients with knee osteoarthritis are
6 months after the last injection. efficient and safe. These results should be confirmed in a
Results One hundred and twelve patients were included randomized controlled study.
(women, 66 %). The mean (SD) WOMAC subscore A Level of evidence IV.
decreased from 52.1 (15.2) at inclusion to 20.5 (19.7) at
Keywords Knee osteoarthritis Intra-articular injection
Sodium hyaluronate Chondroitin sulfate
& Fabrizio Rivera
[email protected]
1
Department of Orthopedic Trauma, SS Annunziata Hospital, Introduction
Savigliano, CN, Italy
2
Sant’Anna Clinic, Casale Monferrato, AL, Italy It is estimated that around 250 million people in the world
3
Eporediese Hospital, Ivrea, TO, Italy are affected by knee osteoarthritis [1]. Knee pain has an
4
La Vialarda Clinic, Biella, Italy
important impact by limiting activity and impairing quality
5
of life. Thus, knee osteoarthritis has been identified as one
Sestri Ponente Hospital, Genova, Italy
of the medical conditions (with stroke, depression, hip
6
Department of Orthopedics and Traumatology, IRCCS fracture and heart disease) accounting for more physical
A.O.U. San Martino-IST, Genova, Italy
disability than other diseases in people 65 years of age or
7
Department of Orthopedics and Traumatology, AO CTO older [2]. Moreover, knee osteoarthritis has a major impact
Hospital, Turin, Italy
on healthcare costs [3–6].
8
Santa Rita Clinic, Vercelli, Italy The management of osteoarthritis should be hierarchi-
9
Via Servais 200 A 16, Turin, Italy cal, with non pharmacological methods as first

Effectiveness of intra-articular injections of sodium hyaluronate-chondroitin sulfate in knee osteoarthritis... 61
interventions (weight loss, exercise and braces), followed improvement, together with the changes of the ultrasound
by analgesic drugs [nonsteroidal anti-inflammatory drugs parameters and biomarkers of cartilage metabolism and
(NSAIDs) and other analgesics], local therapies (topical joint inflammation, suggested a non-placebo effect. These
NSAIDs, intra-articular corticosteroids and hyaluronic results prompted us to assess in a prospective multicenter
acid), and surgery as a last resort [7–10]. study the effectiveness and safety of the use of hyaluronic
Administration of exogenous hyaluronic acid (visco- acid when combined with chondroitin sulfate in patients
supplementation) directly into the joint is available as a with knee osteoarthritis.
treatment for the symptoms of knee osteoarthritis. The
purpose of viscosupplementation is to overcome the qual-
itative and quantitative deficiency of hyaluronic acid that is Materials and methods
associated with osteoarthritis. Hyaluronic acid is a
polysaccharide that is the main constituent of cartilage and Study design
synovial fluid; it is responsible for the mechanical prop-
erties of the joint by allowing shock absorption, lubrication This was an open, multicenter, prospective study, assessing
and cartilage protection [11]. In osteoarthritis patients, the effectiveness of three intra-articular injections of
synovial hyaluronate is depolymerized and is cleared at sodium hyaluronate plus chondroitin sulfate (40 mg of
higher rates compared to normal subjects due to inflam- each compound in 2 mL) in the symptomatic treatment of
mation [12]. Intra-articular injections of hyaluronic acid knee osteoarthritis.
have been shown to be as effective as NSAIDs with fewer The study was conducted prospectively by office or
systemic adverse events [13]; this therapy has a delayed hospital specialists (orthopedic surgeons, rehabilitation
onset of action in comparison with intra-articular corti- medicine physicians) from October 2012 to December
costeroids, but a longer-lasting benefit [14]. Younger 2013.
patients and patients at an earlier stage of the disease are Written informed consent was obtained from each
more likely to benefit from viscosupplementation [15]. patient. The protocol was conducted in accordance with the
Arthrum HCSÒ (LCA Pharmaceutical, Chartres, France) Declaration of Helsinki and Guidelines on Good Clinical
is a new specialty for viscosupplementation combining Practice and approved by a local ethics committee.
sodium hyaluronate and chondroitin sulfate. Chondroitin
sulfate—a sulfated glycosaminoglycan—is an important Inclusion criteria
structural component of the extracellular cartilage matrix.
On the articular system, chondroitin sulfate links to Men or women over 40 years of age were eligible to par-
monomers with high molecular weights. The proteoglycan ticipate if they: (1) had documented knee osteoarthritis
aggregate exhibits viscoelastic and hydration properties evidenced with X-rays over the past 6 months with Kell-
and an ability to interact with the surrounding tissue gren-Lawrence score grade II or III [20]; (2) had pain and
through electric charges, leading to protection of the car- functional impairment for at least 3 months and Western
tilaginous tissues. Furthermore, chondroitin sulfates are Ontario and McMaster Universities Osteoarthritis Index
inhibitors of extracellular proteases involved in the meta- (WOMAC) [21] subscore A (severity of pain) C25 (on a
bolism of connective tissues and stimulate proteoglycan scale of 100); and (3) needed hyaluronic acid injections
production by chondrocytes in vitro; they also inhibit car- after the failure or intolerance to first-line analgesics or non
tilage cytokine production and induce apoptosis of articular steroidal anti-inflammatory drugs. The main exclusion cri-
chondrocytes [16]. Preliminary clinical trials were in favor teria were: severe hydrarthrosis; inflammatory rheumatism;
of the effectiveness of intra-articular injections of sodium history of knee trauma in the past 6 months; history of
hyaluronate-chondroitin sulfate. Thus, in a 3-month mul- arthroplasty or major surgery on the target knee in the past
ticentric pilot study, a series of three weekly injections of a 6 months; history of arthroscopy or surgery on the target
combination of hyaluronic acid-chondroitin sulfate was knee in the past 3 months; planned knee surgery during the
well tolerated and decreased pain in patients with knee study; history of septic arthritis of the knee; knee wound or
osteoarthritis [17]. A recent clinical study suggested that a skin condition; crural or sciatic radiculalgia of the lower
single injection of sodium hyaluronate-chondroitin sulfate limb; tendinopathy; symptomatic homolateral or contralat-
in patients with lateral epicondylitis offer better pain ben- eral hip disease; venous or lymphatic stenosis of the lower
efits for 6 months after injection than intra-articular corti- limb; medical history of venous thromboembolism (in-
costeroids [18]. In an exploratory study, the effectiveness cluding pulmonary embolism) or patient with high risk of
of intra-articular injections of a solution combining hya- venous thromboembolism; patient with a history of auto-
luronic acid and chondroitin sulfate was assessed in 40 immune disease; treatment with diacerein, avocado soy
patients with knee osteoarthritis [19]. The clinical unsaponifiables, glucosamine sulfate/chondroitin starting

less than 3 months previously or with dosage modified Comparisons were made using Student’s t test for
during the past 3 months; recurrent episodes of chondro- quantitative criteria and Chi2 test for non-ordinal qualita-
calcinosis; previous treatment with viscosupplementation; tive variables (or Fisher’s exact test) and Wilcoxon’s test
injection of corticosteroids into the knee under study less for ordinal data. The threshold for significance was set at
than 3 months previously; known hypersensitivity to hya- 5 %.
luronic acid or substances with similar activity; ongoing The analyses were performed using SAS version 9.2
anticoagulant therapy; pregnant or breastfeeding women. (SAS Institute, Cary, NC).
Treatment and clinical assessments

Results
Demographic, description and history of knee osteoarthri-
tis, concomitant treatments and WOMAC subscore A were Patient characteristics
recorded at inclusion visit. The WOMAC (Western Ontario
and McMaster Universities) index is used to assess patients Among 132 screened patients, 112 were analyzed (20
with osteoarthritis of the hip or knee [21]. The subscore A patients were \40 years of age and/or had a WOMAC
of the index is for pain assessment in five different cir- subscore A \25).
cumstances: during walking (A1), using stairs (A2), in bed The characteristics of patients at inclusion are summa-
(A3), sitting or lying (A4) and standing (A5). For each rized in Table 1. Two out of three patients were women
item, pain is graded from 0 (none) to 100 (extreme). The and the mean age was 65.4 years (range from 44 to
sum for the five items is divided by five to give WOMAC 88 years). Two-thirds of patients had a body mass index
subscore A. (BMI) above 25 kg/m2. The most frequent locations of
Patients received three intra-articular injections of knee osteoarthritis were the medial compartment (34.8 %;
Arthrum HCSÒ (40 mg hyaluronic acid and 40 mg chon- 39/112), tricompartmental (32.2 %; 35/112), and patello-
droitin sulfate in 2 mL) 1 week apart. Assessment of
treatment effectiveness and safety was performed during Table 1 Characteristics of patients at inclusion
follow-up visits at 1 month, 3 months and 6 months after
Characteristic N = 112
the last intra-articular injection.
The effectiveness assessment during the follow-up visits Age, years
included: WOMAC subscore A, relief of pain using a Mean (SD) 65.4 (10.6)
visual analogic scale (VAS) ranging from 0 (‘‘maximum Median (range) 66 (44–88)
relief’’, i.e., no pain) to 100 (‘‘no relief’’, i.e., maximal Female gender, n (%) 74 (66.1)
pain) and consumption of analgesic drugs using a VAS Body mass indexa (kg/m2)
ranging from 0 (‘‘no consumption of analgesics’’) to 100 Mean (SD) 26.4 (4.1)
(‘‘maximal consumption of analgesics)’’. \18.5 3 (2.8)
Adverse events were recorded immediately after the (18.5–25) 32 (29.4)
injections and during the follow-up visits. (25–30) 57 (52.3)
C30 17 (15.6)
Statistical analysis Study knee, n (%)
Right 67 (59.8 %)
Data from previous studies were used to estimate the Left 44 (3.3 %)
sample size [22, 23] With a loss to follow-up equal to Right and left 1 (0.9 %)
10 %, it was estimated that a sample size of 122 patients Duration of knee osteoarthritisb, years, mean (SD) 3.0 (3.5)
would provide 50 % power to detect a significant change of Radiological stage, n (%)
WOMAC subscore A (with alpha-risk at 5 %). Grade II 64 (57.1 %)
The primary endpoint was the change of WOMAC
Grade III 48 (42.9 %)
subscore A from inclusion to end of study. The secondary b
Prior knee surgery n (%) 29 (27.9 %)
endpoints were the change of WOMAC subscore A from
Prior physical medicine and rehabilitationc 48 (43.6 %)
inclusion to month 1 or month 3, relief of pain at months 1,
At least one analgesic drug within 3 months 87 (77.7 %)
3 and 6, consumption of analgesic drugs from baseline to
a
months 1, 3 and 6 and global assessment by the investigator Missing data for three patients
b
at the end of the study for the three criteria: pain reduction, Missing data for eight patients
c
improved mobility and consumption of analgesics. Missing data for two patients

femoral/medial femoro-tibial (10.7 %; 12/112). Radiolog- statistically significant compared to baseline (Wilcoxon
ical evaluation of osteoarthritis showed a Kellgren–Lawr- signed rank test, P \ 0.0001).
ence stage 2 in 64 (57, 1 %) cases, and a Kellgren–
Lawrence stage 3 in 48 (42, 9 %) cases. Pain relief
The mean duration of knee osteoarthritis was 3 years;
27.9 % (29/104) of patients underwent knee surgery (the One month after the last intra-articular injection, the mean
main operation was meniscectomy; 35.7 %, 10/28). The pain relief was assessed at 35.9 (25.6) on VAS by patients
median time between operation and inclusion was 7 years (0, maximal pain relief; 100, no relief). Pain relief con-
(n = 28). Just under half of patients had benefited from tinued to decrease at 3 and 6 months: 28.1 (23.1) and 26.1
physical treatment and rehabilitation medicine (43.6 %; (25.4), respectively. Compared to the values at 1 month,
48/110). the values of pain relief at 3 and 6 months were statistically
About four out of five patients (77.7 %; 87/112) were significant (P \ 0.0001 and P = 0.0048, respectively).
taking at least one analgesic treatment during the 3 months
prior to the intra-articular injections: NSAIDs for 74.1 % Consumption of analgesic drugs
(83/112) of patients and an analgesic other than NSAIDs
for 34.9 % (38/109) of patients. There was a moderate One month after the last intra-articular injection, the
relief due to the analgesic treatment: on a VAS from 0 (no patients assessed on a VAS their mean consumption of
relief) to 100 (maximal relief), the mean relief due to the analgesic drugs from 0 (no consumption of analgesics) to
analgesic treatment was 50.8 and 53.9 according to the 100 (maximal consumption of analgesics). The mean (SD)
investigator and the patient (n = 87), respectively. consumption decreased with time: 28.6 (24.4) at 1 month,
19.9 (21.7) at 3 months and 17.1 (22.3) at 6 months.
Severity of pain—WOMAC subscore A Compared to the values at 1 month, the scores of the
consumption of analgesic drugs at 3 and 6 months were
On inclusion, the mean WOMAC subscore A was 52.1 significantly decreased (P \ 0.0001 for both times).
(range 26–86). At 6 months, the mean WOMAC subscore
A was 20.5 (range 0–80). Thus, the decrease of the sub- Global assessment by investigators
score was -31.4 (P \ 0.0001; Wilcoxon signed rank test).
The change in the WOMAC subscore A during the study The investigators were satisfied or very satisfied as regards
is summarized in Fig. 1, and the changes in the five items the therapeutic effectiveness of sodium hyaluronate-chon-
of WOMAC subscore A (A1, walking; A2, using stairs; droitin sulfate in reducing pain (77 %), improving mobility
A3, in bed; A4, sitting or lying; A5, standing) are detailed (78 %) and reducing the consumption of analgesics (74 %)
in Table 2. One month after the last injection, the mean (Fig. 2).
score decreased to 25.7 and pain continued to decrease Overall, about 80 % of investigators stated that the
with a mean score of 20.4 at 3 months. This decrease in the results of the intra-articular injections of sodium hyalur-
WOMAC subscore A at 1 month and 3 months was onate-chondroitin sulfate combination were satisfactory or
very satisfactory.
80 Complications
Pain severity (WOMAC subscore A)
* p < 0.0001 vs. baseline

70
One adverse effect was reported by one patient. This
60
adverse event was knee tumefaction, which lasted 3 days
52.1
50 after the first intra-articular injection.
*
40 * *
30 Discussion
25.7
20.4 20.5
20
Viscosupplementation with hyaluronic acid alone has
10
demonstrated moderate but significant effectiveness vs
0 placebo in terms of pain and function in knee osteoarthritis
0 1 3 6
[11]. Our study is the first, to our knowledge, to assess the
Months after last injection
effectiveness and safety of injections of Arthrum HCSÒ in
Fig. 1 Pain severity (WOMAC subscore A) after three injections of a relatively large population of patients with knee
sodium hyaluronate-chondroitin sulfate in knee osteoarthritis osteoarthritis. We observed that the severity of pain

Table 2 Pain severity

Time (months) 0 1 3 6
(WOMAC subscore A) after
N 112 111 111 109
three intra-articular injections of
sodium hyaluronate-chondroitin Items of WOMAC A
sulfate. Results are given as
mean (SD). For each item, pain A1 (walking) 46.0 (22.2) 23.3 (19.8) 17.7 (18.3) 16.7 (20.3)
is graded from 0 (none) to 100 A2 (using stairs) 72.1 (17.3) 37.0 (22.9) 31.0 (22.8) 31.0 (26.7)
(extreme) A3 (in bed) 34.9 (25.2) 15.5 (19.0) 10.9 (17.7) 10.6 (18.0)
A4 (sitting or lying) 60.6 (21.4) 30.1 (21.7) 24.6 (19.4) 25.4 (23.8)
A5 (standing) 46.8 (23.5) 22.4 (22.5) 17.7 (18.8) 18.8 (22.2)
WOMAC A 52.1 (15.2)a 25.7 (17.4)a 20.4 (16.3)a 20.5 (19.7)a
a
No treatment baseline score
80% systematic review and meta-analysis of randomized saline-

Inadequate Moderate Satisfactory Very satisfactory controlled trials for US-approved intra-articular hyaluronic
70%
acid [24]. There were no statistically significant differences
60%
between hyaluronic acid and saline controls for any safety
Patients (%)
50%
50% outcome.
42% 41%
40% 35%
37% It is now debated that surgical procedures in knee
30% osteoarthritis should be avoided as far as possible or at
24%
20% 18% least delayed [25, 26]. The restoration of the viscoelas-
14%
10%
13% 12% ticity of the synovial fluid in order to protect cartilage, if
10% 4% possible during the early states of the disease, is an
0% attractive therapeutic option. Moreover, a medico-eco-
Pain reduction Improved mobility Consumption of
analgesics decreased nomic evaluation showed that, together with clinical
Overall assessment by the investigator benefits, costs of knee osteoarthritis decreased after
hyaluronic acid injections due to the decreased need for
Fig. 2 Overall assessment of efficacy of three injections of sodium other treatments [23]. However no therapies have been
hyaluronate-chondroitin sulfate in knee osteoarthritis by the
investigator shown to alter the natural history of osteoarthritis. In the
absence of disease modifying osteoarthritis drugs,
assessed with the WOMAC subscore A decreased signifi- treatment of osteoarthritis is focused on controlling
cantly from 52.1 to 20.5 at 6 months. The relief was symptoms, especially pain [10, 27, 28]. Until prospective
already significant 1 month after the last injection. These studies on the efficacy of hyaluronic acid on knee
results were confirmed by the assessment of pain relief and arthroplasty delay are completed, intra-articular treat-
the decrease in the consumption of analgesics with VAS. ment must be considered an additional non-operative
Approximately three out of four investigators were satis- strategy for relief of symptoms.
fied/very satisfied as regards to the therapeutic effective- There are some limitations of the study. First, there was
ness of the injections of sodium hyaluronate-chondroitin no control group. Indeed, there is a debate on the effec-
sulfate in reducing pain, improving mobility and reducing tiveness of viscosupplementation in osteoarthritis, but
the consumption of analgesics. some meta-analyses found an advantage of viscosupple-
The pilot study of Maheu et al. [17] in 41 patients with mentation over sham intervention [14, 24, 25]. Therefore, it
femoro-tibial knee osteoarthritis also reported an was difficult to justify a sham control in one group. Nev-
improvement 3 months after three 2-mL injections of ertheless, the kinetics of the effect observed in the present
hyaluronic acid (12 mg/mL) plus chondroitin sulfate study conform to the conclusions of a meta-analysis on
(30 mg/mL). The mean VAS score decreased from 61 at viscosupplementation that showed that effectiveness
baseline to 29 after 3 months (60 % of patients reported an became significant at 4 weeks, peaked at 8 weeks and
improvement above 50 %). Although the dosages of the persisted for 6 months [14]. Another limitation was the
compounds were slightly different, these results are con- absence of demonstration of the benefit of the addition of
sistent with those of the present study. chondroitin sulfate to hyaluronic acid. With a comparable
The very low proportion of patients with adverse events total number of patients, the statistical power of the trial
confirms the safety of viscosupplementation in knee would decrease with an additional treatment group (hya-
osteoarthritis. The harmlessness of viscosupplementation luronic acid alone). This issue should certainly be addres-
in knee osteoarthritis has been confirmed in a recent sed in further studies.

Arthrum HCSÒ is a new intra-articular treatment com- 9. Zhang W, Nuki G, Moskowitz RW, Abramson S, Altman RD,
bining in the same injection two compounds that are defi- Arden NK (2010) OARSI recommendations for the management
of hip and knee osteoarthritis: part III: changes in evidence fol-
cient in osteoarthritis. Chondroitin sulfate is an essential lowing systematic cumulative update of research published
component of cartilage and is present also in synovial fluid. through January 2009. Osteoarthr Cartil 18:476–499
Our results suggest that intra-articular injections of 10. Jevsevar DS (2013) Treatment of osteoarthritis of the knee:
Arthrum HCSÒ (sodium hyaluronate plus chondroitin sul- evidence-based guideline, 2nd edition. J Am Acad Orthop Surg
21:571–576
fate) in patients with knee osteoarthritis allows a safe and 11. Legre-Boyer V (2015) Viscosupplementation: techniques, indi-
effective control of pain. These results should be confirmed cations, results. Orthop Traumatol Surg Res 101:S101–S108
in a randomized controlled study. 12. Balazs EA, Denlinger JL (1993) Viscosupplementation: a new
concept in the treatment of osteoarthritis. J Rheumatol Suppl
Compliance with ethical standards 39:3–9
13. Pederzini LA, Milandri L, Tosi M, Prandini M, Nicoletta F
Conflict of interest All authors received a consultancy fee reim- (2013) Preliminary clinical experience with hyaluronan anti-ad-
bursement from LCA Pharmaceutical, Chartres, France, to contribute hesion gel in arthroscopic arthrolysis for posttraumatic elbow
to completing this study. This research received no specific grant from stiffness. J Orthopaed Traumatol 14(2):109–114
any funding agency in the public, or not-for-profit sectors. 14. Bannuru RR, Natov NS, Obadan IE, Price LL, Schmid CH,
McAlindon TE (2009) Therapeutic trajectory of hyaluronic acid
Ethical standards All patients gave informed consent prior being versus corticosteroids in the treatment of knee osteoarthritis: a
included into the study. All procedures involving human participants systematic review and meta-analysis. Arthritis Rheum
were in accordance with the 1964 Helsinki Declaration and its later 261:1704–1711
amendments. The study was approved by the Research Ethics 15. Wang CT, Lin J, Chang CJ, Lin YT, Hou SM (2004) Therapeutic
Committee. effects of hyaluronic acid on osteoarthritis of the knee. A meta-
analysis of randomized controlled trials. J Bone Joint Surg Am
86:538–545
16. Bali JP, Cousse H, Neuzil E (2001) Biochemical basis of the
pharmacologic action of chondroitin sulfates on the osteoarticular
system. Semin Arthritis Rheum 31(1):58–68
17. Maheu E, Zaı̈m M, Appelboom T, Bensaber M, Cadet C, Saurel
A (2010) Evaluation of intra articular injections of hyaluronic
acid and chondroitine sulfate for knee arthritis treatment: a
multicentric pilot study with 3 monts follow-up. National
rheumatolgy meeting 2010 (Société Française de Rhumatologie).
http://www.rhumatologie.asso.fr/data/ModuleProgramme/Page
Site/2010-1/Resume/6007.asp. Accessed date 2015
References 18. Tosun HB, Gumustas S, Agir I, Uludag A, Serbest S, Pepele D
(2015) Comparison of the effects of sodium hyaluronate-chon-
droitin sulphate and corticosteroid in the treatment of lateral
1. Murray CJ, Vos T, Lozano R, Naghavi M, Flaxman AD, Michaud
epicondylitis: a prospective randomized trial. J Orthop Sci
C et al (2012) Disability-adjusted life years (DALYs) for 291
20:837–843
diseases and injuries in 21 regions, 1990–2010: a systematic
19. Henrotin Y, Hauzeur JP, Bruel P, Appelboom T (2012) Intra-
analysis for the Global Burden of Disease Study 2010. Lancet
articular use of a medical device composed of hyaluronic acid
380:2197–2223
and chondroitin sulfate (Structovial CS): effects on clinical,
2. Guccione AA, Felson DT, Anderson JJ, Anthony JM, Zhang Y,
ultrasonographic and biological parameters. BMC Res Notes
Wilson PW et al (1994) The effects of specific medical conditions
5:407–411
on the functional limitations of elders in the Framingham study.
20. Ahlbäck S (1968) Osteoarthrosis of the knee: a radiographic
Am J Public Health 84:351–358
investigation. Acta Radiol Stockholm suppl 277:70–72
3. Gupta S, Hawker GA, Laporte A, Croxford R, Coyte PC (2005)
21. McConnell S, Kolopack P, Davis AM (2001) The Western
The economic burden of disabling hip and knee osteoarthritis
Ontario and McMaster Universities Osteoarthritis Index
(OA) from the perspective of individuals living with this condi-
(WOMAC): a review of its utility and measurement properties.
tion. Rheumatology 44:1531–1537
Arthritis Rheum 45:453–461
4. Le Pen C, Reygrobellet C, Gerentes I (2005) Financial cost of
22. Ehrich EW, Davies GM, Watson DJ, Bolognese JA, Seidenberg
osteoarthritis in France. The ‘‘COART’’ France study. Joint Bone
BC, Bellamy N (2000) Minimal perceptible clinical improvement
Spine 72:567–570
with the Western Ontario and McMaster Universities
5. Hunter DJ (2011) Osteoarthritis. Best Pract Res Clin Rheumatol
osteoarthritis index questionnaire and global assessments in
25:801–814
patients with osteoarthritis. J Rheumatol 27:2635–2641
6. Hunter DJ (2015) Viscosupplementation for osteoarthritis of the
23. Mazieres B, Bard H, Ligier M, Bru I, d’Orsay GG, Le Pen C
knee. N Engl J Med 372:1040–1047
(2007) Medicoeconomic evaluation of hyaluronic acid for knee
7. Bennell KL, Hunter DJ, Hinman RS (2012) Management of
osteoarthritis in everyday practice: the MESSAGE study. Joint
osteoarthritis of the knee. BMJ 345:e4934
Bone Spine 74:453–460
8. Hochberg MC, Altman RD, April KT, Benkhalti M, Guyatt G,
24. Miller LE, Block JE (2013) US-approved intra-articular hya-
McGowan J et al (2012) American College of Rheumatology
luronic acid injections are safe and effective in patients with knee
2012 recommendations for the use of nonpharmacologic and
osteoarthritis: systematic review and meta-analysis of random-
pharmacologic therapies in osteoarthritis of the hand, hip, and
ized, saline-controlled trials. Clin Med Insights Arthritis Mus-
knee. Arthritis Care Res (Hoboken) 64:465–474
culoskelet Disord 6:57–63

25. Lo GH, LaValley M, McAlindon T, Felson DT (2003) Intra- medical management of osteoarthritis of the hip and knee.
articular hyaluronic acid in treatment of knee osteoarthritis: a Arthritis Rheum 43(9):1905–1915
meta-analysis. JAMA 290:3115–3121 28. Zhang W, Nuki G, Moskowitz RW, Abramson S, Altman RD,
26. Abbott Thomas, Altman Roy D, Dimeff Robert, Fredericson Arden NK et al (2010) OARSI recommendations for the man-
Michael, Vad Vijay, Vitanzo Peter et al (2013) Do hyaluronic agement of hip and knee osteoarthritis: part III: changes in evi-
acid injections delay total knee replacement surgery? [abstract]. dence following systematic cumulative update of research
Arthritis Rheum 65(Suppl 10):2139 published through January 2009. Osteoarthr Cartil 18(4):476–499
27. American College of Rheumatology Subcommittee on
Osteoarthritis Guidelines (2000) Recommendations for the

10
Congenital idiopathic talipes equinovarus
before and after walking age: observations and strategy
of treatment from a series of 88 cases
Cesare Faldini1,2 • Francesco Traina2 • Matteo Nanni2 • Ilaria Sanzarello3 •
Raffaele Borghi1 • Fabrizio Perna1
Abstract open surgery was higher in cases of delayed treatment. In

Background We reviewed a series of newborns, toddlers cases of relapse, re-casting and/or more extensive surgery
and ambulating children affected by idiopathic congenital was considered.
talipes equinovarus (clubfoot). Taking into account the Conclusions Early treatment enables a high rate of good
time of diagnosis, stiffness of the deformity and walking correction to be obtained with serial casting and limited
age, nonsurgical or surgical treatment was considered. This surgery. Conversely, if the deformity is observed after
study reports clinical outcomes, early complications and walking age surgery should be considered. Serial casting in
relapse at mid-term follow-up. cases of late observation and relapse have demonstrated
Materials and methods Fifty-two clubfeet were diag- encouraging results.
nosed at birth, 12 in non-ambulating children aged between Level of evidence IV.
4 and 12 months and 24 in ambulating children. Feet were
classified using the Pirani score. Newborns and toddlers Keywords Clubfoot Casting Re-casting Surgical
were treated with serial casting (Ponseti); however, tod- treatment Walking age
dlers also underwent open Achilles tendon lengthening (2
feet) and posteromedial release (3 feet). In all ambulating
children, surgical treatment was always performed: selec- Introduction
tive medial release combined with cuboid subtraction
osteotomy (1 foot), posteromedial release (6 feet), and Congenital idiopathic talipes equinovarus, also known as
posteromedial release combined with cuboid subtraction clubfoot, is a deformity present at birth that is characterized
osteotomy (17 feet). by a permanent alteration of the morphology of the foot
Results The average follow-up was 5 years (1–6 years). and its relationship with the leg so the foot cannot lean on
In newborns treated with Ponseti, the results were excellent the ground in a physiological way. Therefore, the treatment
in 42 feet, good in 6, and poor in 4. In non-ambulating should aim to correct all the components of the deformity
children, the results were excellent in 9 feet, and good in 3. (cavus, forefoot varus, hindfoot varus, equinus), in order to
In ambulating children, the results were excellent in 5 feet, restore as much as possible of the physiologic morphology
good in 16, and poor in 3. No major complications were and function of the foot to allow plantigrade stance and
reported. No overcorrections were observed. The need for proper gait.
If diagnosed at birth, clubfoot can be successfully trea-
& Cesare Faldini
ted nonsurgically. When performed on reducible deformi-
[email protected] ties in newborns or toddlers, manipulations and serial
casting, as described by Ponseti et al. [1], or manipulations
1
University of Bologna, Bologna, Italy and functional taping, as described by Masse et al. [2], can
2
Dipartimento Rizzoli-Sicilia, Istituto Ortopedico Rizzoli, achieve successful correction in a high percentage of cases.
Strada Statale 113 km 246, 90011 Bagheria, PA, Italy In most cases, percutaneous Achilles tendon section often
3
University of Messina, Messina, Italy completes this kind of treatment in order to obtain

correction of the equinus deformity. Over recent years Six non-ambulating children (10 feet) who came to our
nonsurgical treatment has been popularized due to excel- observation later (between 4 and 12 months) were previ-
lent results reported by several authors [3–6]. ously treated at other hospitals with serial casting, whereas
However, if the deformity is diagnosed later or observed two children with unilateral deformity were not previously
after unsuccessful conservative treatment, correction may treated before our observation. Of the 24 clubfeet that came
require surgical treatment because of soft-tissue retraction, to our observation after walking age, 10 were previously
joint stiffness and bony changes that make the deformity treated with serial casting and 8 with serial casting and
more rigid as the child ages. A neglected or uncorrected posteromedial release at other hospitals, while 6 feet were
deformity forces the child to start walking with the foot not previously treated at all.
leaning on its lateral aspect [7]. The weight bearing Clinical assessment of the deformity focused particu-
worsens the equinus and the supination, and the lateral larly on reducibility of the talo-navicular joint and severity
column of the foot (calcaneus and cuboid) grows more and of the equinus. In ambulating children, care was taken to
becomes longer than the medial column (talus, navicular, assess the stiffness of the ankle, subtalar and midtarsal
and cuneiform bones), eventually making the foot very stiff joint, and muscular function, strength and atrophy was also
and the deformity no longer reducible [8]. In cases of evaluated. In these children, pain, skin problems, gait and
neglected clubfoot, unsuccessful previous surgery and shoe wearing impairment were also noted.
recurrent deformity, it could be challenging to obtain the Each foot was classified using the Pirani score [18].
correction with manipulation and casting. Therefore, Radiographic assessment was not performed in newborns
extensive surgery is often required to obtain adequate and toddlers at the time of diagnosis, whereas it was per-
correction in these cases [9–13]. Different surgical tech- formed in all feet in ambulating children in order to assess
niques have been advocated for the treatment of unre- skeletal maturity and to plan surgery.
ducible congenital clubfoot, including soft-tissue release, In all newborns the treatment consisted of manipulation
tenotomy and tendon elongation and transfer, joint release, and serial casting according to the Ponseti technique [19].
osteotomy and joint fusion; however, the choice of correct At the end of casting, residual equinus was corrected by
surgical procedure for each case is still a cause of concern percutaneous Achilles tendon section in all cases but one.
[14–16]. Moreover, the optimal timing for surgery still During treatment, radiographs in forced ankle dorsiflexion
remains controversial, although some authors suggest sur- were performed in a 3-month-old child in order to better
gery within the first year of life, just before the child starts investigate reducibility of the equinus and confirm that the
to walk [17]. foot did not require Achilles tenotomy.
We reviewed a series of patients with clubfoot who had Furthermore, in non-ambulating children between 4 and
undergone nonsurgical and/or surgical treatment. Accord- 12 months of age who were observed later, Ponseti serial
ing to the time of diagnosis and the reducibility of the casting was performed according to recent literature [20].
deformity, from reducible deformity observed at birth to In 7 feet, the deformity was corrected only by casting and
unreducible deformity observed after walking age, serial percutaneous Achilles tendon section. However, in 5 feet,
casting and minor or extensive surgical procedures were Ponseti protocol alone was not sufficient and surgery was
performed. The aim of this study is to present the treatment performed in order to correct residual deformity after
of a series of 88 clubfeet observed in newborns, toddlers casting around the first year of life: open Achilles tendon
and ambulating children, and to report the results, early lengthening combined with posterior ankle and subtalar
complications and relapse at mid-term follow-up. joint release was performed in 2 cases in order to correct
isolated residual equinus, whereas posteromedial release,
performed through a single medial approach according to
Materials and methods the technique described by Codivilla [21] was performed in
3 cases to correct residual supination (equinovarus).
Eighty-eight clubfeet in 58 patients were observed (bilat- Surgical treatment was performed in all ambulating
eral deformity in 30 children). Fifty-two clubfeet (34 children. In one case that presented isolated relapse of
patients) were diagnosed at birth, 12 (8 patients) were varus after a previous posteromedial release, a selective
observed later in non-ambulating children aged between 4 medial release, including tibialis posterior tendon elonga-
and 12 months, and 24 (16 patients) were observed in tion, talonavicular joint reduction, and abductor hallucis
ambulating children aged between 1 and 3 years. Clubfeet tenotomy was performed combined with cuboid subtraction
associated with neurologic or syndromic disorders were not osteotomy. In 6 feet that presented with partially reducible
considered in this study. The study was approved by the equinus-varus-supination, with complete reducibility of
Ethics Committee and the parents of all the patients pro- varus, a posteromedial release was performed. In 17 feet
vided informed consent to the treatment. that presented with completely unreducible deformity due

Congenital idiopathic talipes equinovarus before and after walking age: observations and strategy of treatment... 69
to excessive joint stiffness and excessive elongation of the
Clinical evaluation was summarized with the Pirani score. Radiographic assessment was not performed in non-ambulating children except in one case. During treatment with serial casting,
release ? cuboid
radiographs were performed in a 3-month-old child, forcing the ankle in dorsiflexion in order to better evaluate the reducibility of the equinus. In this case, radiographs showed reduction of the
equinus and thus there was no need for Achilles tenotomy. All newborns and non-ambulating children were treated with serial casting before performing surgery. All ambulating children
lateral column of the foot, posteromedial release combined
posteromedial
subtractive
with lateral release of the subtalar joint and cuboid sub-
and lateral
osteotomy
traction osteotomy fixed with a percutaneous Kirschner
wire [22] was performed (Table 1).
17
In children who underwent Ponseti serial casting after
postero-
release
medial
Achilles tenotomy, a foot-abduction brace was applied full-
time for the first 4–6 months, and then only during sleep
6
(continued until 4 years of age) and symmetrical straight
medial release
last shoes were prescribed when the child started to walk.
subtractive
osteotomy
Furthermore, children who underwent open Achilles tendon
? cuboid
Selective
elongation followed the same protocol. Children who
underwent medial release and posteromedial release (with
1
or without cuboid osteotomy) after surgery wore an above-
subtalar joint release

posterior ankle and
knee cast for 10 weeks (which was changed after 5 weeks)
when the percutaneous Kirschner wire was also removed,
lenghtening ?
Open Achilles
followed by application of a foot-abduction brace during the
night and symmetrical straight last shoes during walking.
At the last available follow-up each foot was evaluated
2
in terms of plantigrade position, joint stiffness, muscular
Percutaneous
function, pain, shoewearing and walking. We experienced
tenotomy
difficulty in finding a suitable scoring system to assess the
Achilles
results in this case series; however, we decided to use the
51
Pirani score to summarize the results. In cases of planti-
7
Treatment
grade feet, results were considered as excellent for scores

(Ponseti)
casting
between 0 and 1, good between 1.5 and 2.5, and poor if C3;
Serial
all cases presenting residual equinus were considered poor.
52
12
The need for further or revision surgery was also noted.
Radiographic
1 (forced ankle
dorsiflexion
X-rays at 3
assessment
months)
Results
24
–
Table 1 Population data, clinical and radiographic assessment and treatment
C3
The average follow-up time was 5 years (range 1–6 years).

42
10
21
In newborns who underwent Ponseti serial casting, 4–8
1.5–2.5
Pirani score
casts (average 6 casts) were necessary to obtain the cor-

10
rection. In non-ambulating children, 6–12 casts (average 8

2
casts) were made until correction was obtained or surgical

0–1
treatment was considered necessary to obtain complete

correction.
posteromedial
No major complications were reported in children dur-

casting ?
release
ing treatment with serial casting. Sometimes, after cast

Serial
removal, skin redness or slight swelling of the foot was

Previous treatment
observed and application of the following cast was delayed

casting
Serial
by 24–48 h. Neurovascular injury or profuse bleeding after

10
10
percutaneous Achilles tenotomy were not observed. No

underwent surgical treatment
infection or wound problems were reported including

None
children who had undergone more extensive surgery.

52
Newborns and toddlers learned to walk between 10 and

feet
No.
19 months and at last follow-up, every child was able to

of
52
12
24
wear shoes and walk without limping, even in cases of

child (4–12
Newborn (0–4
observation
ambulating
child ([12
Ambulating
residual deformity.
months)
months)
months)
Age of
Two ambulating children who had undergone postero-

Non-
medial release combined with cuboid osteotomy

experienced a delay of wound healing; one of these and correction was obtained after 4 casts which were
required secondary plastic surgery to restore adequate skin changed every 3 weeks without need of further surgery. In
coverage. At last follow-up, two ambulating children who the other two feet (children age 18 and 20 months,
had undergone surgical treatment and presented with respectively) presenting with isolated relapse of equinus,
residual deformity, reported mild pain during prolonged open Achilles tendon lengthening combined with posterior
walking. ankle and subtalar joint release was performed. Further
According to the Pirani score, the results in terms of surgery was also necessary in cases with poor results (3
plantigrade position and residual deformity in newborns feet) in ambulating children surgically treated with pos-
treated with Ponseti serial casting were excellent in 42 feet teromedial release. In these cases, relapse of equinovarus
(Fig. 1), good in 6, and poor in 4. In non-ambulating was treated with revision of posteromedial release com-
children observed and treated within 4 and 12 months of bined with cuboid subtraction osteotomy. No cases of
age, the results were excellent in 9 feet, and good in 3. In overcorrection were observed in this series.
ambulating children, the results were excellent in 5 feet
(Fig. 2), good in 16, and poor in 3 (Table 2).
In cases that were treated with only the Ponseti method Discussion
and presented poor results (4 feet), further treatment was
performed because of partial relapse observed after the Currently, most authors suggest conservative treatment for
children started walking. In two feet (child age 18 months) clubfoot, restricting the indication for surgery in cases of
with isolated relapse of varus, serial casting was attempted recurrent or resistant deformity [23–28]. In newborns,
conservative treatment with the Ponseti method is effec-
tive in most cases with a high rate of good results [29].
Besides manipulation and casting, minimally invasive
surgery consisting of percutaneous Achilles tenotomy, as
reported in the literature, very often completes the Ponseti
method; in our series all cases but one were treated with
this technique. In cases of severe or recurrent equinus, as
well as in older children, a greater Achilles retraction
could be suspected together with ankle and subtalar joint
stiffness. In these cases we assumed that percutaneous
Achilles tenotomy alone could not obtain satisfactory
correction; hence, a little more extensive surgery was
performed in some cases. Some authors also suggest per-
forming percutaneous Achilles tendon section in ambu-
lating children and in cases of relapse of equinus while
others advise open lengthening or use of percutaneous
incisions as described by Hoke [30–32]. We preferred to
perform open Achilles tendon lengthening in older chil-
dren or in cases of relapse of the equinus in order to also
appreciate the possible tightness of the ankle and subtalar
joints and then perform a posterior release of these joints
along with tendon lengthening.
In our series some children were observed and treated
later, after 4 months of age and before they started walk-
ing. In these patients, serial casting was performed
according to Ponseti as well as in newborns. However,
these cases revealed stiffer deformities which were more
challenging to treat. The number of casts necessary to
obtain correction was higher than in newborns (6–12 casts
versus 4–8 casts) and in some cases (5 of 12) complete
correction was not achieved. Therefore, we decided to
perform surgery before the children start walking. As
Fig. 1 Clinical aspect of monolateral right clubfoot at birth (a) and at
18 months after treatment with serial casting and percutaneous reported, isolated equinus was treated with open Achilles
Achilles tendon section (b) tendon lengthening and posterior ankle and subtalar

Fig. 2 Clinical aspect of bilateral neglected clubfoot in a 3-year-old child presenting with severe rigid deformity (a, b). Posteromedial release
combined with cuboid osteotomy allowed an excellent correction with plantigrade foot (c, d, e)
Table 2 Results
Age of observation (months) No. of feet Pirani score
Excellent 0–1 Good 1.5–2.5 Poor C3
Newborn (0–4) 52 42 (80.7 %) 6 (11.5 %) 4 (7.6 %)

Non-ambulating child (4–12) 12 9 (75 %) 3 (25 %) –
Ambulating child ([12) 24 5 (20.8 %) 16 (66.6 %) 3 (12.5 %)
According to the Pirani score, the results were excellent for scores between 0 and 1, good between 1.5 and
2.5, and poor for C3. All cases presenting with residual equinus were considered poor
release, while in cases of residual supination (equinovarus) allows the release of retracted tissues on the medial aspect
we decided to perform a posteromedial release according to of the foot, restores congruency of the talonavicular joint
the technique described by Codivilla . This technique (and naviculocuneiform and cuneometatarsal joints if

necessary), along with correction of the equinus by few complications and low rate of deformity relapse at
Achilles tendon lengthening and posterior release of the mid-term follow-up.
ankle and subtalar joint. It is still difficult to determine appropriate timing and
As suggested by some authors, conservative treatment technique for the surgical treatment of congenital club-
with manipulation and casting should be attempted not foot, and the approach for both primary and revision
only in older children but also in cases of neglected or surgery has been often defined as ‘a la carte’ [39]. The
relapsed clubfoot, even if this treatment does not always age of the patient at the time of diagnosis, walking age,
ensure a complete correction and prevent the need for stiffness of the deformity and previous treatment are all
surgery [33–37]. In our series, serial casting in the treat- major concerns in the choice of appropriate treatment in
ment of partial relapse after Ponseti treatment proved to be order to determine the need for surgery rather than non-
an encouraging option, even though our experience with surgical treatment, particularly within the first year of age.
this kind of approach is still very limited to date. There are some remarkable limitations to this study,
We always performed surgery in ambulating children. mostly concerning the non-homogeneous series of
Posteromedial release resulted adequate to obtain complete patients, the different grades of deformity, untreated and
correction of the deformity, when complete correction of previously treated patients, and the relatively short fol-
the forefoot varus was achieved using medial release. low-up. In our experience, newborns and toddlers can be
Conversely, in severe neglected clubfeet, posteromedial successfully treated nonsurgically with serial casting, even
release alone was not sufficient to completely correct the though this treatment seems to become more challenging
deformity. Excessive growth of the lateral column of the as the child ages or if it is not started just after birth.
foot compared with the medial column, that is shorter as a Conversely, we advise surgical treatment after walking
consequence of tissues retraction and joint dislocation, age. Nevertheless early reports about re-casting with the
produces a more severe equinovarus deformity. Further- Ponseti technique seem encouraging, extending the
more, in older children, ambulation enhances stiffness of opportunity for nonsurgical or less-invasive surgical pro-
the deformity. In these cases, posteromedial release alone cedures (i.e., Achilles tendon lengthening) in older chil-
was not sufficient to obtain a complete correction. Lateral dren and in cases of relapse and more severe deformities.
release and cuboid subtraction osteotomy combined with
posteromedial release rebalanced the length of medial and Compliance with ethical standards
lateral columns of the foot, thus helping complete reduc- Ethical standards The present study was performed in accordance
tion of the talonavicular joint and allowing complete cor- with the ethical standards of the 1964 Declaration of Helsinki as
rection of forefoot varus. Posteromedial release combined revised in 2000. The study was approved by the responsible Ethics
with cuboid subtraction osteotomy also proved to be a Committee and the parents of all the patients provided informed
consent.
viable option in cases of relapse after extensive surgical
treatment. In cases of less severe deformity, as well as in Conflict of interest The authors declare that they have no conflict
those feet still presenting residual or relapsed isolated of interest.
varus, following the same principles of the described
techniques, we believe that cuboid subtraction osteotomy
can be combined with selective medial release in order to
obtain the correction.
Many complications have been reported after extensive
surgical treatment of clubfoot, mainly incomplete correc-
tions or overcorrections, skin problems and neurovascular
injuries. Moreover, some authors have reported a loss of
correction over time and residual deformity after skeletal References
maturity, stiffness and/or early degenerative changes
involving the ankle, subtalar and midtarsal joints, pain and 1. Ponseti IV, Smoley EN (1963) Congenital clubfoot: the results of
muscle weakness; therefore, they do not recommend sur- treatment. J Bone Joint Surg Am 45:261–344
2. Masse P (1977) Le traitement du pied bot par la methode
gery as a primary treatment for clubfoot [38]. Based on our ‘‘fonctionelle’’. In: Cahier d’enseignement de la SOFCOT, vol 3.
experience, surgical treatment should be considered in Paris: Expansion Scientifique, pp 51–56
walking children because of the greater stiffness of the 3. Morcuende JA, Abbasi D, Dolan LA, Ponseti IV (2005) Results
foot, mainly in cases of obvious deformity in which the of an accelerated Ponseti protocol for clubfoot. J Pediatr Orthop
25:623–626
lateral column of the foot is much longer than the medial as 4. Colburn M, Williams M (2003) Evaluation of the treatment of
a consequence of an unbalanced growth. Surgical treatment idiopathic clubfoot by using the Ponseti method. J Foot Ankle
allowed us to obtain a satisfactory rate of good results with Surg 42(5):259–267

5. Gupta A, Singh S, Patel P, Patel J, Varshney MK (2008) Eval- 22. Faldini C, Traina F, Di Martino A, Nanni M, Acri F (2013) Can
uation of the utility of the Ponseti method of correction of selective soft tissue release and cuboid osteotomy correct
clubfoot deformity in a developing nation. Int Orthop neglected clubfoot? Clin Orthop Relat Res 471(8):2658–2665
32(1):75–79 23. Ippolito E, Farsetti P, Caterini R, Tudisco C (2003) Long-term
6. Jowett CR, Morcuende JA, Ramachandran M (2011) Manage- comparative results in patients with congenital clubfoot treated
ment of congenital talipes equinovarus using the Ponseti method: with two different protocols. J Bone Joint Surg Am 85:1286–
a systematic review. J Bone Joint Surg 93(9):1160–1164 1294
7. Changulani M, Garg NK, Rajagopal TS, Bass A, Nayagam SN, 24. Zhao D, Li H, Zhao L, Liu J, Wu Z, Jin F (2014) Results of
Sampath J, Bruce CE (2006) Treatment of idiopathic clubfoot clubfoot management using the Ponseti method: do the details
using the Ponseti method. Initial experience. J Bone Joint Surg matter? A systematic review. Clin Orthop Relat Res
88(10):1385–1387 472(4):1329–1336
8. Herold HV, Torok G (1973) Surgical correction of neglected 25. Siapkara A, Duncan R (2007) Congenital talipes equinovarus: a
clubfoot in the older child and adult. J Bone Joint Surg [Am] review of current management. J Bone Joint Surg Br 89(8):995–
55:1385–1395 1000
9. Loza ME, Bishay SN, El-Barbary HM, Hanna AA, Tarraf YN, 26. Uglow MG, Kurup HV (2010) Residual clubfoot in children. Foot
Lotfy AA (2010) Double column osteotomy for correction of Ankle Clin 15(2):245–264
residual adduction deformity in idiopathic clubfoot. Ann R Coll 27. Mehrafshan M, Rampal V, Seringe R, Wicart P (2009) Recurrent
Surg Engl 92(8):673–679 clubfoot deformity following previous soft-tissue release: mid-
10. Turco VJ (1971) Surgical correction of the resistant club foot. term outcome after revision surgery. J Bone Joint Surg Br
One-stage posteromedial release with internal fixation: a pre- 91(7):949–954
liminary report. J Bone Joint Surg Am 53:477–497 28. Hallaj-Moghadam M, Moradi A, Ebrahimzadeh MH (2015)
11. Turco VJ (1979) Resistant congenital club foot one-stage pos- Clinical outcome of posteromedial versus posteromedial-lateral
teromedial release with internal fixation. A follow-up report of a release for clubfoot. J Pediatr Orthop B 24(1):24–27
fifteen-year experience. J Bone Joint Surg Am 61:805–814 29. Radler C (2013) The Ponseti method for the treatment of con-
12. Mahadev A, Munajat I, Mansor A, Hui JH (2009) Combined genital club foot: review of the current literature and treatment
lateral and transcuneiform without medial osteotomy for residual recommendations. Int Orthop 2013(37):1747–1753
clubfoot for children. Clin Orthop Relat Res 467(5):1319– 30. Aydin BK, Senaran H, Yilmaz G, Acar MA, Kirac Y (2015) The
1325 need for Achilles tenotomy in the Ponseti method: is it pre-
13. Hassan FO, Jabaiti S, El Tamimi T (2010) Complete subtalar dictable at the initiation or during the treatment? J Pediatr Orthop
release for older children who had recurrent clubfoot deformity. B 24(4):341–344
Foot Ankle Surg 16(1):38–44 31. Bergerault F, Fournier J, Bonnard C (2013) Idiopathic congenital
14. Kuo KN, Smith PA (2009) Correcting residual deformity following clubfoot: initial treatment. Orthop Traumatol Surg Res 99:S150–
clubfoot releases. Clin Orthop Relat Res 467(5):1326–1333 S159
15. Thompson G, Hoyen H, Barthel T (2009) Tibialis anterior tendon 32. Hoke M (1931) An operation for the correction of extremely
transfer after clubfoot surgery. Clin Orthop Relat Res relaxed flad feet. J Bone Joint Surg Am 13(4):773–783
467:1306–1313 33. Garg S, Dobbs MB (2008) Use of the Ponseti method for
16. Elgeidi A, Abulsaad M (2014) Combined double tarsal wedge recurrent clubfoot following posteromedial release. Indian J
osteotomy and transcuneiform osteotomy for correction of Orthop 42(1):68–72
resistant clubfoot deformity (the ‘‘bean-shaped’’ foot). J Child 34. Bor N, Herzenberg JE, Frick S (2006) Ponseti management of
Orthop 8(5):399–404 clubfoot in older infants. Clin Orthop Relat Res 444:224–228
17. Holt JB, Oji DE, Yack HJ, Morcuende JA (2015) Long-term 35. Verma A, Mehtani A, Sural S, Maini L, Gautam VK, Basran SS,
results of tibialis anterior tendon transfer for relapsed idiopathic Arora S (2012) Management of idiopathic clubfoot in toddlers by
clubfoot treated with the Ponseti method: a follow-up of thirty- Ponseti’s method. J Pediatr Orthop B 21(1):79–84
seven to fifty-five years. J Bone Joint Surg Am 97(1):47–55 36. Parsa A, Moghadam MH, Jamshidi MH (2014) Relapsing and
18. Pirani S, Outebridge H, Moran M, Swawtsky B (1995) A method residual clubfoot deformities after the application of the ponseti
of evaluating the virgin clubfoot with substantial inter-observer method: a contemporary review. Arch Bone Jt Surg 2(1):7–10
reliability. POSNA Meeting, Miami 37. Alves C, Escalda C, Fernandes P, Tavares D, Neves MC (2009)
19. Ponseti IV (1992) Treatment of congenital clubfoot. J Bone Joint Ponseti method: does age at the beginning of treatment make a
Surgery Am 74(3):448–454 difference? Clin Orthop Relat Res 467(5):1271–1277
20. Lourenco AF, Morcuende JA (2007) Correction of neglected 38. Dobbs MB, Gurnett CA (2009) Update on clubfoot: etiology and
idiopathic club foot by the Ponseti method. J Bone Joint Surg Br treatment. Clin Orthop Relat Res 467:1146–1153
89:378–381 39. Bensahel H, Csukonyi Z, Desgrippes Y, Chaumien JP (1987)
21. Codivilla A (1905) On the means of lengthening, in the lower Surgery in residual clubfoot: one-stage medioposterior release ‘‘a
limbs, the muscles and tissues which are shortened through la carte’’. J Pediatr Orthop 7:145–148
deformity. Am J Orthop Surg 22(4):353–369

11
A modified Austin/chevron osteotomy for treatment of hallux
valgus and hallux rigidus
Michele Vasso1 • Chiara Del Regno1 • Antonio D’Amelio1 • Alfredo Schiavone Panni1
Abstract The purpose of this brief paper is to present the Keywords Austin/chevron Modified chevron
preliminary results of a modified Austin/chevron osteot- Osteotomy Hallux valgus Hallux rigidus
omy for treatment of hallux valgus and hallux rigidus. In
this procedure, the dorsal arm of the osteotomy is per-
formed orthogonal to the horizontal plane of the first Introduction
metatarsal, the main advantage being that this allows much
easier and more accurate multiplanar correction of first Numerous corrective osteotomies have been described for
metatarsal deformities. From 2010 to 2013, 184 consecu- surgical treatment of hallux valgus (HV), but none of them
tive patients with symptomatic hallux valgus and 48 addresses all cases. One of these procedures, known for its
patients with hallux rigidus without severe metatarsopha- optimal intrinsic mechanical stability, is the Austin/chevron
langeal joint degeneration underwent such modified chev- procedure [1, 2], a V-shaped distal osteotomy, traditionally
ron osteotomy. Mean patient age was 54.9 (range 21–70) indicated for correction of mild to moderate HV, in which the
years, and mean follow-up duration was 41.7 (range 24–56) hallux valgus angle (HVA) is less than 30° and the inter-
months. Ninety-three percent of patients were satisfied with metatarsal angle (IMA) is less than 15° [2]. Recent studies
the surgery. Mean American Orthopaedic Foot and Ankle have demonstrated that chevron osteotomy, associated with a
Society (AOFAS) score improved from 56.6 preoperatively distal soft-tissue procedure with or without the Akin proce-
to 90.6 at last follow-up, and mean visual analog scale dure, can increase the amount of correction, making this
(VAS) pain score decreased from 5.7 preoperatively to 1.6 combination appropriate for more severe deformities [3].
at final follow-up (p \ 0.05). In patients treated for hallux Many variants of the standard chevron have been proposed.
valgus, mean hallux valgus angle decreased from 34.1° All these variants require that the two arms of the osteotomy
preoperatively to 6.2° at final follow-up, and mean inter- be oblique to the horizontal plane of the first metatarsal.
metatarsal angle decreased from 18.5° preoperatively to However, this may generate technical difficulties and inac-
4.1° at final follow-up (p \ 0.05). One patient developed curacy in executing multiplanar osteotomies, regardless of
postoperative transfer metatarsalgia, treated successfully whether or not dedicated instrumentation is available.
with second-time percutaneous osteotomy of the minor A new modification of the chevron osteotomy was
metatarsals, whilst one patient had wound infection that therefore proposed, requiring that the dorsal arm of the
resolved with systemic antibiotics. osteotomy be performed orthogonal to the horizontal plane
Level of evidence Level IV. of the first metatarsal. The main advantage is that this
allows dorsal closing trapezoidal wedges to be performed
much more easily and accurately, especially for correcting
the proximal articular set angle (PASA), without compro-
& Chiara Del Regno mising the intrinsic stability of the traditional chevron.
[email protected]
Other advantages are the possibility of executing dorsal
1
Department of Medicine and Health Sciences, University of closing rectangular wedges, especially to obtain decom-
Molise, Via Francesco De Sanctis, 86100 Campobasso, Italy pression of the metatarsophalangeal joint (MTPJ) in cases

A modified Austin/chevron osteotomy for treatment of hallux valgus and hallux rigidus 75
of moderate hallux rigidus (HR) [4], and of fixing the final preserved. The soft-tissue rebalancing involves release of
osteotomy with a single screw. the adductor tendon from its insertion along the lateral base
of the proximal phalanx, release of the deep transverse
metatarsal ligament, and mobilization with recentering of
Materials and methods the sesamoids. The medial eminence is then removed.
For the modified chevron, the apex of the osteotomy
Surgical technique should remain in the center of the metatarsal head, posi-
tioned 5–10 mm proximal to the MTPJ line. As mentioned
A medial longitudinal incision is deepened to the capsule above, the dorsal arm of the osteotomy is performed per-
of the first MTPJ. The capsule is carefully dissected from pendicular to the horizontal plane of the first metatarsal
the head of the metatarsal and base of the proximal pha- (Fig. 1). The plantar oblique (long) arm of the osteotomy is
lanx, allowing adequate visualization of the joint. The cut just proximal to the capsular attachment to the meta-
dorsal and plantar vascular bundles are isolated and tarsal head fragment, since this carries part of the blood
supply (Fig. 2). The long plantar arm of the osteotomy
makes it possible to obtain greater correction of the IMA,
with the result that this technique is appropriate for more
severe deformities too. When the PASA has to be
Fig. 1 The dorsal arm of the osteotomy is performed orthogonal to

the horizontal plane of the first metatarsal (approximately parallel to Fig. 3 Performance of the dorsal cuts orthogonal to the horizontal
the MTPJ surface) plane of the first metatarsal allows one to easily obtain a precise
trapezoidal wedge for accurate PASA correction
Fig. 2 The plantar arm of the osteotomy is cut proximal to the

attachment of the joint capsule to preserve the plantar vascular bundle Fig. 4 This technique allows easy performance of a precise rectan-
directed to the metatarsal head fragment gular wedge according to the desired shortening of the first metatarsal

Fig. 5 Shortened first metatarsal in hallux rigidus for MTPJ decom- c

pression. In this case, a trapezoidal wedge was necessary for PASA
correction, without shifting the distal fragment laterally (therefore not
affecting the IMA)
corrected, a second osteotomy proximal to the dorsal cut is

performed, also orthogonal to the horizontal plane of the
first metatarsal; this osteotomy is obviously performed to
obtain a trapezoidal wedge with medial basis (Fig. 3). The
inclination of the second dorsal osteotomy relative to the
sagittal plane of the first metatarsal depends on the pre-
operative PASA value, and therefore on the amount of
correction that the surgeon desires to obtain. In the pres-
ence of HR without severe MTPJ degeneration, the second
dorsal osteotomy is performed parallel (if no correction of
the PASA is needed) and proximal to the dorsal cut, again
orthogonal to the horizontal plane of the first metatarsal.
The rectangular wedge is thus removed to decompress the
MTPJ (Fig. 4). The amount of bone removed depends on
the amount of shortening (and therefore decompression)
that the surgeon desires to obtain (Fig. 5).
Following the osteotomies, the head fragment is trans-
lated laterally (when the IMA needs to be reduced)
approximately one-quarter to one-half the width of the
metatarsal bone. In all patients, only one screw, from
medial to lateral, dorsal to plantar, and directed towards the
apex of the osteotomies, is used to stabilize the fragments.
MTPJ capsulorrhaphy is then performed (as part of soft-
tissue balancing) after deflating the tourniquet. In HV
patients, a percutaneous Akin osteotomy is always per-
formed. No device is used to stabilize the Akin osteotomy.
Postoperative care
A functional bandage is applied in the surgery room, with

the hallux in overcorrected position in case of HV treat-
ment. The first bandage is changed after 7 days, then every
2 weeks until its definitive removal, usually at 5 weeks
postoperatively. Full weight bearing is allowed immedi-
ately using a talus orthopedic shoe for 5 weeks, which is
generally sufficient time to obtain complete consolidation
of the osteotomies. Patients treated for HR are encouraged
to begin passive motion of the hallux in the immediate
postoperative period. No crutches or other devices are
generally necessary.
Methods
From 2010 to 2013, the modified chevron osteotomy was

performed in 184 patients with symptomatic HV and in 48
patients with HR without severe MTPJ degeneration.
Patients with posttraumatic HV, previous failed HV

A modified Austin/chevron osteotomy for treatment of hallux valgus and hallux rigidus 77
surgery or infection, Charcot arthropathy, or severe MTPJ Table 2 Detailed data of HR patients
degeneration were excluded from this cohort. There were Preoperatively Final follow-up p-Value
189 women and 43 men. Mean patient age was 54.9 (range
21–70) years, and mean follow-up duration was 41.7 (range Mean AOFAS 58.5 ± 6.3 91.9 ± 5.9 0.027
24–56) months. All surgeries were performed by the same Mean VAS 5.5 ± 1.0 1.4 ± 0.7 0.043
surgeon (M.V.). Patients were clinically assessed using the
American Orthopaedic Foot and Ankle Society (AOFAS) persistent pain; none of them requested reoperation. In
[5] hallux score, visual analog scale (VAS) pain score, and particular, the two patients with preoperative HV com-
patient self-reported subjective satisfaction. Patient sub- plained of mild pain with stiffness of the first MTPJ, while
jective satisfaction was assessed with respect to pain, the patient with preoperative HR developed mild transfer
function, and cosmetic appearance, and the responses were metatarsalgia under the second metatarsal head. There
graded as very satisfied, satisfied, improved, and dissatis- were no cases of delayed union or nonunion, metatarsal
fied. The HVA, IMA, and PASA were calculated radio- head necrosis, MTPJ stiffness (defined as range of move-
graphically before and after surgery in HV patients. The ment \30°), displacement after fixation, or complex
paired Student’s t test was used to compare the preopera- regional pain syndrome. In a patient with rheumatoid
tive and postoperative outcomes. Statistical significance arthritis, two screws were necessary to fix the osteotomy
was accepted at p \ 0.05. because of poor bone stock.
Results Discussion
At last follow-up, 161 (70 %) patients were very satisfied The success of Austin/chevron osteotomy for correction of
with the surgery, 54 (23 %) were satisfied, 14 (6 %) were HV is well established [6]. Although introduced for cor-
improved, and 3 (1 %) were dissatisfied. Mean AOFAS rection of mild to moderate deformities, many studies have
score improved from 56.6 (range 49–64) points preopera- recently demonstrated that chevron osteotomy, associated
tively to 90.6 (range 81–94) points at last follow-up with a soft-tissue procedure with or without the Akin
(p = 0.027), and mean VAS pain score decreased from 5.7 procedure, can allow correction of more severe deformities
(range 4–8) preoperatively to 1.6 (range 0–3) at final fol- [7].
low-up (p = 0.043). In HV patients, mean HVA decreased Multiplanar osteotomies of the first metatarsal are
from 34.1° (range 14–44°) preoperatively to 6.2° (range indicated to correct each component of the deformity, and
-2° to 15°) at final follow-up (p = 0.036), and mean IMA to avoid recurrence of the deformity itself or a noncon-
decreased from 18.5° (range 10–28°) preoperatively to 4.1° gruent joint leading to arthritis [8, 9]. However, even when
(range 1–12°) at final follow-up (p = 0.041). Mean PASA dedicated instrumentation is available, multiplanar osteo-
decreased from 16.3° (range 12°–18°) preoperatively to tomies can prove to be technically demanding and ulti-
6.5° (range 4–9°) at final follow-up (p = 0.039). Detailed mately inaccurate, with the result that oblique cut
data for HV patients and HR patients are presented in orientation may make articular fragment displacement and
Tables 1 and 2, respectively. repositioning quite difficult [9].
One patient developed severe postoperative transfer The presented technique was, instead, found to be
metatarsalgia, treated successfully with a second-time technically simple and highly reproducible, especially in
percutaneous osteotomy of the minor metatarsals, whilst the presence of multiplanar osteotomies, because the
one patient had wound infection that resolved with sys- geometry of the osteotomy makes corrections easy and
temic antibiotics. The three dissatisfied patients (two with precise, and allows considerable versatility. Performing the
preoperative HV and one with preoperative HR) claimed dorsal arm of the osteotomy orthogonal to the horizontal
plane of the first metatarsal makes it easier to obtain
accurate trapezoidal or rectangular bone wedges to correct
Table 1 Detailed data of HV patients the PASA and/or decompress the MTPJ. Compensation for
Preoperatively Final follow-up p-Value structural deformity in three planes is possible; in fact, the
procedure may be combined with an Akin osteotomy in
Mean AOFAS 54.7 ± 6.0 89.3 ± 6.2 0.027
cases of significant interphalangeal valgus and/or residual
Mean VAS 5.9 ± 1.1 1.8 ± 0.9 0.043
pronation deformity. Therefore, this modified chevron
Mean HVA 34.1 ± 4.1° 6.2 ± 1.9° 0.036
procedure is effective and appropriate for correction of
Mean IMA 18.5 ± 3.2° 4.1 ± 2.9° 0.041
mild to severe HV deformities. Finally, the shape of this
Mean PASA 16.3 ± 3.1° 6.5 ± 1.7° 0.039
osteotomy does not affect the intrinsic stability of the

traditional chevron, as sufficient stability is provided by References

impaction of the cancellous head fragment on the shaft.
Additionally, unlike traditional osteotomies, suggested in 1. Robinson AH, Limbers JP (2005) Modern concepts in the treat-
ment of hallux valgus. J Bone Joint Surg Br 87:1038–1045
cases needing MTPJ decompression for HR [9, 10], this
2. Fakoor M, Sarafan N, Mohammadhoseini P et al (2014) Com-
technique requires a single screw instead of two, as parison of clinical outcomes of scarf and chevron osteotomies
reported by other authors [11]. and the McBride procedure in the treatment of hallux valgus
The present modification of the standard Austin/chevron deformity. Arch Bone Jt Surg 2:31–36
3. Ferrao PN, Saragas NP (2014) Rotational and opening wedge
osteotomy allows easy and reproducible multiplanar cor-
basal osteotomies. Foot Ankle Clin 19:203–221
rection of HV deformities, with high patient satisfaction, 4. Giannini S, Ceccarelli F, Faldini C, Bevoni R, Grandi G, Vannini
even in the presence of severe deformities. This procedure F (2004) What’s new in surgical options for hallux rigidus?
is also chosen in cases of HR with mild arthritis to J Bone Joint Surg Am 86-A(Suppl 2):72–83
5. Kitaoka HB, Alexander IJ, Adelaar RS, Nunley JA, Myerson MS,
decompress the MTPJ. In all these cases, only one screw is
Sanders M (1994) Clinical rating systems for the ankle-hindfoot,
needed to fix the osteotomy. midfoot, hallux, lesser toes. Foot Ankle Int 15:349–353
6. Trnka HJ, Zembsch A, Weisauer H et al (1997) Modified Austin
Compliance with ethical standards procedure for correction of hallux valgus. Foot Ankle Int 18:119–127
7. Bai LB, Lee KB, Seo CY, Song EK, Yoon TR (2010) Distal
Ethical standards The authors declare that they have no conflict of chevron osteotomy with distal soft tissue procedure for moderate
interest. The study complies with the 1964 Helsinki Declaration and to severe hallux valgus deformity. Foot Ankle Int 31:683–688
its later amendments; the study was approved by the local Ethical 8. Nery C, Barroco R, Réssio C (2002) Biplanar chevron osteotomy.
Review Board; all patients provided informed consent before being Foot Ankle Int 23:792–798
enrolled. 9. Freeman BL, Hardy MA (2011) Multiplanar phalangeal and
metatarsal osteotomies for hallux rigidus. Clin Podiatr Med Surg
28:329–344
10. Polzer H, Polzer S, Brumann M, Mutschler W, Regauer M (2014)
Hallux rigidus: joint preserving alternatives to arthrodesis—a
review of the literature. World J Orthop 5:6–13
11. Dickerson JB, Green R, Green DR (2002) Long-term follow-up
of the Green-Watermann osteotomy for hallux limitus. J Am
Podiatr Med Assoc 92:543–554

12
The use of a dual mobility cup in the management of recurrent
dislocations of hip hemiarthroplasty
Christian Carulli1 • Armando Macera1 • Fabrizio Matassi1 • Roberto Civinini1 •
Massimo Innocenti1
Abstract may be strongly considered in cases of revisions of

Background Dislocation is one of the most frequent unstable hemiarthroplasties.
causes of failure of hemiarthroplasties of the hip, which is Level of evidence IV.
the most common treatment for femoral neck fractures in
elderly patients. A revision with conversion to total hip Keywords Dislocation Hemiarthroplasty of the hip
arthroplasty is the gold standard in case of failure of closed Dual mobility cups Revision
reduction: however, the use of standard or modular com-
ponents shows variable outcomes. The use of a dual
mobility cup has been evaluated in patients with unstable Introduction
implants, given the good outcomes obtained in primary and
revision surgery. The aim of this study was to assess the Dislocation is one of the major causes of failure of a
results of revisions by dual mobility cups in unstable hemiarthroplasty of the hip (HAH). Its incidence is rated at
hemiarthroplasties. 6–10 % with respect to 2–3 % for total hip arthroplasty
Materials and methods Thirty-one patients (mean age (THA) [1, 2]. Dislocations occur typically within 6 months
75.4 years) were retrospectively evaluated between 2006 after surgery [3], particularly in the first 2–6 weeks. Sev-
and 2010 after conversion to total hip arthroplasty with eral factors have been advocated, such as sex, cognitive
dual mobility cups for recurrent dislocations. The mean status, anatomy of the acetabulum (related to patients);
number of dislocations was 2.6 (range 2–5). The evaluation femoral head diameter, femoral stem rotation and off-set,
was performed by the American Society of Anesthesiolo- surgical approach and excessive removal of joint capsule
gists physical function score (ASA) and the Harris hip (related to surgeons) [4, 5]. It is crucial to understand the
score, and several radiologic criteria. causes of dislocation before facing surgery with an ade-
Results The mean follow-up was 3.8 years. No recurrence quate strategy, in order to limit the recurrence of the
of dislocation was recorded. The ASA score remained instability. Several procedures have been proposed
unchanged, and the mean Harris hip score improved from depending on the cause of the dislocation: repositioning of
62.2 before dislocation to 76.0 points postoperatively. femoral stem [6], conversion to THA [6, 7], revision with
Conclusions Dual mobility cups may be a useful option traditional or modular neck components [7–10], use of
in the treatment of a hemiarthroplasty dislocation. No risk constrained components [11, 12], trochanteric advance-
of a new revision due to instability after insertion of dual ment [13], removal of acetabular or femoral osteophytes
mobility cups resulted in our experience, and this option [6], and repair of the abductor muscles and of the joint
capsule [14, 15]. However, all these procedures showed
rates of success ranging from 60 to 80 %, independently by
& Christian Carulli the cause leading to instability [6, 10, 13, 16–19]. Partic-
[email protected]
ularly, the conversion of HAH to THA demonstrated dis-
1
Orthopaedic Clinic, University of Florence, couraging results with reports of even worse failure rates
Largo P. Palagi 1, 50139 Florence, Italy than a full revision [6, 7]. The implant of constrained

acetabular inserts also showed variable results, with a high radiographic study by anteroposterior and lateral views was
risk of increased wear, osteolysis, and instability in THA conducted to study the femoral stem position according to
[11, 12]. Revisions of unstable THAs are generally con- Loudon and Charnley [35], and the stability of the compo-
sidered technically demanding procedures [20–22]. nents as described by Engh et al. [36]. The presence of
Recently, good results have been obtained by the use of radiolucent lines and osteolysis of periprosthetic bone were
‘‘dual mobility’’ cups for revisions of unstable THAs [23– assessed by the criteria of DeLee and Charnley, and Gruen
31] and primary implants after femoral neck fractures [32], et al. [37, 38]. Cup inclination was assessed in the anterior–
in terms of limitation of dislocation recurrence and posterior projection, measuring in degrees the angle formed
preservation of a wide range of motion (ROM): low wear is by a line drawn along the bottom of the acetabular com-
also expected. To date, no report addresses similar out- ponent intersecting with the horizontal inter-teardrop line.
comes for the management of unstable HAHs treated by Hip centre restoration was assessed by calculating the per-
revisions with dual mobility cups. pendicular distance from the prosthetic centre of rotation to
The purpose of this study was to assess the short-term a horizontal line drawn between the tips of the teardrops.
results of a series of patients affected by unstable HAHs Limb length was evaluated. Finally, the presence of peri-
managed by a conversion to THA with dual mobility cups. articular ossification was also evaluated by Brooker’s
classification [39]. Collaborative patients, or relatives of
poorly oriented subjects were adequately informed, and
Materials and methods approved the treatment and follow-up. Surgery was per-
formed by two surgeons, in all cases by a direct lateral
We retrospectively reviewed 31 patients (31 hips) affected approach through the previous surgical scars. In 19 cases a
by recurrent dislocations of HAH, treated by a conversion to general anaesthesia was performed (ASA score: IV in six
THA with dual mobility cups between 2006 and 2010. All patients, III in 13); in 12 cases, a locoregional anaesthesia
patients had been given bipolar cemented implants for was chosen. In 25 cases, a capsular laxity was present, while
femoral neck fractures: the index operation was performed in the remaining patients the capsule was mostly absent.
with a mean interval of 2.4 days (range 1–3) after patient When possible, capsulae were sutured and soft tissues
admission to the emergency room. Eighteen patients were reconstructed after the cup positioning. In all cases a dual
female and 13 male, with a mean age of 75.4 years (range mobility acetabular cup was implanted as porous coated
71–86) at the time of fracture. The right side was affected in press-fit or cemented (AvantageÒ, Biomet, Warsaw, IN,
17 cases; the left side in 14 cases. Eleven patients were USA). This component consisted of a metal cup with a
operated on in other hospitals, while 20 were operated on at polished inner surface articulating with a high molecular
the authors’ institution. All patients were operated on by a weight polyethylene bipolar insert (acting as a large diam-
lateral approach at the time of HAH. The mean interval to eter head) containing a 28-mm chrome–cobalt head. In 20
the first dislocation after HAH was 23.2 days (range 1–46). cases, a press-fit cup was implanted (Fig. 1): three cups
The mean number of dislocations was 2.6 (range 2–5). needed a further fixation by two or three acetabular screws.
Dislocations were mostly posterior (29 cases); one subject In the remainder, a cemented cup was implanted (Fig. 2).
showed a dislocation in an anterior direction; only one case Criteria leading to the use of a cemented cup were poor bone
was multidirectional (a single patient with five episodes of quality or a significant enlargement of the native diameter of
instability).An evaluation of the associated risk factors of the acetabulum as tested intraoperatively during acetabular
patients was made before proceeding to revision. The mean preparation. Cups sizes between 44 and 56 mm were used.
time between the HAH and the revision in arthroplasty was Actually, in a single case we also proceeded to the revision
3.2 years (range 7 months–6 years). The American Society of the cemented femoral stem, given the remarkable rota-
of Anesthesiologists physical function (ASA) score based tional malposition of the component and the length dis-
on the severity of patients’ comorbidities was evaluated crepancy (2 cm): a new larger cemented femoral stem was
[33]. The ASA score at the time of revision was III in 19 used (MS-30Ò, Biomet, Warsaw, IN, USA). In 12 patients, a
patients, IV in six subjects, and II in the remainder. Several long (eight cases) or extra-long (four cases) 28-mm head
pathologies were present, and a high risk of dislocation was was implanted to ensure an adequate offset and further
considered in some patients: three cases of Parkinson&s dis- stability. The prophylaxis of heterotopic ossifications was
ease, three cases of diabetes mellitus with severe peripheral made by Indometacin 25 mg t.i.d. for 3 weeks in patients
neuropathy, one case of critical peripheral arterial disease, without any contraindications related to other comorbidities
two severe cognitive impairments related to Alzheimer&s or concomitant therapies. Parameters such as blood loss,
disease, one hemiparesis as the result of a previous stroke, following the criteria of Liu et al. [40], surgical time, and
and one of severe pluriarticular rheumatoid arthritis. The early postoperative complications were recorded. Postop-
Harris hip score (HHS) was also recorded [34]. A erative care consisted of a short period of immobilization

The use of a dual mobility cup in the management of recurrent dislocations of hip hemiarthroplasty 81
Fig. 1 A left femoral fracture of a 72-year-old male patient, treated by a hemiarthroplasty of the hip (a); 3 weeks postoperatively, a dislocation
of the implant occurred (b), and conversion to total hip replacement by a pressfit dual mobility cup was performed (c)
Fig. 2 A left femoral fracture of a 79-year-old female patient, closed reduction under anaesthesia (b). A second dislocation recurred
affected by Alzheimer’s disease, and treated by a hemiarthroplasty of after 5 days, thus a cemented dual mobility cup was implanted (c)
the hip (a); 4 days after surgery, a dislocation occurred, treated by
with a pillow between the legs in order to limit adduction of surgical time was 57.8 min (range 45–120). Seven patients
the hips. An assisted passive motion protocol from the 3rd were assisted after surgery in an intensive care unit for
postoperative day was then performed. Active exercises, 24–48 h. No intraoperative complication was recorded.
partial weight-bearing, and assisted gait activities were then Postoperative complications were present in six cases
specifically prescribed for each case, depending on pain and (19.3 %): three deep vein thromboses (one unilateral, one
patients’ collaboration. All patients were clinically and bilateral) managed by a mechanical compression and ther-
radiographically evaluated at 1 month after surgery, and apeutic doses of low-molecular-weight heparin; one case of
after 3, 6, and 12 months. After this follow-up, all the urinary tract infection, treated by antibiotics; one case of
subjects were encouraged to attend a yearly follow-up. superficial wound infection, managed by an advanced
Considering the small size of the study population, only wound care treatment and oral antibiotics; and one case of
the Wilcoxon signed rank test was used to compare pre- an acute imbalance in diabetes mellitus, managed by tai-
and postoperative HHS scores. lored insulin therapy.
No case of dislocation was recorded during the mentioned
follow-up. Radiographic studies revealed radiolucent lines in
Results zone 2 according to DeLee and Charnley in three patients (all
with cementless cups). However, these were not progressive
All patients were followed at least for 2 years, with a mean and were less than 2 mm in width: these cups were correctly
follow-up of 3.8 years (range 2–7 years). The average implanted. In three additional cases radiolucent lines of about
blood loss was 210 cc (range 100–400), and the mean 1 mm without progression around the femoral component

were found in zone 1 (the only patient with the stem revision) Figved et al. [20] reported a lower risk of complications,
and zone 5 (two patients) according to Gruen et al. The mean including instability, based on the Norwegian Arthroplasty
cup inclination was 45.4° (range 42–49°). An adequate hip Register, in cases of conversion of HAH to THA with stem
centre restoration was achieved in 23 cases. A suboptimal hip revisions, compared to stem retaining procedures. More-
centre was achieved in the remaining subjects; however, due over, in the same series, modular implants for revision
to good stability, the patients accepted well the residual presented more advantages related to head size, neck
length discrepancy (in all cases \1.5 cm). No osteolysis, length, and worn head replacement. However, no mention
significant subsidence, or cement mantle fractures were of dual mobility cups has been described.
noted, according to the criteria of Loudon and Charnley. No Only a few studies showed no relationships or even
implant was found to be unstable or poorly stable according to higher rates of dislocation between large diameter heads
Engh’s classification. We recorded three cases (9.6 %) of and the risk of instability in primary and revision implants
heterotopic ossifications grade 1 and one grade 2 (the patient [41, 42]. Llinas et al. [21] reported the long-term outcomes
with the revised stem), without, however, referred symptoms of a series of failed HAHs treated with THA with tradi-
or functional impairments: two of them did not undergo tional components: higher rates of earlier radiologically
prophylaxis due to clinical contraindications. detected loosening of acetabular components inserted fol-
The pillow was maintained for an average interval of lowing HAH failure were found with respect to primary
2.8 days (range 2–4). The mean HHS improved from 62.2 THAs. No mention of dual mobility cups was made in this
points (range 34–75) before the dislocation to 76.0 points series.
(range 71–80) postoperatively with a significant difference Constrained cups and liners have been proposed over the
(p = 0.002). The ASA score remained basically stable after years with variable results [11, 12]. Reduction of ROM
surgery in all the patients. Symptoms and functional dis- related to component impingement, increased wear related
ability progressively decreased over the follow-up period, to high local stresses, and higher risk of loosening were
allowing all patients without neurologic impairments to considered the reasons related to significant rates of failure
return to their common daily activities. Poorly or uncol- of these implants [23–25].
laborative patients were not substantially able to complete a Dual mobility cups and large femoral heads have their
full functional recovery, however, without further episodes rationale in limiting instability, ensuring a wide ROM with
of dislocation. respect to traditional implants, and maintaining low wear in
primary and revision hip arthroplasties. Satisfactory long-
term outcomes have been reported in several series in
Discussion primary and revision hip arthroplasty [23–31, 45]. A single
multicentre study reported the use of this type of implant
Dislocations of HAHs are generally associated with an for the primary replacement in patients affected by a
insufficient restoration of the centre of rotation or other femoral fracture: a dislocation occurred in three cases out
mechanical problems due to a wrong primary implantation. of 214 patients (1.4 %) within the first 3 months [46]. The
The conversion of an unstable HAH to a standard THA is a authors found no recurrence of the dislocation in these
procedure with a high risk of further dislocations, with an patients treated by closed reduction under general anaes-
incidence often higher than revision THA itself [2, 20–22, thesia, even if they used a posterior approach, generally
41, 42]. Several reasons have been advocated: the reduc- associated with a higher risk of dislocation with respect to
tion of the diameter and offset of the femoral head, which the direct lateral approach [47, 48]. However, to date there
may produce an inadequate soft tissues tension; the inap- has been no significant experience regarding series of
propriate positioning of a retained femoral stem, frequently HAHs failed for instability and managed by revision with
maintained to avoid long surgical procedures in critical dual mobility cups. Bouchet et al. reported a statistically
patients; and the insufficient retaining properties of the lower risk of dislocation for the dual mobility cup com-
acetabular cup/liner complex. Several other options such as pared to a conventional 28-mm head and polyethylene
the use of a cemented cup with a structural bone graft fixed inserts implanted through a posterior approach. The insta-
with screws, threaded cups with or without bone grafting, bility rate was 0 % compared with 4.63 % for the con-
constrained cups, reinforcement rings, or ‘‘anti-protrusio’’ ventional prostheses [25]. In our series, we recorded
cages have been proposed over the decades. Variable improvements in the HHS, and complication rates were
results have been obtained in cases of acetabular discon- comparable to other reports in the literature. Nonetheless,
tinuity or severe bone loss, poor acetabular rim coverage, we had no recurrence of dislocation, and no specific failure
and substantial alterations of shape of the acetabulum [43, related to choice of implants. A specific mechanism of
44]. In the remaining cases, outcomes were not failure of dual mobility cups is effectively represented by
satisfactory. the intraprosthetic dislocation [49–51]. It consists of the

The use of a dual mobility cup in the management of recurrent dislocations of hip hemiarthroplasty 83
loss of the polyethylene retentive rim, with escape of the 9. McGann WA, Welch RB (2001) Treatment of the unstable total
femoral head from the liner that may manifest particularly hip arthroplasty using modularity, soft tissue, and allograft
reconstruction. J Arthroplast 16(Suppl. 1):19–23
in younger, high-demand patients undergoing a primary 10. Toomey SD, Hopper RH Jr, McAuley JP, Engh CA (2001)
THA with this implant [28, 51]. No similar complication Modular component exchange for treatment of recurrent dislo-
was recorded in our series. cation of a total hip replacement in selected patients. J Bone Joint
The present study has some limitations. It is a retro- Surg Am 83:1529–1533
11. Shapiro GS, Weiland DE, Markel DC, Padgett DE, Sculco TP,
spective analysis with a small number of patients, and Pellicci PM (2003) The use of a constrained acetabular compo-
without a control group. However, we do not usually per- nent for recurrent dislocation. J Arthroplast 18:250–258
form revisions with standard or constrained cups for 12. Yun AG, Padgett D, Pellicci P, Dorr LD (2005) Constrained
unstable HAHs, using in most cases a dual mobility com- acetabular liners: mechanisms of failure. J Arthroplasty
20:536–541
ponent: related costs are similar to other choices of treat- 13. Ekelund A (2003) Trochanteric osteotomy for recurrent disloca-
ments. Nevertheless, at short-term follow-up we had no tion of total hip arthroplasty. J Arthroplasty 8:629–632
recurrence of instability, with both versions (cemented and 14. Hughes AW, Clark D, Carlino W, Gosling O, Spencer RF (2015)
cementless) of the dual mobility cup. Capsule repair may reduce dislocation following hip hemi-
arthroplasty through a direct lateral approach: a cadaver study.
We feel that dual mobility cups may be a useful and Bone Joint J 97-B:141–144
effective option worth considering in the treatment of HAH 15. Weber M, Berry DJ (1997) Abductor avulsion after primary total
dislocations. hip arthroplasty: results of repair. J Arthroplast 12:202–206
16. Ali Khan MA, Brakenbury PH, Reynolds I (1981) Dislocation
Compliance with Ethical Standards following total hip replacement. J Bone Joint Surg Br
63:214–218
Ethical standards The authors state that the study conforms to the 17. Lind M, Krarup N, Petersen LG, Mikkelsen S, Horlyck E (2002)
1964 Helsinki declaration and its later amendments; the study was Acetabular revision for recurrent dislocations: results in 14 cases
approved by the local or institutional Ethical Review Board; all the after 3 years of follow-up. Acta Orthop Scand 73:291–294
patients provided informed consent before being enrolled. 18. Morrey BF (2004) Results of reoperation for hip dislocation: the
big picture. Clin Orthop Relat Res 429:94–101
19. Robbins GM, Masri BA, Garbuz DS, Greidanus N, Duncan CP
(2001) Treatment of hip instability. Orthop Clin North Am
32:593–610
20. Figved W, Dybvik E, Frihagen F, Furnes O, Madsen JE, Havelin
LI, Nordsletten L (2007) Conversion from failed hemiarthro-
plasty to total hip arthroplasty: a Norwegian Arthroplasty
Register analysis of 595 hips with previous femoral neck frac-
tures. Acta Orthopaedica 78:711–718
21. Llinas A, Sarmiento A, Ebramzadeh E, Gogan WJ, McKellop HA
(1991) Total hip replacement after failed hemiarthroplasty or
mould arthroplasty. Comparison of results with those of primary
References replacements. J Bone Joint Surg Br 73:902–907
22. Diwanji SR, Kim SK, Seon JK, Park SJ, Yoon TR (2008) Clinical
1. Iorio R, Healy WL, Lemos DW, Appleby D, Lucchesi CA, Saleh results of conversion total hip arthroplasty after failed bipolar
KJ (2001) Displaced femoral neck fractures in the elderly: out- hemiarthroplasty. J Arthroplast 23:1009–1015
comes and cost effectiveness. Clin Orthop Relat Res 23. Guyen O, Chen QS, Bejui-Hugues J, Berry DJ, An KN (2007)
383:229–242 Unconstrained tripolar hip implants: effect on hip stability. Clin
2. Sah AP, Estok DM II (2008) Dislocation rate after conversion Orthop Relat Res 455:202–208
from hip hemiarthroplasty to total hip arthroplasty. J Bone Joint 24. Philippot R, Camilleri JP, Boyer B, Adam P, Farizon F (2009)
Surg Am 90:506–516 The use of a dual articulation acetabular cup system to prevent
3. Blewitt N, Mortimore S (1992) Outcome of dislocation after dislocation after primary total hip arthroplasty: analysis of 384
hemiarthroplasty for fractured neck of the femur. Injury cases at a mean follow-up of 15 years. Int Orthop 33:927–932
23:320–322 25. Bouchet R, Mercier N, Saragaglia D (2011) Posterior approach
4. Meek RMD, Allan DB, McPhillips G, Kerr L, Howie CR, From and dislocation rate: a 213 total hip replacements case-control
the Scottish Arthroplasty Project (2008) Late dislocation after study comparing the dual mobility cup with a conventional
total hip arthroplasty. Clin Med Res 6:17–23 28-mm metal head/polyethylene prosthesis. Orthop Traumatol
5. Holmes JD, House AO (2000) Psychiatric illness predicts poor Surg Res 97:2–7
outcome after surgery for hip fracture: a prospective cohort study. 26. Vielpeau C, Lebel B, Ardouin L, Burdin G, Lautridou C (2011)
Psycol Med 30:921–929 The dual mobility socket concept: experience with 668 cases. Int
6. Daly PJ, Morrey BF (2002) Operative correction of an unstable Orthop 35:225–230
total hip arthroplasty. J Bone Joint Surg Am 74:1334–1343 27. Mohammed R, Hayward K, Mulay S, Bindi F, Wallace M (2015)
7. Berry DJ (2001) Unstable total hip arthroplasty: detailed over- Outcomes of dual-mobility acetabular cup for instability in pri-
view. Instr Course Lect 50:265–274 mary and revision total hip arthroplasty. J Orthop Traumatol
8. Earll MD, Fehring TK, Griffin WL, Mason JB, McCoy T, Odum 16:9–13
S (2002) Success rate of modular component exchange for the 28. Langlais FL, Ropars M, Gaucher F, Musset T, Chaix O (2008)
treatment of an unstable total hip arthroplasty. J Arthroplasty Dual mobility cemented cups have low dislocation rates in THA
17:864–869 revisions. Clin Orthop Relat Res 466:389–395

29. Grazioli A, Ek ET, Rudiger HA (2012) Biomechanical concept 42. Tarasevicius S, Kesteris U, Robertsson O, Wingstrand H (2006)
and clinical outcome of dual mobility cups. Int Orthop Femoral head diameter affects the revision rate in total hip
36:2411–2418 arthroplasty: an analysis of 1720 hip replacements with
30. Hailer NP, Weiss RJ, Stark A, Karrholm J (2012) Dual-mobility 9–21 years of follow-up. Acta Orthop 77:706–709
cups for revision due to instability are associated with a low rate 43. Chatrath V, Beaulé PE (2013) The changing paradigm of revision
of re-revisions due to dislocation. Acta Orthop 83:566–571 of total hip replacement in the presence of osteolysis. Instr Course
31. Civinini R, Carulli C, Matassi F, Nistri L, Innocenti M (2012) A Lect 62:215–227
dual mobility cup is effective for reducing dislocations in isolated 44. Paprosky WG, Perona PG, Lawrence JM (1994) Acetabular
acetabular revisions. Clin Orthop Relat Res 470:3542–3548 defect classification and surgical reconstruction in revision
32. Tarasevicius S, Robertsson O, Dobozinskas P, Wingstrand H arthroplasty. A 6-year follow-up evaluation. J Arthroplast
(2013) A comparison of outcomes and dislocation rates using 9:33–44
dual articulation cups and THA for intracapsular femoral neck 45. Cross MB (2015) Dual mobility acetabular components for
fractures. Hip Int 23:22–26 revision THA. J Orthop Traumatol 16:21–22
33. Dripps RD, Lamont A, Eckenhoff JE (1961) The role of anes- 46. Adam P, Philippe R, Ehlinger M, Roche O, Bonnomet F, Molé D,
thesia in surgical mortality. JAMA 178:261–266 Fessy MH (2012) Dual mobility cups hip arthroplasty as a
34. Harris WH (1969) Traumatic arthritis of the hip after dislocation treatment for displaced fracture of the femoral neck in the elderly.
and acetabular fractures: treatment by mold arthroplasty. An end- A prospective, systematic, multicenter study with specific focus
result study using a new method of result evaluation. J Bone Joint on postoperative dislocation. Orthop Traumatol Surg Res
Surg Am 51:737–755 98:296–300
35. Loudon JR, Charnley J (1980) Subsidence of the femoral pros- 47. Masonis JL, Bourne RB (2002) Surgical approach, abductor
thesis in total hip replacement in relation to the design of the function, and total hip arthroplasty dislocation. Clin Orthop Relat
stem. J Bone Joint Surg Br 62:450–453 Res 405:46–53
36. Engh CA, Massin P, Suthers KE (1990) Roentgenographic 48. Demos HA, Rorabeck CH, Bourne RB, MacDonald SJ, McCal-
assessment of the biologic fixation of porous-surfaced femoral den RW (2001) Instability in primary total hip arthroplasty with
components. Clin Orthop Relat Res 257:107–128 the direct lateral approach. Clin Orthop Relat Res 393:168–180
37. DeLee JG, Charnley J (1976) Radiological demarcation of 49. Lecuire F, Bernareau I, Rubini J, Basso M (2004) Intra-prosthetic
cemented sockets in total hip replacement. Clin Orthop Relat Res dislocation of Bousquet dual mobility socket. Rev Chir Orthop
121:20–32 Reparatrice App Mot 90:249–255
38. Gruen TA, McNeice GM, Amstutz HC (1979) Modes of failure 50. Banzhof JA, Robbins CE, van der Ven A, Talmo CT, Bono JV
of cemented stem-type femoral components: a radiographic (2012) Femoral head dislodgement complicating use of a dual
analysis of loosening. Clin Orthop Relat Res 141:17–27 mobility prosthesis for recurrent instability. J Arthroplast
39. Brooker AF, Bowerman JW, Robinson RA, Riley LH Jr (1973) 28:543.e1–543.e3
Ectopic ossification following total hip replacement: incidence and 51. Philippot R, Boyer B, Farizon F (2013) Intraprosthetic disloca-
a method of classification. J Bone Joint Surg Am 55:1629–1632 tion: a specific complication of the dual-mobility system. Clin
40. Liu X, Zhang X, Chen Y, Wang Q, Jiang Y, Zeng B (2011) Hidden Orthop Relat Res 471:965–970
blood loss in hip arthroplasty. J Arthroplast 26:1100–1105
41. Lachiewicz PF, Heckman DS, Soileau ES, Mangla J, Martell JM
(2009) Femoral head size and wear of highly cross-linked poly-
ethylene at 5 to 8 years. Clin Orthop Relat Res 467:3290–3296

13
Biocompatibility of magnesium implants in primary human
reaming debris-derived cells stem cells in vitro
Olga Charyeva1,2 • Olga Dakischew2 • Ursula Sommer2 • Christian Heiss3 •
Reinhard Schnettler3 • Katrin Susanne Lips2
Abstract magnesium alloys did not affect the ALP content. Exposure
Background Use of magnesium for resorbable metal of HRD to magnesium increased the amount of lysosomes
implants is a new concept in orthopaedic and dental med- and endocytotic vesicles. Cellular attachment was gener-
icine. The majority of studies on magnesium’s biocom- ally the best for those crystals that formed on the surface of
patibility in vitro have assessed the short-term effect of all materials. A decrease was observed in Ca2? in the
magnesium extract on cells. The aim of this study was to medium from day 1 to day 14.
evaluate the influence of direct exposure to magnesium Conclusions In terms of cell morphology, cell viability
alloys on the bioactivity of primary human reaming debris- and differentiation, cell density and the effect on the sur-
derived (HRD) cells. rounding pH, Mg2Ag showed the most promising results.
Materials and methods Pure Mg, Mg2Ag, WE43 and All magnesium materials induced calcification, which is
Mg10Gd were tested for biocompatibility. The study con- beneficial for orthopaedic and dental applications.
sisted of assessment of cell viability by 3-(4,5-
dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide Keywords Biocompatibility Magnesium Human
(MTT) test, evaluation of alkaline phosphatase (ALP) reaming debris-derived cells
content, and study of cell morphology under light micro-
scopy, scanning electron microscopy (SEM) and trans-
mission electron microscopy (TEM), along with Introduction
determination of calcification and pH changes induced by
magnesium. Temporary metal implants that would resorb after bone
Results The number of viable cells in the presence of healing has completed would minimise secondary surgery
Mg2Ag was high over the entire observation period. Inhi- for implant removal and thus decrease postoperative
bition of ALP content in osteogenic differentiating HRD infections [1–3]. This in turn would decrease high costs
was caused by pure Mg at day 14 and 28. All other related to repeated surgery, reduce recovery periods and
thus promote higher quality of life for patients.
A perfect resorbable metal implant should provide
& Olga Charyeva enough strength to the healing bone; it should resorb after a
[email protected] set time period, preferably no earlier than 12 weeks [4],
1
and should be non-toxic and cause no damage to the body.
aap Biomaterials GmbH, Lagerstrasse 11-15, 64807 Dieburg,
Germany
Magnesium is a suitable material for biodegradable metal
2
implants because it is biocompatible [5] and does not cause
Laboratory for Experimental Trauma Surgery, Justus-Liebig
toxicological tissue response. Magnesium occurs naturally
University Giessen, Schubertstrasse 81, 35392 Giessen,
Germany in humans since our organism contains about 25 g of this
3 element, about 50–60 % of which is found in bone [6].
Department of Trauma Surgery, University Hospital of
Giessen-Marburg, Rudolf-Buchheim-Str. 9, 35385 Giessen, Magnesium has been shown to stimulate bone formation,
Germany since its ions enhance cell attachment and proliferation [7].

In previous studies, new bone was seen forming in direct The materials were cast at Helmholtz-Zentrum Geesthacht
contact with magnesium [8, 9]. Finally, magnesium’s Magnesium Innovation Center (HZG-MagIC).
mechanical properties are closer to those of healthy bone The three magnesium alloys (Mg2Ag, Mg10Gd and
than titanium’s [10]. WE43) were produced by permanent mould gravity cast-
However, the problem with magnesium is the formation ing. After melting pure Mg, the melt was held at 720 °C
of hydrogen gas on contact with fluids [4]. In water, and the preheated alloying elements were added with
magnesium hydroxide accumulates on the surface of the continuous stirring for 15 min. The melt was poured into a
magnesium implant to form a corrosion layer, also known preheated (550 °C) permanent steel mould treated with
as a degradation layer. While this film slows corrosion boron nitride. During the casting process, cover gas was
under aqueous conditions, it reacts with chlorine ions used (SF6 and Ar mixture). The alloys were homogenised
present in blood to produce highly soluble MgCl2 and with a T4 heat treatment prior to extrusion in Ar atmo-
hydrogen gas [4]. Hydrogen gas is undesirable as it inter- sphere at 550 °C (Mg10Gd and WE43) or 420 °C
feres with normal tissue healing and affects primary (Mg2Ag) for 6 h. Afterwards, the alloys were extruded
implant stability in bone. indirectly with extrusion ratio of 4:25. The chamber of the
At the same time, it has been shown that magnesium extrusion machine was set to 370 °C, and the 30-mm-di-
facilitates calcification and formation of calcium phos- ameter billets were preheated for 1 h at 370 °C (Mg2Ag),
phates [11]. Calcium is needed by the body to ensure that 390 °C (WE43) or 430 °C (Mg10Gd). The extrusion speed
bone laid down by osteoblasts is normally mineralised [12]. was between 3 and 4.5 mm/s. Pure Mg was cast by per-
It has been shown in previous studies that magnesium manent mould direct chill casting.27 The cast billet
increases the pH and that high pH promotes Ca2? binding (d = 110 mm) was extruded indirectly with an extrusion
[13, 14]. This is especially important for implants that are ratio of 1:84. The billet temperature was 340 °C, and the
to be used for bone fractures. speed of the extrusion was 0.7 mm/s. Discs (10 mm
The majority of studies on biocompatibility of magne- diameter, 1.5 mm thickness) were machined from the
sium in vitro have assessed the short-term effects of extruded bars.
magnesium extract on cells [15–17]. The aim of this study
was to evaluate the influence of magnesium alloys on the
Sample sterilisation and pre-incubation
bioactivity of primary human reaming debris-derived
(HRD) cells with up to 21 days of direct exposure. This, in
The samples were sonicated for 20 min in dry isopropanol,
our opinion, will mimic in vivo conditions more closely.
dried and gamma-sterilised at the BBF Sterilisation Service
GmbH facility (Kernen, Germany) with total dose of
29 kGy. Before seeding the cell culture, the samples were
pre-incubated to form a protective degradation layer. Per
0.2 g of sample, 3 mL of F12 K (GibcoÒ, Life Technolo-
Sample production
gies, USA) was used. The samples were kept at 37 °C in
5 % CO2 atmosphere for 12 h.
The following materials were used to produce alloys for
this study: magnesium (99.99 %, Xinxiang Jiuli Magne-
sium Co. Ltd., China), yttrium (99.95 % Y, Grirem Isolation of human reaming debris-derived (HRD)
Advanced Materials Co. Ltd., China), gadolinium cells
(99.95 % Gd, Grirem Advanced Materials Co. Ltd.,
China), rare-earth mixture (Grirem Advanced Materials HRD cells were cultured from various patients, with the
Co. Ltd., China) and silver (99.99 % Ag, ESG Edelmetall- approval of the local Ethics Commission, as described by
Handel GmbH & Co. KG, Germany). Wenisch et al. [18]. The adult patients were of different
Three magnesium-based materials were produced: genders and ages and did not display any disease related to
Mg2Ag (1.89 % Ag, the rest was Mg), Mg10Gd (8.4 % bone metabolism. In total, cells from six different patients
Gd, the rest was Mg) and WE43 (3.45 % Y, 2.03 % Nd, were taken for this study.
0.84 % Ce, the rest was Mg). Pure magnesium (99.99 % The reaming debris was cultured in Petri dishes with
Mg) was used as control. The concentrations of magnesium F12 K medium including 20 % foetal calf serum (FCS),
(Mg), Y, neodymium (Nd) and cerium (Ce) were deter- 100 U/ml penicillin and 100 lg/g streptomycin. After
mined by spark emission spectrometer (Spectrolab M, 4–7 days, HRD cells started to grow out of the debris.
Spektro, Germany), and the concentrations of Ag and Gd When the cells reached confluence after 2–3 weeks, they
were determined by X-ray fluorescence spectrometer were trypsinised and transferred to cell culture flasks. All
(Bruker AXS S4 Explorer, Bruker AXS GmbH, Germany). cells were kept at 37 °C in 5 % CO2 atmosphere.

Biocompatibility of magnesium implants in primary human reaming debris-derived cells stem cells in vitro 87
Cell viability PicoGreenÒ assay were centrifuged, and the supernatants

were diluted in a specific assay buffer. ALP substrate was
To determine cell viability, an MTT assay was conducted applied to the diluted samples, and the absorbance was
according to Mosmann et al. [19]. Briefly, 10,000 cells/ measured at 405 nm. The absolute amounts of ALP were
cm2 were seeded into 12-well plates containing pre-incu- correlated with the cell numbers obtained from the
bated magnesium discs and F12 K medium with 20 % FCS PicoGreenÒ assay. Both ALP and PicoGreenÒ assays were
and 100 lg/g streptomycin. Controls with no magnesium additionally performed for magnesium discs not seeded
discs but only HRDs were used as well. The cell medium with cell culture to exclude the material’s effect on the test
was changed every second day during the experiment. and see only how the cells reacted during the assays.
Duplicates were used for each material and patient,
resulting in testing of 12 wells per specimen. After 24 h, 7 Transmission electron microscopy
and 21 days, MTT solution was added to cell medium. The
cells were then incubated in the dark for 4 h at 37 °C. HRD cells seeded in chamber slides (Nalge Nunc Inter-
Subsequently, the cell medium was discarded, and the cells national, Rochester, NY) were incubated with and without
were lysed with 0.004 N HCl in isopropanol. The cell magnesium discs for 21 days. The cell layer was fixed for
lysates were centrifuged, and supernatants were transferred 30 min with 2 % paraformaldehyde in 0.1 M sodium
as triplets to a 96-well plate. Adsorption was measured at phosphate buffer (pH 7.2–7.4) with 2 % glutaraldehyde
570 and 630 nm using a Synergy HT microplate reader and 0.02 % picric acid, followed by 20 min fixation with
(BioTek, Bad Friedrichshall, Germany). The MTT assay 1 % osmium tetroxide in 0.1 M sodium cacodylate buffer
was also performed for magnesium discs not seeded with (pH 7.2–7.4). The samples were dehydrated and embedded
cell culture to exclude the material’s effect on the test and in Epon before ultrathin sections (80–100 nm) were
see only how the cells reacted during the assay. applied to collodion-coated copper grids. Analysis was
Additionally, cell morphology was studied by inverted done with a LEO 912 transmission electron microscope
light microscopy using a Leica microscope type (Carl Zeiss AG, Oberkochen, Germany) at 80 kV accel-
090-135.002 (Leica Microsystems GmbH, Wetzlar, Ger- erating voltage, equipped with a TRS SharpEye slow-scan
many) equipped with a Nikon Ds-Fi1 digital camera dual-speed charge-coupled device (CCD) camera (Albert
(Nikon, Düsseldorf, Germany). Troendle Prototypentwicklung, Moorenweis, Germany).
Alkaline phosphatase (ALP) content Scanning electron microscopy
As an indicator of changes in the differentiation behaviour HRD cells were cultivated on magnesium discs for 7 and
of the bone-forming cells caused by the test substances, a 21 days. Controls with no magnesium discs but only HRDs
SensoLyteÒ pNPP alkaline phosphatase assay (AnaSpec, were used as well. Subsequently, the cells were fixed in
Fremont, CA) was applied after 24 h and 7, 14, 21 and 2 % glutaraldehyde in 0.1 M Na-phosphate buffer (pH 7.4)
28 days of culturing in Dulbecco’s modified Eagle’s for 1 h at room temperature, followed by dehydration in a
medium (DMEM), low glucose with L-glutamine, 10 % graded series of ethanol and critical-point drying. The
FCS, 100 U/ml penicillin, 100 lg/g streptomycin, 0.1 lM specimens were mounted together on aluminium pin stubs
dexamethasone, 0.005 lM ascorbic acid and 10 mM b- with the help of adhesive carbon pads. The specimens were
glycerol phosphate to induce osteogenic differentiation. then sputter-coated with gold/palladium (SC7640 sputter
The cell medium was changed every second day during the coater, VG Microtech, Uckfield, East Sussex, UK) and
experiment. Duplicates were used for each material and assessed using a LEO 1530 (LEO Elektronenmikroskopie
patient, for a total of 12 wells per specimen. Controls with GmbH, Oberkochen, Germany) field-emission scanning
no magnesium discs but only HRDs were used as well. electron microscope operated at 7.5 or 15 kV.
The cells were washed and frozen at -80 °C. After
thawing, the cell number was measured using a Determination of Ca21 consumption and pH
PicoGreenÒ dsDNA quantitation assay (Invitrogen,
Eugene, OR) according to the manufacturer’s protocol. At established time points, medium was collected and
Cells were lysed with 1 % Triton X-100 in phosphate- analysed for Ca2? in solution and pH. The concentration of
buffered saline. The cell lysates were centrifuged, and the Ca2? was measured using a calcium analyser (9180 elec-
supernatants were mixed with PicoGreenÒ working solu- trolyte analyzer, Roche, Mannheim, Germany), and pH
tion in a 96-well plate. The samples were excited at measurements were carried out using a pH-meter (Titan X,
485 nm, and the fluorescence emission intensity was Fisher Scientific GmbH, Schwerte, Germany) for each time
measured at 528 nm. The cells that were lysed for the point. The control group for this investigation consisted of

Table 1 MTT results for HRDs after exposure to different magne-

sium materials over time
Material Viability Viability Viability (% of
(% of control) (% of control) control) ± SD
± SD (%), day 1 ± SD (%), day 7 (%), day 21
Pure Mg 93.4 ± 25.3 13.9 ± 5.0 24.0 ± 19.5

WE43 57.1 ± 14.4 30.5 ± 9.2 26.3 ± 3.6
Mg10Gd 37.0 ± 13.1 18.1 ± 7.1 54.7 ± 7.2
Mg2Ag 113.4 ± 29.8 63.3 ± 11.0 98.5 ± 12.0
well plates that contained only HRDs and medium but no

magnesium.
Data were analysed using the Statistical Package for the

Social Sciences (SPSS, v18, SPSS Inc., Chicago, USA).
The significance level was set at 5 %. The treatment groups
were compared with the control group without magnesium
by Mann–Whitney U-test. The results are expressed as
means. The graphs were plotted with SPSS.
Results
Cell viability
Cell viability values are presented in Table 1. It was

observed that Mg2Ag had the highest cell viability of all
materials during the whole observation period. Cells with
pure Mg had high viability at day 1 (93.4 ± 25.3 %), but
then the viability decreased and reached 24.0 ± 19.5 % at
day 21. The same pattern was observed for WE43 with
initial viability of 57.1 ± 14.4 %, reaching 26.3 ± 3.6 %
at day 21. Mg10Gd had the lowest cell viability of all
tested materials over the entire observation period.
Alkaline phosphatase content
ALP content is an important factor in bone mineral for-

mation and shows a scale of changes during differentiation.
No inhibition of ALP activity caused by Mg2Ag, Mg10Gd
and WE43 was observed in osteogenic differentiating HRD
at days 14 and 28 (Fig. 1). At day 1, the ALP content was
significantly higher for Mg2Ag (p = 0.004) and WE43
(p = 0.003) compared with control. Significantly low
values for ALP content compared with the control group
were observed for pure Mg at day 14 (p = 0.002) and
day 28 (p = 0.004). Fig. 1 ALP content at day 1, 14 and 28

Fig. 2 Morphology of HRD at 7 days: a control group, the well is in appearance to pure Mg and Mg10Gd with very few cells observed
densely covered with cells; b pure Mg, many fewer cells compared in material’s proximity; f for pure Mg, Mg10Gd and WE43, the cells
with control; c Mg2Ag, the well is densely covered with cells; were found mainly at the well edge (the image shows Mg10Gd’s well
d Mg10Gd, many fewer cells compared with control; e WE43, similar edge). Asterisk magnesium disc, arrows products of degradation
Cell morphology regarding cell number (Fig. 4c). More cells appeared
around Mg10Gd and WE43 at day 28 compared with other
Changes in cell morphology were detected by inverted time points for these materials (Fig. 4d, e).
light microscopy for HRD and for osteogenic differentiat-
ing HRD. After 7 days of exposure to pure Mg, Mg10Gd
Transmission electron microscopy
and WE43, HRD showed a reduction in cell number and an
increasing amount of cell debris in the medium (Fig. 2).
Intracellular structure was examined after exposure of
Few or no cells were seen in direct contact with pure Mg,
HRD to magnesium samples for 21 days. It was observed
Mg10Gd and WE43. Instead, the cells were found on the
that the number of lysosomes and endocytotic vesicles was
edge of the wells (Fig. 2f). The reduction in cell number
higher in the HRD exposed to magnesium alloys than in the
was more apparent for pure Mg than in any other group.
control (Fig. 5). In Mg2Ag, degraded material particles
The cell morphology in the presence of Mg2Ag was similar
were found inside the lysosomes (Fig. 5e) and in the
to control, and the cells were directly contacting the
cytoplasm (Fig. 5f). Degradation particles were not
Mg2Ag discs (Fig. 2c).
observed in other groups.
At 21 days, the HRD in WE43 and Mg10Gd started to
appear closer to the disc, although their number was still
low compared with control or Mg2Ag (Fig. 3). For pure Scanning electron microscopy
Mg, cells were still only found around the edge of the well
and not in proximity to the material. For Mg2Ag, the cell Cellular attachment to magnesium specimens was studied
morphology was most similar to the control with high cell under SEM after incubating HRD with magnesium for 7
density directly contacting the discs. and 21 days. It was observed that the cells attached to the
The osteogenic differentiating HRD showed a similar degradation layer and to the crystals forming on the surface
pattern of cell morphology and cell number as HRD of magnesium (Fig. 6). Cell pseudopodia were numerous
(Fig. 4). No cells were found in direct contact with pure whenever crystals formed on material surfaces (Fig. 6b, c,
Mg (Fig. 4b). Mg2Ag was the most similar to control at all e). Few pseudopodia were seen on smoother surfaces
of the time points regarding morphology and cell density. (Fig. 6d). No difference between 7 and 21 days was
At day 28, the osteogenic differentiating HRD in the observed regarding the number of attached cells or their
Mg2Ag group were still most similar to the control morphology.

Fig. 3 Morphology of HRD at 21 days: a control group, the well is e WE43, similar appearance to Mg10Gd with somewhat more cells
densely covered with cells; b pure Mg, almost no cells compared with than at day 7; f crystal formation was observed for all materials (the
control; c Mg2Ag, the well is densely covered with cells; d Mg10Gd, image shows the pure Mg well). Asterisk magnesium disc. Scale bar
many fewer cells compared with control but more than at day 7; 100 lm
Ca21 consumption Discussion
Calcification induced by magnesium specimens was stud- This study looked at the direct long-term effect of mag-
ied by measuring Ca2? consumption from the surrounding nesium alloys on primary HRDs. Cell viability, differen-
medium. A decrease in free calcium ions was observed in tiation and morphology as well as pH and calcium uptake
the medium around all magnesium materials from day 1 to were analysed to assess the overall biocompatibility of the
day 14. Between day 14 and day 21, Ca2? was released tested materials. We evaluated the long-term effects of
into the medium. Between day 21 and 28, Ca2? levels were magnesium on human cells to simulate the in vivo situation
stable for pure Mg, WE43 and Mg10Gd but decreased for as closely as possible.
Mg2Ag (Fig. 7a). Ca2? consumption levels were stable in Mg2Ag had high cell viability from day 1
the control over the whole study period. The values for (113.4 ± 29.8 %) to day 21 (98.5 ± 12.0 %). It was also
Mg2Ag were most similar to those of the control out of all shown in previous works [15] that Mg-Ag alloys have
groups starting on day 21. negligible cytotoxicity and sound cytocompatibility. Pure
Mg had high viability at the very first day (93.4 ± 25.3 %),
but then the viability decreased to 24.0 ± 19.5 % at
pH measurements day 21. Mg10Gd and WE43 impaired cell viability in this
study. Previous studies have shown higher values for cell
It was seen that all magnesium materials increased the pH viability measured by MTT test compared with the present
of the medium compared with the control group. The fol- study [15–17]. The difference between this work and pre-
lowing general pattern was observed for all groups: the pH vious publications is that the present study applied the
values were stable up to day 7, a sudden pH drop occurred longest in vitro incubation times for magnesium alloys
on day 14, then the pH tended to increase slightly up to tested up to now. The HRDs were kept in direct contact
day 28. The pH values for Mg2Ag were most similar to the with the magnesium samples and not in magnesium extract,
control. Pure Mg caused the greatest increase in pH among as done in most studies [15–17].
all groups, and this increase was statistically significant An important drawback of tetrazolium-based tests is that
compared with Mg2Ag (p B 0.003) but not compared with the difference between cytotoxic (cell death) and cytostatic
Mg10Gd (p C 0.02) or WE43 (p C 0.02). No correlation (reduced growth rate) effects cannot be distinguished [20].
between pH and Ca2? consumption was found in this We thus looked at cell morphology under light microscopy,
study. TEM and SEM.

Fig. 4 Morphology of osteogenic differentiating HRD at 28 days: covered with cells; d Mg10Gd, the well is densely covered with cells;
a control group, the well is densely covered with cells; b pure Mg, e WE43, the well is densely covered with cells. Asterisk magnesium
few cells compared with control; c Mg2Ag, the well is densely disc, arrows products of degradation. Scale bar 100 lm
After examination under SEM and light microscopy, it Interestingly, the crystal distribution was not homogeneous
was revealed that the number of cells decreased in the throughout the corrosion layer. In this sense, our results are
presence of pure Mg, Mg10Gd and WE43. These materials similar to earlier findings [11, 15].
seem to have long-term cytotoxic effects on HRD when The fact that the cells attached to the crystalline struc-
placed in direct contact with the cells. This explains the tures more readily than to the overall material surface and
low viability values. developed numerous pseudopodia can be explained by the
The cell number was high and the cells had normal rough structure of crystals, and by the chemical composi-
morphology in Mg2Ag groups. However, the cell viability tion of these crystals. It was previously shown that cells
was lowest at day 7 (63.3 ± 11.0 %) for Mg2Ag. TEM attach better to certain surfaces with preferable average
analysis revealed an elevated amount of lysosomes which surface roughness of *0.5 lm up to *8.5 lm [23].
contained degraded magnesium particles. Degradation Values below or above this range diminish the cells’ ability
particles were also found in the cytoplasm. The presence of to bind to the surface.
high amounts of degradation products inside the HRDs The chemical composition of the crystals and the
could explain the lower cell viability values for Mg2Ag at degradation layer formed on the magnesium’s surface can
day 7. It was shown in previous studies that uptake of also explain the better attachment of the cells to these
material particles leads to induction of cell stress which structures. Their chemical composition consists of calcium,
triggers cytotoxicity [21]. phosphorus, magnesium and oxygen [14]. Thus, the cells
ALP activity in HRDs is an important factor in bone attach to already reacted material where they are not
mineral formation and shows a range of changes during mechanically disturbed by hydrogen gas produced as a by-
differentiation. Inhibition of ALP activity in osteogenic product of degradation. The formation of the degradation
differentiating HRD was caused by pure Mg at day 14 and layer could also explain the increase in cell density around
28. All other magnesium alloys did not affect the ALP Mg10Gd and WE43 after 21 days of incubation.
activity. In this respect, our study shows similar results to Based on previous findings, the following model for the
previous research in this area [22], despite the fact that we formation of a corrosion layer has been suggested [14]: (1)
observed osteogenic differentiation over much longer initial metal corrosion based on contact with water mole-
periods and using direct contact of cells with magnesium. cules leads to release of Mg ions, and a thin Mg(OH)2 and
SEM analysis revealed that the cellular attachment was MgCO3 layer is formed; (2) The corrosion slows down, and
generally best to crystals generated by degradation prod- a second layer consisting of amino acids and organic matter
ucts on the material surface. Crystals have been seen is formed; (3) Both layers together shield the sensitive
forming on magnesium alloys such as Mg-Ag in previous environment around the material and enable cells to grow
studies [15]. Formation of calcium phosphates [Cax(PO4)x] on the material [14]. Such a complex process of corrosion
was also observed in previous publications [11]. layer formation could explain the cell viability increase at

Fig. 5 TEM analysis at day 21: a control, b Mg2Ag, c Mg10Gd, Mg2Ag (note the degradation particles, arrows). Asterisk lysosomes,
d WE43, e lysosome of HRD cultured with Mg2Ag (note the endocytotic vesicles, n nucleus. Note the high amount of lysosomes
degradation particles, arrows); f cytoplasm of HRD cultured with and endocytotic vesicles in b–d
day 21 for Mg2Ag, Pure Mg and Mg10Gd observed in our promotes formation of calcium phosphates and conse-
experiments. quently decreases the amount of free Ca2? ions in the
All magnesium-based materials decreased the amount of medium [11, 14]. In this regard, our results are consistent
Ca in this study. As shown in previous studies, Mg2?
2?
with earlier works. Sufficient supply of calcium is vital to

Fig. 6 SEM analysis at day 7: a control, densely covered with HRDs the crystal compared with the smooth surface in d. Cells are marked
(scale bar 100 lm); b pure Mg, c Mg2Ag, d Mg10Gd, e WE43. Note with an asterisk. Scale bar in b–e 5 lm
crystal formation on material surface in b–e, and cell attachment to
ensure that bone laid down by osteoblasts is normally However, no statistical correlation was observed between
mineralised [12]. Calcification is thus advantageous for pH and consumed Ca2?.
bone implants that are to be used in the orthopaedic and In conclusion, our study reveals the long-term effects of
maxillofacial fields. magnesium materials on human HRDs seeded directly onto
It was shown in previous studies that magnesium magnesium discs. In respect to cell morphology, cell den-
increases the pH and that high pH promotes Ca2? binding sity and the effect on the surrounding pH, Mg2Ag showed
[13, 14]. In this study, it was also revealed that pH shifts to the most promising results. However, the mechanism of
alkaline values in the presence of magnesium, but to cell stress induction and cytotoxicity needs to be further
somewhat different degrees for the different alloys. studied to enable prediction of possible health risks.

standards of the 1964 Declaration of Helsinki as revised in 2000. All

patients gave informed consent prior to being included in the study.
The patients gave informed consent for publication of the case study.
References
1. Quinn A, Hill AD, Humphreys H (2009) Evolving issues in the

prevention of surgical site infections. Surgeon 7(3):170–172
2. Hachenberg T, Sentürk M, Jannasch O, Lippert H (2010) Post-
operative Wundinfektionen. Pathophysiologie, Risikofaktoren,
und Praventive Konzepte. [Postoperative wound infections.
Pathophysiology, risk factors and preventive concepts]. Anaes-
thesist 9(59):851–866 (in German)
3. Jannasch O, Lippert H (2011) Perioperative Prophylaxe und
Therapie von Infektionen-Postoperative Wundinfektionen. [Sur-
gical site infections]. Anasthesiol Intensivmed Notfallmed Sch-
merzther 46(10):664–673 (in German)
4. Staiger MP, Pietak AM, Huadmai J, Dias G (2006) Magnesium
and its alloys as orthopedic biomaterials: a review. Biomaterials
27:1728–1734
5. Sandler G, Nguyen L, Lam L, Manglick MP, Soundappan SS,
Holland AJ (2011) Trampoline trauma in children: is it pre-
ventable? Pediatr Emerg Care 27(11):1052–1056
6. Goyer RA, Clarkson TW (2001) Toxicity of metals. In: Klaassen
CD (ed) Casarett and Doull’s Toxicology: The Basic Science of
Poisons. McGraw-Hill, USA, pp 811–868
7. Institute of Medicine Standing Committee on the Scientific
Evaluation of Dietary Reference Intakes (1997) Dietary Refer-
ence Intakes for Calcium, Phosphorus, Magnesium, Vitamin D,
and Fluoride. National Academy Press, Washington DC
8. Li Z, Gu X, Lou S, Zheng Y (2008) The development of binary
Mg–Ca alloys for use as biodegradable materials within bone.
Biomaterials 29(10):1329–1344
9. Xu L, Yu G, Zhang E, Pan F, Yang K (2007) In vivo corrosion
behavior of Mg–Mn–Zn alloy for bone implant application.
J Biomed Mater Res A 83(3):703–711
10. Witte F, Kaese V, Haferkamp H, Switzer E, Meyer-Lindenberg
A, Wirth CJ, Windhagen H (2005) In vivo corrosion of four
Fig. 7 a Consumed ionised calcium at different observation points. magnesium alloys and the associated bone response. Biomaterials
b pH over time 26(17):3557–3563
11. Feyerabend F, Drücker H, Laipple D, Vogt C, Stekker M, Hort N,
Willumeit R (2012) Ion release from magnesium materials in
Acknowledgments This project receives funding from the People physiological solutions under different oxygen tensions. J Mater
Programme (Marie Curie Actions) of the European Union’s Seventh Sci Mater Med 23(1):9–24
Framework Programme FP7 (2007–2013) under REA grant agree- 12. Reid IR (2014) Should we prescribe calcium supplements for
ment no. 289163. Special thanks are due to Iris Schütz, Justus-Liebig osteoporosis prevention? J Bone Metab 21(1):21–28
University Giessen, Laboratory for Experimental Trauma Surgery for 13. Witte F, Hort N, Vogt C, Cohen S, Kainer KU, Willumeit R,
technical assistance. Feyerabend F (2008) Degradable biomaterials based on magne-
sium corrosion. Curr Opin Solid State Mater Sci 12:63–72
Compilance with Ethical Standards 14. Willumeit R, Fischer J, Feyerabend F, Hort N, Bismayer U,
Heidrich S, Mihailova B (2011) Chemical surface alteration of
Conflict of interest The authors declare that they have no conflict biodegradable magnesium exposed to corrosion media. Acta
of interest related to the publication of this manuscript. Biomater 7(6):2704–2715
15. Tie D, Feyerabend F, Müller WD, Schade R, Liefeith K, Kainer
Ethical standards The study was authorized by the local ethical KU, Willumeit R (2013) Antibacterial biodegradable Mg-Ag
committee and was performed in accordance with the ethical alloys. Eur Cell Mater 16(25):284–298

16. Feyerabend F, Fischer J, Holtz J, Witte F, Willumeit R, Drücker 21. Pauksch L, Hartmann S, Rohnke M, Szalay G, Alt V, Schnettler
H, Vogt C, Hort N (2010) Evaluation of short-term effects of rare R, Lips KS (2014) Biocompatibility of silver nanoparticles and
earth and other elements used in magnesium alloys on primary silver ions in primary human mesenchymal stem cells and
cells and cell lines. Acta Biomater 6(5):1834–1842 osteoblasts. Acta Biomater 10(1):439–449
17. Yang L, Hort N, Laipple D, Höche D, Huang Y, Kainer KU, 22. Li RW, Kirkland NT, Truong J, Wang J, Smith PN, Birbilis N,
Willumeit R, Feyerabend F (2013) Element distribution in the Nisbet (2014) The influence of biodegradable magnesium alloys
corrosion layer and cytotoxicity of alloy Mg-10Dy during in vitro on the osteogenic differentiation of human mesenchymal stem
biodegradation. Acta Biomater 9(10):8475–8487 cells. J Biomed Mater Res A 14(12):4346–4357. doi:10.1002/
18. Wenisch S, Trinkaus K, Hild A, Hose D, Herde K, Heiss C, jbm.a.35111 (Epub ahead of print)
Kilian O, Alt V, Schnettler R (2005) Human reaming debris: a 23. Stout KJ, Sullivan PJ, Dong WP, Mainsah E, Luo N, Mathia T,
source of multipotent stem cells. Bone 36(1):74–83 Zahouani H. Development of Methods for Characterisation of
19. Mosmann T (1983) Rapid colorimetric assay for cellular growth Roughness in Three Dimensions. European Community Contract
and survival: application to proliferation and cytotoxicity assays. No 3374/1/0/170/90/2 Birmingham: Univ of Birmingham, 1993
J Immunol Methods 65(1–2):55–63
20. Plumb JA (2004) Cell sensitivity assays: the MTT assay. Methods
Mol Med 88:165–169

14
Correlating preoperative imaging with intraoperative fluoroscopy
in iliosacral screw placement
Jonathan G. Eastman1 • Milton L. Chip Routt Jr.2
Abstract (8°–31°) and an average outlet of 43.2° (31°–56°). The

Background Percutaneous iliosacral screw placement can average difference from preoperative to intraoperative was
successfully stabilize unstable posterior pelvic ring inju- 4.4° (-21° to 5°) for the inlet and 0.45° (-9° to 7°) for the
ries. Intraoperative fluoroscopic imaging is a vital com- outlet. The average difference between the preoperative
ponent needed in safely placing iliosacral screws. and postoperative CT was 2.04° (0°–6°) for the inlet and
Obtaining and appropriately interpreting fluoroscopic 2.54° (0°–7°) for the outlet.
views can be challenging in certain clinical scenarios. We Conclusion There is significant anatomic variation of the
report on a series of patients to demonstrate how preop- posterior pelvic ring. The preoperative CT sagittal recon-
erative computed tomography (CT) imaging can be used to struction images allow for appropriate preoperative plan-
anticipate the appropriate intraoperative inlet and outlet ning for anticipated intraoperative fluoroscopic inlet and
fluoroscopic views. outlet views within 5°. Having knowledge of the desired
Materials and methods 24 patients were retrospectively intraoperative views preoperatively prepares the surgeon,
identified with unstable pelvic ring injuries requiring aids in efficiently obtaining correct intraoperative views,
operative fixation using percutaneous iliosacral screws. and ultimately assists in safe iliosacral screw placement.
Utilizing the sagittal reconstructions of the preoperative CT Level of evidence IV, Retrospective case series.
scans, anticipated inlet and outlet angle measurements
were calculated. The operative reports were reviewed to Keywords Iliosacral screw Preoperative planning
determine the angles used intraoperatively. Postoperative
CT scans were reviewed for repeat measurements and to
determine the location and safety of each screw. Introduction
Results Preoperative CT scans showed an average inlet of
20.5° (7°–37°) and an average outlet of 42.8° (30°–59°). Percutaneous iliosacral screw fixation of unstable posterior
Intraoperative views showed an average inlet of 24.9° pelvic ring injuries has become a common successful
(12°–38°) and an average outlet of 42.4° (29°–52°). Post- treatment method [1–4]. In order to place iliosacral screws
operative CT scans showed an average inlet of 19.4° safely, a thorough understanding of the possible osseous
fixation pathways is paramount [5, 6]. Recognizing sacral
dysmorphism and accommodating anatomic variations of
the posterior pelvic ring requires detailed knowledge of the
& Jonathan G. Eastman osteology [7–11]. In addition to obtaining an accurate
[email protected] reduction, combining the osteological details with the
1 corresponding intraoperative fluoroscopic imaging is nec-
Department of Orthopaedic Surgery, University of California,
Davis Medical Center, 4860 Y Street, Suite 3800, essary to safely perform percutaneous fixation. Inlet and
Sacramento, CA 95817, USA outlet fluoroscopic views are utilized to safely place ilio-
2
Department of Orthopaedic Surgery, University of Texas, sacral screws. An intraoperative lateral fluoroscopic view
Health Sciences Center at Houston, Houston, TX, USA can be extremely helpful by providing a third dimension

Correlating preoperative imaging with intraoperative fluoroscopy in iliosacral screw placement 97
that helps verify the osteology seen on the inlet and outlet from preoperative CT scans are the same angles utilized
views [1–3, 12, 13]. with intraoperative fluoroscopy. We hypothesize that pre-
The varying degrees of sacral kyphosis or lordosis as operative CT imaging can successfully be used to accu-
well as the presence of any degree of sacral dysmorphism rately plan and anticipate the exact inlet and outlet angles
leads to a wide range of angles required to achieve actually used intraoperatively during percutaneous iliosa-
appropriate inlet and outlet radiographs as well as intra- cral screw fixation of unstable posterior pelvic ring injuries.
operative fluoroscopic views [14–16] (Fig. 1). In addition
to the details of the fracture, the preoperative CT scan can
be used to measure the ideal inlet and outlet angles. These Materials and methods
measurements can be taken to the operating theater to help
obtain the appropriate fluoroscopic views. This process can After obtaining Institutional Review Board approval, a
help surgeons quickly obtain satisfactory intraoperative 3-month review was performed from our prospectively
imaging and in attaining adequate imaging for all patients. collected trauma database at a regional level 1 trauma
This could be very helpful in difficult clinical situations center. This database has recorded all operatively managed
including morbid obesity, bowel gas, and the presence of fractures since 1989. Fractures are entered and coded
contrast. The purpose of this study was to determine according to the Arbeitsgemeinschaft für Osteosynthese-
whether the anticipated inlet and outlet angles obtained fragen/Orthopaedic Trauma Association (AO/OTA)
Fig. 1 Preoperative CT scan with sagittal reconstruction. The image pelvic ring is demonstrated above in four different patients. Note the
has been rotated 90° from vertical to simulate the patient lying supine near vertical orientation of the S1 sacral body in a. There is a gradual
on the operating room table. The yellow line parallels the anterior increase in the lordotic alignment in b and the S1 body is nearly
cortex of the S1 body with which the fluoroscopic beam would horizontal in d (color figure online)
parallel for an inlet view. The anatomic variability of the posterior

Fracture Classification System by orthopaedic trauma fel- associated fixation strategy. Each patient’s chart was
lows trained in this classification system [17]. Data is reviewed for patient gender, age, mechanism of injury, and
stored and manipulated using a commercially available AO/OTA injury classification.
software program (Microsoft Access). Inclusion criteria The preoperative CT scan of each patient was reviewed
required skeletal maturity, a complete medical chart rela- using a picture archiving and communication system
tive to their injury, adequate preoperative and postopera- (PACS) using Centricity Version 2.1 (GE Medical Sys-
tive radiographic imaging including CT scans, and to have tems, Waukesha, WI, USA). Utilizing the midline view of
had their definitive surgical procedure performed at our the sagittal reconstruction images, inlet and outlet angle
center. From 29 May 2012 to 31 July 2012, 24 consecutive were calculated. One surgeon made both the preoperative
patients with unstable pelvic ring injuries who underwent and postoperative CT measurements. The process to make
operative fixation using percutaneous iliosacral screws the measurement included rotating the entire image 90°
were identified. Pelvic ring injuries and associated insta- clockwise to simulate a supine position on the operating
bility was identified by preoperative radiographic and CT table. A horizontal line parallels that surface of the CT
imaging which demonstrated combinations of anterior and gantry. This line simulates how the patient will be lying
posterior pelvic ring disruptions. Anterior ring injuries supine on the operating table. A line is placed at 90° to the
consisted of either unilateral or bilateral superior and horizontal line that simulates a straight up and down
inferior rami fractures or complete symphysis pubis disposition of the C-arm that would produce an anteroposte-
ruptions. Posterior pelvic ring injuries consisted of a rior (AP) view. The inlet view angle is measured as a line
complete sacral fracture, sacroiliac joint disruption, or a that parallels the anterior cortex of the S1 body in reference
posterior ilium/sacroiliac joint fracture dislocation. All to the horizontal line. The anticipated inlet angle would be
patients underwent an examination under anesthesia as the difference in angles from the straight up and down
previously described [18]. These examinations demon- position of the C-arm down to the angle measured to obtain
strated and documented the instability present, especially an image that parallels the anterior cortex of the S1 body
in the 61-B injury patterns. All patients had documented (Fig. 2). For the outlet view, the same horizontal and 90°
posterior pelvic ring instability and therefore underwent lines are drawn as noted above. The outlet angle for an S1
appropriate operative fixation with accompanying percu- iliosacral screw is drawn as the line that overlaps the
taneous fixation to stabilize the posterior aspect of their
pelvic ring injury. One orthopaedic traumatologist at a
regional level 1 trauma center treated all patients. All
iliosacral screws were placed using standard and previously
described techniques with inlet, outlet, and lateral fluoro-
scopic imaging only utilizing a C-arm [2, 19, 20]. Once
adequately resuscitated and evaluated, each patient was
sedated and transported to the operating suite where they
surrendered to general anesthesia. The patient was then
transferred onto a radiolucent operating table and placed in
a supine position. The patient was placed onto two folded
blankets beneath the lumbosacral spine. The blanket bump
is precisely placed with the distal aspect of the blankets at
the testicles or labia and in the center of the lumbosacral
spine. The perineum was cleansed and isolated from the
operative field with adhesive drapes. The entire abdomen
and bilateral flanks were then sequentially cleansed with
iodine and isopropyl alcohol. Accurate reductions of the
pelvic ring injuries were achieved by both open and closed
means as guided by the injury patterns and surrounding soft
tissue status. All screws placed in this cohort were with the
patient in a supine position. Posterior ring fixation con- Fig. 2 Preoperative CT scan with sagittal reconstruction at the
sisted of 7.0-mm diameter cannulated screws (Synthes, midline demonstrating the anticipated inlet measurement. The
Paoli, PA, USA) or 7.0-mm diameter cannulated screws horizontal line parallels the surface the patient is lying on.
The oblique line parallels the anterior surface of the S1 body. The
(Zimmer, Warsaw, IN, USA) of varying length; both fully
line at 90° simulates a straight up and down position of the C-arm.
and partially threaded screws were used depending on The anticipated inlet angle, labeled a, would therefore be 23°
injury pattern, available osseous fixation pathways, and (113°-90°) of cephalad tilt of the C-arm

symphysis over the center of the S2 body. The anticipated the spinal canal. Extraosseous was defined as any evidence
outlet angle would then be the difference between the of cortical discontinuity. Postoperative rehabilitation and
straight up and down position down to the angle measured mobilization were guided by each patient’s musculoskele-
to obtain an image placing the superior symphysis over- tal injuries and overall medical condition under the direct
lying the center of the S2 body (Fig. 3) [16, 19]. supervision of licensed physical therapists using standard
The operative reports of each patient were subsequently protocols. Statistical analysis was performed using paired
reviewed to determine the inlet and outlet angles used T-tests for comparison of the preoperative CT measure-
intraoperatively. The difference between the preoperative ments to the intraoperative fluoroscopic measurements as
anticipated inlet and outlet angles and intraoperative fluo- well as the Pearson product-moment correlation coefficient
roscopic inlet and outlet angles was determined. The flu- for assessment of intraobserver reliability in obtaining the
oroscopic angle was defined as the center value and the preoperative and postoperative CT measurements.
preoperative value was either less than (negative number)
or greater than (positive number) that value. A postopera-
tive CT was obtained on each patient within 24 h from Results
surgery. This is standard treatment protocol and verifies
reduction and implant placement. Each postoperative CT The cohort consisted of 24 patients (14 males and 10
was reviewed to repeat the inlet and outlet angle mea- females) with an average age of 47.7 years (20–82). The
surements without direct knowledge of the previously mechanisms of injury included eight patients with falls,
measured angles. The preoperative and postoperative five patients involved in motor vehicle collisions, four
measurements were made by a single surgeon to assess patients sustaining equestrian injuries, three patients
whether the method of measurement was reproducible. The involved in motorcycle collisions, three patients involved
location and safety of each iliosacral screw was determined in automobile versus pedestrian accidents, and one patient
and each screw was defined as intraosseous, juxtaforminal, who sustained a crush injury. AO/OTA classification
or extraosseous. An intraosseous position was defined by showed five 61-B injury patterns—two 61-B1.1, two B2.1,
the presence of cancellous bone completely surrounding and one B3.2. There were 19 61-C injury patterns—three
the screw on all CT cuts. Juxtaforaminal was defined by a 61-C1.2, seven 61-C1.3, two 61-C2.3, one 61-C3.1, two
lack of cancellous bone surrounding the screw but an intact 61-C3.2, and four 61-C3.3. Two patients also sustained
cortical rim at the ala, S1 or S2 neuroforaminal tunnel, and accompanying acetabular fractures. One patient sustained
an open pelvic ring injury with complete symphyseal dis-
ruption and complete sacral fracture medial to the neuro-
foraminal tunnels. The open wound included his scrotum
and perineum and was managed with multiple irrigation
and debridements, closure of his scrotal wound, and
packing to closure of his perineal wound. His posterior
pelvic ring underwent closed reduction and percutaneous
fixation and his anterior ring injury was treated with
external fixation for 6 weeks. The average time until sur-
gery was 4.4 days (1–28). Twenty-two patients were
managed with closed reduction. Two patients required an
open reduction of their displaced sacroiliac joint disloca-
tions through an anterior approach. Of 24 patients, 9
(37.5 %) had some degree of sacral dysmorphism as pre-
viously defined [7, 11].
Utilizing the measurement method described above,
preoperative CT scans showed an average inlet view of
20.5° (7°–37°) and an average outlet view of 42.8° (30°–
59°). The intraoperative fluoroscopic views showed an
Fig. 3 Preoperative CT scan with sagittal reconstruction at the average inlet of 24.9° (12°–38°) and an average outlet view
midline demonstrating the anticipated outlet measurement. The hor- of 42.4° (29°–52°). Postoperative CT scans showed an
izontal line parallels the surface the patient is lying on. The oblique average inlet of 19.4° (8°–31°) and an average outlet of
line overlaps the superior aspect of the symphysis pubis to the S2
43.2° (31°–56°), (Table 1). The average difference
body. The line at 90° simulates a straight up and down position of the
C-arm. The anticipated outlet angle, labeled a, would therefore be 52° between the preoperative to intraoperative inlet view was
(90°-38°) of caudal tilt of the C-arm 4.4° (-21° to 5°), which was statistically significant with

Table 1 Inlet and outlet angles obtained using preoperative and incredibly important and being able to correctly interpret
postoperative sagittal CT imaging as well as the fluoroscopic angles the radiographic landmarks of the pelvis and their rela-
used intraoperatively for all 24 patients
tionship to anatomical structures is mandatory [7, 24, 25].
View Minimal Maximal Arc Average In addition to the fracture pattern and displacement, pre-
operative CT scans demonstrate the osteology of each
Preoperative CT inlet 7 37 30 20.5
patient. Utilizing the preoperative CT, anticipated inlet and
Preoperative CT outlet 30 59 29 42.8
outlet angles can be measured and brought to the operating
Fluoroscopic inlet 12 38 26 24.9
room to help the surgeon and radiology technician obtain
Fluoroscopic outlet 29 52 33 42.2
accurate intraoperative imaging. In our series of 24
Postoperative CT inlet 8 31 23 19.4
patients, the preoperatively measured inlet and outlet
Postoperative CT outlet 31 56 25 43.2
angles were within 5° and 1°, respectively, of the corre-
The minimal and maximal values are listed in addition to the corresponding intraoperative fluoroscopic angles.
sponding angular arcs and averages The anatomic variability of the pelvis has been well
documented [14–16]. Figure 1 exemplify this variation and
p = 0.0003. The average difference between the preoper- how the radiographic or fluoroscopic beam would need to
ative to intraoperative outlet view was 0.45° (-9° to 7°), be angled differently in each patient to obtain an ideal inlet
which was not statistically different. The average differ- view that parallels the anterior cortex of the S1 body.
ence when comparing the preoperative and postoperative Traditionally, inlet and outlet radiographs were obtained by
CT scans was 2.04° (0°–6°) for the inlet and 2.54° (0°–7°) directing the beam 45° caudally and 45° cranially from the
for the outlet view (Table 2). In comparing the preopera- direct AP view [15, 26, 27]. This definition has evolved
tive and postoperative CT inlet and outlet angle measure- over time and several studies have since shown that the
ments, the Pearson product-moment correlation coefficient angles required to obtain inlet and outlet views differ
reached a correlation of [0.9. The inlet and outlet mea- greatly from this [15, 16]. Utilizing a similar measurement
surement angles for each patient for the preoperative CT method as in our series, Graves et al. showed an ideal
scan, fluoroscopic views, and postoperative CT scan are intraoperative inlet fluoroscopic view of 25° (21°–33°) and
listed in Table 3. A total of 58 iliosacral screws were an ideal intraoperative outlet fluoroscopic view of 42°
placed in this patient cohort. Postoperative CT imaging (30°–50°) [16]. Similarly in our series, the average ideal
demonstrated that 51 of 58 screws (88 %) were intraoss- intraoperative inlet fluoroscopic view averaged 24.9° (12°–
eous and 7 of 58 screws (12 %) were juxtaforaminal. There 38°) and an average intraoperative outlet view to S1 of
were no extraosseous screws. Of the 58 screws, 31 (53 %) 42.4° (29°–52°). Standardized views do not account for the
were iliosacral-style screws and 27 (47 %) were considered wide variability of the posterior pelvic ring. While erro-
transiliac transsacral screws. Of the 31 iliosacral screws, 4 neous placement of screws despite apparent appropriate
(12.9 %) were noted to be juxtaforaminal and of the 27 screw positioning on intraoperative fluoroscopy has been
transiliac transsacral screws, 3 (11 %) were noted to be documented, unintentionally utilizing incorrect imaging
juxtaforaminal. could lead to implant malpositioning and unintended
iatrogenic injury to neurovascular structures [28]. While
increasing the technical demands of iliosacral screw
Discussion placement, attaining patient-specific non-orthogonal
imaging leads to a more precise identification of the pos-
While some surgeons advocate relying on computer-as- terior pelvic ring anatomy [16]. The preoperative CT scan
sisted and navigated systems, it is common to utilize inlet, allows for the measurement of each patient’s individual
outlet, and lateral fluoroscopic images to safely instrument posterior pelvic ring alignment. This preoperative mea-
the posterior pelvic ring [1, 2, 7, 12, 21–23]. Obtaining surement can be taken to the operating room and assist in
quality intraoperative fluoroscopic images remains obtaining accurate intraoperative fluoroscopic views.
Table 2 Comparison of preoperative CT inlet and outlet measurements with the fluoroscopic angles used intraoperatively
View Minimal difference Maximal difference Range of values Average
Preoperative CT inlet compared to fluoroscopic inlet -21 5 26 4.4

Preoperative CT outlet compared to fluoroscopic outlet -9 7 16 0.45
Postoperative CT inlet compared to preoperative inlet 0 6 6 2.04
Postoperative CT outlet compared to preoperative CT inlet 0 7 7 2.54

Table 3 Summary of the measured inlet and outlet angles for the preoperative CT, intraoperative fluoroscopic views, and postoperative CT scan
for each patient
Patient Preop CT inlet Fluoro inlet Postop CT inlet Preop CT outlet Fluoro outlet Postop CT outlet
1 13 22 15 47 40 40
2 25 30 23 41 50 45
3 19 20 22 47 45 47
4 37 32 31 44 42 44
5 23 20 23 39 40 42
6 9 30 8 44 43 44
7 18 26 19 59 52 56
8 19 21 20 45 40 44
9 7 12 8 46 40 42
10 22 30 25 38 42 40
11 10 19 8 46 41 47
12 27 31 25 42 42 39
13 32 28 28 47 50 52
14 21 21 18 30 29 33
15 17 20 16 45 40 40
16 24 30 23 43 51 44
17 12 18 9 43 42 41
18 21 27 20 41 45 46
19 10 15 8 56 51 55
20 22 30 21 42 39 44
21 21 27 21 41 40 43
22 28 29 25 30 35 34
23 32 38 31 42 40 44
24 24 23 19 30 38 31
Average 20.5 24.9 19.4 42.8 42.4 43.2
Range 7–37 12–38 8–31 30–59 29–52 31–56
The average and range of values is listed for each measurement
Correlating accurate intraoperative imaging with an accu- screws. Wolinsky et al. demonstrated that by rotating the
rate reduction, surgeon tactile feedback, and detailed C-arm[8° towards the foot away from the ideal inlet view,
knowledge of the available osseous fixation pathways will an out-the-back wire can appear to be contained within the
ultimately lead to safe implant positioning. bony sacrum [29]. Unknowingly relying on imperfect
As recently demonstrated by Miller et al., the excessive views could lead to the placement of unsafe iliosacral
fat density associated with morbid obesity makes visual- screws leading to serious neurological or vascular injury.
ization of the pelvic bony landmarks very difficult [13]. In By performing preoperative measurements that closely
fact, if the preoperative lateral CT scout view does not correlate with the expected intraoperative angles, a surgeon
demonstrate identifiable landmarks, the intraoperative flu- can minimize the incidence of obtaining imperfect images
oroscopic lateral would also not be dependable. In such during the procedure. The inlet view appears to have more
cases, the surgeon must have sufficient information and variability between the preoperative and intraoperative
understanding from the preoperative CT imaging and the measurements. Typically, an ideal fluoroscopic inlet will
intraoperative fluoroscopic inlet and outlet views to pro- have a thickened cortical density that corresponds to the
ceed safely without a confirmatory lateral view. Obtaining overlap of the S1 anterior cortex of the S1 body. At times,
adequate fluoroscopic views is challenging in the obese the S1 and S2 body will have the same orientation and a
patient population and having a detailed preoperative plan very thick density can be appreciated. Often though, the S1
with the knowledge of the anticipated intraoperative views and S2 body will have a different orientation in the sagittal
is invaluable. plane and such a distinct cortical density is not appreciated.
Obtaining suboptimal views by malrotation of the C-arm When this variability is present, multiple views can make
has been shown to effect the safe placement of iliosacral the inlet view appear to have an appropriate density when it

is really just erroneous interpretation that does not truly ring. CT imaging is invaluable in demonstrating the injury
correspond to the correct anterior cortical overlap. patterns, detecting differences in sacral morphology, and
Although the preoperative measurements are not flawless, displaying the available osseous fixation pathways.
it is one more tool a surgeon can use towards performing a Although not flawless, preoperative CT sagittal recon-
safe and successful procedure. struction images allow for appropriate preoperative plan-
The limitations of this study include the small sample ning for anticipated intraoperative fluoroscopic inlet and
size. Our goal was not to compare patients in separate outlet views within 5°. Obtaining quality intraoperative
cohorts, but to see if a measuring technique was repro- images can be difficult in certain patient populations and
ducible to a single surgeon as well as applicable intraop- clinical situations. Possessing an in-depth understanding of
eratively. One surgeon made each angle measurement both each patient’s pelvic anatomy and correctly interpreting the
preoperatively and postoperatively and these measurements corresponding bony landmarks intraoperatively is para-
only differed by approximately 2° and statistically showed mount. Having knowledge of the desired intraoperative
a high degree of intraobserver correlation. At our institu- views preoperatively can prepare a surgeon, aid in effi-
tions, these measurements have been anecdotally very ciently obtaining the correct views intraoperatively, and
reproducible between surgeons. It would also be beneficial ultimately assist in the placement of safe iliosacral screws.
to validate good interobserver reliability to these specific
measurements. Although very close, this overall method is Compliance of ethical standards
not without potential for error. One potential source for Conflict of interest Jonathan G Eastman, MD and Milton L. Chip
error stems from patient positioning. If the lumbosacral Routt Jr., MD have no conflict of interest.
bump is placed either too proximally or distally, the patient
will have increased or decreased lumbar kyphosis, which Ethical standards This study was authorized by the local ethical
committee and was performed in accordance with the ethical stan-
will directly affect the translation of the preoperative dards of the 1964 Declaration of Helsinki as revised in 2000. The
measurement into the operating room. One patient in this ethical committee waived the need for informed consent since the
series had a 20° range of variation from the preoperative rights and interests of the patients would not be violated and their
and intraoperative inlet. It is hypothesized that malposi- privacy and anonymity would be assured by this study design.
tioning could have contributed to this as other patients did
not have such a high degree of change. Another source of
error could stem from misreading of the measurement
intraoperatively by the fluoroscopic technician. The mea-
surements as dictated in the operative notes were reported
to the surgeon by the radiology technician and not directly
visualized by the surgeon. Different fluoroscopic machines
display the degree of inlet and outlet cant with varying
degrees of detail. For instance, some fluoroscopic machines
only have a marking every 15°. This could easily be mis- References
interpreted, documented incorrectly, and ultimately lead to
an improper reading. Care should be taken to identify the 1. Routt ML Jr, Simonian PT (1996) Closed reduction and percu-
correct measurement and correlate this with the necessary taneous skeletal fixation of sacral fractures. Clin Orthop Relat
Res 329:121–128
identifiable osseous landmarks to obtain the correct view. 2. Routt ML Jr, Kregor PJ, Simonian PT et al (1995) Early results of
All patients had iliosacral screws placed in a supine percutaneous iliosacral screws placed with the patient in the
position in this study. In theory, the same method could be supine position. J Orthop Trauma 9(3):207–214
used to preoperatively plan with the patient in a prone 3. Routt ML Jr, Nork SE, Mills WJ (2000) Percutaneous fixation of
pelvic ring disruptions. Clin Orthop Relat Res 375:15–29
position. The surgeon would have to consider how the 4. Nork SE, Jones CB, Harding SP et al (2001) Percutaneous sta-
prone positioning would affect the lumbosacral alignment. bilization of U-shaped sacral fractures using iliosacral screws:
In comparison to supine positioning of the preoperative CT technique and early results. J Orthop Trauma 15(4):238–246
scan, the bolsters placed for appropriate positioning and 5. Carlson DA, Scheid DK, Maar DC et al (2000) Safe placement of
S1 and S2 iliosacral screw: the ‘‘vestibule’’ concept. J Orthop
padding may alter the orientation of the pelvis in space. Trauma 14:4
This potential change could be assessed on a lateral fluo- 6. Bishop JA, Routt MLC Jr (2012) Osseous fixation pathways in
roscopic view if possible and the difference accounted for. pelvic and acetabular fracture surgery: osteology, radiology, and
This was not investigated in this study as no patient was clinical applications. J Trauma Acute Care Surg. 72(6):1502–1509
7. Routt ML Jr, Simonian PT, Agnew SG et al (1996) Radiographic
placed prone during this period. recognition of the sacral alar slope for optimal placement of
In conclusion, there is a significant amount of anatomic iliosacral screws: a cadaveric and clinical study. J Orthop Trauma
variation of the pelvis, particularly the posterior pelvic 10(3):171–177

8. Conflitti JM, Graves ML (2010) Chip Routt ML, Jr. Radiographic 19. Routt ML Jr, Simonian PT (1998) Posterior pelvic ring disrup-
quantification and analysis of dysmorphic upper sacral osseous tions: iliosacral screws. In: Wiss D (ed) Masters techniques in
anatomy and associated iliosacral screw insertions. J Orthop orthopaedic surgery: fractures. Lippincott-Raven, Philadelphia
Trauma 24(10):630–636 20. Routt MLC Jr, Meier MC, Kregor PJ, Mayo KA (1993) Percu-
9. Farrell ED, Gardner MJ, Krieg JC et al (2009) The upper sacral taneous iliosacral screws with the patient supine technique. Oper
nerve root tunnel: an anatomic and clinical study. J Orthop Tech Ortho. 3(1):35–45
Trauma 23(5):333–339 21. Tonetti J, Carrat L, Lavellee S et al (1998) Percutaneous iliosa-
10. Gardner MJ, Morshed S, Nork SE et al (2010) Quantification of cral screw placement using image guided techniques. Clin Orthop
the upper and second sacral segment safe zones in normal and 354:103–110
dysmorphic sacra. J Orthop Trauma 24(10):622–629 22. Zwingmann J, Konrad G, Mehlhorn AT, Südkamp NP, Oberst M
11. Miller AN, Routt ML Jr (2012) Variations in sacral morphology (2010) Percutaneous iliosacral screw insertion: malpositioning
and implications for iliosacral screw fixation. J Am Acad Orthop and revision rate of screws with regards to application technique
Surg 20(1):8–16 (Navigated vs. Conventional). J Trauma 69(6):1501–1506
12. Moed BR, Kellam JF, McLaren A et al (2003) Internal fixation 23. Collinge C, Coons D, Tornetta P, Aschenbrenner J (2005) Stan-
for the injured pelvic ring. In: Tile M, Helfet DL, Kellam JF (eds) dard multiplanar fluoroscopy versus a fluoroscopically based
Fractures of the pelvis and acetabulum, 3rd edn. Lippincott navigation system for the percutaneous insertion of iliosacral
Williams & Wilkins, Philadelphia, pp 217–292 screws: a cadaver model. J Orthop Trauma 19(4):254–258
13. Miller AN, Krieg JC, Routt MLC (2012) Lateral sacral imaging 24. Day CS, Prayson MJ, Shuler TE et al (2000) Transacral versus
in the morbidly obese. J Orthop Trauma. May 29. [Epub ahead of modified pelvic landmarks for percutaneous iliosacral screw
print] placement—a computed tomography analysis and cadaveric
14. Ziran BH, Wasan AD, Marks DM, Olson SA, Chapman MW study. Am J Orthop 9:16–21
(2007) Fluoroscopic imaging guides of the posterior pelvis per- 25. Xu R, Ebraheim NB, Robke JK, Yeasting RA (1996) Radiologic
taining to iliosacral screw placement. J Trauma 62:347–356 evaluation of iliosacral screw placement. Spine. 21:582–588
15. Ricci WM, Mamczak C, Tynan M, Streubel P, Gardner M (2010) 26. Pennal GF, Sutherland GO (1961) Fractures of the pelvis (motion
Pelvic inlet and outlet radiographs redefined. J Bone Joint Surg picture). American Academy of Orthopaedic Surgeons Film
Am 92:1947–1953 Library, Park Ridge
16. Graves ML, Routt ML Jr (2011) Iliosacral screw placement: are 27. Mostafavi HR, Tornetta P 3rd (1996) Radiologic evaluation of
uniplanar changes realistic based on standard fluoroscopic the pelvis. Clin Orthop Relat Res 329:6–14
imaging? J Trauma 71(1):204–208 (Discussion 208) 28. Sagi HC, Lindvall EM (2005) Inadvertent intraforaminal iliosa-
17. Marsh JL, Slongo TF, Agel J et al (2007) Fracture and dislocation cral screw placement despite apparent appropriate positioning on
classification compendium-2007: orthopaedic trauma association intraoperative fluoroscopy. J Orthop Trauma 19(2):130–133
classification, database and outcomes committee. J Orthop 29. Wolinsky P, Lee M (2007) The effect of C-arm malrotation on
Trauma 21(Supplement 10):S1–S133 iliosacral screw placement. J Orthop Trauma 21(7):427–434
18. Sagi HC, Coniglione FM, Stanford JH (2011) Examination under
anesthetic for occult pelvic ring instability. J Orthop Trauma
25(9):529–536

15
Cross-cultural adaptation and validation of the International
Knee Documentation Committee Subjective Knee Form in Greek
George A. Koumantakis1 • Konstantinos Tsoligkas2 • Antonios Papoutsidakis3 •
Athanasios Ververidis4 • Georgios I. Drosos4
Abstract construct validity. No floor/ceiling effects were recorded.

Background Patient-reported outcomes require validation The effect size was large (ES = 1.26).
in a particular language and culture before administration Conclusions The IKDC/SKF-GR has comparable mea-
for clinical use. surement properties to the original form.
Materials and methods A systematic translation of the Level of evidence Level II.
IKDC Subjective Knee Form was initially tested in 30
patients with various knee pathologies to develop the first Keywords International Knee Documentation
Greek version (IKDC/SKF-GR). It was then administered Committee (IKDC) Greek Cross-cultural adaptation
to another 80 patients. The test–retest reliability (n = 35) Validation Knee SF-36
and internal consistency (n = 80) were examined. Con-
struct validity was tested by correlating the IKDC/SKF-GR
with the SF-36 subscales (n = 80) and content validity by Introduction
measuring floor/ceiling effects. Responsiveness was mea-
sured in patients with meniscus pathology (n = 24). Several knee-specific patient-reported outcomes (PROs)
Results Patients filled the form without omissions/ques- have been developed to capture current functional and/or
tions regarding the phrasing of items. Internal consistency symptom status of patients with various knee conditions
was good (Cronbach’s a = 0.87) and test–retest reliability [6]. The International Knee Documentation Committee
very good (ICC2,1 = 0.95, SEM = 4.4 and SDC = 12.2). (IKDC) Subjective Knee Form, in particular, monitors
Correlations with the SF-36 subscales confirmed its symptoms and functional status (both in daily and sports
activities) and has been extensively validated in patients
with various knee pathologies [13, 14] and meniscus
injuries [7]. This form has also been found to have equal or
superior measurement properties to other similar measures
& George A. Koumantakis of knee function in patients with complex knee disorders
[email protected] [1], chondral defects [11], meniscus injury (waiting list and
1
post-surgery) [23], ACL rupture and reconstruction [24].
Department of Physical Therapy, 401 Army General Hospital
of Athens, 1 Panagioti Kanellopoulou Avenue, 11525 Athens,
Translations of the IKDC whole form into other
Greece languages (http://www.sportsmed.org/Research/IKDC_
2
2nd Department of Anesthesiology-Pain Unit, ATTIKON
Forms/) as well as cross-cultural adaptations of the IKDC
University General Hospital, School of Medicine, University Subjective Knee Form in the Italian [20], Dutch [12], Thai
of Athens, Athens, Greece [17], Brazilian [19], Chinese [10], Korean [15], Persian [9]
3
Orthopaedic Surgeon, Rethymno, Greece and Turkish [3] languages have already been reported.
4 The purpose of this study was to provide a valid Greek
Department of Orthopaedic Surgery, University General
Hospital of Alexandroupolis, School of Medicine, version of the widely used IKDC Subjective Knee Form, to
Democritus University of Thrace, Alexandroupolis, Greece inform future knee-related outcome studies performed in

Cross-cultural adaptation and validation of the International Knee Documentation Committee Subjective... 105
Greek-speaking populations, and to provide a common Professionals/Research/Grants/IKDC_Forms/IKDC%20Gre

PRO of knee functional status between populations with a ek(1).pdf.
different native language. A systematic cross-cultural
adaptation process was followed [2], and an evaluation of The IKDC Subjective Knee Form further validation
the internal consistency, between-day reliability, construct procedure
and content validity and responsiveness of this form was
performed according to current recommendations of the The cross-culturally adapted IKDC/SKF-GR form was
minimum standards of testing the psychometric properties administered to 80 consecutive patients, 64 (80 %) of
of PROs [21]. which were male and 16 (20 %) female, between March
and August 2010. The patient population tested presented
with a variety of knee disorders (Table 1) examined in the
Materials and methods orthopaedic clinics of our hospital, and had a mean (SD)
age of 35.3 (11.9) years, height of 175.6 (8.7) cm and body
The IKDC Subjective Knee Form mass of 81.0 (12.7) kg.
To establish the test–retest reliability over a 2-week
The form consists of 10 items assessing ‘symptoms’, interval, the scale was filled in twice by a subgroup of the
‘sports activities’ and ‘function’, covering all knee-related participants (n = 35). Internal consistency was also mea-
injuries. The total score is the sum of the individual item sured, including data of all participants (n = 80). Construct
scores and then the score is transformed to a scale ranging (convergent and divergent) validity was tested by corre-
from 0 to 100. The total score can be calculated if at least lating the IKDC/SKF-GR with a generic quality of life
90 % of the items are completed. scale (SF-36) [25] including all participants (n = 80).
Content validity was tested by measuring the floor and
The IKDC Subjective Knee Form translation ceiling effects.
and cross-cultural adaptation procedure Responsiveness of the scale was tested by administra-
tion of the scale in the 24 patients of our sample with
The systematic translation and cross-cultural adaptation of meniscus pathology on two occasions: on admission and at
the original 2000 version of the IKDC Subjective Knee a 3-month follow-up at the hospital. All patients, depend-
Form has been conducted according to detailed guidelines ing on their surgical management, had received written
[2]. Two separate forward translations from American instructions by the hospital physical therapy staff upon
English to Greek were made by two individuals whose discharge, to perform a home rehabilitation program con-
native language was Greek but who were also proficient in sisting of progressive loading, range of motion and
English. Discrepancies between the two translations were quadriceps strengthening exercises under non-weight-
resolved in a meeting and a synthesis of the two transla- bearing and weight-bearing conditions, and ice application
tions resulted in a common translation. Two individuals for effusion and pain control [18]. Patients were advised to
whose native language was English but were also proficient perform the set program for 20 min, 3 times per week. A
in Greek acted as back-translators of the common transla-
tion in Greek, producing two separate translations. An Table 1 Patients’ frequency of knee pathologies (n = 80)
expert committee of a methodologist, a clinician, a lan-
Knee pathology n (%)
guage expert and all translators reviewed all reports and
resolved any remaining discrepancies, assuring the Injury site
semantic, idiomatic, experiential and conceptual equiva- Right 51 (63.7)
lence between the two language versions. A pre-final ver- Left 27 (33.7)
sion of the scale was initially administered to 30 patients Bilateral 2 (2.5)
with various knee pathologies that were referred for Diagnosis
physical therapy in our hospital (pre-testing), with content ACL injury 41 (51.2)
and face validity between source and target versions ACL/MCL injury 1 (1.2)
assured, as all patients completed the IKDC Subjective ACL/meniscus injury 5 (6.2)
Knee Form in Greek (IKDC/SKF-GR) without omissions Meniscus injury 24 (30.0)
and demonstrated a good understanding of the scale items. Chondral injury 5 (6.2)
Therefore, this version was not modified further and was Osteoarthritis 3 (3.7)
considered the final version, available for download at: Plica 1 (1.2)
http://www.sportsmed.org/uploadedFiles/Content/Medical_

diary was kept and returned to the physical therapy Responsiveness was defined as an indicator of patient-
department at the 3-month follow-up. related change over time due to treatment effect [22]. The
responsiveness index used was the effect size (ES),
Statistical analysis expressed as the differences in the means of baseline and
3-month follow-up data, divided by the standard deviation
All analyses were performed with the IBM SPSS version at baseline [14, 22]. A value between 0.20 and 0.50 is
22 statistical package, with a 5 % level of significance. considered a small effect, between 0.51 and 0.80 a mod-
Questionnaire data were initially checked for normality of erate effect and above 0.80 a large effect [5].
distribution with the Kolmogorov–Smirnov test and were
found normally distributed (P [ 0.05), therefore paramet-
ric statistics were employed. Results
Reliability was assessed with the intraclass correlation
coefficient (ICC), which represents a ratio of the variance Patients
of interest over the sum of the variance of interest plus
error [8]. A two-way random-effects intraclass correlation The majority of the patients who participated in our study
coefficient type agreement (ICC2,1) was calculated, as had an isolated ACL injury (51.2 %) or an isolated
systematic differences were considered to be part of the meniscus injury (30 %); however, patients with various
measurement error, supplemented with calculation of the other knee pathologies were included (Table 1). Most of
standard error of measurement (SEM) and the smallest the patients were also male (80 %).
detectable change (SDC) [8, 16]. Bland–Altman plots were The IKDC/SKF-GR was filled in by all patients in
also constructed to depict in a scatter plot format absolute approximately 10 min, without omissions, and there were
agreement for test–retest measurements with 95 % limits no questions regarding the phrasing of the scale items. The
of agreement (LOA) [16]. Between-day systematic differ- mean (SD) and 95 % confidence interval (CI) data from the
ences were tested with a repeated measures ANOVA. IKDC-SKF/GR and the 8 domains and 2 summary scores
Internal consistency was calculated using the Chronbach a, of the SF-36 questionnaires from all participants are pre-
which addresses the homogeneity of the items comprising a sented in Table 2. The distribution of the IKDC-SKF/GR
questionnaire, with values of 0.70 considered fair, 0.80 scores is presented in Fig. 1.
good and above 0.90 excellent [4].
Construct validity was assessed by correlating the IKDC/
Test–retest reliability and agreement
SKF-GR with the subscales of the SF-36 (Pearson correla-
tion coefficient). Convergent validity is the degree of cor-
To assess the test–retest reliability, the form was admin-
relation of a particular outcome measure with other
istered twice (2-week interval) in a group of 35 patients (26
measures theoretically predicted to correlate with it; and
with anterior cruciate ligament reconstruction, 3 with
conversely, divergent validity is the degree to which an
ACL/meniscus, 1 with ACL/MCL, 1 with meniscus, 1 with
outcome measure does not correlate with other measures
plica syndrome and 3 with osteoarthritis, of which 31
that it is predicted not to correlate with [13]. The SF-36
(88.5 %) were male and 4 (11.5 %) female, with a mean
consists of 8 domains (physical functioning, PF; role limi-
(SD) age of 33.2 (11.6) years, height of 178.6 (5.89) cm
tation due to physical problems, RP; bodily pain, BP; gen-
and body mass of 83.4 (12.8) kg. Test–retest reliability and
eral health perceptions, GH; vitality, VT; social functioning,
agreement indices were considered to have sufficient
SF; role limitation due to emotional problems, RE; and
accuracy and clinical applicability [16], with an ICC2,1
mental health, MH), with each directly transformed into a
(95 % CI) 0.95 (0.91–0.98), SEM = 4.4, SDC = 12.2 and
scale from 0 to 100 (higher scores indicate better health
a mean test–retest difference value of 1.59. A repeated
status), to describe patients’ physical and mental states [25].
measures ANOVA did not demonstrate statistically sig-
The sum of PF, RP, BP and GH subscales generates a
nificant differences between the 2 measurement occasions
physical component summary (PCS) score and the sum of
(P = 0.136). The Bland–Altman limits of agreement ran-
the VT, SF, RE and MH generates a mental component
ged from -10.50 to 13.68 (Fig. 2).
summary (MCS) score [25]. For content validity/inter-
pretability, floor effects exist if a proportion of patients
report the lowest possible score, whereas ceiling effects exist Internal consistency
if a proportion of patients obtain the highest possible score
upon the administration of the questionnaire. Floor/ceiling The internal consistency of the IKDC/SKF-GR was good
effects of \20 % are considered acceptable [12, 13]. (Cronbach a = 0.87).

Table 2 Mean value, standard

Mean Standard deviation 95 % Confidence interval
deviation, and 95 % confidence
interval of the outcome IKDC/SKF-GR 54.21 19.73 49.99–58.59
measures (n = 80)
SF-36 PCS 40.30 10.42 37.95–42.70
SF-36 MCS 48.72 8.47 46.95–50.69
SF-36 PF 62.00 23.82 56.94–67.12
SF-36 RP 30.94 43.07 21.87–41.55
SF-36 BP 56.75 25.93 51.02–62.70
SF-36 GH 74.29 12.41 71.62–77.11
SF-36 VT 63.12 15.41 59.69–66.62
SF-36 SF 65.47 24.78 60.47–71.10
SF-36 RE 53.33 42.63 44.58–63.32
SF-36 MH 73.05 13.76 70.05–76.00
IKDC/SKF-GR International Knee Documentation Committee/Subjective Knee Form in Greek, SF-36 short
form 36, PCS physical component summary, MCS mental component summary, PF physical functioning,
RP role limitation due to physical problems, BP bodily pain, GH general health, VT vitality, SF social
functioning, RE role limitation due to emotional problems, MH mental health
Fig. 1 Distribution of calculated International Knee Documentation Fig. 2 Bland–Altman plots with the 3 bold lines representing
Committee Subjective Knee Form in Greek (IKDC/SKF-GR) scores the ±1.96 SD limits of agreement (superior and inferior) and the
(n = 80) average of the differences (intermediate) of the International Knee
Documentation Committee Subjective Knee Form in Greek (IKDC/
SKF-GR), (n = 35)
Construct validity
Finally, the level of correlation between the IKDC-SKF/
GR and the social functioning subscale of the SF-36
The Greek version of the IKDC-SKF demonstrated the
(r = 0.60, P \ 0.001), was higher than initially hypoth-
strongest correlations with the physical component sum-
esized (Table 3).
mary of the SF-36 as well as the physical functioning
subscale (r = 0.77 for both, P \ 0.001). Correlations
with the bodily pain (r = 0.72, P \ 0.001) and the role Content validity/interpretability
limitation due to physical problems (r = 0.68,
P \ 0.001) subscales were also strong. The weakest No floor or ceiling effects were identified for the IKDC/
associations were observed between the IKDC-SKF/GR SKF-GR, with a minimum value of 18.93 and a maximum
and the mental component summary (r = 0.22) as well as value of 93.67 recorded, therefore the content validity/in-
the mental health subscale (r = 0.26) of the SF-36. terpretability was good (Fig. 1).

Table 3 Pearson correlation coefficients between scores of the IKDC/SKF-GR (n = 80), the original and existing cross-cultural adaptations of
the IKDC and the SF-36 subscales
IKDC Subjective Knee Form
IKDC/SKF-GR Original Italian Dutch Thai Brazilian Chinese Korean Persian Turkish
[13] [20] [12] [17] [19] [10] [15] [9] [3]
SF-36 PCS 0.77** 0.66 0.60 – 0.63 0.79 – – 0.626 0.70

SF-36 MCS 0.22 0.16 0.40 – 0.34 0.51 – – 0.159 0.05
SF-36 PF 0.77** 0.63 0.67 0.71 0.75 0.75 0.64 0.66 0.522 0.69
SF-36 RP 0.68** 0.47 0.56 0.55 0.37 0.54 0.50 0.49 0.391 0.53
SF-36 BP 0.72** 0.64 0.75 0.69 0.76 0.63 0.64 0.30 0.679 0.47
SF-36 GH 0.47 0.30 0.26 0.41 0.21 0.54 0.50 0.11 0.336 0.32
SF-36 VT 0.42 0.39 0.36 0.40 0.29 0.46 0.44 0.15 0.402 0.24
SF-36 SF 0.60** 0.47 0.58 0.42 0.22 0.43 0.41 0.48 0.385 0.40
SF-36 RE 0.44 0.26 0.44 0.30 0.34 0.50 0.24 0.30 0.167 0.22
SF-36 MH 0.26 0.25 0.65 0.21 0.29 0.40 0.41 0.15 0.196 0.13
IKDC/SKF-GR International Knee Documentation Committee/Subjective Knee Form in Greek, SF-36 short form 36, PCS physical component
summary, MCS mental component summary, PF physical functioning, RP role limitation due to physical problems, BP bodily pain, GH general
health, VT vitality, SF social functioning, RE role limitation due to emotional problems. MH mental health
** Correlation is significant at the 0.001 level (two-tailed)
Responsiveness those in several other languages (Table 3). Specifically, the

test–retest reliability and agreement (n = 35), the internal
Of the 24 patients with meniscus pathology, 6 were only consistency, convergent-divergent validity and floor/ceil-
managed conservatively (no surgery), 2 had a meniscus ing effects (n = 80), and the responsiveness (n = 24) of
repair and 16 had a partial meniscectomy. Also, 14 the IKDC-SKF/GR were examined.
(58.3 %) were male and 10 (41.7 %) female, with a mean As can be seen in Table 3, for convergent validity,
(SD) age of 39.0 (12.8) years, height of 172.6 (8.51) cm correlation with the PCS, PF and BP subscales of the SF-36
and body mass of 75.4 (10.8) kg. The mean (SD) number were similar to the original version, while correlation with
of recorded home sessions were 29.6 (4.5) from a maxi- the RP (r = 0.68) and SF (r = 0.60) subscales of the SF-
mum of 36 (3 times per week for 12 weeks), indicating a 36 were higher in our study, compared to the original scale
high compliance level, above 80 % of the required. There (r = 0.47 for both). Divergent validity of the IKDC/SKF-
was no correlation between the number of home sessions GR was confirmed to be similar to the original version, as
performed (ranging between 20 and 36) and the change in correlations with the MCS, GH, VT, RE, MH subscales of
the IKDC-SKF/GR recorded. the SF-36 were equally low. In addition, no floor or ceiling
The IKDC/SKF-GR mean (SD) values on admission to effects were recorded, which is a desired attribute of a
hospital were 49.48 (14.81) and at the 3-month follow-up questionnaire for scores not to be clustered at the top or
68.15 (10.72), recording a mean (SD) increase of 18.67 lower end of a questionnaire [21].
(7.15) units. The effect size was 1.26, considered to be of a The relative reliability index ICC2,1 = 0.95 was almost
high level. the same as in the original validation paper (ICC = 0.94)
[13], as well as in other validation studies. Additional
information is contained in the SEM/SDC absolute agree-
Discussion ment indices, expressing the degree to which scores are
identical, in terms of the original measurement [16]. In our
The process of cross-cultural adaptation of the IKDC-SKF study the SEM = 4.4 and SDC = 12.2 were slightly
in Greek followed the Guillemin criteria [2] and was sub- higher than the original validation of the IKDC/SKF, which
sequently validated in a Greek-speaking population with reported an SDC of 9 points, with improvements (or
various knee pathologies, demonstrating comparable mea- deterioration) beyond this range considered as true change.
surement properties with the original scale [13, 14] and The SDC levels in other validation studies ranged between

6.7 and 16.4 points. In the validation study of the Brazilian Acknowledgments We would like to thank the Orthopaedic Clinics
version of IKDC, the relevant SEM/SDC values were the of 401 Army General Hospital of Athens, for referring patients to our
study.
lowest (SEM = 2.4/SDC = 6.7) [19], and in the Turkish
version the highest (SEM = 6.0/SDC = 16.4) [3], while Compliance with ethical standards
they were not reported in the other cross-cultural validation
studies. In a study performed in patients with isolated Conflict of interest None.
meniscus injury the SEM/SDC were found to be similar to Ethical standards The study conforms to the 1964 Helsinki Dec-
the original validation (SEM/SDC = 3.19/8.8) [7]. In laration and its later amendments; the study was approved by the
another study examining patients with focal articular car- institutional ethical review board; all the patients provided informed
tilage defects, the SEM/SDC values were reported to be consent before being enrolled.
slightly better in the longer-term than in the shorter-term
follow-up (SEM/SDC = 5.6/15.6 in 6 months vs SEM/
SDC = 4.9/13.7 in 12 months) [11].
Alternatively, values beyond the limits of agreement, as
depicted in the Bland–Altman plots, can be considered as a
meaningful change in IKDC scores, signifying an alteration
in a patient’s symptomatology [16]. In our study the mean
difference between the two testing occasions was 1.59 (not
statistically significant), and the LOA was between -10.50 References
and 13.68. The only other cross-cultural adaptation study
reporting LOA values is the Brazilian validation study [19], 1. Agel J, LaPrade RF (2009) Assessment of differences between
the modified Cincinnati and International Knee Documentation
with a mean difference between testing occasions of only
Committee patient outcome scores: a prospective study. Am J
0.50 and LOAs between -6.1 and 7.1. Sports Med 37(11):2151–2157. doi:10.1177/0363546509337698
The internal consistency of the IKDC/SKF-GR scale 2. Beaton DE, Bombardier C, Guillemin F, Ferraz MB (2000)
was found to be good (a = 0.87) and at a slightly lower Guidelines for the process of cross-cultural adaptation of self-
report measures. Spine 25(24):3186–3191
level than the original scale (a = 0.92) [13]. From previous
3. Celik D, Coskunsu D, KiliCoglu O, Ergonul O, Irrgang JJ (2014)
similar studies, the internal consistency Chronbach a index Translation and cross-cultural adaptation of the International
ranged between 0.77 and 0.97 [3, 7, 9, 10, 12, 15, 17, 19, Knee Documentation Committee Subjective Knee Form into
20]. Turkish. J Orthop Sports Phys Ther 44(11):899–909. doi:10.
2519/jospt.2014.4865
Responsiveness in the original validation of the IKDC in
4. Chronbach LJ (1951) Coefficient alpha and the internal structure
patients with a variety of knee conditions had an of tests. Psychometrika 16(3):297–334
ES = 1.13 [14] and in another study involving only 5. Cohen J (1969) Statistical power analysis for the behavioral
patients with meniscus pathology ES = 2.11 [7]. The sciences. Academic Press, New York
6. Collins NJ, Misra D, Felson DT, Crossley KM, Roos EM (2011)
responsiveness of the IKDC-SKF/GR was also found to be
Measures of knee function: International Knee Documentation
high (ES = 1.26) in our study, tested in a subsample of the Committee (IKDC) Subjective Knee Evaluation Form, Knee
whole population used (only in patients with meniscus Injury and Osteoarthritis Outcome Score (KOOS), Knee Injury
pathology). Although the IKDC-SKF/GR change score and Osteoarthritis Outcome Score Physical Function Short Form
(KOOS-PS), Knee Outcome Survey Activities of Daily Living
value exceeded the reliability test–retest error (SDC and
Scale (KOS-ADL), Lysholm Knee Scoring Scale, Oxford Knee
LOA), only indirect inferences can be made, as the test– Score (OKS), Western Ontario and McMaster Universities
retest data were derived from a different subsample, which Osteoarthritis Index (WOMAC), Activity Rating Scale (ARS),
included only 1 patient with meniscus pathology. and Tegner Activity Score (TAS). Arthritis Care Res (Hoboken)
63(Suppl 11):S208–S228. doi:10.1002/acr.20632
Finally, since the IKDC scale can account for differ-
7. Crawford K, Briggs KK, Rodkey WG, Steadman JR (2007)
ences across cultures, it may allow for combining and Reliability, validity, and responsiveness of the IKDC score for
comparing data from populations of different language and meniscus injuries of the knee. Arthroscopy 23(8):839–844.
cultural backgrounds. Such comparisons may also provide doi:10.1016/j.arthro.2007.02.005
8. de Vet HC, Terwee CB, Knol DL, Bouter LM (2006) When to
the possibility of studying differences in healthcare deliv-
use agreement versus reliability measures. J Clin Epidemiol
ery and patient management between different countries. 59(10):1033–1039. doi:10.1016/j.jclinepi.2005.10.015
In conclusion, the IKDC Subjective Form in Greek has 9. Ebrahimzadeh MH, Makhmalbaf H, Golhasani-Keshtan F,
demonstrated comparable psychometric properties to the Rabani S, Birjandinejad A (2014) The International Knee Doc-
umentation Committee (IKDC) Subjective Short Form: a validity
original version, therefore the scale is recommended for
and reliability study. Knee Surg Sports Traumatol Arthrosc.
further use in Greek-speaking patients with knee pathology. doi:10.1007/s00167-014-3107-1

10. Fu SN, Chan YH (2011) Translation and validation of Chinese cartilage lesions. J Orthop Sports Phys Ther 40(6):A1–A35.
version of International Knee Documentation Committee Sub- doi:10.2519/jospt.2010.0304
jective Knee Form. Disabil Rehabil 33(13–14):1186–1189. 19. Metsavaht L, Leporace G, Riberto M, de Mello Sposito MM,
doi:10.3109/09638288.2010.524274 Batista LA (2010) Translation and cross-cultural adaptation of the
11. Greco NJ, Anderson AF, Mann BJ, Cole BJ, Farr J, Nissen CW, Brazilian version of the International Knee Documentation
Irrgang JJ (2010) Responsiveness of the International Knee Committee Subjective Knee Form: validity and reproducibility.
Documentation Committee Subjective Knee Form in comparison Am J Sports Med 38(9):1894–1899. doi:10.1177/
to the Western Ontario and McMaster Universities Osteoarthritis 0363546510365314
Index, modified Cincinnati Knee Rating System, and Short Form 20. Padua R, Bondi R, Ceccarelli E, Bondi L, Romanini E, Zanoli G,
36 in patients with focal articular cartilage defects. Am J Sports Campi S (2004) Italian version of the International Knee Docu-
Med 38(5):891–902. doi:10.1177/0363546509354163 mentation Committee Subjective Knee Form: cross-cultural
12. Haverkamp D, Sierevelt IN, Breugem SJ, Lohuis K, Blankevoort adaptation and validation. Arthroscopy 20(8):819–823. doi:10.
L, van Dijk CN (2006) Translation and validation of the Dutch 1016/j.arthro.2004.06.011
version of the International Knee Documentation Committee 21. Reeve BB, Wyrwich KW, Wu AW, Velikova G, Terwee CB,
Subjective Knee Form. Am J Sports Med 34(10):1680–1684. Snyder CF, Schwartz C, Revicki DA, Moinpour CM, McLeod
doi:10.1177/0363546506288854 LD, Lyons JC, Lenderking WR, Hinds PS, Hays RD, Greenhalgh
13. Irrgang JJ, Anderson AF, Boland AL, Harner CD, Kurosaka M, J, Gershon R, Feeny D, Fayers PM, Cella D, Brundage M, Ahmed
Neyret P, Richmond JC, Shelborne KD (2001) Development and S, Aaronson NK, Butt Z (2013) ISOQOL recommends minimum
validation of the International Knee Documentation Committee standards for patient-reported outcome measures used in patient-
Subjective Knee Form. Am J Sports Med 29(5):600–613 centered outcomes and comparative effectiveness research. Qual
14. Irrgang JJ, Anderson AF, Boland AL, Harner CD, Neyret P, Life Res 22(8):1889–1905. doi:10.1007/s11136-012-0344-y
Richmond JC, Shelbourne KD, International Knee Documenta- 22. Terwee CB, Dekker FW, Wiersinga WM, Prummel MF, Bossuyt
tion Committee (2006) Responsiveness of the International Knee PM (2003) On assessing responsiveness of health-related quality
Documentation Committee Subjective Knee Form. Am J Sports of life instruments: guidelines for instrument evaluation. Qual
Med 34(10):1567–1573. doi:10.1177/0363546506288855 Life Res 12(4):349–362
15. Kim JG, Ha JK, Lee JY, Seo SS, Choi CH, Lee MC (2013) 23. van de Graaf VA, Wolterbeek N, Scholtes VA, Mutsaerts EL,
Translation and validation of the Korean version of the Interna- Poolman RW (2014) Reliability and validity of the IKDC,
tional Knee Documentation Committee Subjective Knee Form. KOOS, and WOMAC for patients with meniscal injuries. Am J
Knee Surg Relat Res 25(3):106–111. doi:10.5792/ksrr.2013.25.3. Sports Med 42(6):1408–1416. doi:10.1177/0363546514524698
106 24. van Meer BL, Meuffels DE, Vissers MM, Bierma-Zeinstra SM,
16. Kottner J, Audige L, Brorson S, Donner A, Gajewski BJ, Hrob- Verhaar JA, Terwee CB, Reijman M (2013) Knee injury and
jartsson A, Roberts C, Shoukri M, Streiner DL (2011) Guidelines Osteoarthritis Outcome Score or International Knee Documen-
for Reporting Reliability and Agreement Studies (GRRAS) were tation Committee Subjective Knee Form: which questionnaire is
proposed. J Clin Epidemiol 64(1):96–106. doi:10.1016/j.jclinepi. most useful to monitor patients with an anterior cruciate ligament
2010.03.002 rupture in the short term? Arthroscopy 29(4):701–715. doi:10.
17. Lertwanich P, Praphruetkit T, Keyurapan E, Lamsam C, Kul- 1016/j.arthro.2012.12.015
thanan T (2008) Validity and reliability of Thai version of the 25. Ware JEJ (2000) SF-36 health survey update. Spine
International Knee Documentation Committee Subjective Knee 25(24):3130–3139
Form. J Med Assoc Thai 91(8):1218–1225
18. Logerstedt DS, Snyder-Mackler L, Ritter RC, Axe MJ (2010)
Knee pain and mobility impairments: meniscal and articular

16
Effectiveness of extracorporeal shockwave therapy in three major
tendon diseases
Christian Carulli1 • Filippo Tonelli1 • Matteo Innocenti1 • Bonaventura Gambardella1 •
Francesco Muncibı̀1 • Massimo Innocenti1
Abstract several chronic musculoskeletal disorders, allowing satis-

Background Extracorporeal shockwave therapy is a factory pain relief and improvement of function ability.
conservative treatment for several painful musculoskeletal Level of evidence Level IV.
disorders. The aim of the study was the assessment of the
relief from pain by the shockwave therapy in a population Keywords Extracorporeal shockwave Calcific
of consecutive patients affected by specific pathologies. tendonitis of the shoulder Achilles tendinopathy
Materials and methods A group of consecutive patients Epicondylitis
were studied and treated. They were affected by calcific
tendonitis of the shoulder (129 patients), chronic Achilles
tendinopathy (102 patients), and lateral epicondylitis of the Introduction
elbow (80 subjects). Each patient had 3 applications with a
monthly interval, and was followed up at 1, 6, and Extracorporeal shock wave therapy (ESWT) is one of the
12 months after treatment. Results were evaluated by the great advances in orthopaedics over the last 20 years [1].
numeric rating scale (NRS) in all cases, the Constant Initially indicated for the treatment of kidney stones [2], it
Murley Score for the assessment of the shoulder function, has been applied in cases of bone non-unions, and then in
the American Orthopaedic Foot and Ankle Society Score several musculoskeletal disorders, given the satisfactory
for subjects affected by chronic Achilles tendinopathy, and clinical outcomes reported in different randomized clinical
the Oxford Elbow Score for those affected by a lateral trials and cohort studies. The main indications have been
epicondylitis of the elbow. the following: lateral epicondylitis of the elbow, calcific
Results One year after treatment, the results were con- tendonitis of the rotator cuff, plantar fasciitis, Achilles and
sidered satisfactory with an almost complete resolution of patellar tendinopathy, and pubalgia [1, 3–13]. A reduction
symptoms. There were statistically significant results at the of pain and a good recovery of articular function have been
12-month follow-ups regarding the mean NRS score (from obtained in most cases [13–17], even if in high-level ath-
6.25 to 0.2), the Constant Murley Score (from 66.7 to letes a more aggressive strategy is recommended to allow a
79.4), the Oxford Elbow Score (from 28 to 46), and the quick return to sports activities [18].
AOFAS (from 71 to 86). The mechanism by which ESWT may produce a clinical
Conclusions Extracorporeal shockwave therapy may be effect is still uncertain. Several theories have been pro-
considered a safe, economic, and effective treatment for posed: a mechanical effect by increasing the pressure in the
calcium deposition causing fragmentation; a molecular
effect with induction of an inflammatory response with
neovascularization and then a chemotactic action and
& Christian Carulli phagocytosis of calcific deposits; an analgesic effect by
[email protected]
inhibiting the activation of the serotonergic system, and
1
Orthopaedic Clinic, University of Florence, Largo P. Palagi peripheral denervation. Probably, a combination of angio-
1, 50139 Florence, Italy genic and analgesic effects explains the overall outcomes

on the target tissues [7, 8, 19–23]. Direct and indirect Pain assessment in all patients was conducted before
biologic effects of ESWT vary proportionally to the treatment by an 11-point numeric scale (numeric rating
amount of energy and to the type of frequency applied; scale, NRS). The clinical evaluation was conducted by
moreover, the shockwave driving tool influences the the Constant Score for the assessment of shoulder func-
induced modifications on the target tissue [24]. tion [25]; the American Orthopaedic Foot and Ankle
The aim of the present retrospective study is the eval- Society Score (AOFAS) [26] for subjects affected by
uation of the clinical outcomes in a population of patients chronic Achilles tendinopathy; and the Oxford Elbow
affected by common musculoskeletal disorders treated by Score [27] for those suffering a lateral epicondylitis of
ESWT. the elbow. All patients gave their consent to the treatment
and follow-up.
A single device generating shockwaves (ReflecTronÒ,
Materials and methods HMT, Switzerland) was used in all cases. The energy level
and number of shots were adapted to the specific pathology
From January 2011 to March 2013, 311 consecutive according to the protocols supplied by the manufacturer.
patients were selected and treated by ESWT for specific Each patient had 3 ESWT applications at monthly inter-
painful musculoskeletal disorders at the authors’ institu- vals. Each session consisted of 2400 shockwave applica-
tion. One-hundred and twenty-nine were affected by a tions with an intensity depending on the site and the
calcific tendonitis of the shoulder, 102 by an Achilles pathology observed (Table 2). No local anaesthesia was
tendinopathy, and 80 by a later epicondylitis of the elbow. given before the treatment. All patients were treated by two
The mean age was 48.5 (range 19–80); 230 were male, experienced orthopaedic surgeons.
and 81 female. Inclusion criteria were: adult patients with All subjects were followed up at 1, 6, and 12 months
clinical and instrumental diagnosis of lateral epicondilytis after the last application. The clinical evaluation consisted
of the elbow, chronic Achilles tendinopathy, and calcific of NRS and function evaluation by the administration of
tendonitis of the shoulder; persistent symptoms for at least the above mentioned specific scores (Table 3). Particular
3 months; failure or partial resolution of symptoms after attention has been focused on the use of analgesic drugs,
conservative (medical and physical) treatment; no recent reported complications after the ESWT applications, and
history of trauma or chronic joint instability; no recent the need for any further instrumental study.
related surgery. The statistical analysis was performed by a sample
Exclusion criteria were: patients with a clinical but not size calculation based on a priori assumption of
instrumental diagnosis of any tendon disease; subjects who p = 0.05. All data were tested for the normal distribution
had not tried any conservative approach; subjects referring using the Kolmogorov–Smirnov test. The Student t-test
an inadequate duration of proper medical or physical was used to perform the analysis for the scores, testing
treatments. The institutional review board allowed the each disease separately. For each parameter, three cou-
retrospective analysis of patients’ data and outcomes. pled samples were calculated (before treatment–1 month,
Demographic data of the selected patients are reported in before treatment–6 months, before treatment–12 months)
Table 1. (Table 3).
Table 1 Demographics and characteristics of the patients

Shoulder calcific tendonitis Achilles tendinopathy Elbow lateral
(n = 129) (n = 102) epicondylitis (n = 80)
Male/female 92/37 46/56 45/35

Mean age (range) 47.5 (19–70) 48 (22–80) 50 (20–76)
Mean duration of symptoms (weeks) 4.3 (3–7) 6.7 (2–9) 3.9 (2–6)
Dominant side affected 72 64 56
Previous treatments (number of subjects)
NSAIDs 31 24 19
Other analgesics 52 41 28
Physical therapy 21 16 12

Effectiveness of extracorporeal shockwave therapy in three major tendon diseases 113
Table 2 Active level of ESWT Discussion

Disease Pulses and energy of ESWT
Shockwave therapy represents an innovative approach for
Calcific tendonitis of the shoulder 2400 pulses at [0.20 mJ/mm2 the management of painful chronic musculoskeletal dis-
Achilles tendinopathy 2400 pulses at 0.08–0.33 mJ/ eases, particularly in the case of failure of a previous
mm2
conservative treatment. This treatment has to be considered
Lateral epicondylitis of the elbow 2400 pulses at \0.12 mJ/mm2
safe, minimally invasive, versatile, and with low costs [28].
In the present study, as reported in the literature, after a
latency of days to a few weeks after treatment, all patients
Results reported a clinical benefit, with a significant decrease of
pain, improvement in function, and a fair use of analgesics.
Two-hundred and eighty-three patients completed the fol- Focusing attention on the specific pathologies, our out-
low-up period. Twenty-eight subjects were lost: none of comes are in line with the latest reports.
them was lost due to conditions or complications related to Lateral epicondylitis of the elbow was treated by ESWT in
the procedures. five recent RCTs, mostly of high quality [29–33]. In two of
No complications were recorded. In 42 cases, the these, no significant differences were found up to 48 weeks
patients reported the presence of cutaneous bruises after the after the treatment between ESWT and placebo [31, 32].
applications. The overall mean NRS score was 6.25 (range Spacca et al. [31] found significant differences between
4–9) before the treatment. One month after the first ESWT and placebo on pain (0.5 versus 6.5) and grip strength
application, the mean NRS score was 4.9 (range 3–9), 1.2 (46 versus 36) 12 weeks after the treatment. Pettrone et al.
at 6 months (range 0–3), and finally 0.2 at 12 months [29] found similar significant differences in pain at the
(range 0–2). Considering single pathologies, patients 12-week follow-up. Collins et al. [33] found significant dif-
showed an improvement in any score: mean NRS, mean ference in pain during activity in favour of the ESWT group.
Constant Murley Score for shoulders, mean Oxford Elbow There is conflicting evidence for the effectiveness of ESWT
Score for elbows, and mean AOFAS Score for feet versus placebo in the short term and evidence of no differ-
(Table 3). Over the months of follow-up we recorded a ence in effect on the mid-term and long-term follow-up.
progressive maintenance of results (Figs. 1, 2). The use of Several studies have confirmed the benefits of ESWT for
pain regulating drugs was reported by 34 patients (12.0 %), the treatment of calcific tendonitis of the shoulder [24, 34,
with peak utilisation on the first 3 days, once daily. In 12 35]. Particularly, it has been reported that high-energy
cases (4.2 %), the pain did not show a significant decrease ESWT (EFD C 0.28 mJ/mm2) are more effective than
so an ultrasound or MRI examination was necessary to low-energy doses (EFD \ 0.28 mJ/mm2) in the improve-
understand the causes of the persistency of symptoms. ment of the shoulder function and pain resolution.
Table 3 Clinical and functional scores

Baseline 1 month 6 months 12 months
Numeric rating scale (NRS)

Calcific tendonitis of the shouldera 6.5 ± 1.4 (4–9) 5.9 ± 1.3 (3–9)* 1.2 ± 0.8 (0–3)* 0.2 ± 0.4 (0–1)*
p = 00.013 p \ 0.001 p \ 0.001
Achilles tendinopathya 6.9 ± 1.2 (5–9) 5.3 ± 1.1 (4–8)* 1.7 ± 0.8 (0–3)* 0.3 ± 0.5 (0–2)*
p \ 0.001 p \ 0.001 p \ 0.001
Lateral epicondylitis of the elbowa 6.6 ± 1.2 (4–9) 4.2 ± 1.0 (3–6)* 0.9 ± 0.8 (0–3)* 0.1 ± 0.3 (0–1)*
p \ 0.001 p \ 0.001 p \ 0.001
Functional scores
Constant Murley Score 66.7 ± 4.3 (56–76) 73.7 ± 3.9 (59–78)* 78.3 ± 2.6 (64–80)* 79.4 ± 1.4 (70–80)*
p = 00.012 p \ 0.001 p \ 0.001
AOFAS 71 ± 5.6 (63–80) 72 ± 3.2 (67–75)* 77 ± 2.4 (72–84)* 86 ± 1.9 (82–90)*
p \ 0.001 p \ 0.001 p \ 0.001
Oxford Elbow Score 28 ± 2.7 (23–35) 35 ± 2.5 (31–38)* 42 ± 2.6 (36–47)* 46 ± 2.6 (42–50)*
p = 0.0016 p \ 0.001 p \ 0.001
* Paired Student t-test, compared to baseline (p \ 0.05)
a
The use of pain regulating drugs was reported by 34 patients (12.0 %), with a peak of utilization in the first 3 days, once daily

Fig. 1 NRS scores after 9

treatment
8 ±1.4 ±1.2 ±1.2
±1.3
7
±1.1
±1.0
5 Calciic Tendonitis
Achilles Tendinopaty
4
Lateral Epicondylitis
3
±0.8
±0.8
2 ±0.8
1
±0.4 ±0.5
±0.3
0
-1
Fig. 2 Constant Murley, 90 ±1.9

AOFAS, and Oxford Elbow
Scores after treatment ±1.4
±2.6 ±2.4
80
±5.3 ±3.9
±3.2
±4.3
70
60
50 ±2.6
Constant Murley Score
±2.6
AOFAS
40
±2.5
Oxford Elbow Score
±2.7
30
20
10
0

Effectiveness of extracorporeal shockwave therapy in three major tendon diseases 115
Gerdesmayer et al. [35] enrolled 144 patients with a ran- References

domized level of energy (high or low). Both types of
ESWT resulted in a significant improvement at the 1. Ogden JA, Alvarez RG, Levitt R et al (2001) Shock wave therapy
6-month evaluation, but high-energy ESWT induced a (orthotripsy) in musculoskeletal disorders. Clin Orthop 387:22–40
2. Farsi HM, Mosli HA, Alzimaity M et al (1994) In situ extra-
higher outcome on the Constant Murley Score. Calcific corporeal shock wave lithotripsy for primary ureteric calculi.
deposits disappeared in the same percentage of patients in Urology 43:776–781
both groups. Cacchio et al. [34] used a different score 3. Rompe JD, Maffulli N (2007) Repetitive shock wave therapy for
lateral elbow tendinopathy (tennis elbow): a systematic and
(University of California–Los Angeles UCLA Shoulder
qualitative analysis. Br Med Bulletin 83:355–378
Rating Scale) to evaluate shoulder function after ESWT 4. Thomson CE, Crawford F, Murray GD (2005) The effectiveness of
treatment versus placebo of calcific tendonitis of the extra corporeal shock wave therapy for plantar heel pain: a sys-
shoulder. Significant differences in favour of ESWT versus tematic review and meta-analysis. BMC Musculoskelet Disord 6:19
placebo were reported at the 6-month follow-up. 5. Mouzopoulos G, Stamatakos M, Mouzopoulos D et al (2007)
Extracorporeal shock wave treatment for shoulder calcific ten-
ESWT is effective as a conservative approach in the donitis: a systematic review. Skeletal Radiol 36:803–811
treatment of chronic Achilles tendinopathy. This has been 6. Roehrig GJ, Baumhauer J, DiGiovanni BF et al (2005) The role
recently confirmed by some important RCTs [12, 36, 37]. of extracorporeal shock wave on plantar fasciitis. Foot Ankle Clin
Rasmussent et al. [12] showed improvements in the 10:699–712
7. Ho C (2007) Extracorporeal shock wave treatment for chronic
treatment with ESWT versus placebo at a 12-week fol- rotator cuff tendonitis (shoulder pain). Issues in Emerg Health
low-up. The mean AOFAS Score increased from 74 to 81 Technol 96:1–4
in the placebo group and from 70 to 88 in the ESWT 8. Farr S, Sevelda F, Mader P et al (2001) Extracorporeal shock-
group (p = 0.05). Better results were seen in the ESWT wave therapy in calcifying tendinitis of the shoulder. Knee Surg
Sports Traumatol 19:2085–2089
group at 8 and 12 weeks (p = 0.01 and p = 0.04, 9. Radwan Y, ElSobhi G, Badawy et al (2008) Resistant tennis
respectively). Rompe et al. [37] showed an improvement elbow: shock-wave therapy versus percutaneous tenotomy. Int
in the VISA-A score (specific for Achilles tendon Orthop 32:671–677
pathologies) which increased in two groups: one with a 10. Saxena A, Ramdath S, O’Halloran et al (2011) Extra-corporeal
pulsed-activated therapy (‘‘EPAT’’ sound wave) for Achilles
treatment by eccentric loading exercises and one with tendinopathy: a prospective study. J Foot Ankle Surg 50:315–319
eccentric loading ? ESWT. The better outcomes were 11. Rompe JD, Furia JP, Maffulli (2008) Mid-portion Achilles
registered for the second group. tendinopathy: current options for treatment. Disabil Rehabil
Despite the positive results, this study has some limi- 30:1666–1676
12. Rasmussen S, Christensen M, Mathiesen I et al (2008) Shock-
tations. First of all, there was no patient randomization or wave therapy for chronic Achilles tendinopathy: a double-blind,
use of placebo for any treatment. Moreover, there was no randomized clinical trial of efficacy. Acta Orthop 79:249–256
control group and the analysis of outcomes was performed 13. Rodriguez-Merchan EC (2013) The treatment of patellar
without blind examiners. Our protocols were based on a tendinopathy. J Orthop Traumatol 14:77–81
14. Wang CJ, Ko JY, Chan YS et al (2007) Extracorporeal shockwave
3-session ESWT application that, in our opinion, represents for chronic patellar tendinopathy. Am J Sports Med 35:972–978
a reasonable approach, even if it is not the only approach. 15. Wang CJ, Wang FS, Yang KD et al (2006) Long term results of
Finally, the diagnosis of each pathology was established by extracorporeal shockwave treatment for plantar fascitis. Am J
all authors, even if the ESWT applications were performed Sports Med 34(4):592–596
16. Cacchio A, Rompe JD, Furia JP et al (2011) Shockwave therapy
by two dedicated surgeons. for the treatment of chronic proximal hamstring tendinopathy in
In conclusion, we believe that treatment with extracor- professional athletes. Am J Sports Med 39:146–153
poreal shockwaves may be a useful option in the manage- 17. Daecke W, Kusnierczak D, Loew M (2002) Long-term effects of
ment of several chronic musculoskeletal pathologies, extracorporeal shockwave therapy in chronic calcific tendinitis of
the shoulder. J Shoulder Elb Surg 11:476–480
particularly after the failure of a conventional approach. 18. Benazzo F, Marullo M, Zanon G et al (2013) Surgical manage-
The wide spectrum of applications, the low related costs, ment of chronic proximal hamstring tendinopathy in athletes: a 2
and its safety represent the strength of this modern approach to 11 years of follow-up. J Orthop Traumatol 14:83–89
to the treatment of disabling musculoskeletal diseases. 19. Wang CJ, Huang HY, Pai CH (2002) Shock wave enhances
neovascularization at the tendon-bone junction. J Foot Ankle
Surg 41:16–22
Conflict of interest The authors declare that they have no conflict

of interest.
Ethical standards The study conforms to the 1964 Helsinki dec-

laration and its later amendments; the study was approved by the
institutional ethical review board; all the patients provided informed
consent before being enrolled.

20. Lian O, Dahl J, Ackermann PW et al (2006) Pronociceptive and 29. Pettrone F, McCall BR (2005) Extracorporeal shock wave ther-
antinociceptive neuromediators in patellar tendinopathy. Am J apy without local anesthesia for chronic lateral epicondylitis.
Sports Med 34:1801–1808 J Bone Joint Surg Am 87:1297–1304
21. Ohtori S et al (2001) Shock wave application to rat skin induces 30. Staples MP, Forbes A, Ptasznik R et al (2008) A randomized
degeneration and reinnervation of sensory nerve fibres. Neuro- controlled trial of extracorporeal shock wave therapy for lateral
science 315:57–60 epicondylitis (tennis elbow). J Rheumat 35:2038–2046
22. Wang FS, Yang KD, Chen RF et al (2002) Extracorporeal shock 31. Spacca G, Necozione S, Cacchio A (2005) Radial shock wave
wave promotes growth and differentiation of bone-marrow stro- therapy for lateral epicondylitis: a prospective randomised con-
mal cells towards osteoprogenitors associated with induction of trolled single-blind study. Eur Medicophys 41:17–25
TGF-beta1. J Bone Joint Surg Br 84:457–461 32. Chung B, Wiley JP, Rose MS (2005) Long-term effectiveness of
23. Chen YJ, Wurtz T, Wang CJ et al (2004) Recruitment of mes- extracorporeal shockwave therapy in the treatment of previously
enchymal stem cells and expression of TGF-beta 1 and VEGF in untreated lateral epicondylitis. Clin J Sport Med 15:305–312
the early stage of shock wave-promoted bone regeneration of 33. Collins EDH, Jafarnia KK (2011) A clinical study of extracor-
segmental defect in rats. J Orthop Res 22:526–534 poreal shock waves (ESW) for treatment of chronic lateral epi-
24. Ioppolo F, Tattoli M, Di Sante L et al (2012) Extracorporeal condylitis. Curr Orthop Pract 22:185–192
shock-wave therapy for supraspinatus calcifying tendinitis: a 34. Cacchio A, Paoloni M, Barile A et al (2006) Effectiveness of
randomized clinical trial comparing two different energy levels. radial shock-wave therapy for calcific tendinitis of the shoulder:
Phys Ther 92:1376–1385 single-blind, randomized clinical study. Phys Ther 86:672–682
25. Constant CR, Murley AH (1987) A clinical method of functional 35. Gerdesmeyer L, Maier M, Loew M et al (2003) Extracorporeal
assessment of the shoulder. Clin Orthop Relat Res 214:160–164 shock wave therapy for the treatment of chronic calcifying ten-
26. Kitaoka HB, Alexander IJ, Adelaar RS et al (1994) Clinical rating donitis of the rotator cuff. JAMA 290:7–10
systems for the ankle-hindfoot, midfoot, hallux, and lesser toes. 36. Lakshmanan P, O’Doherty DP (2004) Chronic achilles
Foot Ankle Int 15:349–353 tendinopathy: treatment with extracorporeal shock waves. Foot
27. Dawson J, Doll H, Boller I et al (2008) The development and Ankle Surg 10:125–130
validation of a patient-reported questionnaire to assess outcomes 37. Rompe JD, Furia J, Maffulli N (2009) Eccentric loading versus
of elbow surgery. J Bone Joint Surg Br 90:466–473 eccentric loading plus shock-wave treatment for midportion
28. Ioppolo F, Rompe JD, Furia JP et al (2014) Clinical application achilles tendinopathy: a randomized controlled trial. Am J Sports
of shock wave therapy (SWT) in musculoskeletal disorders. Eur J Med 37:463–470
Phys Rehabil Med 50:217–230

17
Arthrodesis of proximal inter-phalangeal joint for hammertoe:
intramedullary device options
Matteo Guelfi1,4 • Andrea Pantalone1 • Janos Cambiaso Daniel2 • Daniele Vanni1 •
Marco G. B. Guelfi3 • Vincenzo Salini1
Abstract satisfaction. Radiological arthrodesis was achieved in

Background Proximal inter-phalangeal (PIP) joint art- 60.5–100 % of cases. Three of the publications compared
hrodesis today represents the standard treatment for struc- the new devices with the K-wire. Of these three articles,
tured hammertoes; however, recently, a lot of new two employed the traditional technique and one the buried
intramedullary devices for the fixation of this arthrodesis technique. The AOFAS score, evaluated in three publica-
have been introduced. The purpose of this work is to look tions, showed a delta of 19, 45 and 58 points. Major
at the currently available devices and to perform a review complications, which required a secondary surgical revi-
of the present literature. sion, were between 0 and 8.6 %. The complications of the
Materials and methods A literature search of PubMed/ K-wire and the new devices were similar; also the reop-
Medline and Google Scholar databases, considering works eration rate was close to equal (maximal difference 2 %).
published up until September 2014 and using the key- On the other hand, these kinds of devices definitely have a
words: hammertoe, arthrodesis, PIP joint, fusion, intrame- higher price, compared to the K-wire.
dullary devices, and K-wire, was performed. The published Conclusion The use of these new devices provides good
papers were included in the present study only if they met results; however, their high price is currently a problem.
the following inclusion criteria: English articles, arthrode- For this reason, cost-benefit studies seem to be necessary to
sis of PIP joints for hammertoes with new generation justify their use as standard treatment.
intramedullary devices, series with n [ 10. Studies using Level of evidence Level III systematic review.
absorbable pins or screws that are considered as another
kind of fixation that involved more than one articulation, as Keywords Hammertoe Arthrodesis PIP joint Fusion
well as comments, letters to the editor, or newsletters were Intramedullary devices Review K-wire
excluded.
Results Nine publications were included. Of the patients’
reports, 93–100 % were good or excellent concerning Introduction
Nowadays, the treatment of the hammertoe is still dis-

& Matteo Guelfi putable; indeed, a lot of procedures, both on the soft tissues
[email protected]
and the bone structures are purposed and considered effi-
1
Orthopaedic and Traumatology Division, G. d’Annunzio cient. In the rigid and structured deformities not suited for
University, Via dei Vestini 35, 66013 Chieti, Italy manual correction, arthrodesis of the proximal inter-pha-
2
Department of Plastic, Aesthetic and Reconstructive Surgery, langeal (PIP) joint represents the standard treatment [1].
Medical University of Graz, Augenbruggerplatz 29, This procedure is performed by removing the articular
8036 Graz, Austria surfaces of the proximal and intermediate phalanges. Many
3
Orthopaedic Division, Clinica Montallegro, systems such as cannulated screws or absorbable pins have
Via M.Te Zovetto 27, 16145 Genoa, Italy been designed for the fixation of the arthrodesis, yet still
4
Via Caprera 7/3, 16146 Genoa, Italy the K-wire is the traditional method, and most utilized [2–

6]. However, recently, new intramedullary devices have phalanges, and the other one on the middle, these are
been used persistently, trying to solve problems such as fixed together.
infections [3, 7], traumatic breaks [8, 9] and malalignments
The purpose of this work is to look at the currently
[10] tied to the K-wire.
available devices and to review, from the literature, the
As of today (September 2014), after accurate research,
results of these for PIP fusion.
16 different devices are available on the United States (US)
and European (EU) markets. These were divided into four
categories according to technical features and material
composition (Table 1).
– Shape memory devices: these are composed of a A literature search of PubMed/Medline and Google Scholar
memory metal (Memometal NiTinol), which is acti- databases, considering works published up until September
vated by body temperature, modifying its shape once 2014, and using the keywords: hammertoe, arthrodesis, PIP
implanted. Specifically, these become shorter and joint, fusion, intramedullary devices, and K-wire was per-
enlarge themselves to bestow more stability to the formed. The published papers were included in the present
system. study only if they met the following inclusion criteria:
– Bone allograft devices: since these devices are grafts, English articles, arthrodesis of PIP joints for hammertoes
they have bone inductive and conductive properties, with new generation intramedullary devices, series with
which improve their integration significantly. n [ 10. Studies using absorbable pins or screws that are
– One-piece solid or cannulated devices: thanks to the considered as another kind of fixation that involved more
form of their extremities, these can be anchored to the than one articulation, as well as comment, letter to editor or
cortical of the proximal and middle phalanges. The newsletters were excluded.
cannulated type also permits the use of the K-wire as a The search strategy identified over 455 articles. A total
guide. With these devices the proximal part is threaded of 43 publications describing specifically the arthrodesis of
and screwed onto the proximal phalange, while the the PIP joint for hammertoe could be identified.
distal part is anchored to the middle phalange. They are Thirty-four articles were excluded due to exclusion
available in steel, titanium or polyetheretherketone criteria, these being: studies using absorbable pins, screws
(PEEK). or other kinds of fixation (n = 23), fewer than ten patients
– Two-piece devices: a female and a male part make up (n = 9), non-English language (n = 1), comment (n = 1)
these devices. Once positioned, one on the proximal (Fig. 1).
Table 1 Intramedullary devices available on US and EU markets (up to September 2014)

Category Name Company Material No. of sizes available Plantar angle
Shape memory Smart ToeÒ II StrykerÒ Memometal Nitinol 6 ? 2 9 DIP 0°–10°

HammerlockÒ BMEÒ Memometal Nitinol 4 ? 1 9 DIP 0°–10°
One-piece solid or cannulated ProToe VOÒ WrightÒ Stainless steel 5 0°–10°
TM Ò
Arrow-lok Arrowhead Medical Stainless steel 8 0°–10°
Ipp OnÒ IntegraÒ Stainless steel 2 0°
ProxifuseTM CartivaÒ Nitinol and PEEK 1 0°
PhalinxÒ WrightTM Titanium 4 0°–10°a
TM Ò
Digifuse Metasurg Titanium 2 ? 1 9 DIP 0°–10°
Two Step Imp. Syst. Trilliant Surgical LTDÒ Titanium 3 0°
Ò
DuaFit In 2 Bones PEEK 4 0°–10°–17°a
Ò Ò
Toegrip Synchro Medical PEEK 5 0°–10°–20°
HammerFixÒ Extremity MedicalTM PEEK 3 0°
Bone allograft TenFuseÒ Solana SurgicalÒ Bone allograft 2 0°–10°
Two-piece StayfuseTM TornierÒ Titanium 3 Prox/6mid 0°
Ò Ò
Nextra Nextremity Solutions Titanium 2 Prox/3 mid 10°
Hat-TrickÒ Smith and NephewÒ PEEK 4 Prox/2 mid 0°–10°
a
10° and 17° angolated are solid, not cannulated

Arthrodesis of proximal inter-phalangeal joint for hammertoe: intramedullary device options 119
[10, 13], non union (1.5 %) [16], hardware failures

(3–5 %) [12, 16] and ruptures (5 %) [12].
The major complications, which required a surgical
revision, vacillate between 0 and 8.6 %. These complica-
tions were mainly due to malunion, breaks or recurrence.
In conclusion, only two works took the price of the
devices into consideration; Coillard et al. reported a 20
times higher price of these devices compared to the K-wire
[13]. Ellington et al., instead, reported a price of $225 per
device (StayFuseTM, Nexa Orthopaedics, San Diego, CA)
[14].
Discussion
Although hammertoe is a very frequent disease, the treat-

ment is still heavily disputed. In the structured deformities
not suited for manual correction, PIP fusion is considered,
Fig. 1 Flowchart of eligible study evaluation today, the standard treatment [4]. The K-wire technique is
the most utilized method for performing the fusion, as it is
fast, cheap and simple to implant [3]. On the other hand,
Nine articles, finally, met the inclusion criteria and were this kind of fixation method also has weak points: the
compatible with our review (Table 2) [5, 10–17]. In these exterior communication that predisposes for infections and
papers, patient satisfaction, achievement of arthrodesis, traumatisms, the violation of the distal inter-phalangeal
AOFAS score and the rate of complications were taken into (DIP) joint, the lack of compression and rotational control
consideration. and, finally, discomfort at removal [3, 7–10].
Because of this, the intramedullary devices aim to solve
the weak points of the K-wire technique. Indeed, the results
Results reported above seem to be slightly better than those of the
K-wire, especially regarding patient satisfaction and
The results from the nine articles included in the work are malalignment of the arthrodesis.
reported in Table 2. Considering everything, the type of complications
The satisfaction of the patient, taken into consideration reported for the new devices and the K-wire treatment have
in four publications, reports a good/excellent result in been similar, save the superficial infections. Taking into
93–100 % of the cases [11–13, 15]. consideration the major complications, in other words the
In contrast, radiological arthrodesis is achieved in cases which needed a reoperation in the articles that
60.5–100 % of the cases [4, 11]. This value result is directly compare the new devices to the K-wire, no dif-
heterogeneous and is barely correlated to the review cases, ferences were found [10, 11], or in any case no statistically
demonstrating the frequent establishment of a fibrous significant differences [17].
union. On the other hand, as reported by Ellington and Coillard,
Three publications compare the new devices to the the devices definitely have higher prices compared to the
K-wire: two of these use the traditional technique and one K-wire, which represents a limit to their utilization, espe-
the buried technique. The Angirasa et al. and Roukis et al. cially in the case of multiple toe corrections [13, 14].
publications report more satisfaction for the devices, yet Currently, no evidence exists in the literature which
none of these works cite any cases of revision [10, 11]. The justifies the use of these new devices, especially consid-
Scholl et al. [17] group, instead, reports no significant ering their high price. For this reason cost-benefit studies
difference of revisions utilizing K-wire with the buried are necessary to understand whether lower reoperation
technique (8.6 % against 10.7 % p = 0.754). rates can justify the use of these devices as the new stan-
The AOFAS score (evaluated in three publications) dard treatment in the future for hammertoes.
shows a delta of 19, 45 and 58 points [12, 13, 15]. Regarding reoperation, this can also result in difficulties,
Minor complications, often asymptomatic and radio- especially in the phase of the removal of the device, and cause
logically identified have been: malunion (2.4–7 %) [5, 10], an excessive reduction of the toe length. For this reason new
displaced fixation (1.5–13 %) [10], mallet toe (2–23 %) materials such as PEEK aim to make the revision easier.

surgery (%)
Of the 16 devices currently available on the US and EU
markets, as reported in Table 2, only four are also descri-
revision
Device
bed in the literature (according to the criteria previously
0.6
7.9
3.3
8.6
0
0
0
0
0
mentioned) and did not show significant differences in their
results. For the remaining devices, future studies are still
surgery (%) necessary.
revision
K-wire
In conclusion, the new intramedullary devices represent
10.7
NA
NA
NA
NA
NA
NA
0 an interesting topic because of their continuous evolution
0
and the constant birth of new devices on the market with
58.7 (22.9/81.6)
postoperative)
(preoperative/
45 (40.4/84.3) new characteristics and material compositions.

The use of these devices seem to provide good results;
19 (52/71)
variation
AOFAS
however, the dilemma tied to their high price is not neg-

NA
NA
NA
NA
NA
NA
ligible. For this reason, cost-benefit studies that are still
lacking in the literature seem necessary to justify the
Radiological
supremacy and the use of the new devices in the future as

arthrodesis
standard treatment for hammertoes.

83.8
60.5
93.8
68.9
(%)
100
NA
81
73
93

Acceptable
Conflict of interest The authors declare that they have no conflict

alignment
of interest. The authors did not receive any outside funding or grants
in support of their research for or preparation of this work. No
95.9
81.6
87.9
(%)
100
NA
NA
NA
NA
NA
commercial entity paid or directed, or agreed to pay or direct, any

benefits to any research fund, foundation, division, center, clinical
satisfaction
practice, or other charitable or non-profit organization with which the

Patients’
authors, or a member of their immediate families, are affiliated or

associated.
98.3
(%)
100
NA
NA
NA
NA
NA
98
95
Ethical standards This article does not contain any studies with
human participants performed by any of the authors.
Toes/no.
patients
156/117
150/140
58/NA
63/NA
13/13
42/24
38/27
30/10
65/35
NA not available, AOFAS American Orthopaedic Foot and Ankle Society Forefoot score
Smart ToeÒ
Smart ToeÒ
Smart ToeÒ
Smart ToeÒ
Smart ToeÒ
StayFuseÒ
StayfuseÒ
ProToeÒ
Ipp-OnÒ
Device
retrospective case series

prospective case series
References
case series
case series
case series
1. Zelen CM, Young NJ (2013) Digital arthrodesis. Clin Podiatr

Med Surg 30(3):271–282. doi:10.1016/j.cpm.2013.04.006
evidence
Level of
2. Caterini R, Farsetti P, Tarantino U, Potenza V, Ippolito E (2004)

Arthrodesis of the toe joints with an intramedullary cannulated
IV,
IV,
IV,
IV,
IV,
IV,
IV,
IV,
IV,
screw for correction of hammertoe deformity. Foot Ankle Int

Table 2 Results of the reviewed articles
25(4):256–261. doi:10.1177/107110070402500411
Angirasa et al. J Foot Ankle Surg [11]
3. Coughlin MJ, Dorris J, Polk E (2000) Operative repair of the

Scholl et al. J Foot Ankle Surg [17]
fixed hammertoe deformity. Foot Ankle Int 21(2):94–104. doi:10.

Sandhu et al. Foot Ankle Spec [16]
Roukis et al. Foot Ankle Spec [10]
Ellington et al. Foot Ank Int [14]
1177/107110070002100202
Coillard et al. Foot Ank Int [13]
Scott et al. Foot Ankle Spec [5]

Catena et al. Foot Ank Int [12]
Fazal et al. Foot Ank Int [15]
4. Ellington JK (2011) Hammertoes and clawtoes: proximal inter-

phalangeal joint correction. Foot Ankle Clin 16(4):547–558.
doi:10.1016/j.fcl.2011.08.010
5. Scott RT, Hyer CF, Berlet GC (2013) The PROTOE intrame-
dullary hammertoe device: an alternative to Kirschner wires. Foot
Ankle Spec. 6(3):214–216. doi:10.1177/1938640013487891
References
6. Smith BW, Coughlin MJ (2009) Disorders of the lesser toes. Sports

Med Arthrosc 17(3):167–174. doi:10.1097/JSA.0b013e3181a5cd26
7. Klammer G, Baumann G, Moor BK, Farshad M, Espinosa N (2012)
Early complications and recurrence rates after Kirschner wire

Arthrodesis of proximal inter-phalangeal joint for hammertoe: intramedullary device options 121
transfixion in lesser toe surgery: a prospective randomized study. 13. Coillard JY, Petri GJ, Van Damme G et al (2014) Stabilization of
Foot Ankle Int 33(2):105–112. doi:10.3113/FAI.2012.0105 proximal interphalangeal joint in lesser toe deformities with an
8. Konkel KF, Sover ER, Menger AG, Halberg JM (2011) Hammer angulated intramedullary implant. Foot Ankle Int 35(4):401–407.
toe correction using an absorbable pin. Foot Ankle Int doi:10.1177/1071100713519601
32(10):973–978. doi:10.3113/FAI.2011.0973 14. Ellington JK, Anderson RB, Davis WH, Cohen BE, Jones CP
9. Zingas C, Katcherian DA, Wu KK (1995) Kirschner wire (2010) Radiographic analysis of proximal interphalangeal joint
breakage after surgery of the lesser toes. Foot Ankle Int arthrodesis with an intramedullary fusion device for lesser toe
16(8):504–509. doi:10.1177/107110079501600809 deformities. Foot Ankle Int 31(5):372–376. doi:10.3113/FAI.
10. Roukis TS (2009) A 1-piece shape-metal Nitinol intramedullary 2010.0372
internal fixation device for arthrodesis of the proximal interpha- 15. Fazal MA, James L, Williams RL (2013) StayFuse for proximal
langeal joint in neuropathic patients with diabetes. Foot Ankle interphalangeal joint fusion. Foot Ankle Int 34(9):1274–1278.
Spec 2(3):130–134. doi:10.1177/1938640009336199 doi:10.1177/1071100713485545
11. Angirasa AK, Barrett MJ, Silvester D (2012) SmartToeÒ implant 16. Sandhu JS, DeCarbo WT, Hofbauer MH (2013) Digital
compared with Kirschner wire fixation for hammer digit correc- arthrodesis with a one-piece memory Nitinol intramedullary fix-
tive surgery: a review of 28 patients. J Foot Ankle Surg ation device: a retrospective review. Foot Ankle Spec
51(6):711–713. doi:10.1053/j.jfas.2012.06.013 6(5):364–366. doi:10.1177/1938640013496458
12. Catena F, Doty JF, Jastifer J, Coughlin MJ, Stevens F (2014) 17. Scholl A, McCarty J, Scholl D, Mar A (2013) Smart toeÒ implant
Prospective study of hammertoe correction with an intramedul- versus buried Kirschner wire for proximal interphalangeal joint
lary implant. Foot Ankle Int 35(4):319–325. doi:10.1177/ arthrodesis: a comparative study. J Foot Ankle Surg
1071100713519780 52(5):580–583. doi:10.1053/j.jfas.2013.02.007

18
FiberWire tension band for patellar fractures
Lawrence Camarda1 • Alessandra La Gattuta1 • Marcello Butera1 •
Francesco Siragusa1 • Michele D’Arienzo1
Abstract Keywords Patellar fracture Knee fracture FiberWire

Background Symptomatic hardware represents the most Tension band Suture
frequent complication reported following surgical treat-
ment of patellar fracture. For this reason, some authors
suggested using nonabsorbable sutures to fix the fracture Introduction
with various techniques. The aim of this study was to
evaluate clinical and radiological results of patients treated Patellar fractures are relatively uncommon, representing
following a modified Pyriford technique using a FiberWire approximately 1% of all skeletal injuries [2]. Nonsurgical
suture (Arthrex, Naples, FL, USA). management is recommended for fractures with intact
Materials and methods We retrospectively evaluated a extensor mechanism, minimal intra-articular stepoff, and
case series of seventeen patients with displaced patellar minimal fracture displacement (1–4 mm) [3]. An incom-
fractures treated by open reduction and internal fixation petent extensor mechanism associated with a displaced or
with a modified tension band using FiberWire sutures. comminuted fracture with a torn extensor retinaculum is
Clinical and radiological outcome were evaluated. Union indicated for surgery. Operative treatment must provide
time, complications, and reoperation rate were observed stable patellar fracture fixation to allow early mobilization
and recorded. and prevent fracture displacement. Furthermore, articular
Results All fractures healed (time to union congruity is essential to reduce the increased risk of post-
9.2 ± 2 weeks), and no fixation failure was observed. traumatic osteoarthritis as a result of the high-contact for-
Slight losses of reduction (\4 mm) were noted in two ces in the patellofemoral joint [5].
patients at 4 weeks postoperatively. The average Lysholm Several different techniques for internal fixation have
and Bostman scores at the final follow-up were 91 ± 5.7 been proposed and employed over the years, with different
(range 83–100) and 28.3 ± 1.6 (range 26–30), respectively. rates of success. Historically, the most commonly used
Conclusion Modified tension band using FiberWire technique for managing patellar fracture fixation is repre-
sutures showed satisfactory clinical results, with a low sented by the modified tension-band wiring technique [5,
incidence of complications and reoperations. FiberWire 13, 19, 23]. It involves longitudinal Kirschner wires (K-
tension bands could be used in place of metal-wire tension wires) and 18-gauge stainless steel wire in a figure-of-eight
bands to treat patellar fracture, reducing the rate of pattern looped over the anterior side of the patella. This
symptomatic hardware. technique neutralizes tension forces anteriorly produced by
Level of evidence 4 the extensor mechanism at knee flexion and converts them
into stabilizing compressive forces at the articular surface
[23]. This construct represents the most widely used
& Lawrence Camarda method of fixation for transverse and comminuted patellar
[email protected]
fracture. Over the years, this technique was further modi-
1
Department of Orthopaedic Surgery, DICHIRONS, fied by different authors using either K-wires or cannulated
University of Palermo, Via del Vespro, 90100 Palermo, Italy screws with different stainless steel wire configurations.

FiberWire tension band for patellar fractures 123
Even though substantially good clinical results have been patients presented and open fracture; one patient pre-
reported, all techniques using metallic wires can result in sented with polytrauma that required long hospitalization,
symptomatic hardware, which represents the most frequent compromising the rehabilitation program. Twenty patients
complication, with reported rates varying from 0 % to met inclusion criteria. All patients were contacted by
60 % [12, 14, 20]. In fact, implants can irritate overlying telephone and/or email and were asked to return to the
soft tissue and cause pain, requiring a second surgery to hospital for a final physical examination. In order to
remove the implant. In addition, a high incidence of maximize follow-up, Internet-based searches was used to
infection has been reported [9, 12]. For this reason some contact patients who did not respond to the first mailing.
authors have advocated the use of nonabsorbable sutures, Three patients were lost to follow-up: one patient refused
such as braided polyester, to fix the fracture when using the final clinical assessment because she was living
various techniques [6, 7, 9, 12, 21]. Several advantages of abroad; two patients could not be located. For all patients,
using a suture over a wire have been shown in the litera- medical records and X-ray films were reviewed and
ture, such as lower rate of revision surgery and higher fracture types classified according to the Orthopaedic
patient tolerance of surgical material that does not irritate Trauma Association fracture classification (AO/OTA)
soft tissues as much as wire. Another advantage is repre- system.
sented by easier handling of suture materials for the sur- In all patients, surgery was scheduled 2–3 days after tje
geon, which could determine shorter operating and trauma and was performed by the senior author. Patients
tourniquet time [9, 11]. Few authors reported data con- were operated on in a supine position, and a tourniquet was
cerning the use of tension-band fixations employing non- placed high up on the thigh. In all cases, a longitudinal skin
absorbable sutures. Although no. 5 Ethibond (Ethicon, incision on the knee was performed. An extemporary
Somerville, NJ, USA) and no. 5 Ti-Cron (Davis and Geck, reduction was performed with a reduction clamp and
Gosport, Hampshire, UK) braided polyester sutures have checked under fluoroscopy. In this phase, care was taken to
been clinically and biomechanically studied for patellar obtain the best possible reduction and congruence of the
fracture fixation, to our knowledge there are no data on the articular patellar surface. At this point, a modified Pyriford
use of FiberWire sutures (Arthrex, Naples, FL, USA). technique was used for fracture fixation [8]. A no. 5
The aim of this study was to evaluate clinical outcome FiberWire nonabsorbable suture was used, and a peri-
and rate of postoperative complications of patients who patellar circumferential cerclage was performed in a purse-
underwent an open reduction and internal fixation (ORIF) string fashion close to the bone. This allowed initial frac-
using a tension-band wiring technique with a FiberWire ture stabilization. A second no. 5 FiberWire suture was
suture. The authors hypothesized that a FiberWire tension then placed in a figure-of-eight fashion through the
band could be used for patellar fracture fixation, decreasing quadriceps and patellar tendon to obtain a modified anterior
the rate of symptomatic hardware without reducing the rate tension band. Each of the two FiberWire bands was man-
of fracture healing. ually tensioned and knotted on the superior board of the
patella (Fig. 1). Additionally, retinacular defect was
repaired with an absorbable suture.
Materials and methods All patients were immobilized postoperatively with a
long-hinged knee brace locked at 0° for 3 weeks. Weight
We performed a retrospective study of consecutive patients bearing started as tolerated with crutches, while full weight
hospitalized for patellar fractures in our Institution bearing was allowed from the fourth week postoperatively.
(Orthopaedic and Traumatology Unit, AOUP ‘‘P.Giac- Passive range of motion (ROM) began at the third week
cone’’, Palermo, Italy) from 2008 to 2013. Inclusion cri- postoperatively and was limited at 0–90° for the next
teria for the study were the use of FiberWire suture for 2 weeks, allowing active flexion/extension; full ROM was
fracture fixation, age C18 years, and minimum follow-up allowed from the sixth week postoperatively.
of at least 10 months. Exclusion criteria were open frac- Anteroposterior and lateral radiographs of the knee
tures, stainless steel wire implantation, previous ipsilateral were obtained every 2–4 weeks until achievement of
knee surgery, and patients with polytrauma or head injuries bony union and again at final follow-up. Details of
that definitely influence rehabilitation. mechanism of injury, associated injuries, timing of sur-
Fifty-one patients with patellar fractures were admitted gery, time to fracture healing, postoperative complica-
to the hospital during the mentioned period. Among them, tions, final knee ROM, and implant removal were
37 required ORIF. Seventeen patients were excluded: recorded. Time requested to regain normal daily activity
specifically, eight were treated with a suture other than was also scheduled. In addition, knee function was
the FiberWire; in six patients, a hybrid technique was evaluated according to Bostman and Lysholm/Tegner
used (metallic K-wires and nonabsorbable suture); two scores [1].

Fig. 1 FiberWire tension band.

Peripatellar circumferential
cerclage (a); modified anterior
tension band (b); final construct
(c)
Results Discussion
Of 20 patients, 17 were available for clinical and radio- The most important finding of our study was that Fiber-
logical follow-up (12 men and 5 women). The main Wire tension-band technique showed positive clinical
mechanisms of injury were fall from height on hard outcomes with low implant-related complication rates. Just
ground in ten cases (59 %) and road traffic accidents in one of 17 patients required elective implant removal. No
the remaining seven (41 %). Average patient age at injury other patient complained of symptomatic hardware. Fur-
was 46.6 years (range 20–75). There were nine transverse ther, all fractures healed, and no fixation failure was noted.
fractures, six comminuted, and two of the inferior pole. Different methods have been described for treating
The average follow-up was 33 months (range 8–68). No patellar fractures. Traditionally, transverse fractures are
patient had significant flexion contracture. Average ROM best treated with a tension-band technique using two axial
was 131.1° for flexion (range 120°–140°) and 0.5° for K-wires, with a figure-of-eight wire placed anteriorly [18].
extension (range 0°–3°). No significant ROM differences This technique could be combined with the use of screws in
were noted with the uninjured contralateral knee. Average case of more comminuted fracture types [5]. However,
Bostman scores at 3 months postoperatively and final stainless steel wire is a difficult material to manipulate
follow-up were 25.2 ± 2 (range 20–30) and 28.3 ± 1.6 through tissues and may result in poor fracture fixation [6].
(range 26–30), respectively. Further, the mean Lysholm Further, high rates of metal-implant-related complications
score at final follow-up was 91 points ±5.7 (range have been reported, such as K-wire migrationand skin
83–100). irritation due to prominent wire knots, a very common
All fractures healed (time to union 9.2 ± 2 weeks) and complication, ranging from 0 % to 60 % [4, 17], possibly
no fixation failure was observed in the group studied. necessitating metal implant removal. In a case series of 27
Slight losses of reduction (\4 mm) were noted in two patients, LeBrun et al. reported a hardware removal rate of
patients at 4 weeks postoperatively [19, 22]. Because of approximately 52% at a mean of 6.5 years of follow-up
noncompliance with the postoperative rehabilitation pro- [15]. Recently, Hoshino et al. performed a retrospective
tocol, one patient presented knee stiffness at 2 months study of surgically treated patellar fractures. In this study,
postoperatively that required gentle manipulation under elective implant removal was performed in 37 % and 23 %
anesthesia (Fig. 2). The same patient was the only one in of patients treated respectively with K-wires and cannu-
the study to requrie elective FiberWire removal lated screws [10].
24 months after surgery. This was performed secondary to On the basis of these results, some authors have advo-
a superficial infection, which did not affect the final cated the use of nonmetallic implants, such as nonab-
clinical outcome (ROM 0–125°). Another patient under- sorbable suture and biodegradable cannulated screws [6, 7,
went knee arthroscopy in another Institution because of 9, 11, 21, 25]. Many of these studies reported good out-
anterior knee pain. No patients referred localized pain comes for patellar fractures, with very low rate of post-
deriving from prominent suture knots (Fig. 3). Demo- operative complications. However, the concept of
graphics, fracture type, and outcomes are presented in replacing stainless steel wire with sutures remains contro-
Table 1. versial in the literature, mainly due to uncertainty regarding

Fig. 2 Comminuted patellar fracture. Preoperative X-ray (a); postoperative X-ray (b); final follow-up X-ray (c)
Fig. 3 Transverse patellar fracture. Preoperative X-ray (a); postoperative X-ray (b); final follow-up X-ray (c)
rigid fracture fixation. For this reason, some authors eval- rotator cuff repairs, and Achilles tendon repairs. FiberWire
uated the biomechanical properties of sutures used for was evaluated biomechanically for tension-band fixation of
patellar fractures. Chatakondu et al. found a significantly a transverse patellar fracture: Using a three-point-bend
lower tensile strength for Ti-Cron sutures compared with model, Wright et al. observed that a double-strand Fiber-
stainless steel wire (14.80 vs 34.91 kg) [6]. McGreal et al. Wire presented a significantly higher failure load than
reported that braided polyester suture is an acceptable stainless steel wire. Furthermore, it was observed that,
alternative to wire in tension-band fixation after testing unlike stainless steel, FiberWire maintained its initial
cadaveric patellae for [20,000 cycles of knee flexion and stiffness until failure [24]. This was confirmed by our
extension [16]. Also, Patel et al., comparing different study: Using a no. 5 FiberWire tension band, we found no
techniques including polyester sutures (no. 5 Ethibond), significant fracture displacement following knee mobi-
concluded that braided polyester sutures were comparable lization. Only two patients presented slight losses of
to stainless-steel wire for transverse patellar in terms of reduction (\4 mm). This could be justified by progressive
quality of fixation [19]. The FiberWire represents a par- adhesion and adjustment of the suture through the peri-
ticular suture characterized by a core of several small, patellar tissue that could be present during simple load,
individual strands of ultra-high-molecular-weight poly- such as quadriceps muscle contraction. Further, high losses
ethylene covered with braided polyester suture material. It of reduction ([4 mm) and synthesis failure, which are
is used in a variety of orthopedic procedures, such as potentially linked to FiberWire failure or breakage, were
quadriceps/patellar tendon repairs, ACL reconstruction, not noted in our case series. In addition, all treated fractures

Table 1 Demographic data, clinical outcomes, and postoperative complications of patients who underwent open reduction and internal fixation
using a tension-band-wiring technique with a FiberWire suture
Patient Age Type of patellar Follow-up Bostman Lysholm Complication
(years) fracture (months) scores scores
1 40 Transverse 68 26 83 –
2 36 Transverse 66 30 90 –
3 43 Inferior pole 52 30 99 –
4 64 Comminuted 51 27 89 Slight losses of reduction (\4 mm)
5 49 Comminuted 36 26 85 Slight losses of reduction (\4 mm)
6 44 Comminuted 35 30 95 –
7 71 Comminuted 22 30 100 –
8 30 Transverse 37 30 100 –
9 22 Comminuted 24 26 83 Knee stiffness ? FiberWire removal
10 51 Inferior pole 8 30 90 –
11 60 Transverse 9 28 86 –
12 75 Comminuted 12 26 85 –
13 48 Transverse 10 28 94 –
14 20 Transverse 40 28 89 Anterior knee pain
15 40 Transverse 20 27 90 –
16 65 Transverse 32 30 99 –
17 33 Transverse 43 29 95 –
healed at 3 months postoperatively. The only patient ethical standards of the 1964 Declaration of Helsinki as revised in
requiring FiberWire removal was a 22-year-old man who 2000.
was noncompliant with the postoperative rehabilitation Conflict of interest The authors declare that they have no conflict
protocol. In fact, this patient required gentle manipulation of interest related to the publication of this manuscript.
under anesthesia because of knee stiffness at 2 months
postoperatively. Further, the same patient underwent elec-
tive FiberWire removal secondary to a superficial infection,
probably due to implant-related soft-tissue irritation.
On the basis of the results of this study, we found that
FiberWire tension bands can be used in place of metal-wire
tension bands to treat patellar fracture, reducing the rate of
symptomatic hardware.
Our study has several weaknesses. Major limitations are
References
its retrospective nature and small size. However, our group 1. Böstman O, Kiviluoto O, Nirhamo J (1981) Comminuted dis-
was similar in size to other studies reporting results of placed fractures of the patella. Injury 13:196–202
patellar fractures. Further, we believe that high-quality 2. Bostrom A (1972) Fracture of the patella. A study of 422 patellar
prospective, randomized studies are required to define the fractures. Acta Orthop Scand Suppl 143:1–80
3. Carpenter JE, Kasman R, Matthews LS (1993) Fractures of the
effectiveness of nonabsorbable suture for internal fixation patella. J Bone Joint Sur Am 75:1550–1561
of patellar fracture. 4. Carpenter JE, Kasman R, Matthews LS (1994) Fractures of the
In conclusion, the study demonstrates that a modified patella. Instr Course Lect 43:97–108
tension band using FiberWire sutures showed satisfactory 5. Carpenter JE, Kasman RA, Patel N, Lee ML, Goldstein SA
(1997) Biomechanical evaluation of current patella fracture fix-
clinical results with low incidence of complications and ation techniques. J Orthop Trauma 11:351–356
reoperations. Thus, FiberWire tension bands could be 6. Chatakondu SC, Abhaykumar S, Elliott DS (1998) The use of
considered an alternative solution for treating patellar nonabsorbable suture in the fixation of patellar fractures: a pre-
fracture, thus reducing the rate of symptomatic hardware. liminary report. Injury 29:23–27
7. Chen CH, Huang HY, Wu T, Lin J (2013) Transosseous suturing
of patellar fractures with braided polyester—a prospective cohort
Ethical standards All patients gave informed consent before
with a matched historical control study. Injury 44:1309–1313
inclusion into the study. The study was authorized by the local
8. Curtis MJ. Internal fixation for fractures of the patella (1990) A
ethical committee and was performed in accordance with the
comparison of two methods. J Bone Joint Surg Br 72:280–282

9. Gosal HS, Singh P, Field RE (2001) Clinical experience of 18. Muller ME, Allgower M, Schneider R, Willenegger H (1979)
patellar fracture fixation using metal wire or nonabsorbable Manual of internal fixation. Techniques recommended by the
polyester—a study of 37 cases. Injury 32:129–135 AO-ASIF group. Springer, Berlin, pp 248–253
10. Hoshino CM, Tran W, Tiberi JV et al (2013) Complications 19. Patel VR, Parks BG, Wang Y, Ebert FR, Jinnah RH (2000)
following tension-band fixation of patellar fractures with cannu- Fixation of patella fractures with braided polyester suture: a
lated screws compared with kirschner wires. J Bone Joint Surg biomechanical study. Injury 31:1–6
Am 95:653–659 20. Petrie J, Sassoon A, Langford J (2013) Complications of patellar
11. Hughes SC, Stott PM, Hearnden AJ, Ripley LG (2007) A new fracture repair: treatment and results. J Knee Surg 26:309–312
and effective tension-band braided polyester suture technique for 21. Qi L, Chang C, Xin T et al (2011) Double fixation of displaced
transverse patellar fracture fixation. Injury 38:212–222 patella fractures using bioabsorbable cannulated lag screws and
12. Hung LK, Chan KM, Chow YN, Leung PC (1985) Fractured braided polyester suture tension bands. Injury 42:1116–1120
patella: operative treatment using the tension band principle. 22. Smith ST, Cramer KE, Karges DE, Watson JT, Moed BR (1997)
Injury 16:343–347 Early complications in the operative treatment of patella frac-
13. John J, Wagner WW, Kuiper JH (2007) Tension-band wiring of tures. J Orthop Trauma 11:183–187
transverse fractures of patella. The effect of site of wire twists and 23. Weber MJ, Janecki CJ, McLeod P, Nelson CL, Thompson JA
orientation of stainless steel wire loop: a biomechanical investi- (1980) Efficacy of various forms of fixation of transverse frac-
gation. Int Orthop 31:703–707 tures of the patella. J Bone Joint Surg Am 62:215–220
14. Kumar G, Mereddy PK, Hakkalamani S, Donnachie NJ (2010) 24. Wright PB, Kosmopoulos V, Cote RE, Tayag TJ, Nana AD
Implant removal following surgical stabilization of patella frac- (2009) FiberWire is superior in strength to stainless steel wire for
ture. Orthopedics 12:33 tension band fixation of transverse patellar fractures. Injury
15. LeBrun CT, Langford JR, Sagi HC (2012) Functional outcomes 40:1200–1203
after operatively treated patella fractures. J Orthop Trauma 25. Yotsumoto T, Nishikawa U, Ryoke K, Nozaki K, Uchio Y (2009)
26:422–426 Tension band fixation for treatment of patellar fracture: novel
16. McGreal G, Reidy D, Joy A, Mahalingam K, Cashman WF technique using a braided polyblend sutures and ring pins. Injury
(1999) The biomechanical evaluation of polyester as a tension 40:713–717
band for the internal fixation of patellar fractures. J Med Eng
Technol 23:53–56
17. Melvin JS, Mehta S (2011) Patellar fractures in adults. J Am
Acad Orthop Surg 19:198–207

19
The role of femoral offset and abductor lever arm in total hip
arthroplasty
Filip Bjørdal1 • Kristian Bjørgul2
Abstract Level of evidence Level 3, prospective cohort study.

Background In order to create a well-functioning total
hip arthroplasty (THA), it is important to restore femoral Keywords Hip arthroplasty Minimally invasive hip
off-set and thus the abductor lever arm. The aim of this arthroplasty Femoral off-set Uncemented HOOS
study was to investigate the clinical effect of increasing the Harris Hip Score
abductor lever arm to and beyond the anatomical native
lever arm in minimally invasive total hip arthro-
plasty performed through a direct anterior approach. Introduction
Materials and methods We compared the lever arm of the
operated hip to the lever arm of the contralateral native hip on Total hip arthroplasty (THA) is a well-established treat-
radiographs in 148 patients following THA. The patients were ment in patients suffering from arthritic disease of the hip,
divided in two groups based on whether they kept their reducing pain and improving function [1]. By replacing the
anatomical lever arm or had an increased lever arm. The degenerative joint with a prosthetic stem and cup, one
clinical outcome was assessed using hip osteoarthritis outcome seeks to restore the normal anatomy of the joint, but several
score (HOOS), Harris hip score and UCLA activity score. controversies remain regarding the optimal placement of
Results Patients who kept their anatomical lever arm did the components [2–4].
not experience a significantly better clinical outcome than In order to restore the optimal biomechanical forces of
the patients with an increased abductor lever arm. We the joint, the acetabulum may be medialized, thus
found no significant difference in clinical scores at any of reducing the distance between the center of rotation and
the follow-ups during the first year after THA. the body axis [5], which provides better mechanical
Conclusion The results of this study suggest that an conditions for the abductor muscles of the hip [6]. By
increase in the abductor lever arm does not have major medializing the cup there is a risk of reducing the global
effects on the clinical outcome after THA. To avoid the offset. It is therefore considered important to compensate
potential negative effects of decreasing the lever arm, the with an equivalent increase in the femoral offset to ensure
surgeon should aim for an equal or slightly increased lever the biomechanical benefits [6–8]. However, when
arm. increasing the femoral offset, there is an inherent risk of
exaggerated compensation, which may lead to increased
tension on the abductor muscles and possibly pain and
& Kristian Bjørgul
[email protected]
reduced function. We have not been able to find any lit-
erature investigating the clinical consequences of an
Filip Bjørdal
[email protected]
overcorrection of the femoral offset.
The aim of this study was to investigate any correlation
1
University of Oslo, Hollandveien 26, 1555 Son, Norway between a change in lever arm of the abductor muscles and
2
Østfold Hospital Trust, Aleris Health Oslo, Chr. Svendsens clinical outcome, including the possible consequences of
gate 6, 1771 Halden, Norway an exaggerated offset. To clarify this aspect we

The role of femoral offset and abductor lever arm in total hip arthroplasty 129
investigated whether there were any differences in clinical

outcome between patients who had an increase in lever arm
compared to patients who kept their anatomical lever arm.
During 2010 we performed THA in 166 patients using the

direct anterior approach to the hip through the Smith-Peter-
sen interval. Of these, 148 were included in our study group;
15 of the 166 patients were excluded due to previous con-
tralateral hip arthroplasty, and 3 were excluded due to a
decrease in the abductor lever arm (ALA) beyond 5 mm. All
patients were followed and assessed with Harris hip score,
UCLA activity score and hip disability and osteoarthritis
outcome score (HOOS) with the added dimensions walking
ability and recreational ability. Evaluations were made after Fig. 1 Radiograph demonstrating the abductor lever arm (ALA),
6 weeks, 4 months and 1 year postoperatively. defined as the distance from the center of rotation to the line of action
of the abductor muscles
HOOS is a patient-administered questionnaire that
consists of five subscales (pain, symptoms, activity of daily
living, sport and recreation, function, and hip-related contralateral native hip. Group 1 consisted of patients with
quality-of-life). Each question was answered using a Likert a lever arm restored to within 5 mm of the native lever
scale from 0 to 4 points and a score was calculated for each arm, while group 2 comprised patients with a lever arm that
subscale, where 100 indicate no symptoms and 0 represents was increased to greater than 5 mm of the native lever arm.
extreme symptoms [9]. The two groups were compared in regards to all parameters
The UCLA activity score is a scale ranging from 1 to 10, of Harris hip score and HOOS.
where 1 indicates inactivity and 10 the highest level of activity.
The THA was performed through the anterior approach Statistical analysis
on a fracture table. The method has been described thor-
oughly by several authors [10, 11]. All patients were Statistical analysis was performed using Microsoft Excel.
mobilized on the day of surgery. We recommended partial Data were expressed as mean ± standard deviation (SD).
weight bearing as needed and did not impose any restric- Comparisons were made using the unpaired Student’s t test.
tions on activities or range of motion. The implants used A P value of less than 0.05 was considered to be significant.
were an SL-PLUS MIA stem and a REFLECTION press fit
cup (Smith and Nephew, Memphis, TN, US). The SL-
PLUS MIA stem was available in a high offset version as Results
well as in a normal offset version. In this study, we used
exclusively the high offset stem, which has a CCD angle of The study population consisted of 51 men and 97 women
123°. The standard stem has an angle of 131°, and the with a mean age of 67.7 ± 10.9 years. Mean body mass
difference in femoral offset between the stems is 8 mm index was 27.0 ± 4.3 (Table 1). An analysis of Harris Hip
when a size 6 stem is used with a neutral head. Score and HOOS preoperatively did not show any statis-
tically significant differences between the two groups.
Radiological assessment
Radiological result
A standardized anteroposterior pelvic and hip radiograph
was performed in all patients following THA. The ALA In our sample we found a native ALA of 58.0 ± 6.6 mm,
was defined as the distance from the center of the hip joint whereas the mean lever arm of the operated side was
to the line of action of the abductor muscles (Fig. 1) [12]. 65.4 ± 5.9 mm.
The lever arm and the line of action of the abductor should Group 1 consisted of 56 patients with a mean native
form a 90° angle. The lever arm was then measured in the ALA of 61.6 ± 6.1 mm. The mean lever arm of the
contralateral hip and compared to the operated side. The operated side was 63.0 ± 5.4 mm; 17 of the patients in this
patients were divided into two groups based on the dif- group experienced a shortening of the lever arm, whereas
ference in ALA between the operated hip and the 34 had an increase. Five patients did not experience a

Table 1 Patient demographics

Study population Group 1 Group 2 P value
ALA increase/decrease B5 mm ALA increase [5 mm
Number of patients 148 56 92

Gender (male/female) 51/97 27/29 24/68
Age (years)a 67.7 ± 10.9 66.2 ± 13.0 68.6 ± 9.3 0.23
Body mass index (kg/m2)a 27.0 ± 4.3 27.2 ± 4.5 26.8 ± 4.3 0.57
Preoperative clinical scoring
Harris Hip Score 47.4 ± 18.1 46.4 ± 16.7 48.7 ± 18.6 0.45
HOOS––pain 35.7 ± 16.9 33.2 ± 16.0 37.3 ± 18.0 0.18
HOOS––symptom 40.6 ± 17.9 38.0 ± 18.1 42.2 ± 18.3 0.2
HOOS––ADL 36.7 ± 16.8 34.9 ± 17.3 38.2 ± 17.2 0.29
HOOS––sport/recreation 20.2 ± 18.9 17.9 ± 16.7 21.5 ± 20.6 0.28
HOOS––quality of life 27.8 ± 13.6 25.1 ± 11.9 29.1 ± 14.8 0.09
HOOS––activity 1a 2.7 ± 1.2 2.6 ± 1.0 2.7 ± 1.3 0.81
HOOS––activity 1b 2.6 ± 1.2 2.8 ± 1.2 2.6 ± 1.2 0.21
HOOS––activity 2 3.7 ± 2.0 3.5 ± 1.9 3.9 ± 2.0 0.19
ALA abductor lever arm, HOOS Hip disability and osteoarthritis outcome score
a
Values are expressed as mean ± SD
difference in lever arm between the two hips (Fig. 2a). The of the abductors and increased stability due to increased
mean difference in lever arm between the contralateral soft tissue tension [6, 12, 13]. Failure to restore offset has
native hip and the operated hip was 1.4 ± 3.12 mm. been associated with increased joint reactive force and
Group 2 comprised 95 patients with a mean native lever hence an increase in polyethylene wear [14–16]. However,
arm of 55.8 ± 5.9 mm. The mean lever arm of the oper- Little et al. [17] suggested that an increase beyond 5 mm of
ated side was 66.9 ± 5.8 mm. These patients had a mean the contralateral hip might also result in increased poly-
increase in the lever arm of 11.2 ± 4.3 mm (range ethylene wear.
6–28 mm) (Fig. 2b). Although the importance of femoral offset in THA has
been emphasized in several studies, there is limited research
Clinical outcome directly investigating the role of the abductor lever arm and
its effect on clinical outcome. Studies have reported a
Patients whose lever arm was restored to within 5 mm of correlation between the ALA and abductor muscle strength.
the contralateral native hip did not experience a signifi- McGrory et al. [12] reported that ALA length was among
cantly better clinical outcome than the patients with a the most important factors influencing abductor muscle
greater postoperative increase in lever arm (Fig. 3). After strength. Using a 3-dimensional biomechanical model, Delp
1 year of follow-up there were still no statistically signif- et al. [8] demonstrated that lateral displacement of the hip
icant differences in any parameters of HOOS or Harris hip center adversely affected the function of the abductor
score between the two groups (Table 2). muscles by decreasing the lever arm, thereby decreasing the
capacity to generate hip abduction moments. Recently,
Terrier et al. [18] found that the benefits of cup medial-
Discussion ization varies according to individual patient anatomy and
stated that medialization should be balanced against possi-
Our data showed no significant difference in clinical out- ble disadvantages such as increased bone loss.
come between the two groups at any of the follow-ups Our study provides clinical data that enables us to
during the 1st year after operation. This suggests that a investigate how a change in lever arm affects the outcome
change in ALA does not have a large impact on the clinical after THA in a clinical setting where the surgery was
outcome as measured by HOOS or Harris hip score during performed by two surgeons using the direct anterior
the 1st year after THA. approach through the Smith-Petersen interval in every case.
There is evidence that offset plays an important role The same types of implants were used in all patients.
when it comes to the clinical result following THA. Several The radiological assessments were made using digital
studies have documented that an increase in offset results images from our database, enabling the radiologist to use
in increased range of motion, better mechanical advantage measurement tools with high degree of precision.

Fig. 2 a, b Change in offset in

patient groups 1 (a) and 2 (b).
Each patient is represented by
one bar
Furthermore, all measurements were performed by the Although both HOOS and Harris hip score have shown a
same investigator (J.B.), which eliminated interobserver high degree of validity, it is possible that these instruments
variability. Intraobserver variability was not assessed. are not sensitive enough to demonstrate an underlying
There are some limitations to our study. The patients difference in clinical outcome between the groups [19, 20].
were only followed for 1 year postoperatively. It is possi- In our study population only 17 out of 148 patients
ble that more time is required to demonstrate a difference experienced a shortening of the ALA. Several studies have
in clinical outcome. Another limitation may be that we did reported that a shortening of the lever arm may result in
not perform an intra-observer validation study. weakness of the abductor muscles and reduced stability [6,
It is also possible that the instruments used to score the 8, 12, 21]. It is possible that a higher frequency of patients
clinical outcome in our study lack sufficient sensitivity to with a decreased lever arm would have had a larger impact
demonstrate a significant difference between the groups. on the clinical scores.

Fig. 3 Error bars showing

95 % confidence intervals for
the mean of hip disability and
osteoarthritis outcome score
(HOOS) subgroups and Harris
hip score among group 1
(circles) and group 2
(diamonds). The two groups
display overlap in all clinical
parameters
Table 2 Clinical outcome

Group 1 Group 2 P value
1 year after total hip
ALA increase/decrease ALA increase
arthroplasty (THA)
B5 mma [5 mma
Mean ± SD Mean ± SD
HOOS––pain 86.0 ± 19.0 91.3 ± 12.6 0.16

HOOS––symptom 86.7 ± 18.9 90.0 ± 13.0 0.37
HOOS––ADL 85.1 ± 20.1 87.8 ± 16.0 0.5
HOOS––sport/recreation 74.5 ± 27.5 73.4 ± 22.7 0.85
HOOS––quality of life 78.1 ± 25.0 79.4 ± 21.9 0.81
HOOS––activity 1a 3.7 ± 1.8 3.9 ± 1.5 0.69
HOOS––activity 1b 4.5 ± 1.5 4.2 ± 1.4 0.48
HOOS––activity 2 5.9 ± 2.4 5.8 ± 2.3 0.84
Harris hip score 94.1 ± 9.7 94.4 ± 10.6 0.86
a
Values are expressed as mean ± standard deviation
The results of this study suggest that patients who pre-

serve their anatomical ALA do not experience a signifi-
cantly better clinical outcome than patients that have their
lever arm increased. When considering the potential dis-
advantages of decreasing the lever arm, the surgeon should
aim for an equal or slightly increased lever arm during
THA.
Conflict of interest None. References
Ethical standards Patients gave the informed consent prior to 1. Rorabeck CH, Bourne RB, Laupacis A et al (1994) A double-
being included in the study. The study was authorized by the local blind study of 250 cases comparing cemented with cementless
ethics committee and was performed in accordance with the Ethical total hip arthroplasty. Cost-effectiveness and its impact on health-
standards of the 1964 Declaration of Helsinki as revised in 2000. related quality of life. Clin Orthop Relat Res 298:156–164

2. Karachalios T, Hartofilakidis G, Zacharakis N, Tsekoura M 12. McGrory BJ, Morrey BF, Cahalan TD, An KN, Cabanela ME
(1993) A 12- to 18-year radiographic follow-up study of Charnley (1995) Effect of femoral offset on range of motion and abductor
low-friction arthroplasty. The role of the center of rotation. Clin muscle strength after total hip arthroplasty. J Bone Joint Surg Br
Orthop Relat Res 296:140–147 77:865–869
3. Kurtz WB, Ecker TM, Reichmann WM, Murphy SB (2010) 13. Matsushita A, Nakashima Y, Jingushi S, Yamamoto T, Kuraoka
Factors affecting bony impingement in hip arthroplasty. A, Iwamoto Y (2009) Effects of the femoral offset and the head
J Arthroplast 25:624–634 size on the safe range of motion in total hip arthroplasty.
4. Maruyama M, Feinberg JR, Capello WN, D’Antonio JA (2001) J Arthroplast 24:646–651
The Frank Stinchfield Award: morphologic features of the 14. Sakalkale DP, Sharkey PF, Eng K, Hozack WJ, Rothman RH
acetabulum and femur: anteversion angle and implant position- (2001) Effect of femoral component offset on polyethylene wear
ing. Clin Orthop Relat Res 393:52–65 in total hip arthroplasty. Clin Orthop Relat Res 388:125–134
5. Bonnin MP, Archbold PH, Basiglini L, Fessy MH, Beverland DE 15. Devane PA, Horne JG (1999) Assessment of polyethylene wear
(2012) Do we medialise the hip centre of rotation in total hip in total hip replacement. Clin Orthop Relat Res 369:59–72
arthroplasty? Influence of acetabular offset and surgical tech- 16. Barrack RL (1998) Factors influencing polyethylene wear in total
nique. Hip Int 22(4):371–378 joint arthroplasty. Orthopedics 21:937–940
6. Asayama I, Chamnongkich S, Simpson KJ, Kinsey TL, Mahoney 17. Little NJ, Busch CA, Gallagher JA, Rorabeck CH, Bourne RB
OM (2005) Reconstructed hip joint position and abductor muscle (2009) Acetabular polyethylene wear and acetabular inclination
strength after total hip arthroplasty. J Arthroplast 20:414–420 and femoral offset. Clin Orthop Relat Res 467:2895–2900
7. Dastane M, Dorr LD, Tarwala R, Wan Z (2010) Hip offset in total 18. Terrier A, Levrero FF, Rudiger HA (2014) Benefit of cup
hip arthroplasty: quantitative measurement with navigation. Clin medialization in total hip arthroplasty is associated with femoral
Orthop Relat Res 469(2):429–436 anatomy. Clin Orthop Relat Res 472:3159–3165
8. Delp SL, Wixson RL, Komattu AV, Kocmond JH (1996) How 19. Nilsdotter AK, Lohmander LS, Klassbo M, Roos EM (2003) Hip
superior placement of the joint center in hip arthroplasty affects disability and osteoarthritis outcome score (HOOS)––validity and
the abductor muscles. Clin Orthop Relat Res 328:137–146 responsiveness in total hip replacement. BMC Musculoskelet
9. Klassbo M, Larsson E, Mannevik E (2003) Hip disability and Disord 4:10
osteoarthritis outcome score. an extension of the Western Ontario 20. Soderman P, Malchau H (2001) Is the Harris hip score system
and McMaster Universities Osteoarthritis Index. Scand J useful to study the outcome of total hip replacement? Clin Orthop
Rheumatol 32:46–51 Relat Res 384:189–197
10. Paillard P (2007) Hip replacement by a minimal anterior 21. Bourne RB, Rorabeck CH (2002) Soft tissue balancing: the hip.
approach. Int Orthop 31(Suppl 1):S13–S15 J Arthroplast 17:17–22
11. Matta JM, Ferguson TA (2005) The anterior approach for hip
replacement. Orthopedics 28:927–928

20
Supplemental S1 fixation for type C pelvic ring injuries:
biomechanical study of a long iliosacral versus a transsacral screw
Pooria Salari1 • Berton R. Moed1,2 • J. Gary Bledsoe1,2
Abstract Results Vertical displacement increased significantly

Background A single iliosacral screw placed into the S1 (p \ 0.05) within each group with each increase in the
vertebral body has been shown to be clinically unreliable number of cycles. However, there was no statistically
for certain type C pelvic ring injuries. Insertion of a second significant difference between groups in displacement or
supplemental iliosacral screw into the S1 or S2 vertebral load to failure.
body has been widely used. However, clinical fixation Conclusions Although intuitively a transsacral screw may
failures have been reported using this technique, and a seem to be better than a long iliosacral screw in conveying
supplemental long iliosacral or transsacral screw has been additional stability to an unstable sacral fracture fixation
used. The purpose of this study was to compare the construct, we were not able to identify any biomechanical
biomechanical effect of a supplemental S1 long iliosacral advantage of one method over the other.
screw versus a transsacral screw in an unstable type C Level of evidence Does not apply—biomechanical study.
vertically oriented sacral fracture model.
Materials and methods A type C pelvic ring injury was Keywords Iliosacral screws Transsacral screws Type
created in ten osteopenic/osteoporotic cadaver pelves by C pelvic ring injuries
performing vertical osteotomies through zone 2 of the
sacrum and the ipsilateral pubic rami. The sacrum was
reduced maintaining a 2-mm fracture gap to simulate a Introduction
closed-reduction model. All specimens were fixed using
one 7.0-mm iliosacral screw into the S1 body. A supple- Pelvic fractures account for 1–3 % of all skeletal fractures
mental long iliosacral screw was placed into the S1 body in and comprise a broad spectrum of injuries: from low-en-
five specimens. A supplemental transsacral S1 screw was ergy fractures in osteoporotic patients to high-energy dis-
placed in the other five. Each pelvis underwent 100,000 ruptions of the pelvic ring [1, 2]. Type C pelvic ring
cycles at 250 N, followed by loading to failure. Vertical injuries are vertically unstable due to complete disruption
displacements at 25,000, 50,000, 75,000, and 100,000 cy- of the posterior arch [3–6]. This posterior injury is by ne-
cles and failure force were recorded. cessity accompanied by a second injury site in the ring,
commonly in the anterior arch of the pelvic ring, and
& Berton R. Moed
consisting of disruption of the pubic symphysis, and ipsi-
[email protected] lateral and/or contralateral fractures of the superior and
inferior pubic rami [3, 7, 8]. Posterior ring disruption is
1
Department of Orthopaedic Surgery, Saint Louis University associated with high morbidity and mortality rates [9, 10].
School of Medicine, 3635 Vista Avenue, 7th Floor Desloge
As shown in multiple studies following treatment of the
Towers, St. Louis, MO 63110, USA
2
pelvic injury, residual deformity or associated injuries can
Department of Biomedical Engineering, Parks College of
create significant problems in functional recovery [11, 12].
Engineering, Aviation and Technology, Saint Louis
University, 3450 Lindell Boulevard, St. Louis, MO 63103, Numerous investigators have found that displacement
USA through the weight-bearing arch of the pelvis can lead to

Supplemental S1 fixation for type C pelvic ring injuries: biomechanical study of a long iliosacral versus... 135
long-term problems of pain and inability to regain function 2. An ipsilateral vertical zone 2 sacral fracture was
and resume previous lifestyle [13–17]. Regardless of the created by making a unilateral cut through the sacral
exact location of the posterior disruption, early restoration neuroforamina using an oscillating surgical power saw
of pelvic ring integrity is vital, and surgical management is with a thin blade.
thought to reduce long-term complications, such as malu- 3. The ipsilateral sacrospinous and sacrotuberous liga-
nion, nonunion, neurologic dysfunction, low-back pain, ments were transected to ensure complete disruption of
and gait abnormalities [4, 15–22]. the sacroiliac complex.
Many surgical techniques have been described for
This simulated, completely displaced, vertical zone 2
fixation of the posterior pelvic ring injury, with iliosacral
sacral fracture was then reduced in distraction, maintaining
screw fixation into the first sacral body being in common
a 2-mm fracture gap with a calibrated spacer (Fig. 1).
practice [4, 8, 19, 23–26]. Single iliosacral screw fixation
Using fluoroscopic guidance, each specimen was then fixed
into the S1 vertebral body has been shown to be clinically
using one standard-length 7.0-mm stainless steel cancel-
unreliable for unstable type C vertically oriented sacral
lous fully threaded cannulated iliosacral screw (Zimmer,
fractures [8]. Insertion of a second, supplemental, iliosacral
Inc., Warsaw, IN, USA) into the S1 vertebral body. Next,
screw into the S1 or S2 vertebral body has been widely
again using fluoroscopic guidance, a supplemental long
used [27]. In 2006, Moed and Geer published data on series
iliosacral screw was then placed into the S1 body in five
of patients reporting safe use of S2 screws. However, they
specimens (Fig. 2) and a supplemental transsacral S1 screw
raised the concern about using this type of screw in os-
was placed in the other five specimens (Fig. 3). To ensure
teopenic patients and recommended its use only with good
similarity of bone density between these two groups, spe-
bony purchase after instrumentation [28]. More recently
cimens were matched based on T-score values. The pelves
advocated is the use of a long iliosacral screw (extending
in the long iliosacral and transsacral groups had T scores
from the external surface of the ilium to just short of the
that were not significantly different, with means of -2.28
contralateral sacroiliac joint) or a transsacral screw (ex-
(range -1.4 to -3.4) and -2.38 (range -1.1 to -4.1),
tending from the external surface of the ilium across the
respectively (p = 0.62; Mann–Whitney U test). The os-
contralateral sacroiliac joint and exiting the ilium) [8, 29,
teotomized ipsilateral superior and inferior pubic rami were
30]. To our knowledge, no biomechanical study has been
not fixed in any of the pelvis specimens. After fluoroscopic
performed to differentiate the effect of these two screw
imaging confirmed appropriate sacral fracture reduction
lengths on fixation construct stability in type C, zone 2
and screw placement (Figs. 2 and 3), the spacer was re-
sacral fracture with a residual gap at the fracture site to
moved to mimic the clinical situation of a zone 2 sacral
mimic the clinical situation of a closed reduction in which
fracture percutaneously fixed without compression.
an anatomic reduction of the sacral fracture is not attained.
Using a previously described single-limb stance-testing
The purpose of this study was to biomechanically
model, each pelvis was mounted on a servohydraulic ma-
compare the effect of a supplemental S1 long iliosacral
terials-testing system (MTS 858 Mini Bionix, MTS Sys-
screw versus a transsacral screw in an unstable type C
tems, Inc., Eden Prarie, MN, USA) [32, 33]. The
vertically oriented zone 2 sacral fracture model [3, 5].
acetabulum on the ipsilateral side of the disrupted pelvis
was fitted with a potted femoral arthroplasty component
that was secured to the platform of the MTS machine with
two large C clamps to prevent any side-to-side motion. A T
plate was fixed to the ipsilateral iliac crest, and the pelvis
Ten embalmed cadaver pelves with intact, attached 4th and
was linked to a pulley system by a cable incorporated into
5th lumbar vertebrae were harvested with ligamentous
the jig (Fig. 4). Then, the pelvis was loaded using the MTS
structures (including sacroiliac, sacrospinous, sacrotuber-
hydraulic actuator, with the force being applied through a
ous, and symphyseal ligaments) and sacroiliac joint capsules
stainless steel ball-and-socket articulation attached to the
kept intact. Using a GE Lunar Scanner (GE Healthcare, UK),
superior endplate of the 5th lumbar vertebra [32–34]. This
dual-emission X-ray absorptiometry was performed on each
arrangement, which represents loading in vivo, allowed
specimen. All specimens were osteopenic, with a T score of
free rotation in all planes, thereby not restricting motion or
B-1 [31]. Subsequently, a completely unstable and dis-
causing displacement of the hemipelvis [33].
placed type C pelvic ring injury with a zone 2 sacral fracture
Subsequently, each pelvis was loaded at 250 N and
was created using the following steps:
cycled 100,000 times (equivalent to approximately
1. The right superior and inferior pubic rami were 3 weeks of walking [35, 36]) at two cycles per second
osteotomized in a vertical fashion using an oscillating (2 Hz) and then loaded to failure (Fig. 5). The value of
surgical power saw with a thin blade. 250 N was selected, as it approximates the in vivo force

Fig. 1 Pelvis showing the creation of an unstable type C, zone 2, vertically oriented injury. The arrow points to the 2-mm spacer used to create a
fracture gap
applied through the spine during static single-limb stance U test was used to compare bone density (as noted above),
[37, 38]. In addition, 250 N approximates 20 % of the load displacement, and load to failure of the two fixation groups.
to failure in similar studies [34]. Therefore, we felt this Freidman test was used for displacement comparisons
applied force was sufficient but would allow 100,000 cy- within each group. The level of statistical significance was
cles of loading without causing gross failure of the fixation defined as p \ 0.05.
constructs or disrupting the positioning of the pelvis in the
single-leg-stance setup. Vertical displacements were mea-
sured sequentially from the actuator at 25,000, 50,000, Results
75,000, and 100,000 cycles, and load-to-failure was
recorded for each pelvis using the MTS software. To attain All specimens completed 100,000 cycles with no gross
load to failure, a protocol was designed on the MTS soft- evidence of construct failure. The progressive increase in
ware to lower the actuator on the MTS machine at a rate of displacement between each of these measured intervals
1 mm/s. As the load increased, progressive displacement was significantly different (p \ 0.05) within both groups
resulting in fixation failure was expected to occur at the (Tables 1 and 2). In the group with supplemental long S1
sacral osteotomy site. Load and displacement were iliosacral screw, mean displacements at the sacral os-
recorded using MTS software. Failure was defined as the teotomy site at 25,000, 50,000, 75,000, and 100,000 were
point on the load–displacement curve when force mea- 14 ± 14.1, 18.5 ± 13.4, 20.7 ± 15.1, and
surement declined rapidly toward zero and there was no 22.8 ± 15.7 mm, respectively (Table 1). In the group with
further change in displacement [34]. a supplemental transsacral screw, mean displacements at
Statistics were calculated using SPSS software (SPSS the sacral osteotomy site at 25,000, 50,000, 75,000, and
version 19; SPSS Chicago, IL, USA). The Mann–Whitney 100,000 cycles were 10.6 ± 3.9, 11.3 ± 4.3, 11.5 ± 4.4,

Fig. 2 Fluoroscopic image of a pelvis instrumented in the iliosacral

group. The spacer was removed after screw placement
Fig. 4 Pelvis loaded into the MTS machine using the single-limb-
stance model
75,000, and 100,000 cycles showed no significant differ-

ence between groups (Table 4). In addition, there was no
significant difference between groups in load to failure
(Tables 3, 4). At the end of the study, gross inspection of
each specimen revealed that all screws were intact without
any obvious damage or deformity, and loss of fixation
appeared to be caused by loss of surrounding S1 bone
stock. A post hoc power analysis showed that with our
sample size of 5 pelves in each group, our data had 24 %
power for displacement and 10 % power for load to failure
to detect a difference at p \ 0.05.
Fig. 3 Fluoroscopic image of a pelvis instrumented in the transsacral

Discussion
group. The spacer was removed after screw placement
and 12.3 ± 4.5 mm, respectively (Table 2). After under- Stabilization of posterior pelvic ring injuries with iliosacral
going 100,000 cycles at 250 N and each pelvis was loaded screws inserted into the first sacral body is a commonly
to failure, mean load to failure was 546 ± 174 N for the used technique [19, 20, 32, 39–41]. Yinger et al. and van
long iliosacral screw group and 635 ± 196 N for the Zwienen et al., in their biomechanical studies, showed that
transsacral screw group (Table 3). for a completely unstable pelvic ring injury, using two il-
Despite the fact that mean displacement values for the iosacral screws increases rotational stiffness and load to
long iliosacral group were almost twice that of the trans- failure [26, 28]. Consistent with these findings, two il-
sacral group, analysis of displacements at 25,000, 50,000, iosacral screws inserted into S1, or one each into the S1

Fig. 5 Pelvis following fixation failure
Table 1 Posterior displacement

Number of cycles Displacement (in mm)
for the long iliosacral group
Minimum Maximum Mean Standard deviation P value*
25,000 4.8 38.8 14.0 14.1 –

50,000 9.2 42.0 18.5 13.4 0.025
75,000 9.4 47.3 20.7 15.1 0.025
100,000 9.5 49.3 22.8 15.7 0.025
* Freidman test
Table 2 Posterior displacement

Number of cycles Displacement (in mm)
for the transsacral group
Minimum Maximum Mean Standard deviation P value*
25,000 6.2 16.4 10.6 3.9 –

50,000 6.6 17.7 11.3 4.3 0.025
75,000 6.9 18.3 11.5 4.4 0.040
100,000 7.1 18.7 12.3 4.5 0.025
* Freidman test

Table 3 Load to failure for

Group Load to failure (in N)*
both groups
Minimum Maximum Mean Standard deviation
Long iliosacral screws 398 849 546 174

Transsacral screws 414 931 635 196
* P value = 0.42, Mann–Whitney U test
Table 4 Comparison between

Long iliosacral group Transsacral group P value*
groups
Mean displacement at 25,000 cycles (in mm) 14.0 10.6 0.54
Mean load to failure (in N) 546 635 0.42
* Mann–Whitney U test
and S2 bodies, are used as a preferred method for fixation risk for fixation failure. In addition, to maximize vertical
for these injuries [39]. However, this two-screw construct shear and minimize compression across the posterior pelvic
is clinically unreliable in some situations, especially with arch, a single-limb-stance model was used [32].
percutaneous fixation of unstable type C, zone 2, vertically Our study has a number of limitations. First, despite the
oriented sacral fractures in which a residual gap exists at use of nonparametric statistics, our failure to show a differ-
the fracture site [8]. ence between the two groups may be type 2 error due to the
Matta and Tornetta suggested that longer iliosacral relatively small sample size and lack of sufficient statistical
screws might provide better fixation because they have power. Our selection of five specimens in each group was
greater resistance to toggle and are more resistant to ver- based on the findings of Tabaie et al. [34]. However, post hoc
tical shear stress [39]. However, data to support this con- power analysis indicated low statistical power, which was
tention are wanting. A number of studies were unable to due to the relatively large standard deviations in our results
show any significant differences in fracture stability using (Tables 1–3). Therefore, despite the fact that the mean dis-
different iliosacral screw lengths [8, 30]. Griffin et al., in a placement values for the long iliosacral group were almost
study evaluating percutaneous iliosacral screw fixation of twice that of the transsacral group, there were no significant
62 unstable type C injuries, used four different screw differences between groups. These large variations from
lengths: into the sacral body, to the level of the contralat- specimen to specimen were not found by Tabaie et al. and
eral sacral foramen, to the contralateral sacral ala, and might represent greater variability in our specimens or test-
across the sacroiliac joint [8]. They reported that a vertical ing apparatus. However, it is interesting to note that com-
sacral fracture was the only statistically significant risk parison of our raw data to those of Tabaie et al. revealed that
factor for fixation failure [8]. Tornetta et al. found that a the locked transsacral screw construct reported by Tabaie
construct using a standard iliosacral screw in combination et al. has a significantly greater load to failure than our two
with a transsacral screw performed no better than a stan- fixation constructs or the short iliosacral construct tested by
dard two-screw construct [30]. However, Tabaie et al., in a Tabaie et al. (Table 5). In any case, the issue of specimen-to-
biomechanical study with a design similar to ours, com- specimen variability, compounded by a relatively small
pared standard iliosacral screws to a novel locked trans- sample size, is a common problem in biomechanical studies
sacral screw construct and reported significantly improved [26, 30, 34, 37, 38, 42]. Perhaps the differences between our
fixation using the transsacral locked method [34]. study groups would have reached statistical significance with
The purpose of our study was to assess the potential a much larger number of specimens. Second, using em-
improvement of fixation using one of two alternative long- balmed rather than fresh-frozen cadaver specimens is an-
screw fixation options: a transsacral or a long iliosacral other potential limitation. However, Comstock et al. used
implant. In order to create an ‘‘extreme’’ condition, os- embalmed cadaver specimens in a biomechanical evaluation
teopenic/osteoporotic pelvic specimens were used, and the of fixation of the posterior pelvic ring and found results
anterior fractures were not fixed. This allowed us to focus comparable with studies performed with fresh-frozen spe-
directly on the posterior fixation in a model at the greatest cimens [42]. More recently, van Zwienen et al. found

Table 5 Comparison of load to failure for long iliosacral, transsacral, short iliosacral, and locked transsacral screw constructs
Group Load to failure (in N) P value in comparison with
the other screw constructs
Minimum Maximum Mean Standard deviation
Locked transsacral screws (from [34])* 929 1201 1056 118

Long iliosacral screws 398 849 546 174 0.008
Transsacral screws 414 931 635 196 0.016
Short iliosacral screws (from [34]) 798 874 825 31 0.008
* Significantly different from the other three groups (multiple comparisons using Mann–Whitney U test with Bonferroni correction)
embalmed pelvic specimens to be satisfactory for biome-

chanical evaluation of unstable pelvic ring injuries [38].
Although there have been a number of studies compar-
ing standard iliosacral screws with longer screw constructs,
we know of no study directly comparing these longer screw
methods. Tornetta et al. described the concept of different
modes of failure [30], reporting that standard screws cut
through the sacrum while long screws bent, indicating that
References
the long screw was better anchored at its distal end [30].
However, in our study, after applying load to failure, none 1. Giannoudis P, Xypnitos F (2011) Management of pelvic frac-
of the screws in either group were bent or broken. Our tures. Eur Instr Lect 11:67–76
mode of failure was at the S1 body and alar bone stock. 2. Korovessis P, Baikousis A, Stamatakis M, Katonis P (2000)
This difference may be related to dissimilarities in design Medium- and long-term results of open reduction and internal
fixation for unstable pelvic ring fractures. Orthopedics
between the two studies: ours using a fracture-gap single- 23:1165–1171
stance model; theirs using anatomic reduction in a bilat- 3. Marsh JL, Slongo TF, Angel J, Broderick JS, Creevey W,
eral-stance model. DeCoster TA et al (2007) Fracture and dislocation classification
Although intuitively a transsacral screw may seem to be compendium—2007. J Orthop Trauma 21(10 suppl):S1–S163
4. Sagi HC (2009) Pelvic ring fractures. In: Bucholz RW, Court-
more advantageous than a long iliosacral screw in con- Brown CM, Heckman JD, Tornetta PIII (eds) Rockwood and
veying additional stability to a type C, zone 2, vertically Green’s fractures in adults, 7th edn. Lippincott Williams and
unstable sacral fracture fixation construct, we were not able Wilkins, Philadelphia, pp 1415–1462
to identify any biomechanical advantage of one fixation 5. Tile M (1996) Acute pelvic fractures: I. causation and classifi-
cation. J Am Acad Orthop Surg 4:143–151
method over the other. Further study with a larger number 6. Tile M, Hearn T, Vrahas M (2003) Biomechanics of the pelvic
of samples may be required to more accurately compare ring. In: Tile M, Helfet DL, Kellam JF (eds) Fractures of the
screw configuration in these injuries. pelvis and acetabulum, 3rd edn. Lippincott Williams and Wilkins,
Philadelphia, pp 32–45
Conflict of interest The authors attest that they have no financial or 7. Chenoweth DR, Cruickshank B, Gertzbein SD, Goldfarb P,
personal relationships of any kind with other people or organizations Janosick J (1980) A clinical and experimental investigation of
(such as employment, consultancies, stock ownership, honoraria, paid occult injuries of the pelvic ring. Injury 12:59–65
expert testimony, patent applications/registrations, and other funding) 8. Griffin DR, Starr AJ, Reinert CM, Jones AL, Whitlock S (2003)
that could inappropriately influence (bias) this work, entitled ‘‘Sup- Vertically unstable pelvic fracture fixed with percutaneous il-
plemental S 1 fixation for type C pelvic ring injuries: biomechanical iosacral screws: does posterior injury pattern predict fixation
study of a long iliosacral versus a transsacral screw.’’ However, one of failure? J Orthop Trauma 17:399–405
the authors is one of the owners of a patent for a locked transsacral 9. Egol KE, Koval KJ, Zuckerman JD (2010) Pelvis fracture. In: Egol
implant that is not commercially available. KE, Koval KJ, Zuckerman JD (eds) Hand book of fractures, 4th
edn. Lippincott Williams and Wilkins, Philadelphia, pp 327–343
Ethical standards At our institution, all human and animal studies 10. Peltier LF (1965) Complications associated with fractures of the
must be approved by the appropriate ethics committee and therefore pelvis. J Bone Joint Surg Am 47:1060–1069
be performed in accordance with the ethical standards laid down in 11. Gruen GS, Leit ME, Gruen RJ, Peitzman AB (1994) The acute
the 1964 Declaration of Helsinki and its later amendments. This study management of hemodynamically unstable multiple trauma pa-
is a biomechanical one involving cadaver specimens. Therefore, it tients with pelvic ring fractures. J Trauma 36:706–713
was exempt from formal Institutional Review Board Review, and the 12. Latenser BA, Gentilello LM, Tarver AA, Thalgott JS, Batdorf JW
need for informed consent was waived by the ethical committee since (1991) Improved outcome with early fixation of skeletally un-
rights and interests of those who donated their bodies to science stable pelvic fractures. J Trauma 31:28–31
would not be violated and their privacy and anonymity would be 13. Dunn AWM, Morris HD (1968) Fractures and dislocations of the
assured by this study design. pelvis. J Bone Joint Surg Am 50:1639–1648

14. Huittinen VM, Slätis P (1972) Fractures of the pelvis, trauma 29. Gardner MJ, Routt MLC Jr (2011) Transiliac-transsacral screws
mechanism, types of injury and principles of treatment. Acta Chir for posterior pelvic fixation. J Orthop Trauma 25:378–384
Scand 138:563–569 30. Tornetta P III, Jacofsky DJ, Jaczynski AM (2009) Trans-sacral
15. Kabak S, Halici M, Tuncel M, Avsarogullari L, Baktir A, Basturk screw fixation of stable zone 2 sacral fractures. http://www.hwbf.
M (2003) Functional outcome of open reduction and internal org/ota/am/ota09/otapo/OTP09018.htm. Accessed 20 Oct 2014
fixation for completely unstable pelvic ring fractures (Type C): a 31. Kanis JA (1994) Assessment of fracture risk and its application to
report of 40 cases. J Orthop Trauma 17:555–562 screening for postmenopausal osteoporosis: synopsis of a WHO
16. Metze M, Tiemann AH, Josten C (2007) Male sexual dysfunction report. WHO Study Group. Osteoporos Int 4:368–381
after pelvic fracture. J Trauma 63:394–401 32. Gorczyca J, Hearn T, Tile M (2003) Biomechanics and methods
17. Raf LA (1966) Double vertical fractures of the pelvis. Chir Scand of pelvic fixation. In: Tile M, Helfet DL, Kellam JF (eds) Frac-
131:298–305 tures of the pelvis and acetabulum, 3rd edn. Lippincott Williams
18. Carlson D, Scheid DK, Maar DC, Baele JR, Kaehr DM (2000) and Wilkins, Philadelphia, pp 116–129
Safe placement of S1 and S2 iliosacral screw: the ‘‘vestibule’’ 33. MacAvoy MC, McClellan RT, Goodman SB, Chien CR, Allen
concept. J Orthop Trauma 14:264–270 WA, van der Meulen MCH (1997) Stability of open-book pelvic
19. Cole JD, Blum DA, Ansel LJ (1996) Outcome after fixation of fractures using a new biomechanical model of single-limb stance.
unstable posterior pelvic ring injuries. Clin Orthop Relat Res J Orthop Trauma 11:590–593
329:160–179 34. Tabaie SA, Bledsoe JG, Moed BR (2013) Biomechanical com-
20. Failinger MS, McGainty PLJ (1992) Current concepts review— parison of standard iliosacral screw fixation to transsacral locked
unstable fractures of the pelvic ring. J Bone Joint Surg Am screw fixation in a type C—zone II pelvic fracture model.
74:781–791 J Orthop Trauma 27:521–526
21. Lefaivre KA, Slobogean GP, Ngai JT, Broekhuyse HM, O’Brien 35. Bassett DR Jr, Wyatt HR, Thompson H, Peters JC, Hill JO (2010)
PJ (2014) What outcomes are important for patients following Pedometer-measured physical activity and health behaviors in
pelvic trauma? Subjective responses and psychometric analysis of U.S. adults. Med Sci Sports Exerc 42:1819–1825
three published pelvic specific outcome instruments. J Orthop 36. Tudor-Locke C, Ham SA, Macera CA, Ainsworth BE, Kirtland
Trauma 28:23–27 KA (2004) Descriptive epidemiology of pedometer-determined
22. Matta J, Dickson K, Markovich G (1996) Surgical treatment of physical activity. Med Sci Sports Exerc 36:1567–1573
pelvic non-unions and malunions. Clin Orthop Relat Res 37. Sagi HC, Ordway NR, DiPasquale T (2004) Biomechanical
329:199–206 analysis of fixation for vertically unstable sacroiliac dislocations
23. Dujardin FH, Hossenbaccus M, Duprac F, Biga N, Thomine JM with iliosacral screws and symphyseal plating. J Orthop Trauma
(1998) Long-term functional prognosis of posterior injuries in 18:138–143
high-energy pelvis disruption. J Orthop Trauma 12:145–150 38. van Zwienen CMA, van den Bosch EW, Snijders CJ, Klein-
24. Routt ML, Simonian PT, Mills WJ (1997) Iliosacral screw fixa- rensink GJ, van Vugt AB (2004) Biomechanical comparison of
tion: early complications of the percutaneous technique. J Orthop sacroiliac screw techniques for unstable pelvic ring fractures.
Trauma 11:584–589 J Orthop Trauma 18:589–595
25. Tornetta P III, Matta JM (1996) Outcome of operatively treated 39. Matta JM, Tornetta P III (1996) Internal fixation of unstable
unstable posterior pelvic ring disruptions. Clin Orthop Relat Res pelvic ring injuries. Clin Orthop Relat Res 329:129–140
329:186–193 40. Moed BR, Karges DE (1996) Techniques for reduction and
26. Yinger K, Scalise J, Olson SA, Bay BK, Finkemeier CG (2003) fixation of pelvic ring disruptions through the posterior approach.
Biomechanical comparison of posterior pelvic ring fixation. Clin Orthop Relat Res 329:102–114
J Orthop Trauma 17:481–487 41. Moed BR, Whiting DR (2010) Locked transsacral screw fixation
27. Hasenboehler EA, Stahel PF, Williams A, Smith WR, Newman of bilateral injuries of the posterior pelvic ring: initial clinical
JT, Symonds DL et al (2011) Prevalence of sacral dysmorphia in series. J Orthop Trauma 24:616–621
a prospective trauma population: implications for a ‘‘safe’’ sur- 42. Comstock CP, van der Meulen MCH, Goodman SB (1996)
gical corridor for sacro-iliac screw placement. Patient Saf Surg Biomechanical comparison of posterior internal fixation tech-
5:8 niques for unstable pelvic fractures. J Orthop Trauma 10:517–522
28. Moed BR, Geer B (2006) S2 iliosacral screw fixation for dis-
ruptions of the posterior pelvic ring: a report of 49 cases. J Orthop
Trauma 20:378–383

21
Effectiveness of psychological support in patients undergoing
primary total hip or knee arthroplasty: a controlled cohort study
V. Tristaino1 • F. Lantieri2 • S. Tornago1 • M. Gramazio1 • E. Carriere1 •
A. Camera1
Abstract and knee populations separately, an exact opposite tem-

Background We hypothesised that psychological support poral trend was observed in the experimental group com-
would have a significant improvement on the mental and pared to the control group (p \ 0.0001), with generally
physical recovery of patients undergoing primary total hip higher scores in the experimental group (p \ 0.0001). In
or knee arthroplasty. patients with hip arthroplasty, the average time to reach the
Materials and methods 200 patients were consecutively physiotherapy objective (i.e., the patient ability to walk 50
alternately assigned (1:1) to receive routine care (control metres independently and to climb 10 steps) was
group) or, in addition, psychological support from a pro- 6.7 ± 1.8 days (range 4–12) in the experimental group and
fessional psychologist (experimental group). The psycho- 7.9 ± 2.2 days (range 0–13) in the control group
logical support was provided at the pre-operative visit, (p = 0.0015).
during the hospitalisation period and at the rehabilitation Conclusions In summary, there was a lower incidence of
centre. anxiety and depression and better mental well-being in the
Results Upon discharge, based on the ‘Hospital Anxiety group of patients who received the psychological support.
and Depression Scale, a state of anxiety was observed in Within the hip arthroplasty group, the patients who
12.8 % and 78.9 % of the patients in the experimental and received the psychological support reached the physio-
in the control group, respectively (p \ 0.0001). A state of therapy objective 1.2 days earlier than the patients in the
depression was observed in 12.8 % and 73.7 % of the control group (p = 0.0015).
patients in the experimental and in the control group, Level of evidence Level 3, Non-randomized prospective
respectively (p \ 0.0001). With regard to the ‘Physical controlled cohort.
Component Scale’ of the SF-36 questionnaire, a similar
temporal trend of values was observed in the two study Keywords Psychological support Hip arthroplasty
groups, significantly increasing over time in both groups, Knee arthroplasty SF-36 Hospital anxiety Depression
taking into consideration both the joint population and the scale
two hip and knee populations separately (p \ 0.0001).
With regard to the ‘Mental Component Scale’ of the SF-36
questionnaire, in both the joint population and the two hip Introduction
Primary total hip arthroplasty and primary total knee

& V. Tristaino arthroplasty are established elective operations to resolve
[email protected] most severe arthritic conditions affecting the two major
1 lower limb joints. They are two highly successful ortho-
Department of Prosthetic Surgery, Santa Corona Hospital,
ASL 2 - Savonese, via XXV Aprile 38, 17027 Pietra Ligure, paedic interventions in terms of overall functional recovery
SV, Italy for the patient and the incidence of complications. In spite
2
Health Science Department, Biostatistics Unit, University of of this, however, the journey that the patient must take is
Genoa, via Pastore 1, Genoa 16132, Italy not without difficulties in terms of the emotions that he or

Effectiveness of psychological support in patients undergoing primary total hip or knee arthroplasty: a controlled... 143
she may experience in the months leading up to the oper- Prosthetic Surgery of Santa Corona Hospital (Pietra Ligure,
ation, during the stay in hospital, during rehabilitation and SV, Italy) were enrolled in the study. To be eligible they
in the first few months after the operation. had to meet the following inclusion criteria—(1) first
A patient who makes the choice to have a hip or knee prosthetic hip or knee replacement; (2) no psychiatric
arthroplasty operation experiences periods of anxiety and history at the time of enrolment; (3) no degenerative ner-
depression, as already reported in many recent studies. vous system diseases; (4) aged \80 years; (5) initial
Anxiety and depression are emotions that are already pre- decision to carry out rehabilitation at the physiotherapy
sent in the period before the operation [1] and impact on centre was referred through the hospital; and (6) provided
the post-operative progress [2–6]; however, generally informed consent for participation in the study and pro-
speaking, the satisfaction that results from these two types cessing of personal data.
of operation can be considered as undisputed [7]. Each patient who met the inclusion criteria was con-
In the short term, a patient’s recovery of functionality secutively alternately assigned to one of two groups (1:1),
after the operation is mainly linked to clinical factors, e.g., with the allocation of the first patient chosen at random by
the extent of the surgical trauma, but in the long term it is tossing a coin, before the operation. The experimental
more closely linked to the degree of functionality before the group (EXP) consisted of patients who, in addition to
operation and the patient’s emotional [8] and psychological routine treatment, received psychological support from a
reaction (anxiety) to the operation [9]. The patient’s reaction professional psychologist and the control group (CTR)
is not just understood as his or her physiological response to consisted of patients who only received routine treatment.
the operation from a physical point of view, but it also The surgical team was blinded to the treatment arm.
comprises a component that is already partly present in the After enrolment, the patients who had experienced intra-
periods prior to admission combined with an element of the or post-operative complications or for whom more than one
patient’s psychological disposition. Practical implications item of data was missing were excluded (Table 1). Patient
concern the contemplation of psychological factors and the demographics are documented in Table 2.
treatment of psychological symptoms in rehabilitation [10]
and the person’s social and functional readjustment [11–13]. Routine treatment
Therefore, it seems logical to evaluate whether psy-
chological support therapy which accompanies patients As normal practice at our institution, the surgeon during
from their admission to hospital until their discharge can the pre-operative meeting with the patient provided him/
impact on the surgical outcome during the rehabilitation her with operation-related information, as well as using a
period and in the first few months following the operation. standard information brochure as a guide. The information
Although various controlled clinical studies have explained (1) what arthroplasty is and why arthroplasty is
already documented the effect of psychological support in performed, (2) what a prosthesis is, (3) what type of
patients who have undergone cardiovascular surgery [14], prosthesis is chosen, (4) the surgical planning, (5) some
the removal of breast cancer [15] and gastric band surgery information on the surgery itself, and (6) what to do after
[16], we are not aware of any controlled studies relating to discharge (i.e., physical exercises, lifestyle, clinical follow-
patients undergoing hip or knee arthroplasty operations. up visits). The pre-operative meeting between the surgeon
It is for this reason that this controlled cohort study was and the patient took place before patient allocation to one
planned, with the aim of determining the effectiveness of of the two arms.
psychological support in patients undergoing primary total
hip or knee arthroplasty. We hypothesised that psycho- Psychological support
logical support from a professional psychologist would
significantly improve the mental and physical recovery of The psychological support was provided by a professional
patients. The patients with and without psychological psychologist (author VT) and focused on the type of clin-
support therapy were examined by means of standard ical procedure within the scope of hospital health psy-
questionnaires completed by the patient (‘patient reported chology. The activity was carried out over the course of
outcome measures’) and by measuring rehabilitation time. four sessions between the psychologist and the patient,
lasting about half an hour each time. One session was
carried out in the pre-operative period, two during the
Materials and methods hospital stay and one during the stay at the rehabilitation
centre (Table 3). The protocol for the psychological sup-
Between February 2011 and May 2012, 200 consecutive port activity was developed by the psychologist after
patients on a waiting list for an elective operation for pri- 1 year of non-participant observation at the Department of
mary total hip or knee arthroplasty at the Department of Prosthetic Surgery, aimed at defining the psychological

Table 1 Study population

EXP CTR
Hip arthroplasty Knee arthroplasty Hip arthroplasty Knee arthroplasty
group group group group
No. of patients in the initial cohort 63 37 66 34

No. of patients excluded from the 2 ([1 data item 4 (2 had [1 data item 3 (2 had [1 data item 2 (1 had [1 data item
study (reasons for exclusion) missing) missing; 2 had missing; 1 had missing; 1 had
post-operative post-operative post-operative
complications) complications) complications)
No. of patients with pre-op SF-36 61 33 63 32
available
No. of patients with HADS available 61 33 63 32
No. of patients with physiotherapy 59 33 61 32
assessment available
No. of patients with SF-36 at 61 33 63 32
45 days available
No. of patients with SF-36 at 60 33 63 31
4 months available
Table 2 Patient demographics

All patients Hip arthroplasty group Knee arthroplasty group
EXP CTR EXP CTR EXP CTR
No. of patients 94 95 61 63 33 32
Age at surgery (mean ± SD; years) 61.4 ± 8.7 64.5 ± 8.1 59.9 ± 8.4 63.7 ± 8.7 64.2 ± 8.6 66.1 ± 6.6
Gender (M/F) 45/49 56/39 36/25 31/32 13/20 8/24
Table 3 Study synopsis

Time period\activity Pre-operation Hospital stay Rehabilitation centre stay Upon patient 45 days after 4 months after
discharge surgery surgery
Psychological support EXP (1 session) EXP (2 sessions) EXP (1 session) – – –

HADS compiling – – – EXP – –
CTR
SF-36 compiling EXP° – – – EXP EXP
CTR CTR CTR
Physiotherapic assessment – – EXP (each day) – – –
CTR (each day)
° Following the first session with the psychologist
themes and concepts on which to focus the activity. The it is now increasingly common to find the term ‘psychoe-
protocol can be summarised as follows: ducational’ associated with health care programmes,
including in the specific field of arthroplasty [18, 19].
1. Ascertainment of correct comprehension of the medical
2. The patient’s personal history and discussion of the
and supporting information and clarification of any doubts
psychological experiences linked with the illness and
and misunderstandings (at the time of admission). It must
the prescription/decision to undergo an arthroplasty
be noted that, from the perspective of health psychology
operation (at the time of admission).
[17], the provision of health information about the risk
3. Processing the emotional states associated with the
factors corresponds to an increase in the level of
operation and support to manage them (at the time of
information with a possible increase in anxiety and
admission and during the stay in hospital).
consequent use of dysfunctional strategies. For this reason,

4. Modulation of stress and emotional and behavioural scale is converted into a scale ranging from 0-100, with
reactions associated with the recovery. Reinforcement the assumption that each question carries the same weight
of the awareness of perceived self-efficiency associ- in the final total. The lower the score is, the worse the
ated with the results in the short, medium and long impairment and vice versa (i.e., 0 indicates the maximum
term by explaining to the patient their active role in the impairment, while 100 indicates no impairment). It is
healing process (during the stay in hospital and in the possible to obtain two indices from these 8 sub-scales—the
rehabilitation centre). ‘Physical Component Summary’ (PSC) index, comprised
5. Discussion with the patient regarding his/her discharge of the first four sub-scales listed above and the ‘Mental
from hospital, returning home and the check-up visit Component Summary’ (MCS) index, comprised of the last
schedule (during the stay at the rehabilitation centre). four sub-scales. These indices represent two mathematical
calculations which allow us to establish how important the
The various phases followed on from one another in a
physical and mental components are in the patient to
way which was personalised to each patient’s psychologi-
determine their state of well-being [21].
cal needs and shaped gradually to tackle the various phases
(from admission to rehabilitation). During all of the phases,
Physiotherapy sheet
the psychologist also used as a guide the standard infor-
mation brochure that was already provided to the patient by
During the stay at the rehabilitation centre (8 days fol-
the surgeon during the pre-operative meeting.
lowing the 5 post-operative days spent in hospital), the
physiotherapy evaluation sheet was filled in daily for each
Patient evaluation (evaluation programme
patient as routine practice. The information on this sheet
in Table 3)
regarding the time taken between the start of physiotherapy
at the rehabilitation centre and reaching the physiotherapy
Patient questionnaires
objective, defined as the ability to walk 50 metres inde-
pendently and to climb 10 steps (i.e., objective defined as
The ‘Hospital Anxiety and Depression Scale’ (HADS)
the potential minimum for discharge), was analysed for this
questionnaire [20] was completed by patients from both
study. This parameter was defined in this study as ‘delta
groups at the end of the hospital stay. The HADS is a
autonomy days’. The physiotherapist was blinded to the
widely used questionnaire consisting of 14 items which
treatment arm the patient was assigned to.
comprise 2 scales—7 items relating to the scale to measure
anxiety (HADS-A) and 7 items relating to the scale to
Data analyses
measure depression (HADS-D). Each item is given a score
between 0 and 3, so the total score for each scale ranges
The following were analysed:
from 0-21. Values between 0 and 7 indicate a ‘normal’
state of the patient, while higher values indicate a degree of (a) The presence of anxiety and depression using the
anxiety and depression starting from ‘mild’ (8–10), then HADS questionnaire. The results of each of the two
‘moderate’ (11–14), and lastly ‘severe’ (15–21). scales (anxiety and depression) were divided into
This questionnaire is useful to evaluate problems of anxiety two categories—no anxiety or depression (values
and depression in hospitalised patients and patients affected between 0 and 7) and presence of anxiety or
by any physical disease which forces them to undergo medical depression (values between 8 and 21). The compar-
treatment. The grading of the two variables—anxiety and ison between the experimental group and the control
depression—in this specific study should not be incorporated group was made in the joint population and in the
in a clinical-pathological perspective, but in a perspective that two separate populations of patients with hip
considers anxiety and depression as physiological compo- arthroplasty (referred to here as the ‘hip population’)
nents of the contingent situation experienced by the patient. and the patients with knee arthroplasty (referred to
The SF-36 questionnaire was completed by patients here as the ‘knee population’). The groups were
from both study groups during the pre-operative visit (the compared using the chi-squared test with Yate’s
same day as admission but after the first session with the correction or by Fisher’s exact test where more
psychologist), at the follow-up on day 45 and at the feasible.
4-month follow-up after surgery. The questionnaire con- (b) The scores relating to the SF-36 questionnaire,
sists of 36 questions with multiple-choice answers which collected at various time intervals (pre-operative,
make up 8 sub-scales—‘physical functioning’, ‘role-phys- on day 45 after the operation and at 4 months after
ical’, ‘bodily pain’, ‘general health’, ‘vitality’, ‘social the operation). At each follow-up, a comparison was
functioning’, ‘role-emotional’ and ‘mental health’. Each made between the groups using the student’s t test

for independent samples. Considering the relatively Patient questionnaires

low number of samples, the type of data and their
increased variability, especially in the sub-scales, the The following results were obtained:
Mann–Whitney nonparametric test was also applied, HADS (Table 4): 12 out of 94 patients in the experi-
which fully confirmed the statistical results of the mental group (12.8 %) manifested a state of anxiety,
t-test. The temporal trend of the PCS and MCS compared to 75 out of 95 in the control group (78.9 %)
scales and of all the sub-scales making up the SF-36 (p \ 0.0001). Similarly, a state of depression was observed
score was analysed in the experimental group and in 12 out of 94 patients in the experimental group (12.8 %)
control group by means of a two-way repeated and in 70 out of 95 (73.7 %) in the control group
measures analysis of variance (ANOVA). This (p \ 0.0001). The differences between the experimental
analysis simultaneously compares the difference group and the control group were also significant within
between the samples and between the detection both the hip population and the knee population.
times and highlights any behavioural differences SF-36 (Table 5): With regard to the joint population
(interaction) between the groups. The comparison of (hip?knee), considerably higher average values were
the results was made in the joint population and obtained in all 8 sub-scales in the experimental group
separately within the ‘hip population’ and the ‘knee compared to the control group in the pre-operative stage
population’. and in the two subsequent follow-ups. Furthermore, in the
(c) The ‘delta autonomy days’, separately within the case of the ‘hip population’, the differences were signifi-
‘hip population’ and the ‘knee population’. The cant in all subscales and follow-ups apart from the ‘role-
analysis was carried out using the Student’s t test for physical’ sub-scale at the follow-up on day 45. For the
independent samples and the results were confirmed ‘knee population’ the differences between the two groups
through the Mann–Whitney test. only reached statistical significance in some of the sub-
scales—all 4 sub-scales of the ‘Mental Component Scale’
Considering the type and the distribution of the data and
both in the follow-up on day 45 and at 4 months, and the
given the accordingly similarity of the statistical results
‘Physical Functioning’ and the ‘General Health’ sub-scales
obtained with the parametric test and with the nonpara-
on day 45.
metric test, the data relating to the eight sub-scales, the two
With regard to the ‘Physical Component Scale’, a sim-
SF-36 score indices and the physiotherapy evaluation were
ilar temporal trend of values was observed in the two study
summarised as an average and standard deviation and the
groups, significantly increasing over time in both groups,
p-values reported refer to the parametric test.
taking into consideration both the joint population and the
For all of the comparisons between the groups, a p-value
two populations (hip and knee) separately (p \ 0.0001).
of \0.05 was considered to be significant. The statistical
For the joint population, the values were significantly
analysis was carried out with the SPSS 17.0 software.
higher as a whole in the experimental group compared to
The data were collated by the first author (VT) and
the control group (p = 0.0310) and, in particular, were
analysed by a statistician (a co-author; FL). The statistician
higher in the pre-operative visit (p = 0.0466) and in the
was blinded to the treatment arm and to what the numerical
follow-up at 4 months (p = 0.0135), while there was no
measures meant.
significant difference in the follow-up on day 45. The same
consideration applies for the ‘hip population’ (Fig. 1a),
Sample size determination while there was no significant difference between the
groups in the ‘knee population’ (Fig. 1b) at any follow-up.
This is an original research study in the field of hip and knee With regard to the ‘Mental Component Scale’, in both the
arthroplasty; therefore, it was not possible to refer to other joint population and the two hip and knee populations sep-
studies in literature to perform a sample size calculation. arately, an exact opposite temporal trend was observed in the
experimental group compared to the control group
(p \ 0.0001), with generally higher scores in the experi-
Results mental group (p \ 0.0001). The differences are significant in
the pre-operative stage, on day 45 and at 4 months after the
Of the 200 patients enrolled, 11 (5 from the control group operation in both the joint population (p = 0.0005,
and 6 from the experimental group) were excluded—4 due p \ 0.0001, p \ 0.0001, respectively) and in the ‘hip pop-
to intra- or post-operative complications and 7 due to the ulation’ (p = 0.0001, p \ 0.001, p = 0.0004, respectively)
lack of more than one data item (e.g., subject not available, (Fig. 1c). In the ‘knee population’, a significant difference
lack of cooperation, transfer to physiotherapy centre other was observed on day 45 (p \ 0.0001) and at 4 months
than the one referred) (Table 1). (p = 0.0013) but not at the pre-operative visit (Fig. 1d).

Table 4 Hospital anxiety and

EXP CTR p value
depression scale (HADS) results
Anxiety
All patients 12/94 (12.8 %) 75/95 (78.9 %) \0.0001*,a
Hip arthroplasty group 7/61 (11.5 %) 49/63 (77.8 %) \0.0001*,b
Knee arthroplasty group 5/33 (15.2 %) 26/32 (81.3 %) \0.0001*,b
Depression
All patients 12/94 (12.8 %) 70/95 (73.7 %) \0.0001*,a
Hip arthroplasty group 8/61 (13.1 %) 49/63 (77.8 %) \0.0001*,b
Knee arthroplasty group 4/33 (12.1 %) 21/32 (65.6 %) \0.0001*,b
Calculation performed on 95 patients in the EXP group (63 hips; 32 knees) and 94 patients in the CTR
group (61 hips; 33 knees)
* Significance at p \ 0.05
a
Chi-squared test with Yate’s correction
b
Fisher’s exact test
The average values and the statistical significances are case of knee arthroplasty the physical component (also
stated in Table 5. understood as physical pain and the role it plays in the
perceived quality of health) has more prominence and may
Physiotherapy sheet be less influenced by psychological support.
With regard to the ‘Mental Component Scale’ of the SF-
The following results were obtained: 36 score, the results of the overall population (hip?knee)
‘Delta autonomy days’ (Table 6): with regard to the ‘hip were significantly better in the subjects provided with
population’, a significant difference between the experi- psychological support in the pre-operative stage and in the
mental group and the control group was observed, with the two subsequent follow-ups. These values were already
physiotherapy objective being reached, on average, after higher after the first session with the psychologist, taking
6.7 ± 1.8 days (range 4–12) and 7.9 ± 2.2 days (range into consideration the two populations separately (hip and
0–13), respectively, after the operation (p = 0.0015). The knee), with significant differences in all cases, apart from
difference between the experimental group and the control the pre-operative stage for the patients undergoing knee
group in the ‘knee population’ did not reach the statistical operations. In our opinion, these results indicate that the
significance [8.1 ± 2.4 days (range 5–16) vs 8.8 ± psychological support provided during admission, the
2.3 days (range 5–14)]. hospital stay and rehabilitation led to an improvement in
mental well-being in both the short and long term. In
addition, the fact that the score in patients who received
Discussion psychological support increased at the follow-up on day 45
and then decreased at 4 months (but remained higher than
The study highlighted that the group that received psy- the control group) shows, in our opinion, the effectiveness
chological support presented a significantly lower number and the impact of psychological therapy, especially in the
of patients with a state of anxiety and depression upon initial period after the surgery up to the evaluation on day
discharge compared to the control group. 45. Afterwards, the improvement achieved would build up
With regard to the ‘Physical Component Scale’ of the even more over time from a physical and, consequently,
SF-36 score, an improvement in scores over time was emotional point of view.
observed in both the experimental group and the control Lastly, it was observed that the patients provided with
group, although with generally higher scores in the psychological support who underwent hip arthroplasty
experimental group. As regards the population with hip reached the physiotherapy objective (i.e., the patient ability
arthroplasty, the scores were significantly higher in the to walk 50 metres independently and to climb 10 steps)
experimental group in the pre-operative stage (after the first 1.2 days earlier, on average, compared to the patients who
session with the psychologist) and in the follow-up at did not receive this therapy (p = 0.0015). This improve-
4 months. In the population with knee arthroplasty, a sig- ment was also apparent in the population with knee
nificant difference between the two groups was not arthroplasty, although the difference between the study and
observed in any of the follow-ups. This difference in the the control group was in this case not significant. In our
results between the patients with hip operations and those opinion, the incorporation of psychological support in the
with knee operations could be due to the fact that in the clinical, surgical and rehabilitation procedure could

Table 5 Results of the SF-36 questionnaire
148
Scale Population Pre-operative 45 days

(‘‘pop’’)
No. of EXP CTR p value No. of EXP CTR p value
patients patients
Sub-scales
Physical functioning Joint pop 94 vs 95 45.5 ± 27.5 36.6 ± 21.3 0.0059* 94 vs 95 65.0 ± 21.2 48.2 ± 24.7 \0.0001*
Hip pop 61 vs 63 48.2 ± 26.8 36.9 ± 21.7 0.0111* 61 vs 63 68.4 ± 20.5 49.7 ± 25.0 \0.0001*
Knee pop 33 vs 32 40.5 ± 28.3 36.1 ± 20.9 0.4837 33 vs 32 58.6 ± 21.4 45.2 ± 24.2 0.0203*
Role-physical Joint pop 94 vs 95 22.6 ± 30.2 13.9 ± 28.4 0.0439* 94 vs 95 18.6 ± 34.6 9.5 ± 23.1 0.0343*
Hip pop 61 vs 63 23.0 ± 30.0 11.9 ± 24.9 0.0281* 61 vs 63 19.7 ± 36.0 9.1 ± 23.5 0.0568
Knee pop 33 vs 32 22.0 ± 31.1 18.0 ± 34.3 0.6239 33 vs 32 16.7 ± 32.3 10.2 ± 22.8 0.3523
Bodily pain Joint pop 94 vs 95 40.1 ± 19.3 30.5 ± 18.3 0.0005* 94 vs 95 70.5 ± 23.6 58.6 ± 25.6 0.0011*
Hip pop 61 vs 63 43.3 ± 21.2 30.5 ± 19.1 0.0006* 61 vs 63 77.0 ± 21.2 63.0 ± 24.8 0.001*
Knee pop 33 vs 32 34.2 ± 13.7 30.4 ± 17.0 0.3283 33 vs 32 58.5 ± 23.6 49.9 ± 25.3 0.1588
General health Joint pop 94 vs 95 66.8 ± 18.0 57.4 ± 20.1 0.0008* 94 vs 95 79.4 ± 19.0 66.0 ± 22.4 \0.0001*
Hip pop 61 vs 63 69.8 ± 16.4 56.0 ± 22.1 0.0001* 61 vs 63 80.7 ± 19.9 66.1 ± 24.0 0.0003*
Knee pop 33 vs 32 61.4 ± 19.7 60.2 ± 15.3 0.7787 33 vs 32 77.1 ± 17.2 65.8 ± 18.9 0.0142*
Vitality Joint pop 94 vs 95 66.8 ± 18.0 41.1 ± 19.8 0.0001* 94 vs 95 71.9 ± 19.9 37.3 ± 22.6 \0.0001*
Hip pop 61 vs 63 53.4 ± 16.1 38.6 ± 21.0 \0.0001* 61 vs 63 74.8 ± 18.0 37.9 ± 23.6 \0.0001*
Knee pop 33 vs 32 50.3 ± 23.5 46.1 ± 16.3 0.4030 33 vs 32 66.4 ± 22.2 36.3 ± 20.8 \0.0001*
Social functioning Joint pop 94 vs 95 63.6 ± 23.9 53.2 ± 27.9 0.0065* 94 vs 95 77.5 ± 23.2 45.3 ± 26.9 \0.0001*
Hip pop 61 vs 63 64.5 ± 23.4 49.4 ± 28.3 0.0015* 61 vs 63 80.1 ± 21.3 45.9 ± 29.1 \0.0001*
Knee pop 33 vs 32 61.7 ± 25.2 60.5 ± 25.8 0.8507 33 vs 32 72.7 ± 25.6 44.1 ± 22.2 \0.0001*
Role-emotional Joint pop 94 vs 95 48.6 ± 39.9 35.8 ± 38.4 0.0258* 94 vs 95 79.8 ± 33.6 28.4 ± 35.7 \0.0001*
Hip pop 61 vs 63 50.8 ± 41.1 33.3 ± 37.4 0.0147* 61 vs 63 84.7 ± 29.5 29.1 ± 37.1 \0.0001*
Knee pop 33 vs 32 44.4 ± 37.9 40.6 ± 40.4 0.6955 33 vs 32 70.7 ± 38.9 27.1 ± 33.3 \0.0001*
Mental health Joint pop 94 vs 95 67.1 ± 19.4 54.3 ± 22.6 \0.0001* 94 vs 95 82.3 ± 20.4 46.0 ± 26.9 \0.0001*
Hip pop 61 vs 63 67.9 ± 18.5 51.0 ± 22.4 \0.0001* 61 vs 63 83.3 ± 19.4 47.7 ± 28.4 \0.0001*
Knee pop 33 vs 32 65.7 ± 21.2 60.8 ± 21.8 0.3568 33 vs 32 80.6 ± 22.3 42.8 ± 23.9 \0.0001*
Physical component and mental component summaries
Physical component summary Joint pop 94 vs 95 33.8 ± 9.3 31.3 ± 7.4 0.0466* 94 vs 95 39.2 ± 8.4 39.2 ± 7.8 0.9665
Hip pop 61 vs 63 35.1 ± 9.1 31.5 ± 8.1 0.0192* 61 vs 63 40.5 ± 8.9 39.8 ± 8.0 0.6368
Knee pop 33 vs 32 31.2 ± 9.4 31.0 ± 5.8 0.9075 33 vs 32 36.8 ± 6.9 37.9 ± 7.5 0.5397
Mental component summary Joint pop 94 vs 95 47.8 ± 11.3 41.8 ± 12.2 0.0005* 94 vs 95 56.7 ± 11.6 35.0 ± 13.0 \0.0001*
Hip pop 61 vs 63 48.1 ± 11.1 39.9 ± 11.9 0.0001* 61 vs 63 57.6 ± 10.6 35.4 ± 14.1 \0.0001*
Knee pop 33 vs 32 47.4 ± 11.7 45.5 ± 12.0 0.5291 33 vs 32 55.0 ± 13.3 34.4 ± 10.9 \0.0001*

Table 5 continued
Scale Population 4 months ANOVA
(‘‘pop’’)
No. of EXP CTR p value No. of Main effect Main effect Interaction between treatment
patients patients of treatment of follow-up allocation and follow-up
Sub-scales
Physical functioning Joint pop 93 vs 94 86.6 ± 14.3 74.3 ± 25.5 \0.0001* 93 vs 94 \0.0001* \0.0001* 0.1125
Hip pop 60 vs 63 88.4 ± 13.8 71.8 ± 27.8 \0.0001* 60 vs 63 \0.0001* \0.0001* 0.2807
Knee pop 33 vs 31 83.3 ± 14.7 79.4 ± 19.7 0.3700 33 vs 31 0.0490* \0.0001* 0.2449
Role-physical Joint pop 93 vs 94 79.3 ± 36.4 66.3 ± 45.2 0.0319* 93 vs 94 0.0014* \0.0001* 0.7794
Hip pop 60 vs 63 84.6 ± 35.1 62.9 ± 46.4 0.0040* 60 vs 63 0.0003* \0.0001* 0.3100
Knee pop 33 vs 31 69.7 ± 37.4 73.4 ± 42.3 0.7124 33 vs 31 0.7136 \0.0001* 0.6738
Bodily pain Joint pop 93 vs 94 76.7 ± 23.2 68.6 ± 28.0 0.0326* 93 vs 94 \0.0001* \0.0001* 0.6229
Hip pop 60 vs 63 80.4 ± 23.5 69.0 ± 28.8 0.0181* 60 vs 63 0.0001* \0.0001* 0.8813
Knee pop 33 vs 31 70.1 ± 21.5 67.8 ± 26.8 0.7148 33 vs 31 0.1317 \0.0001* 0.5534
General health Joint pop 93 vs 94 78.9 ± 19.7 67.7 ± 25.8 0.0011* 93 vs 94 \0.0001* \0.0001* 0.4014
Hip pop 60 vs 63 81.1 ± 18.9 64.4 ± 27.5 0.0001* 60 vs 63 \0.0001* \0.0001* 0.6899
Knee pop 33 vs 31 74.9 ± 20.7 74.4 ± 21.1 0.9158 33 vs 31 0.1991 \0.0001* 0.0546
Vitality Joint pop 93 vs 94 74.9 ± 20.7 74.4 ± 21.1 \0.0001* 93 vs 94 \0.0001* \0.0001* \0.0001*
Hip pop 60 vs 63 61.3 ± 23.3 40.1 ± 22.6 \0.0001* 60 vs 63 \0.0001* \0.0001* \0.0001*
Knee pop 33 vs 31 53.8 ± 24.4 39.8 ± 19.8 0.0150* 33 vs 31 0.0001* 0.3827 0.0002*
Social functioning Joint pop 93 vs 94 75.7 ± 23.5 56.0 ± 23.8 \0.0001* 93 vs 94 \0.0001* 0.0009* \0.0001*
Hip pop 60 vs 63 76.9 ± 21.8 56.2 ± 24.7 \0.0001* 60 vs 63 \0.0001* 0.0005* 0.0005*
Knee pop 33 vs 31 73.5 ± 26.5 55.6 ± 22.3 0.0051* 33 vs 31 0.0006* 0.2377 0.0018*
Role-emotional Joint pop 93 vs 94 78.9 ± 33.6 60.3 ± 39.2 0.0006* 93 vs 94 \0.0001* \0.0001* \0.0001*
Hip pop 60 vs 63 81.1 ± 32.1 63.5 ± 40.0 0.0080* 60 vs 63 \0.0001* \0.0001* \0.0001*
Knee pop 33 vs 31 74.8 ± 36.4 53.8 ± 37.2 0.0259* 33 vs 31 0.0005* 0.0031* 0.0065*
Mental health Joint pop 93 vs 94 67.8 ± 26.6 50.1 ± 26.2 \0.0001* 93 vs 94 \0.0001* 0.0130* \0.0001*
Hip pop 60 vs 63 68.7 ± 27.5 50.0 ± 27.8 0.0003* 60 vs 63 \0.0001* 0.0040* \0.0001*
Knee pop 33 vs 31 66.3 ± 25.0 50.3 ± 23.1 0.0101* 33 vs 31 \0.0001* 0.4050 \0.0001*
Physical component and
mental component summaries
Effectiveness of psychological support in patients undergoing primary total hip or knee arthroplasty: a controlled...
Physical component summary Joint pop 93 vs 94 52.2 ± 7.5 48.8 ± 10.9 0.0135* 93 vs 94 0.0310* \0.0001* 0.0850
Hip pop 60 vs 63 53.7 ± 7.6 47.8 ± 11.4 0.0009* 60 vs 63 0.0048* \0.0001* 0.0201*
Knee pop 33 vs 31 49.5 ± 6.6 50.9 ± 9.8 0.5114 33 vs 31 0.5968 \0.0001* 0.8457
Mental component summary Joint pop 93 vs 94 46.6 ± 13.0 37.1 ± 13.6 \0.0001* 93 vs 94 \0.0001* 0.0002* \0.0001*
Hip pop 60 vs 63 47.1 ± 13.1 38.2 ± 14.4 0.0004* 60 vs 63 \0.0001* 0.0053* \0.0001*
Knee pop 33 vs 31 45.5 ± 13.1 35.1 ± 11.6 0.0013* 33 vs 31 \0.0001* 0.0024* \0.0001*

149
Fig. 1 Temporal trend of the ‘Physical Component Scale’ (PCS) and experimental group while the dashed lines indicate the control group.
the ‘Mental Component Scale’ (MSC) of the SF-36 questionnaire for Means are shown as circles while the bars represent the 95 %
the hip population ((a) and (c), respectively) and for the knee confidence interval. An evident overlapping between the bars
population ((b) and (d), respectively). The solid lines indicate the indicates lack of significant statistical difference
therefore also be an economic innovation. In fact, in in the case of this rehabilitation centre, the cost of the stay
addition to determining an improvement in the psycho- amounts to EUR 175 per day for each patient (current cost
physical well-being of the patient, it could bring about a as of 2014), early discharge by 1 day compared to the
reduction in costs of patient treatment as a consequence of current standard would correspond to a saving in rehabil-
the reduction in rehabilitation time at the rehabilitation itation costs of EUR 175 gross per patient. This saving
centre (currently, in the case of our facility, set at 8 days should be compared with the gross cost per patient for
following the 5-day post-operative stay in hospital). Con- psychological support, which is calculated at EUR 63 gross
sidering the outcome obtained in this study and given that, (taking into consideration a gross cost of EUR 31.50 per

Table 6 Physiotherapy results

EXP CTR p value
Delta autonomy days [mean ± SD (days)] All patients 7.2 ± 2.2 8.2 ± 2.3 0.0023*
Hip arthroplasty group 6.7 ± 1.8 7.9 ± 2.2 0.0015*
Knee arthroplasty group 8.1 ± 2.4 8.8 ± 2.3 0.2424
Calculation performed on 92 patients in the EXP group (59 hips; 33 knees) and 93 patients in the CTR
group (61 hips; 32 knees). Data analysed by Student’s t test and confirmed by Mann–Whitney test. p-values
refer to t test
hour and considering that each patient participated in four References

sessions, each lasting approximately half an hour). Making
the calculation with approximately 600 patients who 1. Brownlow HC, Benjamin S, Andrew JG, Kay P (2001) Disability
and mental health of patients waiting for total hip replacement.
undergo primary hip replacement each year at our facility,
Ann R Coll Surg Engl 83:128–133
the gross total annual saving would amount to EUR 67,200. 2. Brull R, McCartney CJ, Chan VW (2002) Do preoperative anx-
In summary, in the patients who received psychological iety and depression affect quality of recovery and length of stay
support, a lower incidence of anxiety and depression and after hip or knee arthroplasty? Can J Anaesth 49:109
3. Kurlowicz LH (1998) Perceived self-efficacy, functional ability,
better mental well-being was observed compared to the
and depressive symptoms in older elective surgery patients. Nurs
patients who did not receive this therapy. In the patients Res 47:219–226
who underwent hip arthroplasty, a reduction of an average 4. Lingard EA, Riddle DL (2007) Impact of psychological distress
of 1.2 days in the period to reach the physiotherapy on pain and function following knee arthroplasty. J Bone Joint
Surg Am 89:1161–1169
objective was observed in the group that received psy-
5. Pacault-Legendre V, Anract P, Mathieu M, Courpied JP (2009)
chological support compared to the control group. Pain after total hip arthroplasty: a psychiatric point of view. Int
This study is significant because, to the best of our Orthop 33:65–69
knowledge, it is the first controlled study in this therapeutic 6. Riediger W, Doering S, Krismer M (2010) Depression and
somatisation influence the outcome of total hip replacement. Int
field. It would be interesting to design a study focused on
Orthop 34:13–18
patients with more complex diagnoses (for example 7. Dorr LD, Chao L (2007) The emotional state of the patient after
patients undergoing revision surgery), or by comparing total hip and knee arthroplasty. Clin Orthop Relat Res 463:7–12
protocols with a different number of psychological support 8. Hall GM, Salmon P (2002) Physiological and psychological
influences on postoperative fatigue. Anesth Analg 95:1446–1450
visits to determine which protocol could be the most cost-
9. Salmon P, Hall GM, Peerbhoy D (2001) Influence of the emo-
effective. tional response to surgery on functional recovery during
6 months after hip arthroplasty. J Behav Med 24:489–502
Acknowledgments The authors thank Giovanni Brusaferri, MSc 10. Caracciolo B, Giaquinto S (2005) Self-perceived distress and
(Clinical Research, Zimmer GmbH, Winterthur, CH) for reviewing self-perceived functional recovery after recent total hip and knee
and editing this manuscript. arthroplasty. Arch Gerontol Geriatr 41:177–181
11. Badura-Brzoza K, Zajac P, Kasperska-Zajac A, Brzoza Z,
Compliance with Ethical standards Matysiakiewicz J, Piegza M, Hese RT, Rogala B, Semenowicz J,
Koczy B (2008) Anxiety and depression and their influence on
Conflict of interest The authors declare no conflict of interest the quality of life after total hip replacement: preliminary report.
regarding this study. Int J Psychiatry Clin Pract 12:280–284
12. Badura-Brzoza K, Zajac P, Brzoza Z, Kasperska-Zajac A,
Ethical standards The Ethics Committee of ‘ASL 2 - Savonese’ Matysiakiewicz J, Piegza M, Hese RT, Rogala B, Semenowicz J,
examined the study protocol and issued a favourable opinion for the Koczy B (2009) Psychological and psychiatric factors related to
start of the study in the meeting held on 20 December 2010. All health-related quality of life after total hip replacement—pre-
procedures were in accordance with the ethical standards of the liminary report. Eur Psychiatry 24:119–124
institutional research committee and were performed in accordance 13. Nickinson RS, Board TN, Kay PR (2009) Post-operative anxiety
with the ethical standards of the Declaration of Helsinki as revised in and depression levels in orthopaedic surgery: a study of 56
2008. Each patient provided their own consent to participate in the patients undergoing hip or knee arthroplasty. J Eval Clin Pract
study and for the processing of personal data. 15:307–310
14. Mumford E, Schlesinger HJ, Glass GV (1982) The effect of
psychological intervention on recovery from surgery and heart
attacks: an analysis of the literature. Am J Public Health
72:141–151
15. McArdle JM, George WD, McArdle CS, Smith DC, Moodie AR,
Hughson AV, Murray GD (1996) Psychological support for
patients undergoing breast cancer surgery: a randomised study.
BMJ 312:813–816

16. Kinzl JF, Trefalt E, Fiala M, Biebl W (2002) Psychotherapeutic which is a better predictor of outcome after total hip or knee
treatment of morbidly obese patients after gastric banding. Obes arthroplasty? Patient Educ Couns 66:92–99
Surg 12:292–294 20. Zigmond AS, Snaith RP (1983) The hospital anxiety and
17. Majani G (1999) Introduzione alla psicologia della salute. Centro depression scale. Acta Psychiatr Scand 67:361–370
Study Erickson, Trento 21. Lupo A, Trabucco G, Boaretti C, Rugiu C, Loschiavo C, Fontana
18. Johansson K, Salantera S, Katajisto J (2007) Empowering L, Bravi E, Magalini A, Abaterusso C, DeBiase V, Gambaro G
orthopaedic patients through preadmission education: results (2008) Quality of life of the elderly patient on dialysis. G Ital
from a clinical study. Patient Educ Couns 66:84–91 Nefrol 25:708–712
19. van den Akker-Scheek I, Stevens M, Groothoff JW, Bulstra SK,
Zijlstra W (2007) Preoperative or postoperative self-efficacy:

22
Blood metal ions after hybrid metal-on-polyethylene
Exeter2Trident total hip replacement
Rohit Singh1 • Gopikanthan Manoharan1 • Pete Craig1 • Simon Collier2 •
Phillip Shaylor2 • Ashok Sinha2
Abstract the corresponding findings in the Trident-Accolade sys-

Background Metal-on-metal total hip replacements tem in our previous study. This study provides relative
(THRs) with large femoral heads have been associated with reassurance that the issue does not lie with the V40 taper
elevated levels of cobalt (Co) and chromium (Cr), which trunnion, but raises suspicion that the issue may be with the
have been attributed to high levels of wear at the articular titanium Accolade stem with large diameter femoral heads.
surface. Our unit recently published data showing a sig- Level of evidence III.
nificant increase in the mean levels of Co ions in patients
with a 36-mm diameter femoral head with the metal-on- Keywords Hip Metal Ions Exeter Hybrid Trident
polyethylene Trident-Accolade system. The aim of this
study is to assess the levels of Co and Cr in the
Exeter-Trident hybrid system, as similar findings would Introduction
raise concern over the V40 taper trunnion.
Materials and methods The study included 83 patients Metal-on-metal (MoM) total hip replacements (THRs) with
(45 male and 38 female with a mean age of 75.6 years) who large femoral heads have been associated with elevated
received Exeter-Trident hybrid metal-on-polyethylene levels of cobalt (Co) and chromium (Cr) [1–7]. These ele-
THRs. The patients were then divided into two groups vated levels have been previously attributed to high levels of
according to the diameter of the femoral head used—38 wear at the articular surface which have been linked to a
patients in the 28-mm group (control), and 45 in the 36-mm causative factor in pseudotumour formation [8–12]. Other
(experimental) group. Serum levels of blood Co and Cr complications associated with increased metal ions include
were analysed for all recruited patients. metal hypersensitivity, cardiomyopathy, aseptic lympho-
Results In the control group (28-mm femoral head) all Co cyte-dominated vasculitis-associated lesions, and tubular
and Cr values were normal (under the abnormal threshold), necrosis [1–7]. There has been recent evidence in the liter-
as were the experimental group (36-mm femoral head). The ature on the possibility of the trunnion as a potential source
data values were below \10 nmol and \40 nmol for Co of metal ions; however, much of this evidence is received
and Cr, respectively. from retrieval analysis of large head MoM THRs [13, 14].
Conclusion Since the National Joint Registry (NJR) All patients with large head MoM THRs had levels of
states that the Exeter femoral stem is the commonest Co and Cr ions measured, as advised by the Medicines and
cemented femoral stem prosthesis used in the UK, we Healthcare products Regulatory Agency (MHRA) [15]. If
found it imperative that these results are documented given these levels were above seven parts per billion (ppb;
equivalent to Co 119 nmol/l and Cr 134.5 nmol/l) on two
& Rohit Singh
samples, collected 3 months apart, then clinical findings
[email protected] along with cross-sectional imaging should be used to
consider revision surgery [15].
1
Robert Jones Orthopaedic Hospital, Oswestry, UK Co is a constituent of Vitamin B12 (cobalamin), a co-
2
Mid Staffordshire foundation trust, Stafford, England, UK factor in haemoglobin metabolism [16] and Cr regulates the

action of insulin, and both are important for general health. Exeter cemented femoral stem with the Trident acetabular
The pathological effects that have been reported in patients cup, as this would imply an issue with V40 stem [28].
with raised levels of these ions following THR include visual There has been no study of our magnitude directly
disturbance, neurological impairment, auditory symptoms, comparing the levels of metal ions in the blood in small
cardiomyopathy, deep vein thrombosis, and pseudotumours and large Exeter-Trident hybrid system MoP THRs. The
[17–21]. There is a risk of contamination during the sepa- aim of this study is to assess the levels of Co and Cr in the
ration process when measuring levels of Co and Cr in serum Exeter-Trident hybrid system as a follow-up study to see
of plasma. Therefore, whole blood collected in ethylenedi- if there is a similar effect of raised metal ions using a
aminetetraacetic acid or heparin is used [18]. combination of different head sizes with the V40 taper
In the tenth Annual Report of the National Joint Registry trunnion as seen in the Trident-Accolade THR.
(NJR) for England and Wales, approximately 51 % of
primary femoral stems in THRs in 2013 were reported as
cemented fixation using stainless steel prosthesis. There Materials and methods
has also been a steady rise in the use of large diameter
(36 mm) femoral heads, from approximately 5 % of all All patients who underwent Exeter-Trident hybrid MoP
heads to [20 % over the past 10 years [22]. This increase THRs in 2009 and 2010 were identified using the depart-
in the use of large diameter heads reflects the literature mental database. Two patients had died and were excluding
which states that larger heads reduce the rate of THR leaving 97 patients who were contacted by letter, which
dislocation [23, 24]. In our unit, hybrid THR is undertaken explained the purpose of the study and requested their
using the Exeter-Trident system (Stryker Orthopaedics). participation. This required them to undergo blood Co and
Nationally, the Exeter femoral stem is the most commonly Cr ion level measurements . A total of 83 patients agreed to
used cemented femoral stem in the UK. It is a polished, take part in our study. All 83 patients had undergone rou-
collarless tapered stem comprised of ‘Orthinox’, a propri- tine follow-up and all had satisfactory outcome at the last
etary stainless steel, which is issued with a V40 taper routine clinical review, which was confirmed by validated
trunnion. The Trident acetabular component is a two-piece outcome measures. Ethical approval was obtained and
component with a hydroxyapatite-coated titanium alloy implied informed consent was completed by all patients
shell which is manufactured for press-fit following line-to- prior to enrolment.
line reaming. The system utilises a unique locking mech- Enclosed with the initial recruitment letter was an
anism providing a secure interface between the poly- information leaflet regarding the blood test that was pro-
ethylene (or ceramic) insert and the metallic shell [25]. duced by the laboratory, and a pre-filled clinical chemistry
There is little in the literature about the incidence of request form [29].
metallosis as a result of head–trunnion wear following The 83 patients involved in our study were separated
metal-on-polyethylene (MoP) THR. Corrosion has been into two groups according to the diameter of the femoral
reported previously at metal junctions in THRs, and cases head which was used. A 28-mm head was used in 38
of pseudotumour formation attributed to metallosis in patients (twenty male and eighteen female with a mean age
uncemented MoP THRs [26, 27]. of 76.7 years), a 36-mm head in 45 (25 male and 20 female
Our unit has recently published data on a series of 69 with a mean age 74.5 years). The patients with a 28-mm
patients who received uncemented Trident-Accolade head were considered as a control group as we did not
metal on polyethylene THRs using 28- and 36-mm heads. expect the levels of metal ions in their blood to be raised, as
This study showed a significant increase in the mean levels our previous study with the Trident-Accolade THR found
of Co ions in the blood of those with a 36-mm diameter the levels for this head size to be normal. Therefore, the
femoral head compared to those with a 28-mm diameter patients with 36-mm femoral heads were considered the
head. The levels of Cr in the blood were normal in all experimental group.
patients. The clinical significance of this study to our unit Venesection was performed either at our hospital or in
was the suspension of use of 36-mm femoral heads for the primary care. The samples were analysed at an external
Trident-Accolade system. The stimulus for carrying out (clinical pathology accredited) laboratory [28]. The
that study was a case of severe pain following Tri- patients treated in 2011 and 2012 were chosen for inves-
dent-Accolade THR with a large head. This pain was tigation as it was felt this was a sufficient period of time for
secondary to severe corrosion at the head-trunnion inter- the ‘wearing-in’ of the prosthesis. The measurement of the
face leading to damage at the abductor insertion on the levels of metal ions in the blood occurred at a mean of
greater trochanter [28]. We then needed to see if this issue 32 months after the THR.
of raised metal ions particularly affects this combination of Results were accessed from the hospital’s electronic
components in other commonly used prosthesis, such as the record system and tabulated using an Excel spreadsheet

Blood metal ions after hybrid metal-on-polyethylene Exeter–Trident total hip replacement 155
(Microsoft, Redmond, WA, USA). Abnormal results were our study [31–33]. Factors that have been previously
defined as Co levels [10 nmol/l (0.59 ppb) and Cr levels shown to affect metal ion levels in patients with hip
[40 nmol/l ([2.07 ppb), with normal ranges for adults resurfacing and MOM THR such as gender and activity
without THR being below these thresholds. It should be level were not controlled in our study [34].
noted that the results produced by the laboratory did not The findings from the recent previous study carried out
state the exact value if the result was \10 nmol/l and at our unit showed increased Co levels in patients with a
\40 mmol/l for Co and Cr, respectively. 36-mm head in comparison to patients with a 28-mm head
To our knowledge the exact mechanism of wear at the in Trident-Accolade MoP THRs [28]. These findings
trunnion has yet to be fully determined. It is likely that it is changed the practice in our unit with the cessation of usage
a combination of both frictional wear and corrosion. Cer- of 36-mm femoral heads for that specific prosthesis. These
tain factors such as the neck offset, head length and outer findings also raised concern as to whether there was an
diameter of the acetabular components, all of which issue with the V40 taper trunnion with larger heads. The
influence the lever arm and frictional torque applied to the findings from our current study provide relative reassur-
trunnion, would benefit from their control. This is a limi- ance that there is no issue with the V40 taper trunnion in
tation of our study. the Exeter orthinox prosthesis [28]. It would seem that the
The data was analysed using the SPSS software v20 results from both our studies suggest that the issue may lie
(IBM, Armonk, NY, USA). with the titanium femoral stem in the Accolade system as
opposed to the V40 taper trunnion, as the same trunnion
with the orthinox (stainless steel proprietary) Exeter stem
Results caused no increase in metal ion level regardless of size of
the femoral head. We intentionally ensured that all the
In the control group (28-mm femoral head) all Co and Cr Exeter femoral stems used in our study were hybrid sys-
values were normal (under the abnormal threshold), as tems with the Trident acetabular cup and can therefore
were the experimental group (36-mm femoral head) group certify as much interrelation between the two studies as
(see Tables 1, 2). possible. These results have been shared with the
manufacturer.
A previous randomised blinded clinical trial was carried
Discussion out to assess polyethylene versus metal bearing surfaces in
THR. This study included forty-one patients with identical
In this small series, patients with a 36-mm diameter femoral and acetabular components. Erythrocyte Co and Cr
modular femoral head following an Exeter-Trident hybrid levels were measured and were significantly higher in
MoP THR did not have higher mean levels of Co than patients with MOM articulations, with an average of 7.9-
those with 28-mm heads. We found no increase in the fold increase in erythrocyte Co and a 2.3-fold increase in
mean levels of Cr ions in the blood of patients with an erythrocyte Cr [35]. However, there were still detectable
increased diameter femoral head. In fact, the values for levels with the polyethylene inserts.
both groups were within the normal range for both Co and The effect of taper design on trunnionosis has been
Cr. extensively evaluated. A comprehensive study looked at all
It has been documented in the literature that there is an hip prosthesis with a 28-mm plus zero length head over
issue of wear at the interface between the head and trun- fourteen years with MoP implants giving a total of forty-
nion which has been associated with large head MoM four sets of retrieved implants, with six different taper
THRs [30]. There is a potential source of metal ion release designs. The Goldberg scale was used to score fretting and
in all metal modular junctions, which means that it is corrosion. This study showed no difference in patient age,
unlikely that the bearing surface is the only contributor body mass index, or length of implantation with regard to
[31]. the trunnionosis. However, Taper design had a significant
Certain factors that have been shown to increase poly- effect on corrosion at the base of the trunnion [36].
ethylene wear such as acetabular inclination angles of[45° The results of this study are consistent with another
and failure to restore femoral offset were not considered in study which only looked at twenty patients one year after
Table 1 Mean and medium

28-mm femoral head 36-mm femoral head
blood ion levels of cobalt
Mean cobalt level (range) nmol/l \10 \10
Median cobalt level (range) nmol/l \10 \10

Table 2 Mean and medium

28-mm femoral head 36-mm femoral head
blood ion levels of chromium
Mean chromium level (range) nmol/l \40 \40
Median chromium level (range) nmol/l \40 \40
surgery with the Exeter V40 stem with a variety of 5. Van Der Straeten C, Van Quickenborne D, De Roest B et al
acetabular components. Whole blood Cr levels were within (2013) Metal ion levels from well-functioning Birmingham Hip
Resurfacings decline significantly at 10 years. Bone Joint J
normal limits and only one patient exhibited mild elevation 95-B:1332–1338
of serum Co [37]. The study carried out at our unit had a 6. Malek IA, King A, Sharma H et al (2012) The sensitivity,
much larger series of patients and only looked at one specificity and predictive values of raised plasma metal ion levels
acetabular component. in the diagnosis of adverse reaction to metal debris in symp-
tomatic patients with a metal-on-metal arthroplasty of the hip.
The clinical significance of the findings of our study J Bone Joint Surg Br 94-B:1045–1050
remain uncertain, but as the NJR states that the Exeter 7. Holland JP, Langton DJ, Hashmi M (2012) Ten-year clinical,
femoral stem is the commonest cemented femoral stem radiological and metal ion analysis of the Birmingham Hip
prosthesis used in the UK, we found it imperative that these Resurfacing: from a single, non-designer surgeon. J Bone Joint
Surg Br 94-B:471–476
results are documented given the corresponding findings in 8. Davies AP, Willert HG, Campbell PA, Learmonth ID, Case CP
the Trident-Accolade system in our previous study. (2005) An unusual lymphocytic perivascular infiltration in tissues
around contemporary metal-on-metal joint replacements. J Bone
Compliance with ethical standards Joint Surg Am 87-A:18–27
9. Campbell P, Ebramsadeh E, Nelson S et al (2010) Histological
Ethical standards This study conforms to the last version of the features of pseudotumour-like tissues from metal-on-metal hips.
Declaration of Helsinki, was approved by the Ethical Committee, and Clin Orthop Relat Res 468:2321–2327
implied informed consent from all the patients prior to their 10. Bosker BH, Ettema HB, Boomsma MF et al (2012) High inci-
enrolment. dence of pseudotumour formation after large-diameter metal-on-
metal total hip replacement: a prospective cohort study. J Bone
Conflict of interest The authors whose names are listed on this Joint Surg Br 94-B:755–761
paper certify that they have no affiliations with or involvement in any 11. Canadian Arthroplasty Society (2013) The Canadian Arthroplasty
organisation or entity with any financial interest (such as honoraria, Society’s experience with hip resurfacing arthroplasty: an anal-
educational grants, participation in speakers’ bureaus, membership, ysis of 2773 hips. Bone Joint J 95-B:1045–1051
employment, consultancies, stock ownership, or other equity interest, 12. Lombardi AV Jr, Barrack RL, Berend KR et al (2012) The Hip
and expert testimony or patent-licensing arrangements), or non-fi- Society: algorithmic approach to diagnosis and management of
nancial interest (such as personal or professional relationships, affil- metal-on-metal arthroplasty. J Bone Joint Surg Br 94-B(Suppl
iations, knowledge or beliefs) in the subject matter or materials A):14–18
discussed in this manuscript. 13. Bolland BJ, Culliford DJ, Langton DJ et al (2011) High failure
rates with a large-diameter hybrid metal-on-metal total hip
replacement: clinical, radiological and retrieval analysis. J Bone
Joint Surg Br 93-B:608–615
14. Malviya A, Ramaskandhan JR, Bowman R et al (2011) What
advantage is there to be gained using large modular metal-on-
metal bearings in routine primary hip replacement? A preliminary
report of a prospective randomised controlled trial. J Bone Joint
Surg Br 93-B:1602–1609
15. No authors listed. (2012): Medicines and health products regu-
latory agency. Management recommendations for patients with
metal-on-metal hip replacement implants. http://www.mhra.gov.
References uk/home/groups/dts-bs/documents/medicaldevicealert/con155766.
pdf. Accessed 20 Sept 2013
1. Vendittoli PA, Roy A, Mottard S et al (2010) Metal ion release 16. Takahashi-Iñiguez T, Garcı́a-Hernandez E, Arreguı́n-Espinosa R,
from bearing wear and corrosion with 28 mm and large-diameter Flores ME (2012) Role of vitamin B(12) on methylmalonyl-CoA
metal-on-metal bearing articulations. J Bone Joint Surg Br mutase activity. J Zhejiang Univ Sci B 13:423–437
92-B:12–19 17. Hummel M, Standl E, Schnell O (2007) Chromium in metabolic
2. Clarke MT, Lee PT, Arora A, Villar RN (2003) Levels of metal and cardiovascular disease. Horm Metab Res 39:743–751
ions after small and large-diameter metal-on-metal hip arthro- 18. Egan K, Di Cesare PE (1995) Intraoperative complications of
plasty. J Bone Joint Surg Br 85-B:913–917 revision hip arthroplasty using a fully porous-coated straight
3. Haddad FS, Thakrar RR, Hart AJ et al (2011) Metal-on metal cobalt—chrome femoral stem. J Arthroplasty 10(Suppl 1):S45–S51
bearings: the evidence so far. J Bone Joint Surg Br 93-B:572–579 19. Shimmin A, Bare J, Back D (2005) Complications associated with
4. Langton DJ, Jameson SS, Joyce TJ et al (2010) Early failure of hip resurfacing arthroplasty. Orthop Clin North Am 36:187–193
metal-on-metal bearings in hip resurfacing and large-diameter 20. Gessner BD, Steck T, Woelber E, Tower SS (2015) A Systematic
total hip replacement: a consequence of excess wear. J Bone Joint Review of Systemic Cobaltism after Wear or Corrosion of
Surg Br 92-B:38–46 Chrome-Cobalt Hip Implants. J Patient Saf [Epub ahead of print]

Blood metal ions after hybrid metal-on-polyethylene Exeter–Trident total hip replacement 157
21. Steens W, von Foerster G, Katzer A (2006) Severe cobalt poi- of a prospective longitudinal study. J Bone Joint Surg Br
soning with loss of sight after ceramic-metal pairing in a hip: a 89(2):169–173
case report. Acta Orthop 77:830–832 30. Hexter A, Panagiotidou A, Singh J, Skinner J, Hart A (2013)
22. Porter M, Borroff, Gregg et al. (2013) National Joint Registry for Corrosion at the head: trunnion taper interface in large diameter
England and Wales 10th annual report. www.njrcentre.org.uk head metal on metal total hip arthroplasty: a comparison of
23. Jameson SS, Lees D, James P et al (2011) Lower rates of dis- manufacturers. Bone Joint J 95-B(Suppl 12):3
location with increased femoral head size after primary total hip 31. Little NJ, Busch CA, Gallagher JA, Rorabeck CH, Bourne RB
replacement: a five-year analysis of NHS patients in England. (2009) Acetabular poly-ethylene wear and acetabular inclination
J Bone Joint Surg Br 93-B:876–880 and femoral offset. Clin Orthop Relat Res 67:2895–2900
24. Bistolfi A, Crova M, Rosso F et al (2011) Dislocation rate after 32. Devane PA, Horne JG (1999) Assessment of polyethylene wear
hip arthroplasty within the first postoperative year: 36 vs 28 mm in total hip replacement. Clin Orthop Relat Res 369:59–72
femoral heads. Hip Int 21:559–564 33. Sakalkale DP, Sharkey PF, Eng K, Hozack WJ, Rothman RH
25. Williams HD, Browne G, Gie G, Ling R, Timperley A, Wen- (2001) Effect of femoral offset on polyethylene wear in total hip
dover N (2002) The Exeter universal cemented femoral compo- replacement. Clin Orthop Relat Res 388:125–134
nent at 8–12 years. J Bone Joint Surg Br 84-B:324–334 34. Heisel C, Silva M, Skipor AK, Jacobs JJ, Schmalzried TP (2005)
26. Svensson O, Mathiesen EB, Reinholt FP, Blomgren G (1988) The relationship between activity and ions in patients with metal-
Formation of a fulminant soft-tissue pseudotumour after unce- on-metal bearing hip prostheses. J Bone Joint Surg Am
mented hip arthroplasty: a case report. J Bone Joint Surg Am 87-A:781–787
70-A:1238–1242 35. MacDonald SJ, McCalden RW, Chess DG, Bourne RB, Rorabeck
27. Cooper HJ, Valle Della CJ, Berger RA, Tetreault M, Paprosky CH, Cleland D, Leung F (2003) Metal-on-metal versus poly-
WG, Sporer SM et al (2012) Corrosion at the head-neck taper as a ethylene in hip arthroplasty: a randomized clinical trial. Clin
cause for adverse local tissue reactions after total hip arthroplasty. Orthop Relat Res 406:282–296
J Bone Joint Surg Am 94(18):1655–1661 36. Tan SC, Teeter MG, Del Balso C, Howard JL, Lanting BA (2015)
28. Craig P, Bancroft G, Burton A, Collier S, Shaylor P, Sinha A Effect of taper design on trunnionosis in metal on polyethylene
(2014) Raised levels of metal ions in the blood in patients who total hip arthroplasty. J Arthroplast. doi:10.1016/j.arth.2015.02.
have undergone uncemented metal-on-polyethylene Trident-Ac- 031
colade total hip replacement. Bone Joint J 96:43–47 37. Vijaysegaran P, Whitehouse SL, Bijoor M, English H, Crawford
29. Daniel J, Ziaee H, Pradhan C, Pynsent PB, McMinn DJW (2007) RW (2014) Metal ion levels post primary unilateral Exeter total
Blood and urine metal ion levels in young and active patients hip arthroplasty. Hip Int 24(2):144–148
after Birmingham hip resurfacing arthroplasty. Four-year results

23
Intraosseous concentration and inhibitory effect of different
intravenous cefazolin doses used in preoperative prophylaxis
of total knee arthroplasty
Chayanin Angthong1 • Pongpaibool Krajubngern1 • Warawut Tiyapongpattana2 •
Boonchana Pongcharoen1 • Piya Pinsornsak1 • Nattapol Tammachote1 •

Wanna Kittisupaluck3
Abstract (r = 0.18, p = 0.52) and distal femur (r = -0.29,

Background The aim of this study was to compare the p = 0.30).
intraosseous concentrations and the inhibitory effects on Conclusion IV cefazolin at a dose of 2 g produced
the growth of Staphylococcus aureus of 1 g versus 2 g of greater intraosseous concentrations overall than a dose of
intravenous (IV) prophylactic cefazolin in total knee 1 g. However, the higher intraosseous concentrations did
arthroplasty (TKA). not correlate with higher inhibitory effects.
Materials and methods Eighteen patients (21 knees) with Level of evidence Level III.
primary knee osteoarthritis were divided into two groups
receiving 1 g (12 patients: 14 knees) versus 2 g (six Keywords Cefazolin Knee arthroplasty Prophylaxis
patients: seven knees) IV prophylactic cefazolin prior to Intraosseous concentration Inhibitory Effect
the incision in TKA. Subchondral bone samples (proximal
tibia, distal femur) were taken during the operation. These
samples were analyzed for intraosseous concentration of Introduction
cefazolin and their inhibitory effects on the growth of S.
aureus, using high-performance liquid chromatography Prophylactic antibiotics are known to reduce the risk of
(HPLC) and agar disc diffusion bioassays. perioperative and/or postoperative infection [1–3]. How-
Results The mean intraosseous concentration in the 2 g ever, some previous studies had reported that systemic
dose group was significantly higher than in the 1 g dose antibiotics may not prevent all postoperative infections [4–
group in the proximal tibia (p = 0.007) and distal femur 6]. Moreover, conventional systemic dosages may not
(p = 0.016). There were no significant differences between provide adequate tissue concentrations against more resis-
the two groups in terms of mean inhibitory effects in the tant organisms, such as coagulase-negative staphylococci
proximal tibia or distal femur (p [ 0.05). No significant [7]. The current literature recommends the intravenous (IV)
correlations were found between the intraosseous concen- administration of cefazolin, 1–2 g [1], within 1 h prior to
trations and inhibitory effects in the proximal tibia making the incision. This antibiotic may be repeated every
2–5 h during the operation and should be stopped within
24 h following the operation [8, 9]. However, little is
known about the differences in the intraosseous concen-
& Chayanin Angthong trations of cefazolin or the inhibitory effects on the growth
[email protected]
of Staphylococcus aureus between an IV dose of 1 g and a
1
Department of Orthopaedics, Faculty of Medicine, dose of 2 g.
Thammasat University, Pathum Thani 12120, Thailand This study aims to compare the intraosseous concen-
2
Department of Chemistry, Faculty of Science and trations and the inhibitory effects on the growth of S.
Technology, Thammasat University, Pathum Thani, Thailand aureus of IV prophylactic cefazolin at dosages of 1 versus
3
The Surgical Unit, Thammasat University Hospital, 2 g in total knee arthroplasty. At this point, we hypothesize
Pathum Thani, Thailand that the intraosseous concentrations and the inhibitory

Intraosseous concentration and inhibitory effect of different intravenous cefazolin doses used in preoperative... 159
effects in the group given cefazolin at a dose of 2 g are Each sample was processed by the antibiotic broth
possibly higher than those in the group given a dose of 1 g. elution assay [10]. The solution from a previous process
was further analyzed for concentration using high-perfor-
mance liquid chromatography–photodiode array detection
Materials and methods [HPLC–DAD; Shimadzu (Nexara LC-30A)]. The unit of
this intraosseous concentration was ‘lg/g’, which was
During the period between May, 2011 and February, 2013, derived from the comparison of a 1 g sample of the sub-
patients with primary knee osteoarthritis were recruited to chondral bone. A validation study was done of the
participate in this study. The inclusion criteria were patients extraction and HPLC–DAD technique using bone samples
with primary knee osteoarthritis while the exclusion criteria analyzed with known concentrations of cefazolin. All
included patients with post-traumatic or post-infectious samples were analyzed by a specialist (W.T.) who was
knee conditions, allergies to cephalosporin or penicillin, blinded to the IV cefazolin dose in each sample.
serum creatinine levels [1.5 mg%, creatinine clearance The bioactivity of each sample was determined using an
\55 ml/min, probenecid intake, post-steroid treatment, agar disc diffusion bioassay with S. aureus (ATCC 25923)
chemotherapy, those who were post-knee arthroplasty or [10]. The antibiotic bioassay was to determine the inhibi-
high tibial osteotomy, immunocompromised hosts, and tory effect of intraosseous cefazolin, which was compared
those with deleterious medical conditions. Eighteen patients with the standard minimal inhibitory concentration (MIC)
were recruited into our study using the inclusion and of serum cefazolin (30 lg/ml). The technique was based on
exclusion criteria. The 18 patients (21 knees) were divided the inhibitory activity of discs (Oxoid, UK) containing a
into two groups in accordance with their IV prophylactic standard concentration of cefazolin. Standard paper discs
cefazolin dosages (cefazolin M. H., M & H Manufacturing and samples were placed on S. aureus-seeded agar (Muller
Co., Ltd., Thailand) of 1 g (12 patients: 14 knees, which Hinton Agar, BD, USA) and incubated for 18 h at 37 °C.
were operated by B.P. and P.P.) versus 2 g (six patients: All samples were analyzed by our researcher (N.M.) who
seven knees, which were operated by N.T.) administered was blinded to the IV cefazolin dose in each sample. The
before the incision was made in their total knee arthroplasty. standard MIC of serum cefazolin would inhibit the growth
All patients’ baseline data were prospectively collected, of S. aureus at least 18 mm from the center of the sample.
including the intraosseous concentrations of cefazolin and All samples were analyzed by a staff person in the labo-
the inhibitory effects on the growth of S. aureus, which ratory who was also blinded to the IV cefazolin dose used
were prospectively collected from further analyses. The in each sample.
group with the IV cefazolin dose of 1 g consisted of nine
females and three males with a mean age of Statistical analyses
70.1 ± 4.6 years (61–79 years of age). During the same
period we retained a group with an IV cefazolin dose of 2 g. Statistical analysis was performed using the SPSS software
This group consisted of five females and one male with a version 13.0 (SPSS Inc., Chicago, IL, USA). ANOVA was
mean age of 68.4 ± 3.0 years (64–73 years of age). There used to analyze the statistical significance of the differ-
were no significant differences in the mean age and genders ences in the values of the intraosseous concentration of
between the two groups (p = 0.40 for age, p = 1.00 for cefazolin (lg/g) and the inhibitory effect on S. aureus
gender). The mean weights of the patients were growth on the agar disc (mm) between the two different
61.2 ± 8.4 kg for the group with a 1 g dose and groups of IV cefazolin doses (1 versus 2 g). The correla-
62.0 ± 8.2 kg for the group with a dose of 2 g (p = 0.84). tions between the intraosseous concentrations of cefazolin
All patients were prepared for the TKA operation with (lg/g) and the inhibitory effects of S. aureus growth on
standard, sterile technique. They were administered IV agar disc (mm) were analyzed and interpreted via Pear-
cefazolin, at a dose of 1 or 2 g, prior to tourniquet inflation son’s correlation coefficient (r). Categorical variables were
and before the incision had been made. After performing analyzed using the Chi-squared test. The level of signifi-
knee arthrotomy and bone cutting at the distal femur and at cance was set at p \ 0.05.
the proximal tibia, our experienced knee surgeons, B.P.,
P.P., and N.T. (with the same level of experience), pro-
vided bone from the segments that had been removed from Results
each patient to a researcher (P.K.) who extracted only
subchondral bone at a size of 2.5 9 2.5 mm for further The mean times between cefazolin injection and sample
analyses while under sterile technique. The period of time collection were 50.1 ± 6.5 min for femur and
between IV cefazolin injection and sample collection was 53.1 ± 6.5 min for tibia in the group with IV cefazolin 1 g,
recorded. compared with 58.9 ± 23.1 min for femur and

Table 1 Intraosseous
Parameters IV cefazolin 1 g IV cefazolin 2 g P value
concentrations of cefazolin in
the two groups Cefazolin concentration at proximal tibia (lg/g) 21.4 ± 11.4 35.5 ± 5.1 0.007*
Cefazolin concentration at distal femur (lg/g) 22.6 ± 8.7 44.1 ± 25.8 0.016*
* Significant difference
Table 2 Mean inhibitory

Parameters IV cefazolin 1 g IV cefazolin 2 g P value
effects in the two groups
Mean inhibitory effect at proximal tibia (mm) 11.1 ± 6.1 8.4 ± 4.2 0.33
Mean inhibitory effect at distal femur (mm) 12.2 ± 6.1 10.9 ± 4.6 0.63
67.9 ± 23.1 min for tibia in the group with IV cefazolin on the present study, we found that the mean intraosseous
2 g. There were no significant differences in these mean concentrations in the group receiving 2 g of cefazolin were
times for the same corresponding areas of sample collec- significantly higher than in the group receiving 1 g, at the
tion between the two groups (femur, p = 0.32; tibia, proximal tibia (p = 0.007) and distal femur (p = 0.016).
p = 0.11). However, the mean intraosseous concentrations of the 1 g
Total mean intraosseous concentrations of cefazolin group (22.6 lg/g) and the 2 g group (44.1 lg/g) at the
were 26.6 ± 11.7 lg/g in the proximal tibia and distal femurs were higher than the concentrations seen in a
30.5 ± 19.5 lg/g for the distal femur. The mean intraoss- previous study (9.2 lg/g) [7].
eous concentrations of the group receiving cefazolin 2 g On the other hand, there were no significant differences
were significantly higher than the group receiving cefazolin between the two groups in terms of the mean inhibitory
1 g in the proximal tibia (p = 0.007) and distal femur effects at either the proximal tibia or distal femur via agar
(p = 0.016) (Table 1). disc diffusion bioassay (proximal tibia, p = 0.33; distal
From Pearson’s correlation analyses, there were no femur, p = 0.63). Although a previous report had shown
significant correlations between the levels of intraosseous that high-dose cefazolin could be used for prophylaxis in
concentration and inhibitory effects seen in the proximal an animal study [11], surgeons should keep in mind that,
tibia (r = 0.18, p = 0.52) and distal femur (r = -0.29, based on our findings, when they consider IV cefazolin at
p = 0.30). a dose of 2 g for preoperative prophylaxis, it should be
The total mean inhibitory effects of cefazolin were used with care. This is especially true for patients with
10.0 ± 5.4 mm for the proximal tibia and 11.6 ± 5.4 mm reduced renal function. In addition to the results that
for the distal femur. The mean inhibitory effects in the two showed no significant differences between the two groups
groups (cefazolin 1 g versus 2 g at both proximal tibia and in terms of mean inhibitory effects, we found that the
distal femur) were less than 34.6 ± 0.5 mm, which was mean inhibitory effects in the two groups (cefazolin dose
shown to be the mean inhibitory effect of the standard MIC of 1 versus 2 g at both the proximal tibia and distal femur)
of serum cefazolin (30 lg/ml) in this present study were less than the mean inhibitory effect of the standard
(Table 2). There were no significant differences between MIC of serum cefazolin at 30 lg/ml. Pearson’s correla-
the two groups in terms of the mean inhibitory effects at the tion analyses showed that there were no significant cor-
proximal tibia or distal femur via the analyses on agar disc relations between the intraosseous concentrations and the
diffusion bioassay (proximal tibia, p = 0.33; distal femur, inhibitory effects in either the proximal tibia (r = 0.18,
p = 0.63). p = 0.52) or distal femur (r = -0.29, p = 0.30). Higher
cefazolin dosages would not provide higher inhibitory
effects. At this point, we hypothesize that the standard
Discussion MIC of intraosseous cefazolin may not be directly related
to the inhibitory effects, which are different from the
Although current literature recommends administration of relationship between the standard MIC of serum cefazolin
a dosage of intravenous (IV) cefazolin of 1–2 g [1] as the and its inhibitory effects via the agar disc diffusion
dose for prophylaxis in TKA, there is, as far as we know, a bioassay. In addition, the intraosseous area may contain a
lack of basic research that has shown the actual intraoss- threshold inhibitory effect of cefazolin that gives no
eous concentrations and inhibitory effects of cefazolin at greater inhibitory effect from a higher dose of cefazolin.
those recommended doses. Our study is possibly one of the Further studies are needed to test the inhibitory effects of
earliest reports on the importance of these aspects. Based intraosseous cefazolin directly with the standard MIC of

Intraosseous concentration and inhibitory effect of different intravenous cefazolin doses used in preoperative... 161
intraosseous cefazolin rather than relative tests with the Ethical standards The present study was authorized by the local
standard MIC of serum cefazolin. ethical committee and was performed in accordance with the ethical
standards of the 1964 Declaration of Helsinki as revised in 2000.
There were some limitations in the present study. First Written informed consent was obtained from all patients prior to
of all, bone resection at the distal femur and at the proximal study.
tibia for the analyses of intraosseous concentrations and
inhibitory effects of cefazolin were performed only once
after knee arthrotomy. There was no subsequent bone
sampling for further analyses of the same parameters at the
end of the operation. Therefore, the change in intraosseous
concentration of cefazolin over the operating time could
not be determined at this point. A possible decrease in
concentration at the end of wound closure, including the
periods when suturing and insertion of drains may be a References
source of contamination, was not clarified in the study. The
protective effect of the higher dose (2 g) of cefazolin may 1. Holtom PD (2006) Antibiotic prophylaxis: current recommen-
be more beneficial than that of the lower dose (1 g) due to dation. J Am Acad Orthop Surg 14:S98–S100
the longer persistence of the inhibitory effect, particularly 2. Ridgeway S, Wilson J, Charlet A, Kafatos G, Pearson A, Coello
R (2005) Infection of the surgical site after arthroplasty of the hip.
in the final period of the operation. A previous study found J Bone Joint Surg Br 87:844–850
that biofilm formation could develop for up to 1–2 days 3. Fletcher N, Sofianos D, Berkes MB, Obremskey WT (2007)
[12]; therefore, hypothetically, the higher dose (2 g) of Prevention of perioperative infection. J Bone Joint Surg Am
cefazolin might be more beneficial than the lower dose of 89:1605–1618
4. Segawa H, Tsukayama DT, Kyle RF, Becker DA, Gustilo RB
1 g. Second, the recommendations for surgeons from the (1999) Infection after total knee arthroplasty. A retrospective
results of the present study might appear to be weak study of the treatment of eighty-one infections. J Bone Joint Surg
because of the lack of evidence concerning relationships Am 81:1434–1445
between the intraosseous concentration of cefazolin, its 5. Wilson MG, Kelley K, Thornhill TS (1990) Infection as a com-
plication of total knee replacement arthroplasty. Risk factors and
inhibitory effect and the related infection. This issue relates treatment in sixty-seven cases. J Bone Joint Surg Am 72:878–883
to the previous limitation mentioned, as there was no 6. Windsor RE, Bono JV (1994) Infected total knee replacements.
clarification of those relationships during the final period of J Am Acad Orthop Surg 2:44–53
the operation. Future study may help to clarify these issues. 7. Young SW, Zhang M, Freeman JT, Vince KG, Coleman B (2013)
Higher cefazolin concentrations with intraosseous regional pro-
IV cefazolin at a dose of 2 g provides greater intraoss- phylaxis in TKA. Clin Orthop Relat Res 471(1):244–249
eous concentrations in both the proximal tibia and distal 8. Prokuski L (2008) Prophylactic antibiotics in orthopaedic sur-
femur than does a dose of 1 g. However, its higher gery. J Am Acad Orthop Surg 16(5):283–293
intraosseous concentration does not correlate with higher 9. American Academy of Orthopaedic Surgeons (2004) Information
statement: recommendations for the use of intravenous antibiotic
inhibitory effects. Surgeons have to weigh the advantages prophylaxis in primary total joint arthroplasty. www.aaos.org/
and disadvantages when considering IV cefazolin at a dose about/papers/advistmt/1027.asp. Accessed 27 Feb 2011
of 2 g for preoperative prophylaxis in total knee 10. Chang Y, Shih HN, Chen DW, Lee MS, Ueng SW, Hsieh PH
arthroplasty. (2010) The concentration of antibiotic in fresh-frozen bone graft.
J Bone Joint Surg Br 92:1471–1474
11. Bruinsma BG, Post IC, van Rijssen LB, de Boer L, Heger M, Zaat
Acknowledgments The authors would like to give special thanks to
SA, van Gulik TM (2013) Antibiotic prophylaxis in (Sub)nor-
Ms. Narissara Mungkornkaew who performed the bioactivity analyses
mothermic organ preservation. In vitro efficacy and toxicity of
in the present study. The authors also gratefully acknowledge the
cephalosporins. Transplantation 95(8):1064–1069
financial support provided by Thammasat University under the
12. Zegaer BH, Ioannidis A, Babis GC, Ioannidou V, Kossyvakis A,
Government budget 2015, Contract No. 006/2558 and the partial
Bersimis S, Papaparaskevas J, Petinaki E, Pliatsika P, Chatzi-
support provided by Central Scientific Instrument Center (CSIC),
panagiotou S (2014) Detection of bacteria bearing resistant bio-
Faculty of Science and Technology, Thammasat University.
film forms, by using the universal and specific PCR is still
unhelpful in the diagnosis of periprosthetic joint infections. Front
Compliance with ethical standards Med (Lausanne) 1:30
Conflict of interest None.

24
External versus internal fixation for bicondylar tibial plateau
fractures: systematic review and meta-analysis
David Metcalfe1,2,3 • Craig J. Hickson4 • Lesley McKee5 • Xavier L. Griffin2
Abstract Results Patients undergoing external fixation were more

Background It is uncertain whether external fixation or likely to have returned to preinjury activities by six and
open reduction internal fixation (ORIF) is optimal for twelve months (P = 0.030) but not at 24 months follow-
patients with bicondylar tibial plateau fractures. up. However, external fixation was complicated by a
Materials and methods A systematic review using Ovid greater number of infections (OR 2.59, 95 % CI 1.25–5.36,
MEDLINE, Embase Classic, Embase, AMED, the P = 0.01). There were no statistically significant differ-
Cochrane Library, Open Grey, Orthopaedic Proceedings, ences in the rates of deep infection, venous thromboem-
WHO International Clinical Trials Registry Platform, Cur- bolism, compartment syndrome, or need for re-operation
rent Controlled Trials, US National Institute for Health between the two groups.
Trials Registry, and the Cochrane Central Register of Conclusion Although external fixation and ORIF are
Controlled Trials. The search was conducted on 3rd October associated with different complication profiles, both are
2014 and no language limits were applied. Inclusion criteria acceptable strategies for managing bicondylar tibial plateau
were all clinical study designs comparing external fixation fractures.
with open reduction internal fixation of bicondylar tibial Level of evidence II.
plateau fractures. Studies of only one treatment modality
were excluded, as were those that included unicondylar Keywords External fixation Internal fixation
tibial plateau fractures. Treatment effects from studies Bicondylar tibial plateau Proximal tibial fracture
reporting dichotomous outcomes were summarised using
odds ratios. Continuous outcomes were converted to stan-
dardized mean differences to assess the treatment effect, Introduction
and inverse variance methods used to combine data. A fixed
effect model was used for meta-analyses. Tibial plateau fractures are uncommon injuries, represent-
ing only 1.2 % of all fractures [1]. They have a bimodal
& David Metcalfe incidence, occurring in young patients suffering high-en-
[email protected] ergy trauma, and as fragility fractures in the elderly [2].
1
Harvard Medical School, 25 Shattuck Street, Boston, Bicondylar tibial plateau fractures (Schatzker types V and
MA 02115, USA VI/Orthopaedic Trauma Association types C1, C2, and C3)
2
Warwick Medical School, Gibbet Hill Road, typically follow high-energy trauma [3, 4]. They are
Coventry CV4 7AL, UK complex intra-articular injuries with implications for
3
Division of Trauma, Burns, and Surgical Critical Care, articular congruity, cartilage integrity and extra-articular
Brigham and Women’s Hospital, 75 Francis Street, Boston, structures [5]. Associated complications include compart-
MA 02115, USA ment syndrome, soft tissue damage, secondary
4
Leicester Royal Infirmary, Infirmary Square, osteoarthrosis (OA), and persistent knee instability. Con-
Leicester LE1 5WW, UK servative treatment is rarely appropriate for these injuries
5
Forth Valley Hospital, Stirling Road, Larbert, Scotland, UK [6].

External versus internal fixation for bicondylar tibial plateau fractures: systematic review and meta-analysis 163
Management aims are anatomic reduction of the artic- 2. ‘‘schatzker 6’’ or ‘‘schatzker VI’’ or ‘‘schatzker type 6’’
ular surface, restoration of axial alignment, and stable or ‘‘schatzker type VI’’ or ‘‘schatzker 5’’ or ‘‘schatzker
fixation to prevent secondary displacement of the fracture V’’ or ‘‘schatzker type 5’’ or ‘‘schatzker type V’’ or
fragments [7]. A commonly employed technique is open ‘‘bicondylar’’ or ‘‘comminuted’’ or ‘‘complex’’
reduction and internal fixation (ORIF), using a plate and 3. ‘‘complex tibial plateau’’
screws through either an extended anterior incision or 4. ‘‘external fix*’’ or ‘‘frame’’
through multiple smaller incisions to preserve the soft tis- 5. 1 and 2
sue envelope. High-energy bicondylar fractures are often 6. 3 or 5
already accompanied by soft tissue damage, and ORIF in 7. 4 and 6.
this setting is associated with wound complications, e.g.,
The Cochrane Library and Open Grey (System for
skin necrosis and infection [8]. Soft tissue considerations
Information on Grey Literature in Europe, http://www.
may also delay operative fixation and/or contraindicate
opengrey.eu) were searched using the term ‘‘tibial
ORIF altogether. In addition, there is evidence to suggest
plateau’’.
that, once alignment is restored, residual articular incon-
Conference proceedings from the British Orthopaedic
gruity may not impair long-term functional results fol-
Association, British Trauma Society, Orthopaedic Trauma
lowing these injuries [9–13].
Association, British Association for Surgery of the Knee,
These observations have driven a search for alternative
and European Federation of National Associations for
interventions, including isolated tension band wire fixation
Orthopaedics and Traumatology were screened using the
[14], minimally invasive plate osteosynthesis (MIPO) [2],
digital archive Orthopaedic Proceedings [18] from 1st
and hybrid external fixation [15]. The latter technique
March 2002 to 3rd October 2014. Titles and abstracts were
involves reduction of the fracture using closed manipula-
searched using the term ‘‘tibial plateau fracture’’.
tion, percutaneously, or through limited incisions. Fracture
Ongoing and recently completed trials were searched
reduction is stabilized with one or more percutaneous lag
using the term ‘‘tibial plateau’’ in the WHO International
screws, and an external fixator (typically a circular frame)
Clinical Trials Registry Platform [19], Current Controlled
is assembled to secure the metaphysis to the tibial
Trials [20], US National Institute of Health Trials Registry
diaphysis.
[21], and the Cochrane Central Register of Controlled
This systematic review sought to compare all forms of
Trials [22].
external fixation (including hybrid techniques) with ORIF
Authors of leading studies were contacted for details of
for bicondylar tibial plateau fractures in terms of radio-
ongoing work. Reviews, editorials, and opinion articles
logical and clinical outcomes as well as their post-operative
were used as potential sources of further references.
complication profiles.
Inclusion and exclusion criteria
All clinical study designs were included that met the fol-
lowing criteria:
A systematic review was performed in line with the
Cochrane Handbook for Systematic Reviews of Interven- • Reporting on human patients with bicondylar (OTA C1,
tions [16] and reported according to the Preferred Report- C2, and C3) tibial plateau fractures.
ing Items for Systematic Reviews and Meta-Analysis • Direct comparison between any form of external
(PRISMA) statement [17]. fixation (including hybrid techniques utilizing percuta-
neous screw fixation) and ORIF.
Search strategy • Reporting outcomes that were radiological (fracture
reduction, union, subsequent OA) or clinical (func-
The following databases were searched using the strategy tional scores, patient-reported outcomes, need for
below: Ovid MEDLINE (1946 to September week 4 2014), subsequent operation including arthroplasty), and/or
Embase Classic (1947–1973), Embase (1974 to 2nd post-operative complications (defined as any deleteri-
October 2014), and AMED (1985 to September 2014). All ous event described by study authors as post-operative
searches were conducted on 3rd October 2014. No limits complications).
were applied in terms of language, publication status, or Criteria for excluding studies were:
study design. The search strategy was:
• Reporting data from patients with peri-prosthetic and/or
1. ‘‘proximal tib*’’ or ‘‘tibial plateau’’ pathological fractures.

• Failure to analyze data on bicondylar fractures sepa- visual inspection of overlapping confidence intervals on
rately, e.g., populations including patients with uni- forest plots and consideration of the I2 with P \ 0.1
condylar fractures. Authors were contacted for interpreted as significant heterogeneity.
unpublished data in all such cases. Except for assessment of heterogeneity, P \ 0.05 was
• Isolated case series of patients undergoing either ORIF used as the threshold for statistical significance. All sta-
or external fixation without distinction between treat- tistical analyses were performed using Stata v.13.1 (Sta-
ment modalities. taCorp, Memphis, TN) or RevMan v.5.2.3 (Nordic
Cochrane Centre, Copenhagen, Denmark). RevMan was
also used to construct forest plots.
Selection of studies
Missing data that could not be retrieved despite con-
tacting study authors was excluded from the analysis.
Two authors (DM and CH) independently screened all
retrieved items by title then abstract and full text as nec-
essary using the pre-determined selection criteria. Dis-
Results
agreements were resolved through discussion.
The initial search retrieved 311 individual items, of which
Quality assessment
ten satisfied the inclusion criteria (Fig. 1). These included
seven full research papers [25–31], and three published
Two authors (DM and LM) independently assessed risk of
conference abstracts [32–34], the characteristics of which
bias. Randomized controlled trials were assessed using the
are described in Table 1. Two registered trials were iden-
Cochrane Collaboration Risk of Bias Tool [16], which
tified, both of which were represented by published studies
considers selection bias (random allocation and allocation
retrieved during the search [28, 30]. Six items [26, 29, 30,
concealment), performance bias (blinding of participants
32–34] described three overlapping datasets and were
and personnel), detection bias (blinding of outcome
analyzed in aggregate form as Boston [26, 29], Chertsey
assessment), attrition bias (incomplete outcome data),
[33, 34], and COTS [30, 32].
reporting bias (selective reporting), and other sources of
There was one RCT and six retrospective studies
bias. Non-randomized studies were assessed using the Risk
reporting data on 419 fractures, of which 220 (52.5 %)
of Bias Assessment Tool for Non-Randomized Studies
were treated with external fixation.
(RoBANS) [23]. This tool considers similar bias domains
to that produced by Cochrane but is modified for non-
Study characteristics
randomized study designs. Both tools assess risk of bias in
each domain as ‘‘high’’, ‘‘low’’, or ‘‘unknown’’. Disagree-
The RCT [30, 32] was a large multi-centre trial in which
ments were resolved through discussion.
patients with bicondylar tibial plateau fractures were ran-
domized to either ORIF (with medial and lateral plates) or
Extraction of data
application of a circular fixator with percutaneous/limited
open fracture reduction. The primary outcome measure was
A single author (DM) extracted data from studies onto a
the Hospital for Special Surgery (HSS) knee score, which
standardized proforma. Study authors were contacted for
incorporates pain, function, range of motion, muscle
clarification and/or additional data when fields could not be
strength, flexion contractures, and instability [35]. In total,
completed from the published reports.
82 patients (83 fractures) were randomized, which was the
number determined by an a priori power analysis designed
to give an 80 % chance of detecting a 25 % mean differ-
ence in the primary outcome measure between the two
Treatment effects from studies reporting dichotomous
groups.
outcomes were summarised using odds ratios and com-
The six retrospective studies [25–29, 31, 33, 34]
bined using the Mantel–Haenszel technique [24]. Contin-
accounted for 336 (80.2 %) of the published cases avail-
uous outcomes were converted to standardized mean
able for analysis. There was substantial heterogeneity in
differences to assess the treatment effect, and inverse
terms of the interventions used between the retrospective
variance methods were used to combine data. Confidence
studies. Each reported on a range of external fixation and
intervals were reported at the 95 % level and a fixed effect
ORIF techniques using multiple devices. The former
model was used for meta-analyses, although we planned to
included Ilizarov circular frames, the Hoffman II (Stryker,
use a random effects model in the event of significant
Kalamazoo, MI), and the Synthes AO fixator (DePuy
heterogeneity. Statistical heterogeneity was assessed by
Synthes, West Chester, PA). ORIF techniques variously

Fig. 1 PRISMA flow diagram

showing selection of studies for
the systematic review
utilized locking plates, non-locking plates, and the Synthes The retrospective studies were all judged to be at high
Less Invasive Stabilization System (LISS) (DePuy Synthes, risk of confounding variables. Four of the retrospective
West Chester, PA). Some patients treated with ORIF also studies addressed known confounders by reporting the
received iliac crest bone grafting or artificial bone patient characteristics of each group. Such reporting was,
substitute. however, limited and variable [25, 26, 28, 29, 33, 34]. Only
Chan et al. described a significant difference between the
Study quality two groups in that alcohol dependency was over-repre-
sented in the external fixation group (4 % vs 20 %). Jansen
The RCT [30] was assessed to be at low risk of bias et al. described demographic characteristics for their whole
across most domains (Table 2), although there was no series but not by treatment modality [27]. Due to their
blinding of patients or personnel and the protocol was not retrospective nature, additional confounders (either unre-
published before recruitment commenced. For this reason, ported or unidentified) are likely to exist and conclusions
the study was judged to be at unclear risk of reporting from these studies should therefore be treated with caution.
bias. Financial support was received from Smith &
Nephew Ltd (London, UK) and the Simon Fraser Ortho- Radiographic outcomes
paedic Fund. Smith & Nephew sell a range of external
fixation devices and it was not possible to determine Two studies (142 fractures) assessed fracture reduction
whether the latter sponsor represented a commercial radiologically [25, 30]. In both studies, a single assessor
interest. There was no explicit statement as to the role of graded post-operative radiographs. Chan et al. additionally
these funders in the study report. scored radiographs using Rasmussen’s system, which is
Table 3 shows the risk of bias assessments for the six based on joint depression, condylar widening, and
retrospective studies using the RoBANS tool [23]. Five varus/valgus angulation [10]. Although designed specifi-
were assessed to be at low risk of selection bias [25, 26, 28, cally for fractures around the knee, there is little published
29, 31, 33, 34] and the remaining study was at unclear risk evidence assessing its reliability and validity [36]. These
[27]. Low risk studies either declared that the series was studies reported no statistically significant differences in
consecutive or that it represented all cases treated over a terms of articular displacement, diaphyseal-metaphyseal
given time period. No study explicitly reported blinding of angulation/translation, condylar widening, or Rasmussen’s
outcome assessors and so all were assessed to be at unclear score.
risk of detection bias. Similarly, the risk of reporting bias Only Krupp et al. [28] reported time to radiographic
(selective outcome reporting) was unclear for all of the union which was comparable between the two groups: 6
retrospective studies. Four studies were at high risk of (range 3–14) months in the ORIF group and 7 (range 3–15)
attrition bias (incomplete outcome data) as a number of months in those managed with external fixation.
cases were lost to follow-up [25, 26, 28, 29, 31]. The Three studies (165 fractures) assessed follow-up radio-
remaining two were judged to be at low risk as outcome graphs for evidence of OA [25, 27, 30]. The COTS and
data was reported for almost all cases [27, 33, 34]. Chan studies both used radiographs taken after the same

Table 1 Key features of included studies
166
Name Design Patients (knees) Interventions Follow-up

External fixation ORIF
Ahearn [14] Retrospective 21 34 External fixation: Taylor Spatial Frame (TSF)a External fixation: mean 31 months (range
ORIF: Incision: Unknown. Fixation: Lateral 12–58 months)
locking plate ± medial plate fixation ORIF: mean 41 months (range
12–64 months)
Boston Retrospective 10 (10) 7 (7) External fixation: Monticelli-Spinellib circular External fixation: mean 10 months (range
Mallik [29] fixator 5–28 months)
Covall [26] ORIF: Incision: Unknown. Fixation: Bilateral ORIF: mean 33 months (range 6–60 months)
buttress, semi-tubular plates, or cannulated
screws
Chertsey Retrospective (79) (45) External fixation: Ilizarov circular frame Unknown
Guryel [34] ORIF: Incision: Unknown. Fixation: Unknown
Nawaz [33]
COTS Randomized controlled trial 42 (43) 40 (40) External fixation: Closed/percutaneous/limited 6, 12, and 24 months post-injury
McKee [30] reduction, percutaneous lag screw, and Ilizarov
circular frame
Pirani [32]
ORIF: Incision: Single anterior or combined
medial/lateral. Fixation: medial and lateral
non-locking buttress plates ± iliac crest bone
grafting
Chan [25] Retrospective 34 (35) 24 (24) External fixation: Ilizarov circular frame (23/35, 3, 6, 12, and 24 months post-injury
65.7 %), Hoffman IIb with limited internal
fixation (13/35, 37.1 %)
ORIF: Incision: Unknown. Fixation: Buttress
plate (21/24, 84 %), Less Invasive Stabilization
Systemc (4/24, 16 %)
Jansen [5] Retrospective 2 (2) 20 (21) External fixation: Synthes AOc fixator or Ilizarov Mean 67 months (range 36–109 months)
circular frame
ORIF: Incision: Unknown. Fixation: LISSc (19/
20, 95.0 %) ± additional plates (7/20,
30.4 %) ± artificial bone substitute (7/20,
30.4 %)
Krupp [28] Retrospective 30 (30) 28 (28) External fixation: Hoffman II Hybridb (16/28, Mean unknown (range 6–53 months)
57.1 %) or circular (14/28, 50.0 %) frames and
interval ORIF (locking plate or LISSc)
ORIF: Incision: Unknown. Fixation: locking
plate (8/28, 28.6 %), LISSc (20/28, 71.4 %)
a
Smith and Nephew Ltd, Brough, United Kingdom
b
Stryker Corporation, Kalamazoo, MI
c
DePuy Synthes Companies, West Chester, PA

Table 2 Risk of bias assessment of randomized studies

Sequence Allocation Blinding of participants, Incomplete Selective Other
generation concealment personnel and outcome outcome outcome sources
assessors data reporting of bias
COTS Low risk Low risk High risk Low risk Unclear risk Low risk
McKee [30]
Pirani [32]
Table 3 Risk of bias assessment of non-randomized studies

Selection of Confounding Intervention Blinding of outcome Incomplete Selective outcome
participants variables measurement assessment outcome data reporting
Ahearn [31] Low risk High risk Low risk Unclear risk High risk Unclear risk
Boston Low risk High risk Low risk Unclear risk High risk Unclear risk
Mallik [29]
Covall [26]
Chertsey Low risk High risk Low risk Unclear risk Low risk Unclear risk
Guryel [34]
Nawaz [33]
Chan [25] Low risk High risk Low risk Unclear risk High risk Unclear risk
Jansen [5] Unclear risk High risk Low risk Unclear risk Low risk Unclear risk
Krupp [28] Low risk High risk Low risk Unclear risk High risk Unclear risk
standardized follow-up period, i.e., 24 months post-opera- in the ORIF group, they provided no indication of statis-
tively. However, they relied on subjective assessment by a tical significance. In general, there were few significant
single unblinded assessor. Jansen et al. increased the reli- differences between the groups on any functional outcome.
ability of their results by using an established radiographic The COTS primary outcome measure (HSS) [35] trended
interpretation tool: the Kellgren-Lawrence score [37]. towards higher HSS in the external fixation group (mean
Unfortunately they reported onset of OA for their series as difference in HSS 11.00, 95 % CI 2.03–19.97, P = 0.06),
a whole without distinguishing between the two treatment which might have reached significance with a greater
groups. Their follow-up period also ranged from 36 to sample size. However, any genuine difference did not
109 months, making it difficult to directly compare patients persist at 12 (mean difference 5.00, 95 % CI -2.59 to
[27]. Pooled results from the remaining two studies (Fig. 2) 12.59, P = 0.406) and 24 months (mean difference 7.00,
found radiographic evidence of OA in 22 (32.8 %) of 95 % CI -1.45 to 15.45, P = 0.307) Similarly, the exter-
external fixation and 18 (31.0 %) of ORIF cases (OR 1.14, nal fixation patients were more likely to have returned to
95 % CI 0.53–2.44, P = 0.740) at 24 months post-injury. pre-injury activities at 6 months (P = 0.030) but not at
later follow-up assessments.
Functional outcomes Jansen et al. reported outcomes for their whole series
using the Lysholm score [38] and Knee Injury and
Three studies reported functional outcomes [30–34]. Osteoarthritis Outcome Score (KOOS) [39] but did not
Although Krupp et al. reported better range of movement distinguish between patients in the two treatment groups.
Fig. 2 A forest plot showing pooled data from studies reporting radiographic evidence of OA at 24 months post-injury

Subsequent knee arthroplasty Re-operation
Two studies (117 fractures) reported on subsequent need for Three studies (196 fractures) described rates of re-opera-
ipsilateral total knee arthroplasty (TKA) [25, 28]. Figure 3 tion, as shown in Fig. 6 [28, 30, 31]. In the pooled external
shows that the pooled rates of TKA in the external fixation fixation group, 25 cases (26.6 %) required an additional 40
and ORIF groups were 7.7 and 11.5 % (OR 0.56, 95 % CI operations whereas, in the ORIF group, 29 (28.4 %)
0.16–2.00, P = 0.69). Chan et al. followed up patients at required 72 operative interventions. The pooled re-opera-
24 months, although it is uncertain whether TKAs occurring tion rate was not statistically significant (OR 0.77, 95 % CI
subsequently were included. For example, they reported 0.40–1.49, P = 0.44). However, no study took planned
cases presenting before March 2005 but published their paper procedures (such as frame removal) into account during
in 2012. The authors do not state whether TKAs were their analyses. In the COTS trial, 27 frames (65.9 %) were
included if performed between 2005 and 2012. The cases removed in the operating theatre under general anaesthetic
reported by Krupp et al. had variable follow-up lengths that or sedation.
ranged from 6 to 53 months. In any event, it is likely that an Substantial re-operations (e.g., knee arthrodesis) in the
unknown proportion of patients developed end-stage post- ORIF group were described in the Boston series, although
traumatic OA requiring TKA outside these follow-up periods. these papers did not describe re-operations systematically.
The COTS report observed that re-operations following
Complications ORIF were more substantial (e.g., above knee amputation,
osteotomy) than in the external fixation group (e.g., pin-
All six retrospective studies (336 fractures) described rates track debridement), although there was no attempt to
of superficial and deep infection [25–29, 31, 34]. The rates quantify this observation.
of superficial infection in the external fixation and ORIF
groups, respectively, were 14.0 vs 4.7 % (OR 1.93, 95 %
CI 0.17–22.53, P = 0.01). The rates of deep infection were Discussion
4.2 and 2.6 % (OR 1.23, 95 % CI 0.44–3.44, P = 0.700),
respectively. Pooled results for any infection (deep or Although ORIF is often successful in restoring articular
superficial) found that patients treated with external fixa- congruity, it may further compromise the soft tissue
tion had greater odds of this outcome (OR 2.59, 95 % CI envelope. Many case series have highlighted the dangers of
1.25–5.36, P = 0.01). The forest plots for these infections wound breakdown and deep infection following ORIF of
are shown in Fig. 4. bicondylar tibial plateau fractures [8, 29, 40]. These
Three studies (238 fractures) described rates of venous problems have persisted, even in modern studies utilizing
thromboembolism (VTE) [25, 31, 34]. There were nine techniques such as delayed surgery and minimal soft tissue
cases of deep vein thrombosis (3.8 %), with no statistically dissection. For example, Baeri et al. reported deep infec-
significant differences between the groups (OR 1.56, 95 % tions in seven (8.4 %) of 83 patients treated with ORIF,
CI 0.49–4.96, P = 0.45), and no reported pulmonary each of whom required a mean 3.3 additional operations as
emboli. As neither study described screening for VTE, these a consequence [41].
cases presumably presented symptomatically. Compart- External fixation devices preserve soft tissues and an
ment syndrome was reported as a complication by two emerging body of evidence suggests they can achieve
studies (81 fractures) [25, 27]. It featured in 5.4 % of lower rates of deep infection [42–44]. Although external
external fixation cases and 9.1 % of those undergoing ORIF fixation might risk sacrificing the quality of fracture
(OR 0.61, 95 % CI 0.12–3.20, P = 0.56). Forest plots for reduction, it is uncertain whether this ultimately affects
VTE and compartment syndrome are shown in Fig. 5. functional outcome [9–12].
Fig. 3 A forest plot showing pooled data from studies reporting need for subsequent total knee replacement

Fig. 4 a A forest plot showing pooled results of studies reporting all post-operative infections, b a forest plot showing pooled results of studies
reporting superficial post-operative infections, and c a forest plot showing pooled results of studies reporting deep post-operative infections
Fig. 5 a A forest plot showing pooled data from studies reporting on rates of venous thromboembolism and b a forest plot showing pooled data
from studies reporting on rates of compartment syndrome

Fig. 6 A forest plot showing pooled data from studies reporting need for subsequent re-operation
Few studies have directly compared external fixation than those following external fixation [30]. Importantly,
and ORIF for treatment of bicondylar tibial plateau frac- infection complicated a greater proportion of cases man-
tures. This systematic review identified seven such studies, aged with an external fixator than with ORIF (OR 2.59,
most of which were poor-quality retrospective case series, 95 % CI 0.49–4.96). This suggests that the soft tissue
although there was one RCT. There was substantial complications of external fixation could be even greater
heterogeneity of study populations and reported outcomes. than ORIF in this setting.
In addition, the retrospective studies, which accounted for The existing evidence suggests that neither ORIF nor
the majority of cases available for analysis (80.2 %), were external fixation is clearly superior in the management of
at high risk of bias caused by confounder variables. Pooled bicondylar tibial plateau fractures. Importantly, external
data from these studies suggests that patients managed with fixation does not offer any clear advantage over ORIF in
external fixation are at greater risk of superficial infection, terms of avoiding soft tissue complications. Although
although other complications (including deep infection) clinicians should be mindful of subtly different complica-
were comparable between the groups. However, patients tion profiles and the possible need to remove external fix-
undergoing external fixation may return to pre-injury ators in theatre, both external fixation and ORIF are
activities faster than those treated with ORIF. The seven acceptable strategies for managing these injuries.
studies identified no other statistically significant differ-
ences across a range of outcomes between ORIF and Compliance with ethical standards
external fixation. Conflicts of interest The authors have no conflicts of interest to
One important limitation of all existing studies is the disclose.
relatively short follow-up duration. Post-traumatic OA is
an important long-term complication of intra-articular Ethical standards This article does not contain any studies with
human participants performed by any of the authors.
fractures through this weight-bearing joint. However, it is
difficult to rely on reported rates of secondary OA and Funding No specific funding was received in respect of this report.
need for subsequent TKA in these studies, given the small
numbers involved, short follow-up durations, and incon-
sistent reporting. Similarly, review of follow-up radio-
graphs for early evidence of OA relied on subjective
interpretation by non-blinded assessors. Although there
are few short-term functional differences between those
undergoing ORIF and external fixation, the long-term
impact on knee OA remains unknown. Importantly, the
three studies assessing quality of articular surface
restoration found no difference between the two groups References
[25, 30, 31].
1. Cole PL, Schatzker J, Watson JT (2009) Tibial plateau fractures.
There is additional uncertainty surrounding the com- In: Browner BL, Jupiter J, Trafton P, Krettek C (eds) Skeletal
plication profile of the two procedures. Although the pro- trauma: basic science management and reconstruction. Saunders
portion of patients requiring re-operation appeared to Elsevier, Philadelphia, pp 2201–2287
2. Biggi F, Di Fabio S, D’Antimo C, Trevisani S (2010) Tibial
favour external fixation, this was not statistically signifi-
plateau fractures: internal fixation with locking plates and the
cant. However, the analysis did not include planned pro- MIPO technique. Injury 41(11):1178–1182
cedures, including the need for frame removal under 3. Schatzker J, McBroom R, Bruce D (1979) The tibial plateau
sedation and/or anaesthesia. It was also suggested that re- fracture. The Toronto experience 1968–1975. Clin Orthop Relat
Res 138:94–104
operations following ORIF may be of greater importance

4. Marsh JL, Slongo TF, Agel J, Broderick JS, Creevey W, DeCo- 24. Mantel N, Haenszel W (1959) Statistical aspects of the analysis
ster TA et al (2007) Fracture and dislocation classification of data from retrospective studies of disease. J Natl Cancer Ins
compendium—2007: Orthopaedic Trauma Association Classifi- 22(4):719–748
cation, Database and Outcomes Committee. J Orthop Trauma 25. Chan CCK, Keating J (2012) Comparison of outcomes of oper-
21(S10):S1–S133 atively treated bicondylar tibial plateau fractures by external
5. Jansen H, Frey SP, Doht S, Fehske K, Meffert RH (2013) Med- fixation and internal fixation. Malays Orthop J 6(1):7–12
ium-term results after complex intra-articular fractures of the 26. Covall DJ, Fowble CD, Foster TE, Whitelaw GP (1994)
tibial plateau. J Orthop Sci 18(4):569–577 Bicondylar tibial plateau fractures: principles of treatment.
6. DeCoster TA, Nepola JV, El-Khoury GY (1988) Cast brace Contemp Orthop 28(2):115–122
treatment of proximal tibia fractures. A ten-year follow-up study. 27. Jansen H, Frey SP, Doht S, Fehske K, Meffert RH (2013) Med-
Clin Orthop Relat Res 231:196–204 ium-term results after complex intra-articular fractures of the
7. Schatzker J, McBroom R, Bruce D (1979) The tibial plateau tibial plateau. J Orthop Sci 18(4):569–577
fracture. The Toronto experience 1968–1975. Clin Orthop Relat 28. Krupp RJ, Malkani AL, Roberts CS, Seligson D, Crawford CH III,
Res 138:94–104 Smith L (2009) Treatment of bicondylar tibia plateau fractures using
8. Young MJ, Barrack RL (1994) Complications of internal fixation locked plating versus external fixation. Orthopedics 32(8):559
of tibial plateau fractures. Orthop Rev 23(2):149–154 29. Mallik AR, Covall DJ, Whitelaw GP (1992) Internal versus
9. Jensen DB, Rude C, Duus B, Bjerg-Nielsen A (1990) Tibial external fixation of bicondylar tibial plateau fractures. Orthop
plateau fractures. A comparison of conservative and surgical Rev 21(12):1433–1436
treatment. J Bone Joint Surg Br 72(1):49–52 30. McKee MD, Pirani SP, Stephen DJG, Feibel R, Powell JN,
10. Rasmussen PS (1973) Tibial condylar fractures. Impairment of McCormack R et al (2006) Open reduction and internal fixation
knee joint stability as an indication for surgical treatment. J Bone compared with circular fixator application for bicondylar tibial
Joint Surg Am 55(7):1331–1350 plateau fractures: results of a multicenter, prospective, random-
11. Marsh JL, Buckwalter J, Gelberman R, Dirschl D, Olson S, ized clinical trial. J Bone Joint Surg Am 88(12):2613–2623
Brown T et al (2002) Articular fractures: does an anatomic 31. Ahearn N, Oppy A, Halliday R, Rowett-Harris J, Morris SA,
reduction really change the result? J Bone Joint Surg Am A Chesser TJ et al (2014) The outcome following fixation of
84(7):1259–1271 bicondylar tibial plateau fractures. Bone Joint J B 96(7):956–962
12. Duwelius PJ, Connolly JF (1988) Closed reduction of tibial 32. Pirani SM, McKee M, Canadian Orthopaedic Trauma Society
plateau fractures. A comparison of functional and roentgeno- (2008) Randomized trial of bicondylar tibial plateau fractures
graphic end results. Clin Orthop Relat Res 230:116–126 treated with AO or ring fixator methods: early reintervention rates
13. Honkonen SE (1995) Degenerative arthritis after tibial plateau for complications. J Bone Joint Surg Br B 90(S1):47
fractures. J Orthop Trauma 9(4):273–277 33. Nawaz SZK, Elliott D, Newman K, Khaleel A (2013) Medium to
14. Ali AM, Burton M, Hashmi M, Saleh M (2003) Outcome of long term follow-up of high energy tibial plateau fractures: the
complex fractures of the tibial plateau treated with a beam- Chertsey experience of internal fixation and Ilizarov fine wire
loading ring fixation system. J Bone Joint Surg Br 85(5):691–699 circular frame. Bone Joint J 95B(S1):157
15. Dendrinos GK, Kontos S, Katsenis D, Dalas A (1996) Treatment 34. Guryel EM, Sharma R, Newman K, Elliott D, Khaleel A (2010)
of high-energy tibial plateau fractures by the Ilizarov circular The Chertsey classification of tibial plateau fractures and a
fixator. J Bone Joint Surg Br 78(5):710–717 comparison of the outcome of treatment with internal fixation or
16. Higgins JS, Churchill R, Cumpston M, Chandler J (eds) (2011) an Ilizarov fine wire circular frame. Injury Extra 41:140
Cochrane handbook for systematic reviews of interventions. 35. Insall JN, Ranawat CS, Aglietti P, Shine J (1976) A comparison
Cochrane Collaboration, Oxford of four models of total knee-replacement prostheses. J Bone Joint
17. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, Surg Am 58(6):754–765
Ioannidis JP et al (2009) The PRISMA statement for reporting 36. Tilley ST, Thomas N (2010) What knee scoring system?. J Bone
systematic reviews and meta-analyses of studies that evaluate Joint Surg Br. http://www.boneandjoint.org.uk/content/focus/
healthcare interventions: explanation and elaboration. BMJ what-knee-scoring-system. Accessed 22 Feb 2015
339:b2700 37. Kellgren JH, Ball J (1963) Atlas of standard radiographs.
18. Dickson KF, Munz J (2007) Locked plating: clinical indications, Blackwell Scientific, Oxford
3rd edn. Lippincott Williams and Wilkins, Philadelphia, 38. Tegner Y, Lysholm J (1985) Rating systems in the evaluation of
pp 181–185 knee ligament injuries. Clin Orthop Relat Res 198:43–49
19. World Health Organization. International Clinical Trials Registry 39. Roos EM, Roos HP, Lohmander LS, Ekdahl C, Beynnon BD
Platform. http://www.who.int/ictrp/en/. Accessed 3 Oct 2014 (1998) Knee injury and osteoarthritis outcome score (KOOS)—
20. Subasi M, Kapukaya A, Arslan H, Ozkul E, Cebesoy O (2007) development of a self-administered outcome measure. J Orthop
Outcome of open comminuted tibial plateau fractures treated Sports Phys Ther 28(2):88–96
using an external fixator, 4th edn. Springer-Verlag, Tokyo, 40. Moore TM, Patzakis MJ, Harvey JP (1987) Tibial plateau frac-
pp 347–353 tures: definition, demographics, treatment rationale, and long-
21. National Institutes for Health. National Institutes of Health Trials term results of closed traction management or operative reduc-
Registry. https://clinicaltrials.gov. Accessed 3 Oct 2014 tion. J Orthop Trauma 1(2):97–119
22. The Cochrane Collaboration. Cochrane Central Register of 41. Barei DP, Nork SE, Mills WJ, Henley MB, Benirschke SK (2004)
Controlled Trials (CENTRAL). http://www.cochrane.org/editor Complications associated with internal fixation of high-energy
ial-and-publishing-policy-resource/cochrane-central-register-con bicondylar tibial plateau fractures utilizing a two-incision tech-
trolled-trials-central. Accessed 3 Oct 2014 nique. J Orthop Trauma 18(10):649–657
23. Kim SY, Park JE, Lee YJ, Seo HJ, Sheen SS, Hahn S et al (2013) 42. Ariffin HM, Mahdi NM, Rhani SA, Baharudin A, Shukur MH
Testing a tool for assessing the risk of bias for nonrandomized (2011) Modified hybrid fixator for high-energy Schatzker V and
studies showed moderate reliability and promising validity. J Clin VI tibial plateau fractures. Strategies Trauma Limb Reconstr.
Epidemiol 66(4):408–414 6(1):21–26

43. Catagni MA, Ottaviani G, Maggioni M (2007) Treatment minimal internal fixation. Eur J Orthop Surg Traumatol
strategies for complex fractures of the tibial plateau with external 23(3):349–355
circular fixation and limited internal fixation. J Trauma
63(5):1043–1053
44. Ali AM (2013) Outcomes of open bicondylar tibial plateau
fractures treated with Ilizarov external fixator with or without

25
Deep peroneal nerve palsy with isolated lateral compartment
syndrome secondary to peroneus longus tear: a report of two cases
and a review of the literature
Kunihiko Hiramatsu1 • Yasukazu Yonetani2 • Kazutaka Kinugasa2 •
Norimasa Nakamura3,4 • Koji Yamamoto5 • Hideki Yoshikawa6 • Masayuki Hamada2
Abstract Drop foot is typically caused by neurologic peroneus longus. The characteristic anatomical situation
disease such as lumbar disc herniation, but we report two between the fibular tunnel and peroneus longus might have
rare cases of deep peroneal nerve palsy with isolated lateral caused deep peroneal nerve palsy in these two cases after
compartment syndrome secondary to peroneus longus hematoma adjacent to the fibular tunnel increased lateral
tears. Both patients developed mild pain in the lower legs compartment pressure.
while playing sport, and were aware of drop foot. As
compartment pressures were elevated, fasciotomy was Keyword Lateral compartment syndrome Deep
performed immediately, and the tendon of the peroneus peroneal nerve palsy Peroneus longus tear
longus was completely detached from its proximal origin.
The patients were able to return their original sports after
3 months, and clinical examination revealed no hypesthe- Introduction
sia or muscle weakness in the deep peroneal nerve area at
the time of last follow-up. The common peroneal nerve Compartment syndrome of the lower extremity is a rare
pierced the deep fascia and lay over the fibular neck, which event and can occur with trauma or occasionally with a
formed the floor of a short tunnel (the so-called fibular sports injury. The diagnosis needs to be established early in
tunnel), then passed the lateral compartment just behind the its course to avoid disabling sequelae, such as neurologic
disorders. The most frequent location of compartment
syndrome in the lower extremity is the anterior compart-
ment. Looking at the literature, lateral compartment syn-
drome of the lower leg is quite rare. Lateral compartment
& Kunihiko Hiramatsu syndrome occurs due to inversion ankle injuries [1, 2],
[email protected] exertion [3, 4], horseback riding [5, 6], a prolonged litho-
1
Department of Orthopaedic Surgery, Yao Municipal
tomy position in general surgical, urologic, and gyneco-
Hospital, 1-3-1, Ryugecho, Yao, Osaka, Japan logic procedures [7], peroneus longus muscle tears or
2
Department of Orthopaedic Surgery, Hoshigaoka Medical
avulsion [8–12].
Center, 4-8-1, Hoshigaoka, Hirakata, Osaka, Japan Early diagnosis and treatment of lateral compartment
3
Institute for Medical Science in Sports, Osaka Health Science
syndrome secondary to peroneus longus tear is difficult
University, Osaka, Japan due to the lack of characteristic clinical symptoms [3, 13].
4
Center for Advanced Medical Engineering and Informatics,
To the best of our knowledge, deep peroneal nerve palsy
Osaka University, Suita, Osaka, Japan with lateral compartment syndrome secondary to com-
5 plete avulsion of the proximal origin of the peroneus
Department of Orthopaedic Surgery, Toyonaka Municipal
Hospital, 4-14-1, Shibahara, Toyonaka, Osaka, Japan longus has not been reported. Two rare cases of deep
6 peroneal nerve palsy with isolated lateral compartment
Department of Orthopaedic Surgery, Osaka University
Graduate School of Medicine, 2-2, Yamada-oka, Suita, syndrome secondary to peroneus longus tear are reported
Osaka 565-0871, Japan herein.

Case reports findings in the lateral compartment (Fig. 1a, b), and pres-
sures in the anterior and lateral compartments were 42 and
Case 1 120 mmHg, respectively. Based on these clinical findings,
fasciotomy was performed based on a definitive diagnosis
A 21-year-old man who had previously experienced no of anterior and lateral compartment syndromes.
pain in the legs, no muscle weakness, and no other disor- The proximal lower leg was exposed through a longi-
ders, developed mild pain in the right lower leg while tudinal incision above the lateral compartment. Hematoma
playing baseball, but he was able to continue to playing. within the lateral compartment was evacuated. The tendon
Three days later, he became aware of drop foot of the right of the peroneus longus was found to be completely
leg, but did not seek medical care because he could tolerate detached from its proximal origin. Although the hematoma
the pain. Two days later, he presented to the orthopedic was evacuated, lateral compartment pressure remained
department complaining of persistent drop foot of the right elevated (120 mmHg). The distal lower leg was exposed
leg. The initial clinical examination revealed mild swelling through an incision above the lateral compartment. The
of the anterior and lateral right lower leg, with focal peroneus muscle was ischemic and swollen, but not
prominence over the lateral muscle compartment, as well necrotic. The skin was closed, because the skin was not
as pain and tenderness. No pain was evident with passive tense. Reduced lateral compartment pressure was con-
plantar flexion of the ankle, and plantar flexion of the digits firmed, and the operation was finished. The day after the
was elicited. Manual muscle testing revealed 0/5 muscle operation, the patient complained of right lower leg pain.
strength of the anterior muscle group (tibialis anterior and The wound was opened because lateral compartment
extensor hallucis longus), 5/5 muscle strength of the pos- pressure was again increased (120 mmHg). After the
terior muscle group, and 0/5 muscle strength of the per- wound was opened, the patient noted pain relief.
oneus muscle. The posterior tibial and dorsalis pedis artery Fourteen days later, he was taken back to the operating
pulses were both palpable. There was decreased sensation room for delayed primary closure. At the time of primary
in the deep peroneal nerve area, but sensation was normal closure, tibialis anterior strength had recovered to 3/5, and
in the superficial peroneal nerve area. Lumbar diseases extensor hallucis longus and peroneus strengths were 1/5.
such as disc herniation were excluded because patella The patient was discharged 18 days after fasciotomy,
tendon and Achilles tendon reflexes were normal, and the requiring an ankle–foot orthosis for ambulation. Three
straight leg raising test yielded negative results. Magnetic months after fasciotomy, he was able to return to play
resonance imaging (MRI) of the right lower leg was per- baseball with almost complete recovery of muscle strength
formed because of swelling of the anterior and lateral lower in the tibialis anterior (5/5) and extensor hallucis longus/
right leg. MRI demonstrated hypointensity on T1-weighted peroneus (4/5). Clinical examination after 2 years revealed
imaging and hyperintensity on T2-weighted imaging that no hypesthesia and no muscle weakness in the territory of
appeared to represent a cystic lesion with fluid–fluid level the deep peroneal nerve.
Fig. 1 MRI of the right knee (case 1). Axial T1-weighted fast spin echo (a), and T2-weighted fast spin echo (b). Arrowheads indicate hematoma
in the peroneus longus muscle, which shows a fluid–fluid level

Deep peroneal nerve palsy with isolated lateral compartment syndrome secondary to peroneus longus tear... 175
Case 2 was found to be completely detached from its proximal

origin (Fig. 2). After confirmation of decreased lateral
A 16-year-old boy with no history of pain in the leg, compartment pressure, the operation was finished. Nine
muscle weakness, or other disorders developed pain in the days later, wound closure was performed without compli-
right lower leg after playing soccer. Sixteen days later, he cations. One month after fasciotomy, muscle testing
presented to the orthopedic department complaining of revealed full strength had been regained, and he could fully
swelling, pain, and numbness in the right leg. The initial return to play soccer.
clinical examination revealed swelling of the right lower
leg, and manual muscle testing showed 4/5 muscle strength
of the anterior muscle group (tibialis anterior and extensor Discussion
hallucis longus), 5/5 muscle strength of the posterior
muscle group, and 3/5 muscle strength of the peroneus This is the first case report of deep peroneal nerve palsy
muscle. Sensation was decreased in the deep peroneal with isolated lateral compartment syndrome secondary to
nerve area, but normal in the superficial peroneal nerve peroneus longus tears. In both cases, it was difficult to
area. Anterior compartment pressure was 42 mmHg, but diagnose because of the few and complex symptoms, such
that of the lateral compartment was 100 mmHg. The results as drop foot, which often occur in lumbar disc herniation.
of these clinical examinations led to the definitive diag- Although the most frequent presentation of compart-
nosis of anterior and lateral compartment syndromes, and ment syndrome of the lower extremity involves the anterior
fasciotomy was immediately performed. compartment, lateral compartment syndrome of the leg is
During fasciotomy through a longitudinal incision in the rare. In addition, as injury to the peroneal muscle–tendon
lateral side of the leg, the tendon of the peroneus longus unit tends to occur more distally, reports of acute rupture of
Fig. 2 Intraoperative photograph of the leg. The peroneus longus is completely detached from its proximal origin and retracted distally out of the
lateral compartment. a Case 1, b case 2

Fig. 3 Schematic diagrams of the normal anatomy around the peroneal nerve entrapment can occur at the level of the fibular tunnel
proximal end of the peroneus longus and peroneal nerve (a). Lateral behind the peroneus longus, because hematoma beside the fibular
compartment syndrome may result from a peroneus longus tear tunnel increases lateral compartment pressure. CPN common per-
leading to peroneal nerve palsy (b). The common peroneal nerve oneal nerve, S superficial peroneal nerve, D deep peroneal nerve, BF
(CPN) pierces the deep fascia and lies over the fibular neck, which biceps femoris muscle, AF apex of the fibula, FT fibular tunnel, AIS
forms the floor of the short ‘fibular tunnel’ (FT), and passes the lateral anterior intermuscular septum, PL peroneus longus, GC gastrocne-
compartment just behind the peroneus longus. Idiopathic deep mius. The oval filled by oblique lines represents hematoma
the peroneus longus muscle from its proximal origin are reports of isolated lateral compartment syndrome sec-
very rare [14]. To the best of our knowledge, only four ondary to peroneus longus tear. Previous reports and the
cases of isolated lateral compartment syndrome secondary features of the course of the peroneal nerve usually lead to
to peroneus longus tear have been described [9–12]. In the conclusion that the neurological symptoms caused by
those previous reports, the pathological processes causing isolated lateral leg compartment syndrome might not result
the peroneus longus to tear from its proximal origin were in a deep peroneal nerve lesion, but rather in a superficial
initiated by overuse of muscles, but the situations of injury peroneal nerve lesion. However, a precise anatomical study
remained unclear [10, 11]. As in previous cases, the present showed that the common peroneal nerve pierces the deep
two cases did not show the injury situations clearly. From fascia and lies over the fibular neck, which forms the floor
these perspectives, peroneus muscle tear should be inclu- of a short tunnel (the so-called fibular tunnel), and passes
ded in the differential diagnosis for patients who play the lateral compartment just behind the peroneus longus
sports intensely and develop lateral lower leg pain. [15]. At this fibular tunnel, the common peroneal nerve
The present two cases had unusual clinical findings for divides into the deep and superficial peroneal nerves. From
the following three pathognomonic symptoms. First, pain these anatomical findings, Ryan et al. showed that idio-
with passive stretching, which is one of the typical clinical pathic nerve entrapment could occur at the level of the
signs of compartment syndrome, was absent. However, Lee fibular tunnel behind the peroneus longus [16]. Therefore,
et al. reported that a lack of pain with passive stretching the characteristic anatomical situation between the fibular
might be due to rupture of the peroneus longus muscle [10]. tunnel and peroneus longus might have caused the deep
Second, neuropathy was the main clinical manifestation in peroneal nerve palsy in our two cases when hematoma
both of our two cases. Generally, when a young patient adjacent to the fibular tunnel increased lateral compartment
presents with deep peroneal nerve disorder, such as drop pressure (Fig. 3).
foot, lumbar disc herniation is usually suspected. Third, In conclusion, we have reported two rare cases of deep
independent deep peroneal nerve palsy was present in both peroneal nerve palsy with isolated lateral compartment
cases, although no case with independent deep peroneal syndrome secondary to peroneus longus tear. In both cases,
nerve palsy has previously been described among the diagnosis was difficult due to the few and complex

Deep peroneal nerve palsy with isolated lateral compartment syndrome secondary to peroneus longus tear... 177
symptoms, such as drop foot, which often occurs with 5. Nicholson P, Devitt A, Stevens M et al (1998) Acute exertional
lumbar disc herniation. Although rare, isolated lateral peroneal compartmental syndrome following prolonged horse
riding. Injury 29:643–644
compartment syndrome secondary to peroneus longus tear 6. Naidu KS, Chin T, Harris C et al (2009) Bilateral peroneal
should be considered in patients who play sports intensely compartment syndrome after horse riding. Am J Emerg Med
and develop leg pain with peroneal nerve palsy. 27(901):e903–e905
7. Meyer RS, White KK, Smith JM et al (2002) Intramuscular and
Compliance with ethical standards blood pressures in legs positioned in the hemilithotomy position :
clarification of risk factors for well-leg acute compartment syn-
Conflict of interest None. drome. J Bone Joint Surg Am Vol 84-A:1829–1835
8. Slabaugh M, Oldham J, Krause J (2008) Acute isolated lateral leg
Ethical standards All patients gave informed consent to publish compartment syndrome following a peroneus longus muscle tear.
the present report. Orthopedics 31:272
9. Mendelson S, Mendelson A, Holmes J (2003) Compartment
syndrome after acute rupture of the peroneus longus in a high
school football player: a case report. Am J Orthop 32:510–512
10. Lee RY, Colville JM, Schuberth JM (2009) Acute compartment
syndrome of the leg with avulsion of the peroneus longus muscle:
a case report. J Foot Ankle Sur: Off Pub Am College Foot Ankle
Surg 48:365–367
11. Davies JA (1979) Peroneal compartment syndrome secondary to
rupture of the peroneus longus. A case report. J Bone Joint Surg
Am 61:783–784
12. Arciero RA, Shishido NS, Parr TJ (1984) Acute anterolateral
References compartment syndrome secondary to rupture of the peroneus
longus muscle. Am J Sports Med 12:366–367
1. Moyer RA, Boden BP, Marchetto PA et al (1993) Acute com- 13. Taxter AJ, Konstantakos EK, Ames DW (2008) Lateral com-
partment syndrome of the lower extremity secondary to non- partment syndrome of the lower extremity in a recreational ath-
contact injury. Foot Ankle 14:534–537 lete: a case report. Am J Emerg Med 26(973):e971–e972
2. Gabisan GG, Gentile DR (2004) Acute peroneal compartment 14. Dombek MF, Lamm BM, Saltrick K et al (2003) Peroneal tendon
syndrome following ankle inversion injury: a case report. Am J tears: a retrospective review. J Foot Ankle Surg: Off Publ Am
Sports Med 32:1059–1061 Coll Foot Ankle Surg 42:250–258
3. Ebenezer S, Dust W (2002) Missed acute isolated peroneal 15. Gloobe H, Chain D (1973) Fibular fibrous arch. Anatomical
compartment syndrome. CJEM 4:355–358 considerations in fibular tunnel syndrome. Acta Anat 85:84–87
4. Blasier D, Barry RJ, Weaver T (1992) Forced march-induced 16. Ryan W, Mahony N, Delaney M et al (2003) Relationship of the
peroneal compartment syndrome. A report of two cases. Clin common peroneal nerve and its branches to the head and neck of
Orthopa Relat Res 284:189–192 the fibula. Clin Anat 16:501–505

26
Isolated reconstruction of the medial patellofemoral ligament
with autologous quadriceps tendon
Giovanni Vavalle1,2 • Michele Capozzi1
Abstract Kujala score increased from 35.8 preoperatively to 88.8

Background Since the role of the medial patellofemoral postoperatively and the Lysholm score improved from 43.3
ligament (MPFL) as the primary soft-tissue restraint preoperatively to 89.3 postoperatively.
against lateral patellar translation has been recognized, Conclusions Isolated MPFL reconstruction using an
several different reconstruction procedures for the treat- autologous quadriceps tendon and not requiring bone tun-
ment of patellar instability have been proposed over recent nels, may be a safe, simple and effective procedure for the
years. Many of these techniques require bony procedures treatment of patellar instability without complications such
and hardware fixation at the patellar and femoral side, as patellar fracture as reported by clinical studies using
leading to complications as described previously in the hamstring grafts. For the same reason it may also be
literature. The purpose of the present study is to describe indicated in skeletally immature patients.
the technique of isolated MPFL reconstruction using the Level of evidence Level IV.
quadriceps tendon and report the results at a mean follow-
up of 38 months. The hypothesis is that this technique, not
requiring drilling of bone tunnels on the patellar and Introduction
femoral side, may be a ‘‘simple and safe’’ mean to manage
patellar instability, giving good clinical results with low Acute patellar dislocation is a common disorder among
complication rate in selected patients with normal osseous young active patients with a 50 % recurrence rate [1].
anatomy. Several bony and soft-tissue etiologic factors can pre-
Materials and methods Sixteen consecutive patients (9 dispose patients to recurrent dislocation or subluxation of
male, 7 female; mean age 22 years) with chronic patellar the patella. Over the past decade, attention has been
instability underwent medial patellofemoral reconstruction directed to the medial patellofemoral ligament (MPFL) as a
with the superficial layer of the quadriceps tendon. All the primary passive soft-tissue restraint to lateral patellar
patients were evaluated preoperatively and postoperatively translation at 0–30° of knee flexion, providing approxi-
by physical examination and subjectively with Kujala and mately 53–67 % of the total medial restraining force [2, 3].
Lysholm scores. When patellar dislocation occurs, the MPFL is involved in
Results The average follow-up was 38 months (range [90 % of cases [4, 5].
28–48 months). No recurrent episodes of dislocation or When conservative treatment fails, different surgical
subluxation and no complications occurred. The mean techniques can be used, including proximal and distal
realignment or a combination of both.
Among the proximal procedures, there has been growing
& Giovanni Vavalle
[email protected]; [email protected]
interest in the MPFL over recent years.
Many different techniques have been advocated for the
1
Department of Orthopaedics, Saint Mary Hospital, reconstruction of the MPFL, but there is no agreement
De Ferrariis 18/D, 70124 Bari, Italy regarding the choice of graft, the type of fixation, the graft
2
Via Martiri della Giustizia, 9, 70016 Noicattaro, Italy positioning, the correct tension and the clinical outcomes.

Isolated reconstruction of the medial patellofemoral ligament with autologous quadriceps tendon 179
Furthermore, most of the techniques are associated with also routinely taken to better evaluate the geometric
several complications, such as patellar fracture [6], stiff- parameters of the patellofemoral joint.
ness [7], hemarthrosis [8], and implant [9] and wound Exclusion criteria were a tibial tubercle–trochlear grove
complications [10]. distance [20 mm, a severe trochlear dysplasia (Dejour
In 2005, Steensen et al. [9] described their original grade B, C, D) [13] or a trochlear sulcus angle [145°,
technique using the central one-third of the most superficial patella alta (Insall-Salvati index[1.2) [14], a Q angle[20°
layer of the quadriceps tendon, and leaving it attached to in female and [17° in male patients and previous surgeries
the superior pole of the patella. Once the graft is harvested, on the involved knee.
it is turned 90° medially and fixed to the femoral side with The Lysholm score [15] and the Kujala score [16] were
transosseous sutures. Their clinical experience which used to evaluate subjective knee function. Furthermore, we
involved 14 knees in 13 patients showed early promising assessed the level of satisfaction in relation to pain relief
results, no episodes of recurrent dislocations and no and functional recovery and whether patients had returned
complications. or not to their pre-trauma sport and, at which level of
After 1 year, Noyes et al. [10] presented their surgical competition, at the time of the latest follow-up
technique using the same graft fixed to the medial inter- examination.
muscular septum with no need for bone tunnels or suture Data analysis was performed with paired Students’ t test
anchors. to evaluate differences in preoperative and postoperative
Overall, the technique of Steensen et al. seemed simple subjective outcome scores.
and less invasive than other procedures previously descri- The investigation was performed according to institu-
bed in the literature, by not involving bone tunnels and tionally approved guidelines, having received informed
hardware implants. consent from all the patients.
Therefore, in 2009 we started our clinical experience
using a quadriceps graft, focusing on anatomic femoral Surgical technique
attachment, which is known to have an important effect on
length change pattern of ligaments [11, 12]. Anchor All procedures were carried out under spinal anesthesia
sutures, which are normally used in arthroscopic shoulder followed by clinical evaluation to confirm the diagnosis of
rotator cuff tendon surgery, were implanted. lateral patellar dislocation and to estimate the tightness of
Here, we describe the mid- and long-term results of the lateral retinaculum (Fig. 1). A manual lateral patellar
isolated MPFL reconstruction with an autologous quadri- translation of [50 % of its width from the center of the
ceps graft, in 16 patients affected by recurrent patellar femoral trochlea was an indirect sign of MPFL
dislocation. The hypothesis was that this technique is a incompetence.
simple and safe way to manage patellar instability, giving With the patient supine and a tourniquet applied to the
good clinical results without any bony procedures or thigh, a routine diagnostic arthroscopy was carried out to
complications, and provides fast recovery. diagnose and treat possible intra-articular lesions,
We retrospectively reviewed 16 patients (9 male, 7 female)

with recurrent patella dislocation at a mean follow-up of
38 months (range 28–48 months) after index surgery. The
average age at the time of surgery was 22 years (range
18–25 years). All patients suffered at least 4 well-docu-
mented episodes of unilateral patellar dislocation (mean 8,
maximum 20 episodes).
Nine patients practiced sports before the trauma (7
soccer and 2 volleyball); 5 at recreational level and 4 at
semi-professional level.
All patients underwent a clinical examination to detect
specific signs of patellofemoral pathology and imaging
studies (anteroposterior, lateral, standing weight-bearing
radiographs of the injured knee and bilateral skyline views Fig. 1 Evaluation under anesthesia before the surgical procedure
at 30° of flexion). Computed tomography (CT) scans were showing complete patellar lateral dislocation

particularly any chondral injuries. Arthroscopic lateral A careful subperiostal dissection was performed just
release was performed only if there was excessive tightness above its attachment on the patella, paying attention not to
of the lateral retinaculum, to avoid the risk of medial amputate the graft, extending further distally on the lateral
patellar translation when the MPFL was reconstructed. side so that the graft could be easily turned 90° medially,
Once the arthroscopy was completed and all surgical and lying flush on the anterior surface of the patella
instruments were removed, a percutaneous 1.8- (Fig. 3).
mm Kirschner wire (K-wire) was inserted at the isometric To avoid accidental detachment from the patellar
femoral point with the use of fluoroscopic guidance as insertion, multiple no. 0 nonabsorbable stay stitches were
described by Steensen et al. [17] in the beginning of the used to fix the graft to the periosteum on the anterior
clinical study and according to the radiographic landmarks patella.
defined by Schoettle et al. [18] afterwards (Fig. 2). A small 2–3 cm longitudinal skin incision was made on
A longitudinal incision was performed from the the medial femoral epicondyle, involving the implanted
superior pole of the patella, extending proximally for K-wire. A blunt dissection was carried out until the bone
almost 6–7 cm. After identification of the quadriceps was exposed. Keeping in mind that the MPFL is located on
tendon, with the knee in flexion, the superficial layer layer 2 of the medial aspect of the knee, distal to the vastus
related to the rectus femoris tendon was carefully har- medialis obliquus (VMO) [17], the graft was passed
vested using a no. 15 scalpel blade, without violating the through a soft tunnel between the deep fascia and the
joint capsule, leaving it attached to the superomedial capsule by using a clamp, avoiding penetration of the joint
border of the patellar pole. We tried to harvest the most [9] (Fig. 4).
medial portion of the tendon, leaving 3–4 mm of medial The tourniquet was deflated and hemostasis was per-
rectus femoris tendon, in order to be closer to the medial formed before the graft was fixed. We believe that an
border of the patella and to reduce the distance from the inflated tourniquet can affect patellar tracking due to its
side of the femoral attachment. Usually, the harvesting binding effect on the extensor mechanism.
starts 2–3 cm above the patellar pole, where the interval The graft was then tied over the guiding pin and isom-
between the superficial and middle layer of the quadri- etry was evaluated by several full flexions and extensions
ceps tendon is better identified. The native graft width of the knee. It is important not to over-constrain the
and thickness measured approximately 10–12 and 5 mm, reconstruction, avoiding increased contact pressures in the
respectively, while the length was estimated by measur- patellofemoral joint, and subsequent development of
ing the distance from the superior pole of the patella to arthrosis. The correct patellar tracking was also assessed
the K-wire which was implanted in the medial femoral dynamically under arthroscopic view (Fig. 5).
site, with the knee in a 30° flexion. The graft was then fixed using a 5.5-mm diameter tita-
Once the appropriate length was reached, the graft was nium screw-type anchor (Fastin-RC; Mitek, Westwood,
cut proximally and an Ethibond (Ethicon, Somerville, NJ, MA, USA) loaded with two strands of no. 2 nonabsorbable
USA) no. 2 was stitched at the free end with a Krackow Orthocord suture (Ethicon) at approximately 30° knee
continuous suture. flexion, where the patella was stabilized into the femoral
Fig. 2 Intraoperative view showing the percutaneous placement of a

1.8-mm K-wire in an isometric femoral insertion Fig. 3 The quadriceps graft is harvested and rotated 90° medially

groove (Fig. 6). In six cases, additional trans-osseus sutures

were placed to reinforce the femoral fixation.
The lower border of the VMO was sutured onto the
superior limb of the graft with multiple stitches to enhance
the reconstruction and improve medial patellar constraint.
The wounds were closed in a standard layered fashion
and a hinged brace was worn for 4 weeks after surgery and
locked in full extension for the first 3 weeks. Weight
bearing was allowed as tolerated with the use of crutches.
Quadriceps muscle strengthening was immediately
initiated.
A progressive recovery of flexion started the day after
surgery, avoiding knee flexion [90° in the first 3 weeks.
Fig. 4 The quadriceps graft is stitched at the free end with no. 2 Results
nonabsorbable sutures and is passed medially through a soft-tissue
tunnel with the use of a clamp
At the final follow-up, no patients reported re-dislocation
or subluxation of the patella and no complications occurred
postoperatively. All patients regained full range of motion
of the knee and no patient showed a positive apprehension
test at the most recent follow-up. The endpoint of lateral
passive patellar translation was firm in all patients.
During the arthroscopic procedure, Outerbridge [19]
stage 2 chondral lesions of the patella were observed in 5
knees (3 in 3 knees and 4 in 2 knees). A free patellar
osteochondral fragment was removed from one knee. No
degenerative cartilage changes were detected.
The average Kujala score improved significantly from
35.8 ± 5.5 to 88.84 ± 4.3 postoperatively (p \ 0.001).
The mean modified Lysholm score improved significantly
from 43.3 ± 6.4 to 89.3 ± 3.1 (p \ 0.001). The mean
improvement in the Kujala score was 53.0 ± 4.3 and the
mean improvement in the Lysholm score was 46.0 ± 3.1
Fig. 5 Arthroscopic view showing an optimal patellofemoral (Table 1).
congruence
Fig. 6 Postoperative radiograph showing the correct insertion of the suture anchor at the isometric femoral point as described by Steensen et al.

Table 1 Functional outcome

Mean preoperative Mean postoperative Mean improvement p value
analysis
score score in scores
Kujala 35.8 ± 5.5 88.8 ± 4.3 53.0 ± 4.3 \0.001

Lysholm 43.3 ± 6.4 89.3 ± 3.1 46.0 ± 3.1 \0.001
Ten patients rated the surgical procedure as ’very sat- as proximal realignment procedures in patellar instability.
isfactory’ (64 %), 4 as ’satisfied’ (21 %) and 2 as ’partially Unfortunately, the VMO and superficial medial retinacu-
satisfied’ (15 %). None were dissatisfied with the surgery. lum serve only as a minor restraint to lateral displacement,
Partially satisfied results were found in the two patients so these procedures do not address the real pathoanatomy
with Outerbridge stage 3–4 chondral damage. of patellar dislocation, which is the MPFL injury.
Five out of 9 patients that practiced sports returned to Multiple techniques have been described recently in the
sports at the same pre-injury level (3 semi-professional and literature for the reconstruction of the MPFL, differing in
2 recreational level), 1 reduced their level (from semi- the type of graft and in the method of fixation to the femur
professional to recreational) and 3 retired from sport (all and patella; however, at present there is no consensus as to
recreational) because they were no longer motivated to the best method. Furthermore, no clear clinical superiority
continue or for fear of new injuries. of one surgical technique over another is evident.
The hamstrings are the most frequent grafts used in the
reconstruction of the MPFL, with different types of fixa-
Discussion tion. Most of them entail bone tunnels to the patella and to
the femoral condyle, sometimes leading to potential com-
Of the many proximal patellar realignment procedures, plications such as patellar fractures. Many authors [26–28]
MPFL reconstruction has become one of the most fre- have already described iatrogenic patellar fractures that
quently used methods since its anatomy and key role in occurred because of violation of the anterior patellar cortex
guiding the patella into the femoral groove in the first 30° during drilling of the bone tunnels.
of knee flexion has been well recognized. Recently, Parikh et al. [29] reported an 16.2 % inci-
For many years, the MPFL was thought to be an dence of complications after MPFL reconstruction in a
inconstant anatomic structure, present in only 29–88 % of large series of young patients, using a hamstring graft and
knees [2, 20, 21], and its importance in stabilizing the performing two bone tunnels on the patella and one on the
patella has been underestimated. In fact, being a relatively femur; 47 % of complications were considered to be rela-
thin fascial band, the MPFL is sometimes difficult to ted to improper technique, including malposition of the
identify in all knees. femoral tunnel and patellar fractures.
Nowadays, a lesion of the MPFL is considered to be an Concerns have also arisen about the relationship
’essential lesion’, comparable to the Bankart lesion in between femoral tunnel and distal femoral physis in
anterior shoulder instability, without which the patella skeletally immature patients with possible growth distur-
cannot laterally dislocate [22]. bance risks.
Nomura et al. [23] showed, in a cadaver knee, that by Furthermore, considering the anatomy of native MPFL
applying a lateral displacing force of 10 N, the patella as a thin fascial band, hamstring tendons fail to reproduce
displaced approximately 6 and 13 mm with an intact and basic anatomic features. Being thick, cord-like and much
transected MPFL, respectively. stronger, reconstruction with hamstring grafts can lead to
Similar results were found by Hautamaa et al. [24]. In medial patellar overload and excessive tensioning of the
two different studies using a tensile machine on cadaver graft, thus causing graft failure with the risk of patellofe-
knees, Conlan et al. [2] and Desio et al. [3] found that moral arthritis.
MPFL resisted 53–60 % of the force needed to cause a Finally, the technique described is also economically
patellar lateral displacement, respectively. They also noted advantageous in terms of not requiring the use of implants
that the role of MPFL in resisting patellar displacement for patellar fixation.
was greater with the knee fully extended, while it became When we started MPFL reconstructions in 2009,
less important for flexions [20°. These findings are sup- Steensen et al.’s technique [9]), using the most superficial
ported by anatomic studies [25] on length change patterns layer of the quadriceps tendon, seemed simpler and less
of the MPFL, suggesting that the retinacular structures invasive than other procedures described in the literature.
tighten in the extended knee and slacken as it flexes. This technique did not involve bone tunnels and the need
Imbrication of the VMO and/or the superficial medial for hardware implants, thus avoiding the complications
retinaculum has been widely performed over the past years mentioned above.

Fig. 7 CT scan showing the correction of patellar maltracking after a surgical procedure
Furthermore, by leaving the graft attached to the patellar To avoid bone tunnels on the femoral side and improve
and then turning it 90° medially, where it is sutured to the the stability of fixation, the graft is fixed using an 5.5-mm
patellar retinaculum, the isometric patellar site of the diameter titanium screw-type anchor, normally used in
MPFL is easily restored, with no need for bone holes and shoulder surgery, loaded with two strands of no. 2 nonab-
hardware, thus simplifying the procedure. sorbable suture, and occasionally, additional trans-osseus
The study by Steensen et al. involved 14 knees in 13 sutures are performed. Thus, this technique can be used
patients, with no episodes of recurrent dislocations and no regardless of skeletal maturity.
complications. In addition, the lower border of the vastus medialis is
Therefore, we adopted their technique and tried to sutured distally to the graft, in order to enhance the MPFL,
improve some critical aspects complying with the most and to give it an active role in the stability of the patella.
recent biomechanical advancements and anatomic knowl- In our current study, this surgical technique has proven
edge arising from the international literature. to be effective. The study included 16 patients with
It is known that femoral insertion of MPFL reconstruc- recurrent patella dislocation at a mean follow-up of
tion is the most sensitive to reproduce proper ligament 38 months who experienced at least 4 well-documented
isometry, similar to the femoral attachment of the anterior episodes of unilateral patellar dislocation (mean 8, maxi-
cruciate ligament, so nonanatomic femoral tunnel posi- mum 20 episodes). There were no recurrent dislocations
tioning can lead to unfavorable outcomes [12, 25]. Proximal after surgery and no patient showed a positive apprehen-
placement of the femoral attachment might cause tightening sion test, with 85 % of patients being satisfied/very satis-
of the MPFL reconstruction in flexion with overload of the fied (Fig. 7). No complications occurred postoperatively.
medial patellar facet. Instead, by placing the femoral origin Our results are similar to previous studies [30–32].
distally, the MPFL graft tightens in extension, causing non- The limited number of patients involved in the current
physiologic patellar motion or stretching the graft. study is related to highly careful preoperative selection,
Therefore, using fluoroscopic guidance, the isometric performing isolated reconstruction of MPFL only in cases
femoral point is accurately identified according to the without any predisposing factors such as abnormal Q angle,
radiographic landmarks described by Schoettle et al. [18], trochlear dysplasia, patella alta and malalignment of the
and a percutaneous 1.8-mm K-wire is inserted. It is located lower extremity.
just anterior to the intersection of the posterior femoral Although this is a small study, there are several inter-
cortical line and Blumensaat line on the lateral radiograph. esting observations that can be made. First, all procedures
Once the graft is harvested and placed under the first were performed by one surgeon (GV). Second, the average
layer, it is temporarily tied over the guiding pin and its follow-up rate was 38 months, i.e., a period long enough to
isometry is evaluated by several full flexions and exten- evaluate the effectiveness of a surgical procedure in the
sions of the knee, avoiding an over-constrained recon- treatment of patellar instability.
struction. It is important to verify at the end of the Some limitations to this study must also be considered.
procedure that there is some lateral patellar translation with The postoperative evaluations were not blinded and there
the knee extended, with a firm lateral endpoint. Further- was no control group. In addition, being retrospective in
more, the correct patellar tracking is also assessed nature, our study has limitations similar to other retro-
dynamically under arthroscopic view. spective studies.

In conclusion, MPFL reconstruction using a quadriceps 11. Bicos J, Carofino B, Andersen M, Schepsis AA, Fulkerson JP,
tendon graft is a relatively simple procedure which elimi- Mazzocca A (2006) Patellofemoral forces after medial patello-
femoral ligament reconstruction: a biomechanical analysis.
nates the need for bone tunnels. It is safe (given that we J Knee Surg. 19(4):317–326
observed no complications) and effective with no cases of 12. Elias JJ, Cosgarea AJ (2006) Technical errors during medial
dislocation or subluxation and with a patient satisfaction patellofemoral ligament reconstruction could overload medial
rate [80 %. Correct indications and overall careful pre- patellofemoral cartilage: a computational analysis. Am J Sports
Med 34(9):1478–1485
operative evaluation to rule out any bony abnormalities, are 13. Dejour D, Le Coultre B (2007) Osteotomies in patello-femoral
key to successful long-term outcomes. As the technique instabilities. Sports Med Arthrosc 15(1):39–46
does not require bone tunnels to be drilled, it may be 14. Insall J, Salvati E (1971) Patella position in the normal knee joint.
indicated in skeletally immature patients [33]. Radiology 101(1):101–104
15. Tegner Y, Lysholm J (1985) Rating systems in the evaluation of
knee ligament injuries. Clin Orthop Relat Res 198:43–49
16. Kujala U, Jaakkola LH, Koskinen SK, Taimela S, Hurme M,
Nelimarkka O (1993) Scoring of patellofemoral disorders.
Conflict of interest The authors declare that they have no conflict Arthroscopy 9:159–163
of interest. 17. Steensen RN, Dopirak RM, McDonald WG III (2004) The
anatomy and isometry of the medial patellofemoral ligament:
Ethical standards The authors state that the patients gave the implications for reconstruction. Am J Sports Med 32:1509–1513
informed consent prior being included into the study; The study was 18. Schoettle PB, Schmeling A, Rosenstiel N, Weiler A (2007)
authorized by the local ethics committee and was performed in Radiographic landmarks for femoral tunnel placement in medial
accordance with the Ethical standards of the 1964 Declaration of patellofemoral ligament reconstruction. Am J Sports Med
Helsinki and its later amendments. 35(5):801–804
19. Outerbridge HK, Outerbridge AR, Outerbridge RE (1995) The
use of a lateral patellar autologous graft for the repair of a large
osteochondral defect in the knee. J Bone Joint Surg Am
77(1):65–72
20. Reider B, Marshall DVM, Koslin B, Ring B, Girgis FG (1981)
The anterior aspect of the knee joint. J Bone J. Surg Am
63:351–356
21. Placella G, Tei MM, Sebastiani E, Criscenti G, Speziali A,
Mazzola C, Cerulli N (2014) Shape and size of the medial
References patellofemoral ligament for the best surgical reconstruction: a
human cadaveric Study. Knee Surg Traumatol Arthrosc
1. Colvin AC, West RV (2008) Patellar instability. J Bone Joint 22:2327–2333
Surg Am 90(12):2751–2762 22. Sillanpää PJ, Peltola E, Mattila VM, Kiuru M, Visuri T, Pihla-
2. Conlan T, Garth WP, Lemons JE (1993) Evaluation of the medial jamäki H (2009) Femoral avulsion of the medial patellofemoral
soft-tissue restraints of the extensor mechanism of the knee. ligament after primary traumatic patellar dislocation predicts sub-
J Bone Joint Surg Am 75:682–693 sequent instability in men. Am J Sports Med 37(8):1513–152123
3. Desio SM, Burks RT, Bachus KN (1998) Soft tissue restraints to 23. Nomura E, Horiuchi Y, Kihara M (2000) Medial patellofemoral
lateral patellar translation in the human knee. Am J Sports Med ligament restraint in lateral patellar translation and reconstruc-
26:59–65 tion. Knee 7:121–127
4. Sallay PI, Poggi J, Speer KP, Garrett WE (1996) Acute dislo- 24. Hautamaa PV, Fithian DC, Kaufman KR, Daniel DM, Pohlmeyer
cation of the patella: a correlative pathoanatomical study. Am J AM (1998) Medial soft tissue restraints in lateral patellar insta-
Sports Med 24:52–60 bility and repair. Clin Orthop Relat Res 349:174–182
5. Sanders TG, Morrison WB, Singleton BA, Miller MD, Cornum 25. Amis AA, Firer P, Mountney J, Senavongse W, Thomas NP
KG (2001) Medial patellofemoral ligament injury following acute (2003) Anatomy and biomechanics of the medial patellofemoral
transient dislocation of the patella: MR findings with surgical ligament. Knee. 10:215–220
correlation in 14 patients. J Comput Assist Tomogr 25:957–962 26. Christiansen SE, Jacobsen BW, Lund B, Lind M (2008) Recon-
6. Panni SA, Alam M, Cerciello S, Vasso M, Maffulli N (2011) Medial struction of the medial patellofemoral ligament with gracilis
patellofemoral ligament reconstruction with a divergent patellar tendon autograft in transverse patellar drill holes. Arthroscopy
transverse 2-tunnel technique. Am J Sports Med 39:2647–2655 24:82–87
7. Drez D Jr, Edwards TB, Williams CS (2001) Results of medial 27. Ellera Gomes JL, Stigler Marczyk LR, Cesar PC, Jungblut CF
patellofemoral ligament reconstruction in the treatment of (2004) Medial patellofemoral ligament reconstruction with
patellar dislocation. Arthroscopy 17:298–306 semitendinosus autograft for chronic patellar instability: a follow-
8. Steiner TM, Torga-Spak R, Teitge RA (2006) Medial patellofe- up study. Arthroscopy 20:147–151
moral ligament reconstruction in patients with lateral patellar 28. Dhinsa BS, Bhamra JS, James C, Dunnet W, Zahn H (2013)
instability and trochlear dysplasia. Am J Sports Med 34:1254–1261 Patella fracture after patellofemoral ligament reconstructiom
9. Steensen RN, Dopirak RM, Maurus PB (2005) A simple tech- using suture anchors. Knee. 20:605–608
nique for reconstruction of the medial patellofemoral ligament 29. Parikh SN, Nathan ST, Wall EJ, Eismann EA (2013) Compli-
using a quadriceps tendon graft. Arthroscopy 21(3):365–370 cations of medial patellofemoral ligament reconstruction in
10. Noyes FR, Albright JC (2006) Reconstruction of the medial young patients. Am J Sports Med 41(5):1030–1038
patellofemoral ligament with autologous quadriceps tendon. 30. Goyal D (2013) Medial patellofemoral ligament reconstruction:
Arthroscopy 22(8):904–907 the superficial quad technique. Am J Sports Med 41:1022–1029

31. Fink C, Veselko M, Herbort M, Hoser C (2014) MPFL recon- 33. Nelitz M, Williams SR (2014) Anatomic reconstruction of the
struction using a quadriceps tendon graft. Knee 21:1175–1179 medial patella-femoral ligament in children and adolescents using
32. Kumara K, Renganathan S, Josepf C, Easwar TR, Rajan D (2014) a pedicled quadriceps tendon graft. Arthrosc Tech. 3:303–308
Medial patellofemoral ligament reconstruction in patellar insta-
bility. Indian J Orthop 48:501–505

27
Platelet-rich plasma versus autologous blood versus steroid
injection in lateral epicondylitis: systematic review and network
meta-analysis
Alisara Arirachakaran1 • Amnat Sukthuayat1 • Thaworn Sisayanarane1 •
Sorawut Laoratanavoraphong1 • Wichan Kanchanatawan2 • Jatupon Kongtharvonskul3
Abstract improved effects with unstandardized mean differences

Background Clinical outcomes between the use of pla- (UMD) in pain visual analog scale (VAS), Disabilities of
telet-rich plasma (PRP), autologous blood (AB) and corti- Arm Shoulder and Hand (DASH), Patient-Related Tennis
costeroid (CS) injection in lateral epicondylitis are still Elbow Evaluation (PRTEE) score and pressure pain
controversial. threshold (PPT) of -2.5 (95 % confidence interval, -3.5,
Materials and methods A systematic review and network -1.5), -25.5 (-33.8, -17.2), -5.3 (-9.1, -1.6) and 9.9
meta-analysis of randomized controlled trials was con- (5.6, 14.2), respectively. PRP injections also showed sig-
ducted with the aim of comparing relevant clinical outcomes nificantly improved VAS and DASH scores when com-
between the use of PRP, AB and CS injection. Medline and pared with CS. PRP showed significantly better VAS with
Scopus databases were searched from inception to January UMD when compared to AB injection. AB injection has a
2015. A network meta-analysis was performed by applying higher risk of adverse effects, with a relative risk of 1.78
weight regression for continuous outcomes and a mixed- (1.00, 3.17), when compared to CS. The network meta-
effect Poisson regression for dichotomous outcomes. analysis suggested no statistically significant difference in
Results Ten of 374 identified studies were eligible. When multiple active treatment comparisons of VAS, DASH and
compared to CS, AB injection showed significantly PRTEE when comparing PRP and AB injections. However,
AB injection had improved DASH score and PPT when
compared with PRP injection. In terms of adverse effects,
& Jatupon Kongtharvonskul AB injection had a higher risk than PRP injection.
[email protected]
Conclusions This network meta-analysis provided addi-
Alisara Arirachakaran tional information that PRP injection can improve pain and
[email protected]
lower the risk of complications, whereas AB injection can
Amnat Sukthuayat improve pain, disabilities scores and pressure pain thresh-
[email protected]
old but has a higher risk of complications.
Thaworn Sisayanarane Level of evidence Level I evidence
[email protected]
Sorawut Laoratanavoraphong Keywords Lateral epicondylitis PRP Autologous
[email protected]
blood Corticosteroid Systematic review Network meta-
Wichan Kanchanatawan analysis
[email protected]
1
Orthopedics Department, Police General Hospital, Bangkok,
Thailand Introduction
2
Orthopedics Department, Lerdsin General Hospital,
Bangkok, Thailand Lateral epicondylitis is the most commonly diagnosed
3
Section for Clinical Epidemiology and Biostatistics, Faculty condition of the elbow [21], with a prevalence of 1–3 %
of Medicine, Ramathibodi Hospital, Bangkok, Thailand in the general population [34]. It affects men and women

Platelet-rich plasma versus autologous blood versus steroid injection in lateral epicondylitis: systematic... 187
equally, mainly in the age range of 35-55 years [2, 28]. significant difference. However, these meta-analyses
In most cases of lateral epicondylitis, no obvious under- included too few studies for pooling of the outcomes,
lying etiology can be identified [25]. However, any utilized standardized mean difference, and lacked proper
activity that involves overuse of the wrist extensor or methodological quality required for performing a network
supinator muscles may be incriminating. The most com- meta-analysis. Neither heterogeneity nor sources of
monly affected muscle is the extensor carpi radialis brevis heterogeneity (age, sex, disease duration, preparation of
(ECRB), as originally described by Cyriax [2]. The the intervention and time to assess the outcome) were
pathology of lateral epicondylitis was previously consid- assessed. Moreover, other RCTs [23, 24, 26] have been
ered to be from tendinitis, arising as inflammation of the published since this study was done. Therefore, a sys-
tendon [18]. Histopathologically, it has been shown to tematic review was conducted with a network meta-
have a paucity of inflammatory cells such as macrophages analysis of RCTs at multiple follow-up times with the aim
and neutrophils [7, 11]. The condition is therefore con- of comparing relevant clinical outcomes [visual analog
sidered to be a form of tendinosis, which is defined as a score, DASH score, Patient-Related Tennis Elbow Eval-
degenerative process [2]. The treatment of lateral epi- uation (PRTEE) score, adverse effects and non-response
condylitis includes rest, nonsteroidal anti-inflammatory rates] between AB, PRP and corticosteroids.
medication, bracing, physical therapy, extracorporeal
shock wave therapy and botulinum toxin injection.
Injection of corticosteroids (once the gold standard but Materials and methods
now considered controversial), whole blood and platelet-
rich plasma (PRP), and various types of surgical proce- Search strategy
dures have also been recommended [4, 8, 17, 27, 29, 35].
Injection with corticosteroids has been used since the The Medline and Scopus databases were used to identify
1950s and has been the treatment of choice for many relevant studies published in English from the date of
years. However, several studies have shown no long-term inception to January 18, 2015. The PubMed and Scopus
beneficial effect; several alternative biologic injection search engines were used to locate studies using the fol-
therapies have therefore become available. Complex lowing search terms: ‘lateral epicondylitis’ and ‘platelet-
growth factor preparations, derived from the patients’ own rich plasma’ and ‘clinical trial’. Relevant studies from the
(autologous) blood, are used to drive the body’s own reference lists of identified studies and previous systematic
tissue-healing mechanisms in the hope of stimulating reviews were also explored.
rapid healing mechanisms [5]. Two different preparations
that are most described in the literature are autologous Selection of studies
whole blood (AB) and platelet-rich plasma (PRP) injec-
tion [5, 10, 12, 14, 19, 21, 23, 24, 26, 33]. There have Identified studies were selected by one author (J.K.) and
been several randomized controlled trials (RCTs) that randomly checked by A.A. Their titles and abstracts were
have compared AB with PRP injection [5, 23, 24, 33], initially screened; full papers were then retrieved if a
AB with steroid injection[12, 26] and PRP with steroid decision could not be made from the abstracts. The reasons
injection [10, 14, 19, 21]. However, results as to whether for ineligibility or exclusion of studies were recorded and
PRP, AB or corticosteroids is more beneficial are still described (Fig. 1).
unclear. Previous systematic reviews by Krogh et al. [13]
including 17 studies have shown eight different injection Inclusion criteria
therapies reported by network meta-analysis. The results
showed that AB, PRP and corticosteroids were more Randomized controlled trials or quasi-experimental designs
efficacious than placebo [estimated by standardized mean comparing clinical outcomes between treatments in lateral
difference (SMD)]; however, there were no reports com- epicondylitis patients were eligible if they met the fol-
paring the efficacy of PRP versus AB, PRP versus corti- lowing criteria:
costeroids and AB versus corticosteroids. Ahmad et al. [1]
– compared clinical outcomes between PRP, AB and
showed that PRP was more efficacious than blood injec-
corticosteroid injection
tion in terms of non-response rate and conversion to
– compared at least one of the following outcomes: visual
surgery rates as well as pain visual analog score (VAS),
analog score, DASH score, PRTEE score, pressure pain
and that PRP was more efficacious than corticosteroid
threshold (PPT), adverse effects and non-response rates
injections in terms of pain and Disabilities of the Arm,
– had sufficient data to extract and pool, namely reported
Shoulder and Hand (DASH) score in only one of three
mean, standard deviation (SD), and numbers of subjects
studies, but two other studies showed no clinically

16 studies retrieved 373 studies retrieved

from Medline from Scopus
374 le aer

removed duplicates
365 studies were ineligible

344 non-RCTs
9 no intervenon
12 no disease
- 2 Carlage disease
- 4 Rotator cuﬀ injury
- 3 Achilles tendinis
- 3 Plantar fasciis
1 study from hand

10 studies le for
searching from
full paper
reference list
Visual Analog Score PATIENT-RATED TENNIS ELBOW EVALUATION

Platelet Rich Plasma vs Corcosteroid: 3 studies Platelet Rich Plasma vs Corcosteroid: 1 study
Platelet Rich Plasma vs Autologous Blood:2 studies Platelet Rich Plasma vs Autologous Blood: 1 study
Autologous Blood vs Corcosteroid: 2 study Autologous Blood vs Corcosteroid: 1 study
Disabilies of the arm, shoulder and hand Pain Pressure Threshold

Platelet Rich Plasma vs Corcosteroid: 2 studies Platelet Rich Plasma vs Autologous Blood: 2 studies
Autologous Blood vs Corcosteroid: 1 study Autologous Blood vs Corcosteroid: 1 study
Fig. 1 Flow-chart of study selection
according to treatments for continuous outcomes; Risk of bias assessment

number of patients according to treatment for dichoto-
mous outcomes Two authors (J.K. and T.A.) independently assessed the
risk of bias for each study. Six study quality domains were
considered, namely sequence generation, allocation con-
Data extraction
cealment, blinding (participant, personnel, and outcome
assessors), incomplete outcome data, selective outcome
Two reviewers (J.K. and A.A.) independently performed
reporting, and other sources of bias [15]. Disagreements
data extraction using standardized data extraction forms.
between two authors were resolved by consensus and dis-
General characteristics of the subjects (e.g., mean age,
cussion with a third party (A.T.).
gender, dominant side, duration of disease, pain score,
disabilities scores and PPT at baseline) were extracted. The
Outcomes
number of subjects, mean and SD of continuous outcomes,
namely pain by VAS, DASH score, PRTEE score and PPT
The outcomes of interest were pain VAS, DASH score,
between groups, were extracted. Cross-tabulated frequen-
PRTEE score, complications and non-response rates.
cies between treatment and adverse effects were also
Methods of measurements of these outcomes were used
extracted. Any disagreements were resolved by discussion
according to the original studies. Briefly, this includes the
and consensus with a third party (S.L.).

VAS pain scale of 0–10, the DASH score which consists of Results
30 items with total scores ranging from 0 to 100, the
PRTEE which consists of pain disability and functional Sixteen and 373 studies from Medline and Scopus were
disability with a total score ranging from 0 to 100, and identified, respectively; 15 studies were duplicates, leaving
pressure pain threshold (PPT) which was assessed by an 374 studies for review of titles and abstracts. Of these, nine
algometer with scale units in kg/cm2. Postoperative adverse studies [5, 12, 14, 19, 21, 23, 24, 26, 33] plus one study [6]
effects (skin reaction and local injection site pain) and non- identified from reference lists were reviewed, leaving a
response rates were considered. total of ten studies for data extraction. Characteristics of
the 10 studies [5, 6, 12, 14, 19, 21, 23, 24, 26, 33] are given
Statistical analysis in Table 1. Of seven PRP studies [5, 14, 19, 21, 23, 24, 33],
the comparators included AB in four studies [5, 23, 24, 33],
Direct comparisons of continuous outcomes measured at and steroids in three studies [14, 19, 21]. All three studies
the end of each study between PRP, AB and corticosteroid regarding AB were in comparison with steroids. Most
injection were pooled using an unstandardized mean dif- studies [5, 6, 14, 21, 23, 26, 33] assessed outcomes at more
ference (UMD). Heterogeneity of the mean difference than 2 months; only three studies [12, 19, 24] assessed
across studies was checked using the Q-statistic and the outcomes at 1.5–2 months. Mean age, dominant side,
degree of heterogeneity was quantified using the I2 statistic. duration of disease and VAS before treatment varied from
If heterogeneity was present as determined by a statistically 34 to 50 years, 57 to 85 %, 5 to 18 months and 5.5 to 7.6,
significant Q-statistic or by I2 [ 25 %, the UMD was respectively. The percentage of males ranged from 18 to
estimated using a random effects model; otherwise a fixed 57 %. Various outcomes were compared between treat-
effects model was applied. ment groups (Fig. 1).
For dichotomous outcomes, a relative risk (RR) of
adverse effect of treatment comparisons at the end of each Risk of bias in included studies
study was estimated and pooled. Heterogeneity was
assessed using the previous method. If heterogeneity was The risk of bias assessment is described in Table 2.
present, the Dersimonian and Laird method [3] was applied
for pooling. If not, the fixed effects model by inverse Direct comparisons
variance method was applied. Meta-regression was applied
to explore the source of heterogeneity (e.g., mean age, Data for direct comparisons of all treatments and outcomes
percentage of females, duration of disease, dominant hand measured at the end of each study are given in Table 1.
side and follow-up time) if data was available. Publication Pooling according to outcomes was performed if there
bias was assessed using contour-enhanced funnel plots [20, were at least two studies for each comparison, as clearly
22] and Egger tests [9]. described below. There was no evidence of publication bias
For indirect comparisons, network meta-analyses were by Egger’s test for both pooled effects of all outcomes from
applied to assess all possible effects of treatment if sum- direct comparison.
mary data was available for pooling [16, 30, 31]. A linear
regression model, weighted by inverse variance, was Visual analog score
applied to assess the treatment effects for continuous out-
comes. For postoperative complications, a mixed-effect In seven studies [6, 12, 19, 21, 23, 24, 33], the UMD of
Poisson regression was applied to assess treatment effects -1.7 (95 % CI -2.6, -0.8) and -2.5 (95 % CI -3.5,
[16]. Summary data was expanded to individual patient -1.5) showed that there was significantly lower VAS for
data using the ‘‘expand’’ command in STATA. Treatment PRP and AB, respectively, than for steroids (Table 3). The
was considered as a fixed effect whereas the study variable UMD was homogeneous (I2 = 0) with a value of -1.1
was considered as a random effect in a mixed-effect model. (95 % CI -1.3, -0.8), showing that VAS was significant
The pooled RR and its 95 % confidence intervals (CIs) lower for PRP than AB.
were estimated by exponential coefficients of treatments.
All analyses were performed using STATA version 13.0 Disabilities of the Arm Shoulder and Hand score
[32]. P \ 0.05 was considered statistically significant,
except for the test of heterogeneity where P \ 0.10 was In three studies [12, 19, 21], the UMD of -16.3 (95 % CI
used. -22.3, -10.4) and -25.5 (95 % CI -33.8, -17.2) showed

190
Table 1 Characteristics of included studies

References Years Journal Intervention Comparator Follow-up Preparation of Age Sex Dominant Duration VAS (0–10) Outcome
(months) intervention (years) (male %) side (%) (months) before
treatment
Kazemi et al. [12] 2010 AJPMR AB Steroid 2 Mixed blood with 47.1 18 60 – 6.60 VAS, DASH,
1 ml of 2 % PPT, adverse
lidocaine effect
Peerboom et al. [21] 2010 AMJ PRP Steroid 12 The GPS III system 47.1 48 63 – 6.80 VAS, DASH,
adverse
effect
Thanasas et al. [33] 2011 AMJ PRP AB 6 The GPS III system 36.3 25 86 5.1 6.05 VAS, adverse
effect
Dojode [6] 2012 BJR AB Steroid 6 Mixed blood with 42.5 42 85 8.6 7.6 VAS, adverse
1 ml of 0.5 % effect
bupivacaine
Omar et al. [19] 2012 Egypt PRP Steroid 1.5 Other system 37.5 37 60 1.8 8.40 VAS, DASH
Rheumatologist
Singh et al. [26] 2013 J Health Allied AB Steroid 3 Mixed blood with 34.1 47 57 18.1 – PRTEE
Sci 1 ml of 2 %
lidocaine
Krogh et al. [14] 2013 AMJ PRP Steroid 3 The Recover GPS II 45.8 50 80 – 5.55 PRTEE,
system adverse
effect
Creaney et al. [5] 2014 BJSM PRP AB 6 Other system 50.1 56.5 – – – PRTEE
Raeissadat et al. [23] 2014 BMC Sports PRP AB 12 The Rooyagen kit 43.5 23 67 – 6.95 VAS, PPT
Science
Raeissadat et al. [24] 2014 Pain Research PRP AB 2 The Rooyagen kit 46.3 40 65 – 7.00 VAS, PPT
and Treatment
AB autologous blood injection, Steroid corticosteroid, PRP platelet-rich plasma, VAS visual analog score, DASH Disabilities of the Arm Shoulder and Hand, PRTEE Patient-Rated Tennis Elbow
Evaluation, PPT pressure pain threshold

Table 2 Risk of bias assessment

References Adequate sequence Adequate allocation Blinding Address incomplete Selective Free of Description of
generation concealment outcome data outcome report other bias other bias
Kazemi et al. Y N Y N Y N Did not

[12] mention to
ITT
Peerboom Y Y Y Y Y Y –
et al. [21]
Thanasas Y N Y Y Y Y –
et al. [33]
Dojode [6] U N N Y Y Y –
Omar et al. U N N N Y N Per protocol
[19] analysis
Singh et al. U N N N Y N Per protocol
[26] analysis
Krogh et al. Y Y Y Y Y Y –
[14]
Creaney U Y Y Y Y Y –
et al. [5]
Raeissadat Y N Y N Y N Per protocol
et al. [23] analysis
Raeissadat Y N N N Y N Per protocol
et al. [24] analysis
that there was a significantly lower DASH score for PRP study [21] reported non-response rates. The pooled RR was
and AB, respectively, than for steroids (Table 3). 1.23 (95 % CI 1.01, 1.49), which showed a significantly
higher risk of non-response after PRP injection when
Patient-Related Tennis Elbow Evaluation score compared with steroid injection.
In three studies [5, 14, 26], the UMD of -7.3 (95 % CI Network meta-analysis
-13.8, -0.9) and -5.3 (95 % CI -9.1, -1.6) showed that
there was a significantly lower PRTEE score for PRP and Visual analog score
AB, respectively, than for steroids (Table 3). The UMD of
-11.0 (95 % CI -18.3, -3.7) showing that the PRTEE Seven studies [6, 12, 19, 21, 23, 24, 33] were included in the
score was significant lower for PRP than AB. network meta-analysis. After being adjusted by time, the
regression analysis suggested that for assessment within
Pressure pain threshold 2 months, the mean differences in VAS for PRP and AB
showed that the VAS was lower than for steroid injection,
In three studies [12, 23, 24], the UMD of 9.9 (95 % CI 5.6, but these were not significantly different (as seen in Table 4;
14.2) showed that there was a significantly higher PPT Fig. 2a). For assessment at the last follow-up, the mean
score for AB than steroids (Table 3). The UMD of 2.5 difference in VAS for PRP and AB injection was lower, with
(95 % CI -1.5, 6.5) showing that PPT was higher for PRP statistical significance, than for steroid injection.
than AB, but this was not significant.
Disabilities of the Arm Shoulder and Hand score
Adverse effects (local pain and skin reaction) and non-
response rates Three studies [12, 19, 21] were included in the network
meta-analysis. After being adjusted for time frame, the
In five studies [6, 12, 14, 21, 33], the pooled RR was 1.78 regression analysis suggested that for assessment within
(95 % CI 1.00, 3.17), which showed a significantly higher 2 months, the mean difference in DASH score for AB
risk of complications after AB injection when compared injection was statistically significantly lower than for PRP
with steroids, and no heterogeneity (I2 = 0) was present and steroid injection, with a value of -38.66 (95 % CI
(Table 3). Compared with PRP, the pooled RR for AB and -56.83, 20.48) and -24.27 (95 % CI -40.68, 7.86),
steroids had no statistically significant difference. Only one respectively (Table 4; Fig. 2b). However, assessment at the

Table 3 Summarized results of direct comparisons according to type of interventions

Clinical outcomes No. of studies I2 No. of subjects UMD (95 % CI)
VAS
PRP vs. AB 3 0 72 vs. 72 -1.1 (-1.3, -0.8)*
PRP vs. steroid 2 77.4 66 vs. 64 -1.7 (-2.6, -0.8)*
AB vs. steroid 2 0 60 vs. 60 -2.5 (-3.5, -1.5)*
DASH score
PRP vs. steroid 2 91.6 96 vs. 94 -16.3 (-22.3, -10.4)*
AB vs. steroid 1 – 30 vs. 30 -25.5 (-33.8, -17.2)*
PRTEE score
PRP vs. AB 1 – 80 vs 70 -11.0 (-18.3, -3.7)*
PRP vs. steroid 1 – 20 vs. 20 -7.3 (-13.8, -0.9)*
AB vs. steroid 1 – 30 vs. 30 -5.3 (-9.1, -1.6)*
PPT
PRP vs. AB 2 68 58 vs. 58 2.5 (-1.5, 6.5)
AB vs. steroid 1 – 30 vs. 30 9.9 (5.6, 14.2)*
2
Adverse effects No. of studies I No. of subjects RR (95 % CI)
PRP vs. AB 1 – 14 vs.14 0.44 (0.17, 1.11)

PRP vs. steroid 2 0 71 vs. 69 1.00 (0.31, 3.24)
AB vs. steroid 2 0 60 vs. 60 1.78 (1.00, 3.17)*
Non-response rate No. of studies I2 No. of subjects RR (95 % CI)
PRP vs. steroid 1 – 51 vs. 49 1.23 (1.01, 1.49)*
PRP platelet-rich plasma, AB autologous blood, Steroid corticosteroid, VAS visual analog score, DASH Disabilities of the Arm Shoulder and
Hand, PRTEE Patient-Rated Tennis Elbow Evaluation, PPT pressure pain threshold, I2 degree of heterogeneity, UMD unstandardized mean
differences, CI confidence interval, RR relative risk
* Statistically significant difference (P \ 0.05)
last follow-up of AB injection was statistically significantly respectively. The regression analysis suggested that the
lower than steroid injection but not significantly different mean difference in PPT for AB injection was statistically
when compared with PRP. significantly higher than for PRP and steroid injection with
a value of 2.65 (95 % CI 0.30, 5.00) and 3.67 (95 % CI
Patient-Related Tennis Elbow Evaluation score 1.64, 5.69) when assessed within 2 months and at last
follow-up assessment, respectively; the mean difference
Data from three studies [5, 14, 26] were included in the between PRP and AB was statistically significant and
network meta-analysis of PRTEE score (Table 4). The increased to 7.50 (95 % CI 5.15, 9.85) and 9.87 (95 % CI
lowest mean PRTEE scores were for steroid injection and 7.84, 11.89) (Table 4; Fig. 2c).
PRP injection with a value of 30.82 (95 % CI 18.53, 43.11)
and 29.31 (95 % CI 17.03, 41.60) when assessed within Adverse effects (local pain and skin reaction) and non-
2 months and at most recent follow-up, respectively. There response rates
was no significant difference between the two active
treatments (Table 4). Data from five studies [6, 12, 14, 21, 33] were included in
the network meta-analysis. Compared to AB injection, PRP
Pressure pain threshold and steroid injection had lower risks of having complica-
tions, with borderline statistical significance of 99.6 %
Data from three studies [12, 23, 24] were included in the (RR = 0.004; 95 % CI 0.0002, 0.09) and 53 %
network meta-analysis of PPT (Table 4). The highest mean (RR = 0.53; 95 % CI 0.27, 1.05), respectively. PRP
PPT was for AB injection with a value of 21.23 (95 % CI injection had an approximately 10 % (RR = 0.90; 95 % CI
15.16, 27.31) and 27.53 (95 % CI 21.46, 33.61) when 0.36, 1.27), statistically not significant, lower risk than
assessed within 2 months and at last follow-up, steroid injection (Table 4; Fig. 2d).

Table 4 Comparisons of treatment effects: a network meta-analysis

Treatment Within 2 months At last follow-up
N Mean 95 % CI P value N Mean 95 % CI P value
VAS
PRP 153 3.64 2.84, 4.45 \0.001* 168 2.27 1.51, 3.02 \0.001*
AB 132 2.99 2.19, 3.80 \0.001* 132 2.90 2.09, 3.70 \0.001*
Steroid 79 4.18 3.04, 5.33, \0.001* 94 4.29 3.31, 5.27 \0.001*
N Mean difference 95 % CI P value N Mean difference 95 % CI P value
VAS
PRP vs. steroid – -0.54 -1.76, 0.68 0.386 – -2.02 -3.04, -1.01 \0.001*
AB vs. steroid – -1.19 -2.41, 0.03 0.056 – -1.39 -2.48, -0.30 0.012*
PRP vs. AB – 0.65 -0.21, 1.51 0.138 – -0.63 -1.47, 0.20 0.138

DASH
PRP 51 46.15 35.37, 56.93 \0.001* 66 17.38 8.42, 26.33 \0.001*
AB 30 7.49 -7.61, 22.59 \0.001* 30 7.49 -7.61, 22.59 \0.001*
Steroid 79 31.76 22.81, 40.71 \0.001* 94 35.95 28.10, 43.80 \0.001*

DASH
PRP vs. steroid – 14.39 1.77, 27.00 0.025* – -18.58 -29.08, -8.08 0.001*
AB vs. steroid – -24.27 -40.68, -7.86 0.004* – -24.27 -40.68, -7.86 0.004*
PRP vs. AB – 38.66 20.48, 56.83 \0.001* – 9.88 -7.32, 27.08 0.260

PRTEE
PRP 100 36.37 24.09, 48.66 \0.001* 100 29.31 17.03, 41.60 \0.001*
AB 100 34.12 21.84, 46.41 \0.001* 100 33.87 21.59, 46.16 \0.001*
Steroid 50 30.82 18.53, 43.11 \0.001* 50 32.53 20.24, 44.82 \0.001*

PRTEE
PRP vs. steroid – 5.55 -6.65, 17.76 0.373 – -3.22 -15.42, 8.99 0.605
AB vs. steroid – 3.30 -8.90, 15.51 0.596 – 1.34 -10.86, 13.55 0.829
PRP vs. AB – 2.25 -9.96, 14.46 0.718 – -4.56 -16.77, 7.65 0.464

PPT
PRP 58 18.58 12.66, 24.51 \0.001* 58 20.03 14.11, 25.96 \0.001*
AB 30 21.23 15.16, 27.31 \0.001* 30 27.53 21.46, 33.61 \0.001*
Steroid 88 17.57 11.70, 23.44 \0.001* 88 17.67 11.80, 23.54 \0.001*

Table 4 continued
PPT
PRP vs. steroid – 1.02 -0.48, 2.52 0.184 – 2.37 0.87, 3.87 0.02*
AB vs. steroid – 3.67 1.64, 5.69 \0.001* – 9.87 7.84, 11.89 \0.001*
PRP vs. AB – -2.65 -5.00, -0.30 \0.001 – -7.50 -9.85, -5.15 \0.001*
N IR 95 % CI P value N IR 95 % CI P value
Adverse effects
PRP – – – – 85 0.10 0.03, 0.34 \0.001*
AB – – – – 74 0.20 0.06, 0.65 0.008*
Steroid – – – – 129 0.11 0.03, 0.35 \0.001*

N RR 95 % CI P value N RR 95 % CI P value
Adverse effects
PRP vs. steroid – – – – – 0.90 0.36, 2.24 0.821
AB vs. steroid – – – – – 1.88 0.95, 3.72 0.068
PRP vs. AB – – – – – 0.004 0.0002, 0.09 0.001*
PRP platelet-rich plasma, AB autologous blood, Steroid corticosteroid, VAS visual analog score, DASH Disabilities of the Arm Shoulder and
Hand, PRTEE Patient-Rated Tennis Elbow Evaluation, PPT pressure pain threshold, CI confidence interval, IR incident rate, RR relative risk
Discussion was more efficacious than AB injection in terms of pain

VAS, and that PRP was more efficacious than steroid
The result of the present study was that PRP injection injections in terms of pain VAS. There is additional evi-
significantly improves pain and PRTEE score when com- dence with good methodological quality (RCT) that PRP
pared with AB injection and steroid injection. Compared to injection and AB injection displays an improvement in
AB injection, steroid injection had significantly improved disability scores (DASH, PRTEE) and pressure pain
disability score (DASH) and significantly improved pres- threshold (PPT) when compared with steroid injection.
sure pain threshold (PPT). The chances of adverse effects However, the highest risk of having adverse effects was
from PRP injection and steroid injection were not signifi- with AB injection when compared with PRP and steroid
cantly different but AB injection had a significantly higher injections.
chance of adverse effects when compared with steroid The direct meta-analysis suggests potential benefits of
injection. Multiple active treatment comparisons with time AB injection in reducing pain, improving disabilities scores
adjustment indicated that within 2 months only AB injec- and pressure pain threshold, but increasing the risk of
tion showed an improvement of borderline significance adverse effects when compared with steroids, whereas PRP
(0.0056) in pain VAS, but PRP and AB injection showed a injection can reduce pain, improve disabilities scores and
significant improvement in pain VAS when compared with pressure pain threshold, but has increased rates of non-
steroid injections. AB injection had significantly improved response after injection when compared with steroid
DASH scores and PPT when compared with PRP and injections. However, for other outcomes there was no
steroid injections, but AB injection had a statistically sig- significant difference. There are limitations of direct meta-
nificantly higher risk of adverse effects when compared analysis from the small number of studies that evaluated
with PRP and steroid injections at the last follow-up each particular pair of treatments, but a network meta-
assessment. For PRTEE score, there was no significant analysis circumvents this problem by creating indirect
difference between the two active treatments. comparisons between active treatments and difference in
The results of this study were consistent with previous time of assessment that can identify the most effective
meta-analyses by Ahmad et al. [1] which showed that PRP therapy and the time period that is the most beneficial. In

PRP PRP
n = 168 n = 66
-1.1* -1.7* -9.88 -16.3*
AB steroid AB steroid
n = 132 n = 94 n = 30 n = 94
-2.5* -25.5*
(a) Network meta-analysis of treatment (b) Network meta-analysis of treatment

effects on VAS effects on DASH
A line in the figure represents treatment comparisons, with arrows and tails A line in the figure represents treatment comparisons, with arrows and tails
referring to intervention and comparators, respectively. Bold and dashed referring to intervention and comparators, respectively. Bold and dashed lines
lines refer to direct and indirect comparisons, respectively. The number at the refer to direct and indirect comparisons, respectively. The number at the line
line indicates the mean VAS score of intervention vs comparator, in which < 0 indicates the mean DASH score of intervention vs comparator, in which < 0
indicates favors intervention vs the comparator. indicates favors intervention vs the comparator.
* p <0.05 with Bonferroni correction * p <0.05 with Bonferroni correction
PRP
PRP
n = 85
n = 58
2.5 2.37 0.44 1.00
AB steroid AB steroid
n = 30 n = 88 n = 74 n = 129
9.9* 1.78*
(c) Network meta-analysis of treatment (d) Network meta-analysis of treatment

effects on PPT effects on adverse effect
A line in the figure represents treatment comparisons, with arrows and tails A line in the figure represents treatment comparisons, with arrows and tails referring
referring to intervention and comparators, respectively. Bold and dashed lines to intervention and comparators, respectively. Bold and dashed lines refer to direct
refer to direct and indirect comparisons, respectively. The number at the line and indirect comparisons, respectively. The number at the line indicates the
indicates the PPT of intervention vs comparator, in which < 0 indicates favors complication rate of intervention vs comparator, in which < 1 indicates favors
intervention vs the comparator. intervention vs the comparator.
* p <0.05 with Bonferroni correction * p <0.05 with Bonferroni correction
Fig. 2 a Network meta-analysis of effects of treatment on VAS, b network meta-analysis of effects of treatment on DASH score, c network
meta-analysis of effects of treatment on PPT, d network meta-analysis of effects of treatment on adverse effects
Table 5 Summary of all treatment effects for lateral epicondylitis patients

Treatments Pain VAS DASH score PRTEE score PPT Adverse effects Non-response rate
PRP vs. AB (D* & N) (N*) (D* & N) (D & N*) (D & N*) –
PRP vs. steroid (D* & N*) (D* & N*) (D* & N) (N*) (D & N) (D*)
AB vs. steroid (D* & N*) (D* & N*) (D* & N) (D* & N*) (D* & N) –
D direct, N network
this case, AB injection was the best therapy at the assess- reduce Type I errors. A regression model was used, taking
ment times of within 2 months and over 2 months for into account study effects in order to assess treatment
improvement of DASH score and PPT, as during the sec- effects. The network meta-analysis ‘borrows’ treatment
ond time period it had a cumulative effect. AB injection information from other studies and increases the total
may be the worst therapy in terms of risks of adverse sample size. As a result, treatment effects that could not be
effects when compared with PRP and steroid injections. detected in direct meta-analysis could be identified. All
None of the RCTs compared combined treatments with AB possible treatment comparisons are mapped and displayed
injection or PRP injection and steroid injection. in Table 5. Although the pooled estimates were heteroge-
This study has several strengths. A network meta-anal- neous, the regression model with cluster effect takes vari-
ysis was applied to increase the power of the tests and ations at the study level into account. The limitations

recognized in this review are that some pooled results were 4. Assendelft WJ, Hay EM, Adshead R, Bouter LM (1996) Corti-
heterogeneous but the source of heterogeneity was not costeroid injections for lateral epicondylitis: a systematic over-
view. Br J Gen Pract 46(405):209–216
explored due to limitations in the reported data. 5. Creaney L, Wallace A, Curtis M, Connell D (2011) Growth
Based on the evidence presented, it can be concluded factor-based therapies provide additional benefit beyond phys-
that when comparing three active treatments, PRP injection ical therapy in resistant elbow tendinopathy: a prospective,
was the best treatment for reducing pain VAS after single-blind, randomised trial of autologous blood injections
versus platelet-rich plasma injections. Br J Sports Med
2 months whereas AB injection was the best treatment for 45(12):966–971
improving disabilities scores (DASH, PRTEE) and 6. Dojode CM (2012) A randomised control trial to evaluate the
increasing pressure threshold (PPT) both within and after efficacy of autologous blood injection versus local corticosteroid
2 months. However, AB injection had the highest risk of injection for treatment of lateral epicondylitis. Bone Joint Res
1(8):192–197
adverse effects (injection site pain and skin reaction). 7. Doran A, Gresham GA, Rushton N, Watson C (1990) Tennis
Further research should be done regarding cost-effective elbow. A clinicopathologic study of 22 cases followed for
analysis comparing PRP injection and AB injection or the 2 years. Acta Orthop Scand 61(6):535–538
combination of AB injection and multi-modality physical 8. Edwards SG, Calandruccio JH (2003) Autologous blood injec-
tions for refractory lateral epicondylitis. J Hand Surg Am
therapy, possibly improving outcomes for pain, disabilities 28(2):272–278
scores, and pressure pain threshold as well as lowering the 9. Egger M, Davey Smith G, Schneider M, Minder C (1997) Bias in
risk of adverse effects. meta-analysis detected by a simple, graphical test. BMJ
In conclusion, this network meta-analysis has provided 315(7109):629–634
10. Gosens T, Peerbooms JC, Van Laar W, Den Oudsten BL (2011)
additional information that PRP injection or AB injection Ongoing positive effect of platelet-rich plasma versus corticos-
can be selected for management of chronic lateral epi- teroid injection in lateral epicondylitis: a double-blind random-
condylitis. PRP can improve pain and lower the risk of ized controlled trial with 2 year follow-up. Am J Sports Med
adverse effects whereas AB injection can improve pain, 39(6):1200–1208
11. Kannus P, Jozsa L (1991) Histopathological changes preceding
disabilities scores and pressure pain threshold but has a spontaneous rupture of a tendon. A controlled study of 891
higher risk of adverse effects. patients. J Bone Joint Surg Am 73(10):1507–1525
12. Kazemi M, Azma K, Tavana B, Rezaiee Moghaddam F, Panahi A
Acknowledgments All authors declare no funding source or spon- (2010) Autologous blood versus corticosteroid local injection in
sor involvement in the study design, collection, analysis and inter- the short-term treatment of lateral elbow tendinopathy: a ran-
pretation of the data, in writing the manuscript, and in submission of domized clinical trial of efficacy. Am J Phys Med Rehabil
the manuscript for publication. 89(8):660–667
13. Krogh TP, Bartels EM, Ellingsen T, Stengaard-Pedersen K,
Compliance with ethical standards Buchbinder R, Fredberg U, Bliddal H, Christensen R (2013)
Comparative effectiveness of injection therapies in lateral epi-
Conflict of interest All authors declare that they have no conflicts condylitis: a systematic review and network meta-analysis of
of interests. randomized controlled trials. Am J Sports Med 41(6):1435–1446
14. Krogh TP, Fredberg U, Stengaard-Pedersen K, Christensen R,
Ethical standards Not applicable as no new patients were involved Jensen P, Ellingsen T (2013) Treatment of lateral epicondylitis
in this research. with platelet-rich plasma, glucocorticoid, or saline: a randomized,
double-blind, placebo-controlled trial. Am J Sports Med
41(3):625–635
15. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC,
Ioannidis JP, Clarke M, Devereaux PJ, Kleijnen J, Moher D
(2009) The PRISMA statement for reporting systematic reviews
and meta-analyses of studies that evaluate health care interven-
tions: explanation and elaboration. PLoS Med 6(7):e1000100
16. Lu G, Ades AE (2004) Combination of direct and indirect evi-
dence in mixed treatment comparisons. Stat Med
23(20):3105–3124
17. Mishra AK, Skrepnik NV, Edwards SG, Jones GL, Sampson S,
Vermillion DA, Ramsey ML, Karli DC, Rettig AC (2014) Effi-
References cacy of platelet-rich plasma for chronic tennis elbow: a double-
blind, prospective, multicenter, randomized controlled trial of
1. Ahmad Z, Brooks R, Kang SN, Weaver H, Nunney I, Tyther- 230 patients. Am J Sports Med 42(2):463–471
leigh-Strong G, Rushton N (2013) The effect of platelet-rich 18. Nirschl RP (1973) Tennis elbow. Orthop Clin North Am
plasma on clinical outcomes in lateral epicondylitis. Arthrosc J 4(3):787–800
Arthrosc Relat Surg 29(11):1851–1862 19. Omar AS, Ibrahim ME, Ahmed AS, Said M (2012) Local
2. Ahmad Z, Siddiqui N, Malik SS, Abdus-Samee M, Tytherleigh- injection of autologous platelet rich plasma and corticosteroid in
Strong G, Rushton N (2013) Lateral epicondylitis: a review of treatment of lateral epicondylitis and plantar fasciitis: randomized
pathology and management. Bone Joint J 95B(9):1158–1164 clinical trial. Egypt Rheumatol 34(2):43–49
3. Altman DG, Bland JM (2003) Interaction revisited: the difference 20. Palmer TM PJ, Sutton AJ, Moreno SG (2008) Contour-enhanced
between two estimates. BMJ 326(7382):219 funnel plots in meta-analysis. STATA J 8(2):242–254

21. Peerbooms JC, Sluimer J, Bruijn DJ, Gosens T (2010) Positive 28. Smidt N, van der Windt DA (2006) Tennis elbow in primary care.
effect of an autologous platelet concentrate in lateral epi- BMJ 333(7575):927–928
condylitis in a double-blind randomized controlled trial: platelet- 29. Smidt N, van der Windt DA, Assendelft WJ, Deville WL, Kor-
rich plasma versus corticosteroid injection with a 1 year follow- thals-de Bos IB, Bouter LM (2002) Corticosteroid injections,
up. Am J Sports Med 38(2):255–262 physiotherapy, or a wait-and-see policy for lateral epicondylitis: a
22. Peters JL, Sutton AJ, Jones DR, Abrams KR, Rushton L (2008) randomised controlled trial. Lancet 359(9307):657–662
Contour-enhanced meta-analysis funnel plots help distinguish 30. Song F, Altman DG, Glenny AM, Deeks JJ (2003) Validity of
publication bias from other causes of asymmetry. J Clin Epi- indirect comparison for estimating efficacy of competing inter-
demiol 61(10):991–996 ventions: empirical evidence from published meta-analyses. BMJ
23. Raeissadat SA, Rayegani SM, Hassanabadi H, Rahimi R, 326(7387):472
Sedighipour L, Rostami K (2014) Is platelet-rich plasma superior 31. Song F, Harvey I, Lilford R (2008) Adjusted indirect comparison
to whole blood in the management of chronic tennis elbow: may be less biased than direct comparison for evaluating new
1 year randomized clinical trial. BMC Sports Sci Med Rehabil pharmaceutical interventions. J Clin Epidemiol 61(5):455–463
6(1):12. doi:10.1186/2052-1847-6-12 32. Stata/SE 11.0 for windows (32bit). 13 July 2009 (updated 13 July
24. Raeissadat SA, Sedighipour L, Rayegani SM, Bahrami MH, 2009; cited). (13 July 2009)
Bayat M, Rahimi R (2014) Effect of platelet-rich plasma (PRP) 33. Thanasas C, Papadimitriou G, Charalambidis C, Paraskevopoulos
versus autologous whole blood on pain and function improve- I, Papanikolaou A (2011) Platelet-rich plasma versus autologous
ment in tennis elbow: a randomized clinical trial. Pain Res Treat. whole blood for the treatment of chronic lateral elbow epi-
doi: 10.1155/2014/191525 condylitis: a randomized controlled clinical trial. Am J Sports
25. Shiri R, Viikari-Juntura E, Varonen H, Heliovaara M (2006) Med 39(10):2130–2134
Prevalence and determinants of lateral and medial epicondylitis: a 34. Verhaar JA (1994) Tennis elbow. Anatomical, epidemiological
population study. Am J Epidemiol 164(11):1065–1074 and therapeutic aspects. Int Orthop 18(5):263–267
26. Singh A, Gangwar DS, Shekhar (2013) Autologous blood versus 35. Wong SM, Hui AC, Tong PY, Poon DW, Yu E, Wong LK (2005)
corticosteroid local injection for treatment of lateral epicondy- Treatment of lateral epicondylitis with botulinum toxin: a ran-
losis: a randomized clinical trial. Online J Health Allied Sci domized, double-blind, placebo-controlled trial. Ann Intern Med
12(2):11 143(11):793–797
27. Smidt N, Assendelft WJ, Arola H, Malmivaara A, Greens S,
Buchbinder R, van der Windt DA, Bouter LM (2003) Effec-
tiveness of physiotherapy for lateral epicondylitis: a systematic
review. Ann Med 35(1):51–62

28
Is there a link between the neutrophil-to-lymphocyte ratio
and venous thromboembolic events after knee arthroplasty?
A pilot study
Tyler Barker1 • Victoria E. Rogers1 • Vanessa T. Henriksen1 • Kimberly B. Brown1 •
Roy H. Trawick2 • Nathan G. Momberger2 • G. Lynn Rasmussen2
Abstract Results On postoperative day 1, the NLR increase was

Background This study aimed to identify (1) if the exacerbated (p = 0.02) following TKA compared to UKA
postoperative increase in the neutrophil-to-lymphocyte and predicted (p = 0.02) the occurrence of VTE in TKA
ratio (NLR) is different between contrasting knee arthro- patients prior to hospital discharge.
plasty procedures, and (2) if the NLR predicts venous Conclusion We conclude that the NLR increase is greater
thromboembolism (VTE) after total knee arthroplasty following TKA compared to UKA and could serve as a
(TKA). matrix to predict or identify a patient susceptible of sus-
Materials and methods To address the first objective, we taining VTE after TKA.
retrospectively studied patients who underwent primary Level of evidence 3.
unilateral TKA (n = 111) or unicompartmental knee
arthroplasty (UKA; n = 74) between 2009 and 2012. Keywords Total knee arthroplasty Unicompartmental
Patients who required a blood transfusion, underwent knee arthroplasty Neutrophil Lymphocyte Venous
autologous blood salvage, experienced any postoperative thromboembolic event
complication (such as VTE), or were re-admitted [90 days
were excluded from analysis. For the second objective, we
retrospectively identified patients (cases, n = 10) who Introduction
underwent primary unilateral TKA between 2010 and 2012
and developed postoperative VTE (deep venous thrombo- Osteoarthritis (OA) is the most common degenerative joint
sis, pulmonary embolism, or both) during inpatient care disease [1] and the knee is one of the most commonly affected
(postoperative day 1 or day 2). Cases were matched to sites. Total knee arthroplasty (TKA) is a safe and effective
surgeon, gender, body mass index, age, and date of surgery surgery for treating knee OA, and it is estimated that the
controls (n = 20) who underwent primary unilateral TKA number of knee arthroplasty procedures performed annually
without developing VTE before patient discharge. The in the United States will reach 3.5 million by the year 2030
NLR was calculated from the neutrophil and lymphocyte [2]. Although infrequent (*1–3 %), venous thromboem-
counts extracted from pre- and postoperative (day 1 and bolism (VTE) (including deep venous thromboembolism
day 2) blood chemistry records. [DVT], pulmonary embolism [PE], or both) is a life-threat-
ening complication that can occur after TKA [3–6]. Over the
past two decades, unicompartmental knee arthroplasty
(UKA) has emerged as a less invasive alternative to TKA
& Tyler Barker when knee OA is limited to a single compartment. In com-
[email protected]
parison to TKA, the prevalence of postoperative VTE
1
The Orthopedic Specialty Hospital, 5848 S. Fashion Blvd., (*0.1 %) is apparently lower after UKA [4, 7, 8].
Murray, UT 84107, USA In other pathological conditions, the neutrophil-to-lym-
2
The Orthopedic Specialty Clinic, 5848 S. Fashion Blvd, phocyte ratio (NLR) is developing as a clinical tool that
Murray, UT 84107, USA accurately predicts VTE [9–14]. The NLR is a ratio

Is there a link between the neutrophil-to-lymphocyte ratio and venous thromboembolic events after knee... 199
between the absolute neutrophil and lymphocyte counts, underwent UKA displayed isolated anteromedial compart-
and therefore, is a marker of systemic inflammation rep- ment degeneration, a well-preserved lateral compartment, no
resentative of innate and adaptive immunity. Although greater than mild patellofemoral degenerative changes, and
systemic inflammation modulates the coagulation process an intact anterior cruciate ligament. TKA was performed in
[15], it is surprisingly unknown if the NLR is different patients who did not satisfy the criteria for UKA as deter-
between knee arthroplasty procedures with contrasting risk mined by one orthopedic surgeon. Tolerable pain levels,
levels of sustaining VTE, and more importantly, if the NLR normal venous blood O2 and blood count, and no signs of PE
predicts VTE after TKA. Therefore, the purpose of this were required before inpatient discharge after surgery.
study was two-fold. The first objective was to identify if the In Study 2, we retrospectively identified patients (cases)
NLR is different between arthroplasty procedures that who underwent elective primary unilateral TKA at a single
commonly treat knee OA but possess contrasting risk levels facility (The Orthopedic Specialty Hospital, Murray, UT)
of sustaining VTE. The second objective was to identify if and developed postoperative VTE (i.e., DVT, PE, or both)
the NLR predicts VTE after TKA. We hypothesized that during inpatient care. Patients who underwent elective
the NLR increase is greater following TKA compared to primary unilateral TKA without developing VTE during
that after UKA and that the NLR predicts VTE after TKA. inpatient care served as controls. Cases were identified
We performed two retrospective studies to test our using the Agency for Healthcare Research and Quality
hypothesis. In the first study, we compared the NLR Patient Safety Indicator 12 criteria [16], and the Interna-
between two knee arthroplasty modalities—total and uni- tional Classification of Diseases 9th revision. The presence
compartmental. The second study consisted of a retro- or absence of VTE was confirmed in each patient’s dis-
spective case–control design and investigated the charge notes. In all cases, DVTs were identified using
predictability of the NLR on VTE after TKA. ultrasound and PEs were identified by chest computed
tomography angiogram. Ultrasound and chest computed
tomography angiograms were not performed in asymp-
Materials and methods tomatic patients who were identified as controls.
To minimize the variability in surgical procedures,
To test our hypothesis, we performed two studies that components, and perioperative treatments, data extraction
were approved by an Institutional Review Board (Central was limited from January 1, 2010 to the end of December
Region, Intermountain Healthcare, Salt Lake City, UT). 2012. During this period, eleven patients (from 1,339 pri-
The studies were performed in accordance with the ethical mary unilateral TKAs; 0.82 %) were identified as experi-
standards as laid down in the 1964 Declaration of Helsinki encing postoperative (day 1 or day 2) VTE. Cases were
and its later amendments or comparable ethical standards. matched to surgeon, gender, body mass index (BMI), and
In Study 1, we retrospectively reviewed the medical age controls (two controls for every case). Each case was
charts of 472 patients who underwent elective primary matched to asymptomatic patients who underwent surgery
unilateral TKA (n = 302) or UKA (n = 170) from January by the same surgeon and were of the same gender, age, and
2009 to December 2012 by one orthopedic surgeon at a BMI. Investigators responsible for matching were blinded to
single institution (The Orthopedic Specialty Hospital, blood chemistry results. We were unable to identify controls
Murray, UT). Patients were identified using the International for one case with an exceptionally high BMI (46.7 kg/m2),
Classification of Diseases 9th revision, and all patients were and therefore, this case was excluded from data analysis.
diagnosed with knee OA from X-ray images and voluntarily The final analysis consisted of twenty controls and ten cases
elected for knee arthroplasty. Demographic and surgery (DVT, n = 5; PE, n = 4; DVT and PE, n = 1).
variables were recorded for each patient. Patients were Cases were further identified as having a history of
excluded from analysis if they underwent bilateral proce- diabetes (n = 2), heart problems (e.g., previous heart
dures, were re-admitted within 90 days of surgery, had pre- attack, irregular heartbeat, angina, heart failure; n = 1),
or postoperative infection, had DVT or PE, or received a previous blood clot, transfusion or bleeding tendency
blood transfusion, manipulation, or autologous blood sal- (n = 1), or cancer (n = 2). Controls had a history of dia-
vage. Patients were also excluded from analysis if they were betes (n = 4), heart problems (n = 10), previous blood
lacking an American Society of Anesthesiologists (ASA) clot, transfusion or bleeding tendency (n = 5), stroke
score, or the pain and Knee Outcome Survey scores were not (n = 1), multiple sclerosis (n = 1), or cancer (n = 2).
documented at physical therapy. Controls contained two former and one current smoker,
Following screening, 111 TKA (Zimmer Total Knee; while none of the cases reported a previous or current
Zimmer, Inc., Warsaw, IN, USA) and 74 UKA (OxfordÒ smoking habit.
Partial Knee; Biomet Orthopedics, Warsaw, IN, USA) The ASA physical status classification score was
patients were included in the final analysis. Patients who recorded for each subject. Five subjects (two cases and

three controls) received 4–8 mg of dexamethasone before assessed with separate Pearson chi-squared tests. All sta-
induction of anesthesia. The exclusion of these patients tistical analyses were performed with SYSTAT (version
from the statistical analysis did not change the significance 13.1, Chicago, IL USA). Data are presented as mean (SD)
of the NLR between groups, and therefore, were included unless noted otherwise.
in the final analysis.
All patients were treated with compression stockings
from the time of surgery as well as a prophylactic antico- Results
agulant. Ten cases and sixteen controls were titrated with
5 mg of CoumadinÒ (Bristol-Myers Squibb, New York, Study 1. Blood chemistries and subject characteristics
NY, USA) in order to achieve an INR target of 1.5–2.0. (with the exception of height, p = 0.002) were not sig-
Two controls were treated with XareltoÒ (Leverkusen, nificantly different (p range 0.07–0.90) between TKA and
Germany) (monitored by anticoagulation monitoring UKA patients prior to surgery (Table 1). As expected,
clinic) and two controls with ArixtraÒ (GlaxoSmithKline, tourniquet duration during surgery was significantly
Research Triangle Park, NC, USA) (2.5 mg subcuta- (p \ 0.001) increased in TKA (69.0 [9.2] min) compared
neously). Five cases were administered LovenoxÒ (Sanofi- to UKA group [57.8 [5.2] min). Following surgery, neu-
Aventis, Bridgewater, NJ, USA) when their INR values trophil counts were significantly increased in both groups
were \2.0. (p \ 0.001, Fig. 1a). However, neutrophil counts were not
For the TKA groups in Study 1 and Study 2, preopera- significantly (p = 1.00) different between arthroplasty
tive (morning of surgery and following a 10-h fast) and procedures. Conversely, the lymphocyte decrease
postoperative (day 1 and day 2; the morning after surgery (p = 0.001, Fig. 1b) and NLR increase (p = 0.02, Fig. 1c)
and following a 10-h fast during inpatient care) blood were more prominent following TKA compared to after
chemistries were performed at a hospital laboratory (blin- UKA.
ded) and results were extracted from an electronic database Study 2. Subject characteristics, length of stay, and
warehouse. For the UKA group in Study 1, blood che- blood chemistries prior to surgery were not significantly
mistries were performed preoperative (morning of surgery (p range = 0.25–0.78) different between groups (Table 2).
and following a 10-h fast) and on day 1 (the morning after On postoperative day 1, there was a significant increase in
surgery and following a 10-h fast during inpatient care). neutrophil (p = 0.008, Fig. 2a) and a significant decrease
Blood chemistries for the UKA group in Study 1 were in lymphocyte (p = 0.008, Fig. 2c) counts in the cases and
limited to postoperative day 1 since a postoperative blood controls. However, neutrophil (p = 0.22) and lymphocyte
draw on day 2 is not a standard of care procedure for the (p = 0.32) counts were not significantly different between
participating surgeon. The NLR was calculated from the groups. In contrast, the increase in the NLR was
absolute neutrophil and lymphocyte counts.
Data were checked for normality prior to all statistical
Table 1 Study 1 subject characteristics and results
analyses with the Shapiro–Wilk test. To test the hypothesis
for Study 1, and because the data were not normally dis- TKA UKA P value
tributed, we performed a Friedman two-way analysis of No. (female: male) 111 (80:31) 74 (30:44)
variance test followed by multiple pairwise comparisons Age (years) 62 (9) 62 (9) 0.90
when appropriate to determine if the NLR was different Height (cm) 169 (10) 173 (10) 0.002
between TKA and UKA patients prior to and following Body mass (kg) 98.0 (24.7) 95.9 (18.6) 0.80
surgery. Likewise, separate Friedman two-way analysis of BMI (kg/m2) 34.4 (7.8) 32.1 (5.7) 0.07
variance tests followed by multiple pairwise comparisons
ASA score 2.41 (0.58) 2.35 (0.67) 0.34
when appropriate were performed to determine if the
1 (n) 5 7
neutrophil and lymphocyte counts were different between
2 (n) 56 35
TKA and UKA patients prior to and following surgery. To
3 (n) 50 31
test the hypothesis for Study 2, we performed univariate
4 (n) 0 1
logistic regression to determine if the NLR on day 1 pre-
Length of stay (days) 2.88 (0.44) 1.22 (0.48) \0.001
dicted VTE on day 1 and day 2 post TKA. Separate
0 (n) 0 1
repeated measures analysis of variance followed by Bon-
1 (n) 2 57
ferroni corrections on multiple pairwise comparisons was
2 (n) 12 15
performed when appropriate on other repeated measure
3 (n) 94 1
variables for Study 2. The statistical significance of subject
4 (n) 3 0
characteristics was assessed with separate t tests for Study
1 and Study 2. The length of stay and ASA score were Data are presented as mean (SD) unless noted otherwise

A 12 TKA
B 3.0 *$ *
C 14 *$ *
Lymphocytes (K/µL)
Neutrophils (K/µL)
UKA 12
10 2.5
10
8 2.0
NLR
8
6 1.5
6
4 1.0
4
2 0.5 2
0 0.0 0
Pre 1-d Pre 1-d Pre 1-d
Fig. 1 Blood cell counts and NLR prior to (Pre) and one day (*p \ 0.01 vs Pre). The lymphocyte decrease was significantly more
following (1-d) TKA and UKA. a Neutrophil counts were signifi- pronounced ($p = 0.001) following TKA compared to that after
cantly (bar, p \ 0.001 vs Pre) increased following TKA and UKA. UKA. c The NLRs were significantly increased following TKA
The neutrophil increase was not significantly (p = 1.00) different (*p \ 0.001 vs Pre) and UKA (*p \ 0.001 vs Pre). The NLR increase
between TKA and UKA on 1-d. b Lymphocyte counts were was significantly ($p = 0.02) more pronounced after TKA compared
significantly decreased following TKA (*p \ 0.01 vs Pre) and UKA to UKA. Data are presented as mean (SD)
Table 2 Study 2 subject characteristics and results significantly (p = 0.002) more pronounced in cases com-
Cases Controls P value
pared to controls on day 1 (Fig. 2c).
Consistent with our hypothesis, the NLR on day 1 pre-
No. (female male) 10 (7:3) 20 (14:6) dicted the occurrence of VTE after TKA and during
Age (years) 65 (5) 64 (7) 0.78 inpatient care (p = 0.02; odds ratio (OR) 1.38; 95 %
Height (cm) 169 (9) 168 (12) 0.74 confidence interval (CI) 1.05–1.80). We performed a sec-
Body mass (kg) 98.5 (22.2) 91.4 (22.8) 0.43 ondary analysis excluding day 1 event cases (n = 4) and
BMI (kg/m2) 34.4 (7.7) 32.4 (7.3) 0.50 their controls (n = 8) to confirm that a preceding NLR
ASA score 0.71 predicts a subsequent VTE after knee arthroplasty. Sup-
1 (n) 0 1 porting the earlier finding, the NLR on day 1 predicted
2 (n) 6 9 VTE on day 2 (p = 0.04; OR 1.44; 95 % CI 1.01–2.05).
3 (n) 4 10
Length of stay (days) 3.30 (1.06) 2.90 (0.31) 0.25
2 (n) 1 2 Discussion
3 (n) 7 18
4 (n) 1 0 Knee arthroplasty is an independent risk factor for devel-
6 (n) 1 0 oping VTE [17–19]. Revealing a matrix from a routine
VTE, days from surgery (n) \0.001 blood chemistry performed prior to and following knee
1 day 4 0 arthroplasty could impact the physician’s decision process,
2 day 6 0 improve patient care, and further minimize the risk of
postoperative complications related to VTE without
Data are presented as mean (SD) unless noted otherwise
A 12 B 3.5 C 16
Cases *$ *
Lymphocytes (K/µL)
14
Neutrophils (K/µL)
Controls 3.0
10
2.5 12
8
NLR
10
2.0
6 8
1.5
4 6
1.0 4
2 0.5 2
0 0.0 0
Pre 1-d 2-d Pre 1-d 2-d Pre 1-d 2-d
Fig. 2 Blood cell counts and the NLR in the cases and controls prior Lymphocyte counts were not significantly different between groups
to (Pre) and one and two days following (1-d, 2-d) TKA. a Neutrophil (p = 0.32). c The NLRs were significantly increased in both groups
counts were significantly increased on 1-d in both groups (bar, on 1-d (cases and controls, *p \ 0.001 vs Pre) and 2-d (bar, p B 0.01
p = 0.008 vs Pre). Neutrophil counts were not significantly different vs Pre). The NLR on 1-d was significantly increased ($p = 0.002) in
between the cases and control (p = 0.22). b Lymphocyte counts were the cases compared to the controls. Data are presented as mean (SD)
significantly decreased on 1-d in both groups (bar, p = 0.008 vs Pre).

additional costs. In this study, we provide the first evidence controls developed VTE following discharge. Second, Study
that a more invasive knee arthroplasty procedure with an 1 consisted of a small sample size; however, this was
inherently greater risk of sustaining VTE had a more pro- expected when considering the infrequency of a VTE. Third,
nounced increase in NLR after surgery. Additionally, the in Study 2, 39 % of the subjects underwent UKA and the
NLR after TKA predicted the subsequent occurrence of delineation of performing UKA was determined by a single
VTE prior to hospital discharge. This latter finding sug- orthopedic surgeon with a predefined standard of care
gests that the NLR could serve as an inpatient monitoring screening protocol. This approach strengthens the interpre-
tool to assess the risk of VTE. However, considering its tation between patients with contrasting knee abnormalities,
uniqueness, future research is needed to confirm or refute but it constrains the broad application of these findings to
the present results and to establish the NLR value that other studies conducted in patients who underwent TKA
demarcates an increased risk for a VTE. who could have been candidates for UKA. Finally, although
Previous studies found that an NLR [4.5 predicted the NLR difference between UKA and TKA groups is likely
coronary heart disease mortality in patients without coro- the result of the surgical procedures, it would be remiss to
nary heart disease [20], while an NLR[4.7 predicted death ignore the underlying pathological differences between
or myocardial infarction in patients initially assessed for groups prior to arthroplasty. Future research is also needed
coronary artery disease [14]. In patients screened for or to identify if the type of prosthesis, as well as different types
diagnosed with PE, the optimum NLR cut-off value for of anticoagulant medications, impact the assessment of the
predicting 30-day mortality was 9.2 [9]. Although the NLR on VTE following knee arthroplasty.
optimum NLR cut-off value provided by Kayrak and col- Despite low rates (*1–3 %), the absolute number of
leagues [9] is double the value related to coronary heart patients who will sustain VTEs is anticipated to increase
disease mortality or myocardial infarction [14, 20], it is with the expected continuous growth in the number of
similar to the value reported here. However, the NLR that TKA procedures performed annually. Thus, developing a
predicted VTE after TKA is an acutely elevated value while matrix that improves prognostic potential could alter
the values previously reported in other patient populations patient treatment strategies and further minimize morbidity
(i.e., non-orthopedic) are presumably chronic. Clearly, and mortality rates following knee arthroplasty. In this
additional research delineating the condition-specific NLR study, the NLR increase was exacerbated following a more
that predicts clinical outcomes and the prognostic role of invasive knee arthroplasty procedure (i.e., TKA) charac-
acute NLR fluctuations on VTE is warranted. terized by an increased risk of VTE. Moreover, the NLR
Another unique finding of the present investigation was following TKA predicted VTE prior to patient discharge.
the exaggerated increase in the NLR following TKA Based on these findings, we conclude that NLR increase
compared to UKA. The exaggerated increase in the NLR could serve as a matrix for monitoring the susceptibility of
was mediated by the dramatic and more severe presence of developing VTEs after knee arthroplasty and prior to
lymphopenia following TKA. An increase in the NLR and hospital discharge. Confirmation of the preliminary data
lymphopenia are predictive of cardiovascular events, such reported here is essential and could be paramount in
as acute myocardial infarction (AMI), VTE, and overall guiding the physician’s treatment methods during inpatient
mortality in cardiac patients [10–14]. Despite the use of care following knee arthroplasty.
thromboprophylaxis, the risk of AMI (*0.4 %), VTE
(*1.1 %), and major bleeding (*0.5 %) after TKA sub- Compliance with ethical standards
stantially increase within 90 days of surgery [3, 5, 21]. Conflict of interest The authors declare that they have no conflict
Extrapolating from TKA data, evidence suggests that UKA of interest.
results in fewer complications in terms of DVT [4, 7, 8].
Although the percentage of TKA and UKA patients Ethical standards All procedures performed in studies involving
human participants were in accordance with the ethical standards of
affected by such devastating cadiovascular ailments is low,
the institutional research committee and with the 1964 Helsinki
these are impactful consequences and further research declaration and its later amendments or comparable ethical standards.
investigating the prognostic role of the NLR on such life- This study was approved by the Institutional Review Board and
changing events after knee arthroplasty is warranted. consent was waived for this retrospective chart review study.
Although the present study provides novel results, there
are a few limitations worthy of discussion. First, in Study 2,
computed tomography and ultrasound were not performed
in the asymptomatic control group. Therefore, it is unknown
if the control group had silent DVT or PE. In line with this
limitation, data extraction and analysis were limited to
perioperative outcomes and it is unknown if any of the

References neutrophil count and neutrophil/lymphocyte ratio in predicting

in-hospital mortality and complications after STEMI. J Cardio-
1. Lawrence RC, Felson DT, Helmick CG, Arnold LM, Choi H, vasc Thorac Res 6:35–41
Deyo RA et al (2008) Estimates of the prevalence of arthritis and 12. Arbel Y, Finkelstein A, Halkin A, Birati EY, Revivo M, Zuzut M
et al (2012) Neutrophil/lymphocyte ratio is related to the severity
other rheumatic conditions in the United States. Part II. Arthritis
Rheum 58:26–35 of coronary artery disease and clinical outcome in patients
2. Kurtz S, Ong K, Lau E, Mowat F, Halpern M (2007) Projections undergoing angiography. Atherosclerosis 225:456–460
of primary and revision hip and knee arthroplasty in the United 13. Gibson PH, Croal BL, Cuthbertson BH, Small GR, Ifezulike AI,
Gibson G et al (2007) Preoperative neutrophil-lymphocyte ratio
States from 2005 to 2030. J Bone Joint Surg Am 89:780–785
3. Lalmohamed A, Vestergaard P, Klop C, Grove EL, de BA, and outcome from coronary artery bypass grafting. Am Heart J
Leufkens HG et al (2012) Timing of acute myocardial infarction 154:995–1002
in patients undergoing total hip or knee replacement: a nation- 14. Horne BD, Anderson JL, John JM, Weaver A, Bair TL, Jensen
wide cohort study. Arch Intern Med 172:1229–1235 KR et al (2005) Which white blood cell subtypes predict
4. Mantilla CB, Horlocker TT, Schroeder DR, Berry DJ, Brown DL increased cardiovascular risk? J Am Coll Cardiol 45:1638–1643
(2002) Frequency of myocardial infarction, pulmonary embolism, 15. Levi M, Keller TT, van GE, ten CH (2003) Infection and
deep venous thrombosis, and death following primary hip or knee inflammation and the coagulation system. Cardiovasc Res
arthroplasty. Anesthesiology 96:1140–1146 60:26–39
5. Guijarro R, Montes J, San RC, Arcelus JI, Barillari G, Granero X 16. Agency for Healthcare Research and Quality. Patient safety
et al (2011) Venous thromboembolism and bleeding after total indicators: technical specifications (version 4.5a). (2014) Ref
Type: Online Source
knee and hip arthroplasty. Findings from the Spanish national
discharge database. Thromb Haemost 105:610–615 17. Mantilla CB, Horlocker TT, Schroeder DR, Berry DJ, Brown DL
6. Pedersen AB, Mehnert F, Sorensen HT, Emmeluth C, Overgaard (2003) Risk factors for clinically relevant pulmonary embolism
S, Johnsen SP (2014) The risk of venous thromboembolism, and deep venous thrombosis in patients undergoing primary hip
myocardial infarction, stroke, major bleeding and death in or knee arthroplasty. Anesthesiology 99:552–560
patients undergoing total hip and knee replacement: a 15-year 18. Pedersen AB, Mehnert F, Johnsen SP, Husted S, Sorensen HT
retrospective cohort study of routine clinical practice. Bone Joint (2011) Venous thromboembolism in patients having knee
J 96-B:479–485 replacement and receiving thromboprophylaxis: a Danish popula-
7. Lombardi AV Jr, Berend KR, Tucker TL (2007) The incidence tion-based follow-up study. J Bone Joint Surg Am 93:1281–1287
and prevention of symptomatic thromboembolic disease follow- 19. Brookenthal KR, Freedman KB, Lotke PA, Fitzgerald RH,
ing unicompartmental knee arthroplasty. Orthopedics 30:46–48 Lonner JH (2001) A meta-analysis of thromboembolic prophy-
laxis in total knee arthroplasty. J Arthroplasty 16:293–300
8. Berend KR, Morris MJ, Lombardi AV (2010) Unicompartmental
knee arthroplasty: incidence of transfusion and symptomatic 20. Shah N, Parikh V, Patel N, Patel N, Badheka A, Deshmukh A
thromboembolic disease. Orthopedics 33:8–10 et al (2014) Neutrophil lymphocyte ratio significantly improves
9. Kayrak M, Erdogan HI, Solak Y, Akilli H, Gul EE, Yildirim O the Framingham risk score in prediction of coronary heart disease
mortality: insights from the national health and nutrition exami-
et al (2014) Prognostic value of neutrophil to lymphocyte ratio in
patients with acute pulmonary embolism: a restrospective study. nation survey-III. Int J Cardiol 171:390–397
Heart Lung Circ 23:56–62 21. Pedersen AB, Mehnert F, Sorensen HT, Emmeluth C, Overgaard
10. Erturk M, Cakmak HA, Surgit O, Celik O, Aksu HU, Akgul O S, Johnsen SP (2014) The risk of venous thromboembolism,
et al. (2014) The predictive value of elevated neutrophil to myocardial infarction, stroke, major bleeding and death in
lymphocyte ratio for long-term cardiovascular mortality in patients undergoing total hip and knee replacement: a 15-year
peripheral arterial occlusive disease. J Cardiol retrospective cohort study of routine clinical practice. Bone Joint
11. Ghaffari S, Nadiri M, Pourafkari L, Sepehrvand N, Movasagpoor J 96-B:479–485
A, Rahmatvand N et al (2014) The predictive value of total

29
Calcific tendinitis of the rotator cuff: state of the art in diagnosis
and treatment
Giovanni Merolla1,2 • Sanjay Singh1 • Paolo Paladini1 • Giuseppe Porcellini1
Abstract Calcific tendinitis is a painful shoulder disorder was probably first coined by Plenk [3] in 1952. The disease
characterised by either single or multiple deposits in the subsides spontaneously in the majority of cases and can be
rotator cuff tendon. Although the disease subsides sponta- managed with conservative therapy, but some patients
neously in most cases, a subpopulation of patients continue continue to have a painful shoulder for an extended period
to complain of pain and shoulder dysfunction and the of time with the deposits not showing any signs of reso-
deposits do not show any signs of resolution. Although lution. New conservative treatment modalities such as
several treatment options have been proposed, clinical ultrasound-guided needling (UGN) and extracorporeal
results are controversial and often the indication for a given shock wave therapy (ESWT) have emerged in recent years
therapy remains a matter of clinician choice. Herein, we as additional management options. Incidence varies from
report on the current state of the art in the pathogenesis, 2.7 to 20 %, as reported by various authors [1, 4, 5]. In
diagnosis and treatment of calcific tendinitis of the rotator about 10–20 % of patients, the deposits are bilateral [1, 5,
cuff. 6]. Most studies found higher incidence in women com-
pared with men [1, 6]. Regarding age distribution, the
Keywords Calcific tendinitis Shoulder Rotator cuff average age of presentation in most studies was between 30
Diagnosis Treatment options and 50 years [5, 6]. No deposits were found in the elderly
[5, 7, 8]. Most investigators found the deposits to be more
commonly located in the supraspinatus [1, 3, 4, 6],
Introduction although often the deposits were also located in the
infraspinatus [1, 4, 6] and rarely in the subscapularis and
Calcific tendinitis (CT) is a painful shoulder disorder teres minor [1, 4]. Most patients were sedentary workers or
characterised by either single or multiple deposits in the housewives [6]. The right shoulder was most commonly
rotator cuff (RC) tendon or subacromial bursa [1]. It was affected [6]. The natural history of the disease can be
Codman who, in his book [2], described the deposits as divided into three distinct clinical stages: acute, subacute
being in the RC tendon. The term ‘‘calcifying tendinitis’’ and chronic. The main clinical manifestation is pain, which
may or may not be associated with acute or gradual
restriction of movements [4, 9]. Acute pain is often asso-
& Giovanni Merolla
ciated with the onset of the disease; however, the deposits
[email protected]; may be asymptomatic in 20 % of cases [6]. Muscle spasm,
[email protected] and inflammation of subacromial bursa (bursitis) and the
1
long head of the biceps are determining symptomatic fac-
Unit of Shoulder and Elbow Surgery, D. Cervesi Hospital,
tors. The pain is, in most cases, associated with the acute
Cattolica (RN) - AUSL della Romagna Ambito Territoriale di
Rimini, Italy phase of the disease, but episodes of acute pain are also
2 often related to flare-ups of chronic tendinopathy or onset
Biomechanics Laboratory ‘‘Marco Simoncelli’’, D. Cervesi
Hospital, Cattolica (RN) - AUSL della Romagna Ambito of rare complications not related to the evolution of the
Territoriale di Rimini, Italy disease, such as adhesive capsulitis (AC), rotator cuff tear,

Calcific tendinitis of the rotator cuff: state of the art in diagnosis and treatment 205
pathology of the long head of the biceps or osteolysis of the abnormal pre-existing calcifications can produce or
greater tuberosity (TO) [10, 11]. enhance a complete RC tear, requiring a surgical approach.
Aetiopathogenesis and histopathology Imaging
The aetiopathogenesis of CT remains elusive. Codman [2] Conventional radiology

hypothesised that overuse degeneration of rotator cuff
leads to calcific deposits in the tendon, and this was also Standard radiographs in anterior–posterior (AP), outlet and
supported by Bishop [12], whereas Sandstrom [13] pro- axillary views are used for diagnosis and follow-up of CT,
posed that the degeneration in the tendon follows local because they allow localisation and assessment of the
ischaemia which led to calcium deposition. More recently, texture and morphology of the deposits [22, 23] (Fig. 1).
Mohr and Bilger [14] considered that the process begins Many authors have tried to classify the deposits in terms of
with necrosis of tenocytes due to apoptosis along with size [1] or morphology [6, 24–26] (Table 1). However, the
intracellular accumulation of calcium, but a more detailed fact that there are numerous classifications indicates that no
description was given by Uhthoff et al. [15], who proposed classification perfectly correlates with the radiological
that the disease goes through three stages: precalcific, picture and symptomatology of the patient, and there is
calcific and postcalcific. In the precalcific stage, there is also significant inter-observer variability [27].
fibrocartilaginous metaplasia in the tendon; this stage is The location of the deposits in the tendons also varies [1,
rarely symptomatic. This is followed by the calcific stage, 22] (Table 2).
which is further divided into formative, resting and reab-
sorption phases. It is in the reabsorptive phase that patients Ultrasound
are mostly symptomatic. The postcalcific phase is the
healing phase, in which there is reabsorption of the deposit. Ultrasound (US) examination is a fundamental tool in
Rui et al. [16] postulated that incorrect differentiation of diagnosis and treatment of CT [28, 29]. US has changed
stem cells, tendon-derived stem cells (TDSCs), into from having a purely diagnostic role to become an impor-
osteoblasts or chondrocytes could be the basis of the cal- tant therapeutic tool, especially for carrying out bursal
cification. Disorders of the thyroid (thyroxine) or oestrogen lavage and tendon needling (Fig. 1b, c). Use of high-reso-
metabolism may be related to the onset of the disease. lution US shows the presence of deposits and also defines
Harvie et al. [17] reported endocrine involvement in their locations in the tendon, plus their size and texture. This
64.7 % of cases in their series, whereas Mavrikakis et al. technique shows RC tears in detail, and also enables staging
[18] reported CT incidence in 31.8 % of their diabetic of the deposits by correlation of shadow cones [30, 31]. In
patients, compared with 10.3 % of the control group. the resting phase, the deposits appear hyperechoic and arc
Sengar et al. [19] found an increased frequency of human shaped, whereas they appear non-arc shaped (fragmented/
leucocyte antigen serotype class A1 in patients with CT. punctate, cystic, nodular) in the resolving phase [30]. These
Mutation in the human homologue of the murine progres- appearances can also be correlated with the symptomatic
sive ankylosis gene (ANKH) has been reported in patients and asymptomatic phases of the disease [32]. Farin et al.
with hereditary chondrocalcinosis, leading to alteration of [33] divided the deposits into three types: (1) hyperechoic
the picture of extracellular inorganic pyrophosphate [20]. focus with a well-defined shadow, (2) hyperechoic focus
Oliva et al. [21] found that significantly increased expres- with a faint shadow and (3) hyperechoic focus with no
sion of tissue transglutaminase (tTG)2 and its substrate shadow. Doppler examination during the nodular or cystic
osteopontin was detected in calcific areas compared with phase shows increased vascularity around the deposits [34],
levels observed in normal tissue from the same subject with which correlates well with the histopathological findings of
calcific tendinopathy. They concluded that a variation in Uhthoff et al. [35], who showed how, during the reab-
the expression of these genes could be characteristic of this sorption phase, the deposits are surrounded by phagocytes
form of tendinopathy. The correlation between increased and there was concomitant proliferation of vascular chan-
incidence of endocrine disorders and risk of developing CT nels around the deposits.
remains unclear; similarly, the associations with genetic
mutations, specific antigen serotypes and expression of Magnetic resonance imaging
tissue proteins need to be understood more deeply. One
may speculate that patients with the aforementioned pre- Magnetic resonance imaging (MRI) is an additional but not
disposing conditions may be at greater risk of developing essential imaging tool, because it does not give any addi-
CT. Furthermore, in this subpopulation of subjects, tional information in most cases [36, 37]. Calcific deposits

Table 1 Radiographic classification of calcifying tendinitis of the

shoulder
Author Subtype Description
Bosworth [1] Small \0.5 cm

Medium 0.5–1.5 cm
Large 1.5 cm
DePalma et al. [7] Type I Fluffy, amorphous and ill
defined
Type II Defined and homogeneous
Molè et al. (French Type A Dense, rounded, sharply
Arthroscopy Association) delineated
[27] Type B Multilobular, radiodense,
sharp
Type C Radiolucent,
heterogeneous, irregular
outline
Type D Dystrophic calcific
deposit
Gartner et al. [28, 29] Type I Well demarcated, dense
Type II Soft contour/dense or
sharp/transparent
Type III Soft contour/translucent
and cloudy
Table 2 Percentage of rotator

Tendon Percentage (%)
cuff tendon involvement in cal-
cifying tendinitis of the shoulder Supraspinatus 51
Infraspinatus 44.5
Teres minor 23.3
Subscapularis 3
signal intensity can be misinterpreted as a RC lesion [38,

39]. The accuracy of MRI in identifying calcific deposits is
around 95 %, but it is more useful in cases of chronic CT,
which may be associated with RC tears, AC and TO [10,
38, 40, 41] (Fig 2). All these investigations and a thorough
clinical examination are of critical importance, especially
when the primary disease is associated with signs and
symptoms of other conditions, e.g., the stiffness occurring
in the acute stage of the disease, which should be differ-
Fig. 1 A case with acute calcifying tendinitis of the rotator cuff. entiated from that occurring in AC or secondary stiffness
(a) X-ray shows a large calcium deposit ([1.5 cm) at the insertion of
the supraspinatus tendon in touch with the greater tuberosity; occurring in RC tears. Imaging must be used to differen-
(b) ultrasound image in the same patient as a demonstrates a large tiate chronic forms associated with TO from that occurring
fragmented and punctate calcification (dotted line) with hypoechoic in association with dystrophic calcification or in tumours
area indicating oedema associated with the reabsorptive phase (white [42].
arrows); (c) ultrasound-guided needling and lavage in the same case
as a and b with an abundant leakage of calcium (the window on the
left shows the calcium aspirated in a syringe)
Treatment options
have low signal intensity in all MRI sequences, although
areas of increased signal intensity can be found around Conservative management is always the first line of treat-
deposits in T2 images, signifying oedema around the ment. This includes non-steroidal anti-inflammatory drugs
deposits in the resorptive phase. Such areas of increased (NSAIDs), physiotherapy, UGN and ESWT. The outcome

then, it has been a commonly used intervention, as it is

inexpensive and can be carried out on an out-patient basis
under local anaesthesia (Fig. 1c). Gonzalez et al. [48]
recently published a study of 121 patients with 2-year
follow-up, reporting satisfactory results after UGN at
3 months. de Witte et al. [49] carried out a randomised
controlled trial (RCT) between UGN with subacromial
injection and subacromial injection alone; both groups
showed improvement, but the UGN group fared better as
compared with injection alone. A recent systemic review of
literature [50] for the efficacy of UGN in CT concluded
that, due to the variation in studies and the low quality of
evidence, the efficacy of UGN could not be firmly estab-
lished and additional high-quality studies are required.
Fig. 2 Coronal fatty suppressed MRI reveals a focus of chronic
calcification with associated full-thickness supraspinatus tendon tear
(white arrows) Extracorporeal shock wave therapy
ESWT has been used for medical treatment since the

of conservative treatment was principally studied by Ogon 1990s. Its use for CT is increasing, and like UGN, there is a
et al. [43], who described prognostic factors whose iden- lot of disparity, regarding the dosage (energy flux density),
tification was helpful for tailoring treatment for favourable duration (impulses) and interval of administration of
outcome in the shortest possible time. They defined failure ESWT.
of nonoperative therapy as persistence of symptomatic Low-energy (below 0.08 mJ/mm2), medium-energy
calcific tendinitis of the shoulder after a minimum of (0.08–0.28 mJ/mm2) and high-energy (0.28–0.60 mJ/mm2)
6 months of nonoperative treatment, including a minimum shock waves have been defined [51]. The shock waves can
of 3 months of standardised nonoperative treatment. They be generated through electrohydraulic, electromagnetic or
concluded that the prognostic factors that significantly piezoelectric mechanisms. Farr et al. [52] compared one
increased the probability of failure of nonoperative therapy dose of 0.3 mJ/mm2 versus two doses of 0.2 mJ/mm2,
(negative prognostic factors) were bilateral calcific deposit finding the former to be more effective. Ioppolo et al. [53]
occurrence, localization near the anterior portion of the also published a RCT and found 0.20 mJ/mm2 dosage to be
acromion, medial (subacromial) extension and high volume more effective than 0.10 mJ/mm2. Albert et al. [54] also
of calcific deposit. Prognostic factors that significantly found in favour of high-dose therapy, though their follow-
reduced the probability of failure of nonoperative therapy up was only 3 months and they did not find any significant
(positive prognostic factors) were Gartner type III calcific differences in the size of deposits on X-ray examination.
deposit and lack of sonographic sound extinction of the Various energy doses of ESWT have been reported for
calcific deposit. Treatment can be modulated depending treatment of CT; most authors described good clinical
upon the presence of these prognostic factors. Usually, the outcomes with low- and medium-energy waves [51–53,
acute phase requires NSAIDs to relieve the pain and 55–57]. The authors of a RCT [55] in which the control
appropriate physiotherapy [passive range-of-motion group was given sham treatment opined that the results
(ROM) exercises] to avoid stiffness of the shoulder. Local were better in the ESWT group. The researchers also
steroid injection in the acute phase is a debatable topic, as suggested other forms of treatment for patients who did not
studies have shown it to have positive [35] or no effect respond to ESWT after 6 months. Krasny et al. [56]
[44], or even a negative effect in the form of stopping compared ESWT alone and ESWT combined with UGN,
reabsorption of the deposits [45]. In most cases, conser- finding that the combined treatment was more effective in
vative treatment is sufficient for resolution of symptoms. relieving symptoms and that fewer patients in the com-
Cho et al. [46] reported excellent to good results in 72 % of bined treatment group required surgery. Daecke et al. [57]
their patients. published long-term follow-up of patients managed with
ESWT; although 20 % of all patients required surgery,
Ultrasound-guided needling 70 % of patients were treated successfully and no long-
term complications were seen. Lee et al. [58] carried out a
Although UGN was first demonstrated under fluoroscopy systematic review to determine the midterm effectiveness
control by Comfort et al. [47], it was Farin et al. [33] who of ESWT, but due to the variability of treatment and reli-
described use of US for bursal lavage and needling. Since ability of the available studies, they were not able to come

patients suggesting that complete removal of the deposits is

not essential; they also did not attempt repair of the defects
created following the removal of deposits. Other
researchers [65, 66] have also made similar suggestions. In
their study, Jerosch et al. [67] concluded that repair is not
required following removal of the deposits, but they
insisted on complete removal of the deposits. In contrast,
Porcellini et al. [68] recommended complete removal of
deposits followed by repair of the defect in the tendon,
using simple side-to-side sutures or suture anchors
depending upon the size of the residual defect. They argued
that repair gives similar results without the fear of propa-
gation of the tear and also helps in early patient rehabili-
tation. Tillander et al. [69] compared the outcome of
acromioplasty in 50 patients: 25 with CT and another 25
with other causes of impingement syndrome. They did not
find any significant difference between the Constant scores
of the two groups at 2 years and recommended that the
deposits should be left alone. However, other authors [64–
68] recommended acromioplasty only in cases of visible
mechanical impingement during arthroscopy, characterised
by roughening of the ligament and osteophytes on the
undersurface of the ligaments, as it did not have any
additional benefit and the number of cases requiring
acromioplasty varied in each of the studies.
Most authors [64–66, 68] recommended informing the
patient about delayed recovery post-surgery and were of
the opinion that surgical treatment should be reserved for
Fig. 3 (a) Arthroscopic findings shows a complete insertional patients not responding to conservative treatment for more
supraspinatus tendon tear after complete removal of a calcium than 6 months.
deposit; (b) supraspinatus tendon-to-bone repair with a double suture
anchor at the end of the arthroscopic procedure
Complications
to a conclusion regarding a particular dosage of treatment. In a recent review, Merolla et al. [11] described various
Kim et al. [59] carried out a comparative study between complications associated with CT. They categorised pain
UGN and ESWT, finding better radiological and clinical as a complication, as the majority of patients with CT are
outcomes in the UGN group, though both groups showed asymptomatic. Other complications in their study were
improvement relative to initial findings. secondary AC and RC tears, both of which could occur
during the primary disease or post-surgical intervention.
Surgical treatment They also pointed out ossifying tendinitis, which is an
extremely rare condition occurring following surgical
After failure of conservative treatment modalities, surgical removal of calcium deposits. Many authors [10, 11, 38, 40,
removal of the deposits is the remaining option. Although 41] have described TO of the greater tuberosity as an
favourable results have been described with open removal occurrence along with CT of the RC. Porcellini et al. [10]
of calcific deposits [4, 60–62], arthroscopy has become the suggested that TO should be identified as a different form
preferred technique to treat the chronic formative phase of of CT which is prone to delayed recovery of patients
CT, offering results similar to open surgery but with less managed conservatively and surgically. During UGN, mild
morbidity of the deltoid [63–69] (Fig. 3a, b). However, vasovagal syncope may occur. High-dose ESWT is asso-
many issues remain under debate, such as repairing versus ciated with pain sometimes requiring local anaesthesia, and
leaving the defect created, complete versus incomplete local haematoma, erythema and ecchymosis have also been
removal of the deposits and removal of deposits versus reported. Osteonecrosis of the humeral head has also been
only acromioplasty. Ark et al. [64] published a report of 23 described [70].

Overview 13. Sandstrom C (1938) Peridentinis calcarea: common disease of

middle life. Its diagnosis, pathology and treatment. Am J
Roentgenol 40:1–21
CT of the RC is a controversial topic with several treatment 14. Mohr W, Bilger S (1990) Basic morphologic structures of cal-
options that depend on the biologic stage of the disease. cified tendinopathy and their significance for pathogenesis.
Although reabsorption occurs spontaneously in the major- Z Rheumatol 49(6):346–355
15. Uhthoff HK, Loehr JW (1997) Calcific tendinopathy of the
ity of cases, a subpopulation of patients with persistent
rotator cuff: pathogenesis, diagnosis, and management. J Am
painful shoulder require conservative or operative man- Acad Orthop Surg 5:183–191
agement. In addition, some complications such as TO, AC 16. Rui YF, Lui PP, Chan LS, Chan KM, Fu SC, Li G (2011) Does
or ossifying tendinitis (very rare) may give rise to pro- erroneous differentiation of tendon-derived stem cells contribute
to the pathogenesis of calcifying tendinopathy? Chin Med J
longed pain resistant to common conservative therapies.
(Engl) 124(4):606–610
UGN is indicated in the acute phase, but good results have 17. Harvie P, Pollard TC, Carr AJ (2007) Calcific tendinitis: natural
also been found in patients with chronic calcific deposits. history and association with endocrine disorders. J Shoulder
ESWT can be reasonably used in chronic calcific cases, Elbow Surg 16(2):169–173
18. Mavrikakis ME, Drimis S, Kontoyannis DA, Rasidakis A,
even in combination with UGN. Surgical treatment should
Moulopoulou ES, Kontoyannis S (1989) Calcific shoulder peri-
be considered when conservative measures have failed or arthritis (tendinitis) in adult onset diabetes mellitus: a controlled
in cases with US or MRI evidence of RC tears. study. Ann Rheum Dis 48(3):211–214
19. Sengar DP, McKendry RJ, Uhthoff HK (1987) Increased fre-
quency of HLA-A1 in calcifying tendinitis. Tissue Antigens
29(3):173–174
20. Zhang Y, Johnson K, Russell RG, Wordsworth BP, Carr AJ,
Terkeltaub RA, Brown MA (2005) Association of sporadic
chondrocalcinosis with a -4-basepair G-to-A transition in the 50 -
untranslated region of ANKH that promotes enhanced expression
of ANKH protein and excess generation of extracellular inorganic
pyrophosphate. Arthritis Rheum 52(4):1110–1117
21. Oliva F, Barisani D, Grasso A, Maffulli N (2011) Gene expres-
sion analysis in calcific tendinopathy of the rotator cuff. Eur Cell
References Mater 21:548–557
22. Farin PU (1996) Consistency of rotator-cuff calcifications.
1. Bosworth BM (1941) Calcium deposits in the shoulder and Observations on plain radiography, sonography, computed
subacromial bursitis: a survey of 12122 shoulders. JAMA tomography, and at needle treatment. Invest Radiol
116:2477–2482 31(5):300–304
2. Codman EA (1934) The shoulder. Thomas Todd, Boston 23. Gosens T, Hofstee DJ (2009) Calcifying tendinitis of the shoul-
3. Plenk HP (1952) Calcifying tendinitis of the shoulder. Radiology der: advances in imaging and management. Curr Rheumatol Rep
59:384–389 11(2):129–134
4. Bosworth BM (1941) Examination of the shoulder for calcium 24. Molé D, Kempf JF, Gleyze P, Rio B, Bonnomet F, Walch G
deposits. Technique of fluoroscopy and spot film roentgenogra- (1993) Results of endoscopic treatment of non-broken tendino-
phy. J Bone Jt Surg 23:567–577 pathies of the rotator cuff. Calcifications of the rotator cuff [in
5. Welfling J, Kahn MF, Desroy M, Paolaggi JB, de Sèze S (1965) French]. Rev Chir Orthop 79:532–541
Calcifications of the shoulder. II. The disease of multiple tendi- 25. Gartner J, Heyer A (1995) Calcific tendinitis of the shoulder.
nous calcifications. Rev Rhum Mal Osteoartic 32(6):325–334 Orthopäde 24(3):284–302
6. DePalma AF, Kruper JS (1961) Long term study of shoulder 26. Gartner J, Simons B (1990) Analysis of calcific deposits in cal-
joints afflicted and treated for calcific tendinitis. Clin Orthop cifying tendinitis. Clin Orthop 254:111–120
20:61–72 27. Maier M, Schmidt-Ramsin J, Glaser C, Kunz A, Küchenhoff H,
7. Lippmann RK (1961) Observations concerning the calcific cuff Tischer T (2008) Intra- and interobserver reliability of classifi-
deposit. Clin Orthop 20:49–60 cation scores in calcific tendinitis using plain radiographs and CT
8. McLaughlin HL (1946) Lesions of the musculotendinous cuff of scans. Acta Orthop Belg 74(5):590–595
the shoulder: III. Observations on the pathology, course and 28. Farin PU, Räsänen H, Jaroma H, Harju A (1996) Rotator cuff
treatment of calcific deposits. Ann Surg 124(2):354–362 calcifications: treatment with ultrasound-guided percutaneous
9. McKendry RJ, Uhthoff HK, Sarkar K, Hyslop PS (1982) Calci- needle aspiration and lavage. Skeletal Radiol 25(6):551–554
fying tendinitis of the shoulder: prognostic value of clinical, 29. Papatheodorou A, Ellinas P, Takis F, Tsanis A, Maris I, Batakis
histologic, and radiologic features in 57 surgically treated cases. N (2006) US of the shoulder: rotator cuff and non-rotator cuff
J Rheumatol 9(1):75–80 disorders. Radiographics 26(1):e23
10. Porcellini G, Paladini P, Campi F, Pegreffi F (2009) Osteolytic lesion 30. Lin CH, Chao HL, Chiou HJ (2012) Calcified plaque resorptive
of greater tuberosity in calcific tendinitis of the shoulder. J Shoulder status as determined by high-resolution ultrasound is predictive of
Elbow Surg 18(2):210–215. doi:10.1016/j.jse.2008.09.016 successful conservative management of calcific tendinosis. Eur J
11. Merolla G, Bhat MG, Paladini P, Porcellini G (2015) Compli- Radiol 81(8):1776–1781. doi:10.1016/j.ejrad.2011.05.018
cations of calcific tendinitis of the shoulder: a concise review. 31. Chiou HJ, Chou YH, Wu JJ, Hsu CC, Huang DY, Chang CY
J Orthop Traumatol [Epub ahead of print] (2002) Evaluation of calcific tendonitis of the rotator cuff: role of
12. Bishop WA Jr (1938) Calcification of the supraspinatus tendon: color Doppler ultrasonography. J Ultrasound Med 21(3):289–295
cause, pathologic picture and relation to the scalenus anticus 32. Le Goff B, Berthelot JM, Guillot P, Glémarec J, Maugars Y
syndrome. Arch Surg 39:231–246 (2010) Assessment of calcific tendonitis of rotator cuff by

ultrasonography: comparison between symptomatic and asymp- 50. Vignesh KN, McDowall A, Simunovic N, Bhandari M, Choudur
tomatic shoulders. Jt Bone Spine 77(3):258–263. doi:10.1016/j. HN (2015) Efficacy of ultrasound-guided percutaneous needle
jbspin.2010.01.012 treatment of calcific tendinitis. AJR Am J Roentgenol
33. Farin PU, Jaroma H, Soimakallio S (1995) Rotator cuff calcifi- 204(1):148–152. doi:10.2214/AJR.13.11935
cations: treatment with US-guided technique. Radiology 51. Rompe JD, Kirkpatrick CJ, Kullmer K, Schwitalle M, Krischek O
195(3):841–843 (1998) Dose-related effects of shock waves on rabbit tendo
34. Chiou HJ, Chou YH, Wu JJ, Huang TF, Ma HL, Hsu CC, Chang Achillis: a sonographic and histological study. J Bone Jt Surg Br
CY (2001) The role of high resolution ultrasonography in man- 80(3):546–552
agement of calcific tendonitis of the rotator cuff. Ultrasound Med 52. Farr S, Sevelda F, Mader P, Graf A, Petje G, Sabeti-Aschraf M
Biol 27(6):735–743 (2011) Extracorporeal shockwave therapy in calcifying tendinitis
35. Uhthoff H, Sarkar K, Maynard J (1976) Calcifying tendinitis: a of the shoulder. Knee Surg Sports Traumatol Arthrosc
new concept of its pathogenesis. Clin Orthop Relat Res 19(12):2085–2089. doi:10.1007/s00167-011-1479-z
118:164–168 53. Ioppolo F, Tattoli M, Di Sante L, Attanasi C, Venditto T, Ser-
36. Loew M, Sabo D, Wehrle M, Mau H (1996) Relationship vidio M, Cacchio A, Santilli V (2012) Extracorporeal shock-wave
between calcifying tendinitis and subacromial impingement: a therapy for supraspinatus calcifying tendinitis: a randomized
prospective radiography and magnetic resonance imaging study. clinical trial comparing two different energy levels. Phys Ther
J Shoulder Elbow Surg 5(4):314–319 92(11):1376–1385. doi:10.2522/ptj.20110252
37. Zubler C, Mengiardi B, Schmid MR, Hodler J, Jost B, Pfirrmann 54. Albert JD, Meadeb J, Guggenbuhl P, Marin F, Benkalfate T,
CW (2007) MR arthrography in calcific tendinitis of the shoulder: Thomazeau H, Chalès G (2007) High-energy extracorporeal
diagnostic performance and pitfalls. Eur Radiol 17(6):1603–1610 shock-wave therapy for calcifying tendinitis of the rotator cuff: a
38. Bachmann GF, Melzer CH, Heinrics CM, Möhring B, Rominger randomised trial. J Bone Jt Surg Br 89(3):335–341
MB (1997) Diagnosis of rotator cuff lesions: comparison of US 55. Hsu CJ, Wang DY, Tseng KF, Fong YC, Hsu HC, Jim YF (2008)
and MRI on 38 joint specimens. Eur Radiol 7(2):192–197 Extracorporeal shock wave therapy for calcifying tendinitis of the
39. Rutten MJ, Jager GJ, Blickman JG (2006) From the RSNA shoulder. J Shoulder Elbow Surg Br 17(1):55–59
refresher courses: US of the rotator cuff: pitfalls, limitations and 56. Krasny C, Enenkel M, Aigner N, Wlk M, Landsiedl F (2005)
artifacts. Radiographics 26(2):589–604 Ultrasound-guided needling combined with shock-wave therapy
40. Maier M, Stäbler A, Schmitz C, Lienemann A, Köhler S, Dürr for the treatment of calcifying tendonitis of the shoulder. J Bone
HR, Pfahler M, Refior HJ (2001) On the impact of calcified Jt Surg Br 87(4):501–507
deposits within the rotator cuff tendons in shoulders of patients 57. Daecke W, Kusnierczak D, Loew M (2002) Long-term effects of
with shoulder pain and dysfunction. Arch Orthop Trauma Surg extracorporeal shockwave therapy in chronic calcific tendinitis of
121(7):371–378 the shoulder. J Shoulder Elbow Surg 11(5):476–480
41. Chan R, Kim D, Millet P, Weissman BN (2004) Calcifying 58. Lee SY, Cheng B, Grimmer-Somers K (2011) The midterm
tendinitis of rotator cuff with cortical bone erosion. Skeletal effectiveness of extracorporeal shockwave therapy in the man-
Radiol 33(10):596–599 agement of chronic calcific shoulder tendinitis. J Shoulder Elbow
42. Brown KT, Kattapuram SV, Rosentahl DI (1986) Computed Surg 20(5):845–854. doi:10.1016/j.jse.2010.10.024
tomography analysis of bone tumors: patterns of cortical destruc- 59. Kim YS, Lee HJ, Kim YV, Kong CG (2014) Which method is
tion and soft tissue extension. Skeletal Radiol 15(6):448–451 more effective in treatment of calcific tendinitis in the shoulder?
43. Ogon P, Suedkamp NP, Jaeger M, Izadpanah K, Koestler W, Prospective randomized comparison between ultrasound-guided
Maier D (2009) Prognostic factors in nonoperative therapy for needling and extracorporeal shock wave therapy. J Shoulder
chronic symptomatic calcific tendinitis of the shoulder. Arthritis Elbow Surg 23(11):1640–1646. doi:10.1016/j.jse.2014.06.036
Rheum 60(10):2978–2984. doi:10.1002/art.24845 60. Harmon PH (1958) Methods and results in the treatment of 2580
44. Noel E, Carillon Y, Gaillard T, Bouvier M (1997) Needle aspi- painful shoulders with special reference to calcific tendinitis and
ration irrigation in calcifying tendinitis of rotator cuff. In: the frozen shoulder. Am J Surg 95:527–544
Gazielly DF, Gleyze PT (eds) The cuff. Elsevier, Paris, 61. McLaughlin HL (1963) The selection of calcium deposits for
pp 152–157 operation: the technique and resultant operations. Surg Clin N
45. Tillander B, Franzen LE, Karlsson MH, Norlin R (1999) Effect of Am 43:1501–1504
steroid injections on the rotator cuff: an experimental study in 62. Rochwerger A, Franceschi JP, Viton JM, Roux H, Mattei JP
rats. J Shoulder Elbow Surg 8(3):271–274 (1999) Surgical management of calcific tendinitis of the shoulder:
46. Cho NS, Lee BG, Rhee YG (2010) Radiologic course of the an analysis of 26 cases. Clin Rheumatol 18(4):313–316
calcific deposits in calcific tendinitis of the shoulder: does the 63. Rubenthaler F, Ludwig J, Wiese M, Wittenberg RH (2003)
initial radiologic aspect affect the final results? J Shoulder Elbow Prospective randomized surgical treatments for calcifying
Surg 19(2):267–272. doi:10.1016/j.jse.2009.07.008 tendinopathy. Clin Orthop Relat Res 410:278–284
47. Comfort TH, Arafiles R (1978) Barbotage of the shoulder with 64. Ark JW, Flock TJ, Flatow EL, Bigliani LU (1992) Arthroscopic
image-intensified fluoroscopic control of needle placement for treatment of calcific tendinitis of the shoulder. Arthroscopy
calcified tendinitis. Clin Orthop Relat Res 135:171–178 8(2):183–188
48. Castillo-González FD, Ramos-Álvarez JJ, Rodrı́guez-Fabián G, 65. Seil R, Litzenburger H, Kohn D, Rupp S (2006) Arthroscopic
González-Pérez J, Calderón-Montero J (2014) Treatment of the treatment of chronically painful calcifying tendinitis of the
calcific tendinopathy of the rotator cuff by ultrasound-guided supraspinatus tendon. Arthroscopy 22(5):521–527
percutaneous needle lavage. Two years prospective study. Mus- 66. Balke M, Bielefeld R, Schmidt C, Dedy N, Liem D (2012) Cal-
cles Ligaments Tendons J 4(2):220–225 cifying tendinitis of the shoulder: midterm results after arthro-
49. de Witte PB, Selten JW, Navas A, Nagels J, Visser CP, Nelissen scopic treatment. Am J Sports Med 40(3):657–661. doi:10.1177/
RG, Reijnierse M (2013) Calcific tendinitis of the rotator cuff: a 0363546511430202
randomized controlled trial of ultrasound-guided needling and 67. Jerosch J, Strauss JM, Schmiel S (1998) Arthroscopic treatment
lavage versus subacromial corticosteroids. Am J Sports Med of calcific tendinitis of the shoulder. J Shoulder Elbow Surg
41(7):1665–1673. doi:10.1177/0363546513487066 7(1):30–37

68. Porcellini G, Paladini P, Campi F, Paganelli M (2004) Arthro- 70. Durst HB, Blatter G, Kuster MS (2002) Osteonecrosis of the
scopic treatment of calcifying tendinitis of the shoulder: clinical humeral head after extracorporeal shock-wave lithotripsy. J Bone
and ultrasonographic follow-up findings at two to five years. Jt Surg Br 84(5):744–746
J Shoulder Elbow Surg 13(5):503–508
69. Tillander BM, Norlin RO (1998) Change of calcifications after
arthroscopic subacromial decompression. J Shoulder Elbow Surg
7(3):213–217

30
Is end-stage lateral osteoarthritic knee always valgus? Mechanical
alignment analysis and radiographic severity assessment
Su Chan Lee1 • Viral Gondalia1 • Byoung Yoon Hwang1 • Hye Sun Ahn1 •
Choon Key Lee1 • David J. Hunter2 • Kwang Am Jung1
Abstract differences in lateral compartment eKLG/eJSN (2.3/

Background We hypothesized that not all persons with 2.3 mm in varus, 2.5/1.9 mm in neutral, 2.9/1.6 mm in
end-stage lateral osteoarthritis (OA) have valgus valgus, p = 0.01 and 0.03, respectively), tibia vara angle
malalignment and that full extension radiographs may (4.9° in varus, 4.1° in neutral, 3.0° in valgus, p \ 0.01),
underreport radiographic disease severity. The purpose of and medial compartment eKLG/eJSN (2.1/3.1 mm in
this study was to examine the demographic and radio- varus, 2.0/3.4 mm in neutral, 1.8/4.3 mm in valgus,
graphic features of end-stage lateral compartment knee p \ 0.01 and 0.01, respectively) between MAA groups,
OA. except for the tibial slope angle (9.7° in varus, 10.1° in
Materials and methods We retrospectively studied 133 neutral, 9.8° in valgus, p = 0.31).
knees in 113 patients who had undergone total knee Conclusion Varus alignment was paradoxically shown in
arthroplasty between June 2008 and August 2010. All approximately one-third of those with end-stage lateral
patients had predominantly lateral idiopathic compartment knee OA on both-leg standing hip-knee-ankle radiographs.
OA according to the compartment-specific Kellgren– Films taken in full extension underreported the degree of
Lawrence grade (KLG). The mechanical axis angle OA radiographic severity.
(MAA), compartment-specific KLG and joint space nar- Level of evidence Level IV, observational study.
rowing (JSN) of the tibiofemoral joint at extension and 30°
of knee flexion, tibia vara angle, tibial slope angle, body Keywords Lateral compartment Osteoarthritis Knee
mass index, age, and sex were surveyed.
Results End-stage lateral compartment knee OA has
varus (37.6 %), neutral (22.6 %), and valgus (39.8 %) Introduction
MAA on both-leg standing hip-knee-ankle radiographs.
KLGs at 30° of knee flexion (fKLG) were grades 3 and 4 in Osteoarthritis (OA) in the knee joint is the most common
all patients. However, for KLGs at full extension (eKLG), disorder in orthopedics and is characterized by structural
54 % of all patients had grades 3 and 4. The others (46 %) and functional failure of the synovial joint tissue with loss
showed grades 1 and 2. We observed significant and erosion of articular cartilage, subchondral bone alter-
ation, meniscal degeneration, and bone and cartilage ero-
sion [1].
& Kwang Am Jung Conventional radiography is the most convenient and
[email protected] important imaging examination in a clinical setting when
1
evaluating a patient who has a known or suspected diag-
Joint and Arthritis Research, Department of Orthopaedic
nosis of OA. Radiographs clearly visualize bony features,
Surgery, Himchan Hospital, 20-8, Songpa-dong, Songpa-gu,
Seoul 138-170, Korea including marginal osteophytes, subchondral sclerosis, and
2 bone cysts, but provide only an estimate of cartilage
Rheumatology Department, Royal North Shore Hospital and
Northern Clinical School, University of Sydney, Sydney, thickness and meniscal integrity by joint space narrowing
NSW, Australia (JSN). However, progression of JSN is the most commonly

Is end-stage lateral osteoarthritic knee always valgus? Mechanical alignment analysis and radiographic... 213
used criterion for the assessment of OA progression, and between a line perpendicular to the epiphysis and the
the complete loss of joint space width, characterized by anatomical axis of the tibia, which was measured using
interbone contact is one of the factors considered in the radiographs of the entire lower limb. The line to the epi-
decision for joint replacement. physis was measured perpendicular to the line that con-
Radiography is an indispensable complement to clinical nected both ends of the epimetaphy seal junction (Fig. 1)
examination and plays a key role in diagnosing and mon- [5–10]. Compartment-specific KLG and JSN were mea-
itoring the course of a condition in OA. Lower limb sured at extension and 30° of knee flexion on weight-
alignment and extended and/or semi-flexed knee antero- bearing views. Compartment-specific JSN was measured
posterior radiographs can be used for evaluating the rela- from the center of the medial/lateral condyle to the center
tionship between OA and compartmental pattern and of the medial/lateral tibial plateau [11]. To assess relia-
severity of knee OA. bility, each evaluation (KLG, MAA, tibia vara angle, and
However, most studies have focused on patients with tibial slope angle) was measured by two experienced
early stage OA. In addition, there is insufficient knowledge researchers (BYH, HSA) under the supervision of the
for lateral compartmental OA in comparison to medial coauthor (KAJ, with 10 years’ musculoskeletal radiology
compartment OA. This may raise the question of whether experience), who were blinded to patients’ information
similar findings from different radiographic methods are using the PACS system (INFINITT Healthcare Co Ltd,
found in end-stage lateral compartment OA. Better Seoul, Korea). The average of the two individual mean
understanding of the radiologic characteristics of end-stage values was used.
lateral OA will be helpful in diagnosing and managing
some patients with advanced disease. The purpose of this Statistical analysis
retrospective study was to examine the demographic and
radiographic features of end-stage lateral compartment SPSS (Statistical Package for the Social Sciences, v.12.0,
knee OA (Kellgren–Lawrence grade 3 or 4). We hypoth- Chicago, IL) was used for statistical analyses. For tibia
esized that not all persons with end-stage lateral OA have vara angle and tibial slope angle (continuous data), and
valgus malalignment and that full extension radiographs JSN, ANOVA was used to analyze differences in variables
may underreport radiographic disease severity. with differing MAA. For comparing KLG (categorical
ordinal data), a chi-square test was employed. Bivariate
analysis (Spearman’s correlation coefficient for categorical
Materials and methods data) was used to determine the correlation between radi-
ologic mismatches and variables. The interobserver relia-
We retrospectively reviewed the records of 133 patients bility in measuring variables (KLG, MAA, tibia vara angle,
who had undergone primary total knee arthroplasty JSN, and tibial slope angle) was evaluated using the intr-
between June 2008 and August 2010. All subjects had aclass correlation set at a 95 % confidence interval. A level
lateral compartment OA based on the following criteria: (1) of significance was set at p \ 0.05.
only lateral compartment involvement, (2) only lateral and
patellofemoral involvement, or (3) involvement of the
lateral and medial compartments (with or without patello- Results
femoral involvement) but with the lateral involvement
more severe than the medial involvement according to the There were 119 out of 133 knees from females. End-
compartment-specific Kellgren–Lawrence grade (KLG) [2, stage lateral compartment knee OA has varus (37.6 %),
3]. neutral (22.6 %), and valgus (39.8 %) MAA on both-leg
A detailed retrospective review of the medical records of standing hip-knee-ankle radiographs. KLG at 30° of knee
these patients was conducted to extract all pertinent flexion (fKLG) was grades 3 and 4 in all patients.
information on the body mass index and gender. Preoper- However, for the KLG at full extension (eKLG), 54 %
atively, standing hip-knee-ankle radiographs were taken of all patients had grades 3 and 4. The others (46 %)
and the mechanical axis angle (MAA) was measured. showed grades 1 and 2, which caused mismatches of
MAA is the angle between a line from the center of the KLG between extension and 30° of knee flexion (radi-
femoral head running distally to the mid-condylar point ologic mismatch). Only the MAA had a negative corre-
between the cruciate ligaments (femoral mechanical axis) lation with radiologic mismatch (r = -0.486, p \ 0.01)
and a line from the center of the tibial plateau extending (Table 1). With a more valgus MAA, there was less
distally to the center of the tibial plafond (tibial mechanical radiologic mismatch.
axis) [4]. The neutral MAA was categorized as 0° to 2° of We observed significant differences in lateral compart-
varus. The tibia vara angle was defined as the angle ment eKLG/eJSN (2.3/2.3 mm in varus, 2.5/1.9 mm in

valgus, p \ 0.01 and 0.01, respectively) between MAA

groups except for the tibial slope angle (9.7° in varus, 10.1°
in neutral, 9.8° in valgus, p = 0.31) (Table 2). The post-
hoc test showed increased mean lateral compartment
eKLG/eJSN in valgus MAA compared to the others,
increased tibia vara angle in varus MAA compared to
valgus, increased medial compartment eKLG/eJSN in
varus, and neutral MAA compared to valgus.
Among all patients, 20 patients (15.0 %) had bilateral
severe lateral OA, and 47 patients (35.3 %) had con-
tralateral severe medial OA who underwent TKA. There
was no difference in BMI or sex between the two groups.
However, in the former group, significant older age and
increased mismatch was observed.
The intraclass correlation coefficient for inter-tester
reliability of KLG, MAA, tibia vara angle, joint space
width (JSW), and tibial slope angle was 0.785, 0.833,
0.802, 0.753, and 0.812, respectively.
Discussion
This study focused on the radiologic and demographic

features of end-stage lateral knee OA. In this study, all
patients showed grades 3 and 4 KLGs in their flexion view,
representing bone to bone contact (end-stage). However,
not all patients showed grades 3 and 4 KLGs on extension
views. Valgus and neutral MAA accounted for the majority
of our sample with end-stage lateral knee OA. Varus
alignment was also paradoxically shown in approximately
one-third of those with end-stage lateral knee OA.
Radiographic protocols of the knee in flexion have been
shown to improve the detection of JSN by providing better
exposure of the location of the greater cartilage changes in
the posterior area of the femoral condyles [11–14]. The
flexion weight-bearing radiograph is commonly used and is
reportedly markedly better than the conventional radio-
graph in evaluating detection of JSN and disease severity
[15, 16]. The contact zones of femorotibial articulation
shift in both area and location as flexion occurs. As the
Fig. 1 Mechanical axis angle is the angle between a line from the knee is flexed during the stance phase of gait, the
center of the femoral head to the mid-condylar point between the
cruciate ligaments and a line from the center of the tibial plateau to femorotibial contact area moves posteriorly and decreases
the center of the tibial plafond (a). The tibia vara angle (black arrow) in size. With greater loads per unit of area, the cartilage is
is formed by the line perpendicular to the epiphysis (white arrow- more susceptible to degeneration in the contact zones of
head) and the anatomical axis of the tibia (white arrow) (b). The flexion. Because of this, the sensitivity and specificity of
posterior tibial slope angle is defined as 90° minus the angle made by
the intersection of the line along the longitudinal axis of the tibia and the flexion weight-bearing radiograph is markedly better
the slope of the medial tibial plateau (c) than the conventional extension radiograph [15, 16]. In
other words, the extension view risks underestimation in
diagnosing OA compared to the flexion view.
neutral, 2.9/1.6 mm in valgus, p = 0.01 and 0.03, respec- Despite the generalized loss of articular cartilage in end-
tively), tibia vara angle (4.9° in varus, 4.1° in neutral, 3.0° stage osteoarthritis, JSN is not always found to be consis-
in valgus, p \ 0.01), and medial compartment eKLG/eJSN tent between extension and flexion weight-bearing views,
(2.1/3.1 mm in varus, 2.0/3.4 mm in neutral, 1.8/4.3 mm in contributing to radiologic mismatch. Our study showed that

Table 1 Correlation coefficient

Pearson’s correlation coefficient p value
between radiologic mismatch
and other factors Age, years 0.096 0.29
Sex, M/F 0.014 0.88
Body mass index, kg/m2 -0.058 0.58
Mechanical axis angle, ° -0.486 \0.01
Tibia vara angle, ° 0.196 0.12
Tibial slope angle, ° 0.109 0.23
Lateral compartment joint space width at extension, mm 0.286 \0.01
Medial compartment joint space width at extension, mm -0.129 0.14
Lateral compartment joint space width at flexion, mm 0.002 0.98
Medial compartment joint space width at extension, mm 0.152 0.08
Table 2 Demographic and radiologic features of end-stage lateral knee osteoarthritis with differing alignment
Varus (n = 50) Neutral (n = 30) Valgus (n = 53) p value
Mean age (range), years 69.3 (57–80) 69.3 (60–79) 68.1 (57–83) 0.60
Sex, M/F 4/46 3/27 7/46 0.62
Body mass index, kg/m2 25.1 ± 3.3 26.2 ± 2.8 25.8 ± 3.2 0.44
Lateral compartment Kellgren–Lawrence grade at extension 2.3 ± 0.7 2.5 ± 0.8 2.9 ± 0.8 \0.01
Lateral compartment joint space width at extension, mm 2.3 ± 0.8 1.9 ± 0.9 1.6 ± 0.8 0.03
Radiologic mismatch, % 50.0 60.0 84.9 \0.01
Medial compartment Kellgren–Lawrence grade at extension 2.1 ± 0.4 2.0 ± 0.4 1.8 ± 0.5 0.03
Medial compartment joint space width at extension, mm 3.1 ± 1.2 3.4 ± 1.0 4.3 ± 0.9 0.01
Tibia vara angle, ° 4.6 ± 2.7 3.8 ± 2.3 2.7 ± 2.9 \0.01
Tibial slope angle, ° 9.7 ± 3.1 10.1 ± 2.8 9.8 ± 3.2 0.26
46 % of patients with end-stage lateral OA showed this progression of lateral disease, and an increasing valgus
phenomenon. angle is associated with more severe progression of lateral
Meniscal damage also has an important role in OA. The disease [2, 20, 21]. However, in our study, a large pro-
vast majority of meniscal tears occur in their posterior half portion of patients with end-stage lateral OA showed varus
and thus chondral damage and loss of joint space occurs and neutral alignment. Indeed ‘varus’ alignment was
when the knee is loaded in the flexed position. Anteriorly, shown in approximately one-third of patients with end-
the meniscus and articular cartilage is usually intact, and so stage lateral knee OA on both-leg standing hip-knee-ankle
in the extended position there is less loss of joint space. radiographs. However, it is clear that lateral cartilage loss
Due to a stance phase knee adduction moment, even during is advanced by ‘valgus’ alignment during walking. The
normal gait in healthy knees, more load passes through the knee which is originally valgus is simply seen to be ‘varus’
medial tibiofemoral compartment than through the lateral on both-leg standing hip-knee-ankle radiographs, but it is
compartment [17, 18]. For this reason the extension view is not ‘varus’. Brouwer et al. [22] reported the prevalence of
more suited for observing medial compartment OA than malalignment in knees without OA in 2290 knees, and
lateral disease. However, limb alignment becomes more observed 25 % with varus alignment, and 36 % with val-
valgus angulated with an increase in flexion rather than gus alignment. Interestingly, our study showed that in end-
extension [19], which distributes more weight in the lateral stage lateral OA, the distribution of valgus malalignment
compartment and shows lateral JSN in knee flexion. was similar to the normal knees proportion.
Previous reports noted that lower extremity malalign- Both knees demonstrated end-stage lateral OA (knock
ment increases the rate of progression of knee osteoarthritis knees) in 15 % of all patients. Of enrolled patients who
[20–22]. An increase in the varus angle was associated with showed contralateral medial OA, 35.3 % underwent TKA
a significantly increased adjusted risk of having severe (windblown knees). The patients with contralateral medial
medial disease. Also, valgus alignment increases the risk of compartment OA all had varus MAA in the contralateral

lower limb. There was no difference in BMI or sex

between knock and windblown knees. The former group
showed older age than the latter. This suggested that varus
alignment would affect the progression of medial com-
partment OA more, compared to the valgus alignment
effect on lateral compartment OA. Brouwer et al. [22]
observed a borderline effect of valgus malalignment on the
risk of incident OA, while varus malalignment had a larger
References
effect.
Some reported that limb alignment modulates the effect 1. Hunter DJ (2009) Insights from imaging on the epidemiology and
of standard risk factors for progression of OA of the knee, pathophysiology of osteoarthritis. Radiol Clin North Am
including obesity, quadriceps strength, laxity, and stage of 47(4):539–551
disease [23–25]. In medial compartment OA, limb align- 2. Khan FA, Koff MF, Noiseux NO, Bernhardt KA, O&Byrne MM,
Larson DR et al (2008) Effect of local alignment on compart-
ment has a great effect on OA prevalence. However, in mental patterns of knee osteoarthritis. J Bone Joint Surg Am
lateral OA, another factor beside limb malalignment would 90(9):1961–1969
have more effect on the lateral compartment than the 3. Kellgren JH, Lawrence JS (1957) Radiological assessment of
medial compartment. osteo-arthrosis. Ann Rheum Dis 16(4):494–502
4. Cooke TD, Li J, Scudamore RA (1994) Radiographic assessment
This study did have some limitations. First, we per- of bony contributions to knee deformity. Orthop Clin North Am
formed only a cross-sectional observational study. Second, 25(3):387–393
we attempted to control for this by grouping grades 3 and 4 5. Cooke TD, Sled EA, Scudamore RA (2007) Frontal plane knee
together (severe disease) and grades 1 and 2 together (mild alignment: a call for standardized measurement. J Rheumatol
34(9):1796–1801
to moderate disease). In this way, differences between 6. Park IS, Ong A, Nam CH, Ahn NK, Ahn HS, Lee SC, Jung KA
severe and mild to moderate disease are much more likely (2014) Transepicondylar axes for femoral component rotation
to indicate a real change in radiographic evidence of dis- might produce flexion asymmetry during total knee arthroplasty
ease progression than would have been the case if each of in knees with proximal tibia vara. Knee 21(2):369–373
7. Massin P, Gournay A (2011) Optimization of the posterior
the four grades had been considered separately. Third, condylar offset, tibial slope, and condylar roll-back in total knee
enrolled patients in this study were all symptomatic. The arthroplasty. J Arthroplasty 26:124–130
symptoms of OA of the knee are typically described as 8. Utzschneider S, Goettinger M, Weber P, Horng A, Glaser C,
mechanical—that is, they occur with physical activity. Jansson V et al (2011) Development and validation of a new
method for the radiologic measurement of the tibial slope. Knee
Fourth, this study was based on X-ray findings, not on Surg Sports Traumatol Arthrosc 19(10):1643–1648
MRI. JSW is determined not only by cartilage thickness, 9. Brandon ML, Haynes PT, Bonamo JR, Flynn MI, Barrett GR,
but also other factors such as knee angle, direction of the Sherman MF (2006) The association between posterior-inferior
X-ray beam, and meniscal status. We thought that further tibial slope and anterior cruciate ligament insufficiency. Arthro-
scopy 22:894–899
study based on MRI would enable assessment of the 10. Yoo JH, Chang CB, Shin KS, Seong SC, Kim TK (2008)
articular cartilage without being affected by any of these, Anatomical references to assess the posterior tibial slope in total
and could be helpful in uncovering the reason why such knee arthroplasty: a comparison of 5 anatomical axes. J Arthro-
mismatch occurs. plasty 23:586–592
11. Shelbourne KD, Dickens JF (2006) Digital radiographic evalua-
Valgus and neutral MAA accounted for the majority of tion of medial joint space narrowing after partial meniscectomy
our sample with end-stage lateral knee OA on the both-leg of bucket-handle medial meniscus tears in anterior cruciate
standing hip-knee-ankle radiographs. Varus alignment was ligament-intact knees. Am J Sports Med 34(10):1648–1655
also paradoxically shown in approximately one-third of 12. Merle-Vincent F, Vignon E, Brandt K, Piperno M, Coury-Lucas
F, Conrozier T et al (2007) Superiority of the Lyon schuss view
those with end-stage lateral knee OA. Radiographs taken in over the standing anteroposterior view for detecting joint space
full extension underreported the degree of OA radiographic narrowing, especially in the lateral tibiofemoral compartment, in
severity, with more mismatch being evident with more early knee osteoarthritis. Ann Rheum Dis 66(6):747–753
varus alignment. Varus MAA showed positive correlations 13. Rosenberg TD, Paulos LE, Parker RD, Coward DB, Scott SM
(1988) The forty-five-degree posteroanterior flexion weight-
with increased tibial vara angle and medial compartment bearing radiograph of the knee. J Bone Joint Surg Am
eKLG in end-stage lateral OA. 70(10):1479–1483
14. Piperno M (1998) Hellio Le Graverand MP, Conrozier T, Bochu
Conflict of interest None. M, Mathieu P, Vignon E. Quantitative evaluation of joint space
width in femorotibial osteoarthritis: comparison of three radio-
Ethical standards (1) All the patients gave informed consent prior graphic views. Osteoarthr Cartil 6(4):252–259
being included in the study; (2) the study was authorized by the local 15. Maquet PG, Berg AJ, Van de Simonet JC (1975) Femorotibial
ethical committee and was performed in accordance with the ethical weight-bearing areas. Experimental determination. J Bone Joint
standards of the 1964 Declaration of Helsinki as revised in 2000. Surg Am 57(6):766–771

16. Kurosawa H, Fukubayashi T, Nakajima H (1980) Load-bearing 22. Brouwer GM, van Tol AW, Bergink AP, Belo JN, Bernsen RM,
mode of the knee joint: physical behavior of the knee joint with or Reijman M et al (2007) Association between valgus and varus
without menisci. Clin Orthop Relat Res 149:283–290 alignment and the development and progression of radiographic
17. Morrison JB (1970) The mechanics of the knee joint in relation to osteoarthritis of the knee. Arthritis Rheum 56(4):1204–1211
normal walking. J Biomech 3(1):51–61 23. Felson DT, Goggins J, Niu J, Zhang Y, Hunter DJ (2004) The
18. Andriacchi TP (1994) Dynamics of knee malalignment. Orthop effect of body weight on progression of knee osteoarthritis is
Clin North Am 25(3):395–403 dependent on alignment. Arthritis Rheum 50(12):3904–3909
19. Koshino T, Takeyama M, Jiang LS, Yoshida T, Saito T (2002) 24. Sharma L, Lou C, Cahue S, Dunlop DD (2000) The mechanism
Underestimation of varus angulation in knees with flexion of the effect of obesity in knee osteoarthritis: the mediating role
deformity. Knee 9(4):275–279 of malalignment. Arthritis Rheum 43(3):568–575
20. Sharma L, Song J, Felson DT, Cahue S, Shamiyeh E, Dunlop DD 25. Cerejo R, Dunlop DD, Cahue S, Channin D, Song J, Sharma L
(2001) The role of knee alignment in disease progression and (2002) The influence of alignment on risk of knee osteoarthritis
functional decline in knee osteoarthritis. JAMA 286(2):188–195 progression according to baseline stage of disease. Arthritis
21. Cicuttini F, Wluka A, Hankin J, Wang Y (2004) Longitudinal Rheum 46(10):2632–2636
study of the relationship between knee angle and tibiofemoral
cartilage volume in subjects with knee osteoarthritis. Rheuma-
tology (Oxford) 43(3):321–324

31
Investigation of bone quality of the first and second sacral
segments amongst trauma patients: concerns about iliosacral
screw fixation
Dane Salazar1 • Sean Lannon1 • Olga Pasternak2 • Adam Schiff1 • Laurie Lomasney2 •
Erika Mitchell1 • Michael Stover3
Abstract Conclusion In relatively young, otherwise healthy trauma

Background Iliosacral screw fixation has become a patients there is a statistically significant difference in the
common method for surgical stabilization of acute dis- bone density of the first sacral segment compared to the
ruptions of the pelvic ring. Placement of iliosacral screws second sacral segment. This study highlights the need for
into the first sacral (S1) body is the preferred method of future biomechanical studies to investigate whether this
fixation, but size limitations and sacral dysmorphism may difference is clinically relevant. Due to the relative os-
preclude S1 fixation. In these clinical situations, fixation teopenia in the second sacral segment, which may impact
into the second sacral (S2) body has been recommended. the quality of fixation, we feel this technique should be
The objective of this study was to evaluate the bone quality used with caution.
of the S1 compared to S2 in the described ‘‘safe zone’’ of Level of evidence III
iliosacral screw fixation in trauma patients.
Materials and methods The pelvic computed tomography Keywords Iliosacral screws Pelvic fracture fixation
scans of 25 consecutive trauma patients, ages 18–49, at a Pelvic ring disruptions Regional bone density
level 1 trauma center were prospectively analyzed.
Hounsfield units, a standardized computed tomography
attenuation coefficient, was utilized to measure regional Introduction
cancellous bone mineral density of the S1 and S2. No
change in the clinical protocol or treatment occurred as a Iliosacral screw fixation has become a common method for
consequence of inclusion in this study. surgical stabilization of acute disruptions of the pelvic ring
Results A statically significant difference in bone quality [1–4]. Iliosacral screw placement can be accomplished
was found when comparing the first and second sacral percutaneously in conjunction with closed reduction or
segment (p = 0.0001). Age, gender, or smoking status did after open reduction; providing stability, minimizing de-
not independently affect bone quality. formity, facilitating mobilization and improving outcomes
in patients with posterior pelvic ring injuries [2, 3, 5].
However, loss of fixation, loss of function, neurovascular
& Dane Salazar injury and malunion have all been reported as serious
[email protected] complications following unstable posterior pelvic ring in-
1
juries treated using this method [1, 4–8]. Placement of il-
The Department of Orthopaedic Surgery and Rehabilitation,
Loyola University Health System, 2160 South First Avenue,
iosacral screws into the S1 body is the preferred method of
Maywood, IL 60153, USA fixation, but size limitations and sacral dysmorphism may
2 preclude S1 fixation [4, 9]. In these clinical situations,
The Department of Radiology, Loyola University Health
System, 2160 South First Avenue, Maywood, IL 60153, USA fixation into the second sacral body (S2) has been recom-
3 mended [3, 10, 11].
Department of Orthopaedic Surgery, Feinberg School of
Medicine, Northwestern University, 676 N. Saint Clair, Suite Although safe zones for screw fixation in both normal
1350, Chicago, IL 60611, USA and dysmorphic second sacral segments have been

Investigation of bone quality of the first and second sacral segments amongst trauma patients: concerns... 219
Fig. 1 CT images depicting the cross-referencing technique used for localizing the mid body point of S1 (a) and S2 (b) for measurement in
axial, sagittal and coronal reconstructions
established, challenges exist in achieving proper fixation workup of 25 consecutive trauma patients meeting the in-
into the S2 body because of its smaller size and decreased clusion criteria were obtained between July 2008 and
tolerance of variant screw trajectories [3, 11, 12]. In spite January 2011. All subjects were between the ages of
of multiple studies on surgical techniques for optimal 18–50 years of age to limit the effects of age-related bone
placement of fixation, there is little mention of the quality loss or skeletal immaturity. Subjects were excluded from
of the surrounding bone in the S2 body. The purpose of this this study for the following reasons: previous documented
study is to investigate the bone density of the first and sacral trauma, presence of a zone 3 sacral fracture, neo-
second sacral segments using Hounsfield units, a stan- plasm of the pelvic girdle, documented history of
dardized computed tomography attenuation coefficient. We rheumatoid arthritis, documented history of seronegative
hypothesize that S2 bone density is inferior to that of S1, arthropathies, documented history of osteoporosis or os-
increasing the chances of screw loosening and fixation teopenia, history of paraplegia, non-ambulatory/wheelchair
failure despite screw placement consistent with accepted bound, an inadequate scan technique that would limit
methods in the literature. density determination, including, but not limited to, motion
artifact, streak artifact from internal hardware or external
metallic devices, beam-hardening artifact, or photon de-
Materials and methods privation in the extremely obese patient, known use of
bisphosphonates, steroids and/or hormone medications, or
The study was approved by our institutional review board evident malnutrition. The patient’s age, gender and smok-
and carried out in the radiology suite of a level 1 trauma ing history were recorded from the electronic medical
center emergency department. Pelvic computed tomogra- record. No change in protocol or treatment occurred as a
phy (CT) scans obtained as part of the routine trauma consequence of inclusion in this study.

Fig. 2 Axial CT sections demonstrating the technique for ROI the sacral foramina at S1 (red dashed). d Placement of ROIs at 25 and
placement as described in the methods section. a Horizontal reference 75 % of the vertical midline distance and 50 % of the vertical lateral
lines at S1 (blue). b Vertical midline reference line at S1 (green distance at S1 (white). e–h The same technique at S2 (color figure
dashed). c Vertical reference line tangential to the medial border of online)
Images were viewed using the bone algorithm default from 23.2 to 26.2 mm2. This range was chosen after pilot
windows on picture archiving and communication system testing to maximize the area of trabecular bone tested in
(PACS) viewing software. Using axial images, the midline with the potential screw trajectory, while limiting
body location of S1 and S2 was determined for each sub- overlap of adjacent ROIs. When placing ROIs, one
ject and confirmed by cross-referencing position with horizontal reference line was drawn tangential to the most
coronal and sagittal reconstructions (Fig. 1). To standard- anterior points of both sacral foramina (Fig. 2a, e). One
ize measurement while accounting for normal anatomic transecting vertical reference line was then drawn from the
variation and optimal iliosacral screw trajectory as de- tip of the spinous process through the midpoint of the an-
scribed in the literature, four standardized circular voxel terior cortex of the vertebral body (Fig. 2b, f). ROIs were
regions of interests (ROIs) were drawn at determined mid- then drawn with their center corresponding to 25 and 75 %
body S1 and S2 levels of each subject (Fig. 2). These of the distance from the anterior cortex to the horizontal
standardized circular ROIs were drawn with areas ranging reference line. A vertical reference line was then drawn

Fig. 3 Comparison of mean

aggregate bone density
measurement of S1 vs. S2 in
each subject, as measured in
Hounsfield Units (HU)
tangential to the most medial point of the sacral foramina The values of bone mineral density in Hounsfield units
(Fig. 2c, g). An ROI was then drawn with the center of the for each of the tested points are detailed in Table 1 for each
ROI at 50 % of the distance between the anterior cortex of the patients. All four points were found to have a sta-
and horizontal reference line drawn previously. This tistical difference between S1 and S2 (anterior p = 0.0011,
method was then repeated on the adjacent side. Figure 2d posterior p \ 0.0001, right p \ 0.0001, left p \ 0.0001).
and h demonstrates the placement of ROIs. Hounsfield unit The percentage difference of mean density measured with
(HU) density values for each ROI were then collected and Hounsfield units between S1 and S2 is presented in
averaged to yield the mean value for each segment. Table 2.
Prospective power analysis was conducted and revealed
that a sample size of 25 patients was necessary to detect a
difference in bone density of SI compared to S2 at the 0.05 Discussion
alpha level with 80 % power. Statistical analysis of the
data was performed on the mean values for each segment in Iliosacral screw fixation has emerged as the treatment of
the four ROIs examined using paired Student’s t tests with choice for unstable injuries involving the posterior pelvic
statistical significance being set at p \ 0.05. ring. However, the posterior pelvic anatomy is complex
and variable, and thus placement of fixation can be tech-
nically challenging. A 44 % incidence of sacral dysmor-
Results phism has been reported; therefore, a thorough
understanding of the typical as well as atypical individual
Twenty-five patients, with a mean age of 35.2 years, were anatomy is critical for reliably placing safe iliosacral
studied (ages 18–49 years). Thirteen patients had a positive screws [3, 11]. In dysmorphic sacra, the first sacral safe
smoking history. Nineteen patients were male and 6 fe- zone was 36 % smaller compared to the normal counter-
male. The difference between the average Hounsfield unit parts, and with more oblique orientation from caudal to
(HU) of the first and second sacral segment was 89.9 cranial and posterior to anterior [11]. In the second seg-
(p = 0.0001). Comparisons of the mean bone density of ment safe zone, the cross-sectional area was more than
the first and second sacral segments are presented in Fig. 3. twice as large in the dysmorphic sacra compared to normal
The 13 patients with a positive smoking history had a [11]. Additionally, it was found that a transverse screw
mean HU of 93.7 compared to the mean HU value of non- could be safely placed at the S2 level in 95 % of dys-
smokers, 85.8 (p = 0.66). The average HU difference morphic sacra but only in 50 % of normal sacra [11].
when comparing males, 87.2, versus females, 98.6, was The optimal fixation construct remains unclear; how-
11.4 (p = 0.58). Age had no significant effect on HU dif- ever, injuries with multiplanar instability have increased
ference (p = 0.53). the rates of fixation failure [13]. Biomechanical studies

222
Table 1 Values of bone mineral density in Hounsfield units of analogous ROI locations at S1 and S2 with differences
Patient S1 S2 Difference S1 S2 Difference 2 S1 S2 Difference 3 S1 left S2 left Difference 4
no. anterior anterior posterior posterior right right
1 295.9 226.9 69 261.5 205 56.5 321.2 158.9 162.3 288.1 156.3 131.8
2 240.9 140.9 100 170.3 98.9 71.4 270.2 106.1 164.1 259.7 92.3 167.4
3 1313 1272.4 40.6 1358.7 1298.6 60.1 1369 1294.5 74.5 1310 1288.7 21.3
4 280.9 236.2 44.7 211.8 200 11.8 296.7 186.8 109.9 294.8 174.3 120.5
5 229.5 239.3 -9.8 263 164.9 98.1 263 218.1 44.9 256.8 244.3 12.5
6 230.1 236.3 -6.2 256.7 134.9 121.8 313.5 177.5 136 323.8 180.7 143.1
7 1341.3 1279.4 61.9 1348.8 1262.9 85.9 1335.9 1306.1 29.8 1355.6 1277.3 78.3
8 1342.1 1279.2 62.9 1292.7 1265.5 27.2 1263.3 1237.2 26.1 1258.9 1222.2 36.7
9 1292.3 1237.3 55 1263.8 1146.6 117.2 1148.1 1103.6 44.5 1154.5 1142.2 12.3
10 306.7 268.6 38.1 331.4 200.1 131.3 266.5 222.4 44.1 307 224.7 82.3
11 361.4 373.9 -12.5 547.3 391.2 156.1 447.3 344.5 102.8 411.9 393.5 18.4
12 1392.7 1230.8 161.9 1389.2 1234.3 154.9 1398.8 1191.3 207.5 1417.7 1240.4 177.3
13 293.4 233.4 60 393.3 80.3 313 283.4 160.8 122.6 267.9 153.4 114.5
14 139.5 97 42.5 138.5 41.9 96.6 190 38.7 151.3 216.6 43 173.6
15 287.1 408.7 -121.6 341.9 290.6 51.3 379 256.7 122.3 309.1 296.6 12.5
16 247.8 223 24.8 212.3 151.7 60.6 305.2 136 169.2 254.1 163.2 90.9
17 212.2 167.7 44.5 188.2 106.2 82 150.2 152.6 -2.4 146.1 124.3 21.8
18 248.7 277.9 -29.2 261.7 158 103.7 388.4 197.2 191.2 328.5 173.2 155.3
19 223.4 196.3 27.1 248.9 215.2 33.7 268.2 139.7 128.5 275.3 220.4 54.9
20 271.7 279.3 -7.6 292.4 189.4 103 305.2 217.8 87.4 317.7 176.8 140.9
21 207.9 77.9 130 235 58 177 128.1 39.3 88.8 167.1 44.5 122.6
22 407.5 207.7 199.8 401.9 202.4 199.5 311.9 191.3 120.6 365.2 237.7 127.5
23 155.5 129.4 26.1 162.1 166.2 -4.1 254.9 97.6 157.3 203.3 119.1 84.2
24 367.5 181.4 186.1 172.5 66.4 106.1 232.7 143.9 88.8 258.8 137.9 120.9
25 422.2 306.8 115.4 370.9 223.9 147 384.3 196.2 188.1 367.5 224 143.5
52.14 102.47 110.41 94.6
(p = 0.0011) (p \ 0.0001) (p \ 0.0001) (p \ 0.0001)
All values listed are in Hounsfield units

Table 2 Values of bone mineral density in Hounsfield units of analogous ROI locations at S1 and S2 with differences
Patient no. Difference % Difference % Difference % Difference left %
anterior ROI Difference posterior ROI Difference right ROI Difference ROI Difference
S2 vs. S1 S2 vs. S1 S2 vs. S1 S2 vs. S1
anterior ROI anterior ROI anterior ROI anterior ROI
1 69 77 57 78 162 49 132 54
2 100 58 71 58 164 39 167 36
3 41 97 60 96 75 95 21 98
4 45 84 12 94 110 63 121 59
5 -10 104 98 63 45 83 13 95
6 -6 103 122 53 136 57 143 56
7 62 95 86 94 30 98 78 94
8 63 95 27 98 26 98 37 97
9 55 96 117 91 45 96 12 99
10 38 88 131 60 44 83 82 73
11 -13 103 156 71 103 77 18 96
12 162 88 155 89 208 85 177 87
13 60 80 313 20 123 57 115 57
14 43 70 97 30 151 20 174 20
15 -122 142 51 85 122 68 13 96
16 25 90 61 71 169 45 91 64
17 45 79 82 56 -2 102 22 85
18 -29 112 104 60 191 51 155 53
19 27 88 34 86 129 52 55 80
20 -8 103 103 65 87 71 141 56
21 130 37 177 25 89 31 123 27
22 200 51 200 50 121 61 128 65
23 26 83 -4 103 157 38 84 59
24 186 49 106 38 89 62 121 53
25 115 73 147 60 188 51 144 61
Mean 52 102 110 95
difference (p = 0.0011) (p \ 0.0001) (p \ 0.0001) (p \ 0.0001)
(p value)
Percent 86 68 65 69
difference
of S2
compared to
S1
Average global density of S2 compared to S1 71.9
have suggested improved stability using two points of alternative or adjunct fixation method to the more common
posterior fixation for the treatment of unstable pelvic ring first sacral segment [3, 4, 13].
injuries [14, 15]. Therefore, the placement of two fixation However there is a paucity of data examining the quality
screws has been recommended to aid with stability. Several of bone of the second sacral segment compared to the first
clinical scenarios necessitate the placement of fixation into sacral segment. In one clinical series with 62 patients
the second sacral segment. treated with closed reduction and placement of percuta-
Multiple cadaveric and in vivo studies have investigated neous iliosacral screws for unstable pelvic ring injuries, 2
proving the efficacy and safety of S2 screw fixation using patients were managed with 2 S1 screws, 3 with 2 screws in
both fluoroscopic and computer tomography-based multi- S2, 56 with 1 S1 and another in S2, and 1 patient with 2
planar guidance systems to identify reliable and repro- screws in S1 and a 3rd in S2. Fixation failure occurred in 4
ducible landmarks to establish a safe corridor [10, 12, 16– of 62 patients. Retrospectively, five patients were identified
19]. Several case series have established the placement of as being osteopenic, with two of these five patients having
fixation into the second sacral segment as a dependable early fixation failure. This led the authors to conclude that

S2 screws should be used with caution in patients with fixation in the second sacral segment is warranted and have
suspected pelvic and sacral osteopenia/osteoporosis [13]. proposed that this technique is safe and effective. However,
Additionally, in a series of 49 patients all treated with S2 given our findings of relative osteopenia in the second
screws, 2 had postoperative loss of reduction requiring sacral segment, which may impact the quality of fixation,
revision surgery, both with radiographic evidence of os- we feel this technique should be used with caution.
teopenia. This led to the recommendation of finding al-
ternative fixation methods in those patients with osteopenia Conflict of interest None.
and in patients with questionable intraoperative screw Ethical standard The study was authorized by the local ethical
purchase during placement [3]. To our knowledge, our committee and was performed in accordance with the Ethical stan-
study is the first to specifically compare the bone densities dards of the 1964 Declaration of Helsinki as revised in 2000.
of the first two sacral segments.
Multiple modalities of measuring bone density have
been described and validated, including dual X-ray ab-
sorptiometry (DEXA), plain radiographs and quantitative
computed tomography [20]. More recent studies have
demonstrated that computed tomography examinations
utilizing automatic exposure control are able to accurately
measure regional cancellous bone mineral density [21]. In
our study we utilized Hounsfield units, a standardized
computed tomography attenuation coefficient, which has References
been shown to correlate with both the DEXA and com-
pressive strengths of osseous models. We hypothesized that 1. Matta JM, Saucedo T (1989) Internal fixation of pelvic ring
fractures. Clin Orthop Relat Res 83–97
S2 bone density is inferior to that of S1, increasing the 2. Matta JM, Tornetta P 3rd (1996) Internal fixation of unstable
chances of screw loosening and fixation failure despite pelvic ring injuries. Clin Orthop Relat Res 129–140
screw placement consistent with accepted methods in the 3. Moed BR, Geer BL (2006) S2 iliosacral screw fixation for dis-
literature. ruptions of the posterior pelvic ring: a report of 49 cases. J Orthop
Trauma 20:378–383
We prospectively assessed the pelvic computed to- 4. Routt ML Jr, Simonian PT (1996) Closed reduction and percu-
mography scans of 25 consecutive trauma patients taneous skeletal fixation of sacral fractures. Clin Orthop Relat
evaluated in the Emergency Department of a level 1 trauma Res 121–128
center. We found a statistically significant difference in the 5. Shuler TE, Boone DC, Gruen GS et al (1995) Percutaneous il-
iosacral screw fixation: early treatment for unstable posterior
bone density at all four points and the aggregate of S1 pelvic ring disruptions. J Trauma 38:453–458
compared to S2. Smoking history, gender and age were not 6. Keating JF, Werier J, Blachut P et al (1999) Early fixation of the
found to be independent factors in contributing to this vertically unstable pelvis: the role of iliosacral screw fixation of
difference. the posterior lesion. J Orthop Trauma 13:107–113
7. Tonetti J, Cloppet O, Clerc M et al (2000) Implantation of il-
One of the limitations of our study is that Hounsfield iosacral screws. Simulation of optimal placement by 3-dimen-
units on computed tomography were used as a surrogate sional X-ray computed tomography. Revue de chirurgie
measurements of ‘‘bone density’’ or ‘‘bone quality.’’ This orthopedique et reparatrice de l’appareil moteur 86:360–369
non-invasive method is well described in the literature [21] 8. Cole JD, Blum DA, Ansel LJ (1996) Outcome after fixation of
unstable posterior pelvic ring injuries. Clin Orthop Relat Res
and has previously been utilized as a tool to draw con- 160–179
clusions about bone mineral density; however, it should be 9. Conflitti JM, Graves ML, Chip Routt ML Jr (2010) Radiographic
noted that it is a quantitative and not a qualitative mea- quantification and analysis of dysmorphic upper sacral osseous
surement. To directly calculate bone quality and thus truly anatomy and associated iliosacral screw insertions. J Orthop
Trauma 24:630–636
investigate the local trabecular microarchitecture of bone, 10. Carlson DA, Scheid DK, Maar DC et al (2000) Safe placement of
would require a bone biopsy. S1 and S2 iliosacral screws: the ‘‘vestibule’’ concept. J Orthop
The optimal fixation for posterior pelvic ring injuries Trauma 14:264–269
remains unclear. Our study demonstrates that in relatively 11. Gardner MJ, Morshed S, Nork SE et al (2010) Quantification of
the upper and second sacral segment safe zones in normal and
young, otherwise healthy trauma patients there is a statis- dysmorphic sacra. J Orthop Trauma 24:622–629
tically significant difference in the bone density of the first 12. Gautier E, Bachler R, Heini PF et al (2001) Accuracy of com-
sacral segment compared to the second sacral segment. puter-guided screw fixation of the sacroiliac joint. Clin Orthop
This study highlights the need for future biomechanical Relat Res 310–317
13. Griffin DR, Starr AJ, Reinert CM et al (2006) Vertically unstable
studies to investigate whether this difference has a pelvic fractures fixed with percutaneous iliosacral screws: does
clinically relevant effect on the quality of fixation. Previous posterior injury pattern predict fixation failure? J Orthop Trauma
studies have highlighted clinical scenarios in which 20:S30–S36 (discussion S36)

14. van Zwienen CM, van den Bosch EW, Snijders CJ et al (2004) 19. Ziran BH, Smith WR, Towers J et al (2003) Iliosacral screw
Biomechanical comparison of sacroiliac screw techniques for fixation of the posterior pelvic ring using local anaesthesia and
unstable pelvic ring fractures. J Orthop Trauma 18:589–595 computerised tomography. J Bone Joint Surg Br 85:411–418
15. Yinger K, Scalise J, Olson SA et al (2003) Biomechanical 20. Grampp S, Genant HK, Mathur A et al (1997) Comparisons of
comparison of posterior pelvic ring fixation. J Orthop Trauma noninvasive bone mineral measurements in assessing age-related
17:481–487 loss, fracture discrimination, and diagnostic classification. J Bone
16. Arman C, Naderi S, Kiray A et al (2009) The human sacrum and Miner Res 12:697–711
safe approaches for screw placement. J Clin Neurosci 21. Schreiber JJ, Anderson PA, Rosas HG et al (2011) Hounsfield
16:1046–1049 units for assessing bone mineral density and strength: a tool for
17. Hinsche AF, Giannoudis PV, Smith RM (2002) Fluoroscopy- osteoporosis management. J Bone Joint Surg Am 93:1057–1063
based multiplanar image guidance for insertion of sacroiliac
screws. Clin Orthop Relat Res 135–144
18. Nottmeier EW, Pirris SM, Balseiro S et al (2010) Three-dimen-
sional image-guided placement of S2 alar screws to adjunct or
salvage lumbosacral fixation. Spine J 10:595–601

32
Aseptic lysis L2–L3 as complication of abdominal aortic aneurysm
repair
Federico Mancini • Andrea Ascoli-Marchetti •
Luca Garro • Roberto Caterini
Abstract Osteolytic vertebral erosion is usually related to Introduction

tumours, spondylitis or spondylodiscitis. Few reports in the
literature describe lytic lesions of anterior lumbar vertebral Osteolytic vertebral erosion is usually related to tumour,
bodies resulting from abdominal aortic aneurysm or false spondylitis or spondylodiscitis. Vertebral erosion deter-
aneurysm. We report a case of abdominal aortic false mined by an abdominal aortic aneurysm is rare but already
aneurysm that caused lytic lesions of the second and third described [1–5]. Very few cases of vertebral lesion caused
vertebral bodies in an 80-year-old man who underwent by false aneurysm secondary to prosthetic stent have been
endovascular aneurysm repair. Fluoroscopy guided biopsy reported [5–9], and this is the first case of vertebral erosion
excluded infection or tumour. We performed a posterior due to a false aneurysm in a patient who underwent en-
spinal fusion and decompression because of bone loss of dovascular aneurysm surgery in absence of disco-vertebral
the second and third lumbar vertebral bodies and central infection.
stenosis. Postoperatively the patient showed satisfactory
relief in low-back and thigh pain but, unfortunately, he died
1 month after surgery because of respiratory complications. Case report
This case suggests that when a lytic lesion of a lumbar
vertebral body is discovered in a patient who has under- An 80-year-old man, with multiple co-morbidities was
gone endovascular aneurysm repair, an abdominal aortic admitted to the author’s hospital for severe low-back pain,
false aneurysm may be the cause of the vertebral erosion lower limbs motor impairment and bilateral thigh pain.
even in cases without infective pathogenesis. Nine months before admission to our department, he
underwent endovascular aneurysm repair (EVAR) for an
Keywords Contained aneurysm rupture Vertebral abdominal aortic aneurysm (Fig. 1). A computed tomog-
erosion False aneurysm Endovascular repair raphy (CT) scan performed 1 month after the endovascular
abdominal aortic aneurysm repair did not show any signs
of lumbar vertebral or disc erosion (Fig. 2). Three months
after the endovascular procedure, he developed a progres-
sive lower back pain and bilateral thigh pain that did not
respond to conservative treatments. At this time, in another
hospital, MRI of the lumbar spine showed severe bone loss
F. Mancini L. Garro (&) R. Caterini
in the anterior half of the third lumbar vertebral body and
Department of Orthopaedics and Traumatology,
University of Rome ‘‘Tor Vergata’’, Viale Oxford, 81, L3 vertebroplasty was performed without any significant
00133 Rome, Italy relief of the symptoms (Fig. 3). On admission in our hos-
e-mail: [email protected] pital, 9 months after EVAR, clinical evaluation showed
severe low-back pain, bilateral thigh pain, motor deficit of
A. Ascoli-Marchetti
Department of Vascular Surgery, University of Rome the lower limbs and the patient was not able to walk
‘‘Tor Vergata’’, Rome, Italy (Frankel C) [10]. Laboratory findings were within normal

Aseptic lysis L2–L3 as complication of abdominal aortic aneurysm repair 227
Fig. 1 Intraoperative
angiography before (a) and after
(b) the endoprosthesis
deployment. No signs of rupture
are evident
Fig. 2 Computed tomography scan 1 month after endovascular

abdominal aortic aneurysm repair. No signs of vertebral erosion are
present Fig. 3 Lumbar spine X-ray showing vertebroplasty procedure per-
formed 3 months after aneurysm repair
limits: haemoglobin, 12.5 g/dl (normal 12–16 g/dl), white

blood cell count, 6.3 9 109/l (normal 4.3–10.8 9 109/l) an extensive abdominal aortic false aneurysm, corre-
with a normal differential, erythrocyte sedimentation rate sponding with the prosthetic stents that had eroded the
25 mm/h (normal 2–30 mm/h), C-reactive protein 2.4 mg/l vertebral bodies of L2–L3. The vascular surgeon did not
(normal 0.00–3.00 mg/l). A new MRI of the lumbar spine, consider revision surgery necessary at that time. Labora-
performed at admission, showed the false aneurysm and its tory findings, fluoroscopy guided biopsy and PET-CT
relation to L2–L3 bodies causing vertebral and disc erosion (Fig. 5) had excluded infection, and for this reason we
(Fig. 4). Fluoroscopy guided biopsy was performed, but it avoided draining the fluid at the L2–L3 disc space, but we
was negative as regards microbiological and histopatholo- opted for a posterior decompression of the central canal
gical examination for tumours or infections. A more stenosis, between L2 and L3, to improve the neurologic
accurate evaluation of MRI of the lumbar spine uncovered condition of the patient. In addition a long and extensive

Fig. 4 a MRI 9 months after endovascular abdominal aortic aneurysm repair showed the false aneurysm and its relation to L2–L3 vertebrae.
b L2–L3 central stenosis
Fig. 5 PET-CT section 9 months after endovascular abdominal aor- Fig. 6 Postoperative X-ray of lumbar spine
tic aneurysm repair was not significative of infection
instrumentation of T12–L5 was performed to avoid the risk the disc space L2–L3 we just found abundant serum-
of implant failure due to the presence of severe vertebral hematic fluid. Another biopsy of the L2–L3 disc performed
osteopenia (Fig. 6). Moreover, an autologous iliac bone during surgery confirmed the absence of tumours or
graft was utilized to obtain better postero-lateral fusion. In infections. Postoperatively, laboratory findings were:

Aseptic lysis L2–L3 as complication of abdominal aortic aneurysm repair 229
haemoglobin 9.2 g/dl, white blood cell count 8.5 9 109/l, vertebral or disc erosion even in cases without infective
erythrocyte sedimentation rate 35 mm/h, C-reactive protein complication.
3.00 mg/l. After surgery the patient obtained excellent
relief of low-back and thigh pain with a satisfactory regain Conflict of interest The authors declare that they have no conflict
of interest.
of walking (Frankel D). Unfortunately, 1 month after sur-
gery the patient died because of respiratory complications. Ethical standards The patient provided his consent to the publi-
We excluded sepsis as a possible cause of death because cation of this case report.
the patient in the postoperative period had no fever and
laboratory findings were within normal limits.
Discussion
Osteolytic vertebral erosion is usually caused by neo- References

plasms, spondylitis or spondylodiscitis. Some reports in the
1. Boonen A, Ghesquiere B, Westhovens R et al (1995) Vertebral
literature report that an abdominal aortic aneurysm can fracture induced by chronic contained rupture of aortic aneurysm.
cause erosion of the lumbar vertebral body, due to a pro- Ann Rheum Dis 54:437–438
gressive aneurysmatic sac expansion [1–5]. Few authors 2. Mii S, Mori A, Yamaoka T et al (1999) Penetration by huge
[6, 7] have reported lumbar vertebral erosion resulting from abdominal aneurysm into the lumbar vertebrae: report of a case.
Surg Today 29:1299–1300
abdominal aortic contained rupture aneurysm in patients 3. Grevitt MP, Fagg PS, Mulholland RC (1996) Chronic contained
surgically treated for an abdominal aortic aneurysm by a rupture of an aortic aneurysm mimicking infective spondylitis.
conventional open surgical repair. Other authors have Eur Spine J 5:128–130
reported vertebral lesions resulting from endovascular 4. Jang JH, Kim HS, Kim SW (2008) Severe vertebral erosion by
huge symptomatic pulsating aortic aneurysm. J Korean Neuro-
abdominal aneurysm repair complicated by an infection surg Soc 43:117–118
[8, 9]. To the best of our knowledge, this is the first report 5. Arici V, Rossi M, Bozzani A et al (2012) Massive vertebral
of a case in which an endovascular aneurysm repair destruction associated with chronic rupture of infrarenal aortic
(EVAR) for an abdominal aortic aneurysm was compli- aneurysm: case report and systematic review of the literature in
the English language. Spine 37(26):E1665–E1671
cated by an abdominal aortic false aneurysm which caused 6. Usselman JA, Vint VC, Kleiman SA (1979) CT diagnosis of
severe erosion of two lumbar vertebral bodies and disc aortic pseudoaneurysm causing vertebral erosion. Am J Roent-
through an inflammatory mechanism, without signs of genol 133:1177–1179
infective pathogenesis. It is possible to assume that, despite 7. Diekerhof CH, Reedt Dortland RW, Oner FC (2002) Severe
erosion of lumbar vertebral body because of abdominal aortic
the endovascular procedure, the pseudo-aneurysmatic sac false aneurysm: report of two cases. Spine 27:E382–E384
can cause an inflammatory stimulus that is erosive for the 8. Blanch M, Berjon J, Vila R et al (2010) The management of
adjacent vertebrae and discs. Pre-existing osteopenia, fre- aortic stent-graft infection: endograft removal versus conserva-
quently observed in old patients, can contribute to the tive treatment. Ann Vasc Surg 24:554–555
9. De Koning HD, van Sterkenburg SM, Pierie ME et al (2008)
development of the vertebral erosion. Endovascular abdominal aortic aneurysm repair complicated by
If the patient is in good general condition, an anterior spondylodiscitis and iliaco-enteral fistula. J Vasc Surg 47:
approach with the removal of the prosthesis [11] and 1330–1332
L2–L3 decompression and fusion should be considered. 10. Frankel HL, Hancock DO, Hyslop G et al (1969) The value of
postural reduction in the initial management of closed injuries of
We assume that when a lytic lesion of a lumbar vertebral the spine with paraplegia and tetraplegia. I.Parapleg 7:179–192
body or disc is discovered in a patient treated for an 11. Brinster CJ, Fairman RM, Woo EY, Wang GJ, Carpenter JP,
abdominal aortic aneurysm by endovascular repair, an Jackson BM (2011) Late open conversion and explantation of
abdominal aortic false aneurysm can be the cause of the abdominal aortic stent grafts. J Vasc Surg 54:42–46

33
Malunited extra-articular distal radius fractures: corrective
osteotomies using volar locking plate
Luigi Tarallo • Raffaele Mugnai • Roberto Adani •
Fabio Catani
Abstract other major complications, including non-union and

Background Multiple techniques for corrective osteot- infection, were observed.
omy have been developed in recent years with the same Conclusion The volar approach and locking plate, with-
aims: to improve the radiographic parameters and improve out necessarily the use of bone grafting, proved to be an
motion, pain and grip strength. Volar fixed-angle plates effective approach for addressing symptomatic and even
have added a new concept to the treatment of distal radius severe deformities of the distal radius.
fractures thanks to the low morbidity of the surgical Type of study/level of evidence Therapeutic IV
approach and the strength of the final construct, allowing
early mobilization and return to function. Keywords Malunion Fractures Radius Osteotomy
Materials and methods Between 2005 and 2012, 20 Volar Angulated Locking plate Bone graft DASH
patients with symptomatic dorsally malunited extra-artic- Mayo
ular fractures of the distal radius underwent corrective
osteotomy using a volar locking plate without additional
bone graft. At a mean follow-up of 50 months, all the Introduction
patients were clinically and functionally evaluated.
Results All measurements of pain, final range of motion Fractures of the distal radius are extremely common inju-
and grip strength significantly improved compared with ries and the outcome differs depending on the type of
preoperative measurements. The mean preoperative DASH fracture. Normally stable distal radius fracture is treated
score reduced from 54 points preoperatively to 25 post- non-operatively with a favorable result. In the other hand,
operatively. Based on the modified Mayo wrist score, we unstable fracture easily becomes malunited with inade-
obtained 14 excellent and six good results. Palmar tilt quate treatment. Malunion of the distal radius usually
improved from an average of 23° to 11°. Radial inclination occurs following conservative treatment. The most com-
improved from an average of 29° to 22°, and ulnar variance mon deformity following a malunited extra-articular frac-
decreased from an average of 3.6 mm to 0.9 mm. There ture of the distal radius is the loss of the normal palmar tilt
were two cases of transient median neuroapraxia that of the articular surface in the sagittal plane, and loss of
resolved before the 6-week follow-up appointment. No length relative to the ulna. Once angulation of the distal
articular surface of the radius becomes greater than 25–35°
in the sagittal plane, Fernandez recommended corrective
osteotomy [1, 2]. Several biomechanical studies have
L. Tarallo (&) R. Mugnai F. Catani
Department of Orthopaedic Surgery, University of Modena demonstrated abnormal wrist contact pressure with extra-
and Reggio Emilia, Via del Pozzo 71, 41124 Modena, Italy articular deformity that can predispose to arthritis [3]. An
e-mail: [email protected] increased contact pressure across the ulna is due to the loss
of radial height, especially when the articular surface of the
R. Adani
Department of Hand Surgery, University Hospital of Verona, radius has a dorsal angulation. The anatomical position of
Verona, Italy the distal end of the radius with respect to the carpus and

Malunited extra-articular distal radius fractures: corrective osteotomies using volar locking plate 231
the ulna head is the key to obtaining normal wrist biome- Materials and methods
chanics. A wrist with a normal motion has about 120° of
wrist flexion and extension, 50° of wrist radial and ulnar This prospective study was performed between 2005
deviation, and 150° of forearm rotation. The radius carries and 2012. Inclusion criteria were a malunion following
80 % of the axial load through the wrist, and the distal ulna conservative treatment, with a dorsal tilt of the distal
only 20 % [4]. Malalignment of the distal radius due to an radial articular surface of more than 20°, articular dis-
osseous deformity affects the normal load transmission, placement of more than 2 mm or incongruity of the
causing a limitation in the extension-flexion arc of motion. distal radio-ulnar joint due to shortening of the distal
Multiple techniques for corrective osteotomy have been radius in association with wrist pain and poor wrist
developed in recent years with the same aims: to improve range of movement. A total of 20 patients (8 women, 12
the radiographic parameters and improve motion, pain and men) with a mean age of 40 years (range 17–64) were
grip strength. Authors such as Fernandez [2] have descri- included in this study. The dominant arm was involved
bed the traditional treatment of osteotomy and dorsal in 13 patients and the non-dominant arm in seven.
plating with bone graft for dorsal angulated malunions. Anesthesia was obtained by axillary nerve block. In all
These techniques guarantee good restoration of the anat- patients, a volar approach to the distal radius was per-
omy and relieve pain, but have sometimes been associated formed. A longitudinal incision along the flexor radialis
with irritation or rupture of extensor tendons. Volar fixed- carpi was made. The radial artery was preserved and
angled plates have added a new approach to the treatment dislocated radially. The pronator quadratus was released
of distal radius fractures thanks to the low morbidity of the using an ‘‘L’’ incision from the radial insertion. After
surgical approach and the strength of the final construct, exposure of the volar margin of the distal radius, the
allowing early mobilization and return to function [5]. The distal portion of the DVRÒ plate (Hand Innovation) was
aim of this study is to report the outcomes of a cohort of held against the distal radius with K-wires. At this stage
patients affected by symptomatic dorsally malunited extra- fluoroscopy was necessary to identify the correct posi-
articular fractures of the distal radius who underwent cor- tioning of the plate on the volar surface of the radius
rective osteotomy using a volar locking plate without and for planning the level of the osteotomy. The plate
additional bone graft. was then removed after marking the position of the plate
Fig. 1 Case 1. a Preoperative range of movement evaluation. fluoroscopy. d X-rays 30 days after surgery showing the dorsal gap
b Preoperative X-rays showing the dorsal inclination of articular with early signs of refilling. e Good healing of the osteotomy with
surface of the distal radius. c Intraoperative view showing the complete refilling of the dorsal gap at 3 months after surgery
osteotomy of the radius, the synthesis and the final correction under

(mm) Mean (SD)

and the line of the osteotomy. Once the plate was
Ulnar variance
removed and the osteotomy was performed, the frag-
ments were distracted using a small osteotome as a lever
3.6 (0.3)
0.9 (0.6)
0.000b
to correct the deformity in lateral view under fluoros-
copy. The plate was then fixed to the radius on the distal
fragment, allowing the correct volar tilt and radial
Radial inclination
(mm) Mean (SD)
inclination. Once the plate was fixed on the distal radius
fragment, the oval hole of the plate was used to center
29.0 (6.9)
22.3 (6.1)
the proximal axis of the plate and a screw was inserted
0.000b
Radiographic evaluation
to hold the plate to the radial shaft. In this way the
preformed shape of the distal part of the plate helps to
correct the dorsal tilt, and under fluoroscopy the surgeon
Palmar tilt (mm)

can correct the radial inclination and radial lengthening.
Mean (SD)
23.1 (6.5)
11.3 (4.4)
The remaining cortical screws were inserted to complete
0.000b
the implant (Figs. 1, 2). The pronator quadratus was
sutured back into place, covering the plate. All the
surgical procedures were performed by a single surgeon.
Mean (SD)
46.9 (10.4)
91.7 (5.0)
All patients were placed in a palmar plaster splint for
0.000b
Functional outcome
Mayo
15 days before starting rehabilitation.
It is our routine practice to evaluate all patients clini-
cally at 2 weeks, and both clinically and radiologically at 1
Mean (SD)
53.8 (13.5)
25.3 (7.7)
and 3 months. Moreover, between June and October 2013
DASH
0.000b
all the patients were clinically reviewed. The clinical
assessment included the analysis of passive range of
motion (ROM), grip strength, pain level during activities of Mean (SD)
1.1 (1.3)
0.3 (0.6)
daily living evaluated with a 10-cm visual analogue scale
0.028a
VAS
Pain
(VAS), and functional evaluation using disabilities of the

arm, shoulder and hand (DASH) [6] and the Mayo elbow
Pronation (°)
performance score (MEPS). The total MEPS score ranges

Mean (SD)
75.3 (14.5)
84.0 (7.2)
from 5 to 100 points, with higher scores indicating better

0.000b
function. If the total score is between 90 and 100 points,

function can be considered excellent; between 75 and 89
Supination (°)
points, good; between 60 and 74 points, fair; less than 60

Mean (SD)
79.8 (15.5)
points, poor [7].

16.0 (7.5)
0.000b
Grip strength was measured with a Jamar dynamometer

(Asimov Engineering Corp., Santa Monica, CA, USA), and
wrist ROM using a goniometer.
Mean (SD)
44.8 (16.8)
60.3 (17.8)
Extension (°) Flexion (°)
All the radiographic measurements, including palmar

0.000b
tilt, radial inclination and ulnar variance, were performed

on the last follow-up X-rays using a picture archiving and
Range of motion
communication system (PACS software, Fuji Synapse).

Mean (SD)
39.3 (11.2)
Time of union was determined according to both radio-

70.3 (9.2)
logical and clinical parameters. Radiological criteria

0.000b
included: bridging of the fracture site by bone, callus or

trabeculae; bridging of the fracture seen at the cortices; and
Grip strength
Table 1 Summary statistics
obliteration of the fracture line or cortical continuity.

26.7 (10.5)
(kg) Mean
10.9 (2.9)
Clinical criteria were represented by the patient’s ability to

Paired samples t test
0.000b
(SD)
bear weight on the injured limb and perform activities of

daily living, and the presence of pain at the fracture site
Wilcoxon test
Postoperative
upon palpation and physical stress. Moreover, possible

Preoperative
early or late complications such as non-union, infection,

tendon rupture or tendon irritation, and nervous lesions
were documented.
P
b
a

Fig. 2 Case 2. a Range of motion evaluation before surgery with c Postoperative X-rays showing the correction of the dorsal deformity
significant reduction of flexion. b Preoperative X-rays showing the achieved. d Good functional recovery at 2 months after surgery
dorsal inclination of articular surface of the distal radius.
Results an average of 29° to 22°, and ulnar variance decreased

from an average of 3.6 mm to 0.9 mm. There were no
All osteotomy sites united at a median of 4 months (range intraoperative complications noted. There were two cases
2–5 months) after surgery. The mean duration of follow-up of transient median neuroapraxia that resolved before the
was 50 months (range 20–75 months). All measurements 6-week follow-up appointment. No other major complica-
of pain, final range of motion and grip strength significantly tions, including non-union and infection, were observed.
improved at the last follow-up compared with preoperative
measurements (Table 1). The mean preoperative DASH
score reduced from 54 points (range 27–70) preoperatively Discussion
to 25 (range 12–36) postoperatively.
Based on the modified Mayo wrist score, we obtained 14 Many studies have demonstrated that corrective osteotomy
excellent and six good results. Palmar tilt improved from which restores anatomical configuration can effect an
an average of 23° to 11°. Radial inclination improved from improvement in wrist function, forearm rotation, grip

strength and pain [8]. Usually, an opening wedge osteot- quality and with increased medical co-morbidities that may
omy using dorsal plates and bone grafting has been per- contraindicate the harvesting procedure, due to the longer
formed for malunited Colles’ fractures [9]. However, when operative time and the higher risks of bleeding and
dorsal plates are used, a high incidence of plate removal infection.
has been reported because of painful hardware, tendon
rupture and/or irritation [10–13]. There are several Conflict of interest None.
advantages to using a volar approach in the treatment of a Ethical standards The study was authorized by the local ethical
malunited Colles’ fracture. If the dorsal compartments are committee and was performed in accordance with the ethical stan-
not disturbed, the volar cortex can be fixed directly with a dards of the 1964 Declaration of Helsinki as revised in 2000. The
volar plate. Moreover, according to Malone et al. [14], the patients provided informed consent before being enrolled.
rigid characteristics of the volar locking plates are strong
enough to avoid the requirement of structural bone
grafting.
Donor site morbidity, especially at the iliac crest, has
been well described and minor complications such as
persistent pain at the harvest site, superficial sensory nerve
injury, superficial hematoma or seroma and superficial References
infection have been reported [15]. Moreover, a volar
1. Fernandez DL (1988) Radial osteotomy and Bowers arthroplasty
approach is easier than a dorsal approach and the reduction for malunited fractures of the distal end of the radius. J Bone
of the volar cortex is simple because of less comminution Joint Surg Am 70:1538–1551
and the advantage of direct vision [16]. The present study 2. Fernandez DL (1993) Malunion of the distal radius: current
showed that a corrective osteotomy using a volar locking approach to management. Instr Course Lect 42:99–113
3. Pogue DJ, Viegas SF, Patterson RM, Peterson PD, Jenkis DK,
plate without the use of bone grafting could effectively Sweo TD, Hokanson JA (1990) Effects of distal radius frac-
produce a significant improvement in wrist function in ture malunion on wrist joint mechanics. J Hand Surg Am
patients treated for extra-articular distal radius malunion. 15:721–727
We obtained an excellent correction of deformity based on 4. Werner FW, Glisson RR, Murphy DJ, Palmer AK (1986) Force
transmission through the distal radioulnar carpal joint: effect of
radiographic parameters, with low morbidity and no non- ulnar lengthening and shortening. Handchir Mikrochir Plast Chir
unions, hardware failure or need for hardware removal. Our 18:304–308
results are in line with those reported by Mahmoud et al. 5. Tarallo L, Mugnai R, Zambianchi F, Adani R, Catani F (2013)
[17], who treated 30 malunited dorsally-angulated radii Volar plate fixation for the treatment of distal radius fractures:
analysis of adverse events. J Orthop Trauma 27:740–745. doi:10.
using fixed-angle volar locking plates without bone graft- 1097/BOT.0b013e3182913fc5
ing, obtaining at a mean 18-month follow-up radiological 6. Hudak PL, Amadio PC, Bombardier C (1996) Development of an
evidence of union, correction of the deformity, and clinical upper extremity outcome measure: the DASH (disabilities of the
and functional improvement in all cases. In particular, the arm, shoulder and hand) [corrected]. The Upper Extremity Col-
laborative Group (UECG). Am J Ind Med 29:602–608
improvement in the DASH and Mayo scores obtained in 7. Morrey BF, An KN, Chao EYS (1993) Functional evaluation of
the present study was 28.5 and 42.8 points, respectively, the elbow. In: Morrey BF (ed) The Elbow and its Disorders, 2nd
compared with the 21.6 and 22.7 points reported by edn. W.B.Saunders, Philadelphia, pp 86–89
Mahmoud et al. [17]. These differences in functional out- 8. Jupiter JB, Ring D (1996) A comparison of early and late
reconstruction of malunited fractures of the distal end of the
come can probably be explained by the longer follow-up radius. J Bone Joint Surg Am 78:739–748
period of the present research. Favorable results have also 9. Fernandez DL (1982) Correction of post-traumatic wrist defor-
been reported in numerous studies following volar locking mity in adults by osteotomy, bone grafting, and internal fixation.
plates with additional bone graft [18, 19]. The volar J Bone Joint Surg Am 64:1164–1178
10. Schnur DP, Chang B (2000) Extensor tendon rupture after
approach and the use of locking plates is an extremely internal fixation of a distal radius fracture using a dorsally placed
effective and safe technique; in fact, the use of fixed-angle AO/ASIF titanium pi plate. Arbeitsgemeinschaft für Osteo-
locking plates reduces the risk of postoperative bone dis- synthesefragen/Association for the Study of Internal Fixation.
placement, and requires a shorter immobilization time [20, Ann Plast Surg 44:564–566
11. Simic PM, Robison J, Gardner MJ, Gelberman RH, Weiland AJ,
21]. Moreover, the mechanical strength provided by this Boyer MI (2006) Treatment of distal radius fractures with a low-
construct does not necessarily require the use of bone profile dorsal plating system: an outcomes assessment. J Hand
grafting. We therefore believe that the volar approach and Surg Am 31:382–386
locking plate, without necessarily the use of bone grafting, 12. Kamath AF, Zurakowski D, Day CS (2006) Low-profile dorsal
plating for dorsally angulated distal radius fractures: an outcomes
is an effective technique for addressing symptomatic and study. J Hand Surg Am 31:1061–1067
even severe deformities of the distal radius, and should be 13. Jupiter JB, Fernandez DL (2002) Complications following distal
preferred especially in elderly patients with poor bone radial fractures. Instr Course Lect 51:203–219

14. Malone KJ, Magnell TD, Freeman DC, Boyer MI, Placzek JD 18. Peterson B, Gajendran V, Szabo RM (2008) Corrective osteot-
(2006) Surgical correction of dorsally angulated distal radius omy for deformity of the distal radius using a volar locking plate.
malunions with fixed angle volar plating: a case series. J Hand Hand (N Y) 3:61–68. doi:10.1007/s11552-007-9066-y
Surg Am 31:366–372 19. Sato K, Nakamura T, Iwamoto T, Toyama Y, Ikegami H, Ta-
15. Arrington ED, Smith WJ, Chambers HG, Bucknell AL, Davino kayama S (2009) Corrective osteotomy for volarly malunited
NA (1996) Complications of iliac crest bone graft harvesting. distal radius fracture. J Hand Surg Am 34:27–33. doi:10.1016/j.
Clin Orthop Relat Res 329:300–309 jhsa.2008.09.018
16. Yasuda M, Masada K, Iwakiri K, Takeuchi E (2004) Early cor- 20. Orbay J (2005) Volar plate fixation of distal radius fractures.
rective osteotomy for a malunited Colles’ fracture using volar Hand Clin 21:347–354
approach and calcium phosphate bone cement: a case report. 21. Orbay JL, Badia A, Indriago IR, Infante A, Khouri RK, Gonzalez
J Hand Surg Am 29:1139–1142 E, Fernandez DL (2001) The extended flexor carpi radialis
17. Mahmoud M, El Shafie S, Kamal M (2012) Correction of dor- approach: a new perspective for the distal radius fracture. Tech
sally-malunited extra-articular distal radial fractures using volar Hand Up Extrem Surg 5:204–211
locked plates without bone grafting. J Bone Joint Surg Br
94:1090–1096. doi:10.1302/0301-620X.94B8.28646

34
Complication rates and reduction potential of palmar
versus dorsal locking plate osteosynthesis for the treatment
of distal radius fractures
F. Wichlas • N. P. Haas • A. Disch •
D. Machó • S. Tsitsilonis
Abstract Introduction
Background The aim of this study was to evaluate the
complication rates of volar versus dorsal locking plates and Over recent years an increase in the operative treatment
postoperative reduction potential after distal radius of distal radius fractures has been observed [1]. Despite
fractures. this increase and the high incidence of distal radius
Materials and methods For this study 285 distal radius fractures, several facts have not yet been fully eluci-
fractures (280 patients/59.4 % female) treated with locked dated, especially in terms of surgical approach and
plating were retrospectively evaluated. The mean age of complication rates. The biomechanical advantages of
the patients was 54.6 years (SD 17.4) and the mean follow- locking plates over the traditional plates have resulted in
up was 33.2 months (SD 17.2). The palmar approach was an increase of volar plating [2]. Volar plating is con-
used in 225 cases and the dorsal approach in 60 cases sidered to be a more straightforward surgical procedure,
(95 % type C fractures). which can result in anatomic reduction through indirect
Results Adequate reduction was achieved with both reduction techniques and plate manipulation; however,
approaches, regardless of fracture severity. In the dorsal dorsal articular fragments cannot be directly visualized
group, the complications and implant removal rates were and controlled. On the other hand, the dorsal approach is
significantly higher and the operative time was also longer. surgically more demanding and is thought to be associ-
Conclusions Based on these facts, we advocate the pal- ated with higher complication rates. Tendon ruptures or
mar locking plate for the vast majority of fractures. In cases tenosynovitis due to exposure of the tendons or implant-
of complex multifragmentary articular fractures where no associated soft-tissue irritation appears to be more com-
compromise in reduction is acceptable, and with the bio- mon after the dorsal approach [3]. On these grounds, the
mechanical equality of palmar and dorsal plating remaining introduction of the volar locking plate with the principle
unproven, dorsal plating may still be considered. of subchondral buttressing of the joint surface substan-
Level of evidence Therapeutic level IV. tially questioned the need for dorsal plating. However,
an achievement of anatomic reduction after volar plating
Keywords Distal radius fracture Locking plate is not thought to be always possible, especially in the
Approach Complication case of complex intra-articular ‘pilon radial’ fractures
with central depression fragments and extended dorsal
articular comminution. Under this scope, re-evaluation of
F. Wichlas N. P. Haas A. Disch D. Machó
the indications for volar versus dorsal plating is impor-
S. Tsitsilonis (&)
Center for Musculoskeletal Surgery, Charité,Universitätsmedizin tant when taking complication rates, fracture complexity,
Berlin, Augustenburger Platz 1, 13353 Berlin, Germany and individual patient characteristics into consideration.
e-mail: [email protected] The aim of the present study was the evaluation of
complication rates of volar and dorsal locking plate oste-
S. Tsitsilonis
Berlin-Brandenburg Center of Regenerative Therapies, osynthesis, as well as the evaluation of postoperative
Charité,Universitätsmedizin Berlin, Berlin, Germany radiological fracture reduction.

Complication rates and reduction potential of palmar versus dorsal locking plate osteosynthesis for the treatment... 237
Materials and methods Table 1 Fracture distribution in the study population and in the
subgroups according to the AO classification
For the needs of the present study all patients with distal Type of Group
radius fractures that were operatively treated with a locking fracture
All (n = 285) Volar (n = 225) Dorsal (n = 60)
plate (2.4 and 3.5 mm Locking Compression Plate (LCP), (%) (%) (%)
SynthesÒ, Oberdorf, Switzerland) over a 3-year period
(2005–2007) were included and retrospectively evaluated. A 89 (31.2) 82 (36.4) 7 (11.6)
All patients gave informed consent prior to being included A2 11 (3.8) 11 (4.9) 0 (0)
in the study. The study was authorized by the local ethical A3 78 (27.4) 71 (31.5) 7 (11.6)
committee (EA2/075/11) and was performed in accordance B 16 (5.6) 16 (7.1) 0 (0)
with the Ethical standards of the 1964 Declaration of Hel- B2 8 (2.8) 8 (3.5) 0 (0)
sinki as revised in 2000. An electronic ICD-9 search was B3 8 (2.8) 8 (3.5) 0 (0)
conducted and 285 distal radius fractures (280 patients) C 180 (63.2) 123 (56) 57 (95)
treated with an LCP were identified. The mean age was C1 41 (14.4) 37 (16.4) 4 (6.7)
54.6 years (SD 17.4), and the majority of patients were C2 70 (24.5) 49 (21.8) 21 (35)
female [116 male (40.6 %)/169 female (59.4 %]. The mean C3 69 (24.3) 37 (16.4) 32 (53.3)
follow-up time was 33.2 months (SD 17.2). A 2.4-mm LCP The majority of the fractures were type C. In the dorsal group more
was used in 192 cases (67.4 %) and a 3.5-mm LCP in 93 than half were type C3 fractures
cases (32.6 %). The mechanism of injury in the majority of
the cases was a fall from standing height (172 cases,
60.5 %). The remaining fractures were caused by sports expressed as percentages (%). The Kolmogorov–Smirnov
activities (54 cases, 18.9 %), fall from a greater height (29 test was used in order to assess distribution normality. For
cases, 10.1 %), motor vehicle accident (24 cases, 8.3 %), parametric variables, the Student t test was used for the
and polytrauma (6 cases, 2.1 %). A palmar approach was comparison of two groups; for non-parametric variables the
used in 225 cases and a dorsal approach in 60 cases. The Mann–Whitney test was implemented. Differences for
dorsal approach was used for fractures with a central categorical variables were assessed with the chi-squared
articular depression or which had dorsal joint fragments that test or Fisher’s exact test. Correlations were examined with
were not considered amendable through a palmar approach. either Pearson product moment correlation coefficient or
All patients were operated under general anesthesia and Spearman’s rank correlation coefficient. Differences were
operative steps were fluoroscopically controlled under an considered statistically significant if the null hypothesis
image intensifier. A perioperative single-shot antibiosis was could be rejected with [95 % confidence (p \ 0.05).
given and a pneumatic tourniquet was used. The palmar
approach was located over the flexor carpi radialis tendon
and the dorsal approach located over the third extensor Results
tendon sheath. For the dorsal approach, the retinaculum was
opened in a z-shaped way right above the third extensor The fracture distribution according to the AO classification
tendon sheath and the extensor pollicis longus tendon (EPL) is shown in Table 1. No statistically significant age dif-
was released. An epiperiosteal preparation was conducted ference existed between the two groups [mean age of volar
medially and laterally. The second plate was placed group 55.4 years (SD 18.0); mean age of dorsal group 50.7
between the first and second extensor sheath radially. Both years (SD 16.3) (p = 0.068)]. The dorsal approach group
approaches are described in detail elsewhere [4]. All frac- consisted of 95 % (57 fractures) type C fractures, with
tures except for nine were closed. The evaluated data were more than half being (53.3 %) complex C3 fractures. The
fracture classification according to AO, mechanism of mean operative time for the volar plating group was 97.3
injury, operative time, type of implant, peri- and postoper- (SD 42.5) min and 123.7 (SD 49.3) min for the dorsal
ative complications and the need for an implant removal. group. This difference was statistically significant
Fractures were further subdivided into volar and dorsal (p \ 0.001) (Fig. 1).
plate osteosynthesis groups. Fracture reduction was asses- The preoperative radial inclination for the whole popu-
sed using radial inclination, palmar tilt, and ulnar variance lation was 15.2° (SD 9.2°) and the volar tilt was -13.0°
in posteroanterior (PA) and lateral radiographs according to (SD 17.7°). The preoperative ulna variance was 1.39 mm
the criteria defined by Kreder et al. [5]. These values were (SD 2.96 mm). The postoperative values were 22.1° (SD
measured pre- and postoperatively. 4.8°) for radial inclination and 8.6° (SD 6.4°) for volar
Continuous variables were expressed as mean ± stan- tilt; ulnar variance was -0.35 mm (SD 1.95 mm). The
dard deviation (SD), whereas categorical variables were assessment of the reduction was further analyzed

Fig. 1 Postoperative x-rays of a dorsal (AO 23 C3) (left) and volar locking plate osteosynthesis (AO 23 A3) (right)
Table 2 Pre- and postoperative radiological parameters

Reduction parameters Groups
Pre-operative palmar Pre-operative dorsal Post-operative palmar Post-operative dorsal
Radial inclination 15.1° (SD 8.7°) 15.7° (SD 10.8) 22.3° (SD 4.7°) 21.1° (SD 5°)
p NS 0.044
Volar tilt -13.4° (SD 1.2°) -12.8° (SD 2.1°) 8.1° (SD 6.3°) 10.1° (SD 6.4°)
p NS 0.01
Ulnar variance (mm) 1.63 (SD 2.72) 0.88 (SD 3.12) -0.2 (SD 1.9) -0.8 (SD 2.3)
p 0.001 0.001
separately for the volar and the dorsal groups (Table 2). Table 3 Complication rates in the study population and in the
The difference in the postoperative reduction of the frac- subgroups
tures between the two groups was statistically significant in Complications (n = 18/6.3 %) Groups
both planes, with palmar plating achieving better results for
Palmar (n = 8) Dorsal (n = 10)
radial inclination, and dorsal plating for palmar tilt and
ulnar variance. However, the absolute difference was no Pain/swelling 5 5
more than two degrees; a nearly anatomic reduction was Tenosynovitis 0 2
achieved for both approaches. The observed statistically EPL rupture 0 1
significant difference in the palmar tilt between the two Intra-articular screw 0 1
groups remained even after comparison of type C fractures Fragment displacement 1 0
only (volar tilt: palmar group (n = 112), 7.7° (SD 6.6°)/ Carpal tunnel syndrome 1 0
dorsal group (n = 52), 10.5° (SD 6.2°); p = 0.011). In the Re-fracture 1 0
comparison of the radial inclination of type C fractures CRPS 0 1
only, the difference between the two groups was no longer Incidence 3.6 % 16.7 %
statistically significant. However, a tendency to higher p \0.001
values in the palmar group was observed (palmar group
(n = 112): 22.4° (SD 4.6°)/dorsal group (n = 52): 20.9°
(SD 4.7°); p = 0.055). The mean operative time remained fracture severity and radial inclination was observed
statistically significantly longer for the dorsal group, even (p = 0.004; Spearman’s q -0.376).
in the comparison of the AO type C fractures only (palmar Eighteen complications were recorded overall (Table 3).
group: 105.2 min (SD 49.5 min)/dorsal group: 122.6 min In the majority of cases (13/18) the complications occurred
(SD 47.3 min); p = 0.034). In the palmar group no cor- in type C fractures. In the palmar group, eight complica-
relation was seen between fracture severity according to tions occurred (3.6 %), while in the dorsal group the
AO classification and postoperative radiological outcome. incidence was higher (ten cases, 16.7 %). The difference in
In the dorsal group a weak negative correlation between the incidence was statistically significant (p \ 0.001). This

difference between the two groups remained statistically fractures only; however, this could be attributed to the
significant even after comparison of type C fractures only more demanding surgical technique of dorsal plating with
(p \ 0.001). No significant difference was observed in possible devascularization of soft tissues and bony struc-
complication rates with regard to plate type (2.4 mm/ tures, as well as the iatrogenic tendon injury with the
3.5 mm). addition of longer operative time. Additionally, the posi-
Implant removal was performed in 25 cases in the tioning of dorsal plates right under the tendon sheaths can
overall study population (8.8 %)—15 were performed in further irritate the tendons postoperatively and lead to
the palmar group (6.7 %) and 10 in the dorsal group implant-associated pain. While implant removal rates in
(16.7 %). The difference in the incidence was statistically the dorsal group were also significantly higher, it was
significant (p \ 0.01). The indication for implant removal interesting that in almost half of the cases, implant removal
was implant-associated problems (pain or persistent was initiated by the patients themselves, even in the
swelling located above the plate) in ten cases (five in the absence of objective impairment. The problem of foreign
volar group, five in the dorsal group), as well as tenosyn- body feeling has not yet been overcome, even after plate
ovitis of the EPL tendon in two cases and one intra-artic- design optimization [15]. We generally do not advocate an
ular screw in the dorsal group. In the remaining cases the implant removal unless hardware-associated tendon
implant removal was initiated after patient request. pathology or functional impairment is present.
The high incidence of tendon ruptures after locked
plating reported in the literature, even after palmar osteo-
Discussion synthesis due to oversized screws, was not confirmed in our
study. This is in accordance with other studies [16]. In most
As the trend currently leads towards palmar plating [6], the cases of volar plating, tendon irritations seem to derive from
need for dorsal plating is fundamentally questioned. technical errors and oversized screws [17]. The problem of
Nowadays, [30 different types of locking plates are oversized screws may derive from the traditional idea that
available on the world market, with most of them being bicortical screw purchase is needed for plate fixation; this is
palmar plates. Novel implants with more screw placement not the case for internal fixators such as locking plates. As a
modalities have been introduced; however, the importance recent biomechanical study showed, a screw length of 75 %
of such features remains unconfirmed [7]. of the anteroposterior cortical distance can result in suffi-
In the present study, the postoperative reduction of radial cient buttressing of the joint surface [18]. Nevertheless, if
inclination, palmar tilt and ulnar variance in both groups was dorsal key fragments need to be fixed, meticulous fluoro-
almost anatomic; this was also seen in previous studies [8, 9]. scopical control using dynamic and dorsal tangential views
The absolute value of the observed statistically significant can avoid screw oversizing [19, 20]. Tenosynovitis of the
difference between the two groups postoperatively was flexor was not observed in the present study; however, this
minimal. Radial inclination seems to be better reduced could be attributed to the the smaller plate profile and its
through a palmar approach; the observed negative correla- shape variety (L-, T-plates) with implant placement proxi-
tion between fracture severity according to AO and radial mal to the watershed-line [21–23].
inclination in the dorsal group underlines this fact. However, The main advantage of dorsal plating is the fact that
palmar tilt and ulnar variance were better restored through a centrally depressed and dorsal articular fragments can be
dorsal approach. The observed differences between the directly addressed and anatomically reduced; this is not
groups remained, even after comparison of type C fractures possible through a palmar approach, at least not to that
only. This fact underlines the above-mentioned differences extent. This point finds its implementation mainly in the
in the surgical outcome between the two approaches. treatment of complex multifragmentary intra-articular type
Nonetheless, it is questionable whether such small absolute C3 fractures, or of special fracture types, such as Barton
differences are of clinical relevance. fractures. The question that arises is whether a perfect
The reported complication rates of palmar versus dorsal reduction is needed in every case, especially if it could be
locking plates in the literature remain contradictory. While associated with higher complication rates. It has been
several studies report higher complication rates after pal- shown that in older patients a certain degree of loss of
mar locked plating [9–11], others show no difference anatomic reduction can be tolerated to a certain extent,
between the two approaches [12, 13]. Making the situation without affecting the subjective final outcome [24, 25]. For
even more confusing, other studies report higher compli- younger patients, however, this may not be the case and
cation rates after dorsal plating [3, 14]. In the present study, until proven otherwise, an anatomic reduction in order to
the complication rates encountered in the dorsal group minimize the risk of post-traumatic arthritis should be one
were significantly higher. This difference remained statis- of the main goals of operative treatment in that patient
tically significant even after comparison of the type C group.

The final aspect that should be taken into consideration 4. Jupiter JB, Marent-Huber M (2010) Operative management of
is the biomechanical behavior of different types of plates. distal radial fractures with 2.4-millimeter locking plates: a mul-
ticenter prospective case series. Surgical technique. J Bone Joint
Several studies have confirmed the biomechanical superi- Surg Am 92((Suppl 1 Pt 1)):96–106. doi:10.2106/jbjs.i.01340
ority of locking plates over conventional plates [8, 26]. 5. Kreder HJ, Hanel DP, McKee M, Jupiter J, McGillivary G,
This has also contributed to the increasing number of Swiontkowski MF (1996) X-ray film measurements for healed
fractures treated with palmar locking plates. However, the distal radius fractures. J Hand Surg Am 21(1):31–39. doi:10.
1016/s0363-5023(96)80151-1
biomechanical equality of palmar versus dorsal locking 6. Maschke SD, Evans PJ, Schub D, Drake R, Lawton JN (2007)
plates still remains debatable. While several studies show Radiographic evaluation of dorsal screw penetration after volar
no biomechanical differences between the two implants fixed-angle plating of the distal radius: a cadaveric study. Hand
[27], others still advocate the biomechanical superiority of (N Y) 2(3):144–150. doi:10.1007/s11552-007-9038-2
7. Drobetz H, Schueller M, Tschegg EK, Heal C, Redl H, Muller R
the dorsal plates, which are supposed to be stiffer and (2011) Influence of screw diameter and number on reduction loss
stronger [28]. As long as no undisputable proof of the after plating of distal radius fractures. ANZ J Surg 81(1–2):
biomechanical equality between palmar and dorsal locking 46–51. doi:10.1111/j.1445-2197.2010.05479.x
plates exists, the use of dorsal locking implants for the 8. Konstantinidis L, Helwig P, Strohm PC, Hirschmuller A, Kron P,
Sudkamp NP (2010) Clinical and radiological outcomes after
treatment of fractures in high risk for secondary loss of stabilisation of complex intra-articular fractures of the distal
reduction may be taken into consideration. radius with the volar 2.4 mm LCP. Arch Orthop Trauma Surg
In conclusion, the present study showed that regardless 130(6):751–757. doi:10.1007/s00402-009-0990-x
of fracture severity, an adequate reduction of distal radius 9. Matschke S, Wentzensen A, Ring D, Marent-Huber M, Audige L,
Jupiter JB (2011) Comparison of angle stable plate fixation
fractures is possible through both surgical approaches in approaches for distal radius fractures. Injury 42(4):385–392.
the vast majority of the cases. The higher complication and doi:10.1016/j.injury.2010.10.010
implant removal rates of dorsal locking plates, as well as 10. Knight D, Hajducka C, Will E, McQueen M (2010) Locked volar
the longer operative time needed, are factors in favor of plating for unstable distal radial fractures: clinical and radiolog-
ical outcomes. Injury 41(2):184–189. doi:10.1016/j.injury.2009.
palmar locking plates; therefore, we advocate its use for the 08.024
vast majority of fractures. However, in cases of complex 11. Yu YR, Makhni MC, Tabrizi S, Rozental TD, Mundanthanam G,
multifragmentary articular fractures, where no compromise Day CS (2011) Complications of low-profile dorsal versus volar
in postoperative reduction can be accepted, and as long as locking plates in the distal radius: a comparative study. J Hand
Surg Am 36(7):1135–1141. doi:10.1016/j.jhsa.2011.04.004
the undisputable biomechanical equality of palmar and 12. Zettl RP, Clauberg E, Nast-Kolb D, Ruchholtz S, Kuhne CA
dorsal plating remains unproven, dorsal locking plates can (2009) Volar locking compression plating versus dorsal plating
still be considered as a treatment option in special cases. for fractures of the distal radius: a prospective, randomized study.
Unfallchirurg 112(8):712–718. doi:10.1007/s00113-008-1526-5
Conflict of interest None. 13. Chou YC, Chen AC, Chen CY, Hsu YH, Wu CC (2011) Dorsal
and volar 2.4-mm titanium locking plate fixation for AO type C3
Ethical standards The present study has been approved by the dorsally comminuted distal radius fractures. J Hand Surg Am
appropriate ethics committee and has therefore been performed in 36(6):974–981. doi:10.1016/j.jhsa.2011.02.024
accordance with the ethical standards laid down in the 1964 Decla- 14. Arora R, Lutz M, Zimmermann R, Krappinger D, Gabl M, Pec-
ration of Helsinki and its later amendments. All persons gave their hlaner S (2007) Limits of palmar locking-plate osteosynthesis of
informed consent prior to their inclusion in the study. Details that unstable distal radius fractures. Handchir Mikrochir Plast Chir
might disclose the identity of the subjects under study were omitted. 39(1):34–41. doi:10.1055/s-2007-964922
15. Kwan K, Lau TW, Leung F (2011) Operative treatment of distal
radial fractures with locking plate system-a prospective study. Int
Orthop 35(3):389–394. doi:10.1007/s00264-010-0974-z
16. Hakimi M, Jungbluth P, Windolf J, Wild M (2010) Functional
results and complications following locking palmar plating on the
distal radius: a retrospective study. J Hand Surg Eur 35(4):
283–288. doi:10.1177/1753193409339943
17. Tarallo L, Mugnai R, Zambianchi F, Adani R, Catani F (2013)
References Volar plate fixation for the treatment of distal radius fractures:
analysis of adverse events. J Orthop Trauma 27(12):740–745.
1. Smektala R, Endres HG, Dasch B, Bonnaire F, Trampisch HJ, doi:10.1097/BOT.0b013e3182913fc5
Pientka L (2009) Quality of care after distal radius fracture in 18. Wall LB, Brodt MD, Silva MJ, Boyer MI, Calfee RP (2012) The
Germany. Results of a fracture register of 1,201 elderly patients. effects of screw length on stability of simulated osteoporotic
Unfallchirurg 112(1):46–54. doi:10.1007/s00113-008-1523-8 distal radius fractures fixed with volar locking plates. J Hand Surg
2. Liporace FA, Adams MR, Capo JT, Koval KJ (2009) Distal Am 37(3):446–453. doi:10.1016/j.jhsa.2011.12.013
radius fractures. J Orthop Trauma 23(10):739–748. doi:10.1097/ 19. Sugun TS, Karabay N, Gurbuz Y, Ozaksar K, Toros T, Kayalar M
BOT.0b013e3181ba46d3 (2011) Screw prominences related to palmar locking plating of
3. Rein S, Schikore H, Schneiders W, Amlang M, Zwipp H (2007) distal radius. J Hand Surg Eur 36(4):320–324. doi:10.1177/
Results of dorsal or volar plate fixation of AO type C3 distal 1753193410392869
radius fractures: a retrospective study. J Hand Surg Am 20. Ozer K, Toker S (2011) Dorsal tangential view of the wrist to
32(7):954–961. doi:10.1016/j.jhsa.2007.05.008 detect screw penetration to the dorsal cortex of the distal radius

after volar fixed-angle plating. Hand (N Y) 6(2):190–193. doi:10. 25. Egol KA, Walsh M, Romo-Cardoso S, Dorsky S, Paksima N
1007/s11552-010-9316-2 (2010) Distal radial fractures in the elderly: operative compared
21. Jupiter JB, Marent-Huber M (2009) Operative management of with nonoperative treatment. J Bone Joint Surg Am 92(9):
distal radial fractures with 2.4-millimeter locking plates. A 1851–1857. doi:10.2106/jbjs.i.00968
multicenter prospective case series. J Bone Joint Surg Am 26. Levin SM, Nelson CO, Botts JD, Teplitz GA, Kwon Y, Serra-Hsu
91(1):55–65. doi:10.2106/jbjs.g.01498 F (2008) Biomechanical evaluation of volar locking plates for
22. Soong M, van Leerdam R, Guitton TG, Got C, Katarincic J, Ring distal radius fractures. Hand (N Y) 3(1):55–60. doi:10.1007/
D (2011) Fracture of the distal radius: risk factors for compli- s11552-007-9063-1
cations after locked volar plate fixation. J Hand Surg Am 27. McCall TA, Conrad B, Badman B, Wright T (2007) Volar versus
36(1):3–9. doi:10.1016/j.jhsa.2010.09.033 dorsal fixed-angle fixation of dorsally unstable extra-articular
23. Asadollahi S, Keith PP (2013) Flexor tendon injuries following distal radius fractures: a biomechanic study. J Hand Surg Am
plate fixation of distal radius fractures: a systematic review of the 32(6):806–812. doi:10.1016/j.jhsa.2007.04.016
literature. J Orthop Traumatol 14(4):227–234. doi:10.1007/s10195- 28. Blythe M, Stoffel K, Jarrett P, Kuster M (2006) Volar versus
013-0245-z dorsal locking plates with and without radial styloid locking
24. Gruber G, Zacherl M, Giessauf C, Glehr M, Fuerst F, Liebmann plates for the fixation of dorsally comminuted distal radius
W, Gruber K, Bernhardt GA (2010) Quality of life after volar fractures: A biomechanical study in cadavers. J Hand Surg Am
plate fixation of articular fractures of the distal part of the radius. 31(10):1587–1593. doi:10.1016/j.jhsa.2006.09.011
J Bone Joint Surg Am 92(5):1170–1178. doi:10.2106/jbjs.i.00737

35
Non-operative treatment versus percutaneous fixation
for minimally displaced scaphoid waist fractures in high demand
young manual workers
Haroon Majeed
Abstract However, it was difficult to assess the outcomes of mini-

Background Managing minimally displaced scaphoid mally displaced fractures in isolation. Furthermore, few of
fractures in young individuals doing physically demanding these studies relied on plain radiographs for assessing
work remains an issue of debate due to duration of union and did not report on patients’ work status.
immobilisation and time required off work. Therefore, Conclusion Cast treatment has the disadvantages of
early diagnosis and appropriate treatment are important to longer immobilisation time, joint stiffness, reduced grip
avoid short- and long-term consequences. The literature strength, and longer time to return to manual work. Per-
lacks the exact definition of minimally displaced scaphoid cutaneous fixation is aimed at reducing damage to the
waist fractures. The objective of this review article was to blood supply and soft tissues, allowing early mobilisation
discuss nonoperative and minimally invasive treatment of the wrist and early return to manual work. The best
(percutaneous screw fixation) for minimally displaced available evidence for percutaneous screw fixation versus
scaphoid waist fractures and to systematically review the cast treatment suggests that percutaneous fixation allows a
literature, focussing on young workers with physically faster time to union by 5 weeks and an earlier return to
demanding employment. manual work by 7 weeks, with similar union rates. This
Materials and methods We searched for articles through systematic review indicates a potential requirement for a
the most commonly used portals using appropriate termi- prospective randomised controlled trial to compare these
nologies to identify the most relevant articles in the English two treatment modalities for minimally displaced scaphoid
language comparing nonoperative and percutaneous fixa- waist fractures in workers with physically demanding jobs
tion methods for these fractures in patients between 16 and in order to objectively assess functional outcomes, time to
40 years of age. Strict inclusion and exclusion criteria were union and time to return to work.
observed. Level of evidence Level 3.
Results Sixty relevant published articles were found.
Twenty-one of these were considered valid for inclusion Keywords Minimally displaced scaphoid fractures
and comprised five randomised controlled trials, three Percutaneous fixation Cast treatment Scaphoid waist
prospective studies, four systematic reviews, three meta- fractures
analyses, and six retrospective studies. These studies pro-
vided a reasonable account of information on the managing
undisplaced and minimally displaced scaphoid waist frac- Introduction
tures, with satisfactory clinical and statistical analysis.
Scaphoid fractures account for 50–80 % of all carpal bone
fractures in young and active individuals. Managing sca-
H. Majeed (&) phoid fractures remains an issue of debate because of the
Trauma and Orthopaedics, University Hospital of North
potential risk of delayed union and nonunion. Early diag-
Staffordshire, Newcastle Road, Stoke-on-Trent ST4 6QB,
England, UK nosis and appropriate treatment are important in order to
e-mail: [email protected] avoid avascular necrosis, arthritis and carpal collapse [1].

Non-operative treatment versus percutaneous fixation for minimally displaced scaphoid waist fractures... 243
In young individuals employed in physically demanding patients in each study. All those studies were excluded,
work, it is even more challenging because of the duration which were based on cadavers, or reported the outcomes of
required for immobilisation and thus time required off nonunited fractures, or late-diagnosed fractures, or open
work. Managing scaphoid fractures varies among hospitals fixation methods for acute fractures or distal or proximal
and depends upon local preferences and protocols. How- pole fractures. Duplicate studies were identified and
ever, as a general principle, management involves bal- excluded. Studies published only in English or with Eng-
ancing risk level based on available evidence [2]. lish translation were selected.
Nonoperative treatment is widely accepted and advocated
for acute, undisplaced scaphoid waist fractures [3], and
screw fixation (percutaneous or open) has become an Results
acceptable method for treating displaced fractures [4]. How
to best manage minimally displaced scaphoid waist frac- Based on the above search methodology, 60 relevant
tures remains unclear. Displaced fractures have been published articles were found. Twenty-one were consid-
described in the literature with fracture gap[1 mm [5], but ered valid for inclusion. The selected articles consisted of
the exact description of minimally displaced fracture is not five randomised controlled trials, three prospective studies,
available in the literature. Therefore, we consider a mini- four systematic reviews, three meta-analyses and six ret-
mally displaced fracture as one with \1-mm gap. rospective studies (Fig. 1).
Assessing union may be difficult on plain radiographs Adolfsson et al. [7] compared outcomes of percutaneous
because the scaphoid is composed of [80 % cartilage and fixation (Acutrak screws) with immobilisation in a long
therefore does not develop callus. Radiographic consoli- scaphoid cast in a randomised controlled trial. Fifty-three
dation is often delayed compared with clinical consolida- patients, mean age 31 years) with undisplaced and mini-
tion [6]. The objective of this paper is to discuss mally displaced acute scaphoid waist fractures were
nonoperative and minimally invasive treatment (percuta- recruited. Fixation group consisted of 25 patients and cast
neous screw fixation) for minimally displaced scaphoid group 28 patients. Cast group was immobilised for
waist fractures and systematically review the available 10 weeks; the fixation group was immobilised with a cast
literature, keeping the focus on young individuals for 3 weeks and a removable splint for a further 3 weeks.
employed in physically demanding work. Computed tomography (CT) scan was used to confirm
union in both groups. Results showed a significantly better
range of motion (ROM) in the fixation group (p \ 0.02)
Materials and methods but no differences in union rate or grip strength.
An RCT by Bond et al. [8] compared percutaneous
Articles were sourced from MEDLINE through and Pub- fixation (Acutrak screws) with immobilisation in a long
Med (1970–2013), Embase (1980–2013), Cochrane con- scaphoid cast. Twenty-five military personnel with acute
trolled trials register electronic databases, Thomson undisplaced and minimally displaced scaphoid waist frac-
Scientific Web of Science (1993–2013) and Elsevier Sco- tures were recruited, with an average age of 24 years. The
pus. Primary search terms were scaphoid waist fractures fixation group had 11 and the cast group 14 patients. The
with minimally displaced or cast immobilization or plaster cast group was immobilised in a long cast for 6 weeks,
or minimally invasive surgery or percutaneous surgery. All followed by a short cast until union was achieved, which
types of studies were included in our initial search and final was identified clinically and radiographically. The fixation
selection and included randomised controlled trials group was immobilised with a short cast for 10 days and a
(RCTs), prospective studies, systematic reviews, meta- removable splint until union. All patients were followed up
analyses, retrospective studies and case-series reviews. for 25 months, and all fractures achieved union. Analysis
Abstracts of the relevant searched articles were screened showed significant reduction in time to union in the fixation
first to assess their validity for inclusion. If satisfactory, group (7 vs. 12 weeks, p = 0.0003) and time to return to
then the full-text articles were obtained through online full duty in the fixation group (8 vs. 15 weeks,
access or manual search through the library access. Studies p = 0.0001). McQueen et al. [9], in their RCT, compared
included in the final selection reviewed outcomes of non- percutaneous fixation (Acutrak screws) with immobilisa-
operative treatment for minimally displaced fractures or tion in a Colles’ cast, with the thumb out of cast. Sixty
percutaneous fixation techniques or direct comparison of patients with acute undisplaced and minimally displaced
these two modalities of treatment for this specific group of scaphoid waist fractures, with a mean age of 27 years, were
fractures in young individuals. Age criteria were set to recruited. Thirty patients were allocated to each group after
include results of patients with mean age between 16 and randomisation. The cast group was immobilised for
40 years. There was no minimum or maximum number of 8–12 weeks’ the fixation group was mobilised immediately

Fig. 1 Selection of studies
with physiotherapy. Clinical and radiological assessments results in 30 patients with acute undisplaced scaphoid waist
were performed up to 52 weeks, with outcomes being fractures fixed with Herbert screws. Average age was
measured by a blinded assessor. Analysis showed a sig- 31 years and average follow-up was 41 months. Average
nificant reduction in time to union in the fixation group (9.2 time of cast immobilisation was 3 weeks. ROM was
vs. 13.9 weeks, p \ 0.001). There were also significant assessed using a goniometer and grip strength using a
reductions in the fixation group in times to return to manual dynamometer. Inoue and Shionoya, [13], in a prospective
work (3.8 vs. 11.4 weeks, p \ 0.001) and sport (6.4 vs. study, compared 46 acute scaphoid waist fractures (mini-
15.5 weeks, p \ 0.001). However, there were no signifi- mally displaced and undisplaced) treated with percutane-
cant differences in functional outcomes or union rates ous fixation and 42 fractures treated nonoperatively using
between groups at final follow-up. below-elbow cast, which included thumb. Average age was
Drac and Manak [10] performed a prospective case– 26.5 years. Patients were given the choice of either treat-
control study, with 38 patients in the percutaneous fixation ment. Average follow-up was 10 months. The authors
group and 34 in the cast (control) group for acute, undis- found significantly quicker return to manual work (5.8 vs.
placed and minimally displaced waist fractures. Average 10.2 weeks, respectively; p \ 0.001) and union rate (6 vs.
age was 27 years and minimum follow-up 12 months. Cast 9.7 weeks, respectively; p \ 0.001) in the percutaneous
group had significantly more nonunions (p = 0.024) and fixation group than in the cast group. All fractures united in
restricted ROM and grip strength (p \ 0.0001). Union was the percutaneous fixation group; there was one nonunion in
assessed with computed tomography (CT) scan in all the cast group. A similar study by De Vos et al. [14]
patients in the percutaneous fixation group, making union reported a 97 % union rate after percutaneous fixation of
assessment more reliable, but CT was not done for patients acute scaphoid waist fractures using noncannulated Herbert
in the cast group. screws. This series had 44 patients, including 31 heavy
Haddad and Goddard [11] reported 100 % union rate in manual labourers, with average age of 31 years. Average
50 acute scaphoid waist fractures treated and found an time to union was 6.4 weeks and return to manual work
average duration of 55 days after percutaneous screw fix- 41 days. ROM was 97 %, and power grip and pinch grip
ation. Patients had an average age of 26 years. They were were 96 % each compared with the contralateral side.
allowed full mobilisation immediately after surgery. ROM For nonoperative management of acute scaphoid waist
was equal to contralateral thumb and grip strength (98 %) fractures, different types of casts are used in routine
compared with contralateral thumb at 3 month. All patients practice. These include Colles’ cast with wrist in flexion or
returned to manual work within 5 weeks. Brutus and Be- extension; scaphoid cast below or above elbow; scaphoid
aton [12] reported 90 % union rate and good functional cast including or excluding the thumb. Rhemrev et al. [6],

in their retrospective study, showed that 81.7 % (58/71) of blood supply and soft tissue, allowing early mobilisation of
undisplaced and minimally displaced scaphoid waist frac- the wrist and early return to manual work. It can be per-
tures achieved clinical and/or radiographic union during formed through either the volar or dorsal approach—the
6 weeks of cast immobilisation; another 15.5 % (11/71) former being more popular because of better clinical out-
required an additional 2 weeks of cast immobilisation. comes, easier access and fewer reported complications
Their overall results showed 97 % union rates within [22].
8 weeks. Clay et al. [5] compared union rates in Colles’ For percutaneous screw fixation, among all studies
and scaphoid casts in a prospective randomised trial, with reviewed, there was a total cohort of 274 patients, with an
148 patients in the Colles’ cast group and 143 in the sca- average age of 27.8 years. Union rate was 98.5 %, with an
phoid cast group. No significant difference in union rates average time to union of 46 days and average time to return
were found (p = 0.92). Gellman et al. [15] compared long to manual work of 40 days. Among RCTs comparing the
thumb-spica cast with short thumb-spica cast in a pro- two treatment modalities, the RCT performed by Adolfsson
spective randomised controlled trial. They treated 28 et al. [7] was a poor-quality study providing level 2b evi-
patients in long thumb-spica cast for 6 weeks, followed by dence. They did not report the method of randomisation
a short thumb-spica cast for another 6 weeks; 23 patients and lacked power calculation. Furthermore, a paucity of
were treated with short thumb-spica cast throughout the demographic data did not enable a fair comparison between
duration of treatment. Significantly shorter time to union groups; 25 % of patients were lost to follow-up or excluded
was seen (p \ 0.05) in the long thumb-spica group (9.5 vs. from analysis, and outcome assessors were not blinded,
12.7 weeks). Alho et al. [16] found no significant differ- which may have been a source of bias. However, the RCT
ence in fracture healing in their prospective study on 100 reported by Bond et al. [8] was a high-quality study pro-
patients. They compared above- and below-elbow cast viding level 1b evidence. There were clear inclusion and
immobilisation in a nonrandomised study, with a good exclusion criteria, with adequate study power and 100 %
number of patients who were alternated in each group. follow-up. A limitation of their study was the potential
Results of the Gellman et al. and Alho et al. studies should inaccuracy of detecting union with plain radiographs
be interpreted with caution because there was significant instead of CT scans and the limited times of radiographs
heterogeneity between them. Retrospective studies by being performed. There could be a possibility of observer
Bongers et al. and Papaloizos et al. [17, 18] favoured bias, as two authors themselves assessed the radiographs
operative treatment over cast treatment, reporting for union. The study also had poor generalisability, as all
improvements in ROM, union rates and return to manual participants were full-time military personnel and therefore
work; however, the studies provide limited quality of evi- the times to return to duty may not reflect the time to return
dence due to the nature of the study design. to work in the general population. In comparison, the RCT
by McQueen et al. [9] was also a high quality study pro-
viding level 1b evidence. It is the only study in which the
Discussion outcomes were measured by a blinded assessor, thereby
reducing the risk of bias. A power and sample size calcu-
Traditional cast treatment for minimally displaced sca- lation was, however, not reported. Rehabilitation may have
phoid waist fractures is considered reliable and inexpen- differed between the groups as not all the patients had
sive, with low complication rates. Studies show that physiotherapy. Union rates may be inaccurate as radio-
approximately 85–90 % of these fractures will unite if graphs alone were used to define union due to the reasons
diagnosed early and treated promptly with cast immobili- discussed earlier.
sation. The main disadvantages of cast treatment are longer Among the prospective studies, Drac and Manak [10]
immobilisation time, joint stiffness, reduced grip strength performed a well-structured study with satisfactory statis-
and longer time to return to manual work [19]. Immobili- tical analysis; however, it was limited by the lack of CT
sation may be needed for up to 3 months, and patient scan comparison in the control (cast) group. In comparison,
compliance is thus often unsatisfactory, especially in the Haddad and Goddard [11] described systematic method-
presence of low symptom levels, when plasters may be ology and results in their study, but it was limited by the
discarded early, resulting in delayed union or nonunion method assessing functional outcome, which was sub-
[18, 20]. The advantages of nonoperative treatment have jective rather than employing a standardised tool; more-
been disputed, and some authors found incomplete healing over, the type and extent of patient professions were not
or nonunion in a high proportion of patients at late follow- described. Inoue and Shionoya [13] described good statis-
up. Cast immobilisation may also lead to ongoing pain and tical analysis and homogenous groups, but their study was
reduced ROM and grip strength [21]. Percutaneous fixa- nonrandomised. De Vos et al. [12] performed a retro-
tion, on the other hand, is aimed at reducing the damage to spective review that had a high likelihood of interobserver

variability in measuring ROM and grip strength. Union was nonunion. Early diagnosis and appropriate treatment are
assessed on the basis of plain radiographs, making the important to avoid long-term consequences associated with
assessment of union less reliable. Functional assessment nonunion. Managing minimally displaced fractures of the
was subjective rather than using a standardised tool. Brutus scaphoid waist in young patients employed in physically
and Beaton reported their results of a retrospective study of demanding work is even more challenging due to the issues
results from multiple surgeons, variable follow-up intervals of functional limitations and time off work. The best
and selection criteria and a 40 % dropout rate due to lost available evidence for percutaneous screw fixation versus
follow-up. These factors reduced the number of patients cast treatment suggests that percutaneous fixation results in
and the credibility of results due to lack of standardised a faster time to union by 5 (7 vs. 12) weeks and an earlier
methodology. It was interesting to note here that different return to manual work by 7 (8 vs. 15) weeks, with similar
types of screws used in different studies show similar union union rates [8, 9]. Cast treatment not only results in longer
rates and functional outcomes. duration to union but raises concerns of reduced ROM and
Among studies reporting the outcomes of nonoperative weakened grip strength. Considering the above evidence, a
treatment, a review by Rhemrev et al. provided an excellent young worker in a physically demanding job is likely to
account with regard to type of cast used, but the study lacked benefit from percutaneous fixation, which seems to
good functional outcome score, and functional outcome decrease immobilisation time, help achieve full ROM and
assessment was subjective. Some authors have shown that grip strength and allow earlier return to work. Detailed
immobilisation in slight dorsal extension has a positive effect discussion with the patient is required to explain the pros
on grip strength and wrist joint ROM [5, 15, 23]. The RCT by and cons of each treatment modality. There is a potential
Clay et al. [5] was a strong study in terms of design and requirement for a prospective randomised controlled trial
comparable group size, but union rates were assessed on the to compare these two treatment modalities for minimally
basis of plain radiographs only. In their series, 13 % (37/291) displaced scaphoid waist fractures in workers with physi-
of patients were reported as having probable union. Although cally demanding employment in order to achieve objective
the authors stated that these patients remained asymptomatic assessment of functional outcomes, time to union and
at 12 months, there was no later information. Gellman et al. return to work.
[15] assessed union on plain radiographs in their randomised
study between two different types of casts. Their trial was Conflict of interest None.
underpowered to support their conclusion of significant
benefit of long thumb-spica cast.
This systematic review attempted to focus on a specific
group of patients with a specific type of scaphoid fracture.
Strict inclusion criteria were observed when selecting the
studies, and no limit was applied to the minimum number
of patients in each study. Due to lack of a clear definition of References
minimally displaced fractures, the majority of studies
described outcomes of minimally as well as undisplaced 1. Alshryda S et al (2012) Acute fractures of the scaphoid bone:
fractures. Hence, it was difficult to separately assess the systematic review and meta-analysis. Surgeon 10(4):218–229
outcomes of minimally displaced fractures. In addition, 2. Modi CS et al (2009) Operative versus nonoperative treatment of
acute undisplaced and minimally displaced scaphoid waist frac-
some studies did not patient profession. In some studies, tures—a systematic review. Injury 40(3):268–273
fracture union was assessed by plain radiographs alone, 3. Cooney WP, Dobyns JH, Linscheid RL (1980) Fractures of the
with no CT scan assessment, hence making accurate scaphoid: a rational approach to management. Clin Orthop Relat
assessment of fracture union less reliable [8–10, 13]. Res 149:90–97
4. Filan SL, Herbert TJ (1996) Herbert screw fixation of scaphoid
This systematic review creates a potential requirement fractures. J Bone Joint Surg Br 78(4):519–529
for an RCT in order to compare outcomes of nonoperative 5. Clay NR et al (1991) Need the thumb be immobilised in scaphoid
treatment and percutaneous screw fixation, specifically fractures? A randomised prospective trial. J Bone Joint Surg Br
focusing on minimally displaced scaphoid waist fractures 73(5):828–832
6. Rhemrev SJ, van Leerdam RH, Ootes D, Beeres FJP, Meylaerts
and specifically in young individuals employed in physi- SAG (2009) Non-operative treatment of non-displaced scaphoid
cally demanding work. This will help establish decision- fractures may be preferred. Injury 40:638–641
making guidance for clinicians for appropriate manage- 7. Adolfsson L, Lindau T, Arner M (2001) Acutrak screw fixation
ment of this group of patients with regards to length of versus cast immobilisation for undisplaced scaphoid waist frac-
tures. J Hand Surg Br 26(3):192–195
immobilisation and time taken off work. 8. Bond CD et al (2001) Percutaneous screw fixation or cast
Managing scaphoid fractures remains a debatable issue immobilization for nondisplaced scaphoid fractures. J Bone Joint
because of the potential risk of delayed union and Surg Am 83-A(4):483–488

9. McQueen MM et al (2008) Percutaneous screw fixation versus 17. Bongers KJ, Ponsen RJ (1980) Operative and nonoperative
conservative treatment for fractures of the waist of the scaphoid: management of fractures of the carpal scaphoid: five years’
a prospective randomised study. J Bone Joint Surg Br experience. Neth J Surg 32(4):142–145
90(1):66–71 18. Papaloizos MY et al (2004) Minimally invasive fixation versus
10. Drac P, Manak P, Labonek I (2005) Percutaneous osteosynthesis conservative treatment of undisplaced scaphoid fractures: a cost-
versus cast immobilisation for the treatment of minimally and effectiveness study. J Hand Surg Br 29(2):116–119
non-displaced scaphoid fractures. Functional outcomes after a 19. Vinnars B et al (2008) Nonoperative compared with operative
follow-up of at least 12 month. Biomed Pap Med Fac Univ Pa- treatment of acute scaphoid fractures. A randomized clinical trial.
lacky Olomouc Czech Repub 149(1):149–151 J Bone Joint Surg Am 90(6):1176–1185
11. Haddad FS, Goddard NJ (1998) Acute percutaneous scaphoid 20. Barton NJ (1992) Twenty questions about scaphoid fractures.
fixation using a cannulated screw. Chir Main 17(2):119–126 J Hand Surg Br 17(3):289–310
12. Brutus JP et al (2002) Percutaneous Herbert screw fixation for 21. Herbert TJ, Fisher WE (1984) Management of the fractured
fractures of the scaphoid: review of 30 cases. Chir Main 21(6): scaphoid using a new bone screw. J Bone Joint Surg Br 66(1):
350–354 114–123
13. Inoue G, Shionoya K (1997) Herbert screw fixation by limited 22. Bushnell BD, McWilliams AD, Messer TM (2007) Complica-
access for acute fractures of the scaphoid. J Bone Joint Surg Br tions in dorsal percutaneous cannulated screw fixation of non-
79(3):418–421 displaced scaphoid waist fractures. J Hand Surg Am 32(6):
14. De Vos J, Vandenberghe D (2003) Acute percutaneous scaphoid 827–833
fixation using a non-cannulated Herbert screw. Chir Main 22(2): 23. Hambidge JE et al (1999) Acute fractures of the scaphoid.
78–83 Treatment by cast immobilisation with the wrist in flexion or
15. Gellman H et al (1989) Comparison of short and long thumb- extension? J Bone Joint Surg Br 81(1):91–92
spica casts for non-displaced fractures of the carpal scaphoid.
J Bone Joint Surg Am 71(3):354–357
16. Alho A, Kankaanpaa U (1975) Management of fractured sca-
phoid bone. A prospective study of 100 fractures. Acta Orthop
Scand 46(5):737–743

36
Variable-angle locking plate with or without double-tiered
subchondral support procedure in the treatment of intra-articular
distal radius fracture
Keikichi Kawasaki • Tetsuya Nemoto •
Katsunori Inagaki • Kazunari Tomita •
Yukio Ueno
Abstract Result There were no differences in clinical outcome and

Background Double-tiered subchondral support (DSS) complications. Final volar tilt and ulnar variance were
procedure is two-row fixation in which proximal screws better maintained in the DSS group (P = 0.01 and 0.03).
support the dorsal subchondral bone, whereas distal screws Change in volar tilt of the non-DSS group was more than
support the volar central subchondral bone, using the volar that of the DSS group (P = 0.00).
variable-angle locking plate to achieve better anatomical Conclusion Though there were no significant differences
reduction. We examined whether DSS improves clinical in clinical outcomes, we identified a significant reduction
outcome, complication rate, and loss of correction for in final volar tilt, ulnar variance, and change in volar tilt.
dorsally displaced Arbeitsgemeinschaft für Osteo- DSS procedure is useful to avoid correction loss when
synthesefragen (AO) type C3 distal radius fractures. treating unstable C3 distal radius fractures and thus would
Materials and methods We reviewed dorsally displaced reduce posttraumatic arthrosis.
intra-articular AO C3-type distal radius fractures treated at Level of evidence Level IV.
our institutions with a variable-angle volar locking plate.
We assessed 49 patients (27 DSS; 22 non-DSS) treated Keywords Distal radius fracture Locking plate
with volar locking plates, with a mean age of 59.9 years Double-tiered subchondral support
and average follow-up of 20.2 months (range
12–56 months). We evaluated differences in functional
outcome, complication rates, and loss of correction Introduction
between groups using radiographic parameters.
In intra-articular distal radius fractures, posttraumatic
arthrosis is directly related to the degree of anatomical
reduction [1, 2]. AO type C3 distal radius fractures are the
K. Kawasaki T. Nemoto (&) K. Inagaki K. Tomita most difficult to reduce and stabilize in the anatomical
Y. Ueno position because of the multifragmentary nature, high
Department of Orthopaedic Surgery, Showa University School
incidence of correction loss, and poorer prognosis for these
of Medicine, 1-5-8 Hatanodai, Shinagawa-ku, Tokyo 142-8666,
Japan patients [3]. The double-tiered subchondral support (DSS)
e-mail: [email protected] procedure has been reported to keep fractures of the distal
K. Kawasaki radius in anatomical position when using the volar locking
e-mail: [email protected] plate system [4]. With DSS, proximal screws support the
K. Inagaki dorsal subchondral bone from the proximal screws,
e-mail: [email protected] whereas distal screws support the central subchondral bone
K. Tomita from the distal row (as determined from a lateral view;
e-mail: [email protected] Fig. 1). However, there is no evidence to support the use-
Y. Ueno fulness of the DSS procedure for treating unstable distal
e-mail: [email protected] radius fractures.

Variable-angle locking plate with or without double-tiered subchondral support procedure in the treatment... 249
Fig. 1 Lateral view of distal radius with a variable-angle volar plate

using the double-tiered subchondral support procedure. Proximal
screws support the dorsal subchondral bone; distal screws support the
central subchondral bone Fig. 2 Postoperative lateral X-ray in the double-tiered subchondral
support (DSS) (a) and non-DSS (b) groups
Table 1 Patient characteristics Patients were treated with open reduction and internal
DSS group Non-DSS group
fixation (ORIF) with volar locking plates, with or without
(n = 27) (n = 22) DSS. For all patients, we used the APTUS radius 2.5 fix-
ation system (Medartis, Basel, Switzerland) with variable-
Age* (P = 0.02) 64.3 (23–85) 54.5 (26–83) angle volar locking plate, with a thickness of 1.6 mm and
Gender (male:female) 4:23 7:15 ±15° range of variable locking angles. In the DSS proce-
Dominant hand injured 12 8 dure, proximal screws were inserted toward the dorsal
Follow-up period 20.2 (12–53) 20.1 (12–56) subchondral bone, and distal screws were inserted toward
Ulnar styloid fracture 19 15 the central subchondral bone (Fig. 1). To be included in the
* Significant difference between groups (P = 0.02) DSS group, patients needed to be treated with more than
two proximal and distal screws (Fig. 2a). The remaining
In this study, we examined the hypothesis that the DDS patients formed the non-DSS group (Fig. 2b). Radio-
procedure in variable-angle volar locking plates could graphic parameters, including volar tilt (VT), radial incli-
improve clinical outcome, complications, and loss of cor- nation (RI), and ulnar variance (UV) were assessed on the
rection in the short term for patients with dorsally displaced operative day and at final follow-up. We evaluated range of
Arbeitsgemeinschaft für Osteosynthesefragen (AO) type motion, percent grip power, and Mayo wrist score [5] to
C3 intra-articular distal radius fractures. determine clinical outcome.
During postoperative care, wrists were immobilized in a
volar splint for 1 week, and active and passive finger-
Materials and methods motion exercises were started on the second postoperative
day. Active motion of the wrist was undertaken from
We reviewed all cases of AO type C3 distal radius fractures 1 week postoperatively. Passive movements were permit-
treated at our institutions with volar locking plate from ted after bony union was confirmed by surgeons. We
2008 to 2015 with a follow-up time [12 months. Patients evaluated functional, clinical, and radiographic outcomes
with a period from injury to operation [14 days were and compared findings between groups. Activities of daily
excluded. In addition, one patient who presented with an life (ADL) were assessed using the Mayo wrist scoring
ipsilateral carpal bone fracture was excluded. The 49 system by surgeons. We performed statistical analysis
patients consisted of 11 men and 38 women, with a mean using Fisher’s exact test for categorical data and the
age of 59.9 years (range 23–85). The group treated with independent t test for continuous data.
DSS comprised 27 patients (four men, 23 women), and the
group treated without DSS comprised 22 patients (seven
men, 15 women). Patients characteristics are summarized Results
in Table 1. There were no significant differences between
groups except for age, with DSS patients being signifi- Results are summarized in Table 2, and there were no
cantly older. significant differences between groups.

Table 2 Clinical outcomes measures additional screws and K wires were not sufficient in these
DSS group Non-DSS group P value
cases and that additional bone grafting should be per-
(n = 27) (n = 22) formed for metaphyseal bone loss [10]. In our study,
though there were no significant differences in clinical
Range of motion
outcomes, we identified a significant reduction in final VT,
Flexion 68.0 ± 12.0 65.5 ± 12.3 0.48 RI, and change in VT. Distal screws support the central
Extension 74.8 ± 10.4 74.8 ± 11.6 0.99 subchondral bone and transmit the axial force from the
Pronation 87.0 ± 5.9 86.4 ± 7.1 0.72 subchondral bone to the intact diaphyseal bone [11]. We
Supination 89.6 ± 1.3 88.2 ± 5.7 0.25 believe that this transmission of force maintained ana-
Percent grip power 97.3 ± 24.1 88.3 ± 20.3 0.17 tomical reduction of the lunate facet, thus achieving good
Cooney score 91.1 ± 5.9 91.4 ± 8.8 0.91 clinical outcomes. However, in non-DSS cases in which
Data are mean ± standard deviation. There were no significant dif- distal screws only were used, we observed a VT correction
ferences between groups loss of after early mobilization in AO type C3 fractures
compared with the DSS group. We hypothesize that in
Table 3 Radiological outcomes measures
multifragmentary articular fractures, distal screws alone are
not sufficient to catch the dorsal subchondral bone. As
DSS group Non-DSS group P value such, dorsal articular subchondral bone should be sup-
(n = 27) (n = 27)
ported with proximal screws. In the DSS group, we believe
Postoperative volar tilt (°) 7.5 ± 4.9 6.2 ± 4.8 0.35 that proximal screws transferred the dorsal articular load to
Postoperative radial 24.0 ± 3.9 22.7 ± 3.1 0.21 the implant and the diaphyseal bone. Although correction
inclination (°) loss of VT did not affect short-term clinical outcomes, it is
Postoperative ulnar 0.2 ± 1.2 -0.7 ± 1.6 0.03* likely to affect posttraumatic arthrosis; a longer follow-up
variance (mm)
period is required to determine if this is the case.
Final volar tilt (°) 8.0 ± 5.8 3.8 ± 5.0 0.01* In this study, RI in the DSS group was better maintained
Final radial inclination (°) 24.8 ± 4.2 23.0 ± 3.3 0.09 than in the non-DSS group. Stanbury et al. showed the
Final ulnar variance (°) 1.0 ± 1.1 0.2 ± 1.6 0.03* superiority of a variable-angle locking plate for capturing a
Change in volar tilt (°) ?0.5 ± 1.9 -2.4 ± 3.4 0.00* distal radial styloid compared with a fixed-angle plate in
Change in radial ?0.8 ± 2.4 ?0.2 ± 2.4 0.41 biomechanical study [12]. A variable locking plate using
inclination (°)
the DSS procedure might capture a distal radial styloid
Change in ulnar variance ?0.8 ± 0.9 ?0.8 ± 1.2 0.94
(mm)
fragment better than a non-DSS procedure. The overall
complication rate in our study was 4.1 %, with complica-
Data are mean ± standard deviation tions occurring in each group and no significant difference
* Significant difference observed. A similar complication ratio of 3 % was
observed in another short-term study with a mean follow-
Radiographic parameters are summarized in Table 3. up of 12 months [13]. Flexor tendon rupture after volar
Postoperative UV in the DSS group was smaller and final plate fixation has been reported by some authors [14, 15].
VT better maintained in the DSS group; change in VT was In some severe intra-articular fractures, it was necessary to
greater in the non-DSS group. place the plate distally from the watershed line [16], which
Postoperative complications occurred in two patients: is a transverse ridge that closes the concave surface of the
one in the DSS group had a mild subcutaneous infection, volar radius distally. If the plate is placed above or distally
and one in the non-DSS group developed complex regional from this line, it increases the risk of flexor tendon rupture.
pain syndrome type 1. Both patients were treated conser- We believe that the risk of flexor pollicis longus (FPL)
vatively. None of the patients had delayed union or rupture would be decreased by using a thinner volar plate.
nonunion. In cadaveric models, insertion of screws in all available
holes in the distal fragment of a variable-angle volar plate
showed the highest biomechanical stability [17]. In our
Discussion series, we routinely removed the plate system after bony
union to take into consideration the risks to flexor tendons.
AO type C3 distal radius fractures are the most unstable There were several limitations to this study. We could
fractures, and some groups have indicated the limitation of not compare operation time between groups because we
treatment with volar locking plate alone, using additional K simultaneously performed operations targeting accompa-
wires, and dorsal-plating fixation for patients with these nied disorders, such as ulnar styloid fracture. The DSS
types of fractures [6–9]. Another group reported that procedure required longer fluoroscopy time because of the

Variable-angle locking plate with or without double-tiered subchondral support procedure in the treatment... 251
careful attention required to position screws. Also, three of 4. Orbay JL, Fernandez DL (2004) Volar fixed-angle plate fixation
27 patients in the DSS group and one in the non-DSS group for unstable distal radius fractures in the elderly patient. J Hand
Surg Am 29(1):96–102
underwent artificial bone grafting, which is useful for AO 5. Amadio PC, Berquist TH, Smith DK, Ilstrup DM, Cooney WP
type C3 distal radius fractures for articular surface recon- 3rd, Linscheid RL (1989) Scaphoid malunion. J Hand Surg Am
struction [6]. Finally, this was a retrospective comparative 14(4):679–687
study; a prospective randomized study is required to 6. Arora R, Lutz M, Hennerbichler A, Krappinger D, Espen D, Gabl
M (2007) Complications following internal fixation of unstable
examine the effectiveness of the DSS procedure. A larger distal radius fracture with a palmar locking-plate. J Orthop
patient cohort and longer follow-up are necessary to eval- Trauma 21(5):316–322
uate the usefulness of the DSS procedure. 7. Osada D, Kamei S, Takai M, Tomizawa K, Tamai K (2007)
In conclusion, patients with AO type C3 distal radius Malunited fractures of the distal radius treated with corrective
osteotomy using volar locking plate and a corticocancellous bone
fractures treated with the DSS procedure had a reduced graft following immediate mobilisation. Hand Surg 12(3):
change in VT than patients treated with the non-DSS 183–190
method. We believe that the DSS procedure improves loss 8. Benson LS, Minihane KP, Stern LD, Eller E, Seshadri R (2006)
of VT correction and would help prevent posttraumatic The outcome of intra-articular distal radius fractures treated with
fragment-specific fixation. J Hand Surg Am 31(8):1333–1339
arthrosis in the long term. 9. Rikli DA, Regazzoni P (1996) Fractures of the distal end of the
radius treated by internal fixation and early function. A pre-
Conflict of interest Each author certifies that he or she has no liminary report of 20 cases. J Bone Joint Surg Br 78(4):588–592
commercial associations (e.g., consultancies, stock ownership, equity 10. Schneeberger AG, Ip WY, Poon TL, Chow SP (2001) Open
interest, patent/licensing arrangements, etc.) that might pose a conflict reduction and plate fixation of displaced AO type C3 fractures of
in connection with the submitted article. the distal radius: restoration of articular congruity in eighteen
cases. J Orthop Trauma 15(5):350–357
Ethical standards All patients gave informed consent prior being 11. Figl M, Weninger P, Jurkowitsch J, Hofbauer M, Schauer J,
included in the study. The study was authorized by the local ethical Leixnering M (2010) Unstable distal radius fractures in the
committee and performed in accordance with the ethical standards of elderly patient–volar fixed-angle plate osteosynthesis prevents
the 1964 Declaration of Helsinki as revised in 2000. secondary loss of reduction. J Trauma 68(4):992–998
12. Stanbury SJ, Salo A, Elfar JC (2012) Biomechanical analysis of a
volar variable-angle locking plate: the effect of capturing a distal
radial styloid fragment. J Hand Surg Am 37(12):2488–2494
13. Kamano M, Koshimune M, Toyama M, Kazuki K (2005) Palmar
plating system for Colles’ fractures—a preliminary report. J Hand
Surg Am 30(4):750–755
14. Valbuena SE, Cogswell LK, Baraziol R, Valenti P (2010) Rup-
References ture of flexor tendon following volar plate of distal radius frac-
ture. Report of five cases. Chir Main 29(2):109–113
15. Drobetz H, Kutscha-Lissberg E (2003) Osteosynthesis of distal
1. McQueen M, Caspers J (1988) Colles fracture: does the ana- radial fractures with a volar locking screw plate system. Int
tomical result affect the final function? J Bone Joint Surg Br Orthop 27(1):1–6
70(4):649–651 16. Orbay J (2005) Volar plate fixation of distal radius fractures.
2. Trumble TE, Schmitt SR, Vedder NB (1994) Factors affecting Hand Clin 21(3):347–354
functional outcome of displaced intra-articular distal radius 17. Mehling I, Muller LP, Delinsky K, Mehler D, Burkhart KJ,
fractures. J Hand Surg Am 19(2):325–340 Rommens PM (2010) Number and locations of screw fixation for
3. Kamei S, Osada D, Tamai K, Kato N, Takai M, Kameda M, volar fixed-angle plating of distal radius fractures: biomechanical
Nohara Y (2010) Stability of volar locking plate systems for AO study. J Hand Surg Am 35(6):885–891
type C3 fractures of the distal radius: biomechanical study in a
cadaveric model. J Orthop Sci 15(3):357–364

37
Predictive factors of hospital length of stay in patients
with operatively treated ankle fractures
Matthew R. McDonald • Vasanth Sathiyakumar • Jordan C. Apfeld •
Benjamin Hooe • Jesse Ehrenfeld • William T. Obremskey • Manish K. Sethi
Abstract Charts were reviewed for gender, length of operative pro-

Background Operative fixation of ankle fractures is cedure, method of fixation, American Society of Anesthe-
common. However, as reimbursement plans evolve with siologists (ASA) physical status score, medical
the potential for bundled payments, it is critical that comorbidities, and postoperative LOS. Both univariate and
orthopedic surgeons better understand factors influencing multivariate models were developed to determine predic-
the postoperative length of stay (LOS) in patients under- tors of patient LOS. Financial data for an average 24-h
going these procedures to negotiate appropriate reim- inpatient stay were obtained from financial services.
bursement. We sought to identify factors influencing the Results Six hundred twenty-two patients were included.
postoperative LOS in patients with operatively treated In a linear regression analysis, a statistically significant
ankle fractures. relationship was demonstrated between ASA status and
Materials and methods Six hundred twenty-two patients LOS (P \ 0.001). Multiple regression analysis further
with ankle fractures between January 1st, 2004 and characterized the relationship between ASA and LOS: a
December 31st, 2010 were identified retrospectively. 1-U increase in ASA classification conferred a 3.42-day
increase in LOS on average (P \ 0.001). Based on an
average per-day inpatient cost of $4,503, each unit increase
in ASA status led to a $15,490 increase in cost.
Conclusions Our study demonstrates that ASA status is a
powerful predictor of LOS in patients undergoing operative
M. R. McDonald V. Sathiyakumar J. C. Apfeld B. Hooe fixation of ankle fractures. More complete understanding of
J. Ehrenfeld W. T. Obremskey M. K. Sethi (&) these factors will lead to better risk adjustment models for
The Vanderbilt Orthopaedic Institute Center for Health Policy,
measuring outcomes, determining fair reimbursement, and
Vanderbilt University, Suite 4200, South Tower, MCE,
Nashville, TN 37221, USA potential improvements to the efficiency of patient care.
e-mail: [email protected] Level of Evidence Level III retrospective comparative
M. R. McDonald study regressing length of stay with many variables,
e-mail: [email protected] including ASA physical status.
V. Sathiyakumar
e-mail: [email protected] Keywords Ankle fracture Postoperative length of
J. C. Apfeld stay ASA score Payment and reimbursement model
e-mail: [email protected] Surgical outcomes Healthcare costs
B. Hooe
e-mail: [email protected]
J. Ehrenfeld Introduction
e-mail: [email protected]
W. T. Obremskey Musculoskeletal injuries have become increasingly more
e-mail: [email protected] common in the USA, with approximately three out of every

Predictive factors of hospital length of stay in patients with operatively treated ankle fractures 253
five injuries occurring to this system [1]. Among these procedure was an emergency. Other data acquired from the
injuries, sprains/strains represent the most frequent (44 %), patients’ charts were total LOS, days from admission to
and fractures represent the second most frequent (25 %) surgery, days from surgery to discharge, and any docu-
[1]. Between 2006 and 2007, an estimated 16.2 million mented complications. In addition, a history of other
outpatient and inpatient fractures were treated [1]. One medical comorbidities for each patient was acquired,
population-based study reports that ankle fractures occur in including prior myocardial infarction (MI), dysrhythmia,
187 per 100,000 persons every year [4]. The CDC esti- atrial fibrillation, congestive heart failure (CHF), heart
mated that, in 2003 alone, over 1.8 million people visited block, cerebrovascular disease, chronic obstructive pul-
the emergency room because of ankle and lower leg inju- monary disease (COPD), emphysema, current smoking
ries [9]. status, prior smoking history, renal disease, dialysis, can-
A prospective cost analysis study of 30 patients with cer, and diabetes. No patient identifiers were included in
unstable ankle fractures found the total inpatient hospital this deidentified database. In addition, ASA classification
cost to be $1,801 per patient and the total outpatient cost to was also obtained for each patient. This value was assigned
be $333 per patient [2]. Given the prevalence of these to each patient by the anesthesiologist just prior to the start
operations, it is imperative to seek ways to reduce these of the operative procedure. Incomplete charts were exclu-
costs. Healthcare costs continue to be a central issue in the ded from the analysis.
US economy, especially with the signing of healthcare The average total cost to the hospital of an inpatient day
reform legislation in 2010. With impending changes such at the institution was obtained from hospital financial ser-
as the transition to a bundled payment system, it is crucial vices, and the average cost of an inpatient day was treated
for orthopedic trauma surgeons to develop a better under- as a unit cost per inpatient day. The length of stay for each
standing of cost drivers for the treatment of ankle fractures. patient was multiplied by this unit cost to estimate the cost
One avenue of interest in the pursuit of such variables is of inpatient postoperative care per patient for a given visit.
the role of the American Society of Anesthesiologists Both univariate and multivariate models were developed to
(ASA) classification score, which is based on the anes- determine predictors of length of stay and thus cost of the
thesiologist’s evaluation of the patient’s health status and length of stay postoperatively.
comorbidities prior to an operation [7]. The ASA score has
been proven to be an effective predictor of patients with an
increased risk of complications, including perioperative Results
risk assessment, perioperative mortality, as well as post-
operative outcome [10–12], with higher scores associated After exclusion of incomplete charts, 622 patient charts
with increased risk of complications. ASA scores have also were available for analysis. Basic demographic information
been shown to be a significant predictor in length of stay is provided in Table 1. The average patient age was
(LOS) and therefore cost management of patients under- 44.58 years, and the average LOS was 5.59 days. All
going operative fixation of hip fractures [5] as well as patients were ASA class 1 through 4, with the majority of
return-to-function status [3]. Therefore, this retrospective patients being class 2.
study sought to elucidate the relationship between various In a linear regression analysis, a statistically significant
patient variables, such as ASA score, length of operative relationship was demonstrated between ASA score and
procedure, method of fixation, medical comorbidities, and length of stay (P \ 0.001). Multiple regression analysis
postoperative LOS, in patients undergoing open treatment was conducted to further characterize the relationship
of an ankle fracture. between ASA score and LOS: a 1-U increase in ASA score
conferred a 3.42-day increase in LOS on average
(P \ 0.0001). The average total cost to the hospital of an
Materials and Methods inpatient day at the institution was found to be $4,530.
Treating the average inpatient daily cost as a unit cost, a
After obtaining approval from our institutional review 1-U increase in ASA led to a $15,490 increase in cost to the
board, all patients who underwent open treatment of ankle institution. Table 2 summarizes the multivariate regression
fractures between January 1, 2004 and December 31, 2010 of comorbidities on LOS. No single comorbidity reached
were identified through a current procedural terminology statistical significance as a predictor of LOS.
(CPT) code search (Appendix Supplementary Table 1). Table 3 describes the length of stay data for each ASA
The following metrics were extracted from the selected classification 1 through 4. A predicted total cost to the
patients’ records: date of birth, height, weight, date of institution for the inpatient stay of a patient with ASA
admission, age at time of procedure, start and stop times of score 1 through 4 was calculated utilizing the data pro-
procedure, duration of procedure, and whether or not vided by financial services for the cost of a single-day

Table 1 Population demographic information Table 3 Comparison of ASA physical status, mean LOS, mean
operative duration, predicted costs, and actual costs
N %
ASA Patients Mean Mean Predicted Actual
Age (years) status LOS operative cost cost
0–9 0 0.00 (days) duration (min)
10–19 46 7.40
1 57 2.58 127.37 $3,418 $11,687
20–29 116 18.65
2 328 3.84 134.64 $19,536 $17,395
30–39 99 15.92
3 196 8.07 147.29 $36,118 $36,557
40–49 104 16.72
4 41 11.95 121.88 $49,423 $54,134
50–59 114 18.33
60–69 77 12.38
70–79 42 6.75
80–89 16 2.57
90–99 1 0.16
[99 7 1.13
Gender
Female 305 49.04
Male 317 50.96
Race
Caucasian 497 79.90
African American 77 12.38
Hispanic 19 3.05
Asian 2 0.32
Fig. 1 Relationship between ASA physical status and both mean
American Indian 1 0.16 predicted costs and mean actual costs
Other/unknown 26 4.18
ASA status
inpatient stay. A relatively small difference in mean LOS
1 57 9.16
between ASA score of 1 and ASA score of 2 was observed,
2 328 52.73
as would be expected as both of these groups represent
3 196 31.51
generally healthy patients. However, a steep rise in the
4 41 6.59
mean LOS from 3.84 to 8.07 days was observed between
Other
ASA scores 2 and 3, respectively.
Mean length of stay (days) 5.59
Figure 1 depicts the trend in the actual average cost of
Mean duration of surgery (min) 137.11
postsurgery hospitalization for an ankle fracture patient
based on the ASA classification.
Table 2 Results of regression analysis of comorbidities on LOS

Comorbid condition LOS Costa P value Discussion
MI 0.64565 $2,924.79 0.7829
Our data show that a 1-U increase in the ASA score of
Dysrhythmia 0.54559 $2,471.52 0.7704
patients undergoing operative repair of an ankle fracture is
Hypertension -0.00485 -$21.97 0.9968
associated with a 3.42-day increase in average LOS. These
CHF 2.10027 $9,514.22 0.4892
results complement those previously published regarding
Heart block 1.42492 $6,454.89 0.7458
the positive association between ASA physical status and
Renal insufficiency 0.55606 $2,518.95 0.8498
average LOS in hip fracture patients, confirming that the
Dialysis 3.02694 $13,712.04 0.6605
ASA physical status classification system is a powerful
Diabetes -1.31132 -$5,940.28 0.4759
predictor of LOS. Given that a particular patient’s ASA
Cancer -1.7441 -$7,900.77 0.3589
physical status can seldom be altered prior to surgery, the
Cocaine use -1.30578 -$5,915.18 0.7092 utility of this information is more beneficial in terms of
Alcohol use -2.71025 -$12,277.43 0.1523 budgeting and planning for patients undergoing open
Opiate use 2.58596 $11,714.40 0.6979 treatment of ankle fractures. ASA physical status and daily
a
Cost is a direct multiplication of average per diem inpatient cost by costs are commonly collected, which makes tracking of
LOS these data for real-time budgeting and bed management

Predictive factors of hospital length of stay in patients with operatively treated ankle fractures 255
easily employable at essentially any institution. Further- Conflict of interest W.T.O. has previously consulted for biomet-
more, as reimbursement systems change from a fee-for- rics, done expert testimony in legal matters, and has a grant from the
Department of Defense. For the remaining authors no conflicts of
service model to a fixed-reimbursement model, this study interest were declared.
highlights the utility of a tiered reimbursement model for
each diagnosis based on relatively static patient factors Ethical standards The study was authorized by the local ethical
such as ASA physical status. committee at Vanderbilt and was performed in accordance with the
ethical standards of the 1964 Declaration of Helsinki as revised in
While LOS was the only covariate found to be predic- 2000.
tive of LOS, one of the limitations of this study is that not
all possible patient variables could be accounted for, and
thus it remains possible that some other single factor not
included in this analysis better predicts LOS variation
rather than ASA physical status. Furthermore, an intricate
interplay undoubtedly exists between multiple comorbidi-
ties and patient variables that predict LOS. ASA physical
status was designed expressly for the purpose of integrating References
all of a patient’s comorbidities into a single value, and this
is why the authors believe it is such a good predictor of 1. Andersson G, Bouchard J, Bozic KJ et al (2011) The burden of
musculoskeletal diseases in the United States. American Acad-
LOS in orthopedic trauma patients.
emy of Orthopaedic Surgeons, Rosemont, pp 129–179
In conclusion, this study demonstrates the powerful 2. Bhandari M, Sprague S, Ayeni O, Hanson B, Moro J (2004) A
nature of the ASA classification for explanation of variance prospective cost analysis following operative treatment of
of length of stay for hospitalized patients. This method can unstable ankle fractures. Acta Orthop Scand 75(1):100–105
3. Colvin A, Walsh M, Koval K et al (2009) Return to sports fol-
easily be integrated into almost any hospital, since ASA
lowing operatively treated ankle fractures. Foot Ankle Int
and costs are universally collected. This model may be 30(4):292–296
utilized to predict a patient’s postoperative course more 4. Daly PJ, Fitzgerald RH, Melton LJ, Ilstrup DM (1987) Epide-
accurately, which has benefits for predicting costs, bud- miology of ankle fractures in Rochester, Minnesota. Acta Orthop
Scand 58(5):539–544
geting, and management of patient beds. Additional
5. Garcia AE, Bonnaig JV, Yoneda ZT et al (2012) Patient variables
research may also be done to investigate associations which may predict length of stay and hospital costs in elderly
between ASA physical status and other cost drivers in patients with hip fracture. J Orthop Trauma 26(11):620–623
addition to LOS. 6. Haynes S, Lawler P (1995) An assessment of the consistency of
ASA physical status classification allocation. Anaesthesia
Other limitations to this study include the fact that this
50(3):195–199
particular patient population was from a single, tertiary 7. Keats AS (1978) The ASA classification of physical status—a
medical center, which may receive ‘‘sicker’’ patients, and recapitulation. Anesthesiology 49(4):233–236
therefore may not be indicative of the general population. 8. Mak P, Campbell R, Irwin M (2002) The ASA physical status
classification: inter-observer consistency. Anaesth Intensive Care
Other factors such as surgeon, surgical technique, delayed
30(5):633–640
discharge, pain control, or insurance status could not be 9. Mccaig LF, Burt CW, Ed D, Care H (2005) National Hospital
included in the multivariate regression and may all also Ambulatory Medical Care Survey: 2003 Emergency Department
contribute to LOS. Finally, one of the major limitations of Summary
10. Menke H, Klein A, John K, Junginger T (1993) Predictive value
utilizing ASA score is the fact that research has shown it to
of ASA classification for the assessment of the perioperative risk.
have little to moderate interrater reliability [6, 8]. Int Surg 78(3):266–270
11. Praus G, Ratzenhofer-Comenda B, Pierer G et al (1997) Can ASA
Acknowledgments No grant money or other support (technical, grade or Goldman’s cardiac risk index predict peri-operative
corporate) was involved for this research study. mortality? A study of 16,277 patients. Anaesthesia 52(3):203–206
12. Wolters U, Wolf T, Stutzer H, Schroder T (1996) ASA classifi-
Copyrighted material/consent forms This study used no previ- cation and perioperative variables as predictors of postoperative
ously copyrighted materials or signed patient consent forms. This outcome. Br J Anaesth 77:217–222
study was retrospective in nature.

Permissions
All chapters in this book were first published in JOT, by Springer; hereby published with permission under the
Creative Commons Attribution License or equivalent. Every chapter published in this book has been scrutinized
by our experts. Their significance has been extensively debated. The topics covered herein carry significant findings
which will fuel the growth of the discipline. They may even be implemented as practical applications or may be
referred to as a beginning point for another development.
The contributors of this book come from diverse backgrounds, making this book a truly international effort. This
book will bring forth new frontiers with its revolutionizing research information and detailed analysis of the
nascent developments around the world.
We would like to thank all the contributing authors for lending their expertise to make the book truly unique.
They have played a crucial role in the development of this book. Without their invaluable contributions this book
wouldn’t have been possible. They have made vital efforts to compile up to date information on the varied aspects
of this subject to make this book a valuable addition to the collection of many professionals and students.
This book was conceptualized with the vision of imparting up-to-date information and advanced data in this field.
To ensure the same, a matchless editorial board was set up. Every individual on the board went through rigorous
rounds of assessment to prove their worth. After which they invested a large part of their time researching and
compiling the most relevant data for our readers.
The editorial board has been involved in producing this book since its inception. They have spent rigorous hours
researching and exploring the diverse topics which have resulted in the successful publishing of this book. They
have passed on their knowledge of decades through this book. To expedite this challenging task, the publisher
supported the team at every step. A small team of assistant editors was also appointed to further simplify the
editing procedure and attain best results for the readers.
Apart from the editorial board, the designing team has also invested a significant amount of their time in
understanding the subject and creating the most relevant covers. They scrutinized every image to scout for the
most suitable representation of the subject and create an appropriate cover for the book.
The publishing team has been an ardent support to the editorial, designing and production team. Their endless
efforts to recruit the best for this project, has resulted in the accomplishment of this book. They are a veteran
in the field of academics and their pool of knowledge is as vast as their experience in printing. Their expertise
and guidance has proved useful at every step. Their uncompromising quality standards have made this book an
exceptional effort. Their encouragement from time to time has been an inspiration for everyone.
The publisher and the editorial board hope that this book will prove to be a valuable piece of knowledge for
researchers, students, practitioners and scholars across the globe.

List of Contributors
Alessandro Aprato, Ferdinando Tosto, Marco Favuto, Federica Bergamin, Ezio Nicola Gangemi, Claudia
Alessandro Massè and Alessandro Stucchi Cerato, Alessandra Clemente and Marco Borsetti
Medical School, University of Turin, Turin, Italy Department of Surgical Activities, Division of Plastic
Surgery and Hand Surgery, Maria Vittoria Hospital, ASL
Vasiliki Kalampoki, Alexander Joeris and Elke Rometsch TO2, Via Cibrario 72, 10144 Turin, Italy
Clinical Investigation and Documentation (C.I.D.)
Department, AO Foundation, Du¨bendorf, Switzerland Adolfo Suriani and Stefano Taraglio
Department of Laboratory Diagnostics, Division of
Matheus Azi Pathology, Maria Vittoria Hospital, ASL TO2, Via
Manoel Victorino Hospital, Salvador, Brazil Cibrario 72, 10144 Turin, Italy
Ahmed Ezzat Siam Adeel Aqil, Fahad Hossain, Hassaan Sheikh, Joseph
Spine Unit, El Hadara University Hospital, Alexandria, Aderinto, George Whitwell and Harish Kapoor
Egypt Level 1 Trauma Centre, Leeds General Infirmary, Great
Department of Spinal Surgery with Scoliosis Centre, George St, Leeds LS1 3EX, UK
Schön Klinik Vogtareuth, Krankenhausstrasse 20, 83569
Vogtareuth, Germany Fabrizio Rivera
Department of Orthopedic Trauma, SS Annunziata
Hesham El Saghir Hospital, Savigliano, CN, Italy
Department of Spinal Surgery, Zentralklinik Bad Berka, Via Servais 200 A 16, Turin, Italy
Bad Berka, Germany
Spine Unit, El Hadara University Hospital, Alexandria, Luca Bertignone
Egypt Sant’Anna Clinic, Casale Monferrato, AL, Italy
Eporediese Hospital, Ivrea, TO, Italy
Heinrich Boehm
Department of Spinal Surgery, Zentralklinik Bad Berka, Giancarlo Grandi and Roberto Camisassa
Bad Berka, Germany La Vialarda Clinic, Biella, Italy
Fabrizio Rivera Guido Comaschi

Department of Orthopedic Trauma, SS Annunziata Sestri Ponente Hospital, Genova, Italy
Hospital, Via Ospedali 14, 12038 Savigliano (CN), Italy
Diego Trentini
Donato Rosa, Giovanni Balato, Giovanni Ciaramella, Department of Orthopedics and Traumatology, IRCCS
Giovanni Improta and Maria Triassi A.O.U. San Martino-IST, Genova, Italy
Department of Public Health, School of Medicine, Federico
II University, Via S. Pansini 5, Bl. 12, 80131 Naples, Italy Marco Zanone, Giuseppe Teppex and Gabriele Vasario
Department of Orthopedics and Traumatology, AO CTO
Ernesto Soscia Hospital, Turin, Italy
Institute of Biostructure and Bioimaging, National
Research Council, Via S. Pansini 5, 80131 Naples, Italy Giorgio Fortina
Santa Rita Clinic, Vercelli, Italy
Jason Wong, Jared M. Newman and Konrad I. Gruson
Department of Orthopaedic Surgery, Albert Einstein Cesare Faldini
College of Medicine, 1250 Waters Place, 11th Floor, University of Bologna, Bologna, Italy
Bronx, NY 10461, USA Dipartimento Rizzoli-Sicilia, Istituto Ortopedico Rizzoli,
Strada Statale 113 km 246, 90011 Bagheria, PA, Italy
Hemanth Kumar Venkatesh
Department of Orthopedics, Basildon and Thurrock Francesco Traina and Matteo Nanni
University Hospital, Nethermayne, Basildon, Essex Dipartimento Rizzoli-Sicilia, Istituto Ortopedico Rizzoli,
SS165NL, UK Strada Statale 113 km 246, 90011 Bagheria, PA, Italy
S. S. Maheswaran Ilaria Sanzarello

Department of Orthopedics, University Hospital of North University of Messina, Messina, Italy
Tees & Hartlepool, Stockton On Tees TS19 8PE, UK
Fabrizio Perna and Raffaele Borghi
University of Bologna, Bologna, Italy

258 /LVWRI&RQWULEXWRUV
Michele Vasso, Chiara Del Regno, Antonio D’Amelio Matteo Guelfi

and Alfredo Schiavone Panni Orthopaedic and Traumatology Division, G. d’Annunzio
Department of Medicine and Health Sciences, University University, Via dei Vestini 35, 66013 Chieti, Italy
of Molise, Via Francesco De Sanctis, 86100 Campobasso, Via Caprera 7/3, 16146 Genoa, Italy
Italy
Andrea Pantalone, Daniele Vanni and Vincenzo Salini
Christian Carulli, Armando Macera, Fabrizio Matassi, Orthopaedic and Traumatology Division, G. d’Annunzio
Roberto Civinini and Massimo Innocenti University, Via dei Vestini 35, 66013 Chieti, Italy
Orthopaedic Clinic, University of Florence, Largo P.
Palagi 1, 50139 Florence, Italy Janos Cambiaso Daniel
Department of Plastic, Aesthetic and Reconstructive
Olga Charyeva Surgery, Medical University of Graz, Augenbruggerplatz
aap Biomaterials GmbH, Lagerstrasse 11-15, 64807 29, 8036 Graz, Austria
Dieburg, Germany
Laboratory for Experimental Trauma Surgery, Justus- Marco G. B. Guelfi
Liebig University Giessen, Schubertstrasse 81, 35392 Orthopaedic Division, Clinica Montallegro Via M.Te
Giessen, Germany Zovetto 27, 16145 Genoa, Italy
Olga Dakischew, Ursula Sommer and Katrin Susanne Lawrence Camarda, Alessandra La Gattuta, Marcello
Lips Butera, Francesco Siragusa and Michele D’Arienzo
Laboratory for Experimental Trauma Surgery, Justus- Department of Orthopaedic Surgery, DICHIRONS,
Liebig University Giessen, Schubertstrasse 81, 35392 University of Palermo, Via del Vespro, 90100 Palermo,
Giessen, Germany Italy
Christian Heiss and Reinhard Schnettler Filip Bjørdal

Department of Trauma Surgery, University Hospital of University of Oslo, Hollandveien 26, 1555 Son, Norway
Giessen-Marburg, Rudolf-Buchheim-Str. 9, 35385 Giessen,
Germany Kristian Bjørgul
Østfold Hospital Trust, Aleris Health Oslo, Chr. Svendsens
Jonathan G. Eastman gate 6, 1771 Halden, Norway
Department of Orthopaedic Surgery, University of
California, Davis Medical Center, 4860 Y Street, Suite Pooria Salari
3800, Sacramento, CA 95817, USA Department of Orthopaedic Surgery, Saint Louis
University School of Medicine, 3635 Vista Avenue, 7th
Jr L. Chip Routt. Milton Floor Desloge Towers, St. Louis, MO 63110, USA
Department of Orthopaedic Surgery, University of Texas,
Health Sciences Center at Houston, Houston, TX, USA Berton R. Moed and J. Gary Bledsoe
Department of Orthopaedic Surgery, Saint Louis
George A. Koumantakis University School of Medicine, 3635 Vista Avenue, 7th
Department of Physical Therapy, 401 Army General Floor Desloge Towers, St. Louis, MO 63110, USA
Hospital of Athens, 1 Panagioti Kanellopoulou Avenue, Department of Biomedical Engineering, Parks College
11525 Athens, Greece of Engineering, Aviation and Technology, Saint Louis
University, 3450 Lindell Boulevard, St. Louis, MO 63103,
Konstantinos Tsoligkas USA
2nd Department of Anesthesiology-Pain Unit, ATTIKON
University General Hospital, School of Medicine, V. Tristaino, S. Tornago, M. Gramazio, E. Carriere and
University of Athens, Athens, Greece A. Camera
Department of Prosthetic Surgery, Santa Corona Hospital,
Antonios Papoutsidakis ASL 2 - Savonese, via XXV Aprile 38, 17027 Pietra Ligure,
Orthopaedic Surgeon, Rethymno, Greece SV, Italy
Athanasios Ververidis and Georgios I. Drosos F. Lantieri

Department of Orthopaedic Surgery, University General Health Science Department, Biostatistics Unit, University
Hospital of Alexandroupolis, School of Medicine, of Genoa, via Pastore 1, Genoa 16132, Italy
Democritus University of Thrace, Alexandroupolis,
Greece Rohit Singh, Gopikanthan Manoharan and Pete Craig
Robert Jones Orthopaedic Hospital, Oswestry, UK
Christian Carulli, Filippo Tonelli, Matteo Innocenti,
Bonaventura Gambardella, Francesco Muncibì and Simon Collier, Phillip Shaylor and Ashok Sinha
Massimo Innocenti Mid Staffordshire foundation trust, Stafford, England,
Orthopaedic Clinic, University of Florence, Largo P. UK
Palagi 1, 50139 Florence, Italy

/LVWRI&RQWULEXWRUV 259
Chayanin Angthong, Pongpaibool Krajubngern, Giovanni Vavalle

Boonchana Pongcharoen, Piya Pinsornsak and Nattapol Department of Orthopaedics, Saint Mary Hospital, De
Tammachote Ferrariis 18/D, 70124 Bari, Italy
Department of Orthopaedics, Faculty of Medicine, Via Martiri della Giustizia, 9, 70016 Noicattaro, Italy
Thammasat University, Pathum Thani 12120, Thailand
Michele Capozzi
Warawut Tiyapongpattana Department of Orthopaedics, Saint Mary Hospital, De
Department of Chemistry, Faculty of Science and Ferrariis 18/D, 70124 Bari, Italy
Technology, Thammasat University, Pathum Thani,
Thailand Alisara Arirachakaran, Amnat Sukthuayat, Thaworn
Sisayanarane and Sorawut Laoratanavoraphong
Wanna Kittisupaluck Orthopedics Department, Police General Hospital,
The Surgical Unit, Thammasat University Hospital, Bangkok, Thailand
Pathum Thani, Thailand
Wichan Kanchanatawan
David Metcalfe Orthopedics Department, Lerdsin General Hospital,
Harvard Medical School, 25 Shattuck Street, Boston, MA Bangkok, Thailand
02115, USA
Warwick Medical School, Gibbet Hill Road, Coventry Jatupon Kongtharvonskul
CV4 7AL, UK Section for Clinical Epidemiology and Biostatistics,
Division of Trauma, Burns, and Surgical Critical Care, Faculty of Medicine, Ramathibodi Hospital, Bangkok,
Brigham and Women’s Hospital, 75 Francis Street, Thailand
Boston, MA 02115, USA
Tyler Barker, Victoria E. Rogers, Vanessa T. Henriksen
Craig J. Hickson and Kimberly B. Brown
Leicester Royal Infirmary, Infirmary Square, Leicester The Orthopedic Specialty Hospital, 5848 S. Fashion Blvd.,
LE1 5WW, UK Murray, UT 84107, USA
Lesley McKee Roy H. Trawick, Nathan G. Momberger and G. Lynn

Forth Valley Hospital, Stirling Road, Larbert, Scotland, Rasmussen
UK The Orthopedic Specialty Clinic, 5848 S. Fashion Blvd,
Murray, UT 84107, USA
Xavier L. Griffin
Warwick Medical School, Gibbet Hill Road, Coventry Giovanni Merolla
CV4 7AL, UK Unit of Shoulder and Elbow Surgery, D. Cervesi Hospital,
Cattolica (RN) - AUSL della Romagna Ambito Territoriale
Kunihiko Hiramatsu di Rimini, Italy
Department of Orthopaedic Surgery, Yao Municipal Biomechanics Laboratory ‘‘Marco Simoncelli’’, D. Cervesi
Hospital, 1-3-1, Ryugecho, Yao, Osaka, Japan Hospital, Cattolica (RN) - AUSL della Romagna Ambito
Territoriale di Rimini, Italy
Yasukazu Yonetani, Kazutaka Kinugasa and Masayuki
Hamada Sanjay Singh, Paolo Paladini and Giuseppe Porcellini
Department of Orthopaedic Surgery, Hoshigaoka Medical Unit of Shoulder and Elbow Surgery, D. Cervesi Hospital,
Center, 4-8-1, Hoshigaoka, Hirakata, Osaka, Japan Cattolica (RN) - AUSL della Romagna Ambito Territoriale
di Rimini, Italy
Norimasa Nakamura
Institute for Medical Science in Sports, Osaka Health Su Chan Lee, Viral Gondalia, Byoung Yoon Hwang,
Science University, Osaka, Japan Hye Sun Ahn, Choon Key Lee and Kwang Am Jung
Center for Advanced Medical Engineering and Informatics, Joint and Arthritis Research, Department of Orthopaedic
Osaka University, Suita, Osaka, Japan Surgery, Himchan Hospital, 20-8, Songpa-dong, Songpa-
gu, Seoul 138-170, Korea
Koji Yamamoto
Department of Orthopaedic Surgery, Toyonaka Municipal David J. Hunter
Hospital, 4-14-1, Shibahara, Toyonaka, Osaka, Japan Rheumatology Department, Royal North Shore Hospital
and Northern Clinical School, University of Sydney,
Hideki Yoshikawa Sydney, NSW, Australia
Department of Orthopaedic Surgery, Osaka University
Graduate School of Medicine, 2-2, Yamada-oka, Suita, Dane Salazar, Sean Lannon, Adam Schiff and Erika
Osaka 565-0871, Japan Mitchell
The Department of Orthopaedic Surgery and
Rehabilitation, Loyola University Health System, 2160
South First Avenue, Maywood, IL 60153, USA

260 /LVWRI&RQWULEXWRUV
Olga Pasternak and Laurie Lomasney F. Wichlas, N. P. Haas, A. Disch and D. Machó
The Department of Radiology, Loyola University Health Center for Musculoskeletal Surgery,
System, 2160 South First Avenue, Maywood, IL 60153, Charite´,Universitätsmedizin Berlin, Augustenburger
USA Platz 1, 13353 Berlin, Germany
Michael Stover S. Tsitsilonis

Department of Orthopaedic Surgery, Feinberg School of Berlin-Brandenburg Center of Regenerative Therapies,
Medicine, Northwestern University, 676 N. Saint Clair, Charité,Universitätsmedizin Berlin, Berlin, Germany
Suite 1350, Chicago, IL 60611, USA
Haroon Majeed
Federico Mancini, Luca Garro and Roberto Caterini Trauma and Orthopaedics, University Hospital of North
Department of Orthopaedics and Traumatology, Staffordshire, Newcastle Road, Stoke-on-Trent ST4 6QB,
University of Rome ‘‘Tor Vergata’’, Viale Oxford, 81, England, UK
00133 Rome, Italy
Keikichi Kawasaki, Tetsuya Nemoto, Katsunori Inagaki,
Andrea Ascoli-Marchetti Kazunari Tomita and Yukio Ueno
Department of Vascular Surgery, University of Rome Department of Orthopaedic Surgery, Showa University
‘‘Tor Vergata’’, Rome, Italy School of Medicine, 1-5-8 Hatanodai, Shinagawa-ku,
Tokyo 142-8666, Japan
Luigi Tarallo, Raffaele Mugnai and Fabio Catani
Department of Orthopaedic Surgery, University of Matthew R. McDonald, Vasanth Sathiyakumar, Jordan
Modena and Reggio Emilia, Via del Pozzo 71, 41124 C. Apfeld, Benjamin Hooe, Jesse Ehrenfeld, William T.
Modena, Italy Obremskey and Manish K. Sethi
The Vanderbilt Orthopaedic Institute Center for Health
Roberto Adani Policy, Vanderbilt University, Suite 4200, South Tower,
Department of Hand Surgery, University Hospital of MCE, Nashville, TN 37221, USA
Verona, Verona, Italy

Orthopedic Trauma (2016)

Uploaded by

Copyright:

Available Formats

Orthopedic Trauma (2016)

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

Orthopedic Trauma (2016)

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What topics are covered in the book chapters?

What topics are covered in the book chapters?

Orthopedic Trauma

Orthopedic Trauma: Diagnosis, Operative Techniques and Management

© 2017 Academic Pages

International Standard Book Number: 978-1-9789-2749-0

Chapter 2 Adjacent segment infection after surgical treatment of spondylodiscitis 6

Chapter 3 Single intra-articular injection of high molecular weight hyaluronic

Chapter 5 Outcomes of intramedullary nailing for acute proximal humerus fractures:

Chapter 6 Mid-term results of Miller-Galante unicompartmental knee replacement

Chapter 7 An unusual case of neurothekeoma of the arm in an adult 49

Chapter 9 Effectiveness of intra-articular injections of sodium hyaluronate-chondroitin

__________________________ WORLD TECHNOLOGIES __________________________

Chapter 11 A modified Austin/chevron osteotomy for treatment of hallux valgus

Chapter 13 Biocompatibility of magnesium implants in primary human reaming

Chapter 14 Correlating preoperative imaging with intraoperative fluoroscopy in iliosacral

Chapter 15 Cross-cultural adaptation and validation of the International Knee

Chapter 16 Effectiveness of extracorporeal shockwave therapy in three major

Chapter 17 Arthrodesis of proximal inter-phalangeal joint for hammertoe: intramedullary

Chapter 18 FiberWire tension band for patellar fractures 122

Chapter 21 Effectiveness of psychological support in patients undergoing primary total hip

__________________________ WORLD TECHNOLOGIES __________________________

Chapter 22 Blood metal ions after hybrid metal-on-polyethylene Exeter—Trident total

Chapter 23 Intraosseous concentration and inhibitory effect of different intravenous

Chapter 26 Isolated reconstruction of the medial patellofemoral ligament with autologous

Chapter 28 Is there a link between the neutrophil-to-lymphocyte ratio and venous

Chapter 30 Is end-stage lateral osteoarthritic knee always valgus? Mechanical alignment

Chapter 33 Malunited extra-articular distal radius fractures: corrective osteotomies using

__________________________ WORLD TECHNOLOGIES __________________________

Chapter 35 Non-operative treatment versus percutaneous fixation for minimally displaced

Chapter 36 Variable-angle locking plate with or without double-tiered subchondral support

__________________________ WORLD TECHNOLOGIES __________________________

__________________________ WORLD TECHNOLOGIES __________________________

Elke Rometsch2 • Marco Favuto1 • Alessandro Stucchi1 • Matheus Azi3 •

Abstract not associated with any of the factors investigated. Poly-

__________________________ WORLD TECHNOLOGIES __________________________

__________________________ WORLD TECHNOLOGIES __________________________

Gender 0.639* 0.717*

__________________________ WORLD TECHNOLOGIES __________________________

Table 3 Univariate linear

__________________________ WORLD TECHNOLOGIES __________________________

__________________________ WORLD TECHNOLOGIES __________________________

Abstract Conclusions Adjacent segment infection after surgical

__________________________ WORLD TECHNOLOGIES __________________________

__________________________ WORLD TECHNOLOGIES __________________________

Table 1 Demography, comorbidities, primary presentation, neurological function and FU outcomes

__________________________ WORLD TECHNOLOGIES __________________________

level of the lesion but motor paralysis complete below

the lesion but of no practical use to the patient.

4. ‘Motor Useful’ (D). Useful motor power below the

5. ‘Recovery’ (E). Free of neurological symptoms.

The final functional outcome was completed by ques-

tionnaires including Odom’s criteria [12] which catego-

rized patients’ satisfaction into four grades: excellent,

good, fair and poor.

– Excellent: all preoperative symptoms relieved, abnor-

– Good: minimum residual of preoperative symptoms not

requiring medication or limiting activity, and abnormal

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES

WORLD TECHNOLOGIES