Pediatric Orthopedic Deformities, Volume 2: Developmental Disorders of the Lower Extremity: Hip to Knee to Ankle and Foot [1st ed.] 978-3-030-02019-4, 978-3-030-02021-7

Volume 2 of this comprehensive and state-of-the-art text on pediatric orthopedic deformities focuses on conditions of th

470 28 48MB

English Pages XXIX, 813 [839] Year 2019

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

Lower Extremity Joint Preservation: Techniques for Treating the Hip, Knee, and Ankle 3030573818, 9783030573812

This book describes and discusses the available joint preservation techniques for maintaining the stability homeostasis

712 136 16MB Read more

Imaging Anatomy. Knee. Ankle. Foot [2 ed.] 9780323477802

469 34 58MB Read more

Systemic Disease Manifestations in the Foot, Ankle, and Lower Extremity 9781496388483, 9781451192643

Recognize when lower extremity disorders signal a larger problem. Systemic Disease Manifestations in the Foot, Ankle, an

164 53 16MB Read more

Imaging Anatomy: Knee, Ankle, Foot [2 ed.] 0323477801, 9780323477802

Designed to help you quickly learn or review normal anatomy and confirm variants, Imaging Anatomy: Knee, Ankle, Foot , b

967 183 36MB Read more

The Pediatric Foot and Ankle: Diagnosis and Management 9783030297886, 3030297888

The Pediatric Foot and Ankle is a comprehensive overview of the current evaluation, diagnosis, and management strategie

716 100 109MB Read more

Orthopaedic Trauma Surgery: Volume 2: Lower Extremity Fractures and Dislocation [1st ed. 2023] 981160214X, 9789811602146

This book adopts the principle of guiding surgery by anatomy, fixation by biomechanics, and clinical procedures by funct

342 117 79MB Read more

Foot and Ankle Radiology [2 ed.] 1451192835, 9781451192834

Clearly written and abundantly illustrated, Foot and Ankle Radiology, 2nd Edition provides a solid understanding of diag

1,469 165 100MB Read more

Personalized Hip and Knee Joint Replacement [1st ed.] 9783030242428, 9783030242435

This open access book describes and illustrates the surgical techniques, implants, and technologies used for the purpose

498 82 24MB Read more

AOSpine Masters Series, Volume 9: Pediatric Spinal Deformities 9781626234543, 162623454X

An estimated 9 million children every year are affected by pediatric spinal deformities, encompassing a broad spectrum o

366 102 106MB Read more

Foot and Ankle Biomechanics 9780128154496, 0128154497

Foot and Ankle Biomechanics is a one source, comprehensive and modern reference regarding foot and ankle biomechanics. T

215 49 104MB Read more

Pediatric Orthopedic Deformities, Volume 2: Developmental Disorders of the Lower Extremity: Hip to Knee to Ankle and Foot [1st ed.]
978-3-030-02019-4, 978-3-030-02021-7

Author / Uploaded
Frederic Shapiro

Table of contents :
Front Matter ....Pages i-xxix
Developmental Dysplasia of the Hip (Frederic Shapiro)....Pages 1-182
Legg-Calvé-Perthes Disease (Frederic Shapiro)....Pages 183-322
Slipped Capital Femoral Epiphysis: Developmental Coxa Vara (Frederic Shapiro)....Pages 323-434
Femoroacetabular Impingement (Frederic Shapiro)....Pages 435-472
Developmental Disorders of the Knee (Frederic Shapiro)....Pages 473-604
Torsional, Angular, and Deficiency Disorders of the Lower Extremity (Frederic Shapiro)....Pages 605-664
Developmental Disorders of the Foot and Ankle (Frederic Shapiro)....Pages 665-797
Back Matter ....Pages 799-813

Citation preview

Frederic Shapiro

Pediatric Orthopedic Deformities Volume 2

Developmental Disorders of the Lower Extremity: Hip to Knee to Ankle and Foot

123

Pediatric Orthopedic Deformities, Volume 2

Frederic Shapiro

Pediatric Orthopedic Deformities, Volume 2 Developmental Disorders of the Lower Extremity: Hip to Knee to Ankle and Foot

Frederic Shapiro, MD Visiting Scholar, Stanford University School of Medicine Department of Medicine/Endocrinology (Bone Biology) Palo Alto, CA USA Formerly, Associate Professor of Orthopaedic Surgery Harvard Medical School, Boston Children’s Hospital Boston, MA USA

ISBN 978-3-030-02019-4 ISBN 978-3-030-02021-7 (eBook) https://doi.org/10.1007/978-3-030-02021-7 Library of Congress Control Number: 2015943442 © Springer Nature Switzerland AG 2019 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

To my wife, Carol Ann Satler

Preface

Pediatric Orthopedic Deformities, Volume 2: Developmental Disorders of the Lower Extremity – Hip to Knee to Ankle and Foot is composed of an Introduction and seven chapters. It focuses on the hip with chapters on developmental dysplasia of the hip (DDH), Legg- Calvé-Perthes disease (LCP), coxa vara including slipped capital femoral epiphysis (SCFE), and femoroacetabular impingement (FAI); disorders affecting the knee; rotational and angular deformities of the lower limb including lesions centered at the diaphyseal-metaphyseal regions; and disorders of the foot and ankle including club foot and congenital vertical talus. Volume 1 of Pediatric Orthopedic Deformities covered several topics1, including lower extremity length discrepancies, as well as a detailed overview of the developmental biology of the skeletal system, an overview of how altered biology contributes to causation of deformity, and how the utilization of biologic and mechanical principles leads to correction of those deformities. Understanding epiphyseal and physeal biology is essential owing to its contribution to normal growth and development, to pathologic deformity, and to correction of deformity with growth. Volume 3 will discuss pediatric neuromuscular disorders and the treatment of neuromuscular, congenital, and syndromic scoliosis. In the Introduction to Volume 2, we have provided a Definition of Deformity, a formal list of the 20 General Principles Regarding Pediatric Orthopedic Deformity (40 including subdivisions), and management Implications of the General Principles of Deformity. For each deformity in the seven chapters, we provide a definition (terminology), detailed review of the pathoanatomy, experimental biological investigations (where applicable), natural history, review of the evolution of diagnostic and treatment techniques, results achieved with the various approaches, and the current management approaches (in text and tabular form) including detailed descriptions of surgical technique. The book is extensively illustrated to show the range of deformity for the various disorders, the underlying histopathology from human cases (and experimental models where available), imaging findings, and treatment approaches. This broad approach provides an extensive knowledge base regarding differing diagnostic methods, a detailed review of the underlying pathoanatomy of the disorder, the stage in its progression, the range of treatments, and their effectiveness. This combined information for each disorder improves the likelihood that the specific procedure or management approach chosen is applied at the correct time. The two underlying premises of this volume remain the same as expressed in the preface to Volume 1. These are that (i) current orthopedic treatments of deformities of the developing musculoskeletal system are most effective when based on understanding and relating to the underlying pathobiology and (ii) future treatments are best developed by directly addressing the primary pathobiology.

Chapters in Volume 1: (1) Developmental Bone Biology; (2) Overview of Deformities; (3) Skeletal Dysplasias; (4) Bone and Joint Deformity in Metabolic, Inflammatory, Neoplastic, Infectious, and Hematologic Disorders; (5) Epiphyseal Growth Plate Fracture-Separations; and (6) Lower Extremity Length Discrepancies. A complete listing of the chapter content subsections can be found on the Springer website [springer.com] listing for Pediatric Orthopedic Deformities, Volume 1. 1

vii

viii

These premises are by no means original; as long ago as 1843, William Little, MD of London, England, stressed repeatedly in his “Course of Lectures on Deformities of the Human Frame” that “…. you can never treat a deformity with advantage to the patient or to your own satisfaction….unless you thoroughly understand the pathology of the case” (Lancet 1843; 41: 382-386). He published his course of 18 lectures on deformities in the Lancet in 1843–1844 and collected them in book form in 1853 (Lectures on Deformities of the Human Frame, London, Longman, Brown, Green and Longmans). For each entity discussed throughout our book, the underlying pathology is described in detail. In the current environment, however, these basic premises while generally adhered to verbally are at risk of becoming overwhelmed by the flood of information published in a proliferating number of journals and discussed at innumerable courses. The current concentration on “best practices,” “expert opinions,” “evidence-based recommendations,” “committee recommendations,” “consensus reports,” “peer-review committees,” et cetera all provide meaningful direction for practitioners but risk taking the focus away from more primary studies. Discussion and formulation of “best practices,” “evidence-based” approaches, etc. are to be encouraged; they are in fact derived from most of the same reference sources in the various chapters of the book and are included in the information base provided in the book. Bearing these considerations in mind, Volume 2 of Pediatric Orthopedic Deformities is designed (as was Volume 1) to provide the pediatric orthopedic surgeon, and those managing pediatric patients with orthopedic deformities, with the detailed knowledge base needed to manage patients independent of simply following consensus profiles. It also provides the detailed pathobiologic background needed to guide the evolving molecular, cellular, and biophysical approaches to managing pediatric orthopedic deformity. The biologic and biophysical focus of the book provides clear understanding of investigations directed at major sites of clinical deformity. For example: • Over the past two decades, significant strides have been made in understanding the pathogenesis and effects of avascular necrosis of the femoral head, primarily using experimental piglet models where ischemia is induced by intracapsular circumferential ligation at the base of the femoral neck. Subsequent studies with the model have improved diagnostic methods, led to understanding of both femoral head and secondary acetabular malformation, and helped assess molecular treatment interventions. • Appreciation of malformations at the femoral head/acetabular interface leading to femoroacetabular impingement (FAI) has had major treatment implications for several disorders, particularly slipped capital femoral epiphysis. There was increasing awareness for several decades that hip osteoarthritis was rarely idiopathic but secondary to childhood hip deformity, even if mild; structural definition of the altered femoral-acetabular relationship, however, clarified the causes and led to the development of corrective interventions. • Osteochondritis dissecans at the knee and talus are now addressed primarily via arthroscopy. Earlier intervention allows for limiting the damage done and, in many instances, for primary repair; severe involvement can now be addressed by attempting to induce articular cartilage repair by biologic cellular and tissue approaches. The book is constructed to allow for inclusion of a knowledge base of the underlying pathoanatomy, natural history of the various disorders, and awareness of treatments that have had some effectiveness in the past as well as a detailed presentation of current treatment programs. While using treatments that experts or multicenter committees are recommending can be comforting and tends to raise the consistency of results, awareness of the history of management trends shows that using this approach to management exclusively can be shortsighted. It is a combination of application of knowledge of the underlying pathology of a disorder and appropriate utilization of biomechanical and biological principles in treating the disorder that will ultimately improve the results.

Preface

Preface

ix

The continual change of management profiles in essentially all pediatric orthopedic disorders over relatively short periods of time cannot be attributed solely to a positive unidirectional flow of improvement. Considerable effort has been made in the book reviewing the course of management over several decades. This is not done as a simple historical exercise; rather it indicates the evolution of treatments pointing out where previous efforts were inadequate. Even where surgical techniques from one era are found not to be required as frequently now, awareness of a technique and its value can be applied fully or partially where newer approaches still leave deformity uncorrected. While some of the older operative procedures are rightly abandoned, others remain of value and need to be understood. Historical review becomes even more meaningful by also showing the cyclical nature of many management approaches where treatments abandoned as inadequate resurface decades later as treatments of choice. For example: • Percutaneous tendoAchilles tenotomy for clubfoot deformity, following initial repetitive manipulation to correct the varus/adduction component and followed by lengthy periods of splinting and gentle daily manipulation to maintain the correction, was widely used by Stromeyer in Germany, beginning in the early 1830s, followed very shortly by Little in England, Guérin in France, and many others. This approach had considerable success over several decades and included the procedure that effectively launched the surgical component of pediatric orthopedic surgery. This treatment program subsequently fell into disregard to be followed by several decades of forceful manipulation for clubfoot deformity and a series of nonphysiologic open surgical procedures that, while resulting in apparently straight feet, caused considerable stiffness and deformity necessitating repetitive procedures. Even when Ponseti revived the initial manipulative/casting approach, the almost invariable use of percutaneous tendoAchilles tenotomy for correcting clubfoot equinus, followed by 2–3 years of night splinting, it again took a couple of decades for it to gain its current wide acceptance. • For symptomatic Osgood-Schlatter disease of the knee nonresponsive to conservative management during the growth years, Makins in 1905 described a good result with a simple surgical procedure at skeletal maturity where the loose “osteocartilaginous nodule” of the tibial tubercle was removed and the soft tissues were re-apposed and sutured to the tibia (Lancet 1905; 166: 213-216). Over the next several decades, and continuing to the present, innumerable operative approaches other than simple loose ossicle removal described by Makins were used. These included bone drilling or autogenous bone grafting to get the ossicle to heal, insertion of ivory pegs at the tibial tubercle site to enhance fusion, excision of the tibial tubercle, longitudinal incision in the patellar tendon to relieve venous hypertension, decompression of tissue in the tubercle by arthroscopy, and (once again) simple removal of the ossicles at open incision. At present, most will now perform the procedure essentially described by Makins (removing loose ossicles) that yields rapid repair. This continuing circularity of management approaches is a feature of several of the conditions discussed in the book. • Extensive clinical and experimental efforts beginning with Ollier in France in 1867 and prominent from the 1930s to the 1960s were done to stimulate long bone growth for limb length discrepancies by several methods: irritating the periosteum on the shorter side by subperiosteal stripping, elevating it with foreign objects, or cutting it circumferentially. (See Volume 1, Chap. 6, Sect. 6.9.3.) The resulting repair with increased vascularity stimulated the physes of the operated bones to overgrowth, and results were sometimes effective (0.5–2 cm overgrowth), but good responses were irregular and unpredictable and the techniques were abandoned. At present, there is renewed experimentation that cuts the periosteum circumferentially by non-operative means to induce overgrowth for unilateral limb discrepancies.

x

The questions raised by these examples relate to why, when results were seemingly so good in occasional cases, or at least worked to a certain degree in many, the profession widely abandoned the approaches instead of continuing to refine them with modifications. It is knowledge of the underlying pathoanatomy and the ability to deal with it in appropriate biological and biomechanical (biophysical) ways that sooner or later allows the correct approach to be used. The natural history of the various deformities is described in detail. This too is not provided in a routine or automatic fashion; rather, combined with considerations of the underlying pathoanatomy, the two provide major signals regarding the timing for specific interventions as well as indications that observation alone may allow for spontaneous repair. Operations for a specific disorder may be “correct” (based on current understanding) in that they are applied for that specific disorder but could be considered to have been performed at the wrong time, too late to yield a meaningful long-term result or too early, where good evidence exists that spontaneous growth-related correction alone would most likely have caused improvement. It is the application of biological and biomechanical treatment principles that allows for optimal management. This especially applies to understanding the growth mechanism at each region that plays a major role in both causing and correcting pediatric musculoskeletal deformity.

Preface

Acknowledgments

The author gratefully acknowledges the efforts of several individuals who helped bring this work to publication: Thy Thy Le, Phison Le, and Theresa Bui for the manuscript preparation; Michael Griffin for his continuing detailed and extensive work as Springer Developmental Editor; Kristopher Spring, Senior Editor, Springer Nature, for editorial management and direction at each stage of the publication process; the Springer artists for excellent help with the illustrations; and the Springer publication team for the final production of the book. He thanks Joy Wu, MD, PhD, Department of Medicine/Endocrinology, for appointment in her bone biology laboratory as a Visiting Scholar at Stanford University School of Medicine, Palo Alto, CA. The author is a pediatric orthopedic surgeon who worked at the Boston Children’s Hospital, Boston MA USA, clinically in the Department of Orthopaedic Surgery (attending orthopedic surgeon) and doing basic science research in the Orthopaedic Research Laboratory (Laboratory for the Study of Skeletal Disorders). His Harvard Medical School, Boston MA USA, appointments progressed from Research Fellow to Instructor, then Assistant Professor of Orthopaedic Surgery, and ultimately Associate Professor of Orthopaedic Surgery.

xi

Contents

1 Developmental Dysplasia of the Hip�� 1 1.1 Terminology�� 1 1.1.1 Change in Terminology�� 1 1.2 Development of the Hip: Embryonic and Fetal Periods �� 2 1.2.1 Earliest Developmental Biology of the Hip. Chick Embryo Studies�� 2 1.2.2 General Aspects of Human Hip Development�� 2 1.2.3 Embryonic, Fetal, and Postnatal Development of the Femur �� 10 1.2.4 Embryonic, Fetal, and Postnatal Development of the Acetabulum �� 11 1.2.5 Embryonic, Fetal, and Postnatal Development of the Acetabular Labrum (Glenoid of the Hip)�� 13 1.3 Primary Etiologies of Hip Maldevelopment �� 14 1.4 Etiology and Pathoanatomy of Developmental Dysplasia of the Hip�� 15 1.4.1 Early Clinical-Pathoanatomic Descriptions�� 15 1.4.2 Later Clinical-Pathoanatomic Descriptions�� 35 1.4.3 Subsequent Clinical-Pathoanatomic Descriptions with Emphasis on Early Capsular Laxity�� 39 1.4.4 Multifactorial Causes of DDH Involving Late-Stage Structural Modifications of the Hip, Mesenchymal Tissue Abnormalities, and Intrauterine Mechanical Stresses due to Positioning (Wilkinson, Dunn, Seringe et al.)�� 48 1.5 Experimental Reproduction of Hip Dislocation �� 50 1.5.1 Developmental Changes in the Acetabulum Following Experimental Displacement of the Femoral Head During Early Growth�� 50 1.5.2 Breech Malposition and Hormonal Laxity Causing Hip Dislocation in Young Rabbits�� 51 1.5.3 Mechanical Induction of Hip Deformation and Dislocation In Vitro�� 53 1.6 Epidemiology and Its Relation to Pathophysiology�� 53 1.6.1 Sex Incidence�� 53 1.6.2 Incidence and Side of Hip Instability�� 53 1.6.3 Effects of Intrauterine Environment�� 53 1.6.4 Extrauterine Postnatal Environment �� 55 1.6.5 Genetic Considerations �� 55 1.6.6 Ethnic Considerations �� 55 1.6.7 Spontaneous Stabilization of Hips Without Treatment�� 55 1.6.8 Absence of Ligamentum Teres in DDH�� 56 1.7 Summary of Intrinsic and Extrinsic Environmental and Pathoanatomic Findings in DDH: Discussion of Pathogenetic Sequences �� 56 1.8 Natural History of Hip Dislocations, Subluxations, and Dysplasia�� 59 xiii

xiv

Contents

1.8.1 Natural History of Complete Dislocations�� 59 1.8.2 Natural History of Dysplasia and Subluxation �� 60 1.8.3 Osteoarthritis in Adult Life Following Childhood CDH/DDH�� 60 1.9 Brief History of Treatment Approaches in Developmental Dysplasia of the Hip�� 61 1.9.1 Gradual Development of Reasonably Effective Closed and Open Treatments�� 62 1.10 Progressively Earlier Diagnosis and Treatment of Congenital Dislocation of the Hip�� 69 1.10.1 Hilgenreiner�� 70 1.10.2 Putti�� 71 1.10.3 Ortolani �� 71 1.10.4 Von Rosen and Barlow�� 71 1.10.5 Widespread Adoption of Neonatal Hip Examination �� 72 1.11 Assessments of DDH Treated by Closed Reduction�� 72 1.11.1 Radiographic Classification System (Severin)�� 72 1.11.2 Measurement of the CE Angle�� 72 1.11.3 Severin Classification for Radiographic Assessment of Long-Term Results �� 73 1.12 The Development of Modern Treatment for CDH and DDH �� 74 1.12.1 Hip Reduction and Stabilization in the Early Weeks of Life: “Functional” Method �� 74 1.12.2 Treatment by Closed Reduction�� 78 1.12.3 Treatment by Open Reduction�� 80 1.12.4 Acetabular Corrective Procedures for Treatment of Hip Dysplasia �� 83 1.12.5 Proximal Femoral Osteotomies�� 97 1.12.6 Comparison of Acetabular Development Following Open Reduction Combined with Innominate or Femoral Varus-Derotation Osteotomy�� 97 1.12.7 Combined Acetabular and Proximal Femoral Osteotomies�� 98 1.12.8 Timing of Osteotomies in Relation to Closed or Open Reductions�� 99 1.13 Imaging Techniques Used to Assess Hip Position�� 99 1.13.1 Plain Radiographic Indices �� 99 1.13.2 Arthrography in Assessing Hip Position and Anatomy�� 108 1.13.3 Ultrasonography in the Diagnosis of Newborn Developmental Dysplasia of the Hip�� 112 1.13.4 Long-Term Studies of Sonographic Indices in Normal and Abnormal Hip Development�� 126 1.13.5 CT Scan to Assess Hip Structure and Position �� 127 1.13.6 MR Imaging to Assess Position and Vascularity of the Femoral Head Postreduction�� 128 1.14 Assessments of Hip Growth and Development Following Closed and Open Treatments�� 129 1.14.1 Growth and Development of the Hip Following Closed Reduction in Early Infancy �� 129 1.14.2 Acetabular Development Following Hip Reduction by Closed, Open, or Varus Osteotomy Treatments �� 129 1.14.3 Acetabular Dysplasia and Its Implications for Early Degenerative Joint Disease �� 131 1.14.4 Acetabular Development After Removal of the Limbus in Infancy�� 131

Contents

xv

1.14.5 Acetabular Growth and Positioning Following Acetabular Surgery �� 132 1.14.6 Growth Disturbance Lines in Proximal Femur: O’Brien�� 133 1.14.7 Proximal Femoral Growth Following Femoral Osteotomy�� 133 1.15 Current Treatment Based on Underlying Pathoanatomy, Including Secondary Changes�� 133 1.15.1 General Overview �� 133 1.15.2 Diagnosis Made in the Newborn Period �� 134 1.15.3 Diagnosis Made at 3 Months of Age�� 135 1.15.4 Diagnosis Made at 6 Months of Age�� 135 1.15.5 Diagnosis Made at 12 Months of Age�� 136 1.15.6 Diagnosis Made at 18 Months of Age�� 136 1.15.7 Diagnosis Made Between 18 Months to 2 Years and 4.5 Years of Age �� 138 1.15.8 Imperfect Hip Structure After 5 Years of Age�� 138 1.16 Avascular Necrosis as a Complication of Treatment of Developmental Dysplasia of the Hip�� 138 1.16.1 Blood Supply of the Proximal Femur �� 138 1.16.2 Epiphyseal Blood Supply: Cartilage Canals �� 147 1.16.3 Recognition of the Problem of Avascular Necrosis as a Complication of Treatment for Developmental Dysplasia of the Hip�� 148 1.16.4 Efforts at Understanding and Treating the Causes of Avascular Necrosis�� 149 1.16.5 Classification of Patterns of Avascular Necrosis Following Treatment of Developmental Dysplasia of the Hip �� 154 1.16.6 Avascular Necrosis Associated with Immobilization Devices Other Than a Hip Spica�� 158 1.16.7 More Recent Reports of the Incidence of AVN in Developmental Dysplasia of the Hip�� 158 1.16.8 Reports of AVN After 2000�� 161 1.16.9 MR Imaging to Detect Hip Ischemia due to Extreme Immobilization Positioning�� 164 References�� 168 2 Legg-Calvé-Perthes Disease�� 183 2.1 Definition�� 183 2.2 Original Recognition of Disorder �� 183 2.2.1 General Review�� 183 2.2.2 Legg�� 184 2.2.3 Calvé �� 184 2.2.4 Perthes�� 186 2.2.5 Waldenström �� 186 2.2.6 Sourdat�� 187 2.3 Clinical Profile�� 187 2.3.1 General Features �� 187 2.3.2 Epidemiology of Legg-Calvé-Perthes Disease �� 189 2.4 Early Pathologic Reports of Cell and Tissue Changes in Legg-Calvé-Perthes Disease �� 193 2.4.1 Zemansky, 1928�� 193 2.4.2 Schwarz, 1914�� 194 2.4.3 Phemister, 1920�� 194 2.4.4 Axhausen, 1923�� 194

xvi

Contents

2.4.5 Heitzmann, 1923 �� 195 2.4.6 Riedel, 1923�� 196 2.4.7 Konjetzny, 1926/1934 �� 196 2.4.8 Delchef, 1926�� 196 2.4.9 Rockemer, 1927�� 196 2.4.10 Lippmann, 1929�� 196 2.4.11 Nagassaka, 1930 �� 198 2.4.12 Summary of Histopathologic Changes After Two Decades of Study (Zemansky)�� 198 2.5 Subsequent Pathologic Reports with Better Defined Clinical and Radiographic Correlations�� 199 2.5.1 Ferguson and Howorth, 1934�� 199 2.5.2 Gall and Bennett, 1942 �� 200 2.5.3 Haythorn, 1949 �� 200 2.5.4 Jonsater, 1953�� 201 2.5.5 Ponseti, 1956�� 202 2.5.6 Mizuno, Hirayama, Kotani, and Simazu, 1966�� 202 2.5.7 Dolman and Bell, 1973 �� 203 2.5.8 Larsen and Reimann, 1973 �� 203 2.5.9 McKibbin and Ralis, 1974, and McKibbin, 1975�� 204 2.5.10 Jensen and Lauritzen, 1976�� 205 2.5.11 Inoue, Freeman, Vernon-Roberts, and Mizuno, 1976 �� 206 2.5.12 Inoue, Ono, Takaoka, Yoshioka, and Hosoya, 1980�� 206 2.5.13 Catterall et al., a and b, 1982�� 206 2.5.14 Ponseti et al., 1983�� 208 2.6 Experimental Piglet Models Reproducing Femoral Head Avascular Necrosis�� 208 2.6.1 Overview�� 208 2.6.2 Piglet Model: Structural Changes, Histologic Changes, and Magnetic Resonance Imaging Assessment�� 215 2.7 Pathoanatomic Changes and Their Relation to the Clinical, Radiologic, and Other Imaging Findings�� 225 2.7.1 Overview of Plain Radiographic Changes in Legg-Perthes Disease�� 225 2.7.2 Pathologic Changes and Their Demonstration by Varying Imaging Modalities Including Ultrasonography, Scintigraphy, Magnetic Resonance Imaging, and Computerized Axial Tomography�� 229 2.7.3 Subsequent Pathologic Changes Presenting as a Relative Decrease in Size of the Involved Secondary Ossification Center �� 234 2.7.4 Nutrition of the Proximal Femoral Epiphysis and Its Bearing on Legg-Calvé-Perthes Disease�� 234 2.7.5 Gage Sign/Catterall Sign: Lateral-Proximal Neck Convexity/Lateral Epiphyseal Lysis �� 234 2.7.6 Subchondral Fracture: Crescent Sign�� 234 2.7.7 Increased Radiodensity of the Secondary Ossification Center �� 238 2.7.8 Alternating Areas of Radiodensity and Radiolucency�� 238 2.7.9 Responses of the Cartilage Model of the Femoral Head�� 240 2.7.10 Assessment of Cartilage Model of Proximal Femur Using Arthrography�� 244 2.7.11 Responses of the Physis�� 245 2.7.12 Sagging Rope Sign�� 246 2.7.13 Responses of the Femoral Neck (Metaphysis)�� 246

Contents

xvii

2.7.14 Femoral Neck Anteversion�� 249 2.7.15 Responses of the Greater Trochanter�� 249 2.7.16 Responses of the Acetabulum �� 249 2.7.17 Acetabular Retroversion �� 252 2.7.18 Remodeling in the Residual Phase of the Disease Between the Termination of Healing and Skeletal Maturity�� 252 2.7.19 Imperfect Healing of Legg-Calvé-Perthes with Persistence of an Osteochondritis Dissecans Lesion at Skeletal Maturity�� 253 2.7.20 Hinge Abduction: Imperfect Healing with a Flattened Femoral Head and a Superolateral Prominence Impeding Smooth Abduction�� 253 2.7.21 Femoral Shortening as a Sequel to Legg-Perthes Disease�� 253 2.8 Lower Extremity Length Discrepancies with Legg-Perthes Disease�� 255 2.8.1 Maximum Total Femoral and Tibial Discrepancy During Growth Years�� 255 2.8.2 Femoral and Tibial Discrepancy at Skeletal Maturity�� 256 2.8.3 Maximum Femoral Discrepancy�� 256 2.8.4 Maximum Tibial Discrepancy�� 256 2.8.5 Developmental Patterns of Discrepancies in Legg-Perthes Disease�� 256 2.9 Prognostic Indicators During the Active Disease Process�� 257 2.9.1 General Considerations�� 257 2.9.2 Age of Occurrence of the Disease�� 258 2.9.3 Plain Radiographic Classifications �� 259 2.9.4 Comparison of Classification Schemes�� 266 2.9.5 Expansion of Waldenström Grading System: Joseph et al.�� 267 2.10 Classifications Defining Results at Skeletal Maturity at the End of Repair�� 268 2.10.1 General Considerations�� 268 2.10.2 Sundt Classification�� 268 2.10.3 Stulberg Classification�� 268 2.10.4 Butel, Borgi, and Oberlin Grading System�� 269 2.10.5 Quantitative Indices of Femoral Head: Acetabular Repair �� 269 2.10.6 Additional Long-Term Studies of Adult Results of Childhood Perthes�� 274 2.11 Treatment Approaches to Legg-Perthes Disease�� 276 2.11.1 Early Major Reviews of Treatment Approaches �� 276 2.11.2 Range of Approaches to the Disorder �� 278 2.11.3 Other Factors Concerning Results�� 301 2.11.4 Late-Stage Surgical Intervention to Treat the Sequelae of Legg-Perthes Disease �� 303 2.11.5 Summary of Treatment Approaches�� 309 References�� 314 3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara �� 323 3.1 Coxa Vara�� 323 3.1.1 Terminology�� 323 3.1.2 Causes of Coxa Vara �� 323 3.1.3 Clinical Presentation of Coxa Vara �� 325 3.1.4 Imaging Assessments in Coxa Vara�� 325 3.2 Slipped Capital Femoral Epiphysis�� 327 3.2.1 Terminology�� 327 3.2.2 Evolving Clinical Awareness and Description of the Disorder�� 327 3.2.3 Etiology of Slipped Capital Femoral Epiphyses�� 329

xviii

Contents

3.2.4 Pathoanatomy�� 331 3.2.5 Interpretation of the Studies on Pathogenesis and Pathoanatomy�� 340 3.2.6 Medical Disorders Predisposing to Slipped Capital Femoral Epiphysis�� 342 3.2.7 Types of Classification for SCFE�� 346 3.2.8 Epidemiologic Characteristics of Slipped Capital Femoral Epiphysis: Age, Sex, Weight, Symptom Time, Bilaterality, and Associated Disorders�� 349 3.2.9 Diagnostic Imaging Studies�� 356 3.2.10 Treatment�� 358 3.2.11 More Detailed Review of Complications of Treatments for Slipped Capital Femoral Epiphysis �� 396 3.2.12 Long-Term Follow-Up Studies �� 401 3.3 Developmental Abnormalities of the Femur: Including Proximal Femoral Focal Deficiency (PFFD), Congenital Short Femur, and Infantile Coxa Vara�� 402 3.3.1 Terminology�� 403 3.3.2 Proximal Femoral Focal Deficiency �� 403 3.3.3 Clinical Characteristics �� 409 3.3.4 Pathoanatomic Findings�� 410 3.3.5 Treatment�� 411 3.3.6 Infantile Coxa Vara �� 412 3.3.7 Clinical and Radiographic Presentation of Infantile Coxa Vara �� 413 3.3.8 Pathoanatomy of Infantile Coxa Vara �� 414 3.3.9 Evolution of Radiographic Change�� 417 3.3.10 Pathomechanics of Deformity in Infantile Coxa Vara�� 418 3.3.11 Clinical-Radiographic Correlations�� 420 3.3.12 Management of Infantile Coxa Vara �� 420 References�� 426 4 Femoroacetabular Impingement�� 435 4.1 Brief Introduction to Femoroacetabular Impingement (FAI) �� 435 4.1.1 Terminology�� 435 4.1.2 Overview of Pathogenesis�� 435 4.1.3 Types of Impingement�� 435 4.2 Earliest Descriptions of Hip Disorders Consistent with FAI�� 435 4.2.1 Smith-Petersen�� 435 4.2.2 Royal Whitman �� 436 4.2.3 Vulpius and Stoffel�� 437 4.2.4 Heyman, Herndon, and Strong�� 437 4.3 Growing Awareness of Association Between Imperfect Healing of Childhood and Adolescent Hip Disorders and Their Association with Osteoarthritis of the Hip �� 438 4.3.1 Murray�� 438 4.3.2 Solomon�� 438 4.3.3 Harris, Stulberg, and Colleagues�� 438 4.3.4 Ganz and Colleagues�� 439 4.4 Development of Femoroacetabular Impingement into a Formal Pathomechanical Entity �� 439 4.5 Etiology�� 440 4.5.1 Structural Abnormality in Hip�� 440 4.5.2 Basic Groups of Disorders Predisposing to FAI in Childhood and Adolescence�� 440

Contents

xix

4.6 Pathomechanical Mechanisms Underlying FAI and the Pathoanatomic Sequelae�� 441 4.6.1 Acetabular Retroversion with Pincer Impingement�� 441 4.6.2 Head-Neck Offset with Cam Impingement�� 442 4.6.3 Pathoanatomy�� 447 4.7 Diagnosis�� 448 4.7.1 Clinical Awareness�� 448 4.7.2 Physical Examination�� 448 4.8 Imaging in Assessment of FAI�� 448 4.8.1 Plain Radiographs �� 449 4.8.2 CT Scans �� 451 4.8.3 MR Imaging�� 451 4.8.4 Direct Observation of Femoroacetabular Region �� 453 4.9 Specific Anatomic Studies in Relation to the FAI Entity�� 453 4.9.1 Acetabular Retroversion �� 453 4.9.2 Structure of the Labro-Acetabular Complex�� 453 4.9.3 Blood Supply of the Labrum�� 454 4.9.4 Blood Supply of the Proximal Femur and Its Relation to Open Surgical Dislocation of the Hip�� 454 4.9.5 Anatomic Study of the Acetabulum in Relation to Clinical Applications to Hip Arthroscopy�� 456 4.9.6 Frequency Associations of Abnormal Anatomic Findings and Development of Osteoarthritis�� 458 4.10 Additional Considerations of the FAI Entity Regarding Diagnosis and Treatment�� 458 4.10.1 Regarding Diagnosis�� 458 4.10.2 Regarding FAI Morphology on Hip Radiographs in Asymptomatic Persons�� 459 4.10.3 Regarding Relationship of FAI Findings to the Need for and Timing of Corrective Surgery �� 460 4.11 Treatment�� 461 4.11.1 Overview of Management Profile �� 461 4.11.2 Types of Surgical Treatment for Adolescent and Young Adult FAI�� 461 4.12 Additional Detail on Surgical Approaches at the Hip for FAI�� 462 4.12.1 Surgical Dislocation of the Hip�� 462 4.12.2 Hip Arthroscopy�� 468 4.12.3 Periacetabular Osteotomy �� 468 4.12.4 Trimming of the Anterior Acetabular Bony Rim�� 468 4.12.5 Refixation of Labrum to Acetabular Rim�� 468 4.12.6 Biologic Resurfacing of Damaged Articular Cartilage�� 470 4.13 Complications of Surgical Treatments for FAI �� 470 4.14 Results of Interventions�� 470 References�� 471 5 Developmental Disorders of the Knee�� 473 5.1 Normal Development�� 473 5.1.1 Knee Components�� 473 5.1.2 Histologic Features of Knee Development �� 473 5.1.3 Clinically Relevant Features of Knee Development �� 476 5.2 Normal Developmental Variability �� 479 5.2.1 Physiologic Genu Varum and Genu Valgum in Childhood�� 479 5.2.2 Normal Radiographic Developmental Variants of the Distal Femoral and Proximal Tibial Epiphyses �� 480

xx

Contents

5.3 Osteochondritis Dissecans of Distal Femur�� 483 5.3.1 Disease Profile�� 483 5.3.2 Historical Recognition and Definition of Osteochondritis Dissecans�� 483 5.3.3 Original Descriptions by Paget, Teale, and Koenig�� 483 5.3.4 Three Stages of the Disorder�� 484 5.3.5 Age of Occurrence�� 485 5.3.6 Regions of Involvement of Distal Femur�� 486 5.3.7 Etiology�� 486 5.3.8 Pathogenesis and Pathoanatomic Findings �� 488 5.3.9 Assessment for Stable and Unstable Lesions�� 495 5.3.10 Imaging for Osteochondritis Dissecans�� 495 5.3.11 Classifications for Osteochondritis Dissecans�� 498 5.3.12 Age at Occurrence, Treatment, and Relation to Healing�� 501 5.3.13 Treatment in Childhood OCD�� 503 5.3.14 Treatment in Young Adult OCD�� 516 5.4 Infantile Tibia Vara (Blount Disease)�� 516 5.4.1 Terminology�� 516 5.4.2 Clinical Profile of Infantile Tibia Vara�� 516 5.4.3 Description of the Disorder by Blount�� 516 5.4.4 Clinical-Radiographic Grading Scheme of Langenskiold: Stages I–VI�� 517 5.4.5 Pathogenesis of Varus Deformity�� 519 5.4.6 Pathoanatomy�� 521 5.4.7 Imaging Assessments in Relation to Tibia Vara�� 523 5.4.8 General Management Considerations �� 529 5.4.9 Recurrent Deformity Following Osteotomy �� 531 5.4.10 Spontaneous Correction�� 532 5.4.11 Surgical Approaches to Tibia Vara�� 533 5.4.12 Adult Sequelae of Childhood Tibia Vara�� 540 5.5 Late-Onset Tibia Vara (Juvenile and Adolescent Tibia Vara)�� 541 5.5.1 Terminology�� 541 5.5.2 Clinical Profile�� 541 5.5.3 Variable Opinions on Whether Late-Onset Tibia Vara is Superimposed on a Pre-existing Varus Deformity�� 542 5.5.4 Physeal Height and the Question of Distal Femoral Varus Tilt�� 542 5.5.5 Association of Femoral Varus with Tibia Vara in Late-Onset Blount Disease �� 543 5.5.6 Radiographic Assessments�� 543 5.5.7 Pathoanatomy of Adolescent Tibia Vara �� 544 5.5.8 Treatment�� 544 5.6 Osgood-Schlatter Disease (Tibial Tubercle Chronic Traumatic Apophysitis)�� 546 5.6.1 Terminology and Description �� 546 5.6.2 Pathophysiology�� 548 5.6.3 Tibial Tuberosity Development�� 549 5.6.4 Structure of the Mature Tibial Tuberosity�� 551 5.6.5 Pathoanatomic Changes in Osgood-Schlatter Disease �� 552 5.6.6 Clinical and Radiologic Features of Osgood-Schlatter Disease �� 555 5.6.7 Clinical Symptoms�� 555 5.6.8 Management�� 556 5.6.9 Complications of Osgood-Schlatter Disease�� 558

Contents

xxi

5.6.10 Findings in Young Adults Following Adolescent Osgood-Schlatter Disease �� 559 5.6.11 Follow-Up of Patients with No Specific Treatment in Acute Phase �� 559 5.7 Congenital Dislocation of the Knee�� 559 5.7.1 Terminology�� 559 5.7.2 Historical Recognition�� 559 5.7.3 Clinical Profile�� 560 5.7.4 Etiology�� 560 5.7.5 Classification�� 561 5.7.6 Pathoanatomy�� 561 5.7.7 Diagnostic Considerations�� 563 5.7.8 Treatment�� 563 5.8 Discoid Lateral Meniscus �� 567 5.8.1 Terminology�� 567 5.8.2 Clinical Overview �� 567 5.8.3 Early Recognition of the Discoid Meniscus as a Pathologic and Clinical Entity �� 567 5.8.4 Pathoanatomy and Theories of Pathogenesis of Discoid Meniscus �� 568 5.8.5 Current Classifications of Discoid Menisci�� 570 5.8.6 Clinical Treatment Profile�� 571 5.8.7 Technique of Arthroscopic Saucerization and Repair�� 573 5.8.8 Clinical/Arthroscopic Approaches and Results�� 573 5.8.9 Meniscus Allograft Transplantation�� 577 5.8.10 Discoid Medial Meniscus �� 577 5.9 Developmental Abnormalities of the Cruciate Ligaments�� 577 5.10 Valgus Angulation Following Proximal Tibial Metaphyseal Fractures in Childhood�� 577 5.10.1 Description and Clinical Profile�� 577 5.10.2 Etiological Considerations Underlying Valgus Deformation�� 578 5.10.3 Guidelines for Treatment�� 579 5.11 Developmental (Congenital) Dislocation of the Patella�� 579 5.11.1 General Considerations Regarding the Patella�� 579 5.11.2 Terminology: Developmental Dislocation of the Patella �� 579 5.11.3 Pathoanatomy of Patellar Displacement �� 582 5.11.4 Imaging for Patellar Disorders�� 582 5.11.5 Treatment Approaches�� 584 5.11.6 Difficulties Maintaining Correction in All Types of Patellar Dislocations �� 589 5.12 Other Childhood Disorders of the Patella �� 590 5.12.1 Bipartite Patella�� 590 5.12.2 Sinding-Larsen-Johansson Disorder �� 592 5.12.3 Abnormal Patellar Position and Size�� 592 5.13 Disorders of the Proximal Fibular Epiphysis�� 594 5.13.1 Congenital Proximal Tibial-Fibular Synostosis�� 594 5.13.2 Proximal Fibular Elongation�� 594 5.13.3 Hypoplasia of Fibula�� 594 5.13.4 Hereditary Multiple Exostosis�� 594 5.13.5 Proximal Fibular Overgrowth Secondary to Damage to the Proximal Tibial Physis�� 594

xxii

Contents

5.14 Disorders of the Knee in Skeletal Dysplasias �� 594 5.14.1 Knee Disorders Are a Common Feature of Many Skeletal Dysplasias �� 594 References�� 595 6 Torsional, Angular, and Deficiency Disorders of the Lower Extremity�� 605 6.1 In-Toeing and Out-Toeing�� 605 6.1.1 Overview�� 605 6.1.2 Clinical and Imaging Assessments Used to Describe In-Toeing and Out-Toeing�� 605 6.1.3 Descriptive Terminology�� 606 6.1.4 Clinical Assessments of the Hip (Regarding Anteversion) and Leg, Ankle, and Foot (Regarding Tibial Torsion)�� 608 6.1.5 Ranges of Rotational/Torsional Values �� 610 6.1.6 General Overview of Rotational Deformities Based on Age Groups�� 614 6.1.7 Specific Management Considerations for In-Toeing�� 614 6.1.8 Internal Tibial Torsion (ITT)�� 617 6.1.9 Out-Toeing�� 619 6.1.10 Out-Toeing Due to Markedly Everted, Pronated Flat Foot�� 620 6.1.11 Combined Rotational Deformities (of the Femur and Tibia, in Opposite Directions); Torsional Malalignment Syndrome �� 620 6.1.12 Diagnostic Categories in Large Series�� 621 6.2 Marked Diaphyseal/Metaphyseal Curvature of the Femur and/or Tibia with Normal Joint Alignment at the Hip, Knee, and Ankle�� 622 6.3 Congenital Pseudarthrosis of the Tibia �� 622 6.3.1 Overview�� 622 6.3.2 Clinical and Radiographic Descriptions �� 622 6.3.3 Findings at First Manifestation of the Disease �� 624 6.3.4 Pathology�� 624 6.3.5 Classifications �� 625 6.3.6 Evolution of Management Approaches�� 627 6.3.7 Complications of Surgical Management�� 633 6.3.8 Lower Extremity Length Discrepancies �� 633 6.4 Posteromedial Tibial and Fibular Bowing�� 636 6.4.1 Terminology and Clinical Overview�� 636 6.4.2 Patient Studies�� 637 6.5 Congenital Lower Extremity Limb Deficiencies�� 638 6.5.1 Terminology�� 638 6.5.2 Congenital Abnormalities (Limb Deficiencies) of the Femur�� 640 6.5.3 Congenital Developmental Abnormalities (Limb Deficiencies) of the Fibula: Fibular Hemimelia�� 647 6.5.4 Congenital Developmental Abnormalities (Limb Deficiencies) of the Tibia; Tibial Hemimelia�� 654 References�� 662 7 Developmental Disorders of the Foot and Ankle�� 665 7.1 Development of the Foot and Ankle; Embryologic, Fetal, and Postnatal �� 665 7.1.1 Early Development �� 665 7.1.2 Vascularization and Ossification of the Foot Bones �� 665 7.1.3 Additional Observations on Ankle and Foot Embryonic and Fetal Development �� 666 7.1.4 Descriptive Divisions of the Foot �� 666

Contents

xxiii

7.1.5 Angular Radiographic Measurements of Talus and Calcaneus�� 669 7.1.6 Anatomy of the Talocalcaneal Joint�� 670 7.1.7 Tibiofibular Torsion�� 670 7.2 Terminology of Foot and Ankle Deformities�� 672 7.2.1 Deformities of the Foot Alone�� 672 7.2.2 Deformities of Both Foot and Ankle�� 672 7.3 Metatarsus Adductus�� 673 7.3.1 Terminology and Deformity �� 673 7.3.2 Developmental Pattern of Deformity�� 674 7.3.3 Non-operative Treatment�� 674 7.3.4 Surgical Treatment�� 676 7.4 Clubfoot�� 677 7.4.1 Terminology�� 677 7.4.2 Incidence �� 677 7.4.3 Etiology�� 678 7.4.4 Types of Clubfeet�� 680 7.4.5 Overview of Responses to Therapy and Possible Relationship to Etiology �� 681 7.5 Pathogenesis and Pathoanatomy of Congenital Clubfoot �� 682 7.5.1 Scarpa�� 682 7.5.2 Little�� 682 7.5.3 Guérin �� 683 7.5.4 Adams �� 683 7.5.5 Evans�� 686 7.5.6 Irani and Sherman �� 686 7.5.7 Settle �� 687 7.5.8 Gross and Histological Abnormalities of the Talus in Clubfoot �� 687 7.5.9 Muscle, Nerve, and Connective Tissue Abnormalities in Clubfoot Studies �� 689 7.5.10 Rotational Abnormalities in the Clubfoot Deformity �� 691 7.5.11 Structural Abnormalities of the Calcaneus in Clubfoot�� 691 7.5.12 Initial Clinical Assessment of Clubfoot Deformities�� 691 7.5.13 Imaging for Clubfoot Deformities�� 697 7.5.14 Historic Overview of Treatment, Ancient Times to Present: Manipulation (Kite, French Functional, Ponseti); Entire Range of Surgical Procedures; Current Approaches to Management�� 699 7.6 Flatfoot Deformities�� 737 7.6.1 Terminology�� 737 7.6.2 Flexible Flatfoot�� 738 7.6.3 Rigid Flatfoot: Tarsal Coalition, Congenital Vertical Talus�� 740 7.7 Calcaneovalgus Foot �� 758 7.8 Pes Cavus�� 758 7.8.1 Terminology�� 758 7.8.2 Pathogenesis�� 758 7.8.3 Clinical Presentation �� 762 7.8.4 Radiographic Indices�� 762 7.8.5 Treatment�� 762 7.9 Sesamoid and Supernumerary Bones of the Foot �� 771 7.9.1 Accessory Navicular Bone�� 772 7.10 Foot Discomfort Accompanied by Radiographic Osteochondroses �� 776 7.10.1 Kohler’s Disease �� 776 7.10.2 Sever’s Disease �� 776 7.10.3 Freiberg Disease (Kohler 2 Disease)�� 777

xxiv

Contents

7.11 Deformities of the Big Toe�� 778 7.11.1 Hallux Valgus�� 778 7.11.2 Hallux Varus�� 780 7.12 Other Toe Deformities�� 781 7.12.1 Overlapping Toes�� 781 7.12.2 Syndactyly�� 782 7.12.3 Polydactyly �� 782 7.12.4 Deformities of the Lesser Toes �� 782 7.12.5 Macrodactyly�� 783 7.13 Osteochondritis Dissecans of the Talus�� 784 7.13.1 Terminology�� 784 7.13.2 Clinical Profile�� 784 7.13.3 Classification�� 785 7.13.4 Progression of Disorder�� 786 7.13.5 Management Profile�� 787 References�� 788 Index�� 799

Introduction. Volume 2

Deformity is a deviation in structure and/or position from normal.

General Principles Regarding Pediatric Orthopedic Deformity Several general principles underlie the development and management of musculoskeletal deformity in the pediatric age group. Awareness of this listing of 20 basic principles (40 including subdivisions) provides guidance for the diagnosis of deformity, following patients, timing treatments, and continuing assessments until skeletal maturity. 1. The shape of individual bones and the alignment between adjacent bones and regions often change normally with growth especially in the first few years after birth. It is important to be aware of developmental patterns that are normal physiologic variants and not pathologic deformities. For example, (a) 40° anteversion of the femoral head and neck at birth is common and will decrease with normal growth to 12–15° in late adolescence, while 40° anteversion in an adolescent regardless of cause will not correct spontaneously and can be clinically problematic for the future; (b) bow leg positioning (genu varum) of 30° is almost always a normal, self-correcting physiological position in a 14-month-old but is highly likely to be pathologic in the adolescent; and (c) kyphosis is the normal sagittal plane position at cervical, thoracic, and lumbosacral regions in the newborn spine; the cervical region begins to develop its normal lordosis at 3–6 months of age (as the infant crawls holding the head upright), while lumbar lordosis develops after 1 year of age (with walking in the standing position); both cervical and lumbar kyphosis in juvenile and adolescent years are true deformities with negative clinical consequences. 2. There is a range of normal angles and rotations within individual bones and between adjacent bones and regions. Variations within these normal ranges can be relatively wide and should not be interpreted as deformities. 3. Deformity may be evident by clinical observation alone (such as moderate to severe scoliosis or clubfoot deformity), by plain radiography (such as coxa vara of the hip or varus of the distal femur and valgus of the proximal tibia causing knee joint obliquity), or by more sensitive imaging modalities such as ultrasound, magnetic resonance (MR) imaging, or computerized tomographic (CT) scanning. 4. Gene mutations cause intramolecular and intermolecular malalignments (molecular deformation) leading to abnormal tissue patterning and development and eventual musculoskeletal deformation. While these abnormalities are not commonly considered to be deformities, they do deform the involved molecules and are increasingly viewed as intracellular and extracellular deforming forces; examples affecting the skeletal system are signaling molecules of the Notch group (affecting early patterning) and structural molecules like collagen. Common examples of gene mutations causing significant clinical deformation in the musculoskeletal realm include skeletal dysplasias, osteogenesis imperfecta, peripheral neuropathies (Charcot-Marie-Tooth disease), congenital scoliosis, and muscle disorders (Duchenne muscular dystrophy). In children with deformities, it is now xxv

xxvi

necessary to consider abnormalities of chromosomes, such as Down syndrome with an extra copy of chromosome 21 (Trisomy 21) or genetic point mutations, deletions, or insertions on specific molecules as primary causative deformities; examples include collagen abnormalities leading to osteogenesis imperfecta, dystrophin abnormalities leading to Duchenne and Becker muscular dystrophy, peripheral myelin protein 22 (PMP22) abnormalities leading to Charcot-Marie-Tooth peripheral neuropathy, and delta-like 3 (Dll3) mutations of the Notch family leading to congenital scoliosis with rib malformations. 5. A pathologic deformity during the growing years may proceed along one of the three pathways: it may correct spontaneously with growth, it may remain unchanged with growth, or it may worsen with growth. 6. Each of the main mechanisms of spontaneous correction of a bone deformity in a growing child can be seen with remodeling following a physeal, metaphyseal, or diaphyseal fracture that heals with angulation or malrotation but an intact functional physis. Repair mechanisms involve (a) differential physeal growth tilting and rotating the physis to the normal plane, (b) periosteal new bone formation on the concavity of the deformity, and (c) bone resorption on the convexity. These mechanisms are more effective the greater the number of years of growth remaining, the closer the angular deformity is to the physis, the more the physis involved contributes to the growth of the bone, and the presence of deformity in the plane of motion of the adjacent joint. 7. If persistence of a deformity or its worsening with growth is likely, the deformity must be assessed as problematic or non-problematic over the short, intermediate, and long (adult years) time frames. Problems resulting from deformity include some or all of pain, abnormal function such as gait disturbance, associated organ impingement (cardiac, pulmonary), or appearance. Also to be considered for any deformity is what degree of deformity is associated with what amount of clinical problem. 8. There is a constant interplay of forces during the growing years between the soft tissues [muscle, tendon, ligament, intervertebral disc] and the developing bones [composed of cartilage and bone tissue]. Any abnormality in either the soft tissue or the developing bone due to imperfect structure or function can lead to deformity. 8.1. For any deformity it is essential to determine whether it is flexible or rigid. If it is flexible, there is a need to know if it fully straightens, or at a joint fully reduces, with change of position or passive manipulation or if it only partially corrects. Conversely, if rigid there is need to determine whether it is completely rigid or if manipulation partially straightens or reduces the deformity. The more rigid the deformity, the more extensive the surgery needed to correct it and, in the case of the spine, the less correction attainable. 8.2. Primary soft tissue deformities, including neuromuscular disorders, are initially associated with normal cartilage model/bone model development. The longer the deformity persists and the more severe the deformity, the greater the likelihood that asymmetric pressure on the skeletal elements involved and the lack of normal functional stress will lead to structural deformity of cartilage and bone (and intervertebral joints in the spine). 8.2.1. Excess muscle pull may be due to localized muscle weakness (with the adjacent normal muscles continuing to function) or to localized muscle over-reaction due to spasticity (that overpowers the adjacent normal muscles). Either occurrence leads to malposition of joints and distal structures or disturbances of normal spinal alignment. The malposition or abnormal alignment may correct with muscle relaxation or muscle-tendon transfer to balance strength (e.g., across a joint) but will recur with continuing asymmetric activity and eventually lead to cartilage and bone deformity with growth. 8.2.2. Ligamentous laxity leads to malpositioned joints that are initially present only in the upright position or with weight-bearing but correct to normal anatomic positioning with the recumbent position or non-weight-bearing. If left uncor-

Introduction. Volume 2

Introduction. Volume 2

xxvii

rected in the presence of continuing growth, the malposition can progressively become rigid, not actively or passively correctable, and even associated with misshapen bones. 8.2.3. Asymmetric soft tissue tightness can lead to rigid joints (contractures) and distal structure deformation that with continuing malposition eventually leads to developing bone deformity with growth. 8.3. Bone/cartilage deformity has varying relations to the orientation of adjacent cartilage surfaces. 8.3.1. Angular bone deformity centered at metaphyseal or diaphyseal regions leads to asymmetric joint alignment (obliquity) at one or both ends of an individual bone with a predisposition to abnormal intra-articular pressure and articular cartilage osteoarthritis. 8.3.2. Asymmetric growth plate function due to focal physeal abnormality leads to angular growth of the bone and joint obliquity distal and sometimes proximal to the involved physis that predisposes to abnormal intra-articular stresses. 8.3.3. In rare instances there is bidirectional angular deformity (or bowing) within a bone owing to curvatures within the metaphyseal-diaphyseal regions. This can occur in patients with osteogenesis imperfecta or rickets. For example, a valgus angulation of the tibia centered at the proximal metaphysis can reverse direction by curvature in the mid-diaphysis leading to varus angulation centered at the distal metaphysis. In some instances there can be normal articular surface alignment persisting at both ends, while in others the angular deformity also tilts the physis, epiphysis, and articular surfaces into an oblique plane. 8.4. Correction of deformity due to soft tissue abnormalities will often allow for spontaneous growth-related correction of mild to moderate cartilage model/bone deformities. 8.5. Neuromuscular deformities with both soft tissue and cartilage model/bone deformation usually require soft tissue correction (balancing of muscle groups) and might need bone correction by osteotomy or asymmetric physeal stapling. 8.6. Repositioning of dislocated or partially dislocated (subluxed) joints by non-operative or operative means often allows mild to moderate cartilage model/bone deformities to correct with growth. The younger the patient is and the more years there are to skeletal maturation, the greater the likelihood for cartilage/bone correction with growth after repositioning. 8.7. Normal function or improved function posttreatment allows the normal stresses on immature tissues to further correct the remaining deformity by differential growth toward a normal position. 8.8. Deformities whether congenital or acquired of intra-articular soft tissue structures, such as knee menisci and cruciate ligaments and the hip labrum, if allowed to persist can damage articular cartilage. The damage can occur directly by causing abnormal contact or indirectly by causing joint instability. These deformities can be suspected by clinical history and examination but are diagnosed definitively by direct arthroscopic visualization or by MR imaging. 9. Trauma to a developing bone can lead to deformity in several ways. 9.1. Acute diaphyseal or metaphyseal fractures that heal with angulation, malrotation, translation, or length discrepancy represent deformities that, with growth, may correct partially or fully, persist unchanged, or (rarely) worsen and require surgical intervention for correction. 9.2. Acute physeal (growth plate) fractures (or fracture-separations) can heal with no subsequent growth sequela or with partial or complete growth plate fusion predisposing to angulation and/or shortening with growth.

xxviii

9.3. Chronic repetitive stresses on growth plates can lead to growth damage and premature growth plate fusion. Chronic repetitive stresses on tendon insertions in growing bones can lead to non-displaced tendon-cartilage-bone interface avulsions with painful tendinitis and insertion site swelling. 9.4. Articular cartilage damage either by displaced osteochondral fractures or linear oblique intra-articular fractures with persisting surface irregularity or gaps predisposes to arthritic changes. 10. Bone deformities are not passively correctable. They may, however, spontaneously remodel with growth, be correctable with growth with the use of serial casting or bracing, or eventually require surgical intervention by metaphyseal/diaphyseal osteotomy, asymmetric physeal stapling, or vertebral body tethering for correction. 11. Once a deformity is identified, the presence of a specific disease or an underlying causative disorder must be considered. Treatment can differ significantly for the same deformity: (a) in a patient who is otherwise normal and in one who has an underlying disorder and (b) between patients with different specific diseases. 12. It is important to distinguish between a primary deformity and a secondary or compensatory deformity. The primary (or initial) deformity is at the site of pathologic abnormality and tends to be at least partially fixed or rigid. The secondary (or compensatory) deformity tends to be fully flexible initially and for a longer period of time since it represents a process within normal adjacent regions designed to maintain stability, balance, and alignment. A long-standing secondary deformity may become rigid with time. In the spine, if the secondary curves above and below the primary curve remain flexible, trunk balance compensation occurs (curve in the opposite direction), and the head above and pelvis below remain level; if the primary curve above involves the entire cervical spine, compensation cannot occur and the head is tilted, and if the primary curve below involves the entire lumbar spine, compensation cannot occur and the pelvis is tilted (pelvic obliquity). 13. A static joint deformity is present at all times and in all positions (e.g., supine, standing) owing to its associated bone or soft tissue rigidity. A dynamic joint deformity is present only with muscle activity since it is due to muscle imbalance. Dynamic deformities commonly occur with gait or attempted upper extremity activity with associated asymmetric muscle under- or overactivity. 14. Some deformities are sufficiently mild that no treatment is warranted since they cause no current problems and there is little convincing evidence that they will cause problems in the future. 15. Treatments for some childhood deformities may not be warranted even if they are moderate to severe in view of the severity and progressive nature of the primary underlying pathologic process such as a neurodegenerative disorder even though treatment of the same deformity in an otherwise normal or only mildly involved child would be warranted. 16. Deformity even if marked may be a secondary compensatory deformity necessary to maintain function such that its correction to a normal anatomic position may significantly decrease function. Marked lumbar lordosis in a patient with neuromuscular disease with gluteus maximus muscle weakness can allow a person to continue walking, whereas spinal bracing or fusion straightening the spine can make the patient anatomically normal but unable to walk. Compensatory deformities are common in the spine above and below regions of primary scoliosis or kyphosis. 17. Treatment in childhood is warranted for deformities that will not correct on their own if they are already symptomatic, causing discomfort or altered function, or if they are asymptomatic but there is scientific evidence that they will become symptomatic with time, even if that will be during the adult years. 18. Once a deformity is considered treatable, two general approaches follow. The first is to treat any existing primary underlying disorder by medical means, for example, infection with antibiotics, hemophilic arthropathy with factor replacement, rickets with vitamin D,

Introduction. Volume 2

Introduction. Volume 2

xxix

and spastic cerebral palsy with muscle relaxant therapy. In the future, this principle will also apply to direct treatments for gene and molecular deformation. The second is to treat the deformity itself by orthopedic means. 19. Orthopedic treatment can be non-operative or operative. 19.1. The variety of non-operative therapies is great and includes rest, range of motion and stretching exercises, pharmacologic agents for pain relief and muscle relaxation, and serial casting or bracing to stretch tightened soft tissues and position the deformed part straight to allow normal cartilage and bone growth to occur. 19.2. The ranges of operative therapy are also great. Some operations cure deformity permanently with one intervention. Other procedures correct one aspect of the deformity and then rely on spontaneous growth or repositioning for full correction of adjacent structures. Some deformities need a series of operations over a short or longer period of time in the growing years since (a) the underlying cause may persist, (b) the damage done to the growing structure may itself lead to recurrence, or (c) soft tissue and bone procedures may be needed for correction but are best done at different times in the growth period. 20. Once a deformity in childhood has been corrected, it may follow one of the three pathways during the remaining years of growth: (a) the region involved may remain anatomically normal, and no further management is ever needed; (b) the deformity may recur due either to the fact that it had not been fully corrected or the underlying disorder that caused it persists; and (c) the deformity may overcorrect leading to deformation in the opposite direction due to such factors as a new pattern of muscle imbalance, asymmetric growth plate function, or continuing growth correction resulting in overgrowth. This variability leads to the advisability for following patients with a pediatric orthopedic deformity until skeletal maturity.

Implications of the General Principles of Deformity 1. Close relationship between skeletal growth and pediatric orthopedic deformity can be either beneficial or detrimental. Some deformed positions are actually physiologic and correct spontaneously with growth; pathologic deformities can correct, remain the same, or worsen with growth; and deformities that have apparently been corrected can remain straight, worsen with recurrence (due to unrecognized under-correction or persistence of the underlying disorder), or even overcorrect (due to altered muscle or bone formation balance) with growth. 2. Several concepts must be assessed when managing pediatric orthopedic deformities. These include a range of biologic variation within a bone and the angular relationship between adjacent bones; primary versus secondary/compensatory deformity; rigid versus flexible deformity; static deformity (present regardless of position) versus dynamic deformity (present with muscle function); interplay between soft tissues (muscle, tendon, ligament, intervertebral disc) and growing bone/cartilage models; and relationship between primary disease and secondary orthopedic deformity, such as rickets and bowed femur/ tibia; hemophilia and knee, elbow, and ankle synovitis; and septic arthritis of the hip causing avascular necrosis of the femoral head. 3. Presence of deformity can lead to variable approaches to management. Major surgical intervention may be warranted even with minimal deformity based on natural history studies of invariable worsening with time; the same deformity seen in different patients may warrant differing approaches: surgical correction in an otherwise normal person versus no surgery in one with a progressive neurodegenerative disorder; and surgery to correct major secondary compensatory deformities may be contraindicated such as spinal fusion for lumbar lordosis in an ambulatory neuromuscular patient that worsens function while achieving anatomic straightening.

1

Developmental Dysplasia of the Hip

1.1

Terminology

Developmental dysplasia of the hip is a general term referring to a spectrum of deformities, usually diagnosed in the neonatal period, in which the structural relationship of the proximal femur to the acetabulum is intermittently or continuously abnormal. The spectrum includes (i) a subluxatable or dislocatable hip associated with capsular laxity in which the head of the femur moves partially or totally out of the acetabulum with extension and adduction and back into it with flexion and abduction, (ii) a subluxated hip in which there is a partial but persisting loss of the normal relationship of the head of the femur to the acetabulum in extension with the head more lateral than normal in the acetabulum and the acetabulum more shallow than normal with its lateral roof angled outwardly and upwardly, and (iii) a dislocated hip in which there is a complete and persisting loss of any femoral head-acetabular relationship, regardless of the position of the hip. Developmental dysplasia of the hip (DDH), as currently defined, is not associated with clinically evident connective tissue, neuromuscular, or other diseases. The single most important initial pathoanatomic change appears to be a capsular laxity which renders the hip unstable at birth with all subsequent abnormalities being secondary phenomena which develop an increasing variation from the norm the longer a hip is allowed to grow with any persisting malposition. The terminology used to describe this condition has always been variable and imprecise primarily due to the imperfect understanding of the pathoanatomy and timing of its initial occurrence. Congenital dislocation of the hip (CDH) was used previously to describe the entity, although some used the term congenital dysplasia of the hip to encompass the entire spectrum of the disorder. Dunn defined congenital dislocation of the hip as an “anomaly of the hip joint, present at birth, in which the head of the femur is, or may be, partially or completely dislocated from the acetabulum” [1]. The entity is now referred to as developmental dysplasia of the hip (DDH). Developmental has replaced congenital since (i) it focuses on abnormalities in development which predispose to the

condition and which ‘worsen in the absence of normal hip positioning and (ii) it is not definite that all dysplastic hips were structurally abnormal and/or detectable at the time of initial postnatal examination. Dysplasia is a vague general term referring to a poorly defined disease process. Delayed, and thus imperfect, development of the acetabulum and of the proximal femur is referred to as a dysplastic process. Acetabular dysplasia and proximal femoral dysplasia themselves are either primary disorders and/or disorders which occur secondary to growth in the presence of undetected and untreated developmental hip disease. Developmental dysplasia of the hip therefore encompasses a spectrum of hip abnormality. These include (i) an initial subluxatable or dislocatable hip in which the femoral head is located in a normal relation to the acetabulum in certain positions (generally flexion and abduction) but has a partial or complete loss of continuity in other positions; this situation can spontaneously correct itself within a few days of birth or it can progress if untreated to persisting deformity, (ii) a subluxation of the hip which refers to a partial loss of continuity between the femoral head and acetabulum where the abnormal relationship is present throughout the entire range of movement, and (iii) a dislocated hip with complete loss of continuity between joint surfaces at all times regardless of the position of the hip. Some refer to an unstable hip detected clinically on initial screening in the newborn nursery as having “neonatal hip instability” (NHI). Terminological distinctions are not merely a semantic issue; imprecise use of terms implies imprecise understanding of the underlying pathoanatomy that can lead to investigations and treatments which are not fully appropriate.

1.1.1 Change in Terminology Klisic wrote a brief report in 1989 strongly supporting the use of DDH (which he defined as developmental displacement of the hip) to refer to the entire entity of hip dysplasia, subluxation, and dislocation [2]. He felt that the widely used

© Springer Nature Switzerland AG 2019 F. Shapiro, Pediatric Orthopedic Deformities, Volume 2, https://doi.org/10.1007/978-3-030-02021-7_1

1

2

1 Developmental Dysplasia of the Hip

term CDH (congenital dislocation of the hip) was inaccurate since it suggested a gross prenatal malposition demanding orthopedic correction. In reality the term DDH was preferable since it indicated a dynamic disorder capable as the child developed of getting better or worse. Klisic acknowledged the role of Michele who had used the term developmental hip dislocation as the title of his chapter on hip dysplasia in his book Iliopsoas: Development of Anomalies in Man in 1962 [3]. Michele recognized that a small number of dislocations (~2%) were congenital originating as embryologic defects in the germplasm but that the vast majority (~98%) occurred in an otherwise normal fetus at 6–9 months of uterine life due to a failure of stimulus to normal growth leading to what was in effect a “developmental dislocation.” He felt that “congenital” should refer only to the atypical teratological cases, while the typical acquired environmental-anthropological cases should be referred to as “developmental” dislocations. The term DDH is now widely accepted, but CDH or CDH/DDH will be used in discussing articles written using the CDH terminology.

1.2

evelopment of the Hip: Embryonic D and Fetal Periods

1.2.1 E arliest Developmental Biology of the Hip. Chick Embryo Studies As long ago as 1883, Johnson outlined the earliest development of the pelvic girdle, hip region, and hind limb in the chick embryo [4]. “The future cartilage is only just distinguishable from its surroundings of indifferent (undifferentiated) mesoblastic cells.” “We can clearly distinguish three elements in the girdle meeting in the broad acetabular region, which passes on without a break into the femur….. the cartilage of the femur is continuous with that of the girdle, as are the three elements of the girdle with one another (ilium, ischium, pubis).” Only after structural development of the femoral-acetabular components was relatively well established “the femur begins to be separated from the girdle by an intervening tract of tissue” (meaning the cellular interzone which then is removed by the joint cavitation process). Chevallier using chick and quail embryonic transplants demonstrated that the bones of the pelvic girdle originated from the somatopleural mesoderm which was shown to be regionalized as early as 2 days incubation, even prior to somatic segmentation [5]. The eventual cartilage centers of the three bones composing the acetabulum were preformed as a uniform mesenchymal condensation at 5 days but separate centers for the ilium, ischium, and pubis formed at 8 days which was shown to be regionalized as early as 2 days incubation. Malashichev et al. performed two studies on the early embryogenesis of the pelvic region and its genetic components [6, 7]. One study demonstrated that ectodermal signals occurred

at pre-limb bud stages for pelvis formation and that the regulation of ilium development differs from that for pubis and ischium [6]. Emx2 was shown to be required for formation of the ilium but not the other two components. When the ectoderm over the somatopleure was removed, there were severe defects in the pelvic skeleton, but the defects differed depending on the time of intervention. The diverse pelvic elements appeared in temporal sequence of the ilium, pubis, and ischium. Emx2 was expressed in regions giving rise to the ilium and Pax1 in regions for the pubis, but these were restricted to times prior to chondrogenesis. In a second study [7], the entire pelvic girdle originated from the somatopleure with no somitic cell contribution to the pelvic skeleton. Ectodermal signals controlled development of the pelvis however, especially pubis and ischium. Pax1 and Alx4 modulated normal ischial and pubic development. It became evident that while Emx2 expression helped direct formation of the ilium, signals from both ectoderm and somites were needed to complete development of the ilium. The chick pelvis thus originates from lateral plate mesoderm, but its development requires signals from overlying ectoderm and (for the ilium) paraxial mesoderm/somites. Nowlan and Sharpe assessed chick embryo hip joint morphology not only by histology but also by a more sensitive technique of direct 3D capture using optical projection tomography (OPT) assessing tissue-specific markers allowing for earlier evolving structural assessments [8]. This enabled them to determine that major anatomical features of the developing hip were present a full day prior to joint cavitation. This included demonstration that rotation of the pelvis with respect to the femur in advance of cavitation (which allows for the effects of motion).

1.2.2 G eneral Aspects of Human Hip Development The embryonic period in the human refers to the first 8 weeks of development during which time each of the organs including the cartilaginous models of the long bones and vertebrae have formed. By the end of the embryonic period, the average embryo is of 3 cm crown-rump (CR) length. The fetal period from 8 weeks of age to birth is associated with increase in size and organ differentiation. Watanabe outlined hip development in 288 hips from 144 embryos and fetuses from 14 to 300 mms CR length ending at 24 weeks gestation [9]. The femoral head diameter around 11 weeks was 2 mms at which time the joint space was formed and the head could be dislocated by cutting the capsule. The femoral head diameter by 24 weeks was around 8 mms. The diameter of the femoral head increases in size in a linear pattern and parallels the growth of the entire body. The femoral head is spherical at the beginning of development and remains so throughout growth. The neck-shaft angle was 130 during fetal development. Femoral anteversion averaged −4° from 10 to 15 weeks, 5° from 15 to 20 weeks, and 11° from 20 to

1.2 Development of the Hip: Embryonic and Fetal Periods

24 weeks, but there was a wide range of variability at these times of both positive and negative values. At birth femoral anteversion had increased to 35°. Watanabe’s study found no examples of full dislocation, but there were 26 dysplastic hip joints characterized by “an overall hypoplasia of the entire hip joint with a shallow acetabulum.” The femoral head was always stable with flexion and tended to subluxate with extension. The femoral head and acetabulum had reached infantile shape prior to joint space formation such that dislocation could not occur during the embryonic period. Strayer studied hip development from human embryos 6.5–237 mms crown-rump length [10]. He concluded, in agreement with other observers, that all elements of the hip joint differentiate in situ in a mass of blastema. The head of the femur is globular (spherical) in shape at all times during its development, and the relative proportions of the blastemal and early cartilage developmental segments of the pelvic bones entering into formation of the acetabulum are the same in early embryos as in fetal stages and postnatal life. The ligamentum teres develops in situ within the joint. Congenital dislocation of the hip cannot occur before the opening of the joint cavity. The synovial lining does not develop as a cellular ingrowth but rather from cells in situ as part of the original blastemal mass. The synovium forms along the line of cleavage that appears between cells as the interzone tissues are liquefied. The acetabulum develops by growth and fusion of processes from the iliac, ischial, and pubic cartilages. Each of these develops around the femoral head, and their fusion initially produces a shallow acetabulum. The portion of the acetabulum to which each pelvic cartilage contributes is approximately the same as those later furnished by the corresponding pelvic bones being 2/5 ischium, 2/5 ilium, and 1/5 pubis. Each of the pelvic cartilages has a centrifugal growth pattern within the blastema. The region that will become the hip joint is composed initially of dense blastemal tissue referred to as the interzonal tissue, while the embryo is growing from 20 to 30 mms in length. Cavity formation begins in the tissue between the cartilage of the acetabulum and the cartilage of the head of the femur. The interzone tissue other than the ligamentum teres becomes looser in texture with time and ultimately is resorbed to leave the joint cavity. Other Studies The greater trochanter is evident at 30 mms and the femoral neck and lesser trochanter at 34 mms [10]. The hip joint cavity according to Moser appears first in the lateral part of the joint at 34 mms [11]; Haines described an initial cavity at 34 mms [12]. The ligamentum teres develops in situ with Moser describing it as early as 20 mms and Strayer noting it at 23 mms. The glenoid labrum of the hip joint was noted at 30 mms as a transition with the outer region of the acetabular cartilage. Dimeglio et al. reviewed prenatal hip development stressing the unique interrelationship of the pelvis, femur, and associated muscles on normal structure [13]. They stressed in particular

3

three aspects of growth: (1) enlargement and full development of the acetabulum, (2) harmonious spherical enlargement of the femoral head and its secondary ossification center, and (3) elongation of the femoral neck in the postnatal period. Detailed observations on the prenatal development of the human hip joint were provided by Gardner and Gray [14] in a study based on 52 human embryos and fetuses ranging from 12 mm crown-rump length (6 weeks) to 370 mm (term) and by Andersen [15] in a study of 30 human embryos-fetuses from 20 mm (7–1/2 weeks) to 121 mm (16 weeks). Their observations are in good agreement and are combined below.

1.2.2.1 Origin of Limb Bud The lower extremity limb bud is seen in embryos 3–4 mms in length as a small protuberance on the anterior and lateral aspect of the body wall at the level of the lumbar and first sacral segments. The specific tissue differentiation for each bone then follows from blastemal tissue or undifferentiated mesenchymal cells, to precartilage, to cartilage, and then to bone. The region of the future hip joint appears as a group of densely packed undifferentiated cells in the form of a cone with an oblique base applied to the side of the body. The first appearance of the acetabulum is in the 14–15 mm embryo as a line of cells of diminished density proximal to the head of the femur. This region was initially felt to represent an arc of 65–70° which subsequently deepened to enclose a full half circle of 180° as the joint cavity formed. The interzone demonstrates increased cell density by 15–22 mms. Early differentiation of the ligamentum teres and periarticular capsular structures is noted around 23 mms. As development and growth proceed from 23 to 45 mms, the cartilage of the ilium grows out over the head of the femur with the labrum continuous with its outer margin. Increases in the extent of the elements of the acetabulum are responsible for the relative lateral displacement of the labrum. The acetabulum is never flat; from the earliest stages, it extends, together with the labrum, beyond the midway point of the head and always has a concave shape. Differentiation of blastema in the innominate region begins in the ilium just above the acetabulum at the 15 mm stage. This most lateral region lags behind the shaft and head of the femur in differentiation at all stages. The three cartilage centers become vascularized separately and serve to outline the triradiate (or Y) cartilage. Chondrification radiates from the three centers of these regions of the ilium, ischium, and pubis. Endochondral ossification then occurs within the central regions of these cartilage masses: ossification starting at 9 weeks in the ilium, at about the fourth month of gestation in the ischium, and a few weeks later (fifth month) in the pubis [16]. The triradiate cartilage for endochondral growth lies between the bony centers. 1.2.2.2 Acetabular Labrum and Transverse Acetabular Ligament The acetabular labrum (often referred to in the older literature as the glenoid labrum of the hip joint) is formed at the earliest stages of the formation of the acetabulum as early as

4

19 mms and appears histologically as a condensation of blastema at the cartilage periphery [10, 14, 17]. By 25 mms it is clearly differentiated. It becomes vascularized at 61 mms. The transverse acetabular ligament also forms during this time period; the site of the transverse ligament of the hip joint is considered by Strayer to be the weakest point of structure. By 28 mms a condensation for the transverse acetabular ligament is seen, and by 30–33 mms the ligament is well defined. The superior labrum covers the widest diameter of the femoral head. The anteroinferior part of the acetabulum, which is known as the acetabular notch, is covered by the transverse acetabular ligament [17]. This ligament is the support for the acetabular labrum as it crosses the notch.

1.2.2.3 Joint Capsule and Synovium In 12–15 mm embryos, avascular blastemal tissue in the region of the future joint is denser than that of the adjacent anlagen. This density is more pronounced at 17 mms with a definite interzone present. The interzone is more definite by 20 mms, and it is possible to define a three-layered interzone, the middle layer of which is directly continuous with the mesenchymal tissues surrounding the joint except in those areas of capsular condensation. The outer layers of the interzones are directly continuous with the perichondrium of the femoral and acetabular anlagen. The capsule surrounding the joint is defined. Contained within it is a portion of the mesenchyme surrounding the joint that is structurally a part of the interzone. This intra-articular mesenchyme is the first indication of what will become synovial mesenchyme. The intermediate layer of the interzone is continuous with the blastemal synovial mesenchyme, and both are vascularized. The three-layered arrangement of the interzone is more pronounced at 22–25 mms. Early spaces form within the middle layer. By 30–33 mms a clear cavity is present around the periphery of the joint. Even at the time of opening of the joint space, it is not possible morphologically to distinguish between the cells of the inner margin of the capsule that will eventually form the synovial membrane and the capsule itself. The first indication of the fibrous capsule is seen at 20 mms with a condensation appearing as a direct continuation of the perichondrium of the femur and pelvis. 1.2.2.4 Joint Cavity Joint cavity formation represents a programmed degenerative and mechanical process with no evidence of ingrowth of tissue from the outside to provide a lining for the joint. Early evidences of degeneration are seen at 23 mms with increases in the intercellular spaces in the interzonal cells between the head of the femur, the ligamentum teres, and the acetabulum. At 36–42 mms spaces filled with fluid are formed. Andersen times formation of the joint cavity between 34 and 42 mms [15]. Vascularization of the interzone is an integral part of joint cavitation. Joint cavitation begins in the central area of the joints and then moves toward the periphery [12]. Cavity formation at the hip takes place as an annular rim, limited medially by the head

1 Developmental Dysplasia of the Hip

of the femur and laterally by the acetabular/glenoid labrum. The ligamentum teres remains in the middle of the developing joint cavity. In later stages of cavitation, the space is enlarged centrally around the ligamentum teres and peripherally passing beyond the tip of the labrum and surrounding the head in its entirety and also the neck distally to the capsular insertion.

1.2.2.5 Retinacula of Weitbrecht The extension of the joint space down the neck of the femur leaves as intracapsular structures the perichondrium, the retinacula of Weitbrecht, and the ascending cervical vessels. The retinacula of Weitbrecht are intracapsular flattened band structures of the hip joint present on the interior of the capsule and passing toward the margin of the femoral head. The retinacula are synovial-covered capsular reflections or prolongations [18]. The blood vessels eventually supplying the proximal femur perforate the capsular attachment at the base of the neck and pass along the surface of the neck entering the metaphysis of the neck and the epiphysis of the head through small foramina. Walmsley continues: “From the points where they perforate the capsule these vessels derive and carry inwards indefinite fibrous prolongations of the capsule wall which are covered over or are completely invested by reflections of synovial membrane. These elements constitute the retinacula of Weitbrecht.” The fibrous prolongations terminate at varying distances from their origins, while the synovial reflections covering the vessels continue toward the cartilaginous margin of the head. The retinacula are reflections or continuations of the synovial membrane combined with fibrous sheath prolongation of the capsular wall which carry within themselves the blood vessels of the head and neck. 1.2.2.6 Ligamentum Teres At 22 mms the first suggestion of the ligamentum teres is found. The ligamentum teres is present in 22–25 mm specimens as a region of greater cellularity but is not sharply demarcated from the neighboring interzone. There is never any evidence of a depression in the head to receive the ligamentum. The separation of the ligamentum teres to form a free mass within the joint occurs simultaneously with the opening of the remainder of the cavity which is characterized by peripheral vascularization, degeneration, and splitting between the cells along its margin. The ligamentum teres is well defined in the 30–33 mm fetuses. Blood vessels are first noted within the ligamentum teres at 60 mms. The ligamentum teres originates broadly from each side of the acetabular notch and from the transverse acetabular ligament. It is attached to a depression on the femoral head just below and posterior to the center of the head [11, 17]. 1.2.2.7 Extra-articular ligaments The hip receives additional stabilization from its extra-articular ligaments. Anteriorly and superiorly support is derived from the iliofemoral ligament, referred to as the Y-ligament of Bigelow or the ligament of Bertrand (Fig. 1.1a). Posteriorly, support comes from the ischiofemoral ligament, the lower

1.2 Development of the Hip: Embryonic and Fetal Periods

a

5

b Crista ilica 4 Linea glutaea dorsalis 10

Spina ilica dorsalis cranialis

Tuberculum ilicum Linea glutaea cranialis

7 6 7

Lig. sacroilicum dorsale longum Spina ilica ventralis

11 Foramen ischiadicum maius

Linea glutaea supraacetabularis Tuberculum ilicum

Lig. sacrospinale Lig. iliofemorale

Spina ossis ischii Foramen ischiadicum minus

2

Zona orbicularis

Lig. ischiocapsulare Trochanter maior

Lig. sacrotuberale 1 12

4

Fossa trochanterica Collum femors

Tuber ossis ischii

10

articuli coxae

5

Crista intertrochanterica 8 Trochanter minor 9

3 Linea pectinea

Fig. 1.1 The basic structures of the normal hip in a growing child are outlined in a–j. (a) Anterior view of the hip joint (bottom) demonstrates the iliofemoral ligament (the inverted “Y” ligament of Bigelow). The ligament extends from above the acetabular rim to the intertrochanteric line. The iliofemoral ligament diverges into medial and lateral bands distally (the inverted “Y”). Medially the pubocapsular band is now referred to as the pubofemoral ligament. Posterior view of the hip joint (top) shows the ischiofemoral ligament. The proximal lateral fibers are a continuation from the anterior iliofemoral ligament. The posterior capsule and ligaments insert part way up the neck leaving the distal part uncovered. The lower part of the ischiofemoral ligament is thickened and often referred to as the orbicular band, zone, or ligament. (Figures reprinted from Morris’s Human Anatomy (ed. H Morris, J Playfair McMurrich), 4th edition, part 1, Philadelphia, P. Blakiston’s Son and Co, 1907). (b) Posterior view of the hip joint also showing (above) the superior and posterior part of the iliofemoral ligament as well as the ischiofemoral ligament. The lower margin of the ischiofemoral ligament is almost a discrete structure itself referred to as the orbicular zone or ligament. It is evident on normal hip arthrograms. The synovial protrusion at the lower margin of the orbicular ligament (another arthrographic finding) is shown. The capsule and ligaments of the hip joint insert more distally anteriorly along the intertrochanteric line compared to their posterior insertion that leaves the most distal portion of the neck extracapsular. (Reprinted with permission from Praktische Anatomie by T von Lanz and W Wachsmuth, Springer-Verlag, 1955). (c(i)) Illustration of the partial pelvis at left from the outer, lateral aspect shows the three component parts of the acetabulum that grow from ileal, ischial, and pubic centers of ossification. The triradiate cartilage is seen linking the three during the growing years. Upper right drawing shows the three bones and triradiate cartilage as they appear on the inner view of the pelvis; lower right drawing shows components in anteroposterior view. (c(ii)) Growth from components of the triradiate cartilage is shown by directional arrows in the same projections as shown in c(i) above. The triradiate cartilage lengthens, widens, and deepens the acetabulum with growth. (c(iii)) Drawing of the acetabular cartilage complex from (a) medial-inner pelvic aspect, (b) posterolateral aspect, and (c) lateral outer aspect. There is cartilage tissue communication between the triradiate cartilage and the hemispheric articular cartilage. (Reprinted with permission from TJ Harrison,

Tuberositas glutaea

Journal of Anatomy). (c(iv)) Drawing shows the functional specificity of the cartilage components of the developing acetabulum and pelvis, all of which appear only as radiolucent regions on plain radiographs. Where two bone regions are adjacent to each other, the triradiate cartilage separating them is physeal, epiphyseal, and physeal from bone to bone. Where a bone region is adjacent to the joint, the cartilage between bone and joint is physeal, epiphyseal, mini-plate, and articular. The full code is listed on the illustration. AC, articular cartilage; EC epiphyseal cartilage; and PC, physeal cartilage. (d) At puberty, the depth of the acetabulum is increased by three secondary ossification centers at the periphery of the acetabular cartilage. The os acetabuli (OA) is the epiphysis of the pubis and helps form the anterior wall of the acetabulum. The acetabular epiphysis (AE) is the epiphysis of the ilium and forms a major part of the superior wall of the acetabulum, while a third smaller epiphysis in the ischium is also formed. (Reprinted with permission from Ponseti, JBJS Am). (e) Three anterolateral views of the pelvis and acetabulum, following removal or displacement of the proximal femur, demonstrate that the acetabulum is spherical, deepened by the acetabular labrum (glenoid in older terminology), and given further support inferiorly and anteriorly by the transverse acetabular ligament across the acetabular (condyloid) notch. Articular cartilage does not cover the entire interior of the acetabular socket, being present in a lunate shape covering primarily the superior, posterior, and lateral aspects of the socket. It is relatively deficient medially where it is replaced or covered by the synovial membrane, the fibro-fatty tissue (pulvinar), and the origin of the ligamentum teres. (Illustrations reprinted with permission from Morris’s Human Anatomy 1907). (f) Coronal section drawing illustrates the main features of the developing hip. There is lateral extension of the acetabulum by the fibrocartilaginous labrum. The capsule inserts laterally and superiorly above the acetabular labrum and cartilage onto the side of the ilium. This recess is a normal anatomic feature and is outlined in a normal hip arthrogram. A similar attachment of the capsule inferiorly beyond the acetabular labrum occurs. Medially, the floor of the acetabulum is covered by fibro-fatty tissue, the synovial pad, and the origins of the ligamentum teres, leaving the articular cartilage present superiorly, posteriorly, and laterally. The trabecular orientations within the bones outline the direction of bone deposition responding to regions of heightened stress.

6

1 Developmental Dysplasia of the Hip

c i

ii

Ilium

Pubis Ischium

iii

E A

F

a

iv C

D

b 1 mm.

B

c

Fig. 1.1 (continued) (Reprinted with permission from Praktische Anatomie by T von Lanz and W Wachsmuth, Springer-Verlag, 1955). (g) Anteroposterior pelvic (hip) radiograph from a child at the same age corresponds to the illustration in (f). (Reprinted with permission from T von Lanz and W Wachsmuth). (h) A series of drawings of proximal femurs shows progressive decrease of the mean head-neck-shaft angles with normal growth from 3 weeks of age (150°) (far left) to 15 years of age and then in adulthood (120°) at far right. (Reprinted with permission from T von Lanz and W Wachsmuth). (i) The ranges of anteversion that occur with normal development in the proximal femur are shown. The proximal head and neck of the femur are shaded dark gray, and the distal femur at the condyles (knee) is outlined but clear within. The two femoral segments are drawn as if visualized along the same plane with the proximal part superimposed on the distal. The head-neck axis is the

darkest line and the transcondylar axis the lightest. The angle between these lines indicates the extent of anteversion or retroversion of the head/neck in relation to the distal condyles. The middle drawing (c) shows the normal with a mean angle of 12° anteversion. The images above (b) and below (d) are progressing toward the outer ranges of normal, (b) increasing the anteversion to 20°, and (d) decreasing the anteversion to 4°. At top, (a) demonstrates increased anteversion beyond normal to 37°, and, at bottom, (e) demonstrates clear retroversion of −25°. (Reprinted with permission from T von Lanz and W Wachsmuth). (j) Diagrammatic representation of the labrum in relation to the articular cartilage of the acetabulum and the lateral edge of acetabular bone. Note the continuous transition zone between articular cartilage and labral fibrocartilage. (Reprinted with permission from Field and Rajakulendran, JBJS Am)

1.2 Development of the Hip: Embryonic and Fetal Periods

7

d

e

Posterior limit of external oblique

i Insertion of external oblique Internal oblique

Latissimus dorsi

AN TE RI OR

Tensor fasciæ latæ AL TE LU G

Sartorius

Gluteus minimus

GL UT EA L

Rectus femoris

POSTERIOR GLUTEAL LINE

E N

ARTICULAR PORTION OF ACETABULUM

AE

OA

Synovial membrane

COTYLOID FOSSA

LESSER SCIATIC NOTCH Gemellus inferior COTYLOID NOTCH

OBTURATOR NOTCH Semimembranosus

OBTURATOR FORAMEN

Quadratus femoris Semitendinosus and biceps

DESCENDING RAMUS OF PUBIS Gracilis

Adductor magnus RAMUS OF ISCHIUM

ii Articular capsule, cut Glenoid lip Articular capsule Ligamentum teres

Articular capsule

Glenoid lip Transverse ligament Transverse ligament

Ligamentum teres attached to ischium outside the acetabulum

Fig. 1.1 (continued)

Piriformis POSTERIOR INFERIOR ILIAC SPINE GREATER SCIATIC (ILIO-SCIATIC) NOTCH

Gemellus superior SPINE OF ISCHIUM

TERMINAL LINE Pectineus Rectus abdominis Pyramidalis Adductor longus Adductor brevis

POSTERIOR SUPERIOR ILIAC SPINE

ISCHIUM

Capsule

PB

Gluteus maximus

LI

INFERIOR ILIAC NOTCH

2

CREST OF ILIUM

LINE

INFE RIO R

Gluteus medius

Obturator externus

8

1 Developmental Dysplasia of the Hip

f

g Epiphysis marginalis

Facies lunata

Corpus ossis ilei Labium articulare Os acetabuli

Zona orbicularis Knochenkern des Caput femoris

Spina ossis ischii Corpus ossis pubis

Fossa trochanterica

Lig. capitisfemoris

Knochenkern des Trochanter maior

Symph. oss. pubis Tuber. ossis ischii Synchondrosis ischio-pubica Linea intertrochanterica Knochenkern des Trochanter minor

h

150˚

Fig. 1.1 (continued)

148˚

145˚

142˚

138˚

133˚

120˚

1.2 Development of the Hip: Embryonic and Fetal Periods

i

9

j +37˚

+25˚

Bone +20˚ +8˚

Blood vessels Clacified cartilage layer Labrum Transition zone

+12˚

Femoral head

Hyaline cartilage

-8˚ +4˚

-37˚ -25˚

Fig. 1.1 (continued)

border of which is an almost discrete thickening at the back of the neck referred to as the orbicular ligament or zone (Fig. 1.1b).

1.2.2.8 Skeleton The outline of the developing skeleton of the hip joint consists largely of condensed blastemal tissue in the 12 mm embryo. The acetabulum is barely indicated at 12 mms but is slightly more apparent at 14 and 15 mms. Increase in cartilage depth is particularly noted in the 30–33 mm stages as the pubic, iliac, and ischial cartilages clearly enter into formation of the acetabulum. Both trochanters are well defined but the neck is quite short. By 49–50 mms a center of ossification is present in the ilium near the acetabulum, and vessels have begun to penetrate the acetabular/glenoid labrum and the adjacent cartilage. The femoral head is beginning to be vascularized by vessels from the perichondrium of the neck. Much more extensive vascularization of the femoral head and neck is noted in 85 and 95 mm fetuses. The perichondrium is evident as a distinct condensed layer of cells

enclosing not only the extracapsular part of the cartilages but also the intracapsular portion continuous with the chondrogenous layers of the blastema. The interzone is primarily the blastemal disk between the cartilages but is later continuous at its periphery with the synovial mesenchyme. Later the interzonal and synovial mesenchyme tissue gives rise to the various intra-articular structures and articular cavities formed within it. The rim of cartilage tissue associated with the fibrous glenoid labrum is a constant finding in fetuses 140 mms and larger and is continuous with the cartilaginous glenoid labrum.

1.2.2.9 Summary After 8–9 weeks of embryonic development, the general form of the hip joint resembles that of the adult. As early as 1878, Bernays concluded that the development of the knee and hip joints up until this time is determined genetically rather than being due to mechanical features [19]. Most observers feel that the tissue cutoff by the developing hip capsule persists as the synovial mesenchyme from which the

10

intra-articular structures arise. This tissue is continuous with the interzone. Intra-articular structures arise in situ. The majority of studies also support the fact that the general form and major components of joints are present before cavity formation begins. Smaller irregular spaces are present at the periphery of the joint at 8–9 weeks of age, and these rapidly coalesce to form a singular articular cavity. The appearance of the joint cavity marks the beginning of the period designated by Bernays as the stage of completion. The formation of synovial tissue takes place soon afterward. Concavity of the acetabulum is apparent throughout the fetal period of development from 8 weeks to term. The future acetabulum is indicated as early as 13–15 mms and the acetabular fossa by 22–25 mms. The femoral head and neck are present in 22–25 mm specimens, and both trochanters are shaped by 28 mms. The interzone is seen between the anlagen of the femur and acetabulum. By 20 mms its middle portion becomes thinner and lighter in cell density, and a three-layered arrangement is evident. The outer layers are continuous and serve as chondrogenous layers for the adjacent cartilage. The intermediate layer is continuous with the adjacent synovial mesenchyme that becomes intra-articular in position with formation of the capsule and in which clefts appear by 22–25 mms. Cavities begin to appear at this time first in the synovial mesenchyme. Cellular condensations for the capsule make their appearance in some regions of the joint as early as 20 mms. Ligamentum teres and acetabular (glenoid) labrum appear as cellular condensations at 22–25 mms and the transverse acetabular ligament by 28 mms. All three of these structures arise in situ. Vascularization of epiphyseal cartilage begins by 50 mms.

1.2.3 E mbryonic, Fetal, and Postnatal Development of the Femur Development of the normal shape of the proximal femur is very important clinically. Assessment of several developmental dimensions demonstrates the dynamic aspects of growth that involve not only increases in length and width but also changes of shape and angular position in particular at the proximal end. Felts studied 53 femurs from crown-rump (CR) lengths of 31–485 mms (ninth embryonic week to term plus 1 infant of 3 weeks) (74). The joint cavities form at the hip at 30 mms CR length and at the knee at 33–37 mms CR length. He assessed 15 different growth parameters, the following of which are particularly important to an understanding of developmental dysplasia of the hip: (1) diameter of the head, (2) inclination of the head and neck angle with the longitudinal axis of the femur (head-neck-shaft angle), and (3) torsion (anteversion or retroversion) involving the angle relating the proximal to the distal segments with the proximal angle along the

1 Developmental Dysplasia of the Hip

head-neck axis and the distal along the posterior intercondylar plane. Felts measured the angular dimensions involving proximal femoral torsion, head-neck-shaft, and inclination in a large series of femurs [20]. There have been extensive reports of values for torsion of the adult femur by which is meant the angular displacement of the head and neck anterior (anteversion) or posterior (retroversion) to the frontal plain measured in relation to the bicondylar axis of the distal femur. Of equal importance to the actual degree of angulation is the change during development from embryonic to fetal to postnatal periods. There is generally no anteversion initially in the embryonic and early fetal period; anteversion of the head and neck increases in the few months before birth, then decreases appreciably during the first year of life, and continues to decrease until skeletal maturity. There is a marked tendency for diminution of anteversion angle in the postnatal period. The femoral torsion is much more variable in the late prenatal and early postnatal period than in the adult. From the data presented in several series, femoral torsion has a perinatal value of 30–40° with anteversion decreasing to 12° in adulthood. Several studies have shown that the torsion is actually negative (retroversion) in many in the embryonic and early fetal stages. In studies measuring torsion with embryo lengths less than 29 mms CR, values reported were − 10°, −9°, − 6°, −4°, −9°, −10°, −4°, −22°, and −15°. Le Damany observed no torsion (0°) in the early fetal period with increase toward term [21, 22]. Von Lanz showed a mean anteversion of 10° at 4 months fetal life increasing to 34° at birth and then diminishing throughout childhood and adolescence to a mean value of 11° at maturity [23]. Torsion at any given fetal time shows a high degree of variation between individuals, and its increase in the prenatal period is even greater than its postnatal decrease. Increase in torsion is characteristic of the late fetal period in all studies. Felts refers to studies as early as 1879 with comparable adult data with mean numbers for anteversion of 11.67°, 11.33° (right) and 14.07° (left), 11.63° (right) and 14.71° (left), 11.76° (right) and 9.73° (left), and a bilateral average of 11.23° and 8.02° [20]. In virtually all studies, the standard deviations are fairly wide although the differences between right and left are very small. In one of the largest series using a standardized technique, Fabry et al. documented anteversion radiologically in 864 hips in 432 normal children from 1 to 16 years of age [24]. The average measurement was 31.13° at age 1 year with gradual diminution to 15.35° at 16 years. Harris defined a normal range of anteversion from 35° at birth decreasing to 11° in adulthood [25]. Hoaglund and Low studied Caucasian and Chinese proximal femoral anteversion in 294 adult cadavers [26]. The average anteversion in 112 male Caucasian femurs was 7° (range 2–35°), in 31 female Caucasian femurs 10°

1.2 Development of the Hip: Embryonic and Fetal Periods

(−2° to 25°), in 116 male Chinese 14° (−4° to 36°), and in 35 female Chinese 16° (7–28°). Other mean adult values reported in the older literature are 15.3°, 14.3°, 11.6°, and 11.9°. Efforts to measure anteversion in patients are still associated with problems of accuracy and interobserver variability. Plain radiographic techniques, computerized axial tomography, and clinical examination of hip internal and external rotation in extension all provide an indication of the extent of anteversion, but little appears to be gained clinically in childhood from precise radiographic documentation [27]. A major question has been the following: Where does the torsion occur between the proximal and distal femoral epiphyseal regions? Pitzen felt that torsion is not restricted to the proximal region of the fetal femur but is present throughout its length [28]. Felts also was of the opinion that torsion is not restricted to a localized area but is present throughout most of the shaft [20]. This is different than the humerus where torsion occurs at the junction of the proximal epiphysis and metaphysis. The head-neck-shaft angle of inclination of the proximal femur as measured in the anteroposterior plane has also been studied extensively. The normal neck-shaft angle also decreases with age, but the range of variability is considerable. Harris measured the average at birth as 137°, 145° at 1–2 years, 143° at 2–4 years, 135° at 4–6 years, 134° at 6–8 years, 133° at 8–12 years, and 120°–125° in adulthood [25]. Von Lanz showed a similar pattern of change with the highest mean angle of inclination 145° at 1–3 years of age followed by diminution to a mean of 126° at skeletal maturity [23]. The angle was 135° at 4 months fetal, decreasing to the lowest prenatal value of 122° at 8 months of fetal life. Adult values documented are similar with studies from the late 1800s to the mid-twentieth century indicating values of 124°, 126°, 129.6°, and 126.4°. Humphrey published one of the earliest detailed studies in 1888 [29]. The average angle in 30 adult femurs was 124° (range 113–135°), in 15 additional femurs from individuals greater than 70 years of age it was 123.7°, and in 30 adult bones from Germany it was 128°. He also documented the higher, more valgus, neck-shaft angle in patients with diminished to absent weight bearing capability and in fetuses. Hoaglund and Low noted adult angles of 135° in Caucasian and Chinese femurs with no sex variances [26]. Measurement of this angle in the fetus is difficult because of the variable amount of anti-torsion which tends to increase the angle in the anteroposterior plane, the relative shortness of the neck in the fetus compared with the postnatal period, and the difficulty of measurements since radiographic studies cannot be used owing to the lack of bone in the femoral head. The proportion of growth at the ends of the femur in the postnatal period is 30% proximally and 70% distally. Growth in the embryonic and fetal period is less well studied, but there appears to be equal contribution from each end.

11

1.2.4 E mbryonic, Fetal, and Postnatal Development of the Acetabulum The acetabulum forms initially as a cartilaginous mass of tissue differentiated from the mesenchymal blastema. Laurenson studied the development of the acetabulum in the fetus using arthrograms, plain radiography, gross inspection, and histology in 14 fetuses from 14 weeks of age to full term [30]. At 14 weeks, the acetabular roof is entirely cartilaginous and a well-formed labrum, referred to by Laurenson as a limbus, of characteristic shape projects laterally where it is separated from the joint capsule by the typical lateral recess. The zona orbicularis fits closely around the neck of the femur, the ligamentum teres is present as are the acetabular fossa and transverse acetabular ligament. The bony roof of the acetabulum is beginning to form but is much less extensive than in the newborn. The acetabular bone becomes more prominent with time, but the basic relationships remain unchanged. Lee et al. studied acetabular development from 6 to 20 weeks of gestation. Acetabular anteversion changed little in the early fetal period [31]. The cartilaginous femoral head in the fetus is almost spherical and fits deeply into the acetabulum. There is no marked change in the relative size of the acetabulum and the femoral head with early fetal development. As development proceeds, however, Laurenson showed in the two oldest specimens, one at 300 CR length in mms and one at term, that there was a clear increase in size of the femoral head in relation to that of the acetabulum. Measurements from the 90 mms size onward were essentially equal for the acetabulum and femoral head in relation to the depth of each, but at the latter two time periods, the femoral head was relatively larger than the acetabulum at 11.0–8.5 mms and 10.0–8.5 mms. Le Damany felt that these differences served as a possible precondition for hip subluxation and dislocation in the newborn [21, 22, 32–34]. The relative shallowness of the acetabular socket in late prenatal and early postnatal development has also been noted by Ralis and McKibbin and is one of the possible predisposing causes of neonatal hip instability [35]. The lateral parts of the acetabulum grow both by endochondral expansion and ossification and, at the inner and outer iliac cortices, by periosteal intramembranous bone that, similar to what is seen in the long bone, always forms slightly in advance of the contained endochondral bone. At birth ossification of the acetabular roof has not yet completed its progress from the primary center of ossification at the greater sciatic notch. From that center, ossification spreads inferiorly toward the triradiate cartilage, anteriorly toward the anteroinferior iliac spine, and at later times laterally toward the limbus (labrum). In terms of the developing perichondrial bone in the pelvis, the advancing edges appear separately on the inner and outer surfaces of the ilium. With hip dysplasia, the lateral superior acetabular spur is less prominent than the medial spur, and endochondral bone formation in the lateral part of

12

the roof lags behind that in the medial part. Abnormal pressure of the laterally and proximally subluxed femoral head (i) directly on the labrum and (ii) indirectly on the lateral edge of perichondrial bone and adjacent acetabular cartilage both retard cartilage development laterally and secondarily retard endochondral ossification in the lateral part of the acetabular roof and perichondrial ossification of the outer iliac wall. Diminished pressures on the acetabular cartilage associated with a completely dislocated head also disturb the normal developmental sequence especially laterally. With early development of the human (and other vertebrate) hip in utero, the triradiate cartilage forms, and acetabular development receives contributions from iliac, ischial, and pubic segments. While not seen on single two- dimensional histological sections or on plain radiographs, there is contiguity of acetabular articular and adjacent pelvic cartilage (acetabular cartilage hemisphere) and the triradiate cartilages. The intervening pelvic cartilage (between the physeal triradiate and the acetabular cartilage) is essentially epiphyseal cartilage (within which the iliac, ischial, and pubic bone centers appear) homologous with the epiphyseal cartilages at the ends of long bones. Bucholz et al. referred to the acetabular articular cartilage and the adjacent pelvic cartilage as the acetabular hemisphere [36]. Ponseti referred to yet another region of acetabular cartilage at its periphery as the ring apophysis in continuity with the growth plates of the ilium, ischium, and pubis and with the three flanges of the triradiate cartilage [37]. Fabricant et al. pointed out the normal ossification center along the edge of the posterior wall of the acetabulum appearing around 8 years of age and fusing just before closure of the triradiate cartilage [38]. Harrison and others refer to this area as a natural continuation of more internal tissue, calling it “the articular cartilage at the acetabular rim” rather than an implied specific structure [39, 40]. Some studies have demonstrated these cartilage continuities. Portinaro et al. performed a detailed histologic study on two acetabula from a 3-month-old infant quantifying relative contributions to growth based on histomorphometric measurements of the involved growth plates [41]. As noted in the sections above on the acetabular labrum, the inner fibrocartilaginous part of the labrum is fully continuous with the acetabular articular cartilage and provides significant support to the femoral head. Graf points out that the acetabular roof cartilage is composed of two histologic subdivisions the “acetabular labrum” and the “hyaline- preformed cartilaginous acetabular roof” [42, 43]. Ponseti studied postnatal acetabular development by histologic and radiographic techniques in ten normal full-term infants and three children 7, 9, and 14 years of age [37]. In infancy the cartilage of the acetabular socket is continuous medially with the triradiate cartilage. The acetabular cartilage forms the outer two-thirds of the acetabular cavity with the ilium above the horizontal plane, the ischium below it,

1 Developmental Dysplasia of the Hip

and parts of the triradiate cartilage forming the medial wall of the acetabulum. The pubis is separated from the acetabular cavity by intervening cartilage. Fibro-adipose tissue, referred to as the pulvinar, is interposed between the femoral head and the non-articular depth of the acetabulum. The fibrocartilaginous labrum is at the peripheral margin of the acetabular cartilage, and the joint capsule actually is attached several millimeters above the most peripheral rim of the labrum into the fibrous tissue covering the outer surface of the acetabular cartilage. At the inner and outer margins of the ilium, a characteristic perichondrial groove of Ranvier forms as the intramembranous periosteal bone extends slightly beyond and covers the physis of the acetabular cartilage. In the postnatal period, bone formation gradually increases at the expense of acetabular cartilage. The acetabular cartilage preceding development of the bony acetabulum originally contains elements from the ilium, ischium, and pubis. At the medial depth of the acetabulum, these three cartilage growth plates intersect to form the triradiate cartilage that is composed of three linear components: one anterior and slanted superiorly, one posterior and oriented horizontally, and one oriented vertically (Fig. 1.1c). Interstitial growth within the triradiate cartilage allows for expansion of the hip socket. At puberty the depth of the acetabulum is increased further by three secondary centers of ossification at the periphery of the cartilage (Fig. 1.1d). The os acetabuli, the epiphysis of the pubis, forms the anterior wall of the acetabulum; the acetabular epiphysis, the epiphysis of the ilium, forms a major part of the superior wall of the acetabulum; and a third small epiphysis in the ischium is also formed. The concavity of the acetabulum develops in response to the presence of the spherical femoral head. This is evident in cases of hip displacement where acetabular development is correspondingly abnormal. It has also been shown in a more controlled environment by Harrison who observed that following excision of the femoral heads in rats, the socket failed to develop in terms of depth and area [40]. The growth of the acetabulum involves both interstitial growth of the acetabular cartilage, appositional growth from the perichondrium at the periphery of the cartilage, and eventually intramembranous periosteal new bone formation at the acetabular margin much as occurs in the developing cortex of a long bone [13]. Walker studied histologic development of 74 paired acetabulae from normal human fetuses from 12 to 42 weeks [43]. At a gross morphologic level, the labrum is noted to contribute to a minimum of one-fifth of socket depth and often more. Histologic sections show the labrum to be increasingly fibrous, as distinct from fibrocartilaginous, the closer the fetus is to full term. Cartilage cells intermingle with fibroblasts primarily at the acetabular cartilage-labrum junction. Histologic sections from the developing superior quarter of the acetabulae show bone development beginning from the medial ischial side and the posterior areas adjacent

1.2 Development of the Hip: Embryonic and Fetal Periods

to the sciatic notch. The bone of the superior roof develops first with that of the walls of the socket following. The posterior and medial bone development of the acetabular socket precedes the anterior and lateral bone development. Severin noted the following arthrographic criteria of a normal joint: (1) the labral thorn should lie under or possibly early on 1–2 mms above the horizontal “Y” line of Hilgenreiner; (2) the cartilaginous acetabulum should cover at least one-half of the femoral head; (3) there should not be a great quantity of contrast medium in the bottom of the acetabulum; and (4) the shape of the head of the femur should be practically spherical [44]. The limbus (labrum) lies lateral and slightly superior to the head of the femur. All the articular structures identifiable on the arthrogram of the normal hip of the young infant including the labrum and capsule are present in the 14-week-old fetal hip as well.

1.2.5 E mbryonic, Fetal, and Postnatal Development of the Acetabular Labrum (Glenoid of the Hip) The acetabular labrum is an integral structure of the acetabulum serving to increase the depth of the socket and stability of the hip. Early in the fetal period, the labrum is composed of fibroblasts and loosely arranged bundles of collagen with vascularization throughout. Labral microstructure changes to a well-structured pattern in the later fetal and early postnatal period. Three layers of tissue compose the labrum, transitioning gradually from one to another. (i) The articular internal part is continuous with the articular acetabular cartilage and is composed of fibrocartilage. Chondrocytes are embedded between the collagen fibrils that lie in a plane for the most part parallel to the joint surface. This layer is approximately 200– 300 micrometers thick when the individual is fully grown. (ii) The external part directed toward the hip joint capsule consists of dense connective tissue with flattened longitudinal fibroblasts between collagenous bundles. (iii) A transitional tissue zone lies between the internal and external regions, becoming progressively more fibrous away from the articular surface [45]. The labrum is triangular in shape in vertical sections, being narrowest further away from its acetabular attachment. Vascularity is greatest in the early fetal stage, progressively diminishing but still present even in adults. Regional vascularity of the acetabular labrum in 21 fetuses (5–10 months of gestation) has been studied by Maslon et al. [46]. Blood vessels were counted with microscopic examination of four spaced quadrants. The outer capsular-facing part contained more vessels than the articular-facing part by a ratio of about 3:2 but were evenly spaced in the quadrants. With increasing age the number of vessels in the labrum seemed to decrease. Similar findings in 18 fetuses 11–24 weeks of age were also found revasculature by Turker et al. [47].

13

The blood supply of the labrum originates from radial branches of a periacetabular vascular ring that crosses the osseolabral junction on its capsular side and continues toward the free edge of the labrum [48]. Putz and Schrank note that the anastomotic ring derives from the superior gluteal vessels, obturator artery, and an ascending branch of the medial femoral circumflex artery [49]. The vessels enter exclusively from the capsular-facing surface and then pass toward the articular-facing surface. Vessels do not traverse from acetabular cartilage or bone into the labrum. Kelly et al. outlined the adult pattern of vascularity well [50]. They stressed what one notes visually that while the labrum has blood vessels, it is still relatively poorly supplied. The side was better supplied than the articular side, but anterior, superior, posterior, and inferior labral regions had similar vascular patterns. The orientation of the collagenous fibrils is not uniform throughout the labrum, but there is good agreement in the patterns present, including the fibrillar attachments via the labral-chondral junction [51, 52]. The posteroinferior labrum has very densely packed collagen fibers oriented perpendicular to the junction, while in the anterosuperior quadrants, the fibrils run parallel. Interdigitation of fibers is extensive posteriorly (from labrum to acetabular cartilage), while anteriorly the transition zone is abrupt with minimal interdigitation of fibers since the collagen fibers of the superior (and anterior) labrum are arranged parallel to the labral-chondral junction. This has been interpreted as explaining the stronger anchorage and less tearing posteriorly, while most acetabular labral tears are located anterosuperiorly. It has been estimated that the labrum increases the articular surface area by 22% and acetabular volume by 33% [52]. In a mechanical sense, the labrum creates a seal for the femoral head-acetabular “compartment” keeping synovial fluid localized at this region [53]. The seal also resists distraction of the femoral head (suction cup effect) improving joint stability. By maintaining joint fluid at this region, nutrition and gliding are enhanced. Ferguson and colleagues specifically studied the biomechanical functions of the acetabular labrum seal. In both theoretical and experimental studies, they indicated that the labrum could seal a layer of pressurized fluid between the femur and acetabulum, thus preventing contact of the articulating surfaces. With this sealing effect loads were transferred across the joint predominantly by uniform pressurization of the interstitial fluid of the cartilage layers, thus protecting against increased strains within the solid matrix of the cartilage layers. The labrum added resistance in the flow path of the fluid being expressed from the cartilage layers. When the labrum was removed, the solid-on-solid contact stresses between femoral head and acetabular cartilage layers were greatly increased (to 92% higher) increasing friction between the joint surfaces [54].

14

They considered that the labrum seals the hip joint creating a hydrostatic fluid pressure in the intra-articular space and limiting the rate of cartilage layer consolidation [55]. Consolidation refers to gradual compression of the cartilage layers as interstitial fluid is expressed from the collagenous/ proteoglycan-rich solid matrix of the tissue. The intact labrum maintains a pressurized fluid layer within the joint. Damage to or resection of the labrum lessened its protective effort on the cartilage surfaces, predisposing to frictional wear of the cartilage surfaces and eventual fibrillation and determination. The acetabular labrum has innervation via a branch of the nerve to the quadratus femoris muscle and from the obturator nerve. Many types of mechanoreceptors are present in the labrum. The labrum is wider and thinner in the anterior region and thicker posteriorly. The anterior and posterior edges of the labrum are attached inferiorly to the transverse acetabular ligament over the acetabular notch. The labrum is widest anteriorly and thickest superiorly (laterally) An overall view of the acetabulum is seen in Fig. 1.1e. The bone and soft tissue components of the developing hip are shown in a coronal section drawing (Fig. 1.1f). Pertinent elements of hip formation are summarized in Table 1.1a, b. Table 1.1a Embryonic, fetal, and postnatal hip development Hip joint structures form by differentiating in situ from one mass of undifferentiated mesenchymal cells. The femoral head and acetabulum reach infantile shape prior to joint space cavitation such that dislocation cannot occur in the embryonic period Proximal femur Anteversion Early fetal period: 0° (neutral version) with many studies showing retroversion Middle-late fetal period: increasing anteversion to ~30–35° at birth Postnatal period: fairly rapid decrease from birth to 3 years of age, gradual decrease thereafter to ~10°–12° at skeletal maturity Neck-shaft inclination Maximal at fetal stages at 150°; ~140–145° at birth; progressive postnatal diminution to ~120–125° at skeletal maturity Acetabulum Many studies (but not all) show late fetal growth of the femoral head to be relatively greater than that of the acetabulum causing slight acetabular shallowness at birth Acetabulum forms from iliac, ischial, and pubic cartilage masses with triradiate cartilage in depths of acetabulum Initial acetabular roof bone forms from the posterior, medial region of the ilium adjacent to sciatic notch by endochondral ossification, followed by increasing formation in anterior, inferior, and eventually lateral directions Radiolucent roof over femoral head moving from medial to lateral is composed of acetabular cartilage, fibrocartilaginous labrum, and capsule Acetabular obliquity (anteversion) remains unchanged during development ranging from 15° to 30° but averaging 20°

1 Developmental Dysplasia of the Hip Table 1.1b Formation of the acetabulum Three bones combine to form the acetabulum: ilium, ischium, and pubis. Cartilage models of each bone form in embryonic period, and structure of cartilaginous acetabulum is formed by 8 weeks, the end of the embryonic time frame Ossification begins in each bone via endochondral ossification within the cartilage model. Ossification occurs first in the central ilium of the fetus at 8–10 weeks, positioned approximately halfway between developing hip joint and iliac crest. Ossification front then moves progressively outward toward transverse triradiate cartilage below and iliac crest apophysis above and meets up with inner and outer cortical intramembranous periosteal bone formation that accompanies the central endochondral bone in both directions. The other centers also form by the endochondral sequence; ischial bone center forms slightly before pubic bone center at 4–6 fetal months. Triradiate cartilage continues to separate each bone segment and is a characteristic feature of acetabular growth. The three bone centers and triradiate cartilages are fully formed around 7–9 years of age and then grow proportionately throughout the remainder of the growth period. The ilio-ischial growth plate is oriented in the horizontal plane, is the largest of the three growth plates, and contributes the most to growth in length Toward the end of the first decade, three additional ossification centers form at the periphery of the acetabulum, one in each bone. These are analogous to secondary ossification centers of long bone epiphyses. Os acetabula forms at 8 years of age as the epiphysis of the pubis and contributes to the anterior wall of the acetabulum; acetabular epiphysis forms in the ilium also at 8 years and contributes to the superior surface of the acetabulum; and a small unnamed ischial epiphysis forms in most in adolescence at the posterior acetabular surface Fusion of all acetabular bone centers and cartilages becomes complete at a wide range between 20 and 25 years of age The ilium and ischium both contribute to forming about two-fifths of the acetabulum and the pubis contributes to forming about one-fifth

1.3

rimary Etiologies of Hip P Maldevelopment

The hip can become subluxed or dislocated late in utero, in the immediate postnatal period, in infancy, or in childhood. The secondary effects of displacement are dependent on the primary cause, the time during the developmental period that the displacement occurs, the age at which it is detected, and the effectiveness of treatment. In the newborn period, the distinction of an abnormal hip currently lies between (i) an idiopathic developmental dysplasia of the hip and (ii) a teratologic hip dysplasia caused by an associated disorder which affects hip development prior to birth. Idiopathic developmental dysplasia of the hip refers to a perinatal hip displacement in an otherwise normal child presenting as a subluxatable or dislocatable hip owing to an isolated hip capsular laxity with proximal femoral and acetabular changes then occurring secondarily. Teratologic hip dysplasia refers to the presence of developmental abnormalities other than capsular laxity, which originate in the embryonic or early fetal time period, are generally more severe than the idiopathic entity leading to a subluxed or dislocated hip prior to birth, and are associated with neurologic,

1.4 Etiology and Pathoanatomy of Developmental Dysplasia of the Hip

muscular, or connective tissue disorders. Recognizable teratologic abnormalities leading to hip dysplasia in the newborn are spinal dysraphism syndromes including meningomyelocele, severe myopathies, skeletal dysplasias, joint laxity syndromes (Ehlers-Danlos), and a wide array of dysmorphic syndromes, most of which are chromosomal or genetic in origin. There appears to be a spectrum of teratological disorders ranging from the obvious association of hip dislocation with gross structural abnormalities throughout the body to what may well be localized hip region abnormalities of embryonic or early fetal derivation in an otherwise normal appearing child. A small percentage of dislocatable hips in the newborn are associated with unrecognized neuromuscular disorders, generally myopathies, which may not become clinically apparent or be diagnosed for several months or even years after birth. There may be an overlap between mild and currently undetected teratologic abnormalities localized to the proximal femur and acetabulum that date to the embryonic and early fetal time periods and predispose to perinatal subluxation and dislocation. At pathoanatomic assessment, this appears as a capsular laxity and not as a primary bone or cartilage abnormality. There may be associated soft tissue abnormalities in a teratologic sense that are present throughout the fetal period and render the hip nonresponsive to early therapy. The terms idiopathic and teratologic are imprecise but at present provide for a reasonable indication of cause and expected response to basic treatments. Many neuromuscular abnormalities cause a weakness and imbalance in the hip musculature and the secondary development of femoral-acetabular structural abnormalities postnatally over several months to years. The cerebral palsies, spinal muscular atrophies, and myopathies often present with hips normally positioned at birth that proceed to develop subluxation and dislocation owing to muscle imbalance and delayed or abnormal gait patterns. Proximal muscle weakness, prolonged non-ambulation due to neurologic or neuromuscular disorders, and even limited ambulation with an inability to run all predispose to retention of the newborn proximal femoral characteristics of increased anteversion and inclination angles and coxa valga. These femoral findings especially with asymmetric weakness and abductor muscle tightness predispose to hip subluxation and dislocation. The structural development of the proximal femur, particularly the normal diminution of anteversion and inclination angles, is highly dependent on normal gait patterns.

1.4

Etiology and Pathoanatomy of Developmental Dysplasia of the Hip

This section will refer in detail to descriptions of hip pathology and only briefly to theories of pathogenesis, most of which have been proposed with little or no experimental

15

verification. We will clarify where possible whether the observations made were representative of idiopathic developmental dysplasia of the hip (a hip disorder in the absence of any other apparent abnormality) or of teratologic dysplasia (where other abnormalities were found). Such distinctions, often not made by the original authors, are important. Many pathologic dissections seemingly describing the underlying pathoanatomy of idiopathic developmental dysplasia of the hip have been made on babies who were either stillborn, died in the newborn period, or lived only for a few months, so the descriptions would appear to represent teratologic hips rather than an isolated idiopathic developmental dysplasia of the hip. Our knowledge of the etiology of hip dysplasia is incomplete, and this distinction may be arbitrary, but it is one that has persisted for some time.

1.4.1 Early Clinical-Pathoanatomic Descriptions 1.4.1.1 Palleta, 1788 The earliest detailed clinical and pathoanatomic descriptions of congenital dislocation of the hip were written by Palleta of Milan, Italy, in 1788 [56] and 1820 [57]. Palleta described five cases, two clinical and three at autopsy including bilateral dislocation of the hips in a child who died at 14 days of age. Both femoral heads that remained spherical were situated above the acetabulum but had not yet been surrounded by any new socket. They were present in the area of the anteroinferior iliac spine. The acetabulae were completely filled with a fatty substance, the anterior portion of the socket was closed by the acetabular ligament that had been turned on itself, and the joint capsule was much broader and looser than normal and very thick. The intra-articular ligament was longer than normal. Examination of the enlarged capsule and lengthened ligaments allowed one to appreciate the increased movements of the femoral head in several directions. Palleta’s work established the existence of the hip dislocation in a child of only 14 days of age that based on the pathoanatomy was of congenital and not traumatic origin. He attributed the disorder to an original defect of the germ, that is, a genetic disorder. Delpech referred to and quoted extensively from Palleta’s work in his book De l’Orthomorphie, published in 1828 [58]. 1.4.1.2 Dupuytren, 1826 One of the early clinical descriptions of congenital dislocation of the hip was made by Dupuytren [59]. He clearly defined the entity in terms recognizable today, discussed the possible underlying causes, distinguished it from other abnormalities of the hip, and commented on treatment. Until that time it had not been widely recognized as a specific entity. The disorder involved a transposition of the head of

16

the femur from the acetabulum onto the external wing of the ilium, a transposition which one observed from birth and which appeared to be the result of a defect in the acetabulum that was not as deep or as complete as normal rather than being due to trauma or disease. The disorder belonged to the class in which dislocations of the femur were upward and outward. Some practitioners had previously been aware of two varieties of dislocation defined as traumatic or pathologic. Dupuytren sets out to describe the third type which he called original or congenital dislocation to distinguish it from the others. He recognized the displacement of the head of the femur upward and outward, shortening of the involved limb, rising of the head of the femur onto the external iliac wing, prominence of the greater trochanter, shortening of almost all the muscles of the superior part of the thigh toward the iliac crest where they formed around the head of the femur, uncovering of the ischial tuberosity owing to the displacement of the muscles, rotation of the limb internally, displacement or telescoping of the thigh up and down and inward and outward, a slanting or obliquity of the femur which was greater the larger and older the patient, a sharp angle of the thigh with the pelvis, and a thinness of the entire limb in particular at its superior part. He noted the limited movements of the involved hip especially those of abduction and rotation, all of which led to difficulty with standing, walking, and lower extremity exercises. The involved lower limb was atrophied in relation to the opposite unaffected side, trunk, and upper extremities. The pelvis was large and prominent; there was a horizontal position of the pelvis on the femur owing to the displacement and a lumbar lordosis. He described what would later be referred to by Trendelenburg as a waddling gait. He distinguished the congenital dislocation from traumatic dislocations or dislocations owing to disease since there was a lack of swelling, abscess formation, fistulas, or evidence of scarring. Those affected with congenital dislocation did not experience any discomfort in childhood either of the hip or the knee but did notice fatigue and numbness with walking. He noted frequent bilaterality. The clinical signs were due to elevation of the head of the femur onto the iliac fossa and shortening and prominence of the muscles drawn toward the iliac crest. If attention was called to the disorder early, at birth, several clinical signs were already present: widening of the hips, prominence of the greater trochanters, and the obliquity of the femur. The tendency, however, was for patients to present for assessment only after they had begun walking when awkwardness of gait was noted. In many individuals, presentation for medical attention was not until 3 or 4 years of age. Basic understanding of the disorder was limited since there was little opportunity for pathologic study as the patients were otherwise well. In a few cases he studied however, the muscles were always pulled up toward the iliac crest; some were remarkably developed, but others were thin

1 Developmental Dysplasia of the Hip

and atrophied. The hypertrophic muscles were those which continued their activities; others had their activities limited owing to the change of position and often were so fibrotic and yellow that one could scarcely note any muscle tissue persisting. The superior part of the femur for the most part preserved its form although on occasion the internal and anterior part of the head of the femur lost its roundness owing primarily to positioning against parts that were not shaped to receive it. The acetabulum of the iliac bone was either absent completely or offered only a small irregular bony prominence with no trace of articular cartilage or synovial capsule. The acetabulum was filled with a fibrous resistant tissue covered by muscles inserting onto the lesser trochanter. The round ligament (ligamentum teres) was greatly lengthened, flattened superiorly, and worn away in certain areas owing to pressure and rubbing of the head of the femur. The head was lodged into a cavity analogous to that which developed in femoral head dislocations following trauma; the new cavity was superficial, bereft of a rim, and situated in the external iliac fossa above and behind the acetabulum, a position which was proportionate to the shortening of the limb. In summary, the findings in these subjects were similar to those seen in cases of pathologic or traumatic dislocation with the difference being that those he examined had findings that appeared to date from a more remote time and to have been so positioned from the earliest time of life. Dupuytren listed possible causes of femoral head displacement: (i) contracted from the mother illness affecting the hip region during the fetal time period, (ii) trauma which had displaced the head of the femur from the acetabulum after which the hip developed abnormally in the absence of normal function, and (iii) maldevelopment of the acetabular socket as an evolutionary problem in particular since the socket was a complex union of three pieces of bone. He did not place any faith in the first theory of prenatal disease. The second possibility of a force that caused the head of the femur to displace from the cavity was felt to be credible. He felt that this was possible owing to the position of the fetus in utero which was one of marked flexion of the lower extremity which forced the head of the femur continually against the posterior and inferior capsule of the joint causing a strain which was without effect in healthy individuals but perhaps caused a problem in others less well constituted whose tissues were less resistant. This relatively weakened region therefore would be susceptible to passage of the head of the femur out of the socket allowing a dislocation to happen. The final possibility of maldevelopment of the acetabulum was also feasible; such a possibility was related to embryologic and anatomic studies on embryo-fetal development which indicated that the final regions of hip development were those of the joint cavities and in particular those where several regions of the bone were required to unite such as in the acetabulum. It was known that the acetabulum was com-

1.4 Etiology and Pathoanatomy of Developmental Dysplasia of the Hip

posed of three separate segments and that the formation of this cavity appeared to be one of the final aspects of hip development. If development of the hip socket did not form in relation to the femur, the femoral head was then displaced to the external region of the ilium. Dupuytren indicated that “in the three preceding hypotheses, the displacement of the head of the femur was not only congenital; in each that we have come to examine it was original and dated from the first organization of the parts. It is a defect of original conformation, a defect in the organization of germs.” He presented several examples of multigenerational familial involvement indicating that a hereditary predisposition existed, and he also pointed out that almost all the individuals affected were female.

1.4.1.3 Sedillot, 1835 Sedillot recognized dislocation of the femur and its presence on the external iliac fossa [60]. He defined the formation of the new capsule in a dislocated hip which he described as continuous with the old such that the new articulation on the side wall of the iliac fossa was in continuity with the previous acetabulum. He commented on the production of the false acetabulum and indicated that the view that congenital dislocations were due to the absence of the acetabulum was incorrect. He felt that his work had established that “the most frequent cause of congenital displacement of the femur is the looseness and relaxation of the ligaments which remained intact and allowed the great mobility of the thigh.” He described the late pathoanatomy of two hip dislocations detailing cadaver studies on a young woman with bilateral hip dislocation who was between 20 and 25 years of age and a woman of 35 years of age with a unilateral dislocation. In the first patient, he noted that the capsular ligament had remained intact and that the original acetabular cavity that no longer contained the femoral head was triangular in shape and filled with synovial tissue. The round ligament was thin and flattened. The femoral head was smaller than normal and flattened in that region which corresponded to its relationship to the iliac bone. It had entirely lost its former spherical shape. The femoral neck was quite short. The appearance was felt to be indicative of a complete relaxation of the ligamentous apparatus that thus played a causative role of extreme importance. The dislocation or at least the disposition to it was present at birth and was thus congenital, and many of the subsequent abnormalities were secondary owing to the displacement of the femur. The appearance of the original acetabulum that was smaller and shallower than normal could easily be explained by the long duration of the dislocation and by the fact that any bony cavity would become obliterated when it did not contain the body which it was naturally destined to contain. Similar findings were present in his second dissection. The original acetabulum was smaller and less deep than normal. The round ligament was intact. The femo-

17

ral head was atrophied particularly at its superior part being somewhat conical in shape and flattened against the iliac region. The femoral neck was short. The external iliac fossa had formed a cavity that was quite deep to serve as a false articulation for the femoral head. He concluded with 12 observations: (1) dislocation was long-standing; (2) dissection revealed the possibility of reduction; (3) there was proportional atrophy of the head of the femur and of the acetabulum; (4) the head was accompanied by formation of a new capsular ligament which was continuous and united with the old capsule in a manner which produced a large capsule embracing both the old and the new articulations and leaving a free communication between them; (5) the persistence of the round ligament (ligamentum teres); (6) development of a new fiber-cartilage on the portion of the external iliac bone which formed the depth of the new false articular cavity for the femoral head; (7) deposit of bone on the external wing of the ilium surrounding the head of the femur in such a manner as to represent a new acetabulum; (8) formation of a new joint between the lesser trochanter and the anterior surface of the ischium; (9) alterations in shape which were, however, less deep than the original in relation to the pelvis; (10) femoral atrophy which involved primarily thinness of the bone which was less marked distally than proximally; (11) the description of the best positions to lead to repositioning of the dislocated bone into its natural cavity; and (12) the immediate re-establishment of movement of articulations following reduction. The work of Sedillot, even though it was based on studies on young adults, pointed out the key role played by capsular laxity and secondary deforming changes in the appearance of congenital dislocation of the hip.

1.4.1.4 Pravaz, 1847 Pravaz of Lyon, France, wrote a detailed treatise on congenital dislocation of the hip (femur) covering the entire spectrum of the disorder [61]. He reviewed several pathoanatomic descriptions quoting in detail from the works of Palleta and of French physicians. In summary: • The congenital hip dislocations were recognized as a distinct entity from traumatic dislocations and could be either bilateral or unilateral. • The dislocated femoral head rested free on the external surface of the ilium with motion limited only by the surrounding capsule soft tissues and attached muscles, or, in some cases, the femoral heads were partially stabilized by a reactive pseudarthrosis that formed. • The hip joint capsule had elongated but was intact, and its thickness was augmented superiorly and posteriorly. • The capsule assumed a cone-like shape with a communication through a narrow channel between the original acetabulum and the displaced femoral head in the

18

•

•

•

•

•

•

•

1 Developmental Dysplasia of the Hip

expanded capsular portion in the early years of life, but this became obliterated with advanced age. The inter-articular ligament (ligament teres) was stretched and flattened, passed through the narrow channel in the capsule while still attached to the femoral head, was squeezed between the femoral neck and the acetabular sourcil, and eventually wore out and was destroyed. A false articulation formed above the normal socket with the orbicular ligament eroded at the new sit; the new cavity enclosed the head of the femur in a receptacle lined partially by bone and membranous tissue with a tubal membranous connection between new and original cavities that eventually was interrupted. The original acetabulum, after displacement of femoral head, was not completely obliterated but was only deformed becoming oval or triangular and reduced in size but retaining a sufficient capacity to accept the femoral head with little ability to retain it. The dislocated head of the femur loses its sphericity, becoming oval and flattened with the neck shorter and horizontal and the entire femur shorter and thinner. The muscles of the hip (pelvic-femoral) were stretched, had altered direction, and most tended to a fibrous or fatty appearance. In the early years after birth, head displacement was not always complete; the head was sometimes retained on the rim of the socket, depressing it and hollowing out a supplementary socket adjacent to the original with some continuity between them. With time, the continuing action of gravity and the shocks of walking augmented the initial displacement with the head becoming more elevated onto the pelvis and leaving traces of variable depth that it had created.

1.4.1.5 Cruveilhier, 1849 Cruveilhier, a prominent professor of pathologic anatomy in Paris, wrote on congenital dislocation of the hip, summarizing the findings of Palleta and Dupuytren and discussing seven cases he had observed [62]. Most of his cases were bilateral which tended to rule out any traumatic episode. The capsule was invariably enlarged without rupture again leading to the non-traumatic etiology; it remained enlarged and thickened in the congenital form of dislocation. There were invariably abnormalities in development of the acetabulum although the original socket could always be identified. The varieties of acetabular change were great. In some instances it maintained its size and shape reasonably well such that it would have been able to receive the femoral head if reduction could have been brought about. In other instances, the adjacent ligaments obstructed the entrance to the acetabulum, and often the acetabulum was filled with a fibro-fatty tissue. The femoral head was generally smaller than normal

with flattening of one part of the surface of the head often seen. The round ligament or the ligamentum teres was invariably longer than normal. Cruveilhier recognized that during fetal life individuals, who were subsequently born with a congenital hip dislocation, had presented a normal conformation of the affected joint with the acetabulum and femoral head in perfect relationship. Some cause served to remove the femoral head from the acetabulum, and owing to the laws of bone development, the empty acetabular socket became narrower, deformed, and filled with fat, while the head of the femur also deformed owing to the abnormal and unequal pressures to which it was subjected in its new position. The major problem was to determine the cause of the dislocation in the hip that previously appeared to be developing normally. After much discussion, which focused on the reality of an enlarged capsule and round ligament, he felt that external force placed on the fetus via uterine compression in association with certain intrauterine positions of the hip and limited amniotic fluid all conspired to lead to the dislocation. The external pressure was not considered to be a violent, one time phenomenon but rather something lasting over a considerable period of time in association with the other findings.

1.4.1.6 Carnochan, 1850 Carnochan (1817–1887) was an American orthopedic surgeon who spent several years at the start of his career studying and working in New York, London, and Paris, where he concentrated on defining congenital dislocation of the hip [63]. He then returned to a lengthy practice in New York City. He defined the disorder as “…..a transposition of the head of the femur from the cotyloid cavity (acetabulum) upon the external iliac fossa of the os innominatum occurring during intrauterine existence, generally not so fully manifested in the early period of childhood as it becomes in adult age…...” His first description was published in an article in Lancet 1844; 43 #1099: 781–785 titled “On congenital luxation of the head of the femur upon the dorsum ilii.” He described a 19-year-old male with bilateral congenital hip dislocation who was otherwise healthy and showed no signs in childhood of infection or severe trauma. His description of the disorder was the first clear recognition of the entity in the English literature both the United States and England. He pointed out: “this congenital affection…..is of much more common occurrence than the slight notice, or rather entire silence observed, regarding it by authors in Great Britain, or my own country (USA).” At the time he described the entity, there was no existing awareness of it in English-speaking nations, and it was Carnochan who definitively defined it as quite common, pointing out also that there were 25 cases described by Dupuytren and more than 30 by Guérin in France. Carnochan indicated he alone had seen 24 cases in New York, Paris, and London by the time

1.4 Etiology and Pathoanatomy of Developmental Dysplasia of the Hip

his book was published in 1850. His book on the disorder published in 1850, A Treatise on the Etiology, Pathology, and Treatment of Congenital Dislocations of the Head of the Femur, was the first devoted entirely to discussing the entity. It showed a clear understanding of the disorder discussing: normal and abnormal fetal developmental anatomy of the acetabulum, proximal femur, hip joint capsule (“hourglass deformity”), ligaments, and controlling musculature; etiology; symptomatology and diagnosis, with detailed drawings and descriptions of clinical examination and altered gait with unilateral and bilateral dislocations; prognosis; pathology, accompanied by detailed descriptions of adolescent/ adult fully developed deformities accompanied by pathoanatomic drawings of the skeletal deformations of femoral head, empty acetabulum, and false acetabulum on the dorsum of the ilium; and beginning treatment approaches at the time. He considered the disorder to be caused by “a pathological spasmodic retraction of the muscular tissue, resulting from a perverted or distorted condition of the excito-motor apparatus of the medulla spinalis.” While this mechanism is no longer considered the cause of what is still referred to as idiopathic congenital (developmental) dislocation of the hip, his overall description of the mechanism is still recognizable. He considered the disorder to be congenital “indicating the existence of the disease when the child comes into the world.” He felt that the disorder began intrauterine at about 3 or 4 months of gestation; as the hip developed in a position of flexion excess muscle activity “will induce the head of the femur to slip from its cavity over the posterior part of the margin of the acetabulum….” “Having once passed the border of the cotyloid cavity (acetabulum), the extension of the limb after birth will throw the head of the femur still more upon the ilium and the retracted muscles, continuing to act with other causes……..will…….. induce the head of the femur to glide still further (away from the socket onto the dorsal surface of the ilium)….” He recognized that pathoanatomic changes were minimal at time of birth and that it was only after several years that the severe deformities with growth rendered the deformity essentially incorrectable at that time. “If the examination of this displacement be made during the foetal period or when extra-uterine life has been of short duration, the cotyloid cavity is found to be but little altered in its normal shape and dimensions, and to retain the capability of receiving the head of the femur. The period of life at which the cotyloid cavity begins to assume an alteration of shape and dimensions is not the same in all cases…….(but certainly by 12 or 14 years of age)…..the cotyloid cavity in the progress of the affection tends to become contracted and to assume an oval or even triangular shape…..(and become)…..nearly filled up with softish adipo-osseous tissue.” He recognized that the deformities worsened with time the longer after birth they persisted. Carnochan also recognized and described

19

“congenital subluxation” where “the head of the bone rests on the margin of the acetabulum.” His descriptions of the abnormal gait, other than being extra lengthy, recognized the abnormalities later defined more precisely by Trendelenburg. “In walking owing to the want of fixedness of the heads of the femur and the displacement which they must undergo of alternate depression and elevation according as the weight of the body is transferred from one inferior member to the other and also owing to the strain which is put upon the psoas and the internal iliac muscles upon the side where for the moment the weight of the trunk is thrown a kind of double lameness is produced somewhat resembling the hobbling motion of the duck.” “The subjects so affected when about to commence walking, are seen…….to incline the superior part of the trunk towards the member (limb) which is about to support the weight of the body and to lift the other with an effort in order to bring it forward…” Carnochan described the early efforts at treatment by others in the book but appears not to have been involved with any innovative treatments the rest of his career. He defined the entity clearly however and also pointed out the need for early diagnosis and treatment if there was to be any chance of restoring normal or close to normal function. “From a consideration of the serious evils which a continuation of this displacement will entail upon the sufferers,……it behoves the surgeon to be prepared to form a correct diagnosis……..in relation to this affection……..in order that the proper therapeutic means, as yet unknown,…..may…… remedy the deformity before the approach of adult age would proscribe the utility of such approach.”

1.4.1.7 Roser, 1864 and 1879 Roser as early as 1864 pointed out that congenital dislocation of the hip resulted from an abnormal adducted position of the legs during fetal life [64]. He “based this belief on observations in children whose flail hips could be dislocated by adduction of the leg and then reduced again by abduction.” In a second article in 1879, he again made these points and pleaded with his obstetrical colleagues to actually perform this test on the newborn child [65]. He indicated: “that children no longer be allowed to reach the age of 2 years before their hip dislocations are diagnosed.” Roser was accurate in his impression that most if not all cases of congenital dislocation of the hip could be diagnosed by a newborn exam. He also recommended a treatment, not widely adopted for several decades, which could bring about reduction and cure of the disorder. He indicated that “I believe that many, even most, of these cases would still be curable if the disorder were detected in the newborn and if the necessary abduction appliance were applied at once. I believe that with plaster boots held apart by a cross bar or cross board the object would be most simply obtainable.”

20

1.4.1.8 Verneuil, 1866 Verneuil studied and wrote on childhood hip dislocation throughout a long surgical career. Hip Dislocation due to Infantile Paralysis His primary theory of hip dislocation, propounded initially in 1866 [66, 67], was that the large majority of hip dislocations seen in clinical practice and referred to as congenital really represented dislocation after birth secondary to some form of infantile paralysis. He attributed congenital dislocation of the hip to paralysis of the gluteal muscles. Although recognizing that occasional cases of hip dislocation occurred secondary to pathologic intra-articular disease, he considered them to be rare. He and his colleagues had been searching specifically during the course of their pathoanatomic dissections in newborns for isolated congenital dislocations over a period of several decades and had never subsequently seen the lesion in question. The conclusion that Verneuil reached therefore was that “really the luxation did not exist” and therefore the term congenital was inaccurate since dislocation was essentially never seen at the moment of birth but was produced afterward. He did not accept the viewpoint that in most cases congenital dislocation was due to defects of hip development such that the dislocation was present at the time of birth but offered no signs to allow it to be detected. According to that opinion, the head or the acetabulum was misformed but still appropriately related and did not displace until walking began. Verneuil reasoned that if there were predisposing features to delay the luxation, they should be demonstrable either in the bones, ligaments, or muscles at pathoanatomic assessment. True congenital or intrauterine dislocations directly observed at the moment of birth were extremely rare, while the large majority of dislocations became evident around the beginning of the second year of life and then increasingly during the course of the first decade. In his clinical practice, he had seen over 300 cases of dislocations referred to by others as congenital. He did not accept the delayed appearance of a dislocation in a predisposed, but slightly abnormal hip. He made the not unreasonable conclusion that if the anatomist did not find any hip dislocation in the fetus at full term at dissection, perhaps it really did not exist. If the dislocation was not observed at birth and only was diagnosed later, it was not truly congenital in the accurate sense of the word. Although Verneuil was even then swimming against the tide, his argument was not without merit. It addressed several medical and legal concerns in terms of the time of diagnosis of the disorder and in a sense has been incorporated into the current terminology of developmental dysplasia of the hip in which the profession now recognizes, at least on a partially scientific basis, that the displacement may not be either present or even diagnosable at birth but only appears later with increased stresses. Verneuil indicated that many adopting the congenital term in

1 Developmental Dysplasia of the Hip

their writings were continually referring to rare isolated cases from the literature rather than performing newer studies of their own. Verneuil summarized his view indicating that a certain number of displacements of the femur are due to the paralysis more or less complete of the groups of muscle that surround the hip joint, in other terms to the weakening of the pelvi-trochanteric muscles and particularly the gluteal muscles. He indicated that it was necessary to change the name of the disorder: the term “congenital” was not acceptable since displacements often occurred a long time after birth, while others occurred during intrauterine life, “original” was too vague and didn’t signify much, and “spontane” could refer to other mechanisms such as external violence or pathologic disorders. He indicated that since the displacement was due to a preceding (antecedent) disease state and was a pathologic dislocation, it should be referred to as a paralytic dislocation of the femur (“luxation paralytique”). He clearly recognized however that severe teratologic dislocation occurred including those of proximal femoral focal deficiency and multiple congenital anomalies in stillborn fetuses. He concluded his work by commenting on the accuracy of the paralytic luxation etiology [68].

1.4.1.9 Reclus, 1878 Reclus accepted and clarified the views of Verneuil on the causes of what was widely referred to as congenital dislocation of the hip [69]. Since the disorder was not recognized at birth and the dislocation did not occur or at least become evident until walking age, they were reluctant to refer to it as “congenital.” They considered that it was a localized or regional paralysis of the hip region muscles (expressed as the muscles of the buttocks and the pelvi-trochanteric muscles) that combined with the stresses on the hip of walking that led to the dislocation. They clearly indicated that it was a combination of atrophy/paralysis of the gluteal and adjacent peritrochanteric musculature along with the persistent functioning and integrity of the adductor muscles of the thigh that led to the common displacement. This disorder therefore was most commonly an acquired paralytic luxation rather than a congenital malformation. It was the unrecognized infantile paralysis that was the causative factor leading to what was in effect a “luxation paralytique.” Reclus described several cases, and his article was really one of the earliest and clearest descriptions of what we refer to today as a neuromuscular dislocation. From his examples, the direction of the dislocation was dependent on which groups of muscles were paralytic and which antagonists remained functional. He indicated that even if their observations did not explain all cases considered as “congenital,” they did explain many more than the skeptical orthopedic community accepted. During that era there was extensive

1.4 Etiology and Pathoanatomy of Developmental Dysplasia of the Hip

clinical and neurophysiologic definition of entities, and Reclus refers to the work of Duchenne and Charcot to strengthen his arguments. He pointed out that awareness of these paralytic luxations was commonest between 6 months and 3 years of age. Reclus derived several conclusions about their observations on the entity widely referred to as congenital dislocation of the hip (femur): • From the group of dislocations referred to as congenital, it was necessary to separate out those that were paralytic dislocations. • These dislocations follow muscle paralysis (“amyotrophies”) and can occur at any age even if they were barely noticed or overlooked in infancy. • For a dislocation to occur, two conditions were necessary: atrophy (paralysis) of one muscle group and persisting function of its antagonists. If all the regional muscles are paralyzed, there are looseness and exaggerated mobility of the joint but no dislocation. • In the hip, iliac dislocation of the femoral head is most frequent due to the muscle pull of the functioning adductor thigh muscles (antagonists) and atrophy (paralysis) of the muscles of the buttocks (gluteal muscles) and other pelvi-trochanteric muscles which provided no counterbalancing effects. • The influence of gait and unbalanced muscle function can take several months to years to produce displacement which includes overcoming resistance from the joint capsule and the (Y) ligament of Bertin.

1.4.1.10 Brodhurst, 1876 Brodhurst noted that the vast majority of the congenital hip dislocations were displaced upward and outward with the head lying on the dorsum of the ilium [70, 71]. A marked 3:1 female-male preponderance was noted. He discounted the theory of spasmodic muscular retraction that had been offered by Guérin and Carnochan in separate works. He felt that “the cause of congenital dislocation of the hip, as it usually presents itself, is a purely mechanical cause.” The dislocation occurred with non-routine or difficult labor and especially with breech presentations. In this position the head of the femur must press against the posterior and interior portions of the capsule of the joint so that traction in this position (associated with birth) will readily cause the head of the bone to escape from a shallow acetabulum. The dislocation was “produced at birth” through downward force supplied to the thigh in “endeavoring to hasten the birth in breech presentations.” Some very rare instances of congenital dislocation had occurred where the head of the femur was misformed, the cavity of the acetabulum was imperfectly developed, and other deficiencies and abnormalities

21

existed, a situation that refers to what we now call a teratologic dislocation. He indicated that the latter were rare and were quite separate from the types of congenital hip dislocation he was discussing. In the idiopathic variety that he was discussing, children were healthy, well developed, and well nourished. Pathological Anatomy In the idiopathic congenital hip dislocation, the acetabulum at birth was never altered in shape or dimension, and the head of the femur retained its normal appearance. Changes took place, however, secondarily with persistence of the dislocation both in the acetabulum and in the head of the femur as the cartilage wasted and the acetabulum became filled with cellulo-osseous material, while the head of the bone became somewhat irregular in shape and its cartilage became thin. The capsular ligament retained its integrity but became elongated, and the ligamentum teres was stretched and eventually became slender and finally gave way. The head of the bone came into direct contact with the ilium. A false articulation was formed which ultimately developed a new capsule, while “a cavity is formed to receive the head of the bone by deposition of osseous matter upon the ilium.” Brodhurst stressed that “when dislocation occurs without other abnormality both the acetabulum and the head of the thigh bone are usually perfect at birth.” He indicated that treatment at birth should be relatively simple but rarely occurred since the diagnosis was not made at that time. By the time the diagnosis was made after months and years had elapsed, secondary changes had already taken place which tended to impede reduction. In the third edition of his book Observations on Congenital Dislocation of the Hip, Brodhurst (1896) defined four types of congenital dislocation with type I being what we refer to today as idiopathic or developmental dysplasia of the hip [71]. He again commented on the mechanical causation in which the high incidence of breech presentation and the trauma associated with delivery displaced the femoral head. The dislocation was overlooked commonly at birth and came to clinical evidence only when the child began to stand and walk. A waddle was seen in association with the tilted pelvis and lumbar lordosis. He recognized that “when dislocation is recognized early, the head of the femur may be immediately restored and retained in the acetabulum. When it remains displaced for several years changes take place, such as retraction of muscle, filling up of the acetabulum, and some flattening of the head in the bone.” He also indicated that “in those cases in which displacement had not occurred in utero…… the parts are fully formed so that neither is the femoral head so much flattened nor is muscular retraction so great; the acetabulum also is found to be fully developed. In this series not only is development complete, but there is no other abnormality present.” The second type of hip dislocation was “produced in utero” being what we refer to as tera-

22

tologic. He felt in agreement with ideas common at the time, such dislocations occurred as a result of accident, shock, and spasmodic action. Others had referred to this type of development as arrested both in the acetabulum and in the head of the femur, but he suggested: “it occurs in consequence of the parts being displaced in utero before the development is complete.” In the third category, he placed dislocation resulting from inflammation and destruction of the joint that we recognize as septic dislocation. In the fourth class were placed malformations such as those associated with spina bifida or clubfoot that we recognize as a teratologic dislocation. He pointed out that dislocations had many causes and that the pathoanatomy was bound to differ in various types. After several years of dislocation, “the acetabulum becomes more or less filled up with fatty or with fatty and fibrous matter with retraction of the trochanteric muscle so that even though the head of the femur may be reducible difficulty will be experienced in retaining it in position.” He again contrasted the neonatal dislocation produced at birth in which the femoral and acetabular components were normal with dislocation that occurred in utero where the acetabulum and femur will not be fully developed. In those dislocations in utero where development is imperfect, it depended on the stage of development at which the parts had arrived when dislocation took place as to whether there was hope of normal restoration.

1.4.1.11 S ummary of Theories of Causation in the Nineteenth Century (Reeves, 1885) Reeves summarized well the theories of causation of congenital dislocation or malposition of the hip that had evolved over several decades [72]. In reviewing the theories and relating them to his extensive clinical practice, he reached the conclusion that “different cases have different etiology.” Reeves noted that many of the deformity descriptions considered to be the primary cause of the disorder were in fact secondary deformations. Dupuytren and others observed that the hip capsule was too large, which allowed for displacement, a view commonly held today as being a primary etiologic component [59]. Sedillot also regarded the looseness of the hip joint ligaments to be a primary causative factor [60]. Abnormal position of the fetus in utero was also felt by many to lead to an abnormality of development which itself produced the hip dislocation deformity. Among those of this opinion were Dupuytren, Cruveilhier, and Roser. It remains well recognized today that the breech position in particular has a relatively high incidence of associated hip dysplasia although that position is still associated with the minority of cases. A direct mechanical force was considered by many to be causative, for example, the view of Brodhurst in which the trauma of delivery led to the physical displacement of the hip [28]. Reeves and others felt there was insufficient evidence

1 Developmental Dysplasia of the Hip

for that. Neuromuscular abnormalities were felt to be causative in many instances. Guérin, Carnochan [63], and others related the disorder to muscular retraction associated with pathological conditions of the nervous system. While many recognized associations with neuromuscular disorders, it was widely felt that muscular tightness was secondary as is the usual belief today. Many including Verneuil [66–68] and Reclus [69] implicated a muscle paralysis with atrophy of the hip muscles causing most cases. Many physicians at that time recognized infantile paralysis as being associated with hip dysplasia that occurred gradually over several years following birth such that congenital paralytic displacement was rarely seen. Another group of physicians attributed the malformation to primary developmental abnormalities of the hip region some placing these in the acetabulum, some in the proximal femur, some in the capsule, and some in the entire hip complex. Reeves also felt this was a fairly common source of the dysplasia. Certainly these descriptions can be recognized as essentially descriptive of what we refer to today as teratologic hip dysplasia. The increased incidence of hip dysplasia in females was recognized very early and has been a consistent observation ever since.

1.4.1.12 Sainton, 1893 Sainton wrote a two-part study of the anatomy of the childhood hip and the pathogenesis of congenital dislocation of the femur [73, 74]. (a) Anatomy of the Childhood Hip. Anatomic studies were performed on more than 30 prenatal and postnatal hip joints with the largest number in children between birth and 1 year of age [73]. The study involved assessments of the external form of the hip joint, measurements of the relative dimensions of the femoral head and acetabulum, and examination of the interior of the bone on decalcified sections to assess the development of ossification. At 2.5 months the embryonic head of the femur had formed its regular shape and was contained by the acetabular concavity in the iliac cartilage, but there was as yet no indication marking the femoral neck or differentiating the head from the shaft. By 3.5 months the shape of the articular surface resembled that of the postnatal period, but the neck was still quite short. Around 3.5 months the greater trochanter began to form. During this time period of early and initial hip joint development, the presumptive joint area separating the primitive head from the acetabulum was filled with tissue rather than being an empty space. The cellular substance in the intermediate zone later became subject to a process of resorption that passed from the central parts of the joint toward the periphery. During intrauterine life the hip cavity was sufficiently deep to be able to contain the convex head of the femur. The peripheral fibrocartilage

1.4 Etiology and Pathoanatomy of Developmental Dysplasia of the Hip

labral rim of the acetabulum was also developing. Important observations in view of subsequent hip disorders were that prior to birth the acetabulum was deep enough and the neck almost completely absent throughout the embryonic period such that the hip articulation favored retention of the anatomic position. Significant changes had occurred, however, by the time of birth and in the immediate years thereafter. The normal acetabulum was so constructed as to hold the femur in place. Three elements of difference were noted between the proximal femur of the newborn and the adult: (i) the head was relatively much larger in the infant than in the adult; (ii) the neck was longer in the adult than in the child; and (iii) the greater trochanter was more prominent in the adult than in the child. (b) Femoral-Acetabular Articulation. Studies in the infant indicated that the femoral head was relatively larger than in the adult during the first year of life, whereas the acetabulum was relatively shallow. Three epiphyseal regions formed in the upper part of the femur, these being the secondary ossification center of the femoral head, that of the greater trochanter, and that of the lesser trochanter. It was already known that the femoral head epiphysis was entirely intrasynovial in the developing hip. During the first year of life, a single uniform growth plate underlies the head-neck region and the greater trochanter. By 1 year, however, an angle is formed between the growth plate underlying the greater trochanter and that underlying the head-neck region. By 3 years of age, the neck is almost completely ossified, and the head is quite large. The secondary center of the greater trochanter appears around 3 years of age. (c) Acetabulum. The relative amplitude of the acetabulum in the infant is not nearly as great as that in the adult and thus is slightly less able structurally to maintain the head. The triradiate cartilages are well developed. The depth of the cartilage was measured in the studies, and the part of the head completely contained in the acetabulum was defined by cutting the head/neck-acetabular junction. This enabled measurements of the diameter of the segment of the head within the acetabulum and the portion augmented by the cartilage rim. The relationship between the head and the acetabulum was thus established. In the very young subjects aged 4–6 weeks, the acetabulum was not yet very deep. At 36 days a similar relationship persisted as seen shortly after birth, and the acetabulum was not able to receive the major portion of the head. The femoral head-acetabular relationship at the time of birth presented a cavity that predisposed to some instability. Comments were made on the development of the triradiate cartilage. Three cartilaginous branches separate the points of ossification forming the ilium, the ischium, and the pubis. The cartilaginous regions are

23

almost as wide as they are long. The capsule itself is well formed from birth and offers considerable stability. In summary, (1) the anatomic neck of the femur is very short in the infant, (2) the diameter of the neck is relatively large in relation to its appearance in the adult, (3) the diameter of the head is relatively larger in the infant than in the adult, (4) the acetabulum is less deep in the infant than in the adult, and (5) the head is particularly contained by the posterior part of the cavity. The anatomy was such that luxation of the femoral head, while difficult in the adult, was relatively easy to produce in the infant owing to the structure of the joint. ( d) Pathogenesis of Congenital Luxation of the Femur. Sainton then summarized studies of the previous several decades and provided a pathoanatomic assessment of three cases of dislocation, two of which were in newborns with teratogenic hips and one in a girl 12 years of age with a long-standing dislocation [74]. Traumatic Theory of Congenital Hip Dislocation Two types of trauma were postulated to cause hip subluxation, these being intrauterine trauma and obstetrical trauma at birth. By the late nineteenth century, few maintained that intrauterine trauma was a cause of hip dislocation, although until a few decades previously, it was a widely held view. Sainton himself placed no faith in that theory because it failed to conform to the known facts. The concept of obstetrical trauma, however, was more commonly held relating the occurrence of dislocation in association with traction on the lower extremities at the time of birth in particular with breech presentations. The occurrence of hip dislocation was felt to be much greater however than either the occurrence of breech presentation or of excessively traumatic deliveries. Even at this time, it was becoming apparent to many that traumatic deliveries were generally associated either with epiphyseal separations or actual limb fracture rather than true joint dislocations. Inflammatory/Pathologic Theories Two commonly held theories of causation involved hydrarthrosis and sepsis. The dislocation was felt to be the secondary to either intrauterine or postnatal hip joint inflammation. This theory evolved since there were undoubtedly many cases of septic arthritis of the hip that did dislocate. Malgaigne thought most luxations, which he felt were rather rare, were secondary to joint fluid accumulation or joint infection. These arguments, however, were soon combated in relation to the clinical situation. Sainton objected to the idea of dislocation associated with intra-articular fluid because such fluid had never been seen in the pathoanatomic specimens. In addition, cases of intrauterine sepsis that were known to occur were never associated with congenital dislocation.

24

Myogenic and Neurogenic Causes of Dislocation As early as the first half of the nineteenth century, abnormalities of the nervous system were felt to be associated with congenital hip dislocation. Delpech of Montpellier in particular stressed the importance of muscle development in relation to bone formation and developed the theory that bony, ligamentous, and muscular malformations were indicative of pathologic anatomy and thus also an indication of embryonic maldevelopment. One specific theory of abnormality was the theory of muscular contracture proposed by Guérin (1841) who felt that virtually all orthopedic deformities were caused by contractures. He placed such abnormalities as scoliosis, clubfeet, torticollis, and dislocation of the hip into a general law of etiology of orthopedic deformities that for him was almost universal. He was one of the early practitioners of tenotomy for the correction of such deformities. He also noted pathoanatomic findings that allowed one to establish a frequent relationship between certain malformations of the central nervous system and those of the joints with congenital dislocation of the hip being particularly prominent. Sainton felt that this theory depended too much on the assessment of patients with teratologic abnormalities since most of the patients with the congenital hip disorder were otherwise normal. Any muscle tightness was felt by Sainton to be secondary in the large majority cases rather than being the primary cause of the disorder. The other neuromuscular theory was of muscular paralysis as presented in 1866 by Verneuil [66–68]. The theory of muscular or infantile paralysis as the cause of dislocation of the hip in the newborn was hotly debated at the time. Verneuil felt that previous practitioners had studied only old cases of dislocation but that in reality the dislocation was present from the intrauterine period and was caused by a partial paralysis of muscles in the peritrochanteric region. The argument was not well supported. Bilhaut, commenting in 1896, felt that paralysis of the gluteal muscles was not found on studies [75]. Although it was thinner than normal, this was due to “absence of several of its fasciae” with those present demonstrating “absolutely normal development” both by microscopic examination and the electrophysiologic demonstration of “normal contractability” with no reaction of degeneration found. Some components of the gluteus muscles were stretched or lengthened by the displaced femoral head but were otherwise normal and functioned well after successful relocation of the femoral head. Sainton indicated that there were clear examples where infantile paralysis led to hip dislocation but these occurred well after birth and did not represent the large majority of early hip displacements, a view still accurate today. Theory of Primary Developmental Malformations Sainton supported this theory as underlying the primary cause of congenital hip dislocation based on his own pathoanatomic studies and those of others in particular Grawitz [76]. He felt

1 Developmental Dysplasia of the Hip

that if during the course of operative management of such dislocated hips (which was becoming increasingly common at that time) one did not find abnormalities of the hip or found only alterations which were insignificant, then it would be necessary to search for other causes of the congenital dislocation. He stated, however, that in the disorder “articular abnormalities were to the contrary quite pronounced and capable in themselves of leading to the displacement of the femur onto the iliac bone.” Pathoanatomic abnormalities had been discussed and were reviewed. The alterations of form varied with age. The acetabulum was narrower and less deep than normal; the head of the femur was larger; in other cases it was smaller and almost conical in shape. The round ligament was lengthened when it existed but was often absent; the capsule was elongated, deformed, quite large, and capable of receiving the displaced head; the neck was often shorter than normal; and its direction was changed to increased anteversion. In a word each of the regions of the hip was to some extent modified, and since these alterations were primary, it seemed unnecessary to search for other causes of displacement; it was necessary simply to indicate the embryologic causes of these different malformations. Objections to this theory could still be raised, however, and the questions commonly asked still concerned whether these changes were primary or secondary and whether they preceded or followed the dislocation. If they were only an epiphenomenon, they would lose a great part of their interest, and one could really not attribute to them a primary pathogenic role. If they were indeed primary, then the question to be asked next would be their cause and whether they were capable alone of producing the dislocations or at least rendering them imminent. One problem that existed was that most of the pathoanatomic studies had been done on subjects a few years old and had rarely been done on the newborn. Many authors including Sedillot had clearly noted that the pathologic anatomy was composed of two types of findings, those that were primary and truly congenital and others which were secondary and occurred later. It was felt that in those cases assessed early, the primary lesions found did not appear particularly extensive either in the acetabulum or in the femoral head such that most of the changes described were indeed secondary. There was information available even at that time that in the newborn hip with dislocation, the acetabular cavity itself looked good, and the femoral head had no modifications. The round ligament was almost always intact and was never lacking. (e) Pathoanatomy. Detailed analyses of dissections from two newborn infants with hip dislocation were given. The first case was a fetus that was born with many congenital malformations and died within an hour of birth. There was a flexion and adduction deformity of one hip along with bilateral clubfeet, cystic kidneys, and an

1.4 Etiology and Pathoanatomy of Developmental Dysplasia of the Hip

appearance consistent with a decrease of amniotic fluids such that the fetus had been compressed during intrauterine life. One hip was normal and the other abnormal. The round ligament was intact on both sides, but on the dislocated side it was longer. The acetabulum was shallower on the dislocated side. The acetabular rim did not exist especially at the superior region on the involved side, and the head of the femur that was much smaller than that on the normal side no longer had a hemispherical shape. It was flattened on one side and rested its deformed and flattened segment on the superior rim of the acetabular cavity. The entire proximal region of the thigh on the subluxed side was atrophied. Measurements of the acetabulum were smaller on the involved side than on the normal side. The Y (triradiate) cartilage did not show any modification of its normal shape or size. The second observation was made in a female fetus born spontaneously following an apparently normal pregnancy who died several hours after birth. One lower extremity was shorter than the other and was held in abduction and external rotation. There was also a bilateral clubfoot deformity. The opposite hip was normal. The acetabulum was scarcely seen and the femoral head was markedly atrophied. The round ligament was quite long, and a pseudo-capsule had formed on the outer wing of the iliac fossa into which the capsule and synovium were attached. The neck was markedly shorter on the involved side, and the head in relation to the shaft was implanted at essentially a right angle (anteversion) that was also indicative of shortness of the neck. Sainton felt that the acetabular cavity did not even exist on the involved side although where the head rested against the fossa, a small new cavity was formed within which an articular capsule had also formed. The Y cartilage of the base of the acetabulum was normal. These studies were presented as proof that at the moment of birth, very pronounced bony deformation was present both of the proximal femur and of the acetabulum. It was evident that these deformities would be increased later when the individual began to walk but that would only represent an exaggeration of modifications of form that already existed at the time of birth. The final dissection was performed on a 12-year-old girl who had signs of bilateral congenital dislocation of the hips and who died during the process of therapeutic reductions. The femoral head was found displaced onto the external iliac fossa. The capsule was considerably thickened by associated fibers except at its superior area where it rested against the head of the femur. The insertions were present and normal. The position of the head was maintained in flexion and adduction in relation to the pelvis. In this particular case, there was total absence of the round ligament, and the area of the femoral head where the ligament normally inserted was associated with a round depression. The head itself was

25

somewhat irregular in shape being ovoid and flattened at the top. The articular cavity was divided into three portions involving the new joint superiorly which contained the femoral head, a second part which had developed into a flattened region referred to as the pseudo-acetabulum, and a third part which corresponded to the original acetabulum and which was triangular in shape and the smallest of the three regions. The Y cartilage was normally conserved, and Sainton felt that the theory that attributed congenital dislocation to premature fusion of the Y cartilage was false.

1.4.1.13 Kirmisson, 1894 Kirmisson wrote extensively on the etiology and pathogenesis of congenital hip dislocation [77]. The vast majority of disorders were noted to occur in females, and unilateral dislocations were more frequent than bilateral. He supported the existence of congenital dislocation of the hip evident at the time of birth. Two dissections were reported. In one of the teratologic cases, the round ligament was present, but it was much longer than that on the opposite side, and the acetabular cavity was much more shallow than normal. The rim of the acetabulum had no prominence and did not exist especially at the posterior part. The head of the femur was smaller than that on the opposite side and did not have the normal hemispheric shape being flattened on one side. The femoral head was not positioned in the acetabulum that was far too small to receive it; it rested instead with its deformed side on the superior part of the socket. In a second case, the round ligament was also present but longer than normal. The capsule was inserted at the base of the fossa in a region where there was neither a border nor a prominence of any acetabulum. Kirmisson commented that his study of both specimens left no doubt about the existence of the displacement. In the large majority of cases, however, hip displacement was not recognized by physicians or family either at the moment of birth or in the first few weeks of life. The luxation was perhaps about to appear but did not in effect yet exist. In reviewing eight cases under his care, Kirmisson noted that in only two was the deformity noticed at the time of birth; in the large majority, the diagnosis was made only when the children began to walk. It was evident that the deformity, associated with the pathoanatomic abnormalities, existed at the joint from the moment of birth but was completed and magnified under the influence of walking. Paralysis or contracture of muscles was rejected as a general theory for causation of congenital dislocation and the contractures appeared as secondary phenomena. The contractures led to hip flexion and adduction deformities and tightness of the fascia lata. In the young infant with hip dislocation, the hip was very flexible, and one could put it in any number of positions. He negated the hypothesis of muscular paralysis although he based this opinion on operative interventions done to perform open reduction of the hip in which it was noted that the

26

muscles appeared clinically normal. The disorder represented a primitive malformation of the joint. Kirmisson reported that Lorenz had noted absence of the round ligament in 40 of 57 open reduction surgical cases, and above the age of 5 years, it was rarely seen [77]. In some cases it was present, and in others it was absent such that one could not propose a theory of hip dislocation on that structure alone and changes in it were also felt to be secondary. Kirmisson discussed the possibility that the dislocations were due to an arrest of development solely of the acetabular cavity as had been suggested by others. Examination of the two cases demonstrated abnormalities not only of the acetabulum but also of the proximal femur; “each of the elements constituting the hip joint articulation participated in the deformity.” The major question was the origin of the malformation, but he was objective enough to conclude: “we are completely ignorant of the cause of this malformation.”

1.4.1.14 D eveloping Awareness of Teratologic Congenital Hip Dislocation: Examples Assessed in the Fetus and Newborn (a) LePage and Grosse. LePage and Grosse described the congenital dislocation of one hip in a child who died at 14 days of age [78]. The infant was born at term and had shortening of one lower extremity and multiple congenital anomalies including a cleft palate, facial asymmetry, ectopic testicle, and a large hernia. Both thighs were hyperflexed and rigidly held against the abdominal wall with the legs flexed on the thighs. The right femoral head was clearly dislocated and the limb on that side shorter than the opposite side. At autopsy the pelvis was asymmetric with the right side smaller and less well-developed than the left. The ischium and sacrum on the involved side were also less prominent. There were no anomalies or irregularities of the musculature about the displaced right hip. The articular capsule on the normal side inserted circumferentially about the acetabular cavity. On the involved side, it was greatly stretched out being large at either extremity opposite the acetabular cavity and the femoral head but narrow in its middle portions. The inferior part of the capsule was stretched out over the empty acetabular cavity. The superior part that was very thick covered completely and enveloped almost in its circumference the dislocated femoral head above just under the anterior superior iliac spine. The normal acetabular cavity on the opposite side was deep, regular in shape, and with a well-formed cartilaginous rim. It completely contained the femoral head and measured 11 mms in its anteroposterior diameter and 13 mms in vertical diameter. The round ligament (ligamentum teres) was 6 mms long and did not permit any displacement of the head from the acetabular cavity. On the right side, the

1 Developmental Dysplasia of the Hip

acetabular cavity was atrophied, not very deep, without a bony rim, and without a cartilaginous rim. It measured only 6 mms in anteroposterior diameter and 8 mms in vertical diameter. At its posterosuperior part, the bony rim that in the normal forms a support against which the femoral head rests was flattened by the dislocated head. The displacement of the head outside the acetabular cavity was also made possible by the abnormal length of the ligamentum teres that measured 10 mms and inserted at the bottom of the acetabular cavity. There was no premature ossification of the triradiate cartilage. The normal femoral head was hemispheric in shape with a diameter of 13 mms. The femoral neck was well formed. The distance between the base of the greater trochanter and the top of the head was 17 mms. On the involved right side, the entire proximal portion of the femur was atrophied involving the head, neck, and greater trochanter. The femoral head was deformed, conical in shape, and smaller measuring only 9 mms in diameter. The femoral neck was scarcely seen to exist, and the difference from the base of the greater trochanter to the top of the head was only 11 mms. The entire femur on the right was thinner than that on the left although it was of the same length and also presented a curvature convex to the lateral side. An illustration of both proximal femurs and the pelvis with the capsule removed and the head displaced showed the atrophied femoral head on the involved side and the empty acetabular cavity on the involved side covered by the stretched capsule whose most superior part had been stretched and distended by the displaced femoral head resting up against the ilium. The ligamentum teres was thin and stretched, and there was a depression just above the acetabulum where the femoral head had been resting. LePage and Grosse felt that the dislocation of the hip resulted from an arrest of development that had been present for some time during the fetal state. The diagnosis was made in the newborn that was markedly different from most instances of congenital dislocation of the hip which tended to be made only when the infant began to walk and was noted to have an abnormal gait. They felt that at the time of birth, there was a considerable difference between the case they were describing, that we would now refer to as a teratologic dislocation, and the more common displacement since the latter was rarely diagnosed in the newborn period. They felt the common congenital dislocation of the hip was not truly dislocated at birth. At birth there was not a displacement of the femoral head but rather an anomaly of articulation that consisted in the fact that the femoral head was not fitted deeply into the cavity but rather placed opposite an immature model of the cavity. It was only under the influence of walking that the position of the head changed and rose to position itself adjacent to the iliac fossa. The displacement was based on the early articular malformation but really only occurred

1.4 Etiology and Pathoanatomy of Developmental Dysplasia of the Hip

and magnified itself with gait. They indicated that when the displacement was recognized at birth, it was almost always due to a difference in the position, length, or size between the two lower extremities that allowed the diagnosis to be made. In the case they were describing, the displaced hip was diagnosed based on the shortening of the involved side and the multiple malformations present which led to a careful examination. Since congenital displacement was rarely diagnosed at birth, it was even rarer to examine the skeleton of a newborn who had this malformation. The case of Barr and Lamotte was referred to in which a similar congenital dislocation was associated with an arrest of development of the acetabular cavity, the iliac bone, the femoral head, and the entire superior part of the femur. A stoppage of development was not necessarily required for a dislocation; it was sufficient if there was a disproportion in the dimensions between the elements comprising the hip. An acetabular cavity that had slightly atrophied or was smaller in size than expected could then be incapable of receiving and maintaining the femoral head for its normal development. Many observers felt that the smallness of the acetabular cavity was due to premature fusion of the triradiate cartilage, but the studies of LePage and Gross, Grawitz, Kirmisson, Broca, and Lorenz did not show this. In addition, Sainton and Delanglade always found preservation of the triradiate cartilage such that there was no evidence for its abnormality. They felt that it would be very rare in the presence of an open triradiate cartilage that there would be significant atrophy of the acetabulum to allow for inappropriate fit with the adjacent femoral head. The authors felt that an arrest of development was the cause of the dislocation. Trauma had previously been listed as a cause of the displaced hip, but there was no trauma in their case based on the details of delivery that occurred in a very small infant over the course of only 10 min to a mother who had four previous pregnancies. The capsule was intact at autopsy without evidence of tearing or bleeding. They also referred to multiple experiments in which individuals attempted experimentally to produce dislocation of the hip in autopsy specimens by manipulation and managed only to produce either fractures of the proximal femur or proximal femoral growth plate fracture-separations. (b) Cautru. Cautru described a case of congenital dislocation of the hip in a patient with multiple congenital malformations who died several hours after birth [79]. There was shortening of the left lower extremity, and the hip was held in abduction and external rotation. There were bilateral clubfeet and a complete anterior dislocation of each radius at the elbow. The right hip was normal, but the left was clearly subluxed with a small atrophied or underdeveloped femoral head and a poorly developed acetabulum.

27

(c) Kirmisson. Kirmisson reported on a dislocation of the hip in a stillborn fetus [80]. There were asymmetries of the pelvic region including the sacrum and ilium and considerable atrophy on the affected side. There was a posterior subluxation of the femoral head, and the acetabular cavity on the affected side pointed almost completely forward and also was markedly atrophied. The femoral head rather than being fully contained in the acetabulum was straddling the posterior ridge of the acetabulum. The femoral head had a vertical posterior depression where it rested against the acetabular ridge. At that area the acetabular rim was flattened. The anterior and middle regions of the acetabular cavity were empty and did not relate to the femoral head. The capsular attachments were such that the femoral head remained within the capsule but the capsule was notably elongated. The involved femoral head was smaller in each of its dimensions. Vertically it measured 1.3 cm compared to 1.6 on the opposite normal side; the transverse diameter was 1.4 cm in comparison with the normal 1.7; and the anteroposterior diameter was 1.3 cm compared to the normal 1.6. The ligamentum teres was only slightly elongated, and the surrounding musculature was normal. Each of the elements of the hip region was affected; the lesions did not simply involve the femoral head but rather all structures constituting the joint. (d) Potocki. Potocki reported a detailed analysis of a congenital dislocation of the hip in 1905 [81] from a child stillborn at approximately 7.5 months. The thigh was shortened on the involved side, and there was an associated clubfoot. The head was larger than normal. There was a flexion and adduction contracture of the left hip with limited movements into extension and abduction. The involved thigh was flexed and rotated internally. Based on the appearance of the limb, a diagnosis of congenital dislocation of the hip was made and confirmed radiographically. The musculature of the hip, thigh, and leg on the involved side was less bulky than the normal. The muscular insertions were normal, however, except for the pyramidal muscle that was attached to the superior aspect of the hip joint capsule. There were some structural abnormalities of the sciatic nerve. There was an arrest of development quite pronounced of the pelvis on the affected side. A large part of the external iliac fossa on the affected side was filled with the articular capsule of the hip. The capsule was thin in its anterior and superior part, but the ischio- and pubofemoral ligaments were very thick. Once the capsule was incised, there was a spacious articular cavity for the femoral head. The cavity was divided by the rim of the acetabulum into two parts. The superior part had for its edge the insertion of articular capsule opposite the external face of the ilium and below the acetabular rim. At this level,

28

1 Developmental Dysplasia of the Hip

however, the iliac cartilage was not covered by articular cartilage. The true acetabular cavity was empty and poorly developed. The rim of the acetabulum was missing except in its most inferior portion. The depths of the acetabulum were almost completely occupied by the ligamentum teres. The triradiate cartilage persisted and was of normal large size. The empty acetabular cavity was more elongated along the longitudinal dimension; its transverse diameter was 8 mms, while its vertical was only 4 mms. On the normal side, the transverse diameter of the articular cavity was 11 mms and its height 13 mms. The femoral head was not hemispherical in shape but rather flattened, in particular, that part which was relating to the iliac bone. The head was smaller than normal in terms of height and diameter. The ligamentum teres was much longer and thinner than normal. On the normal side, the round ligament (ligamentum teres) was thick and short and did not permit any displacement of the femoral head from the acetabular cavity. The neck of the femur on the involved side was shortened. In summary, each of the bony elements entering into the hip joint was notably smaller and deformed, while the articular capsule and round ligament were markedly elongated, and the rim of the acetabulum was scarcely protruding (i.e., markedly flattened). The author denied trauma as the cause showing no difficulty during pregnancy or delivery that was vertex in nature. There were neither contractures nor paralysis of muscles. The most common cause was felt to be an arrest of development affecting each of the elements of the hip joint. The development had not been affected from the earliest stages, but was due rather to an illness in midpregnancy.

1.4.1.15 Clarke, 1896 Clarke presented a series of illustrations from a case of bilateral congenital dislocation of the hip which was stillborn at full term and in which bilateral hip dissections had been performed [82]. Both hips were fully flexed with the knees in the extended position. Abnormalities of development of the acetabulae are shown and compared with the appearance in the normal. Malposition of the femur in relation to the acetabulum and the tightened capsule was noted. The joint deformity was reduced, after opening of the capsule, by downward pressure on the femur and internal rotation. The misshapened proximal femur and pelvis were also illustrated. 1.4.1.16 Bilhaut, 1896 Bilhaut described the prevailing observations of the pathoanatomy of the congenital dislocation of the hip from dissections, many at the time of surgical open reductions. These descriptions are from complete dislocations, in patients a few years old. A major pathologic component of the disloca-

tion was the iliopsoas muscle/tendon that was firmly bound to the capsule, running across it inferiorly. The capsule was thus deformed into an ampullar shape, with the superior part, large and containing the head shaped as a flashlike or saccular dilation. The capsule was thickened both above and below the iliopsoas and also internally “where large fibrinous bundles stretch themselves when the femur is abducted.” The iliopsoas was “intimately united” with the capsule; the formation of pericapsular fibrinous thickening around the displaced femoral head is described forming a kind of secondary capsule. The head of the femur communicates with this fibrinous thickening “by means of the superior part of the capsule which during the ascension of the head comes between the iliac bone and the femur.” There was however no true new joint formed. The ligamentum teres was present with its normal insertions but was lengthened and thin. The capsule was enlarged. The femoral head was not round but had become flattened in front and back. Reduction was prevented by the tight iliopsoas muscle and the adductors, especially the superior thickened fascia of the adductor magnus, and by the tension of the inferior part of the capsule [75].

1.4.1.17 Keetley, 1900 Keetley also remarked on the controversy as to whether or not the disorder was truly congenital and recognized that although the preconditions might exist at the time of birth, the actual dislocation appeared to occur afterward [83]. His text had a detailed section on the etiology of congenital dislocation of the hip. Grawitz had studied seven patients with teratologic hip dislocations of whom five had bilateral and two unilateral dislocations. All had multiple congenital anomalies including spina bifida, clubfeet, clubhands, and scoliosis. The development of dislocation of the hip in association with infantile paralysis also occurred. Those individuals described were not in the idiopathic developmental dysplasia category. The Y-shaped cartilage was relatively undeveloped and the acetabulum disproportionately small in relation to the head of the femur. Among the reported causes of hip dislocation, which were hypothetical without experimental or pathoanatomic evidence, were position in utero (Dupuytren), deficiency of liquor amnii (Roser), intrauterine injuries secondary to trauma to the mother’s abdomen, and spasmodic fetal muscular action. The “position in utero” argument was not felt to be compelling since almost every fetus is placed with its thighs hyperflexed. Another hypothesis related trauma to the fetus during birth in particular with breech presentations. Keetley argued that although breech presentations have a higher incidence of hip dislocation, their absolute numbers were still small in relation to the number of cases of hip dislocation. Other hypotheses related the dislocation to congenital absence of an acetabular rim (observations made by Lockwood [84] and Grawitz), relaxation of the ligaments of the joint (Sedillot), and disease of

1.4 Etiology and Pathoanatomy of Developmental Dysplasia of the Hip

the fetal hip joint. Keetley hypothesized in agreement with Sedillot that “so called congenital dislocation of the hip may sometimes or even often be due to abnormal laxity of the ligamentous structure of the joint, the dislocation of necessity occurring long after birth.” In support of this latter theory, he commented that the ligamentum teres is generally intact but at the same time observations noted laxity of the capsule and external ligaments of the joint. He strongly supported that etiology in a child where dislocation was discovered a long time after birth and in which there had been no symptoms, no history of infantile paralysis or injury, and no other associated congenital deformity. Even at this relatively early date, all data pointed to the strong preponderance of females with nearly nine patients out of ten being female. The dislocation was unilateral only slightly more commonly than bilateral in a proportion of less than 3 to 2. Keetley summarized the then current information well and reached a conclusion which regrettably still exists today that “with regard to the majority of the cases classed together by surgeons under the name of congenital dislocation of the hip it is easy to suggest many theories of their origin and difficult if not impossible at present to prove one.” The pathologic changes found depended on the age of the patient and the variable causes of the deformity.

29

A pathoanatomic study from a full-term stillborn fetus demonstrated the hyperflexed position of the lower extremities. The entire innominate bone was deformed, and the acetabulum was oval and much smaller than in a normal specimen. Joint displacement was demonstrated prior to opening of the capsule; reduction was possible once the capsule had been freed and the femur pulled downward and rotated inward. The femoral head was displaced superior and posterior to the acetabulum, there was proximal femoral anteversion, and the ligamentum teres was elongated and flattened.

1.4.1.18 Le Damany, 1904 Le Damany wrote a series of articles on the pathogenesis of congenital dislocation of the hip which remain of value today in particular his discussion of normal hip development and diagrams illustrating the possible relationship of subtle proximal femur and acetabular abnormalities to the disorder [21, 22, 32–34, 85, 86]. He reviewed the evolutionary development of the hip in man and other species. The human changed hip position from flexion in utero to extension postnatally; this fact if combined with particular variations of acetabular and proximal femoral development could predispose to hip dislocation. During the final stages of intrauterine growth, the relatively large size of the fetus and in particular the (i) Newborn. The head and neck of the femur were smaller femur in relation to other species caused the hip to assume than normal and also altered in shape sometimes being the hyperflexed position and relatively increased pressures short or round and sometimes long and conical. The exerted on the hip region and the femur from uterine pressure acetabulum “was always small” even in proportion to against the knee. These two changes led to altered relationthe diminished size of the head of the femur. It was nar- ships of the proximal femur and acetabulum in comparison row and oval. Fat occupied its cavity, and the deficient to those present in the embryonic and early fetal stage. Le depressed posterior margin was encroached upon by the Damany noted that in the human fetus, the acetabulum was cartilage surface of the new acetabulum. The ligaments hemispheric during the first two-thirds of fetal life but that and the joint capsule were stretched but not torn; the during the last trimester, its depth in relation to its width ligamentum teres was unusually long, thin, and flat. The gradually diminished. He felt that early on the depth was joint capsule enclosed both the old and new acetabulum. one-half that of the diameter but at birth it was only two- The pelvis was also somewhat misshapen. In the terato- fifths. In adult life it further increased to three-fifths. There logic cases described, the cartilage was intact, the joint was therefore a relative lack of depth of the acetabulum capsule and the round ligament were lengthened, and around the time of birth, and this had negative implications luxation could be reduced although it quickly re- in relation to hip stability. The proximal femur also u nderwent displaced. There was defective development of the rotational changes increasingly during the latter stages of Y-shaped cartilage of the acetabulum. intrauterine growth. It went from neutral anteversion (0°) at (ii) Older Children. Once the child began to walk, the 4 months to anteversion as high as 35–40° at birth followed changes in the hip region became more marked. The by diminution to approximately 10° by the time of skeletal acetabulum was narrower, smaller, and shallower, and it maturation (Fig. 1.2a). To Le Damany these changes were began to assume a three-cornered shape, filled with fat, examples of extrinsic pressure affecting growth of the relaand no longer able to receive the femoral head even tively pliable cartilage model of the bones. When the changes when reduction was performed. The capsule and liga- in both the acetabulum and the proximal femur were slightly mentum teres were longer. A regular false joint formed exaggerated, the possibility of dislocation was greatly on the dorsum of the ilia. In many patients operated after increased. The proximal femoral and acetabular abnorseveral years, the original acetabulum could scarcely be malities or more accurately angular-rotational excesses noted. Keetley indicated that for “practical purposes the were therefore maximal at term. He also noted that there acetabulum gradually becomes obliterated.” was no anterior obliquity or tilt of the acetabulum in any

30

1 Developmental Dysplasia of the Hip

a

i

A

B

C

ii

Fig. 1.2 (a (i, ii)) Le Damany demonstrated normal features of developmental hip anatomy and his theory as to how subtle abnormalities of femoral and acetabular development in particular in combination predispose to dislocation when the child assumes the upright posture with hip extension. (a(i)) This illustration shows changing proximal femoral anteversion during development. “A” at left illustrates neutral version in the early fetal stages. In each of “A,” “B,” and “C,” the femoral head is at the middle part of the image, the trochanter at left, and the distal femoral condyle at right. The axis of the proximal femoral head and mid-neck is illustrated by the line from the trochanter through the midneck and head. There is no version in “A,” anteversion as great as 40° during the fetal period in “B,” and diminution of the anteversion to approximately 10° at skeletal maturation in “C”. (Reprinted from Le Damany, Ref. [32, 34]). (a(ii)) The combination of changes predisposing to hip subluxation and dislocation as theorized by Le Damany is shown. At left, normal degrees of anterior acetabular opening and of proximal femoral anteversion allow the head to be seated within the acetabulum. At right there is increased anterior opening of the acetabulum (increased acetabular obliquity) in association with increased proximal femoral anteversion. These two features can lead to relative instability in particular when the child assumes the upright posture. (Reprinted from Le Damany (Ref. [33]) and Z Orthop Chir 1908; 21: 129–169). (b) Developmental changes in the acetabulum as it grew when the hip was dislocated were appreciated in early pathoanatomic studies of CDH. A cross section of the acetabulum in the normal childhood hip is illustrated at right and one from a dislocated hip at left. The

acetabulum is shorter, shallower, and with rounded margins on the involved side. (Reprinted from Le Damany (Z Orthop Chir 1908; 21: 129–169)). (c) Illustration clearly delineating the pathoanatomy of hip dislocation. The capsule (g) is enlarged, elongated, and displaced proximally onto the side wall of the ilium well above the acetabulum although it provides superior support for the displaced femoral head. The acetabulum is shallow and oblique and filled with fibro-fatty material. The capsule and transverse acetabular ligament (e) are pulled across the opening of the acetabulum by the displacement. (Reprinted from Deutschlander 1910, Ref. [90]). (d) The hip on the right is normal and that on the left abnormal. On the right the femoral head has been displaced to show the normal acetabulum (a) and ligamentum teres (b). On the left, the hip is dysplastic with a small misshapened acetabulum (f), false acetabulum superior to the original acetabulum and indented into the lateral iliac wall (d), elongated ligamentum teres (c), and misshapened head-neck (a). (Reprinted from Lepage and Gross 1901, Ref. [78]). (e) The normal side acetabulum is shown at left and the abnormal acetabulum of the dislocated hip at right. The acetabulum on the abnormal side is smaller than that on the normal side, “triangular” in shape, and shows an enlarged capsule (pointer) with flattening of the acetabular rim and a more spacious joint. (Reprinted from Ludloff 1911, Ref. [89]). (f) A dislocated hip is seen with elongated ligamentum teres, stretched capsule, small and shallow original acetabulum, and more spacious pseudo-acetabulum above and laterally. (Reprinted from Ludloff 1911, Ref. [89])

1.4 Etiology and Pathoanatomy of Developmental Dysplasia of the Hip

b

31

c

g

a

h b c e

d

f

d

d

f a

a b

c

Fig. 1.2 (continued)

32

1 Developmental Dysplasia of the Hip

e

f

Fig. 1.2 (continued)

1.4 Etiology and Pathoanatomy of Developmental Dysplasia of the Hip

of the species other than man. Normal anterior obliquity of the acetabulum did not vary with developmental age although it had a relatively wide range of values in the normal between 15° and 30°. He then established a quantitative index which when applied to cadaver studies of available specimens with CDH led to an indication as to what amount of proximal femoral-acetabular deformity led to dislocation. He simply added the degree of femoral anteversion to the degree of anterior acetabular obliquity. Femoral anteversion in the normal varied between 30° and 50°, while acetabular obliquity varied between 15° and 30°. The most extreme angulations therefore would be 50° (femoral) plus 30° (acetabulum) leading to an index of 80° which would favor dislocation, while the safest values would be 30° (femoral) plus 15° (acetabular) leading to an index of 45°. In the adult the normal index would be in the range of 32° involving the proximal femur anteversion of 12° and the acetabular anteversion of 20°. It was those infants with an index greater than 60° that were subject to dislocation (Fig. 1.2b). If all the constituent elements of the hip were within the normal range, the subtle deformities were capable of correction, often spontaneously, but if the additive features led to a sufficient degree of deformity, then a dislocation would result. These relatively subtle developmental variations were an increased femoral anteversion accompanied by an increased anterior opening obliquity of the acetabulum. These were reciprocal occurrences such that each finding could be in the normal range but that when they were both somewhat more marked, the situation was set for a possible dislocation. He further stressed that the definitive dislocation itself would not occur until several months after birth at which time hip function in the extended and then upright position began to have its most negative effect. Positioning of the human hip into extension in the early days and weeks after birth further worsened the stability. He made a series of comparative hip studies throughout other species pointing out that dislocation was virtually unheard of in any other species and that the reason for this was primarily the maintenance of the flexed position of the hip in all but the human. In his outline of normal hip development, embryo, neonatal, and adult femur specimens were assessed. Proximal femoral anteversion was noted to begin only during the second half of pregnancy. There was essentially no anteversion of the proximal femur during the embryonic phase, and it was measured at 0°. By birth it had increased to an average of 40° (ranging from 30° to 50°), whereas by the time of adulthood, it had decreased to a mean of 12°. In association with this was the fact that the acetabulum was relatively most shallow in relation to the position and size of the femoral head in the newborn period that also predisposed to subluxation and dislocation. There was also slight anterior obliquity with the acetabulum opening not only to the side but also anteriorly that worsened stability. The secondary changes in a dislo-

33

cated hip had been described previously, but he felt that the initial subtle pathoanatomic reasons had not been appreciated. Le Damany divided causes of congenital hip dislocation into two basic categories, teratologic and anthropologic, the latter referring to what we now term idiopathic DDH. He divided congenital hip dislocations into group A, which were relatively rare and included those of teratologic, traumatic, and pathologic origin, and group B, which were relatively frequent referred to as anthropologic luxations. The traumatic dislocations were not caused by direct extrinsic damage to the fetus but were occasionally produced by difficulties in birth although even then most hip displacements at birth were felt to be fracture-separations. Among the pathologic disorders were those involving neuromuscular or infectious disorders of the hip and the teratologic abnormalities referred to those associated with spina bifida or true abnormalities of development with many systems malformed. The term anthropologic dislocation was used to refer to the combination of abnormalities that he himself had described and thus otherwise appear as idiopathic congenital or developmental dislocation of the hips since the patients were otherwise normal. He described these as being congenital because the predisposing variations developed during the latter stages of intrauterine life although the actual dislocation occurred after birth. The abnormalities involved extensive anteversion of the proximal femur and increased anterior obliquity of the acetabulum beyond a stable range once the postnatal posture of hip extension was assumed. Both of these were due to the same mechanism of increased force on the hip region owing to the relatively large size of the human fetus, the length of the femur, and the external uterine pressure on the relatively soft developing bones. Postnatally the position of the human hip into extension in association with the femoral and acetabular variations led to the dislocation. He stressed that in the anthropologic type, the patient was otherwise normal, while in the rarer intrauterine disorders, many patients had other abnormalities. The anthropologic types resulted from an exaggeration of the normal variations or imperfections of the human hip and were extremely frequent. The possibility of cure however was quite high since both femoral and acetabular tendencies were to correct to a normal range with development. The displacement is thus not truly congenital, but predisposition for the displacement shortly after birth occurred due to late intrauterine events. Le Damany was quite aware of the existence of the dislocatable hip at birth which he felt was quite frequent and which generally went on to a spontaneous recovery without treatment in a period of time varying from several days to several months. He even described the clinical maneuver that would sublux or dislocate the hip and the reverse maneuver which would allow it to relocate. When the hip was placed in a position of flexion and adduction, and a slight force was

34

applied to the knee along with a force from within outward on the thigh, the head would displace over the posterior border of the acetabulum. Replacement into the acetabulum would occur with a reverse flexion-abduction maneuver. This description predated those of Ortolani and Barlau [87]. Le Damany also pioneered the systematic clinical examination of large groups of newborn infants (1722) for hip instability reporting studies from Rennes, France, and Paris which also included the observation that several unstable hips at birth spontaneously stabilized over the next few days or weeks [85–87]. Le Damany also indicated that the principles of therapy were quite simple since the dislocation in the anthropologic type was only due to a slight deviation from the norm. He referred to the proposed treatment as being almost completely geometrical and mechanical and that those under 3 years of age should be readily curable. The first principle was simply to place the articular surfaces in exact coaptation such that the new acetabulum would be formed exactly in the site of the existing but slightly shallow one. The pressure of the head would lead to deepening of the cavity. “If we wish to reform an articulation it is not to immobilization that we ought to have recourse but to motion.” He went on to indicate: “function makes the organ.” Le Damany noted that at the time of birth, the deformations that provoked the congenital subluxation, torsion of the femur and anterior obliquity of the acetabulum, had not yet produced any deformation in either the head or the acetabulum. It is only with time that the secondary changes occurred as the hip remained subluxed or dislocated. These changes were well recognized and included atrophy of the empty acetabulum, further deformation of the head of the femur and of the neck, proximal femoral anteversion which did not correct from its neonatal level, enlargement of the capsule with or without the hour glass narrowing, lengthening and thickening of the round ligament (ligamentum teres) followed by its thinning and eventual rupture, lengthening of certain muscles and shortening of certain others, and shortening and atrophy of the involved limb. The anatomic abnormalities that he commented on therefore preceded and prepared for the dislocation such that they were necessary for its occurrence but were not the sole cause of it. The positional abnormalities were at their maximum in the newborn. There was little awareness of his anthropologic variant because newborn pathoanatomic assessments, which were extremely rare, did not note any abnormalities since the shape of the head was normal, the shape of the acetabulum was normal, and the dislocation had not yet occurred. Only if the detailed studies that he had described were performed would one notice the fact that the predisposing features for the dislocation were present. The dislocation was rarely present at birth but only occurred after birth owing to the extended position of the hip and the onset of weight bearing.

1 Developmental Dysplasia of the Hip

Although his work was mainly theoretical in one article, he did present pathoanatomic descriptions of four hips varying in age from 3 to 18 months in which the abnormality he described was defined. Three of these he felt would have corrected spontaneously, and one had gone on to subluxation.

1.4.1.19 Bennett, 1908 Bennett discussed congenital dislocation of the hip indicating that birth injury was not a contributing factor but that abnormal laxity of ligamentous structures of the joint was causative [88]. Bennett combined previous theories feeling that in a child who is otherwise normal, the dislocation was a “pure accident, dependent on the position in utero which makes it possible, and the laxity of tissues around the joint which makes it probable. For the accomplishment of the deformity, it needs but a slight movement in the required direction...... either in or outside the uterus.” He agreed with Dupuytren’s hypothesis in which the extremely flexed uterine position in a child with lax tissues served to allow the head to slip onto the posterior acetabular rim which as a result of the pressure exerted did not grow at the same rate as the remainder of the circumference so that at birth the head was lying on an atrophied posterior rim and it was just chance as to whether the early postnatal movements placed it in or out of the socket. Pathology The capsule is necessarily stretched with the head displaced from the acetabulum. With walking the stretching is worsened, and the capsule becomes thickened. As it stretches across the acetabulum, the capsule becomes adherent at the rim so that the acetabulum appears obliterated being covered by thick fibrous tissues reaching from rim to rim. The adductor muscles become secondarily shortened as is the iliopsoas. The glutei are not shortened initially, and the muscles that pass from the pelvis to the greater trochanter (obturators, gemelli, etc.) are lengthened. The hip ligaments adjust to the deformity and shorten along with the adductor muscles and hamstring muscles and thus become factors opposing reduction. Those attached to the lesser trochanter are particularly involved. Bony changes involve a gradual loss of the hemispherical character of the head of the femur that tends to become flattened and an alteration of shape and depth of the acetabulum that tends to become less circular and more triangular in shape with the tissues at the bottom filling it. Anteversion of the femoral head is seen and the neck-shaft angle is increased. The usual position of displacement of the femoral head is upward and backward.

1.4.1.20 Additional Pathoanatomic Studies Large numbers of pathoanatomic studies in the early decades of the twentieth century originated from Central Europe due to the high prevalence of the disorder in those countries and

1.4 Etiology and Pathoanatomy of Developmental Dysplasia of the Hip

the high level of medical investigation. The increasing use of open reduction allowed for better appreciation of the pathoanatomy. (a) Ludloff. Ludloff performed one of the early studies of acetabular abnormalities in association with congenital dislocation of the hip [89]. He made assessments of four individuals in relation to hip abnormalities prior to birth, shortly after birth, in adolescence, and in adulthood. Additional detailed reports from the literature were added. He concluded that the depth of the acetabular socket without exception was significantly decreased in all cases from the embryonic time onward. The width of the upper half of the acetabulum was always more seriously affected (diminished) than the lower half that led to a triangular shape of the acetabulum. The reduction in depth increased during development of the deformity along with flattening of the periphery of the acetabulum and inversion of the limbus. The inversion of the limbus was the main reason for the flattening of the rim of the acetabulum. When the limbus was inverted, it played a major role in causing the protrusion of the head from the acetabular socket. Several illustrations from his own work and the work of others showed the marked under development of the acetabulum in the dysplastic hip compared with the normal. The shallow and flattened acetabulum was illustrated along with the widened and enlarged capsule, the anteversion of the proximal femur, and the elongation of the ligamentum teres. The dysplastic acetabulum took on a triangular shape with its base toward the obturator foramen and its apex superior. Cross-sectional drawings of the acetabulum showed both its diminished width and depth and the loss of sphericity with the floor tending to a flattened position. (b) Deutschlander and Loeffler. Deutschlander clearly outlined the pathoanatomy of the complete dislocation in the process of describing the limitations of closed reduction [90]. Illustrations from his work show the primary difficulties with reduction owing to the shallow and dysplastic acetabulum, the presence of fibro-fatty tissue within it, and the hypertrophied capsule covering the inferior half of the entrance to the socket. He also clearly demonstrated the dangers of capsular interposition in association with closed reduction. Similar illustrations of the various tissue interpositions limiting the effectiveness of closed reduction were shown by Loeffler [91]. (c) Werndorf. Werndorf described the pathologic anatomy from a child who died having surgery on the contralateral hip [92]. Characteristic findings included a socket reduced and flattened, a triangular shape to the acetabulum, the upper acetabular roof completely absent, and the socket base so thickened that barely a fingertip could be inserted. The ligamentum teres was considerably

35

lengthened and broadened and completely filled the socket. The proximal end of the femur was retarded in growth with flattening of the head and considerable anteversion of the femoral neck. The flattening involved the medial and posterior parts of the head. The head was still contained within the capsule that was lengthened, tubular in shape, widened at its upper part, narrowed centrally at the isthmus, and considerably thickened. Most of the muscles passing from the pelvis to the femur were shortened in particular the adductor group. The muscles however passing from the pelvis to the greater trochanter tended to be lengthened. (d) Lance. The secondary changes in hip structures in untreated CDH were also assessed by Lance [93]. The femoral head was small and conical in shape being flattened on one side. The secondary ossification center was late to appear and was not central along the long axis of the neck. The neck tended to be short, thickened, and rotated into increased anteversion. This frequently was in the range of 40–60° at a time where it should have been diminishing to the 15–20° range. The lesser trochanter tended to be markedly hypertrophied owing to changes in weight bearing in relation to its attached muscles. The acetabular changes were similar to those described by Ludloff. The head always rested within the capsule that was enlarged allowing displacement to occur. In some cases the capsule remained quite spacious, while in others it took on an hourglass shape with a narrow central isthmus below the head but above the original acetabulum. At times, the capsule was interposed between the head and the side wall of the ilium, while at other times it always remained superior to the head. There were also examples of attachment of the capsule either to the external wall of the ilium and in many instances to the head and neck of the femur.

1.4.2 Later Clinical-Pathoanatomic Descriptions 1.4.2.1 Fairbank, 1930 Fairbank discussed the pathoanatomy of congenital dislocation of the hip in detail basing his report on 50 open surgical procedures for CDH, 46 dislocated hips from the Dupuytren museum in Paris, and an extensive literature review [94]. His study is particularly strong on the late pathoanatomic changes seen in the adult years. The ultimate structural changes in congenital hip dislocation that had gone untreated or in those diagnosed late are well described. Fairbank attributed the primary pathology even in the affected fetus to the “poor development of the upper margin of the acetabulum.”

36

(a) Acetabular Findings. Acetabular development is markedly abnormal when not accompanied by presence of the femoral head in its normal position. With complete dislocation, the acetabulum becomes triangular in shape with its base toward the obturator foramen and the apex pointing upward and backward. This triangulation is considered to be the result of continued growth of the anterosuperior and posterior boundaries of the socket unchecked by the pressure of the femoral head. The margins of the acetabulum are usually straight and sharp, and the cavity retains some depth, even though the floor is more or less flat. The cavity is filled by cartilage and fibro-fatty tissue. The bones of the adjacent pelvis are also abnormal, the obturator foramen is more triangular than normal, the pubic angle is increased, the ilium is shorter and broader than normal, and the anterior border is prolonged in the vertical direction. The anteroinferior iliac spine is twisted to conform to this outline. A false acetabulum forms on the outer border of the ilium. As a rule, it is larger than the femoral head that rests in or against it and the disproportion suggests considerable mobility of the femur in both the anteroposterior and vertical directions. In some cases, the edges are well developed, and occasionally a deep hemispherical cup with a polished eburnated floor is seen. The appearance of the false acetabulum varied greatly. In 38 hips, Fairbank noted that it was a shallow depression only with little or no margin in nearly one-half (17) and in 9 there was no sign of a false joint, while in 9 a well- marked socket with lipped margins and an eburnated floor was present. (b) Femoral Changes. The head is smaller than normal, even in the young child with dislocation, and the secondary ossification center is late in its appearance. The head becomes flattened by pressure against the ilium on its inner and posterior aspect. In adults, there is often erosion and pitting of the cartilage surface to actual complete disappearance. The shape of the head is also variable in several specimens but almost always small and imperfectly shaped. Anteversion is quite common in patients with congenital hip dislocation. Whitman noted the normal angle of anteversion at 35° at birth with gradual reduction to 10–15°, whereas if the hip was dislocated, this reduction did not take place [95]. In the study by Farrell et al., based on radiograms in 336 cases, in nearly one-half the angle was over 20°, and in these about one-half gave an angle of 20–50°, while in the remainder, the angle was over 50° [96]. (c) Capsule. As the head migrates upward into its dislocated position, it carries in front of it a dome of the capsule that blends with the periosteum above and behind the acetabulum. Where this fusion occurs in the floor of the false

1 Developmental Dysplasia of the Hip

acetabulum, the two are transformed into fibrocartilage. The capsule, although initially lax and thin, becomes thickened with time in particular in those regions where it has a weight-bearing function; it could be as much as onethird of an inch in thickness in a child of 13 years. With complete dislocation, the thickened capsule rides upward and becomes tightened and thus serves as a barrier against easy reduction. The joint cavity develops an hourglass shape. The isthmus between the true and false joints is accentuated by the altered position of the psoas tendon and by the thickened capsule. This well-developed isthmus occurs with time and is generally seen only after 3 years of age. ( d) Muscle. The adductor muscles are always shorter than normal. The iliopsoas tendon plays a major role with displacement. When the femur is displaced posteriorly and upward, the pelvis tilted, and the lordosis marked, this tendon takes a practically horizontal course upon leaving the pelvis. It is quite stressed and generally causes a deep groove in adult specimens below the anteroinferior spine. Most felt that the tendon had to be divided as part of any open reduction. The gluteal muscles tend to be unaltered in length since the trochanter is displaced outward as well as up. The horizontal muscles, the obturators, gemelli, and quadratus, are lengthened, and the direction of their fibers is altered. They no longer run horizontally but pass upward and backward to reach the trochanter. Support of the hip in the dislocated position is almost exclusively due to the soft tissues with the false acetabulum having relatively little support function. The capsule and the muscle share the work of support. Capsular strain is shared by the thickened bands, passing from pelvis to femur in front and below, and by the capsular sling arching over the neck. The muscles that assist the capsule and prevent further stretching of this sling are the iliopsoas in front and the obturator group behind with the gluteus minimus helping to some extent. The abductor muscles in the dislocated hip act under decided mechanical disadvantages. Fairbank concluded that the older the patient at the time of reduction, the greater the chance of an imperfect anatomical result. Manipulative reduction is best “at an early age” defined at that time (1930) as before the age of 3 years and better still before the age of 2 years. He was convinced that closed reduction would be appropriate in the younger patients as many of the capsular changes noted by early proponents of open reduction were relatively late secondary changes. Burghard defined details of open reduction; he divided the psoas and enlarged the isthmus. Fairbank felt, however, that it was rarely necessary before the age of 4 years. The antetorsion was frequently corrected by osteotomy, even at that time, but Fairbank felt that if the hips were reduced early,

1.4 Etiology and Pathoanatomy of Developmental Dysplasia of the Hip

before the fourth year, the antetorsion would correct with resumption of gait over time.

1.4.2.2 Leveuf, 1947 Leveuf made a clear distinction between primary congenital subluxation of the hip and primary congenital dislocation of the hip basing his opinion on multiple arthrographic studies as well as findings from operative interventions [97]. The primary subluxation presented some anatomical characteristics that were distinctly different from those of a dislocation and he considered that the two disorders therefore were distinct rather than representing a spectrum in which a subluxed hip which worsened became a dislocated hip. In a subluxation the limbus is forced upward and outward, whereas in a dislocation the limbus is displaced downward and inward toward the acetabulum. This is particularly well illustrated by a drawing of the disorders. Specific differences were noted in the structure of the acetabulum, the head of the femur, the capsule, and the neck of the femur in each disorder. In subluxation the cartilaginous roof and the limbus, which were forced by the head of the femur against the external iliac wall, and the iliopsoas muscle always appeared atrophied, whereas in a true dislocation, the cartilage roof and the limbus were forced toward the acetabulum which led to the limbus being hypertrophied and interposed between the head and the acetabulum during a closed reduction. The acetabulum in the dislocation “generally retains a satisfactory depth with the well developed group.” In subluxation the head of the femur was deformed very early being enlarged and widened transversely as well as flattened at its superomedial pole. In dislocation the head retained a regular contour for a long time. The joint capsule in a subluxation was enlarged but never interposed between the femoral head and the acetabulum, and the round ligamentum teres was “practically always absent.” In dislocation on the other hand, the capsule generally was interposed between the head and the acetabulum, and the round ligament was present in only about one-third of the cases. In dislocation there was a clear “interposition of soft parts” (the limbus, the round ligament, and the lower fold of the capsule) where they constituted an obstacle to reduction, whereas such interposition of soft parts never existed in a subluxation. The neck of the femur in subluxation often was associated with a valgus position as high as 150–155° (normal 130°). Anteversion was also quite common in subluxation. In dislocation there was no valgus of the neck and anteversion was rare. Leveuf reinforced his opinion that the two disorders were distinct entities indicating “in our opinion not one well established fact can prove that a subluxation, showing the characteristics which have been described, can become a luxation.” This radical distinction has not been widely accepted by the orthopedic community, but his description of the pathoanatomic findings in the variants for the most part appears accurate.

37

1.4.2.3 Badgley, 1949 Badgley reviewed two main theories of etiology that had been prominent for many years, one indicating the lesion to be the result of a primary germinal fault and the other that it was due to faulty development of environmental origin [98]. He felt that acetabular abnormality which was clearly a part of the congenital dislocation hip syndrome was a secondary developmental problem and not primary. Mechanical factors were clearly involved in some instances of congenital hip dislocation. He referred to a discussion in Tubby’s textbook of a report by Tridon that 121 cases of congenital dislocation of the knee (in which the knee is hyperextended) had been associated with 20 instances of congenital dislocation of the hip [99]. While this was indicative of a mechanical etiology, it did not substantiate the mechanical theory for dislocation of the hip except in this unusual circumstance. The hyperextended knee seemed to be analogous to the high incidence of CDH in breech malposition. He also felt that the mechanical concept of Le Damany seemed plausible but did not adequately account for all features. The high incidence of CDH in those with breech presentation is also considered to represent an example of mechanical effects on development. Badgley reviewed the embryologic principles of development concentrating on the importance of “perfect timing” for the development of the constituent parts. Early embryologic development is intrinsic to the specific part, but ultimately, when the gross skeletal model was being refined and perfected, the importance of extrinsic factors increased. Rotation of the limb buds is an important feature of embryonic development. The limb buds and extremities undergo rotational changes during development to the extent that they ultimately twist around their longitudinal axes and rotate through an angle of approximately 90°. The alteration of position of the limb buds starts prior to the separation of the components of the hip joint, and “this postural change of the limb bud prior to motion in the hip joint may be a definite factor in the production of the inclination of the neck of the femur.” Most postural change, however, occurs after commencement of the joint cavitation after 30 mms stage of fetal development. The femoral region must rotate internally approximately 90° at the hip joint as part of the normal developmental sequence. Adaptive changes in the acetabulum and the upper end of the femur are necessitated by the rotation phenomenon as well as the development of the oblique position of the acetabulum. The inclination of the acetabulum is important with 30–40° of forward inclination and 60° of downward inclination. Dega in a review of 100 fetal skeletons showed the angle of forward inclination of the acetabulum to be 29.5° and the downward inclination in relation to the transverse plane 62.8° [100]. These changes in acetabular development were also pointed out by Le Damany who commented on the tilting of the iliac bone of the sacrum in the human, obliquity of

38

the acetabulum, and anteversion of the head and neck of the femur associated with femoral torsion; if the sum of the angles of obliquity of the acetabulum and anteversion of the neck of the femur was greater than 60, dislocation occurred [21, 22, 32–34]. The femoral head and acetabulum developed in close correlation, but perfect adaptation of the component parts was present in the intrauterine position of flexion. Badgley postulated that the fault did not lie in a hereditary failure of one part but, in the embryonic and early fetal stages, there could conceivably be interference in the orderly time development of reciprocal parts after the formation of the joint cavity. Extrinsic factors were more likely to be involved than hereditary genetic factors in both the acetabular structures and the femur with changes occurring on the basis of a secondary adaptive fault from an alteration in the normal timing of development. Badgley then concentrated on the changes at the upper end of the femur that involved increased anteversion of the head and neck. It was inappropriate to ignore the changes in the femur or to call them secondary changes when they were an important part of the deformity as was the acetabulum. He felt that both acetabular and proximal femoral changes were reciprocal faults secondary to a developmental error. Anteversion occurred primarily in the diaphysis with the head and neck in normal relation with the trochanters although anteverted in relation to the shaft. The radiologic evidence of apparent coxa valga is at least partially due to increased femoral anteversion as can be shown by taking a radiograph with the femur in internal rotation at which time the angle of inclination will be found in most to approximate the normal. His concept of congenital dysplasia of the hip was that through a developmental fault, the acetabulum has failed to deepen and the head and neck of the femur have become anteverted. The anteversion tends to rotate the cartilaginous head forward and laterally so that the glenoid labrum and acetabulum cover less of the head than usual. The adaptation of the head and acetabulum requires growth changes altering the intrinsic pattern and is manifested in subluxation or acetabular dysplasia. Rotation of the limb buds may be an early important factor in the abnormal development. Interference with the orderly timing of rotation could produce a failure of the intrinsic design. The altered environment would then produce adaptive features in all structures of the hip joint and not simply a primary change in the acetabulum alone. Loss of the normal dynamic reciprocal relationship of the component parts of the hip joint during the stage of rotational adjustment of the limb buds may produce the secondary adaptive changes which lead to acetabular dysplasia or congenital dislocation, and the extended position of the infant hip after birth would further stress containment depending on the reciprocal proximal femoral-acetabular angles. Dega, on the basis of his fetal studies, noted perfect adaptation of the femoral head and acetabulum only in the uterine position

1 Developmental Dysplasia of the Hip

of flexion. The known embryological development of the hip joint is thus opposed to the theory of a primary inherited failure of development of a portion of the acetabulum alone.

1.4.2.4 Howorth and Associates Howorth noted that displacement was initially lateral and that if present in utero, the displacement was posterior. After birth with extension of the hip and especially with weight bearing, displacement tended to be upward. The only pathology seen in every case was an elongated and lax capsule. All the bony changes were felt to be secondary involving both the acetabulum and the proximal femur. He found no examples of “hourglass” constriction even though the capsule was elongated. The capsule was pulled from below across the opening of the acetabulum, but it was really the transverse acetabular ligament which created the hourglass appearance. The amount of elongation depended upon the degree of displacement. Howorth commented on the inversion of the labrum into the acetabulum with full displacement of the femoral head. His illustrations of progressive changes with increased subluxation and eventual dislocation were quite consistent with those previously presented. Although Howorth and Massie did little pathoanatomic study themselves, their extensive clinical experience and writings were based on efforts to relate to the underlying pathoanatomy [101–107]. They articulated the prominent view that congenital dislocation of the hip was due to a “pathologic relaxation of the joint capsule” and that “all other pathologic changes develop subsequently as the result of simple mechanical stresses.” That viewpoint was widely adopted in mid-century and has remained the cornerstone of most management philosophies of congenital or developmental dislocation of the hip. Bone and cartilage changes involving acetabular dysplasia, increased proximal femoral anteversion, and delayed appearance of the proximal capital femoral secondary ossification center were all felt to be secondary mechanical sequelae of a failure of the femoral head to be appropriately positioned in the acetabulum. The fact that development of these structures occurred normally once the head was definitively relocated supported the impression that they were secondary and not primary phenomena. Howorth consistently maintained that the constant and essential pathoanatomic feature and the primary anatomic cause of displacement of the hip of the fetus or infant is elongation of the capsule. Subluxation and dislocation represented the same disorder with dislocation simply being a more severe variant. The disorder was not embryonic in nature but occurred in the late fetal or immediate postnatal periods. Any deformities begin to regress as soon as the hip is completely reduced such that if complete reduction is maintained in a young infant, complete correction will occur. Bony changes in the acetabulum and femoral head

1.4 Etiology and Pathoanatomy of Developmental Dysplasia of the Hip

were secondary to the displacement and not primary causes of it. Howorth and Smith described 72 cases of congenital dislocation of the hip treated by open reduction with 16 under 3 years of age, 39 from 3 to 5 years of age, 14 from 6 to 9 years of age, and 3 at 10 years old or more [108]. Complete dislocation was seen in 93%. There were no posterior dislocations in this group, all being anterosuperior dislocations. Common findings included proximal femoral anteversion greater than 60° in 20% and greater than 45° in 58%, although 15 hips had anteversion of less than 30°. The ligamentum teres was absent in 5, ruptured in 2, thinned in 7, and elongated and often thicker in the rest. In 56 hips, the capsule was found pulled up inferiorly with the transverse ligament across the lower portion of the acetabulum as the capsule followed the displacement of the head. In 1935, 39 additional patients were reported with Farrell involving 49 hips [109]. At surgery, 9 of the hips were dislocated high posterosuperiorly and 40 anterosuperiorly. The empty acetabulum tended to fill inferiorly with hypertrophied tissue and ligamentum teres. The ligamentum teres was absent in 16, ruptured in 3, and separated in 2. The capsule was elongated and pulled up inferiorly but not constricted superiorly. The pathology appeared consistent with primary elongation or stretching of the capsule of the hip rather than dysplasia of the acetabulum or anteversion of the femur.

1.4.3 Subsequent Clinical-Pathoanatomic Descriptions with Emphasis on Early Capsular Laxity 1.4.3.1 Scaglietti and Calandriello, 1962 Scaglietti and Calandriello provided a detailed description of the pathoanatomy of the congenitally dislocated hip in their description of correction by open reduction [110]. Their material was based on 182 operative procedures in 162 patients in which 48% had the open procedure as a primary initial treatment. They divided the obstacles to reduction into two major groups, extra-articular and intra-articular. (a) Extra-articular obstacles. The two major extra-articular obstacles are the relatively shortened gluteus medius and iliopsoas muscle groups. In long-standing cases of dislocation, in particular when the patient has become ambulatory, the gluteus medius muscle is shortened and must be released from its iliac crest origins in association with correction. The iliopsoas produces great problems since it normally passes over the anterior part of the hip joint capsule to insert into the lesser trochanter. When the head of the femur is not in the acetabulum, the iliopsoas tendon is stretched tightly and crushes the capsule anteriorly against the mouth of the acetabular cavity. As the

39

femoral head is further displaced, the pressure of the iliopsoas tendon against the capsule increases and helps cause the central capsular narrowing referred to as the isthmus, leading to the hourglass deformation. At times adhesions are formed between the capsule and the psoas tendon. ( b) Intra-articular obstacles. Capsule. The capsule is thickened and greatly enlarged owing to the hip displacement that has occurred. In particular the upper part of the capsule bulges upward and outward as it functions as the main support structure for the displaced head. In many patients the capsule becomes adherent to the lateral wall of the ilium. While normally the capsule reaches to the base of the neck in some cases, it becomes attached to the periphery of the head and the adjacent neck by fibrous adhesions. This finding is referred to as the “pericephalic insertion.” The enlarged capsule is not shapeless but often has a central hourglass constriction which prevents passage of the dislocated head into the true acetabulum with closed reduction maneuvers. Limbus. Inversion of the labrum between the femoral head and acetabular cartilages can occur although the structure interposed is not just the labrum but rather a double fold of thickened capsule and the enclosed labrum which we refer to as the limbus. The interposed structure is therefore much thicker and bigger than the labrum alone. During the subluxation phase of any hip displacement, the capsule often becomes adherent to the outer surface of the labrum such that when total head dislocation occurs, the inverted structure is both capsular and labral together (forming the inverted limbus). Ligamentum Teres. The ligamentum teres is seen on some occasions but not in others. When present it is thickened and elongated in some, while in others it is elongated, thin, and atrophic. On occasion it is absent. Head and Neck of the Femur. The commonly described anteversion of the femoral neck is seen in most instances although it alone is not sufficient to prevent reduction of the head into the acetabulum. Acetabulum. The two pathoanatomic aspects to the acetabulum include the presence of an enlarged ligamentum teres with fibro-fatty tissue and a deficient cartilage roof.

1.4.3.2 Stanisavljevic, 1964 Stanisavljevic wrote a monograph on congenital hip pathology in the newborn comparing both normal and dysplastic hips [111]. Three hundred newborn hips were dissected (150 infants) with congenital pathology detected in 12 hips (all female). Four infants had bilateral and four unilateral congenital pathology involving five complete dislocations, four subluxations, and three dysplastic hips.

40

(a) Normal Hip Structure. Studies in the normal hip after removal of all the surrounding muscles but with the articular capsule and ligamentum teres intact revealed that it was impossible to dislocate or subluxate the femoral head from the acetabulum. The “click” which is often noted in normal hip exams was caused not by femoral head subluxation or dislocation but rather by an interposition of an unusually large ligamentum teres or by sliding of the iliopsoas tendon over the enlarged iliopectineal bursa. Anteversion of the femoral head/neck was between 25° and 35°, and the average neck-shaft angle was 140°–145°. Asymmetric thigh folds were often seen in children with hips that were normal. (b) Congenital Hip Subluxation. A positive diagnosis of hip (c) subluxation was made in four hips. Case 1. Bilateral hip abnormalities were found in eight full-term children who died at 6 h. No other congenital deformities were seen. The findings were indicative of prenatal congenital supero-posterior subluxation. The iliopsoas tendon was hypertrophic. The capsule was loose. The acetabulum was of abnormal shape and filled with abundant pulvinar. There was not a normal limbus on the supero-posterior aspect of the acetabulum, and in this location there was a well-developed sulcus through which the femoral head could luxate. The limbus was misshapen almost completely circumferentially and was fused on its outer aspect with the internal side of the capsule with no space present between the capsule and the limbus as in a normal hip. The ligamentum teres was hypertrophied. The femoral head was misshapen. The neck of the femur was shorter than normal. The neck- shaft angle was 145° and the anteversion was 65°. Case 2. Stillborn female with no other congenital anomalies. The tendon of iliopsoas muscle was hypertrophic and the capsule was loose. The femoral head was subluxated supero-posteriorly, and the misshapen and shallow acetabulum was filled with abundant pulvinar. There was no normal acetabular limbus on the superiorposterior aspect. The ligamentum teres was longer and thicker than normal. There was a well-developed sulcus on the supero-posterior aspect of the acetabulum through which the head could subluxate. The femoral head was misshapen and smaller than normal. Case 3. Full-term stillborn female. The iliopsoas tendon was hypertrophic and the capsule loose. The acetabulum was of abnormal shape and filled with abundant pulvinar. The limbus was not normal on the supero- posterior aspect showing a sulcus through which the femoral head could luxate. The limbus was misshapen circumferentially and fused with the internal side of the capsule with no space between limbus and capsule. The ligamentum teres was hypertrophied. The femoral head was misshapen, the neck of the femur was shorter than

1 Developmental Dysplasia of the Hip

normal, the neck-shaft angle was 140°, and anteversion was 48°. The findings were indicative of a prenatal congenital supero-posterior subluxation. The most common prenatal congenital hip pathology in the examples of subluxation was a defect of the supero-posterior region of the acetabulum that allowed the femoral head to sublux. The bone of the acetabulum in this region was also less developed and thus dysplastic. Studies of ranges of motion indicated that with the hip flexed 90° or beyond and abducted 70–75°, the femoral head was well seated in the acetabulum leaving the defect in the supero-posterior aspect free from any pressure by the femoral head. Congenital Hip Dislocation. Congenital hip dislocation was detected in 5 hips of 4 babies among the 300 hips studied. Each of these would appear to fall into the teratologic category. Case 1. Stillborn female at eighth month of fetal life. No other congenital abnormalities were found. There was an unusually thick iliopsoas tendon. The acetabulum was very small, shallow, deformed, and filled with abundant pulvinar that explained the easy “telescoping” of the femoral head noted on clinical examination. The ligamentum teres was longer and thicker than normal. The edges of the acetabulum were flat, and an intact acetabular limbus could not be detected. The capsule was larger than normal, thick at the supero-posterior region, and within the acetabulum anteriorly and inferiorly obliterating a portion of it. On the posterosuperior aspect of the acetabular edge, a sulcus was found which corresponded to the size of the ligamentum teres which was causing pressure. The head of the femur was spherical but smaller than the opposite side. The femoral neck was very short and the head was retroverted 10°. The findings represented a congenital supero-posterior dislocation of the left hip. Case 2. Stillborn female at 5–6 months. The right iliopsoas tendon was thicker than the left. Relocation of the femoral head was not possible by clinical manipulation. The entrance of the acetabulum was small because of the interposition of an enfolded acetabular limbus circumferentially. There was a high attachment of the capsule on the superior and supero-posterior region of the ilium, and no adhesions between the capsule and the external surface of the ilium were present. The capsule was larger and thicker than normal. The head of the femur was smaller although spherical and the femoral neck was short. There was no anteversion or retroversion. The findings represented a prenatal congenital dislocation. Case 3. Stillborn female 4–5 months. No other congenital abnormalities were seen, and there was a normal size to the iliopsoas tendon. It was not possible to reduce the hip with the capsule intact. The entrance into the acetabulum

1.4 Etiology and Pathoanatomy of Developmental Dysplasia of the Hip

was almost entirely restricted by the limbus which left only a small opening for the ligamentum teres. The superior capsule was attached to the external surface of the ilium forming a new acetabulum in which the deformed and flattened femoral head was found. The femoral neck was absent. The opposite hip had the same findings. Case 4. Female child died 12 h after delivery. There was hypertrophy of the iliopsoas muscle and tendon. The capsule was thick and tight anteriorly, and the femoral head was immediately under the capsule anteriorly and was dislocated anteriorly. The anteroinferior portion of the limbus was compressed and folded into the acetabulum and the remaining limbus inverted. Pulvinar tissue was abundant. The femoral head and neck were of normal shape. The neck-shaft angle was 145° and anteversion was 55°. ( d) Dysplasia of the Hip. Dysplasia of the hip was found in two instances without subluxation or dislocation. Following removal of muscle and tendon, in one case, there was an increase of motion of the femoral head in the acetabulum, but it was not possible to sublux or dislocate it. In another case the acetabulum was abnormal in shape, but the femoral head engaged the acetabulum well with the thigh flexed to 80° and beyond. The limbus was compressed and folded upward on the supero- posterior portion of the acetabulum and compressed and folded peripherally in the anterior and anteroinferior part of the acetabulum. The femoral head was of a normal spherical shape.

1.4.3.3 Salter, 1968 Salter reviewed the pathogenesis of congenital hip dislocation stressing the importance of capsular laxity in initiating the displacement with bony changes of acetabular and proximal femoral dysplasia being secondary and thus reversible with early relocation of the joint [112]. His hypothesis of the sequence of events in the etiology and pathogenesis of congenital dislocation of the hip remains active today and forms the basis of the illustration in Table 1.2. 1.4.3.4 Trevor, 1968 Trevor reported on the pathoanatomy of 61 hips with congenital dislocation in otherwise normal ambulatory children between 3 and 11 years of age who were undergoing capsular arthroplasty (Hey Groves-Colonna procedure) for undetected or persisting complete dislocation or severe subluxation with poor acetabular development [113]. The femoral head, contained within the capsule, was superiorly displaced from the poorly developed acetabulum and sometimes also anteriorly or posteriorly positioned. The capsule was often extremely thickened superiorly (often 1 cm or more) due to weight-bearing stress. Other parts of the capsule

41

Table 1.2 Pathogenesis of developmental dysplasia of the hip A. Prenatal/intrauterine factors (i) Epidemiologic 4–8:1 female/male incidence Breech position (especially after 30–34 weeks) Genu recurvatum/knee in extension oligohydramnios first child twins

minimize chances of spontaneous version

Family history (genetic component), 1–5% (ii) Structural capsular laxity (hormonal, genetic) irregularities of labrum increased acetabular anteversion increased proximal femoral anteversion

often 2 or more features, individually at upper range of normal, combine predisposing to instability

B. Postnatal factors Sudden passive extension of hips at birth and immediately postnatally predisposes to subluxation and dislocation. Hips develop in utero in flexed position and are most stable in flexion Infants positioned postnatally with hips extended and adducted are most prone to instability C. Temporal factors (i) Prenatal onset If the hip is imperfectly positioned in the third trimester, structural changes at birth are relatively marked; the earlier the malposition occurs, the worse the secondary changes since growth has occurred in utero in an abnormal position (ii) Perinatal onset If the hip is well positioned throughout the third trimester, hip structure at birth is normal, or so minimally abnormal it is grossly undetectable, other than the capsular laxity that makes it dislocatable (iii) Teratologic hip The more severe variants of DDH are often referred to as teratologic, a term used imprecisely but including two general situations: (i) the abnormal hip occurs with other structural changes implying primarily a mesenchymal cell defect [mesenchymal teratologic hip dysplasia] or (ii) the hip development is normal until malposition occurs prenatally in the third trimester with secondary changes occurring in utero [prenatal teratologic hip dysplasia] (The latter is probably more common) D. Postnatal responses to hip instability (i) Spontaneous stabilization Spontaneous stabilization of dislocatable hips (without treatment) in normal position occurs in about 50% of cases (ii) Dislocatable hips with delay in diagnosis E. General expectations for results Excellent results generally occur with early treatment; the following progression is seen: the later the time of diagnosis, the greater the secondary changes, the more complicated and prolonged the treatment, and often the less excellent the results

however had been stretched thin. In most cases there was a narrow constriction of the capsule between the displaced femoral head above and inferior part adjacent to the acetabulum (referred to as the hourglass constriction). The iliopsoas tendon was drawn tightly across the capsule toward its inferior part. Once arthrotomy was performed, the femoral head was often enlarged, especially in relation to the poorly developed

42

acetabulum, and deformed, generally with flattening where it lay opposed to the acetabulum rim or (if fully dislocated) the ilium. The articular cartilage was often focally eroded. The femoral head and neck were anteverted beyond normal. The adductor muscles were tight further limiting abduction. The acetabulum was always smaller than normal, shallow, triangular in outline, and filled with fibro-fatty soft tissue and an elongated and thickened ligamentum teres.

1.4.3.5 Dunn, 1969 Dunn reported a landmark series of several studies of normal and congenital dislocated hips assessed at necropsy [114]. Twenty-two normal hip joints from fetuses with gestational ages ranging from 13 to 40 weeks were dissected postmortem, and he reported little change observed in the relative depth of the acetabulum or in the general morphology of the joint. “In no case was it possible to provoke subluxation either before dissection or following exposure of the joint capsule. Even after division of the capsule the femoral head remained snugly within the closely fitting limbus of the acetabulum, unless considerable force was applied to the leg with the femur in full adduction and external rotation.” In a second group, 23 joints were dissected whose clinical examination had revealed hip instability [1]. There were 15 infants with 8 cases bilateral and 7 unilateral and right and left joints almost equally affected. All infants had died during labor or shortly thereafter with gestational ages from 27 to 44 weeks. Breech presentation had occurred in eight cases. The variable pathologic changes were divided into three subgroups of grades I, II, and III. Ogden also adopted this classification [115]. CDH Grade I There were seven examples of this type of deformity referred to as dislocatable hip. The head of the femur was located normally within the acetabulum, but dislocation over the posterior or posterosuperior lip of the acetabulum was possible with relatively gentle backward pressure on the head of the femur with thighs flexed and adducted. Dunn felt that the crucial pathology “appeared to lie in the limbus itself” which was unstable being stretched and slightly everted in the posterosuperior aspect giving the acetabulum an elliptical outline instead of the normal circular one. Dislocation was only partial in these cases with further displacement restrained by the capsule and the ligamentum teres. CDH Grade II There were four examples. The limbus was more everted in particular at its posterosuperior margin, the capsule more stretched, and the ligamentum teres further lengthened. Instability was marked, and partial or complete dislocation was usually present at rest. The acetabulum was usually shallower than normal, and the head of the femur had frequently lost some of its sphericity and was reduced in size.

1 Developmental Dysplasia of the Hip

CDH Grade III There were 12 hips in this group. In each case the head of the femur was dislocated upward and backward, and the limbus in particular in its posterosuperior aspect was compressed and inverted into the joint so that it formed a partial floor of a false acetabulum. The ligamentum teres emerged through a crescentic gap bounded by its free margin. The acetabulum was invariably shallow and partially developed and the head of the femur smaller than normal and less spherical. Some of the infants were otherwise normal although many had associated malformation of the neuromuscular system or urinary tract. Dunn felt that the whole spectrum of abnormality is a single pathological entity. He defined congenital dislocation of the hip as an “anomaly of the hip joint, present at birth, in which of the head of the femur is, or may be, partially or completely dislocated from the acetabulum.” Dunn later reported additional observations as his series increased to 48 hips that were dislocated or dislocatable at birth that had been subsequently dissected [116]. The studies were from 31 infants who died within a few days of birth with 16 bilateral and 15 unilateral cases of CDH. The gestational ages range from 27 to 44 weeks, and 16 of the infants had presented by breech. Clinical examination at birth had revealed instability of the hip in every case, and this was confirmed under direct vision after exposure of the joint capsule. Dunn indicated that at least 1% of all newborn infants in Great Britain had congenital dislocation of the hip as determined by a careful neonatal examination. Of these, 85% were present in infants who were otherwise normal and 15% in infants who had many other malformations. In those infants who were normally formed, 90% had a CDH of grade I and only 10% grades II and III; there was an overall perinatal mortality of 5%. In the 15% that were malformed, only 50% had grade I CDH with 50% being grades II and III and perinatal mortality as high as 70% [1]. Dunn used the term malformations to refer to abnormalities forming during the embryonic period (which were therefore teratologic), while the term deformations was used to refer to deformities arising after the embryonic period in a normally formed part and considered secondary to extrinsic intrauterine pressure. The large majority of CDH cases therefore represented postural deformity imposed on an otherwise normal hip during the late periods of intrauterine position. Congenital dislocation of the hip was frequently found with other postural deformities including torticollis; deformities of the skull, face, and mandible; and clubfoot deformity. There was a high degree of association with breech presentation, first pregnancies, and oligohydramnios. In a large prospective study performed over several years, Dunn noted that 56% of infants with CDH were first born, while 50% had presented by the breech position [114]. In the patients with CDH alone, there was a 4 to 1 female-to-male ratio.

1.4 Etiology and Pathoanatomy of Developmental Dysplasia of the Hip

1.4.3.6 McKibbin, 1970, and Ralis and McKibbin, 1973 McKibbin described findings in a child with bilateral CDH who died shortly after birth [117]. Studies were compared with hip dissections in 15 intact pelves obtained from full- term infants who died in the first 2 weeks of life from causes unrelated to the musculoskeletal system. The child with the CDH was born at term from a breech position. The infant had been in the extended leg position with fully flexed hips and fully extended knees. The child died of cerebral hemorrhage with no other congenital malformations seen. The only significant pathoanatomic finding in both hips was excessive laxity of the capsule. Indeed at initial assessment with the muscles intact, it was possible to reduce the dislocation either by abduction and flexion or by abduction, extension, and medial rotation maneuvers. When the hip was flexed, it actually was relatively loose and dislocated. Once the capsule had been removed, it was noted that there was increased length of the ligamentum teres and also increased fibro-fatty tissue in the acetabular base (pulvinar). The acetabular labrum however was normal as were the acetabulum and proximal femur both to gross inspection and by measurements of angulation that placed them in the normal range. McKibbin thus felt that the dislocation in this instance clearly began in utero, the sequence of events was initiated by primary laxity of the capsule allowing the flexed hip to dislocate irrespective of the bony conformations, and the acetabular and proximal femoral changes were secondary. He interpreted the concept of acetabular dysplasia to be entirely secondary to abnormal position of the head both in relation to this case and to other readings and interpretations of the literature. Ralis and McKibbin sought to assess the possible change in the size and shape relationship of the femoral head to the acetabulum from embryo to mid-childhood ages [35]. They studied quantitatively 44 hip joints from 11.5 weeks of embryonic age to 11 years. Measurements of the acetabulum at each age included the greatest diameter and depth of the femoral head, the greatest diameter and height, and the percentage cover of the femoral head. They confirmed the oft- disputed opinions of Sainton [73, 74] and Le Damany [21, 22, 32–34, 86] that the human acetabulum was shallowest at time of birth compared with both fetal and later postnatal findings. The perinatal period was thus the time when the risk of dislocation was greatest and dependence on soft tissue support was highest. In three embryos, the acetabulum was extremely deep-set and almost totally enclosed the femoral head. As growth continued, its shape began to change, and as fetal age increased, the acetabulum became increasingly shallow until at birth it represented as little as one-third of a complete sphere. After birth, the cavity steadily deepened again with growth. The femoral head shape also changed with the least stability observed in the newborn

43

period. In the embryo, femoral head shape was globular, coming closest to representing a complete sphere (about 80% of a complete sphere) than at any other time. As birth approached, it was closer in shape to a hemisphere, while after birth a return to a globular shape recurred partially. The proportion of the head contained within the acetabulum gradually diminished with fetal growth reaching a minimum around the time of birth, but increasing again postnatally. These findings were felt to predispose to dislocation in the perinatal period since a joint consisting of more than half of a sphere was inherently stable, whereas a shallower hip with only hemispherical (or less) relationships was more dependent on soft tissue support.

1.4.3.7 Milgram and Tachdjian, 1976 Milgram and Tachdjian reported on a 10-month-old patient with multiple congenital anomalies who died [118]. The right hip was dislocated and had been untreated. The hip disorder was considered to be teratologic. The severe changes, in particular in the acetabulum, at 10 months of age suggested an early intrauterine dislocation with subsequent malformation of hip development. The study showed a lax redundant capsule, an elongated large ligamentum teres, anterosuperior dislocation of the hip, hypoplasia of the true acetabulum that was filled with fat and fibrous tissue, and an abnormal limbus. The fibrous limbus did not appear to represent an inverted labrum with adherent capsule; rather no distinct tissue planes were present, and the fibrous limbus appeared to emerge from the floor of the false acetabulum and project over the rim of the true acetabulum. It “appeared to have been locally induced by the presence of the dislocated femoral head.” Milgram, 1976 Milgram reported a case of bilateral dislocated hips in a 74-year-old male who had had no treatment for the disorder and also “never had pain referable to the hips and thighs” even though he was employed as a security guard [119]. The left hip joint was disarticulated at postmortem assessment. The femoral head was oval and flattened medially but had a thin layer of fibrocartilage covering the articular surface. There were no degenerative changes of exposed subchondral bone or osteophytes. A thick fibrous capsule completely surrounded the femoral head separating it from the pelvis. There was no false acetabulum and the ligament teres was absent. The atrophic acetabular fossa was filled with fibrous tissue. The right hip joint was assessed following longitudinal sectioning. On this side also there was no bony contact between the femoral head and the ilium. The elongated thick capsule provided support to the femoral head. As on the opposite side, there was no ligamentum teres, the acetabulum was filled with fibrous tissues and was very shallow, and the femoral head while slightly flattened medially and somewhat smaller in size showed no degenerative arthritis.

44

1.4.3.8 Walker, 1980–1983, and Walker and Goldsmith, 1981 Walker and Goldsmith performed a series of studies in human fetuses from 12 to 42 weeks of age to document the developing structure of the proximal femur and acetabulum [120]. Both hip joints from 140 fetuses (280 joints) were assessed following elective abortion (62.2%), stillbirth (23.7%), and death during the perinatal period (14.1%). In an effort to concentrate on normal development and subtle variability, the hips had to demonstrate normal hip joint morphology by classic criteria and no displacement of the femoral head in relation to the acetabulum. The joints were dissected, morphology was inspected, and measurements were taken of the depth and diameter of the acetabulum, the diameter of the femoral head, the length and width of the ligamentum teres, the neck-shaft angle, and the anteversion of the proximal femur. Multivariate analysis showed no significant differences between males and females or between right and left sides. The acetabular depth was the slowest growing hip variable, and acetabular indices of less than 50% indicated a shallow socket at term. There was a strong relationship between the size of the femoral head and the acetabular diameter, but in many joints the femoral head diameter exceeded that of the acetabulum. Findings were interpreted to indicate that the soft tissue structures about the joint of necessity played an important role in neonatal joint stability. Except for the neck-shaft angle, the means for all variables studied increased steadily with time with the strongest increase between 12 and 20 weeks of age. Acetabular depth showed the slowest growth in the period studied, and a consistent linear growth trend was apparent in only the acetabular and femoral head diameters. Maximum values for proximal femoral antetorsion were not observed at term but rather at 32 weeks. In a number of hips, the femoral head diameter exceeded the acetabular diameter such that deep seating of the head within the socket was not possible. In younger fetuses after cutting the capsule, some force was required to displace the head from the socket, but in older fetuses division of the capsule produced immediate subluxation or dislocation of the head from the socket. Socket coverage of the femoral head was increased in the flexed position in utero, and any movement of the femur out of this position decreased socket coverage of the head. One observation not previously appreciated by many was that in 56% of the femurs, the lesser trochanter was more prominent than the greater. The data were then interpreted in relation to congenital hip disease. Although there is a clear clinical female preponderance of hip dysplasia, the study showed no difference in fetal development of the acetabulum or femur of the hip joint either between males and females or between right and left sides. This information was considered to support indirectly the hypothesis that the preponderance of DDH in females was due to a greater influence of maternal sex hor-

1 Developmental Dysplasia of the Hip

mones on female fetuses allowing for capsular laxity and hip displacement. As early as 1905, Le Damany had felt that the acetabular socket was shallower at term than at any other period of fetal life [33, 86]. This appeared true in the present study as well in the sense that whereas femoral head and acetabular diameter increased more than fourfold in the period studied, increasing depth of the acetabulum was less than fourfold. Because of the relative shallowness of the acetabulum, soft tissue structures of necessity played an important role in stability of the hip joint in older fetuses and in the newborn. Many joints demonstrated a position of maximum fit or congruency in which there was maximal coverage of the femoral head by the socket; this corresponded to the normal position in utero. This observation was most frequently made in third trimester fetuses. The neck-shaft angle measurements were somewhat lower than those published previously based on radiographs, and also there was no apparent change in the angle with age. The problem with radiographic measurements is the change with rotation in which internal rotation increases the angle and external rotation decreases it. It thus appears that the greatest increase in this angle occurs during early fetal life and the values of 125° were similar to what is noted in the adult. Femoral torsion changes were seen with the mean value of femoral torsion at birth around 35°. Many had noted neutral values up to 24 weeks of development with Le Damany [33, 86] and Watanabe [9] in particular feeling that antetorsion develops almost exclusively in the second half of pregnancy. With an increase in the amount of positive torsion, there was a change in the structure of the proximal femur that led to the lesser trochanter becoming directed more medially. The opinion was given that torsion appeared to take place in the femoral shaft and not specifically at the head-neck area. The torsion values were distinctly lower at term in this study than those reported from studies using radiographic measurements, and increases were noted to occur in the 12–18- week period. The mean adult value of 11.2° was exceeded in the study by a relatively small 62% of femurs. With a high correlation shown between depth of the acetabulum and acetabular and femoral head diameters, such data could be useful clinically. If so the current most acceptable method of obtaining that data would be by ultrasound. Walker documented a considerable degree of morphologic variability in the human fetal hip joint in his study of 280 hips and 140 normal fetuses [121]. Sixty-five of 92 hips in 46 fetuses between the age of 12 weeks and term showed structural variants although the hips were neither subluxated or dislocated and showed no other statistically significant morphological differences from normal joints. Variants observed included flattening (14) or rounding of the rim of the labrum, localized dips in the labrum (20), folding of the labrum (6), capsular folds (4), extension of the pulvinar pad between the joint surfaces (6), and kinking of the ligamentum teres (7). Most fetuses with variants showed more than one variant. The

1.4 Etiology and Pathoanatomy of Developmental Dysplasia of the Hip

vast majority of these were localized to the anterosuperior quadrant and increased with fetal age. The variant hips formed 55% of the fetuses older than 28 weeks and only 23% of those younger than 28 weeks. These findings are supportive of the belief that these variants would be potentially associated with congenital-developmental hip disease. The structural changes of hips would thus follow a continuum from normal hips to variants of dysplastic-subluxated-dislocated hips. There was rounding of the rim of the labrum with the extent of the rounding variable ranging from the entire circumference to only one quadrant. Flattening of the rim was also seen. The flattening allowed some inward fold and overhang of the labrum on the articular surface. This led to a slight ledge instead of the normal smooth transition at the junction between the inner margin of the labrum and the articular cartilage of the joint surface. A fold of the capsule projected into the acetabulum in four hips. The fold was always in the anterosuperior quadrant of the acetabulum and was associated with flattening or rounding of the labrum. In six hips there was a thin relatively avascular sheet of fibrous tissue that was connected with the pulvinar centrally in the acetabular fossa, extended over the articular cartilage, and had a free peripheral head. The tissue was thus interposed between the socket and the head of the femur and was not connected to the synovial membrane. On occasion kinks were seen in the ligamentum teres sometimes at the acetabular origin and sometimes at the femoral insertion. On many occasions several of the previously listed mild variations were present in a single hip. It was not possible to say whether these would have led to dislocatable hips at birth. Clinical examination of the intact specimens, however, indicated stability, but obvious problems with clinical interpretation are still present. Several variants some of which might well be associated eventually with a dysplastic hip were well documented in a large series. Walker also performed a study on 12 abnormal fetal hip joints that were not described as variant hips since they had more frequent or more complex aberrant features, the impression of dysplasia grossly, or actual malalignment between the femoral head and acetabulum at dissection [122]. Walker found that 10 of the 12 abnormal hip joints had several dimensions less than the range for their comparative normal age group by 2 standard deviations or greater. Two general features characterized the findings. One was that the neck- shaft angle rarely was in the abnormal range indicating that in fetuses and neonatal infants, abnormality of the femoral angles alone was a poor indicator of hip dysplasia. The major irregular finding in virtually every case related to the soft tissue component of the acetabular rim with flattening in particular of the anterior and superior rim and abnormalities of the labrum which was described as indistinct, flattened, folded inward, rounded, or associated with a dip rather than being spherical. Increased amounts of fibro-fatty tissue were also seen in the depths of the acetabulum.

45

1.4.3.9 Ponseti, 1978, and Ippolito et al., 1980 Ponseti studied the acetabulum in six infants who died soon after birth and had unilateral hip dysplasia [123]. Findings of a thickened capsule were similar to many previous descriptions. In most there was a cartilaginous ridge in the acetabulum that separated the hip socket into two sections. They were thus dealing with what some might refer to as a teratologic hip both because of the severity of the findings and the neonatal deaths. In three hips, a ridge was formed by acetabular cartilage (as defined histologically). The labrum was not interposed but was everted and plastered against the capsule. The surface of the acetabular cartilage was covered by a thin layer of fibrous tissue. In the other three hips, the more commonly described finding was noted in that there was an inverted labrum that covered the bulge of acetabular cartilage. In each instance, however, it was noted that the more medial part of the acetabular complex was both anatomically and histologically normal; this encompassed the triradiate cartilage and the adjacent acetabular cartilage and bone. It is thus evident from multiple assessments that the bulk of the acetabular dysplasia in association with CDH/DDH is occurring laterally due to abnormal pressure from a subluxed head or no pressure at all from a fully dislocated head. Ippolito et al. performed histologic, histochemical, and ultrastructural studies on hip joint capsule and ligamentum teres from 9 patients (12 hips) with CDH from 2 months to 4 years 2 months of age [124]. They concluded that the slight changes seen were secondary to mechanical stresses on the tissues caused by the dislocation rather than representing primary pathologic abnormalities. The collagen fiber bundles were thicker than in the normal capsule but appeared of normal shape ultrastructurally. In the ligamentum teres, the elastic fibers were thicker and more numerous in the dislocated hips. 1.4.3.10 Somerville, 1982 Somerville listed the causes of dislocation of the hip in infancy as (1) the effects of muscle imbalance, (2) congenital contractures, (3) intrauterine compression, (4) true teratological abnormalities, or (5) factors which are suspected but still unproved [125]. Only in the last group where the large majority of patients were concentrated was the dislocation an entity unto itself that warranted being called true “congenital dislocation of the hip.” Somerville studied and treated congenital dislocation of the hip from the early 1950s and derived a consistent program based on his initial (and unchanging) observation: “the exploration of twentythree hips with typical dislocations has led to the conclusion that an obstruction, if present, always has the same causenamely an inverted limbus. Of the twenty-three hips operated it has been found in everyone, and its removal has resulted in immediate, unresisted and complete reduction” [126]. He then articulated well how this inversion of the limbus related to the entire spectrum of the disorder as it evolved including acetabular dysplasia, femoral antever-

46

sion, and the end stages outlined by Fairbanks [127]. Somerville discounted the theory of primary acetabular dysplasia as being a common cause of CDH owing to the fact that 60–70% of all cases spontaneously stabilized in the first week of life and the large majority of the others responded to simple splinting over a several-week period. Capsular laxity was felt to be a major cause of the disorder. Capsular laxity can be generalized as in connective tissue disorders such as the Ehlers-Danlos syndrome. Laxity could develop in a part of a capsule only, while the rest remains normal or even contracted. This is the common type with CDH. A hormonal relationship with capsular laxity has been suggested and was theoretically attractive, fitting many of the facts of the disorder although no definite proof of such a relationship has yet been defined. Somerville’s major contention is that displacement occurs as a result of stretching of a previously weakened capsule and that it is a part of the capsule that is stretched rather than the whole. In a large majority of instances, “the only abnormality is the capsular laxity” with all other changes secondary. All subluxations are initially anterior and anterosuperior, and by definition dislocations, which are only a more severe variant of subluxation, are also initially anterior. Subluxation and dislocation are part of the same condition, not two different conditions. Displacement occurs when a newborn hip is placed suddenly into extension in particular if proximal femoral anteversion or lateral rotation of the hip is associated. Any extension of the hips in the newborn child, forceful or not, is potentially dangerous. Displacement thus occurs after birth when the hips are extended for the first time. Posterior positioning develops with time in the untreated as capsular enlargement greatens and the child begins to walk but is rarely seen below 4 years of age. Somerville clearly indicates the difficulties that occur when the limbus is inverted into the joint. He uses the term limbus to define the normal fibrocartilaginous outer rim of the acetabulum. Eversion of the limbus and subluxation take place slowly, whereas inversion of the limbus takes place rapidly and may be established within a few months. The inverted limbus prevents the head from entering the acetabulum fully. Somerville also indicates that the neck-shaft angle at the proximal end of the femur at birth is normal to only minimally altered. The main femoral deformity is the angle of anteversion which is increased generally to as much as 45° or 50° and may be increased as high as 90°. The hips with greater degrees of femoral anteversion are more at risk for dysplasia and dislocation. He records the usual angle of anteversion at birth as being 25–35°. The pathogenesis of the disorder is thus a mechanical one characterized by capsular laxity and femoral anteversion. If capsular laxity does not spontaneously correct itself in the first few weeks of life or splinting in flexion is not done, the disorder proceeds to a hip subluxation and then to anterior dislocation which even-

1 Developmental Dysplasia of the Hip

tually over a few years is converted to a posterior dislocation. Somerville continually stresses the importance of the position of the limbus which in subluxation is not inverted and does not cause an obstruction to reduction but in dislocation is inverted and causes an obstruction to concentric reduction. Arthrography is important in assessing the position of the limbus. With a complete dislocation, the limbus is turned into the joint; the posterior part of the limbus is always inverted. Pooling of the dye in the joint indicates that the femoral head is not in that particular area and that the obstruction must be at the periphery of the acetabulum where dye does not appear. It had been suggested by Severin that if the head is kept in relationship to the acetabulum even with an inverted limbus, the head will wear out the limbus or force it into the appropriate position [44, 128]. Somerville disagreed feeling that a false reduction would lead to damage to the associated tissues and almost certainly to premature cartilage deterioration. Renshaw also showed that soft tissue interposition between femoral head and acetabulum that is accepted following closed reduction represents an inadequate reduction which can lead to poor results [129]. An inverted limbus must thus be removed or repositioned to allow for a concentric reduction. Somerville favored excision of the limbus; at hip arthrotomy “usually the inverted limbus will be seen quite easily, looking rather like a medial meniscus.” The inverted limbus was then excised after which the hip reduced uneventfully when the leg was internally rotated. This closed the gap in the capsule so that no capsular suture was even needed. The patient was placed in hip spica in the internally rotated position for 1 month. In many instances a proximal femoral osteotomy is performed to correct anteversion with some degree of varus also build into the correction.

1.4.3.11 More Recent Pathoanatomic Findings Although the pathoanatomy of DDH has been well defined, observations continue in relation to the magnitude and variability of findings. In two studies of the same patient population, Sankar et al. documented findings in 37 consecutive hips undergoing surgical procedures for DDH. They assessed femoral anteversion [130] and femoral head sphericity [131] in patients at a mean of 33.5 months (6 months to 6 years 7 months (79 months) of age). The mean femoral anteversion was 50.3° ± 17.9°, but the range was wide from 0 to 95.7°. They also established a femoral head sphericity index using multiplanar hip radiographs with a mean sphericity score of 85.2 (range 72.2–97.3). Gross et al. studied femoral anteversion (antetorsion) in patients with DDH treated by the functional harness method [132]. They demonstrated a correlation of higher or persisting anteversion with greater acetabular dysplasia, referred to

1.4 Etiology and Pathoanatomy of Developmental Dysplasia of the Hip

as a flatter front edge of the acetabulum, a steeper acetabular roof, or a smaller CE angle. They felt that antetorsion was secondary to the hip dysplasia.

47

Figure 1.3a–h shows many examples of the pathoanatomy of congenital/developmental hip dysplasia from works of several authors.

a

b

Fig. 1.3 (a) The structural differences between a hip subluxation (right) and dislocation (left) are illustrated. Leveuf felt that each entity was separate in the sense that the subluxed hip did not become a dislocated hip although most others before and since felt that worsening subluxation would lead to dislocation. In the subluxed hip (right), the acetabulum is shallow and oblique, but the labrum remains on top of the upwardly displaced femoral head and is plastered against the capsule; in the dislocated state (left), the labrum and adjacent capsule are inverted into the joint, and the femoral head is even further displaced and supported only by the capsule superiorly. (Reprinted with permission from Leveuf, JBJS). (b) The more commonly accepted view, illustrated here, is that with displacement the hip passes from a normal anatomic arrangement (left) to a subluxed state (center) to a fully dislocated state (right). Crucial to the normal anatomic alignment is the radiolucent roof of the acetabulum that in the normal state is composed of the acetabular cartilage plus the triangular fibrocartilaginous labrum and the adjacent capsule that attaches beyond the labrum into the side wall of the acetabular bone. With subluxation the head moves laterally and upward, but the labrum remains on top of the head although it tends to become flattened. In the fully dislocated state, the head is displaced onto the side wall of the ilium, and both the labrum and part of the capsule are inverted to lie between the acetabular cartilage and the fem-

oral head cartilage. The inverted labrum and capsule together are referred to as the limbus. (Reprinted with permission from Scaglietti and Calandriello, JBJS Br). (c) Dunn’s classification of hip displacement is widely referred to. The normal position of the hip with overlying labrum and capsule is seen at upper left. In grade I, the capsule and labrum are stretched and displaced slightly but are still in normal position to support the head in the socket without interposition. In grade II, the labrum and acetabular rim are deformed, and the inferior labrum, transverse ligament, and lower capsule are stretched and malpositioned as the head further displaces. In grade III, there is complete dislocation of the head from the acetabulum, the acetabulum is further malformed (dysplastic), the labrum and capsule are inverted (=limbus) between head and acetabulum, the ligamentum teres is stretched, and the inferior capsule and transverse ligament further block reentry of the head into the acetabulum. (Reprinted with permission from Tönnis D, Congenital Dysplasia and Dislocation of the Hip, Springer-Verlag, 1987). (d) A teratologic dysplastic neonatal hip is shown. The displaced elongated femoral head is at right and the dysplastic acetabulum at left. The ligamentum teres, increased fibro-fatty tissue in the depths of the acetabulum, and a flattened malpositioned capsule and labrum are interposed between the acetabular cartilage and the femoral head cartilage. (Reprinted with permission from JBJS Am, Milgram and Tachdjian)

48

1 Developmental Dysplasia of the Hip

c

d

Normal

Grade 2

Grade 1

Grade 3

Fig. 1.3 (continued)

1.4.4 M ultifactorial Causes of DDH Involving Late-Stage Structural Modifications of the Hip, Mesenchymal Tissue Abnormalities, and Intrauterine Mechanical Stresses due to Positioning (Wilkinson, Dunn, Seringe et al.) 1.4.4.1 Overview It is increasingly considered that DDH occurs in relation to several associated factors since multiple pathoanatomic dissections along with clinical, radiographic, arthrographic, and sonographic studies almost invariably fail to reveal marked structural abnormalities in the perinatal period. There is widespread acceptance of the fact that there are two prime factors in the mechanism of late intrauterine and perinatal displacement of the hip, these being hip capsule laxity and mechanical stresses placed on the fetal hip joint. Some observers still consider that subtle primary hip abnormalities, often described as variations of normal development, in combination predispose to hip dysplasia. These involve acetabular labrum abnormalities, a more shallow and anteverted acetabulum, and increased proximal femoral anteversion. Wilkinson [133–136], Dunn [1, 114, 116, 137, 138], and Seringe et al. [139–141] have studied and summarized well the multifactorial considerations. Both feel that the mechanical factors are most

important and that DDH is associated with a characteristic position in utero involving hyperflexion of the hip with adduction and external rotation causing an abnormal pressure on the greater trochanter leading to an expulsion of the head upward and backward. Wilkinson further demonstrates that it is in effect the hyperextended position of the knees that places further increased stress on the fetal hip joint. Seringe et al. sought to explain how the mechanical intrauterine forces on the hip could account for the tendency to capsular stretching and subsequent instability regardless of the position of the fetus in relation to the maternal pelvis. They felt that the knee hyperextension theories of Wilkinson and others related only to the breech position. Their theory was considered to be equally valid in full breech presentations whether the knees were flexed or extended and in the cephalic presentations where the knees were almost always flexed. They identified three dislocating postures, each one of which was characterized by the damaging role of lateral rotation of the lower limb as indicated by Wilkinson. The first position had the knees extended and the lower extremities in lateral rotation, the second had the knees semiflexed but also in lateral rotation, while the third had the knees in full flexion and in contact with each other in neutral rotation but with an excessive femoral anteversion which is equivalent to lateral rotation in terms of the effect of the proximal femur on

1.4 Etiology and Pathoanatomy of Developmental Dysplasia of the Hip

the acetabulum and capsule. The laterally rotated position was in some way caused by fetal-maternal mechanical compression. In these situations, which are most common with breech presentations, there is failure of the leg-folding mechanism. In cephalic or vertex presentations, only 1% of babies have the knees extended. Much of the information on the multifactorial influences on CDH/DDH has been drawn from epidemiologic studies on large numbers of patients. These findings and their relationship to the pathophysiology of the disorder are presented in more detail in Sect. 1.5 below. Seringe et al. have reviewed both the intrinsic and extrinsic causes.

1.4.4.2 Intrinsic Causes These include a primary acetabular dysplasia, subtle imperfections of the anatomic structure of the labrum, increased anteversion of the femoral head and neck, and capsular laxity. Primary acetabular dysplasia is accepted only rarely as a primary causative mechanism at the present time since the acetabulum is normal or virtually normal in almost all hip dysplasia patients at birth and, more importantly, concentric and maintained reduction of the femoral head in the first year of life leads to rapid normalization of acetabular development. A primary dysplasia of the acetabulum can be considered to exist only where abnormal development persists even with clear concentric reduction. Proximal femoral anteversion and capsular laxity are considered to be predisposing causes to DDH. Seringe et al. considered capsular laxity to be secondary to displacement of the femoral head and not the primary cause of the dislocation since their studies defined not a diffuse laxity initially but rather “an elongation of the posterosuperior part of the capsule [139, 140].” The capsular laxity may be due to a genetic connective tissue abnormality or to a perinatal hormonal imbalance in which circulating “relaxing hormones” from the mother are super-added to the hormones of the child leading to the tendency to capsular laxity in the latter. This theory is theoretically attractive and correlates well with the much higher female incidence of DDH in the range of 4–8 to 1. It was proposed several decades ago but no truly definitive demonstration has been made. 1.4.4.3 Extrinsic Causes The extrinsic or mechanical causes for DDH refer to intrauterine forces on the fetal hip and help in explaining the association of DDH with many epidemiologic features. The postural-mechanical cause of the disorder is supported by the high incidence of DDH and breech positions; high incidence with genu recurvatum, torticollis, and foot deformities; its association with first births (where the uterine and abdominal muscles are relatively tight); and its association with oligohydramnios where there is diminished intrauterine fluid limiting spontane-

49

ous version in the third trimester, twin pregnancies, and excessive birth weight. The mechanical predisposition can also be defined as due to intrauterine factors, factors at the moment of birth at which time the hip is relatively forcibly hyperextended, and factors in the immediate postnatal period in which those groups which position the infants with the hips flexed and abducted tend to have a much lower incidence of long range problems than those where the child is placed with the hips extended and adducted. The latter two factors are not considered to produce a hip instability or dislocation but rather to diminish the likelihood of a spontaneous stabilization occurring. Seringe et al. defined a unifying theory for the pathogenesis of CDH/DDH [139–141]. Dislocation of the hip was produced mechanically by a combination of two factors. The first involved the position of the proximal femur where the head was not directed toward the depths of the acetabulum but rather toward the rim of the cavity and the capsule; this was the dislocating posture. The second feature required a force pushing on the proximal femur to lead to its displacement; these forces involved pressure on the greater trochanter and also relative over activity of certain muscle groups with the predisposed dislocating posture involving primarily the psoas, adductor, and hamstring muscles. The dislocating posture involved excessive flexion with a certain degree of lateral rotation or an excess of proximal femoral anteversion and some adduction. In summary, it was pressure on the greater trochanter of the femur in lateral rotation or with excessive anteversion that served to drive the femoral head above and behind the acetabulum. In the case of breech presentation, the pressure on the greater trochanter arose from contact with the narrow upper part of the maternal pelvis, while in cephalic presentations contact arose from the maternal lumbar spine which helped explain the greater frequency of left unilateral dislocations since the fetal spine is more often on the left. The dislocation thus develops at the end of fetal life or even during the period of labor under the primary influence of the mechanical factors. At birth the dislocated hip is freed from intrauterine constraints, and two possible pathways of development can occur. If the instability persists, the dislocation is perpetuated and becomes progressively more irreducible leading to a congenital dislocation of the hip. The other pathway involves spontaneous improvement which if complete leads to a normal hip and if incomplete can leave the patient with a subluxation or residual acetabular dysplasia. The majority of cases labeled as teratologic were really just straightforward congenital dislocations which appeared structurally worse because they had occurred earlier in the intrauterine period such that the secondary changes were more marked at the time of birth.

50

1.5

1 Developmental Dysplasia of the Hip

xperimental Reproduction of Hip E Dislocation

A few experimental studies have (i) effectively reproduced the mechanism of dislocation as seen in breech malpositions by holding the knee in extension and (ii) reproduced the secondary adaptive changes of the acetabulum in those situations where the femoral head was not present in the acetabulum during early development. The concept of capsular laxity, defined first by Sedillot [60], was also accepted by Langenskiold and Laurent [142] and Langenskiold et al. [143]. Experimental dislocation of otherwise normal hips in experimental animals was followed shortly by dysplastic changes similar to those seen in humans. The work in rabbits [143] and puppies [144] lent support to the belief that stretching of the soft parts of the joint was a major contributing cause of DDH. (a) Production of subluxation and dislocation of the hip by maintenance of the knee in the extended position in young rabbits. Michelsson and Langenskiold were able to produce subluxation and dislocation of the hip in rabbits less than 3 weeks old by placing the knees in extension in plastic tubes and maintaining that position for several weeks [145]. The hip was in normal position and the animal was allowed to move about. In another group of animals, the knee was held in extension, but the hamstring muscles were surgically released either proximally or distally. In the group with the knee immobilized in extension, 83 of 87 hips developed some abnormality involving hip dysplasia, subluxation, or complete dislocation. In another group of seven rabbits in which the hamstrings were cut, no hip changes were noted. In eight rabbits both hind limbs were immobilized in the same way. On one side the hamstrings were cut leading to no hip changes, while on the other they were left intact leading to dislocation in 4 and subluxation in 4. Investigation clearly demonstrated that in almost all young rabbits less than 3 weeks of age, it was possible to cause subluxation or dislocation of the hip along with most of the secondary changes characteristic of human CDH by simply immobilizing the knee in extension for 3–4 weeks. In these experiments the hip was allowed to move freely, and the animals were noted to keep the hip in a normal position of flexion. Dislocation occurred despite the fact that the hip was free to move in a normal position. They concluded: “the only obvious pathological factor in the region of the hip provoked by immobilization of the knee in extension was increased tension in the hamstring muscles.” That conclusion was further supported by the observation that when the hamstring muscles were cut, immobilization of the knee in extension was

not followed by any positional changes. They concluded that this reproduced well the mechanism of dislocation in the human breech position where hip dysplasia was far commoner than in children born in the vertex position with the knees flexed. ( b) Production of hip displacement in young rats. Canillas et al. produced hip dysplasia and progressive subluxation in young rats by the same mechanism of immobilizing one knee in the hyperextended position using fixation of the tibia and femur with a Kirschner wire [146]. Radiologic, macroscopic, and microscopic changes in the hip were assessed at 4 days, 1 week, and 2 weeks with macroscopic changes alone at 6 weeks. As in rabbit experiments, a progressive hip dysplasia and subluxation was produced with pelvic and acetabular changes accompanying displacement of the femoral head.

1.5.1 Developmental Changes in the Acetabulum Following Experimental Displacement of the Femoral Head During Early Growth It has been unclear whether or not the developmental acetabular and proximal femoral changes in CDH/DDH are primary abnormalities leading to displacement or whether they occur secondary to the displacement owing to growth in an abnormal relationship. Experimental studies have addressed the matter by removing the femoral head surgically from the acetabulum in early life and following development of the bony segments with time. Smith et al. performed an experimental displacement of the hip in young puppies dislocating the right hip surgically in each of 22 animals under anesthesia while leaving the contralateral hip in place [144]. Study was then performed from 4 to 8 weeks following dislocation. As a consequence of experimental dislocation of the hip in puppies 3–5 weeks of age, the following changes were observed: (1) acetabular dysplasia as early as 4 weeks following dislocation, (2) progressive dysplasia to the point of an unrecognizable acetabulum at the time of maturity, (3) normal acetabular development following experimental dislocation and immediate relocation, (4) pronounced changes in the head and neck of the femur in the dislocated state, and (5) no abnormalities in the head and neck of the dislocated hips in which the femurs were relocated. They felt that it was the displacement of the otherwise normal femur in relation to the acetabulum that led to the secondary developmental changes. Similar findings in the rabbit were produced by Langenskiold et al. [143]. The right hips of rabbits 1–5 days

1.5 Experimental Reproduction of Hip Dislocation

of age were dislocated under anesthesia by manipulation. Findings were followed in 101 animals by radiographs at varying intervals. Histologic sections were also made. In the study, 29 of the 101 hips reduced spontaneously and subsequently developed in normal fashion. In those that remained dislocated, many of the features of CDH in the human were seen. These included dysplasia of the acetabulum, dysplasia of the head of the femur, anteversion of the femoral neck, capsular distortion with formation of an isthmus and limbus, and acetabular dysplasia without dislocation. Histological analysis of 23 animals demonstrated changes in the acetabular roof, limbus, and femoral head similar to those seen in human CDH. Their interpretation, similar to that of Smith et al., was that the changes in the femur and acetabulum were secondary to the displacement which itself could be due to some abnormality of the soft tissues. Histologic sections of the displaced femur and acetabulum clearly show that the acetabular dysplasia occurs almost exclusively due to growth abnormalities lateral to the triradiate cartilage owing either to pressure of the femoral head against the lateral acetabular cartilage or to complete displacement of the head leaving the cartilage to grow without any opposing mechanical stimulus. Harrison performed a classic study demonstrating the influence of the femoral head on pelvic and acetabular growth and shape in the rat [40]. He performed several surgical procedures leading to unilateral and bilateral excision of the femoral head, dislocation of the hip, and amputation of the entire lower extremity removing the femur completely leaving the acetabulum to develop in the absence of mechanical pressures. Acetabular development was grossly and histologically abnormal in each of these groups (Figs. 1.4a, b). The acetabulum was narrower, shallower, and smaller on the operated side. The rim of the acetabulum gradually lost its sharp edge and became blunt and ovoid with the long axis running dorsoventrally. Histologic exam showed the capsule to invert into the socket along with the rim of articular cartilage. The acetabulum became filled with fat and sealed laterally with a layer of fibrous tissue. Histologic sections showed a particular maldevelopment of the lateral acetabular cartilage with the triradiate cartilage being relatively unaffected. Even histologically the iliac, ischial, and pubic growth cartilage plates within the triradiate cartilage in the medial wall of the acetabulum were unaffected either in thickness or histological structure. Subtle variations were seen in each of the different approaches. In addition, there were general pelvic growth abnormalities as well although the acetabulum was most severely affected. The major disturbances of acetabular development included failure of the socket to develop in area and depth, blunting irregularity of the acetabular margin, and atrophy and degeneration of the articular cartilage. It was the acetabular cartilage of the lateral regions that was affected with the triradiate growth cartilages medially remaining

51

h istologically normal and for the most part supporting total length growth of the innominate bone. He also concluded: “it is more likely that the dysplasia found in congenital dislocation of the hip is the consequence rather than the cause of the dislocation.” The experiment showed yet another example of alterations in the pattern of mechanical stress leading to growth abnormalities owing to absence of the pressure of the femoral head against the acetabular cartilage. Absence of the femoral head from the socket resulted in marked acetabular dysplasia and pelvic asymmetry.

1.5.2 B reech Malposition and Hormonal Laxity Causing Hip Dislocation in Young Rabbits Wilkinson reproduced the features of hip dysplasia in 6–8-week-old rabbits by splinting one of the hind limbs in full extension to reproduce the breech posture and by giving the animals estrone and then progesterone to induce joint laxity hormonally [133]. Assessments were done after 6 weeks of splintage and hormone therapy. Both male and female rabbits were used. Dislocation was produced preferentially in the females who had the combination of the hormones and the knee extension splinting. Dislocation was also more common with lateral femoral rotation breech malposition along with the presence of hormonal joint laxity. Subsequent anatomic studies showed soft tissue and bone deformities in the dislocated regions quite consistent with those seen in the human disorder. With lateral femoral rotation, there was inversion of the posterior capsule of the hip along with swelling and fibrosis. He felt these were analogous to development of the limbus (invented labrum plus capsule) in the human. The ligamentum teres was often thickened but never ruptured. When the knee extension and hormonal joint laxity were performed in isolation, no dislocations occurred. All displacements occurred when both were acting indicating that both were essential to the dislocating mechanism. Medial and lateral rotation of the hip was dictated by the mode of application of the knee hyperextension splinting. Hormonal induced joint laxity. One of the primary epidemiologic findings in all studies with congenital hip disease is the high female preponderance. This has generally been considered to be associated with the increased circulating female hormones present in the female fetus both acquired from the mother and present in the child. During pregnancy the hormonal level rises in the mother allowing for progressive softening and lengthening of the maternal pelvic ligaments and producing a pliancy of the birth canal that makes labor easier. During the second trimester, placental estrone and progesterone pass into the fetal circulation. These hormones produce minor degrees of fetal laxity by their direct

52 Fig. 1.4 Illustrations are shown from the work of Harrison who induced femoral head dislocation in the young rat and studied secondary developmental changes of the acetabulum. (Reprinted with permission from TJ Harrison, Journal of Anatomy). (a) The shape of the normal acetabulum is shown at left (a) while that of the acetabulum developing in the absence of a femoral head is at right (b). The abnormal acetabulum is much smaller and shallower than the normal and is misshapen with an oval appearance compared to the normal circular shape. (b) Cross sections of the acetabulum in two experimental groups are shown. The changes are similar to those outlined in human developmental dysplasia (Fig. 1.2b). Figures (a) above and (d) below show normal acetabulae. The figures to the right in each instance (b, c, e, and f) are abnormal. There are decreased width of the acetabulum, decreased depth, and a general misshapen appearance

1 Developmental Dysplasia of the Hip

a a

b

b a

b

c

G

G

K L

K

J

I

L

J

I B

H

A H d

action on the developing ligaments. In the female fetus, they stimulate the immature uterus to produce relaxin. Andren and Borglin indicated that diminished hepatic function in children with congenital dislocation of the hip allowed cir-

e

f

culating levels to be even higher and thus increasing the hormonal laxity [147]. Delgado-Baeza et al. also produced pelvic deformity in 2-week-old rats by experimental methods involving

1.6 Epidemiology and Its Relation to Pathophysiology

immobilization of the knee in extension, hormonal alterations, surgical dislocation by capsular release and section of the ligamentum teres, and resection of the femoral head [148].

1.5.3 M echanical Induction of Hip Deformation and Dislocation In Vitro Hjelmstedt and Asplund have studied human hip stability by mechanical means, inducing displacement in vitro in infant autopsy specimens [149]. Loading of hips at 45° flexion induced deformation and dislocation similar to that seen clinically with capsular stretching and no gross structural damage. In addition, periods of time as short as 3 h sufficed to lead to displacement. Their conclusion was that hip displacement could be induced in vitro solely by moderate mechanical forces in a short period of time.

1.6

pidemiology and Its Relation E to Pathophysiology

Certain epidemiologic considerations are important in understanding the underlying bases of congenital/developmental hip abnormalities.

1.6.1 Sex Incidence There is a markedly higher incidence of the disorder in females compared to males. In major series reported over the past several decades, females account for roughly 80% of the cases, a 4:1 ratio. In a large study from New York City, the female/male incidence of unstable hips was 4.15/1 [150]. Many series show an even higher female predominance: Putti, 5.7:1 (in 1879 cases) [151]; Farrell et al., 5.7:1 (in 310 cases) [96]; Lempicki et al. 6.6:1 (in 1010 cases) [152]; and Grill et al. 8:1 female-male predominance in 2636 dysplastic hip patients [153]; and Somerville described typical congenital hip dislocation to be eighth times as common in girls as in boys [125].

1.6.2 Incidence and Side of Hip Instability The incidence of hip instability is quite high and generally varies between 1% and 3% of all newborns depending on the region of the world where the assessments are made and the inclusion criteria used. A large study of 23,408 patients in New York City from 1966 to 1972 showed an incidence of 1.33% involving dislocatable or dislocated hips [150]. The study classified 82% as dislocatable and 18% as dislocated.

53

The incidence of instability is reasonably well known based on large-scale neonatal screening studies. Howorth summarized 11 studies reported from 1950 to 1975 examining 200,000 newborn infants with an instability rate of 0.9% [102]. In five major studies done by investigators with a committed interest in diagnosing hip dysplasia in the newborn, the incidence was 1.4% (of 105,375 infants). It is widely estimated that one of each 100 newborns will show some hip instability with 50–60% of those stabilizing spontaneously without treatment within 1 week of birth [116]. The incidence of DDH is greater in certain European countries than in England and North America. Tönnis has reviewed the literature of the second half of the twentieth century extensively showing higher incidences in Germany, Austria, Poland, and the former Czechoslovakia [154]. The incidence is very low in those of black African and Chinese descent, but is high in Japanese. The criteria used by differing authors and in particular in relation to large screening programs over several decades can make comparative studies difficult, but the trends and tendencies shown by these numbers appear accurate. The left hip is affected more commonly than the right in most series with bilateral involvement intermediate. Putti [151] showed 39% bilateral, Farrell et al. 27% [96], and Coleman 38% [155]. The incidence of DDH appears to be stable over several centuries. A study by Mitchell and Redfern found a 2.7/1000 prevalence of DDH in a large medieval skeletal collection excavated from Britain from the years estimated between 1100 and 1530 ad, a number similar to current figures [156].

1.6.3 Effects of Intrauterine Environment Several mechanical features of the intrauterine environment are related to congenital hip problems. There are many who feel that these mechanical effects are primary causes of DDH. Among the most common of these effects are breech presentation, oligohydramnios, increased incidence in firstborn children owing to the relative tightness of the uterine and abdominal muscles, and twin pregnancies. Breech presentations of firstborn children are most affected with hip instability (10%).

1.6.3.1 Breech Position Congenital/developmental dislocation of the hip has been recognized since the late nineteenth century to have an increased incidence in breech presentations [70, 71]. Studies of CDH/DDH have documented 15–25% as occurring with breech presentations. The incidence of unstable hips in breech presentations was noted to be 6.35 times that of vertex presentations [150]. The large study noted breech presentations to be present in 4% of all pregnancies (with vertex presentations at 96%), while Vartan [157] documented a

54

2.2% breech incidence, Barlow [158] a 4.4% incidence, and Wilkinson a 2.6% incidence [135]. Suzuki and Yamamuro studied the hips in 6559 newborn infants to assess mechanical factors of presentation in relation to congenital hip deformity [159]. The incidence of CDH was 1.1%, 0.7% with cephalic presentation, 2% in footling presentation, and 20% in single-breech (knees extended) position. CDH was also found in six of seven with congenital genu recurvatum (hyperextended knees). These observations are consistent with the fact that a fetal position with the hip flexed and the knee extended predisposes to the development of CDH/DDH. Suzuki and Yamamuro later illustrated this mechanism by ultrasound demonstrating femoral head movement in five patients with hip dislocation 3–4 months of age; with the hip joint flexed, the dislocated head was posterior to the acetabulum and displaced even further posteriorly when the knee was extended [160]. The actions of the hamstrings aided in the displacement with this positioning. Wilkinson studied breech malposition and its relation to CDH [133, 134, 136, 161]. The term breech position simply refers to the lie of the fetus in utero, while the causative association of hip dysplasia was with true breech malposition which involved hip flexion and knee extension. The commonest intrauterine posture at birth is the vertex or cephalic posture that is established by 30–34 weeks of gestation. In the vertex posture, the legs are folded with the hips and knees both flexed. If leg folding never takes place, the primary breech posture is common. The breech posture with knees hyperextended is common during the second trimester, but once the fetal spine, hips, and knees flex, spontaneous version usually converts the breech to a cephalic presentation. This conversion is favored by low uterine muscle tone of the multigravida along with increased intrauterine fluid. Vartan in a classic study of fetal positioning in utero observed that the breech position was common until the 30–34-week period of gestation, being present in 25% of pregnancies [157]. Spontaneous version occurred in 60% of these in association with knee flexion. When therapeutic version was added to the spontaneous corrections, only 8.5% of 1000 infants who at one time were in breech position were formal breech deliveries. In the entire series of 3875 patients, the eventual incidence of breech delivery was 2.2%. Vartan also noted that maintenance of the breech posture was associated with extended knees and diminished amniotic fluid, both of which limited spontaneous and even therapeutic version to the cephalic position. In the primigravida muscle tone is high and fluid relatively less. In frank breech births, Wilkinson noted breech malposition with the knees extended and the legs internally rotated or the knees semiflexed and legs externally rotated. Persistence of these postures beyond the 28th week of gestation constitutes breech malposition, and after the 32nd week, they usually remain unchanged until birth. Vartan noted that prior to the 30th week, the breech posture

1 Developmental Dysplasia of the Hip

was present in one of four but that spontaneous version occurred with knee flexion. The majority of cephalic or vertex presentations show the hips and knees to be fully flexed and the thighs adducted and internally rotated, and only 1% have the knees extended. Firstborn breech deliveries usually have the hips flexed and the knees relatively extended with the lower extremities externally rotated. At a later stage of breech malposition, there is still locked external rotation of the lower extremities but some flexion of the knees present. The high percentage of extended knee postures in breech positions and in those with DDH sustains the theory that failure of the leg-folding mechanism is one of the chief mechanical causes of breech birth. Evidence of delayed leg folding of some extent was present in 80% of newborn infants with hip displacement. Wilkinson noted that in 174 patients with CDH, there was clinical evidence of breech malposition in at least 25% [135], and 60% of patients with CDH were firstborn. In 20 consecutive cases of neonatal CDH, all of the newborns had displayed the locked-breech birth posture. Breech position was noted by Barlow to have a high correlation with CDH/ DDH. In an examination of 8814 normal babies, only 4.4% were breech presentations. Of 139 with hip abnormalities however, 17.3% had breech presentations [158]. Calculations from his data indicate that approximately 6% of babies with breech presentations have hip dysplasia, while the data of Artz et al. [150] showed 7% of breech presentations with unstable hips. Fox and Paton studied the relationship between the mode of delivery and DD in 571 consecutive breech presentations [162]. When all grades of hip dysplasia were included (Graf II, III, and IV), the incidence of dysplasia was the same in the three groups comprising elective Caesarian section (262 patients, 8.4% DDH), emergency section (223, 8.1%), and vaginal delivery (86, 7.0%). When the more clinically involved Graf III and IV cases of DDH were subgrouped however, vaginal deliveries were associated more often with DDH compared to elective sections by a 4:1 ratio. Sarkissian et al. demonstrated that spontaneous stabilization of sonographic instability (on dynamic ultrasound examination) in breech position infants was 3.72 times more likely compared with non-breech infants by regression analysis. The overall incidence of spontaneous stabilization in 122 hips was 74% of a mean age of 9 weeks (4–18 weeks) with 80% in breech versus 66% in non-breech [163]. In summary, the overall incidence of human birth in a breech position is 2–4%, while anywhere from 15% to 25% of children with CDH are born from the breech position. The knees are extended in utero in 56–75% of children born in a breech position, while only 3% of children born in a cephalic/ vertex position have their legs extended in utero. In a newborn infant born in a breech position, the frequency of congenital displacement of the hip is in the range of 4–7%.

1.6 Epidemiology and Its Relation to Pathophysiology

1.6.3.2 Birth Order The incidence of hip dysplasia in most series is shown to be higher in firstborn infants. In the study by Artz et al. of infants with unstable hips, 63% were first born, 21% second, 8% third, 4% fourth, and 4% fifth [150]. The birth order affects the intrauterine environment since muscular forces on the fetus are greatest in the first pregnancy owing to both tighter uterine and abdominal musculature. Dunn noted 58% of patients with CDH to be first born [116]. 1.6.3.3 Prematurity Prematurity does not appear to be a risk factor for DDH. In a well-balanced physical exam and ultrasound study comparing both hips in 221 infants born at a mean age of 31 weeks with 246 infants born at a mean age of 40 weeks (and having no risk factors for DDH), prematurity was not found to be a predisposing factor for DDH. There was a statistically significant prominence of immature hips in the preterm infants (defined as α (alpha) angles less than 60°) but no significant difference of pathological dysplasia (with α angles less than 50°) [164]. 1.6.3.4 Association of DDH with Torticollis There is a definite coexistence of congenital muscular torticollis (CMT) and DDH. Joirer et al. found an incidence of 12% with DDH in 97 patients (12/97) with CMT [165]. They recommended ultrasound or radiographs of the hips as part of evaluation of a patient with CMT. This association has been known for several decades and appears related to the high incidence of breech presentation also associated with CMT [166]. Tien et al. found DDH in 8/47 (17%), 4 of whom patients with congenital muscular torticollis, 4 of whom (8.5%) needed active treatment [167].

1.6.4 Extrauterine Postnatal Environment The positioning of the infants shortly after birth can vary from society to society, and in those groups where infants tend to be positioned primarily with the lower extremities and hips abducted, the incidence of subsequent hip problems is low, and in those where the lower extremities are maintained for relatively long periods of time in the adducted and extended position, the incidence of hip problems is increased [112].

1.6.5 Genetic Considerations Developmental dysplasia of the hip is not a classic hereditary disorder in the sense that clear recessive or dominant patterns of inheritance have not been documented. There appears, however, to be a slightly increased likelihood of the disorder

55

occurring in those where a clear family history of DDH has been documented. The disorder is thus genetically heterogeneous with environmental modifiers. The relatively few studies that have been done in this regard and the imprecise use of terminology and diagnostic criteria make it difficult to interpret studies published, but a slight predisposition to familial occurrence appears real. Stalder and Jani calculated the genetic risk after a boy with CDH was approximately 10% for the next child, after a girl with CDH the risk was 3%, and when one of the parents had CDH, the risk for the first child was 5% [168]. The familial incidence is postulated further to be due to joint laxity syndrome which itself predisposes to the capsular hip joint laxity allowing displacement to occur. Dunn reported a strong family history of 3% [116]. A more recent study from Utah (USA) has further strengthened the concept of a familial predisposition to DDH. In a study of 1649 distinct individuals with DDH, the relative risks (RR) were significantly increased in first-degree relatives (RR = 12.1), siblings (RR = 11.9), and first cousins (RR = 1.7) [169].

1.6.6 Ethnic Considerations Congenital hip problems are extremely common throughout European societies and in groups in North and South America originally of European descent. Once again, one must note the definition used in various studies to allow for the CDH diagnosis and the rigorousness of the examination criteria. There are differences even within the European community. There is a particularly high incidence of congenital or developmental hip disorders in Central Europe. Tönnis et al. [170] report that with ultrasound screening of newborns, as many as 2.7% were found with pathologic hips. The incidence has also been reported as quite high in those from Lapland, Northern Scandinavia, and in North American Indians although the extrauterine environment in which the children are placed may play a major role in this regard. Several studies from a few decades ago also indicated a markedly decreased incidence among Chinese and black children.

1.6.7 S pontaneous Stabilization of Hips Without Treatment Barlow performed detailed studies documenting instability in newborn clinical examinations of hips followed by a tendency to spontaneous stabilization in normal position in as many as 60% without treatment over a short period of time [171]. Using clinical assessment only he reported on examinations of 19,625 children in which 357 abnormally lax hips were found – 168 frankly dislocated and 189 dislocatable. Considerable spontaneous stabilization occurred however as

56

determined by the fact that in 931 babies examined on the day of birth and 1197 babies examined at the age of 1 week, there were far more unstable hips in the former group than in the latter. The percentage of abnormal hips was 2.3% in the first group, and only 1% in the second group. Extrapolating these figures to larger groups, he predicted that if one were to find 100 abnormal hips on the day of birth, there would be only 40 abnormal hips 1 week later and that 60 of the original 100 would have undergone spontaneous recovery. Spontaneous recovery could occur up to the age of about 6 weeks, but most of the stabilization occurred quite early. The occurrence of spontaneous stabilization led to some differing practice approaches in different centers. In some centers all patients with unstable hips at birth are reassessed clinically at 1 week. If the patient has undergone stabilization to clinical assessment, then no further treatment is indicated. If the dislocatable state persists, then treatment is instituted. Use of the Pavlik harness has made treatment much easier, and increasing medical-legal concerns have also led many to treat all hips that are dislocatable even at birth. The widespread use of ultrasound has further refined neonatal assessment, and many are guided by the sonographic indices as well. There are definite complications noted with the Pavlik harness and other forms of abduction splinting such that it would still appear to be medically acceptable to reexamine an unstable hip clinically at 1 week of age at which time ultrasound exam can also be performed. Treatment would then be guided by clinical exam and ultrasound at 1 week rather than by the finding of dislocatability on the day of birth. The tendency of many unstable hips to stabilize spontaneously was commented upon in some detail by Le Damany in 1914 [33]. What he referred to as the “subluxable hip of the newborn” was frequent in France. He clearly described the dislocation and relocation maneuver in which one could detect the head coming out of the socket and then again a repositioning of the head into the acetabular cavity. He noted that “this difficulty is almost spontaneously corrected in a length of time varying from several days to several months”; it was more frequent in girls and was due to a relative diminution in size of the acetabulum. If the torsion of the femur was not too great, recovery would take place, but if the femur had excess anteversion, then the subluxable condition would be the first stage of a congenital dislocation. Le Damany also described the clinical maneuver, referred to widely as the Barlow maneuver. Dislocation occurred when the hip was placed in the flexion-adduction position following which a slight force applied to the knee combined with a force from within outward on the thigh caused the head to half emerge above the posterior border of the acetabulum. Displacement of the thigh toward a flexion-abduction position allowed the head to reenter the cavity. Lance commented on the tendency

1 Developmental Dysplasia of the Hip

for many unstable newborn hips to heal without treatment spontaneously in the normal position [93], and Putti recognized that “congenital dislocation can undergo spontaneous reduction” [172]. This spontaneous stabilization is the reason why many screening programs using ultrasound do not do the initial study until 1–2 weeks of age. Recent assessments of this phenomenon have better quantitated the finding. Molto et al. in Spain did a 3-year prospective study of 103 consecutive patients (137 hips) diagnosed with DDH at birth clinically with a positive Ortolani or Barlow sign [173]. No specific treatment was performed and the patients were reassessed at 2 weeks. When repeat clinical tests showed a stable hip, ultrasound was done to assure a concentric reduction. They noted that 73.8% of the hips had stabilized spontaneously by 2 weeks of age, with most stabilizing in the second week. Holen et al. in Norway made a similar observation; 68 of 99 newborn infants had spontaneously stabilized by 8–15 days (68.7%) with clinical and ultrasound assessment [174]. Mackenzie [175] and McKibbin et al. [176] noted spontaneous stabilization of clinical instability in 65% and 79%, respectively, by 3 weeks.

1.6.8 Absence of Ligamentum Teres in DDH Assessment of the extent of an absent ligamentum teres in DDH requiring open reduction was done by Li et al. [177]. They studied 150 hips in 123 patients; 28 hips (18.67%) demonstrated absence of the ligamentum teres, and 122 hips (81.33%) still had the ligamentum teres intact. There was a strong tendency for the absent ligamentum teres to be associated with severe deformation with 22/28 (78.57%) graded as Tönnis IV. Examination of the femoral heads in hips where the ligamentum teres was absent showed them to be extremely small with cartilage surfaces showing erosion – like changes.

1.7

ummary of Intrinsic and Extrinsic S Environmental and Pathoanatomic Findings in DDH: Discussion of Pathogenetic Sequences

Overview Structural variations from the normal, actual deformities and predisposing epidemiologic features leading to femoral head-acetabular malposition and their time of onset each play a crucial role in defining the extent of hip malformation at birth. Idiopathic developmental dysplasia of the hip is associated with certain predisposing epidemiologic features plus a variety of subtle structural abnormalities, some of which are tolerated without leading to instability,

1.7 Summary of Intrinsic and Extrinsic Environmental and Pathoanatomic Findings in DDH: Discussion of Pathogenetic Sequences

while others combine to lead to an unstable joint; and the ultimate severity of the disorder is dependent on the time of displacement with those occurring in utero predisposing to greater secondary change than those occurring during labor or postnatally with extension of the hip. The earlier in utero the imperfect relationship is established, the worse the secondary changes, and the longer that relationship persists postnatally, the further those changes worsen. Extrinsic mechanical theories of causation and epidemiologic characteristics appear to play a major although probably not exclusive role in predisposing to CDH/DDH. Chief among these are female infant, breech presentation, “dislocating posture” of hyperflexion, lateral extremity rotation and femoral anteversion, genu recurvatum, oligohydramnios, first birth (primigravida), excessive infant weight, twinning, genetic considerations, and ethnic group. The subsequent presentation follows the outlines of Salter’s approach to the pathogenesis of deformity in CDH (DDH) [112]. Capsular Laxity The primary pathoanatomic defect in developmental dysplasia of the hip is considered widely to be capsular laxity. This diminishes the restraint on the femoral head-acetabular relationship and allows the femoral head to sublux or dislocate laterally, posteriorly, and then superiorly in relation to the acetabulum. The capsular laxity in some descriptions is present throughout the hip capsule, but others have localized it initially to the posterosuperior region. With persistence of the dislocated position, the capsule enlarges universally, hypertrophies, and may undergo isthmic narrowing just above the rim of the original acetabulum. This narrowing is partly due to extrinsic capsular pressure by the iliopsoas tendon which must change its normal relationship as its insertion is pulled laterally and then superiorly by the displaced proximal femur and its lesser trochanteric insertion. Many dissections of newborn hips with this condition show both the acetabulum and proximal femur to appear to be structurally normal. Assessments at finer levels of resolution have however attributed subtle changes to underlie a tendency to displacement. The additive presence of two or three of these might lead to a slightly unstable articulation which then causes hip (femoral-acetabular) mobility stressing and stretching the capsule. Among these are the observations of Le Damany [21, 22, 32–34, 85, 86] and others on (i) slight degrees of proximal femoral anteversion combined with (ii) increased anterolateral acetabular obliquity, (iii) a relatively shallow newborn acetabulum, and Walker’s [121] observance of (iv) morphological changes in the labrum. In general, however, the acetabulum is normally shaped and positioned, the labrum is normally shaped and positioned, and the proximal femur is normal in size and shape and in particular does not show increased anteversion in relation to the contralateral normal side. Most of the changes that

57

develop in the acetabulum and proximal femur postnatally appear to be secondary to their growth in an unstable mechanical environment where the femoral head is not continually seated within the acetabulum owing to capsular laxity.

Acetabular Dysplasia As the femoral head develops in an imperfect relationship to the acetabulum, bone and cartilage development in both bones is retarded, and tissue deposition is abnormal. Acetabular dysplasia develops in which the upward slope of the lateral acetabulum persists and even worsens, whereas in the normal hip, the acetabular index diminishes, and the acetabulum deepens. It is important to think of the developing hip in a three-dimensional sense, especially since the various descriptive terms and radiologic indicators refer almost exclusively to a two-dimensional representation. Acetabular dysplasia, which is measured on plain radiographs as the acetabular index, defines an upward slope of the bony acetabulum, but the primary problem is a shallowness and anterolateral tilting of the cartilaginous acetabular socket. Slight anteversion of the acetabular socket, as initially pointed out by Le Damany, may also predispose to minimizing stability. With the altered pressure applied by the proximal femur against the lateral acetabulum, cartilage development is slowed, and bone deposition is secondarily slowed as well. Slowed acetabular cartilage growth appears to occur initially lateral to the triradiate cartilage. Proximal Femoral Dysplasia Abnormal development is also evident in the proximal femur with delayed appearance of the secondary ossification center and failure of postnatal diminution of the proximal femoral anteversion that is normally most extensive in the later stages of the third trimester just prior to birth. If the capsular abnormality leads to altered hip stabilization in a subluxed position, then the abnormal development is relatively slight, and the secondary ossification center on the involved side is present but somewhat smaller than that on the normal side. In completely dislocated hips, the secondary ossification center can often be delayed in appearance for several months compared to the normal side. The second aspect of developmental irregularity in the subluxed or dislocated proximal femur relates to persisting femoral anteversion. Proximal femoral anteversion is as great as 30–35° in the newborn and decreases at a fairly constant rate to 10–15° by early adolescence. Lower extremity positioning in utero that forces the femur into lateral (or external) rotation also effectively positions the head of the femur anteriorly and away from full reduction into the socket. If the hip is allowed to develop in the dislocated position, femoral anteversion persists and may even increase. It is not unusual therefore to find 30° or 40° of anteversion in

58

an 18-month-old child with a unilateral dislocated hip. The clinical correlate of this pathoanatomic development is increased internal rotation needed to seat the femoral head into the acetabulum with reduction maneuvers. A third aspect of proximal femoral dysplasia is an increasing loss of femoral head sphericity (misshapen head) especially with dislocation. Adductor Muscle Tightness As hip dislocation persists, the head remains lateral, posterior, and somewhat proximal to the acetabulum. The soft tissues react in a characteristic way. The adductor muscles tighten as early as 2 months of age, an occurrence that tends to hold the hip in the subluxed or dislocated position. With greater displacement and at later time periods, the iliopsoas and gluteus muscles are relatively shortened as well. Soft Tissue Deformation The medial capsule stretches and elongates to accommodate the displaced femoral head. As the head migrates laterally and superiorly, the capsule is pulled across the inferior segment of the acetabulum. The ligamentum teres remains attached to the femoral head but hypertrophies in both length and width. In long-standing complete dislocations, it is absent, having either ruptured or in the opinion of some been congenitally absent. This structure normally present in the fovea must now traverse a greater distance and comes to fill much of the dysplastic acetabulum. Significant changes occur in what is referred to as the radiolucent roof of the hip. This term refers to the lateral regions of the acetabular cartilage, the labrum, and the capsule. The position of the labrum is crucial to an understanding of the pathoanatomy of developmental dysplasia of the hip, and assessment of labral position is important to determine in considering the advisability of non-operative versus operative reduction in treatment. As the femoral head moves progressively further away from its normal position in the acetabulum, it pushes the labrum outward and upward. In the subluxation phase, the head pushes up against the labrum, but the labrum is still on top of and thus supporting the head. With even further displacement, the labrum is pushed against the capsule by the femoral head but still maintains its relation to the head. As the head completely displaces from the acetabulum, the labrum no longer lies on top of the head but rather slips between the cartilage head and the acetabular cartilage carrying with it part of the capsule which is adherent to its outer surface. The combination of the inverted labrum plus the associated flattened capsule is referred to as the limbus. The limbus is therefore not a normal anatomical structure but rather a pathologic structure composed of two normal anatomic structures in an abnormal relationship, firmly apposed to one another, and position. At this stage, the femoral head is supported only by the capsule that reacts by

1 Developmental Dysplasia of the Hip

increasing its size (enlarging) and thickness (becoming markedly hypertrophic). As the femoral head rides up against the capsule outside the acetabulum, some stability is provided eventually as the head rests against the outer wing of the ilium and forms what is referred to as a pseudo- acetabulum. By this stage of the developing pathoanatomic picture, it is not possible to relocate the femoral head into the acetabulum by closed reduction maneuvers. Adductor tightness limits the range of abduction, and the acetabulum has become filled with soft tissue structures that can no longer be pushed away to accommodate the femoral head. These include the inferomedial capsule which has been stretched across the inferior margin of the acetabulum as it is attached to the lesser trochanter, the ligamentum teres which has lengthened and hypertrophied as it remains attached to the displaced femoral head, increased fibro-fatty tissue (pulvinar) in the depths of the acetabulum occupying the space where the femoral head should have been, and the inverted labrum-capsule or limbus which further blocks normal positioning of the femoral head. False reductions can be achieved by closed means, and the adequacy of any reduction after the first few months of life should be checked by imaging other than plain radiographs. With mild to moderate complete dislocations undergoing closed reduction, the head usually sweeps the inferior capsule away from the joint surfaces although the inverted limbus prevents anatomic reduction superiorly. In severe dislocations treated after walking has begun, as was often the case several decades ago, the inferior capsule can also be inverted with closed reduction maneuvers. Idiopathic and Teratologic Dysplasia The terms idiopathic and teratologic in developmental dysplasia of the hip imply a clear-cut differentiation between two types of hip dysplasia, but in reality while the terms have some validity, they are far from being definitive. In clear-cut examples of either type, we can be comfortable that an accurate impression is left. In the straightforward idiopathic hip dysplasia, there appears to be sufficient evidence to conclude that capsular laxity is the offending occurrence (setting aside what causes the laxity), that the proximal femur and acetabulum are structurally normal initially, and that if diagnosis is made shortly after birth and the child is placed in a Pavlik harness or similar apparatus in which the hips remain flexed beyond 100° and slightly abducted, normal hip development will follow once capsular tightening has occurred. Indeed, 50% of dislocatable hips in the newborn will stabilize spontaneously in normal position without treatment within a week. At the other end of the spectrum are infants which are either stillborn or born with extensive connective tissue abnormalities involving the skeletal system who have displaced hips where stable reduction cannot be performed by closed means even

1.8 Natural History of Hip Dislocations, Subluxations, and Dysplasia

59

at birth and where either autopsy or direct examination at surgery show clear structural developmental abnormalities of the proximal femur and acetabulum. There are many patients, however, who fall between these two extremes such that the pathogenesis of the hip dysplasia is unclear. For example, there are normal appearing patients except for the dislocatable hip who undergo a seemingly uneventful closed reduction but who are found after several weeks to months to still have either a dislocatable or dislocated hip which subsequently is amenable to repair only by surgery. On the other hand, there are patients with clear connective tissue or neuromuscular abnormalities such as Ehlers-Danlos disease, skeletal dysplasia, myopathies, or even arthrogryposis who undergo uneventful closed reduction of the hip and Pavlik harness or hip spica treatment and soon develop a stable hip which is radiographically normal such that the teratologic etiology does not preclude a good result obtained by straightforward means. There are therefore two general possibilities in relation to the underlying pathogenesis of developmental hip dysplasia. The simplest concept indicates that the differences seen in several patients, and in particular those who appear otherwise normal, are related to the time at which displacement occurs. If there is malposition due to capsular laxity during, for example, the final 2 months of intrauterine life, then the secondary changes would be more extensive leading to the failure of treatment with simple closed reduction and flexion-abduction splinting. One is thus not dealing with a teratologic hip but rather with a relatively straightforward DDH that occurred several weeks prior to birth and was not amenable to early diagnosis and treatment. The second occurrence is that simple capsular laxity alone cannot be implicated as the sole etiologic agent in all cases, while other developmental abnormalities affecting the acetabulum, proximal femur, and associated soft tissue structures can be implicated and thus lead to the need for more complicated treatment protocols. In an effort to assess those individuals who fall between the two major extremes of idiopathic and teratologic, the use of imaging techniques plays a major role. These are done increasingly frequently postnatally in r elation to ultrasound primarily but also to MRI and CT scanning.

bone part of the acetabulum, delayed appearance and a smaller-sized secondary ossification center of the femoral head, and one which is more laterally placed than that on the normal side. The position of the secondary ossification center is either close to or on the other side of Perkin’s line that is a vertical line dropped from the outer bone margin of the acetabulum. There is a break in Shenton’s line which is a continuous curvilinear relationship involving the inferior neck of the femur and the inferior surface of the superior pubic ramus at the obturator foramen. The longer the femoral head remains outside the acetabulum, the more misshapened it becomes, and the less uniform the fit between the two when reduction is finally achieved. Table 1.2 summarizes the pathogenesis of developmental dysplasia of the hip.

Worsening of Secondary Changes with Time Secondary changes develop in the postnatal period and are increasingly more marked the longer the hip remains subluxed or dislocated. A dislocated hip shows diminished abduction with a prominent adductor muscle-tendon band felt on abduction. This adaptive change can be seen as early as 6–8 weeks of age, becomes progressively tighter, and may ultimately prevent the performance of relocation by simple clinical manipulation. Radiologic correlates increasingly demonstrate acetabular dysplasia with an increased acetabular index which is an indication of diminished development of the

1.8.1 N atural History of Complete Dislocations

Late Developmental Dislocation of the Hip After Initial Normal Examination Occasional but seemingly well- documented cases have been reported of late hip subluxation or dislocation after clinical and radiographic neonatal assessments were normal in the first 3 months of life. Raimann et al. reported five cases of otherwise normal children with this finding, four of whom required open reduction, which they referred to as “late hip dislocation” [178]. Such cases should be documented in the literature. It remains important to rule out connective tissue disorders and, at the early stage, subclinical neuromuscular disorders such as relatively mild myopathies. “Primary acetabular dysplasia” is also suspected. Although the hip remains stable and located, the acetabular index fails to reach normal configurations even with prolonged abduction splinting. The other evidence supporting this entity is the finding of premature hip pain in young adults, not recognized as having had a DDH but presenting with Wiberg’s CE angle 45° [179] in their teenage years.

1.8

atural History of Hip Dislocations, N Subluxations, and Dysplasia

There have been a limited number of studies over the past few decades following patients with untreated unilateral or bilateral dislocated hips well into their adult years. In a complete dislocation, the opinion appears well grounded that pain is a relatively infrequent occurrence, in particular in the first three decades of life. Patients are sometimes described as doing well with an untreated complete dislocation although these references refer to the minimal to absent dis-

60

comfort and the fact that the patients can walk independently and function in society. The instability and waddling gait limit rapid running and competitive athletic pursuits. These long-term studies defined certain occurrences clearly. Discomfort can be present in the lower back with the associated lumbar lordosis and in the hip on the dislocated side. Generally, however, the onset of discomfort is in the mid- fourth to mid-fifth decades. Wedge and Wasylenko reported on 32 patients with 42 complete dislocations between 16 and 86 years of age [180]. When a modified Harris hip score was applied to their assessment, 41% of the patients had a good rating, 14% were fair, and 45% had poor clinical ratings. The presence or absence of a false acetabulum played a major role in influencing the clinical status. Wedge and Wasylenko demonstrated only a 24% chance of a good clinical result with a well-developed false acetabulum, while in those with a moderately developed or absent false acetabulum, the clinical rating improved to 52%. A false acetabulum, which by definition would also imply imperfect positioning of the head in relation to that acetabulum, would lead to wear and tear changes on the cartilage and subsequent discomfort. It appears better from a clinical sense to have no false acetabulum since the cartilage degeneration and bone eburnation phenomena would not occur and it is these latter two occurrences that lead to discomfort. Some have felt that in patients with bilateral dislocations, the low back pain is greater owing to the more marked compensatory hyperlordosis, while in patients with complete unilateral dislocations, discomfort relates to limb length inequality with ipsilateral knee deformity and pain, scoliosis, and more marked gait disturbances. The flexion-adduction hip deformity on the involved side is compensated by a valgus knee that in itself leads to increased knee stresses and discomfort.

1.8.2 N atural History of Dysplasia and Subluxation The worst position in terms of development of symptomatic discomfort is the subluxed position since the patients remain ambulatory from the early years of life but the wear and tear on the femoral head cartilage and outer acetabulum is great. In those with dysplastic hips with the relatively mildest degree of acetabular malformation, symptoms of degenerative joint disease can occur but generally in mid- to late adult life (Fig. 1.5). The fact that subluxation of the hip is the position which primarily leads to the development of degenerative joint disease and clinical symptomatology was also stressed by Wedge and Wasylenko who noted that only 42% of the 38 subluxed or dislocated hips had a good clinical rating [181]. The degree of subluxation also appeared to correlate roughly with the symptoms. The most severe subluxations led to onset of symptoms during the second decade of life, those with moderate sublux-

1 Developmental Dysplasia of the Hip

Fig. 1.5 Radiograph illustrates an adult hip with osteoarthritis following imperfectly treated developmental dysplasia of the hip of childhood. Note the persisting acetabular dysplasia, lateral acetabular subchondral bone sclerosis, relatively spacious acetabulum, femoral head riding laterally and upward, joint space narrowing, and osteophyte formation

ation presented during the third and fourth decades, and those with minimal subluxation had symptoms even later.

1.8.3 O steoarthritis in Adult Life Following Childhood CDH/DDH It is now widely recognized that any result less than anatomic restoration of the hip in those with CDH/DDH predisposes to osteoarthritis in mid to late adult life. This observation took some time to evolve into clinical certainty however. Over the past few decades, the diagnosis of primary osteoarthritis is being made less frequently based on a more careful assessment of the adult radiograph which increasingly is interpreted to indicate the sequelae of previously unrecognized childhood hip disease. In a report in 1933 Putti summarized his observations on the goals for successful

1.9 Brief History of Treatment Approaches in Developmental Dysplasia of the Hip

treatment of congenital hip dislocation [172]. “It is a complete delusion that one can have a result permanently satisfactory in function in a hip incompletely reduced. Every such hip that has not obtained from the first, or not preserved, normal anatomical relationship between the femoral epiphysis and the acetabulum is inevitably destined to become the subject of that precocious articular senility which is usually diagnosed as osteoarthritis. No complete and permanent restoration of function occurs apart from perfect anatomical reduction.” Morville (1936) showed excellent awareness of the fact that osteoarthritis of the adult hip was frequently and actually almost always preceded by structural hip disorders of childhood [182]. He analyzed 100 cases of osteoarthritis of the adult hip (referred to as arthritis deformans). In only 16 of the 100 had the arthritis developed in an anatomically normal hip joint, and even in 12 of these, an associated cause was noted being either arthritis of childhood, trauma, or infection. The remaining 84 cases showed anatomically abnormal conditions. He divided these into two types with 38% representing congenital hip subluxation and 46% representing subluxations due to acquired hip disorders the two commonest of which were Legg-Perthes disease and slipped capital femoral epiphysis. In those arthritis cases secondary to congenital hip dysplasia, the acetabulum was steep, and the femoral head was subluxated upward and outward which was noted radiographically by interruption of Shenton’s line. The deformity had undoubtedly developed on the basis of the congenital flat acetabulum although it was not until the age of about 30–50 years that the arthritic changes began. In the other group of predisposing hip disorders, the acetabulum was broader than normal but not steep leading to primarily a lateral subluxation of the head. He indicated “it is undoubtedly the deformity - the subluxation - that is primary but for years giving no symptoms until the arthritic changes occur in a secondary fashion.” Since so many of the arthritic conditions in the adult hip developed on the basis of childhood hip disorders, there would be great value in improving treatment of the latter. In relation to congenital hip dysplasia, he felt that the indications for treatment toward an anatomically normal joint should be enhanced. Wiberg (1939) was also one of the early investigators to document definitively that a subluxation persisting from childhood would in mid-adult life frequently lead to osteoarthritis of the hip [183]. He presented 18 cases with x-rays from 15 to 30 years following treatment of a childhood congenital subluxation of the hip where osteoarthritic deformities evolved with typical changes in the femoral head and acetabulum. Murray (1965) assessed 200 radiographs of osteoarthritis and determined that only 35% were truly idiopathic with 65% of cases being due to underlying childhood hip disorders the main ones being a persisting acetabular dysplasia in 25% of the cases and 40% showing a tilt defor-

61

mity felt to be consistent with minimal slipped capital femoral epiphysis [184]. A study by Gade (1974) in Norway in 123 operated cases of hip osteoarthritis had shown an almost 50% incidence of acetabular dysplasia in adult osteoarthritis with only 24% being primary [185]. Harris (1986), an adult hip surgeon in Boston, estimated that more than 90% of patients with “primary or idiopathic” osteoarthritis of the hip clearly showed demonstrable abnormalities of the hip joint which would have been present at the cessation of growth (in those patients where sufficient radiographs were available for assessment) [180]. The disorders involved acetabular dysplasia and what he defined as the “pistol-grip deformity” (deformed femoral head-neck configuration) associated with mild slipped capital femoral epiphysis, Legg-Perthes disease, multiple epiphyseal dysplasia, spondyloepiphyseal dysplasia, or a malpositioned intra-acetabular labrum (many of these disorders unrecognized at time of occurrence). When these abnormalities are considered along with other known precursors of osteoarthritis of the hips such as metabolic (e.g., hemochromatosis) or inflammatory (e.g., rheumatoid arthritis) disorders, it became clear that osteoarthritis of the hip as a primary disease either did not exist or was rare. More briefly stated, a large portion of osteoarthritis previously not felt to be associated with pre-existing hip disease actually develops from childhood hip developmental abnormalities. Primary osteoarthritis of the hip remains a recognized disorder, however, considered to be genetic in origin and occurring in 3–6% in populations of European descent [186]. The primary variant is virtually absent however in non-European populations, being rare throughout Asia and Africa. Primary osteoarthritis is defined as occurring after the age of 55 years and specifically excluding underlying childhood hip disorders, femoral-acetabular impingement, osteonecrosis, trauma, sepsis, rheumatoid arthritis, and other defined causes.

1.9

rief History of Treatment B Approaches in Developmental Dysplasia of the Hip

The four principal aims of treatment for a developmental/ congenital dysplastic hip are straightforward. They include: (i) Reducing the hip anatomically so that the femoral head is normally seated in the acetabulum without the interposition of soft tissues such as the inverted labrum or capsule. This can be achieved by either closed or open methods. (ii) Maintenance of the reduction for a sufficient period of time to allow tissue reconstitution to stabilize the hip such that loss of position does not occur when the hip is no longer immobilized.

62

(iii) Development of a structurally normal proximal femur and acetabulum and a snug capsule. If these do not develop with growth following relocation, surgical methods involving proximal femoral derotation and varus osteotomy, acetabular augmentation or redirection procedures or capsulorrhaphy may be needed. (iv) Avoidance of avascular necrosis of the femoral head.

1.9.1 G radual Development of Reasonably Effective Closed and Open Treatments 1.9.1.1 Evolution of Treatment Approaches in the 1800s and Early 1900s Congenital dislocation of the hip was placed on a firm scientific setting by the works of Palleta [56, 57] in 1788 and 1820 and Dupuytren [59] in 1826. For the next 50 years, little headway was made in successful relocation and maintenance of relocation for dislocated hips. Good results were difficult to obtain owing to the late ages at diagnosis, almost always after walking had begun, and the imperfect means of reduction and immobilization. Several attempts at closed reduction by traction in extension were reported in France in the 1820s and 1830s, but early efforts failed once traction was released and walking resumed. Charles-Gabriel Pravaz of Lyon, France, is credited with first successfully reducing and maintaining congenital dislocation of the hip by closed means in 1835 using a strong recumbent traction apparatus with the hip and lower extremity in extension. The patient was a male 7 years of age treated by Pravaz and considered cured 2 years after the reduction by a committee of several prominent physicians in France who reviewed the case in 1838. Their discussion, however, which included descriptions of patient gait and physical examination made it evident that they did not consider the result to have led to a perfectly normal hip. The nature and length of the treatment, however, were extremely complicated and prolonged, generally lasting more than 1 year. Both Pravaz [61] and Carnochan [63] outlined and illustrated the treatment in detail. This included (i) placing the child in a mechanical apparatus using pulleys, weights, and ropes to apply continuous bi-directional traction with the hips in extension to pull the displaced femoral head down to the level of the acetabulum. The patient was recumbent with gradually inclined positions built into the apparatus with this phase of treatment taking 8–10 months; (ii) reduction was then performed in extension with the limb abducted in traction, the hip manually manipulated, and hip position stabilized with a padded thigh piece of the apparatus including pressure against the greater trochanter; (iii) this position was then maintained in recumbent posture, for several months using the padded, hinged metal plates holding the thigh/hip region to maintain reduction to allow the body

1 Developmental Dysplasia of the Hip

“by the plastic effect of the organism” to construct a stable articulation; (iv) a wheeled carriage was then used to continue non-weight bearing but allow for range of motion exercises; and (v) a wheeled supportive walker was used to allow the patient to be upright and undergo gait training with gradual weight-bearing added progressing (via eventual crutch use) to full activity. While assessment of any result was done by clinical examination only in this era, a reasonable determination of position could be made by such criteria as hip range of motion, limb lengths, bony landmarks, and gait assessment. Pravaz’ contributions demonstrated the possibility of curing what had been considered an incurable deformity. The procedure also followed principles still considered essential today, namely, bringing the head of the femur to the level of the acetabulum, reducing the head into the acetabulum by abduction, maintaining the reduced position continually until the bones and soft tissues remodeled biologically with growth in the correct position so that stabilization would be maintained when treatment was discontinued, range of motion exercises after prolonged immobilization in a protected non-weight-bearing fashion, and gradual resumption of weight bearing and walking. Pravaz indicated that “it is possible by continued and progressive mechanical action to restore sometimes the head of the femur to the rudimentary cotyloid (acetabular) cavity which it had left before birth and to keep it in this situation where, by the plastic recuperative efforts of the organism, assisted by the performance of certain proper movements resembling those of walking, it (the head of the femur) will at last form for itself a sort of artificial joint” [61]. Guérin: Subcutaneous Release of Contractures At the same time that closed reduction in traction was increasingly attempted, Jules Guérin, a prominent surgeon in Paris, developed his theory of musculoskeletal deformation being caused by spasmodic muscular action induced by primitive disorders in the nervous system [187]. Lesions in the lower spinal cord were considered to correlate with lower extremity contractures and dislocations, those in the upper spinal cord with deformities of the upper extremity, and those in one-half of the brain causing deformities on one side of the body. He felt that the position and extent of the neurologic lesion correlated to a great extent with the position, extent, and direction of the musculoskeletal deformity and dislocation. Congenital dislocation of the hip was thus caused by muscular spasmodic action of the hip region muscles. Guérin acted upon his theory by performing subcutaneous surgery for release and correction of the contractures. He indicated that the muscles affected were pathologically shorter than normal and converted to hard deforming structures composed of fibrous tissue rather than muscle. He considered his theory to be universally applicable throughout the body causing clubfeet, scoliosis, torticollis, upper extremity deformities, as well as

1.9 Brief History of Treatment Approaches in Developmental Dysplasia of the Hip

congenital dislocation of the hip. He treated CDH with multiple hip region subcutaneous tenotomies and myotomies as well as scarification in the vicinity of the cotyloid cavity to induce the iliac periosteum to produce an “effusion of plastic ossifiable matter” to create a deepening tissue support structure for the femoral head. He reported releasing the following tight muscle contractures as needed for CDH: long adductor, sartorius, rectus femoris, psoas, iliacus, and glutei as well as the coxafemoral ligament. He also applied traction and some of the other current modalities, but the surgical releases were considered to be the main therapeutic measures. He also treated cases of clubfeet and scoliosis with his subcutaneous tenotomy/myotomy releases. Guérin was a powerful figure in the French medical professor of his time serving as surgeon, professor, writer, and medical journal editor but also had many detractors who felt that his theories were unsupported by scientific evidence and his surgical procedures were not sufficiently effective to be applied throughout the body to correct the entire range of deformities. While clearly effective in many cases of clubfeet, his approach to the hip did serve as a starting approach to treatment of the myotendinous contractures that were making relocation of the dislocated hip so problematic. At the end of the 1800s and into the early 1900s, congenital hip dislocation (as it was commonly referred to then) was actively treated by surgical open reduction before problems with that approach in that era led to a reconsideration of closed means. The works of Poggi (1888) [188], Hoffa (1890, 1892) [189, 190], Lorenz (1892, 1895) [191, 192], Burghard (1903) [193], Bradford (1894, 1904) [194, 195], and Ludloff (1913) [196] will be described below to show the evolution of these approaches. Closed reduction to treat hip dislocation was used by others, however, and the two approaches were widely used until a consensus emerged. Paci in Italy (1887) [197] and Lorenz in Austria (1895) [198] developed closed reduction into an effective treatment. Paci, using more gentle traction, appeared to improve position without succeeding in bringing about a true reduction and stabilization. He did not abduct the limb in an effort to maintain position. Lorenz set about to fully reduce the hip but used extensive physical force to do so since the tight muscles and ligaments needed to be stretched to remove resistance to reduction. The use of force was such an accepted part of the treatment that it was often referred to as “forcible reduction.” Regarding descriptive terminology for this approach, what we now refer to as “closed reduction” was, in the late 1800s and early 1900s, referred to as the “bloodless method,” the “forcible reduction,” or the “manipulative method.” Lorenz then stabilized the closed reduction position of hip flexion at 90° and wide abduction to 90° with hip spica immobilization followed by abduction splinting which at the time revolutionized the therapy for CDH. One-stage

63

reductions under anesthesia, using considerable manual force and stabilizing the hip in wide abduction in spica cast, comprised the Lorenz method. The patients were generally 3–10 years of age; each of the flexor, adductor, abductor, and extensor muscle groups was forcibly removed from their bony origins by manual pressure and stretching so that the limb was flail-like before the actual reduction was able to occur. Although spica casting in the Lorenz position involving hip flexion of 90° and full abduction of 90° has subsequently been shown to lead to a high incidence of avascular necrosis (and is no longer used), his approach was considered of great importance in the treatment profile for CDH since he was able to reduce the head into the socket and keep it there sufficiently long such that ultimate stability ensued [199]. Lorenz published his major work in 1895 and continued to practice, write, and speak on it for the next quarter century, publishing a lengthy treatise on CDH in 1920. He stressed not only the reduction of the hip but also maintenance of reduction until stability followed. He actually developed his approaches to closed reduction based on an extensive experience with open reduction, switching to closed means as a primary approach after several postoperative deaths and infections and the occurrence of considerable rigidity and contractures. Tubby has outlined the methodology for manipulative treatment in great detail in the second edition of his textbook [200]. The two chapters on congenital dislocation of the hip remain of value today, especially regarding pathoanatomy and clinical findings. Bradford and Lovett, 1915 By this time the evolution of treatment described by Bradford and Lovett was that “in a majority of cases manipulative reduction under an anesthetic is the method to be employed” [201]. The manipulative reduction is described and illustrated in great detail, reminding us a century later just how physically forceful the procedure was at that time. The treatment was “based upon the plan of stretching the contracted soft parts, muscles, capsule and ligaments, so that the head can be forced successfully through the distorted capsule into the socket.” Several of the photographs show the physician and assistant each with both hands on the pelvis and lower extremity applying the traction and countertraction needed. We realize the difficulty and force applied to the procedure by the terms used in the description: “the adductor muscle group should be overstretched,” “the limb should be rotated forcibly to both the outer and inner side, and then forcibly abducted both with the knee flexed and straight,” and “after the limb has been brought to nearly a right angle with the thigh, the knee being straight, (further flexion of the hip is done) until the thorax is almost touched by the front of the thigh, thus stretching the hamstring muscles.” Reduction is then attempted with the lower extremity (flexed at knee) flexed and abducted strongly with countertraction applied on the crest of the ilium with

64

pressure to aid reduction on the trochanter. The head is thus pressed into the acetabulum by manipulation “after all contracted tissues are relaxed by overstretching.” It was recognized early that a possible complication was the folding of the capsule into the joint between head and socket. A key maneuver indicates “forced abduction stretching the adductors with blows upon the adductor attachment.” For children of a larger size, reduction was felt to be more effective when assisted by mechanical force applied for traction and countertraction using a specially constructed apparatus. It was felt that the shape of the bone (femur) would be altered by growth if complete reduction has been established. A well-padded plaster hip spica was applied to hold the reduction using Lorenz criteria of flexion of the hip to a right angle and wide abduction. Time in cast varied from 2 to 6 months followed by an intensive physical therapy regimen. The authors reported 80–90% cures between 2 and 10 years of age. Redisplacement led to an occasional need for second reduction. The list of complications was severe at times: fracture of the femoral head, rupture of the femoral artery, temporary and permanent regional paralysis, and occasional death from shock. Bennett (1908) wrote that Lorenz was reporting approximately 50% “good anatomic results” in follow-up of 680 hips [88]. Others used what came to be described as the Lange position for cast immobilization after closed reduction. Forcible longitudinal traction was applied for reduction following which the hip was immobilized in a plaster hip spica with the hip in extension, and thus flexed much less than 90°, but still in full abduction and maximum medial (internal) rotation to compensate for proximal femoral anteversion and fully position the femoral head deeply into the acetabulum (Lange position) [202]. Over the past few decades, the position of hip immobilization in spica casts after reduction, as defined (separately) by Salter [203] and Fettweis [204] and referred to as the “human position,” is hip flexion slightly beyond 100° with abduction 45–50° to protect femoral head blood supply. By the 1880s it was evident that traction methods whether or not augmented with subcutaneous tenotomy/ myotomy were almost always ineffective in fully reducing congenital hip dislocations. Over the next two to three decades, improved open operative and closed manipulative methods evolved, although the operative approach took precedence early on. Open reduction was attempted in several centers but that generally fell into disuse as well primarily due to high surgical morbidity, including death, as well as for less than excellent results. Improved closed reduction methods holding position by casting was being developed as well and also came into wide use. In 1888 Poggi succeeded in replacing the congenitally dislocated femoral head in a 12-year-old female by open reduction followed by 50 days traction in extension and gradual re-

1 Developmental Dysplasia of the Hip

ambulation with crutches [188]. The pathoanatomy was well described including the enlarged thickened capsule, the absent ligamentum teres, the narrow capsular isthmus overlying the small acetabulum, and a small, deformed non-spherical head. Poggi sectioned many muscles to overcome obstacles to reduction. Following enlargement of the capsular foramen with longitudinal incisions, the true acetabulum was uncovered and found filled with fibrous tissue and the remnant of the ligamentum teres. “I deepened the acetabulum, reshaped the deformed head giving it its proper contour, and by means of well directed traction and incisions in the capsular ligaments, I was able with only little difficulty to replace the head into the newly deepened acetabulum.” Closure was associated with removal of excess capsule and its repair. Poggi did not subsequently do many procedures beyond his initial description. Hoffa [189, 190] in Germany and Lorenz [191, 192] in Austria played major roles in developing the techniques for operative open reduction for CDH. The method of Hoffa involved curettage of the acetabulum to deepen it but also extensive release of the surrounding soft tissues including multiple tenotomies (open and closed) as well as subperiosteal release of the muscles attached to the greater and lesser trochanters and opening of the joint capsule. He limited the procedures to younger children, excluding adolescents and older age groups. He noted that it was the iliopsoas tendon passing over the capsule that restricted relocation of the femoral head into the acetabulum. Lorenz’ procedure was less involved as he left the muscles intact and, except for the hip adductors, concentrated on opening the capsule with a “T” incision and reducing the head into the socket. Lorenz recognized that the acetabular socket existed and was often sufficient to accept the femoral head. By 1896 an outline of approaches used to treat a congenitally dislocated hip is recognizable today. Bilhaut outlined a seven-step approach based on recognition that the capsule was too narrow to allow for reduction of the femoral head into the acetabulum, primarily due to external compression by the iliopsoas and by the fibrinous thickness in the inferior part of the capsule running parallel to Bertin’s ligament (iliofemoral ligament, Y-ligament of Bigelow) [75]. These steps included (i) manual reduction as soon as diagnosis is certain, as early as possible, followed by immobilization for 4 months in plaster of Paris cast; (ii) if reduction unsuccessful, apply weights (traction) in extension, recumbent position with lower extremity abducted for several weeks to elongate the muscles followed by manual reduction under (chloroform) anesthesia; (iii) if reduction is achieved but not maintained in good position, resort to open reduction surgically to increase the dimensions of the capsule; (iv) consider reducing the size of the capsule (presumably capsulorrhaphy to retain reduction); (v) if still difficult to maintain reduction, resort to sub-

1.9 Brief History of Treatment Approaches in Developmental Dysplasia of the Hip

trochanteric osteotomy; and (vi) recognition that some cases are too deformed with further surgery contraindicated. Burghard was one of the proponents of open reduction in England – describing approaches similar to those used today [193]. Bradford in Boston, USA, used open reduction as well, favoring a more limited muscle-sparing approach, opening the capsule, widening the narrowed portion, repositioning the head into the joint, and then repairing the capsule (capsulorrhaphy) [194, 195]. He felt that the chief obstacle to reduction was the iliofemoral bands of the capsular ligament (Y-ligament of Bigelow) whose division markedly aided closed reduction. Capsular repair was performed with attempts made to suture the capsule around the head to form, in effect, an artificial cotyloid ligament. The acetabulum rarely needed to be deepened by curette, but adductor muscle release and fascial releases (iliotibial band) were generally done. Femoral derotation osteotomy was done (subtrochanteric) if anteversion was greater than 60°. Ludloff used the medial hip approach for open reduction [196] but applied a plaster cast postoperatively in the Lorenz position. The complications of surgery of that era, including sepsis, were steep however, and closed reduction was again used increasingly.

1.9.1.2 Overview of Approaches Based on Evolving Closed and Open Hip Reduction By 1900 there was sufficient experience with treating CDH by closed and open methods that recommendations for various approaches were expressed. The three methods of treatment were recognized as (i) continuous traction in extension (or recumbent traction) on specially constructed traction/ countertraction frames as originally practiced by Pravaz, Guérin (with multiple tenotomies), and subsequently many others was no longer considered as effective, (ii) open reduction (often called “reduction by incision”) with adherents to both the Hoffa technique or the Lorenz technique with many surgeon specific variations (Tubby, Bradford), and (iii) closed reduction (manipulative reduction) with the Lorenz method with modifications widely used. Treatment was prolonged with pre-procedure traction recommended even in surgical cases to reduce the head of the femur to the level of the acetabulum, and after reduction stabilization of several months in recumbent traction or hip spica cast was needed to allow for remodeling. Similarly, extensive time was needed to remobilize the child to full weight-bearing status to minimize chances of re-dislocation. The discovery and clinical application of radiography in 1895 allowed for clear evidence of the effectiveness of therapy as it progressed. It was still recommended however by many (Bradford and Lovett) that “no attempt at reduction is advisable under two years of age as the tissues are not sufficiently developed to prevent relapse” [201]. In early cases, defined as 2–5 years of age, it was considered that reduction could be “easily accomplished

65

by forcible manipulation.” In resistant or older cases where a firm narrow hourglass capsular constriction and/or alterations of femoral head/acetabular shape existed, they always recommended manipulation first (with mechanical apparatus used) followed by open reduction if that failed. Gradual traction was applied preoperatively to pull the head to the level of the acetabulum. The controlled approach could be appealing to some surgeons. The surgical dissection toward the femoral head was done “with as little injury to muscle as possible.” The capsule was opened, the capsular constriction divided, ligamentous and fibrinous bands cut, head reduced, and redundant capsule repaired after which a hip spica was applied to hold the position.

1.9.1.3 Early Reviews of Treatment Results Results improved gradually in reports collected over 20–25- year intervals. In the early 1900s, detailed reports began to be published although criteria for assessing results were variable. It was after these reports that more careful assessment began to be performed since there was widespread recognition that anatomic remodeling to complete normalcy was not necessary for functional exams in childhood to be reasonably good. A detailed report by Stern (1906) examined the world literature on results of bloodless (closed) reductions of CDH [205]. The author acknowledged the limitations of such studies, but interesting information was still obtained. He assessed reports from 39 surgeons. The number of hips available for follow-up assessment was 2593. There were 7 physicians with 100 or more cases (1835 cases) that accounted for 71% of the hips studied including Lorenz (680) and Hoffa (380). (Although the article referred to joints successfully “operated” the procedures were manipulative with “operated” appearing to mean treated.) Results were categorized as “ideal” which could be either perfect anatomical restitution (limbs of equal length, lordosis and waddling gone, excellent range of motion of joint, and x-ray showing the femur normally positioned within the acetabulum) and/or a perfect functional result (even if there were some clinical/x-ray abnormalities); “very good and good” where transpositions are seen, referring to only partial anatomic correction of the head which comes to lodge securely against the pelvis in the periacetabular region anterior or posterior to the acetabulum on its upper rim but function of the hip joint is improved (in relation to stability, ability to walk long distances, and marked improvement of the Trendelenburg limp); or “bad” with re-dislocation, relapses, failure of any improvement at all, or complications such as paralysis, fractures, or worse all leading to failure to obtain good function. Understandably, as Stern indicates, he needed to place the numbers from the many papers reviewed into some categorizable format so some (about 1%) inaccuracy is present in the figures. He noted however that summation of the findings was about the same as reported by Lorenz from his extensive

66

studies. Of the 2593 hips assessed, ideal results were seen in 49% (either anatomical restitution 1084, 42% plus excellent functional results 187, 7%); good/very good with structural transpositions 1036, 40%; and bad results with poor femoral head position and failure to obtain good function 314, 12%. With our current level of understanding, even most of these results would be expected to deteriorate, since the patients were only followed for 1–3 years but in view of what was considered only a few decades previously to be an incurable disorder, the overall improvement was promising. When functional results alone were considered as good or bad, 89% of the 2593 were good, while the ideal results were 49%, of which 42% were anatomical reconstitutions. Reports on Results Kirmisson of Paris reported anatomical repositioning with perfect to good results in 11/28, 39%, hips with unilateral cases but only 2/18, 11%, favorable in bilateral cases [200] (see Tubby). Lorenz reported 63% cures in 572 cases in 1905 [200] (see Tubby). In the review by Stern discussed above, he lists 680 cases for Lorenz: 401 (59%) anatomical restitution (excellent, ideal), 251 (37%) good functional with transposition (improved but imperfect reduction), and 28 (4%) bad (re-dislocation, poor function) [205]. Hoffa reviewed 315 of his closed reduction (bloodless) cases in 1905 of which 250 were unilateral and 65 bilateral [206]. Of the unilateral cases, 75 were perfect anatomic reductions (30%), 160 anterior transpositions (improved position but not fully relocated, 64%), and 15 re-dislocated (6%). In a functional sense, 100/160 cases were improved; when combined with the perfect anatomic reductions, therefore, 185/250, 75% had a satisfactory functional outcome with 25% considered failures. Bilateral results were much less favorable with only 5/675, 7.7% perfect/ideal reductions, 32 (49%) transpositions (partially reduced), 10 (15.4%) with one side good but the other poor, and 18 (27.7%) re- dislocated. Bradford updated his results in 1909 [207]. Assessing cases done from 1896 to 1908, there were 210 hips treated. Results were reported as “cure” in 79/154 (51%) cases treated by manipulation and 27/56 (48%) by open reduction. The overall rate of cure was thus 106/210 (50%). When the results were assessed by years starting from 1896 forward, there was a steady improvement of results. Separate reports by Bradford et al. on cases treated at his unit showed that the two primary methods of treatment involved open reduction and closed manipulative reduction [195, 207]. The earliest open procedures, which utilized the Hoffa technique in which the acetabulum was deepened with a curette, were ineffective. His group noted that an acetabulum sufficient to receive the femoral head was always present when open surgical approach was done. Subsequent utilization of the Lorenz open approach resulted in much improved results. The muscles of the hip region were spared, the capsule was freed, and any constrictions were divided, and the capsule

1 Developmental Dysplasia of the Hip

was then repaired following reduction of the femoral head. As the techniques evolved, reduction was improved by dividing specific tightened structures around the hip such as the adductor muscles, repairing the enlarged capsule to provide support for the reduced head, and in some instances performing a proximal femoral derotation osteotomy to correct femoral anteversion in particular when it was 60° or greater. Many cases were treated with manipulative reduction using the Paci-Lorenz method. It was recognized that inversion of the distorted capsule into the acetabulum in front of the reduced head could take place during closed reduction, a finding that led many to perform the open procedure. Unsatisfactory cases were considered to be due to imperfect reduction with folding of the capsule in front of the head of the femur, uncorrected persisting femoral anteversion, and defects of the acetabulum. They recognized that “for successful treatment of cases of congenital dislocation of the hip after reduction it is necessary that the capsule should fit closely and not loosely around the head.” Lorenz subsequently indicated that the best results were obtained when reduction was done between the ages of 2 and 3 years, bearing in mind that he did not treat those children less than 2 years of age. He reported on a large number of hips, 1057, in which good to excellent results were obtained in 57% of unilateral cases and 53% of bilateral cases [208]. Criteria for good results involved prevention of upward slipping of the femoral head by an appropriately shaped acetabular roof, free mobility of the joint, and excellent muscle function. Lange reported a large series involving 2200 reduced hips over a period from 1904 to 1925 [209]. Results improved in the later time periods with an indication of 22% good results from 1904 to 1914 that increased to 63.7% between 1915 and 1925. The definition of anatomic healing was somewhat vague by today’s standards but involved good function, good mobility of the joint, and x-rays showing no dislocation, subluxation, or serious deformities of the femoral head and neck. Putti published a report on 523 cases involving closed treatment from 1899 to 1927 [210]. His detailed categorization graded anatomic and functional results from 0 to 10. He showed good anatomic results (grades IX or X) in 34% and good functional results in 40%. There were also improvements during the later time periods such that in the latter group, from 1921 to 1927, good anatomic results (marks of 9 and 10) were seen in 37% of bilateral cases and 59% of unilateral cases, while the highest-grade functional results were seen in 40% of bilateral cases and 68% of unilateral cases. Indications for open reduction became clearer with Deutschlander who indicated that surgery remained warranted when closed reduction was unsuccessful and in particular when there were hindrances to reduction such as a long ligamentum teres or capsular interposition [90]. Many

1.9 Brief History of Treatment Approaches in Developmental Dysplasia of the Hip

surgeons continued to resort to open reduction when closed reduction failed. Galloway [211] was one of the earliest in North America to favor the method, while Farrell and Howorth [109] described 122 procedures from their unit in 1935. The forceful measures of reduction plus the extreme positions of hip immobilization in plaster spicas led to severe sequelae attributable to avascular necrosis. At fault in the latter regard were two positions of immobilization: the Lorenz position (maximal 90° abduction and flexion of 90°) and the Lange position characterized by full hip abduction in extension and marked internal rotation. Once the closed and open methods had at least enabled the head to be repositioned in the socket, attention was then directed (i) to achieving a structurally and functionally normal hip which involved anatomic reduction of the femoral head into the acetabulum without capsular or soft tissue interposition and normal structural shapes of the two bones and (ii) to preventing avascular necrosis as a complication of therapy.

1.9.1.4 Mid-twentieth-Century Overview of Results (a) Leveuf. Leveuf summarized the long-term results being achieved with closed and open reductions of congenital hip dislocation [212, 213]. His interpretation constantly stressed the importance of understanding and demonstrating the underlying pathoanatomic features. (i) Closed Reductions. Leveuf performed an extensive review of the results following closed reduction by analyzing 602 cases from major clinics throughout France with a 10–40-year follow-up [212]. He also reviewed extensively the literature on closed reductions from other countries. He stressed the importance of assessing not only the functional result but also the anatomic result and clearly documented that the long-term result was in fact dependent on the anatomic reconstitution of the hip since the functional results during childhood and adolescence were almost invariably good. He clearly demonstrated that with increasing passage of time, many results originally graded as good and excellent became fair to poor with progressive development of arthritis in the affected hips. His analysis did not attribute the better results in the more recent cases to improvements of technique but indicated that less than perfect anatomic reconstitution early on invariably led to worsening over the succeeding decades. This finding was strengthened by results from several clinics where the same technique by the same individuals had been used over several decades. In one clinic he noted that good results from 10 to 14 years were seen in 58% of the patients;

67

from 15 to 20 years, this had diminished to 30% of the patients; and from 21 to 32 years, this had further diminished to 22% of the patients. Other groups with comparable statistics showed a diminution of good results from 48% at 10–27 years, to 31% at 15–30 years, to 25% at 15–40 years. The long-term prognosis was dependent not on the short-term clinical profile but rather on the anatomic profile as revealed radiographically. It was only the good and excellent results anatomically that lead to good and excellent long-term results. It was not the technique of reduction that played the major role in determining the quality of the result but rather the anatomic reorientation that was achieved. The portion of good long-term results was highest in the groups that were treated at the youngest ages and unilateral dislocations had better results than bilateral. In assessing the fair and poor results, he noted two categories of findings. One involved hips that had remained subluxed in which femoral-acetabular alignment was less than perfect, and in the other group, there were major modifications of the shape and structure of the femoral head and acetabulum as a result of the treatment, findings that are interpreted today as secondary to avascular necrosis of the femoral head. Subluxation was noted radiographically by interruptions of the femoral neckobturator line (Shenton’s line in the English literature), valgus positioning of the head and neck in relation to the shaft, persisting femoral anteversion, and obliquity of the lateral aspect of the acetabulum. (ii) Open Reductions. The results of open reduction for treatment of congenital hip dislocation were slightly improved from the 25% rate with closed reduction, but analysis indicated only a 40% range with good long-term results [213]. Poor long-term results were due to failure to correct the valgus deformity and anteversion of the head and neck region. The slightly improved results over those from closed reduction were due to the removal of interposed limbus and capsular tissue. In some instances imperfect handling of these interposed tissues still constituted a problem. Leveuf pointed out that in a theoretical sense, the approach taken by Zahradnicek of Czechoslovakia was required for marked improvement of results which involved not only open reduction with the removal of interposed soft tissues (limbus and capsule) but also the simultaneous correction by osteotomy of the valgus and anteversion deformities of the proximal femur by resecting a trapezoidal segment at the base of the neck with a wedge based posteriorly to correct the

68

1 Developmental Dysplasia of the Hip

anteversion and medially to correct the valgus [213, 214]. These procedures also led to a certain degree of shortening which eased the reduction. Although the principles of this approach were excellent in reality, there were many problems since with the techniques of that era there was some mortality. ( iii) Summary of Treatment Approaches. Leveuf and Bertrand summarized approaches to treatment outlining many principles that are still valid [97, 215]. Treatment should not be rigidly defined as either closed or open, as was the case in many clinics; the approach that best suited an individual patient was the desired one. It was essential to make a precise anatomic diagnosis of the status of the dysplastic hip prior to deciding whether closed reduction or operative intervention was the desired treatment. Consideration of the age of the patient played a major role since the deformity worsened with time and in particular with the beginning of weight bearing since many of the changes were secondary and not primary abnormalities. Since many patients were seen after walking had begun, between 18 months and 2 years, it was important to determine whether or not the dislocation was reducible, whether some deformations had been induced secondarily following a trauma of reduction, and whether after treatment certain malformations which compromised the eventual stability persisted these being a dysplastic acetabular roof and valgus and anteversion of the proximal femur. In a young infant, difficulty in obtaining closed reduction was unusual. False reduction could be present however and increased with the age of the patient owing to the interposition of the limbus or the capsule or both between the femoral head and the acetabular cartilage. Owing to the technique of arthrography however, they felt that false reduction should be recognized and not accepted. The second group of concerns related to circulatory problems following reduction that led to AVN and subsequent growth abnormalities even if the femoral head was successfully reduced. They recognized that these growth problems were secondary to the various reduction maneuvers since these changes did not exist in adult dislocations that had never undergone any treatment. They did not accept however that these were exclusively cases of AVN since they reported that such findings were infrequently seen to the point of not existing with open reductions. The third group of problems occurred with imperfect long-term development in those cases where acetabular dysplasia or proximal femoral rotational changes had not been corrected.

The value of early treatment was recognized. Early diagnosis by clinical assessment and radiography was possible, and the acetabular index of Hilgenreiner was a particularly sensitive, early indicator of acetabular dysplasia [216, 217]. The value of arthrography was stressed in efforts to note interposed tissues that compromised good end results. They noted that Putti had treated 777 cases of early dysplasia by splinting with the hips in abduction and slight flexion and a review of 478 of these cases indicated 92% excellent results. Even at that time (1941) however, they recognized that many of these hips would have been corrected to a normal range spontaneously. Only those hips truly shown to be dislocated warranted treatment. (b) Gill; Farrell and Howorth. Gill published a detailed review in 1948 of 105 congenitally dislocated hips that required either closed or open reduction for treatment [218]. Of the 105 hips treated by closed reduction, initially 52 subsequently required open reduction for appropriate positioning. Results were divided into perfect, excellent, satisfactory, and failure. Only 25% of dislocated hips could be expected to become perfect or excellent after closed reduction with increase to 35% when the dislocation was reduced in the first 3 years of life. Another 15% became satisfactory for a varying number of years, with increase to 20% when treatment occurred before the end of the third year. Failure however was noted in 60% of all primarily reduced dislocations and 45% of those reduced in the first 3 years of life. Subsequent open reduction and femoral and acetabular procedures were of value but clearly represented failures of closed treatment. Gill also noted that the earlier the treatment began the better the results. In those treated under 2 years of age, there was a 36% rate of perfect and excellent results which diminished to 34% in those treated in the third year, to 21% in those treated in the fourth year, and to 15% in those treated beginning over 4 years of age. Farrell and Howorth, reviewing over 600 cases from the first third of the 1900s, reported 42% “successful” closed reductions and 77% “successful” open reductions [109]. Howorth described the open reduction technique [101]. Steindler reported only 11% “anatomically perfect hips” from his institution in 1950 [219]. He also documented the progressive deterioration of results with time owing to failure to achieve anatomic restoration of joint structure with initial treatment. One to 5 years postreduction in 114 hips good results were 70.6%, fair 14.3%, and poor 15%; at 10–20 years, corresponding percentages were 52.6%/17.4%/30%. (c) Capsular arthroplasty. The capsular arthroplasty was an operative procedure developed originally early in the twentieth century to treat dislocated hips in ambulatory

1.10 Progressively Earlier Diagnosis and Treatment of Congenital Dislocation of the Hip

69

patients older than 3 years of age when straightforward included joint stiffness, avascular necrosis (epiphysitis closed or open reduction was destined to be ineffective. in older papers), re-dislocation, and need for revision. The procedure was designed to restore length and With appropriate attention to detail, the procedure establish some form of femoral head-acetabular relaappeared to provide 15–20 years of reasonable relief. tionship by surgically placing the head into a remade The basic operation was widely used, with variable socket. The operation was described and frequently technique, over the first half of the twentieth century done by Colonna. He reported on its evolution and and even into the 1960s in some centers. Codivilla’s results from 1932 (initially describing 66 operations) original description was in 1901 [222] with refinements [220] until 1965 [221, 222]. The procedure was origiby Hey Groves (1926) in England [224] and Colonna nated by Codivilla in 1901 [223] and Hey Groves in (1932) in the United States [220]. Bertrand reported on 1926 [224] and has recently been revived by Ganz 70 cases in 1955 [226], Laurent 102 (1964) [227], et al., with significant additions for use in patients in Trevor 61 (1968) [113], and Dega et al. 172 [228]. the second and third decades who require major reposiShortening and derotation osteotomies of the femur tioning but are too young for total joint arthroplasty were added increasingly to minimize the need for pre[225]. Colonna defined the two-stage procedures for operative traction to bring the head to the level of the those requiring open reduction where extensive deforacetabulum and to treat the proximal femoral antevermation had occurred owing to age. He applied the prosion. In four separate studies, results were reported as cedure for congenital dislocations in the age range good in 70% [226, 228], satisfactory in 70% [113], and 3–8 years as long as the femoral head shape approached good in 56% [227]. Subsequent papers with more pronormal. The procedure is used for otherwise normal longed follow-up continued to report a similar distribuchildren. Trevor also has used the procedure for hips tion of results: Chung et al., 63 hips, excellent (8), good with subluxation and severe acetabular dysplasia. The (23) = 31/fair (19), poor (6) = 25 with degenerative first stage involves subcutaneous adductor tenotomy changes and pain at 20–25 years post-surgery [229]; followed by skin or skeletal traction to pull the femoral Pozo et al. 50 hips, 70% good at mean follow-up of head to the level of the acetabulum. The second stage, 20 years [230]; and Stans and Coleman, 22 hips, Harris the capsular arthroplasty, involves (i) clearing the acehip score mean 82 (range 52–98) at a mean of 16 years tabulum of soft tissue to reveal the triradiate cartilage; [231]. Much of the persisting use of the procedure indi(ii) using a child hip reamer to create a deep, smooth cates the magnitude of the problem facing the orthopecup-shaped socket; (iii) suturing of the joint capsule dic community in treating the dislocated hip in the over the head; and (iv) thinning of the capsule until the 3–12-year-old range. Many good radiographs and funccapsule-covered head fits easily into the new socket. If tional results were achieved however. more than 60° of proximal femoral anteversion is still present, a third stage derotation femoral osteotomy is There has even been a revival of its use, in occasional done, often distally. A plaster hip spica holds the hip in circumstances with many modifications, in the 13–25-year- place for approximately 4 weeks after which the limb is old age group so as to delay the need for total joint arthromaintained in balanced suspension for 4–6 more weeks. plasty. Ganz et al. have detailed their technique for capsular Range of motion exercises are then done, if possible in arthroplasty to include some or all of relative neck lengthena pool or water bath, to gradually re-establish motion, ing with distal transposition of the greater trochanter, femobut walking is not done until 6 months post-arthro- ral shortening/derotation osteotomy, acetabular roof plasty. The synovium lining the inner capsule nourishes augmentation (shelf procedure), femoral head size reduction, the cartilage of the femoral head, while the outer sur- and use of the surgical hip dislocation vascular preservation face of the capsule is stabilized by healing to the bleed- approach [225]. At a mean follow-up of 7.5 years, the Harris ing acetabular cup after it has been cleared of soft hip score was a mean of 84 (78–94). tissue and reamed to a smooth surface. Some excellent hip radiographs were shown by Colonna post- reconstruction, although many problems such as stiff- 1.10 Progressively Earlier Diagnosis ness and femoral head avascular necrosis occurred. The and Treatment of Congenital prolonged postoperative hospitalization, close superviDislocation of the Hip sion, and physical therapy often produced an excellent biologic and clinical result, but, other than the infre- Roser [65] in 1879 stressed the value of early diagnosis and quent modification and application by Ganz et al., to be early treatment, but little meaningful response to his teachdescribed below, the procedure is not currently used. ings occurred among the orthopedic and surgical professions While many good results were reported, complications for several decades.

70

1 Developmental Dysplasia of the Hip

1.10.1 Hilgenreiner Hilgenreiner in 1925 reviewed the approaches to congenital hip dislocation concentrating on the timing of treatment since most patients were still a few years old when treatment was started [216]. Since the condition could often be diagnosed in the newborn, he felt that treatment in a very young infant, while difficult, was desirable. He pointed out that “with the development of methods for early diagnosis and consequently the possibility of very early recognition of hip disease on one hand, and with the development of the retention splint on the other, there is nothing to prevent the universal early treatment of congenital dislocation of the hip, even in infants.” He concluded that every congenital hip dislocation can be treated as soon as it is recognized and that the earlier the reduction is done, the simpler the method of reduction such that in the infant, it could be accomplished without anesthesia. Good results depended mainly on improved development of the acetabulum and the femoral head, and with early treatment delayed complications could be avoided. The abduction splint was best for retaining the reduced femoral head in the acetabulum in the infant, and even early operation, if needed, was desirable compared with the long-term complications when the hip remained dislocated. Results, seemingly good at the termination of treatment, often degenerated during the later years of growth. He commented on the increasing recognition of avascular necrosis and also on the recognition that the harshness of the reduction in particular in patients greater than 2–3 years of age was the cause of the AVN. He indicated: “today it is far more important that the age limit be moved downward.” Many continued to wait once diagnosis was made until even the third or fourth year of life, but Hilgenreiner noted: “this remarkable attitude is hard to understand because from the beginning, it was quite evident that this lesion is similar to a Fig. 1.6 Illustration from the classic work of Hilgenreiner delineating radiographic changes in dysplastic hips. The acetabular index (alpha) is increased on the dysplastic side (left side of this image), and the H (h) distance is decreased on the dysplastic side. The transverse line described by Hilgenreiner linking the two triradiate cartilages now carries his name. (Reprinted from Hilgenreiner 1925, Ref. [216])

traumatic dislocation. The lesion does not improve by waiting and delay is not harmless. The loss of contact between the acetabulum and head causes secondary effects on the capsule, the ligamentum teres, and the hip musculature changing the shape of the bones and making it more difficult to reduce the hip and to maintain the reduction therefore unfavorably affecting the end result.” One of the problems of early treatment in the first year of life was difficulty with the hip spica casts. Hilgenreiner reviewed the works of several authors who had attempted to promote the earliest possible treatment time. Bade was quoted by Hilgenreiner as stating in 1908 that “if every doctor would carefully examine the hip joints of the newborn, and if he had the slightest doubt consult an orthopaedic surgeon or get x-rays in order to clarify the issue, then it is probable that many cases of congenital limping would not develop and could be avoided even before the infant takes its first step.” Walther, Joachinsthal, and Loeffler also concluded that those treated early, especially before walking began, had the best results (see [216]). Vulpius and Engelmann also recommended that treatment should begin at the time of diagnosis and was preferable before walking began (see [216]). Hilgenreiner noted the relative ease of reduction in the early months of life, even without anesthesia. He commented on the typical “noise” which could be appreciated with reduction. Asymmetry of the inguinal, adductor, and gluteal thigh folds was noted with dislocation, but he also stressed that the formation of infant thigh folds could vary greatly, both in relation to number and symmetry “even when the hips are normal.” He also described the subtle changes in the hip radiograph in the early months of life with hip dislocation prior even to formation of the femoral head secondary ossification center. Particularly important were variations in the slope of the acetabulum, as determined by the angle of the acetabulum (Fig. 1.6). He defined the acetabular angle

α h

d

α1 d1

h1

1.10 Progressively Earlier Diagnosis and Treatment of Congenital Dislocation of the Hip

and felt that in the normal infant, the angle was usually less than 20°. He described the “Y” line, subsequently referred to as the line of Hilgenreiner, in the following way: draw a horizontal line connecting the two Y-shaped cartilages. Determine the acetabular angle that is the angle formed by the bony acetabulum with the abovementioned horizontal line. After the first year of life, the secondary ossification center of the femoral head was below the horizontal line and medial to a vertical line drawn from the outer edge of the bony acetabulum perpendicular to the vertical line as described by Perkins and Ombredanne. He developed a splint that held the hips in abduction but was much less cumbersome than the plaster spica.

1.10.2 Putti Treatment was begun as early as 4 months of age [172, 210, 232]. Although surgeons since the late 1800s had occasionally recommended early treatment for congenital dislocation of the hip, by which was meant during the first year of life, it was really the experience and influence of the Italian surgeon Putti (as well as Hilgenreiner) that placed the concept on a firm footing. Closed reduction in the older patient had improved results somewhat, but there were still many imperfect results. The approach to be followed, he suggested, was one of “lowering of that age limit.” At that time, the earliest age of treatment was considered to be 2 years, although Putti contended there was no theoretical or practical reason to forbid commencing treatment before that age. When treatment was begun in the first few months of life, he was able to abduct the hips gently without anesthesia so that the head repositioned itself into the acetabulum. Therapy then required maintaining this position for a few months in order to secure permanent reduction. He developed an abduction cushion for that purpose. The average duration of treatment was from 8 to 12 months. The treatment was felt to be ideal, however, in the sense that there was no anesthesia and no manipulation that even then was recognized to lead to osteochondritis (avascular necrosis). Since rigid immobilization in plaster hip spica casts was avoided, there was also a marked diminution of atrophy of muscles and joint rigidity. He concluded that the best way of improving the results of treatment was to lower the age limit at which treatment begins. Long-term studies were demonstrating that where a hip was incompletely reduced, the results were always less than ideal such that “no complete or permanent restoration of function occurs apart from perfect anatomical reduction.” The younger the patient, the easier is the reduction and the less the likelihood of soft tissue interposition. He clearly articulated the

71

principle that one must “abolish” the age limit and begin treatment at the very moment the deformity is observed even if that was on the day of birth. He even made the radical statement of the value of submitting every newborn child to a routine x-ray examination of the hips. He was able effectively to treat large numbers of patients during the first year of life; in one series the average age was 4 months. In 119 cases treated by the abduction method from 34 days to 16 months of age (average age 4 months), complete cure was achieved in 113.

1.10.3 Ortolani Ortolani of Italy furthered emphasis on the great value of early diagnosis and early treatment in congenital hip dysplasia [233]. He stressed the value of newborn examination of the hips based on his feeling that congenital dysplasia developed in utero and was present at birth. The hip instability was diagnosable clinically by the “click” sign in which the femoral head was dislocated with adduction and relocated with abduction maneuvers. This sign came to be referred to widely as the Ortolani sign and was considered by Ortolani and others to be most classically seen within the first 2 months of life. The sign was best noted in cases of mild to moderate hip dysplasia. Those with severe dysplastic hips often were not reducible, and the clicking sensation was not detectable. Ortolani felt however that the vast majority of cases of congenital hip dysplasia were mild to moderate and thus detectable in the early weeks of life.

1.10.4 Von Rosen and Barlow Separate reports by Von Rosen [234] and Barlow [158] in 1962 stressed the value of routine neonatal hip examinations encompassing all newborns in detecting hip instability and allowing for careful observation and appropriate early treatment of those with instability. It shortly became accepted that the hip examination was valuable as a routine part of the neonatal assessment, an approach that has minimized considerably the late initial detection of the dysplastic hip. Barlow stressed that the most essential finding was the dislocatable/relocatable hip as determined by the clinical maneuvers. Limited abduction was not of value in the newborn since it was a later finding representing a secondary change. Asymmetry of the thigh folds was also of no specific value since less than one-half of the dislocated hips had asymmetric folds and the large majority of children with asymmetric folds had normal hips.

72

1.10.5 Widespread Adoption of Neonatal Hip Examination Many individual surgeons had recognized that it was possible to detect congenital instability of the hip in the neonatal period and that early examination and treatment of hip displacement would almost certainly lead to improved results and minimize secondary adaptive changes. It was not until the 1950s however that widespread examination of newborns was studied and adopted as common practice. Large centers in Sweden and Great Britain were prominent in this regard. Simple hip abduction splinting for several weeks led to cure in the large majority of cases. Howorth in a report in 1977 summarized several of these studies reported from 1950 to 1975 and indicated that 2010 displacements had been reported in 155,255 examinations [102]. At present, it is widely accepted that hip examination, specifically checking for instability, should be an integral part of the neonatal pediatric examination. Procedural questions involve whether or not the clinical examination should be done by an orthopedic surgeon and whether ultrasound studies should be performed routinely in all children or only where clinical concerns have been raised. Plain radiographs do not have sufficient resolution to effectively diagnose most cases of infantile developmental dysplasia of the hip. Secondary ossification centers of the proximal femur have not formed, the acetabular index is too variable, and there is no dynamic component to the method. There are occasional hip dislocations that can be missed even with careful clinical examination screening programs [235]. Mitchell reports 4 dislocated hips diagnosed after walking began having been either not recognized or not recognizable in a total of 31,961 newborn examinations between 1962 and 1968. The incidence of unstable or luxated hips detected early was 100/260 or 0.7%.

1.11 A ssessments of DDH Treated by Closed Reduction 1.11.1 Radiographic Classification System (Severin) Severin developed a clinical-radiographic categorization that has greatly enhanced the assessment of long-term results [44]. He related the effects of treatment in childhood years to the likelihood of a predisposition to osteoarthritis in mid- to late adult life if perfect anatomic repositioning was not achieved. He performed one of the earliest and most detailed studies of the results of closed reduction treatment of CDH and developed a categorization system based on the radiographic appearance that relied heavily on the determination of the CE (center-edge)

1 Developmental Dysplasia of the Hip Center-Edge (CE) Angle of Wiberg

CE E

C

Head/Neck Shaft Angle

Fig. 1.7 Measurement of the center-edge (CE) angle of Wiberg plays an important role in assessing results in treatment of CDH-DDH. It is most valuable in adolescent and adult assessments but can be used as early as 6 years of age

angle of Wiberg (Fig. 1.7). He ultimately reviewed treatment in 330 patients with 448 cases of congenital dislocation of the hip. Most patients had been treated in the 1920s and 1930s when the best time for treatment was considered to be between 2 and 3.5 years of age, even if diagnosis had been made prior to that time. As a prelude to his study, he performed additional studies on the CE angle of Wiberg stressing in particular values between 6 and 13 years of age to compare with Wiberg’s studies that assessed men and women between 20 and 35 years of age.

1.11.2 Measurement of the CE Angle Severin defined measurement of the CE angle as follows: “If one draws a line between the center of the head and the outside edge of the acetabulum and another through the center of the head parallel with the longitudinal axis of the body (at right angles to the ‘Y’ line of Hilgenreiner) the 2 lines will normally form an angle, the CE angle.” In a normal hip, the CE angle is positive, but with maldevelopment of the acetabular roof, upward and outward displacement of the head, or some other deformity, the center of the head moves closer to and perhaps beyond the lateral edge of the acetabular roof; the CE angle diminishes and when lateral to the edge of the bony roof becomes negative. Wiberg’s studies combined data from males and females and assessed values between 20 and 35 years of age. He defined an angle in that age group of less than 20° as abnormal, 20–25° as uncertain, and greater than 25° as normal. Severin performed studies between the ages of 6 and 17 years involving 200 normal hips in 100 subjects,

1.11 Assessments of DDH Treated by Closed Reduction

52 females and 48 males. The values for children 14–17 years of age were similar to those of Wiberg in adults from 20 to 35 years of age such that Wiberg’s standards for the CE angle apply down to and including the age of 14 years. Values for the developing hip were represented by those between 6 and 13 years of age (136 hips); a CE angle less than 15° was abnormal, between 15° and 19° uncertain, and greater than 20° normal. In summary, CE values greater than 20° are normal for those between 6 and 13 years of age, and values greater than 25° are normal for those greater than 14 years of age. Abnormal values are less than 15° for those 13 years of age and less than 20° for those of 14 years of age or greater. The CE values have been assessed in other populations. Skirving compared Caucasian and African radiographs, and the distribution of values was similar in the two groups [236]. This was somewhat unexpected since osteoarthritis is rare in the African population. Data have also been documented in a Chinese population [237].

1.11.3 Severin Classification for Radiographic Assessment of Long-Term Results Severin performed a long-term study on 330 patients involving 448 hips which had been treated by closed reduction and who at the end of treatment were considered to be “primarily successful results” by the treating physicians [44]. His follow- up series involved 306 patients with 417 involved hips. Of these, 266 were female and 40 male, a 6.7:1 female/male ratio. He categorized the end result into six groups. The functional results tended to deteriorate as the patient grew older because of the development of secondary changes in the joint. He indicated: “only the anatomically cured cases can reckon with freedom from future trouble.” Those in group I were considered to be normal, but even here he noted that “there is always something to distinguish the hip with CDH from a normal one, even though some of the findings are sufficiently mild that the patient was still considered normal.” Group I Anatomically well-developed hip joints with a spherical femoral head and a normal CE angle. Group I was subdivided into groups Ia and Ib: Ia, CE angle of more than 19° for ages 6–13 years and CE angle of more than 25° for ages from 14 years up and Ib, CE angle of 15°–19° for ages 6–13 years and 20°–25° for ages from 14 years up. Group II Hips exhibit a distinct but moderate deformity of the femoral head or neck or acetabulum but have a satisfactory structure with otherwise normal conditions in the joint

73

overall. This group is also subdivided into group IIa with normal CE values for age and group IIb with uncertain CE values for age. Group III Dysplastic hip joints with a maldeveloped acetabular roof but no subluxation. The CE angles are less than 15° in those 6–13 years old and less than 20° in those 14 years of age and up. Group IV Subluxation. The femoral head is displaced upward and laterally in relation to the acetabulum, and there is an unmistakable break in Shenton’s line such that the hip is subluxated. The acetabulum is always more or less dysplastic and the femoral head is often deformed. Subclassifications are: IVa, moderate subluxation with the CE angle still positive or equal to 0° and IVb, severe subluxation with a negative CE angle. To be considered as a subluxation, the acetabulum must be a direct continuation of the original acetabulum. Group V The femoral head articulates with a secondary acetabulum developed at the margin of the original acetabular roof. This secondary acetabulum is separated from the original joint cavity and does not represent a continuation of it but rather a remolding phenomenon. Group VI Complete re-dislocation of the femoral head. Deformities of the femoral head and neck occur in all of the groups except group I. The deformities are moderate in groups II and III, but in groups IV to VI, they are of major degree and are generally combined with shortening and thickening of the neck. In the long-term study in which patients were categorized from group I to group VI, there were 417 hips assessed with the following results: group I (well-developed hips joints), 4.6%; group II (moderate deformation of the femoral head or neck or acetabulum in an otherwise well-developed joint), 7.7%; group III (dysplasia without subluxation), 8.6%; group IV (subluxation), 47% (of these, group IVa with slight subluxation involved 17.7% and group IVb with severe subluxation involved 29.3%); group V (femoral head is in a secondary acetabulum in the upper part of the original acetabulum), 13.9%; and group VI (redislocation), 18.2%. The prognosis was better for unilateral than for bilateral cases, and the later the hips were reduced, the more deformities there were in the femoral neck and femoral head. The late results for the joint as a whole were best if the reduction was done early which in the then current framework meant before the age of 1 year. Femoral head and neck deformities were invariably seen simultaneously.

74

1.12 T he Development of Modern Treatment for CDH and DDH 1.12.1 Hip Reduction and Stabilization in the Early Weeks of Life: “Functional” Method (a) Pavlik harness. Over the past few decades, treatment by closed reduction along with earlier diagnosis preferably in the newborn period has become the preferred approach. Pavlik harness application is now widely accepted as the initial treatment leading to hip reduction and stabilization in the early weeks and months of life [238]. This approach is referred to as the “functional method” which is the English translation from the European literature where the concept originated. It refers to the allowance of active hip motion, in a position of flexion greater than 90° and abduction of approximately 45–60°, within this restricted range (as controlled by a harness) which predisposes to (i) reduction of the hip without force and based solely on positioning and (ii) maintenance of the reduced position allowing the soft tissues and bones of the hip region to grow or mature into the normal stable conformation. It thus allows for a more physiologic management of hip dysplasia. Ease of management of the child is improved compared to a hip spica, and hip motion in the reduced range is more physiologic than rigid immobilization. Treatment using a harness originated with Arnold Pavlik in Czechoslovakia in the early 1940s with report of an 85% success rate in the reduction of dislocated hips [238]. He reported 1912 hips including 640 dysplasias, 640 subluxations, and 632 dislocations. He came to feel that early treatment preferably starting no later than 8–9 weeks of age was best. The range of success in several series has been reported from 80% to 97%. As a general rule, combining studies from the larger series across a spectrum of the extent of disorders, there appear to be approximately 95% success reported overall with 99% effectiveness for dysplasia (neonatal instability), 98% for Barlow-positive hips, and 83% for Ortolani-positive hips and then decreased effectiveness ranging around 60% with fixed dislocation. In review of 3611 dislocated/dysplastic hips in 2636 patients less than 11 months of age using the Pavlik harness, healing rates were determined depending on the age treatment began and the degree of displacement based on the Tönnis radiographic and Graf sonographic classifications. The average age at initial treatment was 4.1 months (range 2 days to 11 months) [153]. Results were better the lower the degree of displacement defined by the Tönnis 1–4 grading criteria. The healing rate at follow-up at an average of 4.5 years (range 1–9 years) was Tönnis grade I (dysplastic hips) 95.4%, grade II

1 Developmental Dysplasia of the Hip

92.3%, and grade III 52%. In those with dislocated hips, 80% normalized. (Few grade IV hips were treated with the harness.) The authors concluded that the Pavlik harness could be used as the preferred primary treatment from the neonatal period to 6–7 months of age. The rate of AVN was 2.4% (higher in Tönnis 2–4 patients). Results continue to be reported with similar ranges of effectiveness in many countries. Since the extensiveness of involvement in DDH is not uniform, the findings being dependent on the age at diagnosis, it is the extent of the dysplasia at the beginning of treatment that heavily influences the results of treatment. The most favorable results occur with the femoral head positioned in a mildly dysplastic acetabulum and the poorest with a fully dislocated hip which does not completely reduce with initial positioning. Not all reports fully clarify the initial status of each hip being treated. A recent systematic review of 218 studies (of which 62 fit more rigid criteria) found “satisfactory clinical and radiologic outcomes with the use of the (Pavlik) harness at long-term follow-up” although some failures and avascular necrosis were reported [239]. The authors strongly supported use of ultrasound in conjunction with the method. Lerman et al. assessed 137 hips with DDH treated with a Pavlik harness with 19% (26/137) failing treatment and 81% effectively managed. All six patients that were initially irreducible and with an ultrasound coverage of 64°) dysplastic unstable hips (Graf type IIc) and dislocated hips [257]. The mean times at start of therapy and length of treatments were 18 weeks (14–25)/17 weeks (14–20) and 3.5 days (1–8)/52 days (21–87), respectively. (d) Pavlik harness treatment from 6 to 24 months of age. Pollet et al. studied use of the Pavlik harness in those diagnosed late from 6 to 24 months of age [258]. They treated 26 hips with the Pavlik harness at a mean age of 9 months (6–23). Successful management was seen in 46% (12/26). The successful cases were all Graf type III patients; none of the Graf type IV hips had successful relocation. In another study Pavlik harness treatment starting at a mean age of 6 months (5–12) has been reported [259]. Overall 67% (20/31) were successfully reduced with the Pavlik harness. Younger age and lower (better) Tönnis scores were favorable for better results. Seventeen (81%) of Tönnis 2 hips were successfully reduced, while only two (25%) of Tönnis 3 and 4 hips were reduced with this approach. (e) Hip reduction from 6 to 18 plus months using “Functional Method” of Hoffmann-Daimler. Efforts to extend the use of the closed-functional method for DDH to older age groups have not been widely accepted in the orthopedic community, especially in North America, but have been developed and modified by some centers in Europe. Hoffmann-Daimler of Heidelberg, Germany, developed the method and explained the principles along biomechanical and histopathologic lines [260–262]. The treatment was used in the age range primarily from 6 months to 4.5 years of age. A major concern was the finding of relatively high rates of avascular necrosis. More recently, Papadimitriou et al. modified the technique and reported good results [263]. They report on 95 hips in 65 patients with onset of treatment at a mean age of 16 months (range 6 months to 3 years 10 months). The patients had isolated DDH with no other systemic (i.e., neuromuscular, skeletal dysplasia) disorders and no previous operative treatment. The H-D method comprises two stages. In the first or reduction phase (Phase A), a harness is used to hold the hips in full flexion until there is plain radiographic evidence of concentric reduction. Posterior straps were used to filly abduct the hips. The recent modifications, designed to minimize avascular necrosis, flex the hips only to 120° and does not force abduction with posterior straps. The mean time to reduction was 47 days (15–150). Where the hips reduced quickly (e.g., 15 days), the child was placed in hip spica for 6 weeks to encourage the lax soft tissues to tighten before advancing to the second phase. The shoulder and chest components of the harness are similar to the Pavlik harness with the hips flexed by wide thigh slings with the feet and knees free.

1 Developmental Dysplasia of the Hip

The second or acetabular remodeling phase (Phase B) is designed to maintain reduction, while acetabular remodeling to normal depths and angles develops in relation to the now normally repositioned femoral head. In the original approach, the harness was continued for a month with both flexion and abduction at 90° along with use of the abduction splint. In the current modification, abduction is done only to the comfortable levels of abduction but is retained until full acetabular remodeling is seen radiographically. After that abduction is gradually reduced to 45° for each hip as the child walks with it on. The average length of phase B was 13 months (3–30). Papadimitriou et al. noted a 6.3% incidence of avascular necrosis (6/95) with no episodes of re-dislocation. Follow-up was a mean of 11.5 years (6–29). The average acetabular index decreased from 40° ± 7.4° prior to treatment to 24° ± 5.7° at end of treatment. Results were satisfactory in 93% (Severin I 67, Severin II 21) and unsatisfactory in 7% (Severin III 6, Severin IV 1) with no V or VI grades. Prior to treatment there were 45 (47%) Tönnis 2 hips, 23 (24%) Tönnis 3, and 27 (28%) Tönnis 4. Femoral nerve palsy was not seen. These findings are significant for those who remain particularly concerned about problems also associated with closed or open reductions. (f) Mechanisms of action in “functional” treatments of DDH. The principles of action underlying the Pavlik harness and the history of its gradual and now widespread acceptance have been well outlined by Mubarak and Bialik [264] and Bialik [265, 266]. Pavlik considered movement to be the principal necessity for correct treatment of congenital dysplasia of the hip. This was in marked contrast to rigid cast techniques that he termed “passive mechanical” treatments. Pavlik wrote that “The principle of the method is to bring the child’s lower limbs to flexion in the hips and knees, using stirrups. It is well known that neither child nor adult is able to keep the lower limb adducted in flexion. This is non- physiological; the muscle become tired quickly and the limbs go into abduction. And this is what the hip joint needs for the treatment of dysplasia, subluxation and dislocation.” Also, “flexion, abduction and adduction are free! The badly developed hip joint needs movement for healing, because the hip joint is an organ for movement.” (g) Length of treatment and continuing follow-up in functional harness management. Bin et al. address the important question as to the duration of harness therapy in neonates [267]. They questioned whether correction of instability alone was all that was needed since residual acetabular dysplasia would then correct spontaneously over several months or whether treatment needed to continue after stability was reached to protect and enhance

1.12 The Development of Modern Treatment for CDH and DD

acetabular normalization. In 42 abnormal hips, they used the former method, treating the instability only for a mean of 34 days beginning at 5 days of age. Subsequent follow-up to a mean of 6.7 years (5–14) showed a mean acetabular angle of 20° (12–30°) and a mean CE angle of 30° (22–35°) with all hips graded Severin 1a. Gans et al. however felt that abduction bracing helped residual acetabular dysplasia to resolve [268]. They assessed patients who has residual acetabular dysplasia at 6 months of age after earlier treatment; in 70 hips showing an acetabular index >30°, 39 hips went unbraced, while 31 hips were treated for 6 months using an abduction brace at night. Those braced improved their acetabular indices a mean of 5.3°, while those observed (not braced) improved only 1.1°. Part-time abduction bracing was recommended to enhance acetabular remodeling. Sarkissian et al. make an observation worthy of further assessment [269]. They continued to follow a consecutive series of 115 infants with DDH whose hips had become normal after early Pavlik harness treatment at a mean of 3.1 ± 1.1 months based on normal ultrasounds without acetabular dysplasia and no signs of clinical hip instability. Anteroposterior radiographs were done in all at 6 and 12 months of age; at 6 months 17% of all infants had radiographic signs of acetabular dysplasia, and, of those left untreated (n = 106), 33% had dysplasia at 12.5 months. They recommended continuation of radiographic assessment after seemingly full correction at 3–4 months of age by clinical and ultrasound criteria. ( h) Muscle action in functional harness management. For both the Pavlik harness and Hoffmann-Daimler “functional” approaches, muscle action helps redirect the femoral head into the acetabulum once the favorable positions of flexion and abduction of the hip are provided and maintained. Ardila et al. for the Pavlik harness [270] and Papadimitriou et al. for the Hoffmann-Daimler approach [263] discuss the muscle contributions to reduction. Muscle over-reaction is not considered to cause the dysplastic positioning of subluxation or dislocation but can influence reduction and treatment. Ardila et al. have recently done a detailed three-dimensional computer model study simulating hip reduction dynamics. They identify five hip adductor muscles as key mediators of DDH prognosis, working in the direction needed to achieve concentric reductions. The work is based on the flexion/abduction position of the hip as the needed position for effective management. The adductor muscles play key roles in effecting reduction, namely, pectineus, adductor brevis, adductor longus, gracilis, and adductor magnus (minimus [proximal], middle, and posterior [distal] segments). The iliopsoas is relaxed in the flexion/abduction position and does not restrict reduction (although it does in extension by

77

stretching against the anterior hip capsule). The pectineus exhibits the highest component of pull in direction of reduction toward the center of the acetabulum aiding in reduction and maintenance of reduction. Forces are more favorable overall for adductor muscle action with subluxation of Graf III type with pectineus, adductor brevis, adductor longus, and proximal adductor magnus playing major roles. In dislocations (Graf IV) all the muscles are detrimental initially to reduction explaining the need for some “traction” to initially improve positioning. Based on origin and insertion points, the muscles are actually shorter in full dislocation than in the normal or reduced hip, postulating the need for traction to overcome muscle tension and bring the femoral head over the posterior labrum for reduction. The authors define a two-stage series of events as the muscle channel improvement of position; these are (a) release phase as the femoral head is brought from posterior to the acetabulum to the perimeter of the labrum and then (b) reduction phase as the femoral head is pulled from the labrum perimeter into the acetabulum. Even in the older patients, the concept of muscle involvement is stressed. Positioning the hip relaxes certain muscles and places others in the optimal position for them to function in allowing for/or causing reduction and then maintaining position. Hoffman-Daimler illustrates the role of the adductors associated with subluxation. Papadimitriou et al. specifically illustrate how deformation with the hip adducted and the knee fully extended allows the muscles around the hip (adductors, iliopsoas, and hamstrings) to act to dislocate the hip, driving the femoral head over the posterosuperior rim of the acetabulum. Placing the hip in flexion and adduction allows the redirected adductors to function moving the head into the acetabulum while neutralizing (relaxing) the deforming forces of hamstrings and flexors. (i) Femoral nerve palsy in Pavlik harness treatment. Development of a femoral nerve palsy as a complication of Pavlik harness treatment for DDH has been recognized for several decades (Ramsey et al) [271] but generally as isolated cases. Murnaghan et al. however performed a retrospective single institution review of 1218 consecutive patients treated for DDH with a Pavlik harness and found 30 cases of femoral nerve palsy, an incidence of 2.5% [272]. The palsy was defined as an inability of the infant to actively extend the knee either spontaneously or in response to gentile stimulation of the foot. Most of the palsies (26/30, 87%) occurred within the first week of treatment, all involved an affected hip, and all had full return of function with no permanent femoral nerve palsy persisting. There was a clear tendency for occurrence in heavier, larger, and older infants. Bilaterality and birthweight did not correlate. The

78

1 Developmental Dysplasia of the Hip

response to the palsy varied from loosening the harness to decrease the amount of hip flexion to discontinuing its use for a few days to abandoning the harness completely. The condition resolved from 1 to 28 days after detection and management, but those who responded within an average of 5 days had a much better eventual outcome of hip development than those who had return of function at an average of 15 days. The nerve palsy served as an indicator of the extent to successful long-term result (of the DDH); the overall DDH/Pavlik harness group had a 94% success rate which decreased to 47% in the palsy group. The positive results in the palsy group progressively diminished from dysplastic to Barlow positive to Ortolani positive to fixed dislocation categories.

1.12.2 Treatment by Closed Reduction 1.12.2.1 C losed Reduction with General Anesthesia, Gentle Manipulation, and Hip Spica Immobilization This approach is used when a dislocated hip has not responded well by reducing with functional harness or abduction splint management after a variable time period (depending on the protocols established in the particular center where treatment is occurring). The time for stopping ineffective Pavlik harness treatment is as early as 2–3 weeks in some centers although others will persist for 4–6 weeks or occasionally longer. If the hip is not reducing smoothly, the harness or abduction splint can lead to progressively poor results since they are positioning the hip abnormally as growth proceeds. In the age range from 4 to 12 months where optimal reduction has not been achieved, there are great variability from center to center and country to country and little specific documentation in terms of subsequent management. After 1 year of age, open reduction is usually resorted to in most centers. In the interim (4–12) months, most will proceed with closed reduction/gentle manipulation/hip spica immobilization and resort to open reduction at any point in that time frame if closed reduction has been ineffective. The closed reduction maneuvers, while seemingly repeating the methodology that developed in the period from the 1890s to the 1930s, differ significantly, primarily due to the knowledge of the high incidence of avascular necrosis of the femoral head that developed with the forcible reductions and extreme casting positions of the earlier era. More Recent and Current Approaches Efforts to reduce the hip while minimizing damage to the proximal femoral vascularity have included prereduction skin traction (overhead with hips in flexion, Bryant’s traction, or longitudinal traction with the hips in extension) or percutaneous adductor tendon release. Reduction is then performed under general

anesthesia in gentle fashion, as distinct from the reports of forcible reduction with extensive abduction and rotation described from earlier eras. The hips are then immobilized in the “human position” of Salter (flexion to approximately 100–110° but beyond 90° and abduction in the “safe zone” between 45° and 60°) to further minimize avascular necrosis. The positions of Lorenz (maximum abduction of the hips to 90° bilaterally with flexion at 90°) and Lange (hips extended, abducted maximally and internally related to overcome the femoral anteversion) induced high, if not invariable, levels of avascular necrosis (AVN) and are no longer used. The relationship of the femoral head to the acetabulum was assessed using plain radiographs; although these were always difficult to interpret due to the thickness of the overlying cast and the fact that the femoral head ossification center was either small or not present. Arthrography (in the operating room at time of reduction) was often used to assess the position of the femoral head and labrum. CT scanning was then introduced which allowed for three-dimensional projections to better assess the position of the femoral neck in relation to the acetabulum in the appropriate planes. More recently, MRI has been used (often under the same anesthesia used for reduction and cast immobilization) to “immediately” confirm femoral head-neck position in relation to the acetabulum and, in a few centers, to assess vascularity of the femoral head in the hip spica cast with gadolinium enhancement. Hip development is then followed by plain radiographs (often out of cast at hip spica changes with or without anesthesia) with favorable responses including: retained good position of head to acetabulum, °increasing size of the proximal femoral ossification center, and deepening of the bony acetabulum (improved acetabular index which should be decreasing to less than 30°). Immobilization is generally needed for several months to produce a stable, anatomically normal remodeling of the femoral-acetabular relationship. Once the full-time hip spica is discontinued, many will use the bivalved cast at night or switch to use of a plastic hip abduction orthosis for variable periods of daily use to assure normalcy.

1.12.2.2 E ffectiveness of Closed Reduction and Its Relationship to Prereduction Traction Traction has been an integral part of treatment for hip dislocation since the mid-1800s. The original studies on reducing hips in that era used longitudinal traction as a means for placing the hip into the acetabulum; pulling on the lower limb against countertraction to relax and lengthen the tight muscles, tendons, and fascia and position the femoral head at the level of the acetabulum; and then using abduction in traction to reposition it. Even after closed manual reduction and hip spica immobilization became the method of choice around 1900, traction remained part of

1.12 The Development of Modern Treatment for CDH and DD

therapy to pull the femoral head to the level of the acetabulum. In the mid-century (1940s–1960s), increasing awareness of the high incidence of avascular necrosis of the femoral head with closed reduction extended the use of prereduction traction with the reason being to gradually elongate the soft tissues including vasculature and gradually stretch the tightened adductor muscles. Increasing doubt crept into management approaches as to whether prereduction traction was necessary. The widespread use of percutaneous adductor tenotomy as part of the reduction/ casting procedure along with limitation of abduction with the “human position” for cast immobilization convinced many in the 1980s–1990s to discontinue traction as part of management. Weinstein (1997) concluded that it “cannot be proven that traction alters the outcome of developmental dislocation of the hip” [273]. Since the 1980s studies have appeared advocating the use of traction, while others conclude it to be of no value. In North America the commonest approach is to use traction for 2–4 weeks prior to closed reduction to stretch the soft tissues gradually and better position the femoral head in the acetabulum so as to enhance the ease of closed reduction and minimize the occurrence of femoral head AVN. Skin traction is generally used bilaterally with the hips flexed 90° (overhead traction) and weights just elevating the pelvis off the bed. In conjunction with this traction, variable abduction is used. In some European and Asian centers, traction is more prolonged and is used as a primary therapy to actually reduce the hip prior to abduction splinting or hip spica casting. In these situations traction tends to be applied longitudinally with the hips extended with gradual abduction increased to guide the reduction. Studies Finding No Value to Prereduction Traction A major study by Sucato et al. found no difference in obtaining closed reduction or in the rate of AVN in 342 hips treated from 1980 to 2009, 269 with a full fixed dislocation and 73 reducible with the Ortolani maneuver [274]. Bryant’s overhead traction was used in 276 with no difference noted in the 2 parameters studied compared with cases not treated with prereduction traction. Kutlu et al. noticed no effect on the rate of AVN when preliminary traction was used as a single determinant in 52 hips and not used in 40 prior to closed reduction [275]. Quinn et al. found no value to preliminary traction regarding reducibility or AVN with a mean of 3 weeks of traction in 90 hips compared with contemporary series in the literature where no traction was used [276]. Kahle et al. treating 47 hips closed or open without traction noticed a very low incidence of AVN (2, 4%) and felt that preliminary traction was not needed for closed or open reduction in non-teratologic hips younger than 2 years of age [277]. Brougham et al. assessed results following closed reduction in CDH of 210 hips [278]. Incidence of some form

79

of AVN was 99/210, 47%, but the incidence was not influenced by the use of traction, the duration of traction, or performance of adductor tenotomy; patients treated by closed reduction without preliminary traction did not have a higher incidence of AVN. Studies Finding a Positive Value to Prereduction Traction Many reports however continue to support the value of traction prereduction for DDH. Langenskiold and Paavilainen found that prereduction traction decreased the incidence of AVN in children 6–36 months of age [279]. The study compared 86 hips undergoing closed reduction before 1957 without preliminary traction with 176 hips after 1957 treated with prereduction traction. The prereduction traction was found to decrease the incidence of AVN of the femoral head when dislocated hips of children 6–36 months of age were treated with closed reduction with traction as the single variable. Daoud and Saighi-Bououina found that overhead traction for a mean of 23 days (18–72) in 50 hips facilitated closed reduction of the untreated congenitally dislocated hips/DDH [280]. Treatment began at a mean of 33 months. Many of the reports favoring the continued use of traction apply it not just to stretch and lengthen the oft tissues but to actually reduce the hip (or virtually reduce it), and therefore the traction is more prolonged and the positioning more detailed. Kamer et al. reported a large series of 1178 hips. The first stage of treatment using skin traction pulled the lower extremity longitudinally for 2–4 weeks followed by another 2 weeks in abduction [281]. The child was then placed in an abduction splint that allowed for some hip movement. AVN was reported as only 3.4%, and the approach was recommended for use for CDH in those 6 weeks to 2.5 years. Tavares et al. used guided abduction via overhead traction in children 1–28 months of age where the Pavlik harness had failed [282]. Effective closed reduction occurred in 20/27, 74%, after which abduction bracing/casting was used. AVN was seen in only 2/27 but the treatment was only effective up to 24 months of age. The authors felt that the traction approach led to a reduced number of open reductions. Danielsson reported on treatment of 75 hips diagnosed at a mean age of 10 months (2–64) [283]. Tenotomy and longitudinal traction were used and closed reduction failed in only seven. Traction with gradually increasing weight shifted to an abduction phase as the femoral head moved toward the acetabulum. After 3–4 weeks, closed reduction and hip spica immobilization were done. AVN occurred in four hips (5%). While innominate and femoral derotation osteotomies were needed in some later, traction/tenotomy/ closed reduction is favored to prevent the need for open reduction.

80

Yamada et al. treated 62 dislocations of the hip with a mean age at start of therapy of 11.5 months (6–23) [284]. Preliminary prolonged traction, mean 8 weeks, was the key to treatment along with adaptation of traction to position as determined by radiographs and ultrasound. The technique involved use of bilateral longitudinal skin traction of 1.5 Kg for 3 weeks or until the femoral head was pulled opposite Hilgenreiner’s line. Then traction weight was increased to 2 Kg with hips flexed to 90°, slightly abducted with knees extended (overhead traction) for 1 week. Flexion and abduction were then increased to 120°/60°, respectively, for 1 week at which time spontaneous reduction without manual manipulation was often noticed. For the final 3 weeks, traction was maintained using 0.5 Kg, 100° hip flexion, and 50° abduction and knee flexed. Under anesthesia an arthrogram confirmed position, and a hip spica (human position) was applied. There was reduction in 57/62, 92% hips, and only 1 case (1.6%) of AVN. Although acetabular/femoral osteotomies were needed in some, the ability to reduce the hip closed greatly minimized the need for open reduction, AVN, and secondary osteotomies. Sibinski et al. assessed the use of prereduction traction in 107 hips compared with no traction in 48 [285]. Overhead traction for an average of 17 days with gradual abduction to 50° (per side) was done followed by closed reduction under general anesthesia and hip spica application. Their study showed that growth disturbances to the proximal femur were significantly reduced in those greater than 1 year of age who had prereduction traction; 2/31, 6.5%, hips with traction had type III AVN (Bucholz-Ogden), while 7/20, 35%, without traction had type III AVN. They strongly recommended prereduction traction for those who seemed to need it most, patients over 1 year of age, and/or those with high dislocated hips. [Closed reduction is discussed further in relation to methods used to diminish AVN – cast position and adductor releases – in the section below on AVN.]

1.12.3 Treatment by Open Reduction By the middle third of the twentieth century, open reduction was resorted to increase when treatment began after 1 year of age. The use of arthrography made the indications for open reduction much more specific since lack of a fully concentric reduction of the femoral head into the depths of the acetabulum which generally meant interposition of the limbus (labrum plus capsule) following closed reduction was soon recognized by most to mandate the open intervention. There were some, such as Severin, who felt that hip spica immobilization could be continued even with inversion of the limbus with the expectation that pressure of the head against the acetabulum would with time wear away the inverted tissue

1 Developmental Dysplasia of the Hip

[128]. Most felt however that this either was not true or that if it succeeded, it would still have left damage to the articular cartilage surfaces that in the long run would not be ideal. In the opinion of Scaglietti and Calandriello, open reduction is resorted to in three situations: (i) in teratologic dislocations where both the acetabular dysplasia and the displacement are so marked that they are considered to originate during the embryonic or early fetal stages and are not amenable to closed treatment, (ii) in those with increased age and dysplasia in whom the combination of clinical and radiographic findings indicates little to no likelihood of successful closed reduction, and (iii) in all children over 3 years of age [110]. The open reduction addressed each of the extra- articular and intra-articular obstacles to reduction described previously in the section on pathoanatomy. In their review of operative procedures that ranged from those under 1 year of age to those between 4 and 5 years of age, various findings were assessed. In the 162 children operated, only 11% were less than 1 year of age, while there were 32% between 1 and 2 years, 29% between 2 and 3 years, 15% between 3 and 4 years, and 13% between 4 and 5 years. Shortening of the gluteus medius muscle was noted as a formal obstacle preventing reduction in only 3.2% although the actual frequency was probably greater since this muscle attachment was freed from the iliac crest for operative approach for open reduction. Tightness of the iliopsoas muscle presenting as a real obstacle to reduction was seen over 50% of the time, and they almost invariably performed a “Z” elongation of the musculotendinous region. This eased reduction of the femoral head and partially relieved the isthmic constriction of the capsule. Pericephalic insertion of the capsule was seen in one-third of the patients and was often a sequel to previous attempts at closed reduction and hip spica immobilization. Occasionally however it was found in primary reductions. Capsular adhesions both internal and external were seen on occasion and were always felt to be acquired following closed reduction procedures. The inverted limbus (capsule + labrum) was observed in 35%. In those less than 2 years of age, it was rarely seen, and the femoral head slipped easily over the labrum with surgical attention to it not required. In those over 2 years of age who were walking, it appeared that the inverted limbus components were usually hypertrophic. The ligamentum teres appeared congenitally absent in 20% of the hips. In 31% there were remnants of the ligamentum teres usually as an atrophic fragment with fibro- fatty tissue in the depths of the acetabulum. In 49.2% the ligament was intact, sometimes being thin and elongated, while in others it was long and flattened. It was always removed since it obstructed complete reduction. The head of the femur in 21% was deformed usually being pear shaped from flattening against the ilium. Some anteversion of the neck of the femur was almost always evident as determined by the degree of internal rotation necessary to reduce the

1.12 The Development of Modern Treatment for CDH and DD

head completely into the acetabulum. The anteversion rarely if ever prevented the centering of the head and derotation osteotomy was needed at primary reduction in only 1.6% or three hips. Shallowness of the acetabulum was due mainly to obliquity of the roof and was always present to a varying extent. On only 11 occasions (6%) was acetabular reconstruction carried out at the same time as reduction. The depths of the acetabulum were always filled with a mass of fibro-fatty tissue usually containing ligamentum teres, and this tissue was always removed. In only nine hips (5%) was the head too large for the acetabulum or so misshapen that the fit was inappropriate. They concluded that in only 12% of hips was there only a single obstacle to reduction. The most frequently found abnormal structures, usually in combination, were a tight psoas muscle, hypertrophic ligamentum teres, fibro-fatty tissue in the depths of the acetabulum, a pericephalic insertion of the capsule, and an inverted limbus. Bony procedures were rarely done as part of the primary procedure. When indicated owing to continuing imperfect development, they were performed as secondary procedures involving either a proximal femoral varus-derotation osteotomy or reconstruction of the acetabular roof. Their acetabuloplasty consisted of a curved osteotomy about 1 centimeter above the upper rim of the acetabulum into the iliac wall, levering down of the bone to cover the anterolateral part of the head and maintenance of position with bone graft. In 187 hips in 162 patients treated surgically during the years 1947–1959, 72% of the patients were operated before the age of 3 years. The best results were obtained when the patients were operated on at an early age that they defined as up to 3 years. Beyond that the percentage of good results decreased greatly. Their system of assessment observed 68.3% favorable results in 171 hips treated by open reduction. As open reduction was resorted increasingly and by more surgeons, a major consideration involved what to do with the inverted limbus (fold of labrum plus capsule). Some such as Somerville [125–127] recommended excising the limbus, but most, including Salter [286] and Hall [287], strongly recommended saving the tissue, which in reality is an integral part of the normal structure of the hip, by freeing it and replacing it in its normal position supporting the head as a continuation of the lateral acetabular cartilage. Open reduction was accompanied by excision of the ligamentum teres, removal of the fibro-fatty tissue in the depths of the acetabulum, and opening of the inverted limbus so that it could be everted and freed from its previous position allowing the articular cartilage of the head to relate directly to that of the acetabulum with the repositioned labrum serving as a further superior support for the head. It was essential to free the inferior transverse acetabular ligament that was invariably stretched and tightened and served as a major block to reduc-

81

tion of the head. Capsular repair (capsulorrhaphy) was an integral part of the open reduction procedure. Renshaw also stressed the need for concentric reduction removing any interposed tissues between the femoral head and acetabular cartilages, by open reduction, if necessary [129].

1.12.3.1 Open Reduction by Medial Approach Mau et al. reported good to excellent results with the Ludloff anteromedial approach [288] in 1971, and Ferguson reported favorable results in 30 infants and children 2 weeks and 2 years of age using primary open reduction of the dislocated hip with the medial adductor approach [289]. This surgical approach was accepted slowly by many centers due to concern that visualization and correction of deforming soft tissue structures, located at the superior and lateral regions of the hip, were limited. Over the next couple of decades however, the medial approach came to be widely used in many centers and countries, although not in all. Akilapa [290] has recently performed a systematic review of the literature on the medial approach to open reduction of DDH concentrating on papers by Koizumi et al. [291], Matsushita et al. [292], Ucar et al. [293], Okano et al. [294], and Holman et al. [295] from 1996 to 2012 that best fit the criteria for evidence-based conclusions. These papers had sufficient criteria for inclusion on interpretation of the effectiveness of the medial approach. In the discussion, Akilapa et al. also referred to three of the older studies. The studies were retrospective with a wide and variable range of assessment criteria. The age groups at surgery were in the range of mean ages from 10 to 17 months with average follow-up 16–25 years. Assessments of outcome included hip function using validated outcome measurements, avascular necrosis, femoral head/acetabular development, Severin classification, and secondary operative procedures needed. In most of the studies, large numbers of patients were lost to follow-up. Age at surgery greater than 17 months, early re-dislocation, and avascular necrosis were all poor prognostic indicators. In the one study where two relatively large comparative groups were assessed (Ludloff medial exposure in 32 hips and a wide exposure with 360° capsulotomy), satisfactory outcomes (Severin I/II) in the medial approach were at 56% compared to 84% in the wide exposure group with 11/32, 34.4%, in the medial group needing additional operations, while none were needed in the wide exposure group. Avascular necrosis was very variable from low (5.5%) to high (43%). The percentages of Severin I/II outcomes varied from 40% to 60% with one at 80% for the medial approach open reductions. However, these outcomes were even less positive when the large numbers of patients who needed additional operations were classified as “unacceptable” (meaning essentially lower than Severin I/II grading) and lowered successful outcomes in the studies to 23%, 34.3%, 40%, and 59%. The rate of secondary operations in the

82

papers reviewed ranged from 11% to 50% and was 25–65% in three earlier studies. Akilapa correctly indicates that need for additional surgery (femoral or acetabular osteotomies) indicates that lack of formation of a normal hip followed medial approach open reduction. Koizumi et al. [291] and Matsushita et al. [292] “questioned the technical adequacy of the medial approach” since it greatly limited the ability to address extra-articular components such as the tight/ expanded posterosuperior capsule and contracted short external rotator muscles leading to eccentric reductions and poor femoral-acetabular congruity. While documentation of technique was variable, some performed partial labrum excision (in place of labral repositioning best done by the wide anterior approach) and only anteroinferior capsulotomy (since the posterosuperior capsule was not addressed via the medial route). Akilapa concluded based on this systematic literature review that “the medial approach as a solitary procedure” has significant limitations for inducing successful hip maturation [290]. He also stressed the importance of psoas tenotomy in the medial approach to enhance congruent femoral-acetabular repositioning. A review of studies over the past 25 years shows the practical aspects of the medial approach but repeatedly indicates the considerable problems and difficulties that follow open reduction (regardless of the approach). Rates of avascular necrosis, need for additional surgeries to manage re- dislocation and femoral-acetabular deformity requiring osteotomy, and good to fair Severin ratings at medium term follow-up (10–15 years) are cited frequently in the reviews, a few of which follow: Mergen et al. reported on 31 cases of dislocated hips in 1991 treated by open reduction with the Ferguson method [296]. Mean age at surgery was 12 months (3–33 months), and mean follow-up was at 7.9 years of age. AVN was seen in 9.7%, and 25% of hips had already had (or were scheduled for) additional corrective surgery due to insufficient radiologic results. Mankey et al. reported on 66 hips in 1993 treated by the Ludloff approach also at a mean age of 12 months (2–63 months) [297]. AVN resulted in seven hips (11%) with an increased prevalence in those operated after 24 months of age. Other problems included at least one redislocation and two subluxations within 4 weeks of surgery. Acetabular dysplasia failed to resolve in 33% of the hips, and subsequent pelvic osteotomy was performed. The authors felt however that the approach enabled direct access to the iliopsoas muscle and tendon, the transverse acetabular ligament (also a structure impeding reduction), and the constricted capsule. Morcuende et al. reported an open reduction of 93 congenitally dislocated hips in 1997 operated via the anteromedial approach [298]. Average age at surgery was 14 months (2–50 months), and evaluation was at 11 years (4–23 years)

1 Developmental Dysplasia of the Hip

post-surgery. Using the Severin grading system, 66 hips (71%) were good or excellent, 24 (26%) fair, and 3 (3%) poor. An inverted limbus at surgery led to a poor result (although almost all cases with complete dislocation would have the labrum inverted). AVN was extremely common, being present in 43% (Bucholz-Ogden type II 22 hips, 24%; type III 13 hips, 14%; type IV 3 hips, 3%; and unclassified 2 hips, 2%). Operation after 24 months of age was associated with a higher rate of femoral head growth disturbances. In spite of these findings, the authors considered the anteromedial approach useful at 24 months of age or less. Tumer et al. reported more favorable results in 1997 in 56 dislocated hips with medial open reduction [299]. Surgery was at an average of 11.2 months (2–25 months) and followup 8.1 years (3–17 years). Excellent or good (Severin I/II) results were seen in 98% although 11 hips (19%) needed secondary bone procedures and 5 (8.9%) showed AVN. Altay et al. used the medial approach for open reduction in 67 patients at a mean age of 14 months (7–23 months) and divided the patients into 2 groups for assessment; 29 were still not walking at time of surgery and 38 were [300]. The results in 44 were excellent or good (65.6%) and AVN occurred in 20 (24.1%). There were no significant differences in outcome however when the groups were compared for acetabular correction or when comparisons were made for age at surgery, operative side, or presence or absence of ossific nucleus. The authors felt that their results compared favorably with classic anterior/anterolateral approaches as well as being shorter in operative time. Konigsberg et al. reported satisfactory results in 2003 in 75% of 40 hips treated by medial open reduction [301]. Subsequent pelvic osteotomies were done in 8 hips (20%); AVN developed in 11 hips (27.5%). Gradings of Severin I/II were made in 30 hips (75%) with III in 6 (15%) and IV in 3 (7.5%) with 1 unclassified (2.5%). Citlak et al. reviewed 110 hips in 2013 treated with medial open reduction at a mean age of 17.7 months (6–48 months) and followed at 14 years (5–24 years) [302]. Radiological results were excellent or good (Severin I/II) in 86.4%, fair (III) in 10%, and poor (IV) in 3.6%. Acetabular development was better in those treated before 18 months of age. Additional operations were needed in 32 hips (29%); none were needed when surgery was done before age 12 months, but 48% of those operated late (19–24 months) required additional procedures. AVN rate was 17.3%. Gardner et al. performed an extensive systematic review of 14 articles reporting on medial open reduction surgery [303] in 2014. The mean follow-up was 10.9 years (2–28 years). Clinically significant AVN (types II–IV) was 20% (149/734), and the presence of AVN led to a higher incidence of unsatisfactory outcome at skeletal maturity of 55% versus 20% of hips with no AVN. The risk of developing clinically significant AVN was increased by surgery at less

1.12 The Development of Modern Treatment for CDH and DD

than 12 months of age when hips were immobilized at greater than 60° abduction. Hoellwarth et al. compared age-matched cohorts of patients with DDH undergoing medial and anterior hip approaches for open reduction [304]. There were 19 hips in each group (38 hips) with operation at an average age of 6 months (1.4–14.9 months). The outcome was based on incidence of AVN and need for additional surgery. This group had been proponents of the anterior approach and adopted the medial approach much later than other groups. Follow-up was at a mean of 6.2 years (1.8–11.7 years). AVN occurred at the same rates in the 2 groups: 4/18 (22%) in medial and 5/18 (28%) in lateral approaches. Need for additional surgery was also the same with both approaches. It was noted however that failure of closed reduction preceding open reduction led to a significant increase in need for additional corrective surgery, 7/12 (54%) failed closed reduction versus 4/26 (16%) without failed closed reduction.

1.12.3.2 R isk Factors for Failure After Open Reduction While the risk factors are outlined in the previous few sections, some studies directly addressed this issue in their investigations. Gholve et al. assessed results in ambulatory patients treated with open reductions of the hip late, at a mean age of 31.3 months [305]. There were 49 open reductions, 12 (24%) had open reduction only, 15 (31%) had concurrent pelvic osteotomy, 4 (8%) had femoral osteotomy, and 18 (37%) had both femoral and pelvic osteotomy. Repeat open reduction was needed in 4 (8%) at a mean of 5 months. A secondary procedure was needed in 24 (49%), and 8 of these needed 2–3 additional procedures. Of the 27 patients not having a concurrent femoral osteotomy, 19/27 (73%) needed a second procedure, while only 5/22 (23%) patients having femoral osteotomy at primary surgery required a secondary procedure. They concluded that open reduction in this age group should be accompanied by concurrent femoral osteotomy. Holman et al. assessed 66 hips having had medial and lateral open reductions and showed that 22/66 (33%) had Severin IV or worse outcomes. Surgery after 3 years of age had substantially poorer outcomes than those operated at younger ages [295]. All hips with AVN developed Severin IV gradings as did most hips that re-dislocated. Sankar et al. compared 22 hips with successful open reduction with 22 needing revision open reduction [306]. In their population of 421 patients undergoing open reduction, 25/421 (5.9%) developed re-dislocation at a mean of approximately 4 months post-surgery. Patients with right-sided or bilateral DDH, lower abduction angles of hips in postoperative spica casts (mean 39° versus 51°), dysmorphic misshapen femoral heads, and increased proximal femoral anteversion were at increased risk for failure.

83

In those operated beyond 18 months of age, the anteversion of the proximal femur and persisting acetabular dysplasia are taken into consideration. Here again opinions differ as to the approaches to be taken. Some feel that both of these areas of bone/cartilage deformation will correct once the head is appropriately repositioned and stabilized into the acetabulum. Others feel that a safer approach is to correct the bony deformities surgically as well. Salter [286] strongly recommended the innominate osteotomy, while Pemberton [307, 308] used the pericapsular iliac osteotomy after 18 months of age. Some preferred to correct the femur with a varus-derotation osteotomy and allow the acetabular correction to occur spontaneously once function had improved. The timing of such procedures will be discussed in greater detail below.

1.12.4 Acetabular Corrective Procedures for Treatment of Hip Dysplasia 1.12.4.1 Overview of Development of Acetabular Procedures: Three Basic Approaches Salter clarified the differing principles underlying the three surgical approaches designed to improve acetabular dysplasia [309]. (i) Acetabuloplasty refers to an incomplete osteotomy of the pelvis which levers the roof of the acetabulum downward with its new position maintained by a bone graft. The variants of this procedure used most commonly are the Pemberton pericapsular osteotomy, the Mittelmeier type osteotomy, and the Dega osteotomy. (ii) An extracapsular shelf operation refers to an operation in which the existing acetabular roof is extended outside the fibrous capsule of the joint to serve as a more effective buttress for the femoral head. (iii) The final procedure is a complete pelvic osteotomy extending into the sciatic notch, allowing the entire acetabulum to be redirected. Widely practiced variants of this technique are the innominate osteotomy of Salter [286], the medial displacement osteotomy of Chiari [310, 311], and the triple innominate osteotomy described initially by Steel [312]. The various acetabular procedures and the effectiveness of the results have been reviewed by Tönnis [154, 313–316]. Deficiency of the lateral acetabulum has long been recognized as one of the associated features of a CDH/ DDH, in particular one that has been left unreduced for more than a few months. Much attention has been directed to assessment and management of the acetabular deformity. As early as 1892, Koenig performed an acetabular shelf procedure to increase the depth of the acetabulum and its lateral coverage [317]. Many variations of this technical approach were proposed over the next several decades. The shelf procedure does not redi-

84

1 Developmental Dysplasia of the Hip

rect the acetabulum but augments it by providing a bone buttress superiorly and laterally. Until the early 1960s, major technical variations involved improved methods of containment utilizing shelf procedures to provide a buttress effect increasing the lateral extent of the acetabulum and thus increasing femoral head coverage and decreasing the CE angle. A major innovation by Salter in 1961 was development of the innominate osteotomy in which the entire acetabulum was tilted in an anterior and lateral direction to improve femoral head coverage [286]. He utilized the procedure as early as 18 months of age. Around the same time, there were other modifications of acetabular procedures described by Mittelmeier [318, 319] in Germany and Pemberton [307, 308] in the United States which involved transverse osteotomies just above

the capsule and levering down and forward of the acetabulum followed by interposition of bone graft to maintain position and stability.

Fig. 1.8 (a(i, ii)) The technique of the Salter innominate osteotomy after femoral head reduction is illustrated. (i) In figure at left, the Gigli saw is positioned within the greater sciatic notch and the line of osteotomy outlined; at right, the osteotomy has been completed and the repositioning is shown. (Reproduced with permission from Sales de Gauzy). The line of osteotomy is just above the joint. (ii) Postosteotomy the site has been opened to displace the acetabulum anteriorly and laterally to improve coverage of the femoral head. The bone graft from the iliac crest is inserted and stabilized with two Kirschner wires. Anteroposterior and lateral views post-osteotomy are shown. (Reproduced with permission from Wedge et al. from JBJS Am with permission). (iii) An example of the effectiveness of the innominate osteotomy in correcting acetabular dysplasia is shown. (a) Anteroposterior x-ray of the pelvis shows acetabular dysplasia with lateral subluxation of the proximal femur on the left at 6 months of age. The secondary ossification center on the left is positioned laterally in the lower outer quadrant. The acetabular index measures 36° on the left hip. (b) Anteroposterior radiograph at 2 years of age shows persisting lateral subluxation of the left femur with acetabular dysplasia. The secondary ossification center is now considerably smaller on the left than on the right. (c) Innominate osteotomy was performed and stabilized with two K-wires. The distal fragment was displaced somewhat laterally and tilted anteriorly. (d) Anteroposterior x-ray at 3 years of age shows normal containment of the left femoral head, increased size of the secondary ossification center, and correction of the acetabular dysplasia. (e) Lateral radiograph at 8 years of age shows a normal hip on the left with excellent acetabular development. (f) Anteroposterior hip radiograph at 8 years of age highlights normal hip on the left. (b) Plane of osteotomy and iliac crest bone graft for the Pemberton pericapsular innominate osteotomy are illustrated. The lateral iliac entry point is higher than for periacetabular osteotomies more commonly used at present. (Reproduced from Huang et al. JBJS Am with permission). (c(i-iii)) Planes of osteotomy for the Dega/San Diego periacetabular osteotomy are illustrated. Progressive stages of the procedure are shown in figures i, ii, and iii. The curvilinear osteotomy remains above the triradiate cartilage at all times. It begins just above the capsular attachment at the lateral iliac cortex. It is also curved so as to keep close to the acetabular surface without penetrating it. Iliac crest graft holds position after leveraging into corrected acetab-

ular orientation. The full-thickness medial/lateral cortical cuts are performed anteriorly and posteriorly at the sciatic notch with the intervening medial cortex left intact for hinging and post-osteotomy stability. Three pieces of iliac crest cortico-cancellous bone hold the osteotomy site open; their size controls anteroposterior positioning. (Reproduced from McNerney et al. JBJS Am with permission). (d(i-iii)) Planes of osteotomy for the Dega/Pittsburg periacetabular osteotomy are shown along with post-positioning bone grafts for stabilization in figures i, ii, and iii. The posterior inner iliac cortex adjacent to the sciatic notch remains intact in this technique. (Reproduced from Grudziak and Ward JBJS Am with permission). (e(i-v)) The Tönnis periacetabular osteotomy is illustrated. (i) Radiograph demonstrates dysplastic acetabulum. (ii) An arthrogram of a subluxed hip with acetabular dysplasia is shown. (iii) Position of the osteotome in relation to the acetabulum is shown. (iv) Femoral head coverage is markedly improved post-osteotomy and deproteinized bone wedge insertion. (v) Normal radiographic hip anatomy has been restored 4.5 years postsurgery. (ii, iv, and v reproduced with permission from Tönnis, Clin Orthop, and Relat Res, and I and iii reproduced with permission from Tönnis, Congenital Dysplasia and Dislocation of the Hip, SpringerVerlag, 1987). (f) Steps in performance of the Chiari osteotomy are shown. (Reproduced from JBJS Am, Hiroshi et al., with permission). (g) Illustration shows a periacetabular osteotomy for hip dysplasia from 1915 (Albee). Note minimal depth of the osteotomy cut; the lateral acetabular roof is tilted downward to improve coverage and held open with a bone graft for stabilization. The greater trochanter has been returned to a normal position surgically to tighten its muscle attachments. Capsular shortening associated with the correction is illustrated. (Reproduced from FH Albee in Bone-Graft Surgery, Philadelphia, WB Saunders, 1915). (h) Shelf procedure is shown. (Reproduced with permission from Gillingham et al., JAAOS, 1999). (i) The most surgically direct approach to persisting acetabular dysplasia beyond 2 years of age involves both acetabular and proximal femoral surgical corrections. In a combined procedure, the acetabulum is levered downward with the correction held by a bone graft (acetabuloplasty), and the proximal femur is redirected into the depths of the acetabulum by a combined derotation and varus osteotomy. (Reproduced with permission from Fritsch et al., Current Orthopedic Practice (Clin Orthop Relat Res) 1996)

1.12.4.2 Acetabular Procedures (a) Innominate Osteotomy (Salter). The Salter innominate osteotomy can provide excellent femoral head coverage when the procedure is done from 18 months of age up to the age of 6 or 7 years [286, 320, 321]. A pelvic osteotomy is performed, and the entire acetabulum is redirected to improve both lateral and anterior coverage (Fig. 1.8a). Rotation occurs through the cartilaginous symphysis pubis, and only one transverse cut in the pelvic bone is needed. A triangular wedge of bone from the anterior superior iliac spine region is removed, inserted

1.12 The Development of Modern Treatment for CDH and DD

a i

ii

85

86 iii a

1 Developmental Dysplasia of the Hip

b

c

e

Fig. 1.8 (continued)

d

f

1.12 The Development of Modern Treatment for CDH and DD

b bone graft donor site

bone graft

Fig. 1.8 (continued)

87

88

1 Developmental Dysplasia of the Hip

c i

Osteotome 1 2 Osteotomy

110º

ii

Downward leverage

iii

TRIANGULAR BONE GRAFTS

SYMMETRICALLY HINGED ACETABULOPLASTY

Fig. 1.8 (continued)

1.12 The Development of Modern Treatment for CDH and DD

d

e

Fig. 1.8 (continued)

89

90

1 Developmental Dysplasia of the Hip

f Kirschner wire

Joint capsule

Anterior superior iliac spine

Osteotomy line

Greater sciatic notch

Kirschner wire

Cancellous screw

Fig. 1.8 (continued)

Anterior inferior iliac spine Acetabular edge

1.12 The Development of Modern Treatment for CDH and DD

g

91

h

i B

A

C

C

A

B

Fig. 1.8 (continued)

92

1 Developmental Dysplasia of the Hip

into the opened gap, and stabilized with a Kirschner wire to stabilize the pelvis and maintain the corrected position. Salter stressed that “the basic abnormality responsible for instability of the reduced congenital dislocation is the abnormal direction in which the entire acetabulum faces.” Instead of facing downward, it is directed anterolaterally such that the femoral head is inadequately covered anteriorly when the hip is extended and laterally when it is adducted. The principle of innominate osteotomy is redirection of the entire acetabulum to enhance stability in the upright functional position. The p rocedure was widely although not universally adopted after its initial description in 1961 and served as a major improvement in treatment of CDH. The prerequisites for acetabular redirection, as defined by Salter and Dubos, include (1) necessity of bringing the head of the femur to a level opposite the acetabulum, (2) release of contractures of the adductor and iliopsoas muscles, (3) complete concentric reduction of the femoral head within the true acetabulum whether by closed or open reduction, (4) congruity of the hip joint, (5) good preoperative range of motion, and (6) preferred age of intervention between 18 months and 6 years [320]. Below 18 months of age, acetabular operations were felt to be rarely necessary since repositioning of the head into the acetabulum by closed or open means will generally allow for spontaneous acetabular correction to a normal range. By 18 months of age, normal bone development of the acetabulum and femoral head are no longer assured even with prolonged retention of the hip in the reduced position in the opinion of some, including Schwartz. Hall also stressed that the first and most essential prerequisite in association with performance of the innominate osteotomy was complete concentric reduction of the femoral head in the true acetabulum [287]. The innominate osteotomy does not change the shape or the capacity of the acetabulum but changes its direction so that the head of the femur is covered adequately in the standing position. Salter and colleagues have illustrated the technique again [321]. Additional procedures have been developed to overcome the fact that rotation of the acetabulum at the symphysis pubis is more difficult due to diminished flexibility after 6 or 7 years of age. Osteotomies have been devised in which the ilium and the superior and inferior pubic rami are cut to allow a freely floating pelvic segment to be tilted into appropriate position (the Steel osteotomy) [312]. Sutherland and Greenfield performed an osteotomy just lateral to the symphysis pubis and removed a small piece of bone to allow for easier rotation of the acetabulum and some medial displacement [322]. Another innovation was the acetabular “dial procedure,” done at skeletal maturity, which involves cutting the

bone around the entire acetabulum close to the articular cartilage and rotating the bony acetabulum to a better position but not interrupting the medial rim of the pelvis. At skeletal maturity acetabular dysplasia is treated most commonly by a periacetabular osteotomy as described by Ganz et al. [323]. (b) Pericapsular Osteotomy of Ilium (Pemberton). The Pemberton osteotomy not only redirects the sloping dysplastic acetabular roof but also diminishes the capacity of the acetabulum [307, 308]. This is essentially a hinge- type procedure. Pemberton described a pericapsular osteotomy of the ilium to correct directly the acetabular dysplasia and CDH in 1965. His rationale for the procedure was the desirability of rotating the acetabulum forward and laterally to close the anterior defect by shifting the more posterior and medial portions of the socket over the femoral head to provide a good weight-bearing surface. The procedure was designed to make the acetabulum somewhat smaller without undue acetabular distortion to stabilize the hip. The acetabular roof is rotated around the femoral head since one of the rationales for the procedure is the fact that the femoral head is small relative to the acetabulum in CDH/DDH. The procedure is completed posteriorly through the posterior limb of the triradiate cartilage. There is sufficient plasticity in the horizontal arms of the triradiate cartilage to permit the downward, anterior, and lateral displacement of the roof of the acetabulum and to prevent fracture or excessive distortion of the acetabular joint surface. The osteotomy is done separately through both inner and outer tables of the ilium. Pemberton and others have stressed that the osteotomy must not go into the sciatic notch since a fracture into the notch would greatly destabilize the acetabular segment. The earliest recommended time of intervention is 18 months with the upper age limit approximately 8–10 years, dependent on the presence and plasticity of the triradiate cartilage. The best results occurred in those under 4 years of age since the procedure is easier to perform and acetabular remodeling is greater than at later ages. A high level of acceptable results was also achieved in the group from 4 to 7 years of age, while those between 8 and 12 years had a higher proportion of fair results. Pemberton was in agreement with Salter and others that the tendency for acetabular correction based on femoral repositioning was limited by 18 months of age and that operation as early as 18 months and preferably between 18 months and 4 years of age on the acetabulum itself produced excellent results. He felt that “the defect in the acetabulum allowing the head to come out was not one of a relatively shallow acetabulum but was rather an abnormal direction with the acetabular surface directed forward and laterally as well as being comparatively shallow.”

1.12 The Development of Modern Treatment for CDH and DD

The operation was based on the principle that “since the triradiate cartilage is the only flexible structure in which the size and shape of the acetabulum could be changed it was determined that this would be used as a hinge. If the iliac portion of the acetabular roof was detached, the iliopubic and ileo-ischial limbs of the triradiate cartilage could be used as a variable hinge to wrap this iliac portion around the femoral head either anteriorly, laterally, or a combination of both directions.” Once the acetabulum is repositioned, it is locked into position with a bone graft from the iliac crest. Early reviews of the Pemberton procedure by Coleman [324] and McKay [325] were favorable. They stressed the prerequisites of the pericapsular iliac osteotomy: (1) hip must be concentrically reduced or reducible at open operation; (2) hip must have a satisfactory range of motion; and (3) congruity of the femoral head and acetabulum must exist. The incomplete pericapsular osteotomy rotates the anterior and superior portion of the acetabulum forward, laterally, and downward utilizing the triradiate cartilage as its fulcrum. This is different compared to the complete innominate osteotomy that rotates the entire acetabulum with the site of rotation through the pubic symphysis. Salter and others were concerned that the incomplete pericapsular osteotomy of Pemberton that hinged at the triradiate cartilage medially would of necessity produce an angulation of the acetabular articular surface with altered internal congruity. Erturk et al. have recently compared 47 hips treated with Salter’s innominate osteotomy and 50 treated with the Pemberton procedure at 5 years post-surgery [326]. The acetabular index, center-edge (CE) angle of Wiberg, and femoral head migration (Reimer) did not differ between the groups at final follow-up, but the acetabular depth ratio (ADR) was slightly greater in the Pemberton group. Wade et al. felt they had superior results with stability of the Pemberton osteotomy using a contoured iliac crest allograft rather than the autograft [327]. (c) Dega, Chiari, and Other Acetabuloplasties. Several other acetabular osteotomies have been developed in Europe including those of Dega [328–330], Lance [331], Mittelmeier et al. [318, 319], and Chiari [332]. Schulze et al. commented favorably on the Dega procedure but were tending to switch to the Pemberton technique [333]. The Mittelmeier procedure is an acetabuloplasty combined with varus-derotation osteotomy to correct bone deformation once there is complete femoral head reduction into the acetabulum [see Tönnis, [D., 1987. Congenital Dysplasia and Dislocation of the Hip. Springer-Verlag`, Berlin #572]]. Tönnis’ modification involved osteotomy of the acetabulum under fluoroscopic control “5 mm” from the roof directing the osteo-

93

tome to the most posterior point of the triradiate cartilage. Only a small bony bridge was left intact at the innermost posterior wall directly above the triradiate cartilage. The acetabular roof was levered down and held in position by a wedge shaped bone graft. All of these acetabular procedures, often combined with proximal femoral varus-derotation osteotomies, are most effective in the first 6 years of life. Over the past 15 years, the Dega procedure has been widely used as the primary initial acetabuloplasty in European and North American centers. Some of this shift from the innominate osteotomy is due to the relatively recent recognition that the innominate can tilt the acetabulum too far forward leading to adolescent age impingement syndromes. The Dega osteotomy, developed in Poland, was in use in Europe for several years before its acceptance in North America. Part of the reason for this was the unclear and even varying translations, or at least understandings, of the technical aspects of the procedure. Mubarak et al. described their use of the procedure in 1992 with favorable results, and since then it has been increasingly widely used [329]. Grudziak and Ward in 2001 provided a detailed description of their operative technique for the Dega procedure [330]. It is an acetabular redirection procedure similar to the Pemberton. Some descriptions have the medical cut exiting or hinging through the medial margin of the triradiate cartilage (which actually is the Pemberton) but for most of it is a curvilinear cut above the acetabulum passing through both cortices anteriorly above the triradiate cartilage but ending posteriorly 1–1.5 cm in front of the sciatic notch and leaving the adjacent inner cortex intact where the hinging occurs. The osteotomy site is then hinged open to improve anterior and lateral coverage as needed with a bone wedge inserted to hold position from either the anterior iliac crest or the femur if a varus-derotation osteotomy has been done at the same time. The Chiari osteotomy passing obliquely upward from the lateral margin of the acetabulum through the inner iliac wall is designed to medialize the femoral head and pelvis leaving the superior iliac segment to support the head with the capsule interposed [311, 332]. It is infrequently used today in North American centers but continues to be reported on from many European centers. It is primarily indicated for older patients beyond 10–12 years of age and in early adults past skeletal maturity where considerable acetabular deformity is present. Ito et al. report on 173 Chiari procedures done at average age of 20 years (range 9–54) with 72% satisfactory results. The 30-year hip survival rate (prior to total hip arthroplasty) was 85.9% with much better results observed when the procedure was done at time of early osteoarthritis compared with advanced osteoarthritis [334]. Karami et al. reported on 20 Chiari osteotomies done at an average age of 12.6 years with a short follow-up of 54 months [335]. The

94

results were excellent in 11, good in 8, and fair in 1, but graft resorption was seen in 25%. The authors cautioned that there were only rare indications for the procedure and that the Chiari osteotomy in this age group should only be the last treatment option. Combined Pelvic Osteotomy. For severe acetabular underdevelopment in children at a mean of 4 years of age, Rejholec describes a double or combined pelvic osteotomy involving a Salter or Pemberton procedure to which is added a Lance acetabuloplasty [336]. Teot et al. developed an experimental biologic augmentation acetabuloplasty with a vascularized pedicle graft transferred from the iliac crest utilizing both the growth cartilage of the crest plus some adjacent iliac bone [337]. Another experimental approach to acetabular correction was the application of physeal distraction with an external fixator to the triradiate cartilage in immature dogs 2.5–4 months of age. The procedure was designed to enlarge the acetabulum both in depth and width [338]. (d) Results of Acetabular Procedures (i) Pemberton Osteotomy. Reviews by independent groups have supported the value of the Pemberton procedure. Eyre-Brook et al. reported on 37 procedures and favored this type of acetabular procedure when an operative approach to the acetabulum was warranted [339]. A large majority of their procedures, which were either isolated or combined with a primary open reduction, were performed between 1 and 3.5 years of age. One of the concerns with this procedure had been that correction was obtained through the triradiate cartilage with the possible complications of premature fusion and growth irregularity of the acetabulum creating an incongruous shape that would not remodel. Eyre-Brook et al. specifically studied the triradiate cartilage in followup radiographs and felt that early closure or other abnormality was not seen. The hinge effect occurred owing to a fracture-separation at the triradiate cartilage with movement occurring either from the fracture-separation laterally into the acetabulum or medially toward the pelvis. CT scanning or MR imaging would demonstrate the plane of separation most clearly. Faciszewski et al. reviewed 52 Pemberton procedures with an average follow-up of 10 years [340]. The average age of their patients at operation was 4 years (range 3–10 years). The upper age limit was 10 years because the triradiate cartilage was relatively thin after that age. They identified no radiographic evidence of premature arrest of the triradiate cartilage or of acetabular chondrolysis and concluded that the pericapsular osteotomy was a safe and effective procedure for the treatment of residual acetabular dysplasia.

1 Developmental Dysplasia of the Hip

Plaster et al. reported premature arrest of the triradiate cartilage after a Gill acetabuloplasty at 14 months of age, but they attributed it to placement of bone graft across the physis itself [341]. Acetabular dysplasia worsened with time. ( ii) Innominate Osteotomy (Salter): Short- and Intermediate-Term Results. Results by Salter and colleagues in Toronto showed an extremely favorable response to the innominate osteotomy procedure from the initial report in 1961 onward [112, 203, 286, 320]. Many other centers achieved excellent results with this procedure as well although some had difficulty. Salter and colleagues stressed the need to adhere stringently to the prerequisites and indications for the procedure as well as close attention to technical detail. Salter and Dubos using open reduction and innominate osteotomy for complete dislocation noted 94% excellent or good results with treatment from 1.5 to 4 years of age which diminished to 57% excellent/good at the age of 4–10 years [320]. With osteotomy alone for subluxation, excellent results were achieved in 94% of 16 hips 1.5 to 4 years of age and 58% excellent and 33% good in 12 hips 4–16 years of age. Roth et al. had an 85% excellent result rate with primary open reduction/innominate osteotomy in 65 dislocated hips in children 1.5 to 4 years of age at initial treatment and a 92% excellent rate in 12 hips with osteotomy alone for subluxation at 1.5 to 4 years [342]. Salter, in his initial report in 1961, described results in 25 hips between the ages of 18 months and 6 years [286]. Ninety-two percent of those patients had excellent (80%) or good (12%) results by the Severin classification. Heine and Felske-Adler studied 43 hips treated with the innominate osteotomy and open reduction under 4 years of age and found the hips to have normal or slightly abnormal acetabular indices although when assessed in terms of the CE angle and other hip indices, the results were somewhat disappointing with only 34% considered good or fair [343]. Some of the less than excellent results were due to the relatively high rate of preoperative AVN and to the fact that many patients had had other procedures. Blamoutier and Carlioz studied the results of their Salter procedures in 43 hips with the average follow-up of 10 years [344]. Overall results were satisfactory in 60% of cases with much better results in those under 5 years of age at surgery. Many of the failures were related to technical difficulties although imperfect results seemed to follow some cases where no intraoperative problem had been identified. They eliminated patients who had previously had AVN as a result of earlier treatment. The average age at operation was 3 years

1.12 The Development of Modern Treatment for CDH and DD

10 months with a range between 18 months and 10 years. Complications were infrequent and no case of AVN attributable to the procedure was seen. According to the classification of Severin, 83.3% of the hips were classified as an excellent group I result. With a more rigorous classification, however, based on several radiographic indices, the number of normal hips diminished to 60.4%. The problematic results involved decreased ranges of motion, persisting dysplasia of the acetabulum, and excessive coverage of the hip. Blamoutier and Carlioz concluded that the Salter osteotomy was useful over the long-term for correction of residual acetabular dysplasia in particular if operation was done before 5 years of age. The operation was technically demanding, however, and if imperfectly performed was unlikely to lead to a good result. Even with the procedure, however, a considerable number of patients could be seen with persisting dysplasia several years afterward. When patients were followed for 10 years or more, many reports of lesser results were seen. Mader et al., assessing 20 cases, did not observe any with fully normal sphericity and concentricity [345]. Morel studied 23 cases and found 60.8% with good results [346]. Fournet-Fayard et al. noted 76.7% good and excellent results [347]. In a review by Hall and colleagues, 29 hips operated between the ages of 18 months and 5 years for CDH were assessed [348]. All were felt to have good or excellent radiographic results using the Severin classification. Three patients required reoperation for failed initial procedures. The average length of follow-up was 9 years, 3 months. The immediate postoperative mean acetabular index was 12°, the same as the long-term index. The center- edge angle increased immediately following surgery to a mean of 36° and the long-term result had a mean of 39°. In the Severin radiographic classification, 25 of 29 hips were class I, 3 were class II, and only 1 was class IV. McKay reported 73% of 26 patients between the ages of 18 months and 6 years with Severin class I or II radiographic results after innominate osteotomy [325]. Crellin reviewed 25 hips in 21 patients treated by innominate osteotomy between 14 months and 5 years, 3 months with 72% having Severin I and 24% Severin II gradings [349]. Gallien et al. assessed 43 hips treated between the ages of 14 months and 4 years with 68% having good or excellent results [350]. They reported a 5% incidence of AVN. Barrett et al. assessed 42 hips treated between 18 months and 4 years 2 months by either innominate osteotomy alone or combined with open reduction and noted 88% with either Severin I (62%) or II (26%) gradings [351]. Hansson et al. reported on 83 procedures with overall radiographic results good or excellent in only 41% and fair or poor in 59% [352]. The study reviewed the first 15 years of experience in Sweden with the procedure in which 26 surgeons were involved. Their best results were obtained in hips with subluxation that had not been previously treated or treated only with closed reduction. The poor-

95

est results were obtained in hips with residual subluxation or dislocation after previous operation. The results in this Swedish series, however, were best when the procedure was performed before the age of 5 years. Windhager et al. reviewed 63 innominate osteotomies done at a mean age of 4.1 years (range 18 months to 18 years) and followed for a mean of 15.7 years [353]. The failure rate was 29%: 25% in those operated before 4 years of age and 41% in those operated at a later age. Improvement with time helped some in the younger group but not in the older group. Those with mild or moderate dysplasia pre-surgery did better than those with severe pathologic dysplasia. Long-Term Results Salter and colleagues did a very long- term review of 80 hips in patients who had same stage open reduction and innominate osteotomy a minimum of 40 years previously [321, 354]. The same operative procedure was used consistently. The average age at time of surgery was 2.8 years (1.5–5 years). The average follow-up was 43.3 years (40–48 years), and the average age at follow-up was 45.8 years (42–51 years). Clinical and radiographic scoring systems were used, but the definitive index of a poor result was performances of a total hip arthroplasty. Total hip arthroplasty had been done in 24/77 patients (3 died from unrelated causes) or 31% surviving beyond 40 years from the index procedure. The first hip failure (arthroplasty) was at 30 years with rapid decline between 40 and 45 years post-surgery; hip survival rates 30, 40, and 45 years post-surgery were 99%, 86%, and 54%, respectively. Bilateral cases were at 2.9 times greater risk of subsequent hip replacement than unilateral cases. The rate of postoperative complications was 46% in hips needing a joint replacement and only 15% in hips surviving. The complications involved recurrent dislocation needing repeat open reduction, residual subluxation needing abduction casting or revision innominate osteotomy, supracondylar femoral fracture, and avascular necrosis. In summary, good results were maintained for 30 years after which osteoarthritis progressed quickly in many leading to a 31% hip replacement rate with 54% surviving at 45 years. Body mass was not related to outcome nor was age at surgery, but bilateral involvement and postoperative complications predisposed to poorer results. Bohm and Brzuske. A long-term study of the results of innominate osteotomy was done by Bohm and Brzuske in Germany [355]. Between 1963 and 1972, 74 innominate osteotomies were done for DDH using Salter’s technique with 73 available for long-term review. The mean age at surgery was 4.1 years (1.3–8.8), and the mean follow-up was 30.9 years (26.2–35.4). While 48 had innominate osteotomy alone, others had variable combinations of simultaneous open reduction (12) and various timings for intertrochanteric osteotomies. There were seven true revisions (reoperations)

96

involving acetabuloplasty at 5.4 years, triple innominate osteotomy at 25.9 years, and five total hip arthroplasties at 18, 23, 24, 28, and 28 years. These 7 true revisions plus 8 hips with a Harris hip score 75th %

>50th %

48

65

87

95

45

62

79

89

47

64

84

93

Weight charted on the Ross Laboratory charts derived from the Hamill PVV, et al. National Center for Health Statistics percentiles. Am J Clin Nutr. 1979; 32:607–29. [Based on 167 patients with accurate data]

84% were above the 75th percentile; and 93% were above the 50th percentile. The percentages were virtually the same in males and females. The determinations were made on 167 patients, 114 males and 53 females (2.2 to 1) (Table 3.5). There was no difference in the weight percentile distribution

352

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

Table 3.6 Weight percentile distribution of patients with unilateral and bilateral involvement Weight percentile at initial involvement Unilateral Male Female

Bilateral Male Female

>95th

90–95th

60–89th

60°) were seen compared to ent, is a sign of chronicity greater than 3 weeks [13, 14]. the group with a normal head-neck offset [154]. These findings help determine acute versus chronic versus 3.2.9.5 Pistol-Grip Deformity in Relation acute-on-chronic hip status. [See Sect. 3.2.7.3].

3.2.9.3 Computerized Axial Tomography (CT Scans) CT is not routinely used in SCFE. CT scanning defines with great clarity the slippage of the head and the retroversion of the neck. Guzzanti and Falciglia showed, however, that careful application of plain radiographic techniques produced measurements of slip severity with a high level of concordance with CT measurements [152].

to Femoroacetabular Impingement (FAI) and Osteoarthritis (OA) of the Hip An osteological study of 2665 adult human skeletons assessed the number of femurs with the proximal head-neck deformity (pistol-grip/Stulberg or femoral head tilt/Murray) indicative of a childhood SCFE [155]. It also related the proximal femurs to associated acetabular changes indicative of an osteoarthrosis. The prevalence of post-slip morphology in the large series was 8% (215/2665), and severe osteoarthrosis was prevalent with

358

post-slip morphology in 38% (116/306) of hips versus 26% in matched controls (79/306). High-grade osteoarthrosis was seen in 68% (63/93) of hips with post-slip morphology compared with 48% (45/93) of control hips. Flattening of the acetabulum anteriorly and cystic degeneration in the neck were other signs of osteoarthrosis. Giles et al. studied preoperative CT images from 81 hips with OA undergoing hip resurfacing at an average age of 52 years [156]. Evidence of childhood SCFE was seen in 90% of hips on the oblique axial view and 95% of hips on the radial view based on pathologically increased alpha angles. In the oblique axial and radial planes, 60% and 68%, respectively, were abnormal for the alpha angle, head-neck tilt, and anterior offset ratio that strongly suggested the SFE morphology. The findings supported the concept of the association of SCFE with cam-type FAI leading to OA.

3.2.9.6 Other Methods Bone scintigraphy can detect early femoral head avascularity. Arthrography is rarely if ever used at present to assess slips.

3.2.10 Treatment 3.2.10.1 Goals of Treatment. Detailed Overview The initial treatment in slipped capital femoral epiphysis is designed primarily to stabilize the head on the neck, preventing further slippage by inducing premature growth plate fusion. A second goal for some but not all practitioners is to reposition the head to restore partially or fully the anatomy of the proximal femur in those instances where displacement has been moderate, severe, or complete. Treatment should not induce avascular necrosis of the femoral head or chondrolysis of the articular cartilage of the hip joint so that, over the longer term, treatment leads to a hip at skeletal maturity with a femoral head-neck-shaft relationship as close to anatomic normal as possible, normal head sphericity, viable well vascularized femoral head bone, normally shaped and structured articular cartilage, and no clinically significant limb shortness. In patients with a pre-slip or mild to moderate slipping, prevention of further slippage has appeared until recently to be sufficient. The most common approach to causing premature fusion of the proximal femoral capital epiphysis has been primarily by in situ pinning using a single screw for transphyseal fixation to stabilize the femoral head on the neck. In severe and complete slips, the stabilization procedure was being sometimes performed (i) in association with open reduction and cervical resection (of physis and adjacent metaphysis/neck), sometimes referred to as wedge or shortening osteotomy or femoral head realignment, designed to restore the anatomic position of the femoral head to the femoral neck and shaft or (ii) with another approach to repositioning, compensatory basicervical or intertrochanteric osteotomy, that could be done at the same time as or

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

several months after, in situ stabilization. It is widely recognized that at time of diagnosis, distinction should be made as to whether the slip is acute/unstable or chronic/stable and whether there is an element of an acute-on-chronic slippage. These distinctions affect the treatments chosen. Variable opinions continued as to what degree of slippage mandated anatomic repositioning of the head in relation to the neck, which technique should be used to reposition the head, and when, if ever, repositioning should be done. By the year 2000, there was good recognition of the need in each case of SCFE to diagnose an acute/unstable hip or a chronic/stable hip, to recognize an acute-on-chronic slip, and then to treat these variants in distinct ways. With greater recognition of femoroacetabular impingement (FAI) in SCFE, more aggressive management of it, and the fact that many early-adult and certainly mid-adult cases of hip osteoarthritis needing total hip arthroplasty were secondary to childhood SCFE, there was a reassessment in several centers of the treatment approach to the various types and severities of the disorder. Many have continued to use in situ pinning for the much more frequent (85% of cases of SCFE) stable/chronic slips with mild, moderate, and even severe slippage. Others have felt that many cases of moderate and even mild slips, as well as severely displaced slips, predisposed to an early onset FAI and that more aggressive management in adolescence at time of initial treatment regardless of the severity of the slip could be beneficial long term. Techniques of hip arthroscopy for osteochondroplasty of head-neck offset in mild cases and open reduction with the modified Dunn technique using surgical hip dislocation and anatomic reduction with or without cervical osteotomy for moderate and severe slips were used increasingly by many. These were considered by others to still have risks too extensive to take, and they continued to opt for in situ pinning, reserving the open approaches either only for severe cases, or, in many instances resorting to repositioning osteotomies at subcapital/cervical, basicervical, or inter/ subtrochanteric levels a considerable period of time after in situ pinning and only when symptoms of FAI were becoming apparent. For the acute/unstable hips (15% of cases of SCFE), some continued with in situ pinning as an emergency within 8–24 h with no attempt at reduction, while others performed gentle (closed or open) reduction with hip joint aspiration or arthrotomy to restore anatomy and then proceed with in situ pinning. Over the next several sections, we will outline the evolution of treatment approaches over several decades.

3.2.10.2 E volution of Treatment Modalities up to the 1960s Surgical Intervention, Closed Manipulation, and Hip Spica Immobilization The first nail used for internal fixation of SCFE is attributed to Sturrock in 1894, but owing to infection, the device had to

3.2 Slipped Capital Femoral Epiphysis

be removed at 2 days so early effectiveness was not defined [157]. The mention of nailing was in an article on traumatic separation of epiphyses, and the proximal femoral epiphyseal separation was either a displaced acute slip or a pure traumatic fracture-separation. The first osteotomy to correct deformity was reported by Keetley in 1888; he performed subtrochanteric wedge osteotomy of the femur [21]. A later paper by Keetley on coxa vara in 1900 outlined evolving concepts of the disorder, most of which were cases of SCFE, and the surgical approach of valgus osteotomy for correction [22]. He took great care in outlining his claims to priority for surgical correction by osteotomy and for reporting the first case “in which the nature and seat of the disease [coxa vara] were diagnosed correctly during life and before operation.” There was early awareness of the corrective value of osteotomy, and procedures were done in the neck, intertrochanteric region, and shaft followed by positioning of the distal fragment in abduction, internal rotation, and flexion. Helbing illustrated several of the osteotomy procedures described by his contemporaries [3]. Whitman was an early advocate of manipulation to reduce the displacement either with general anesthesia or in traction followed by hip spica immobilization [158]. Closed manipulative reduction with the goal of complete reduction was still being recommended by some decades later [159]. The tendency to attempt repositioning of the head as though the patient had a displaced fracture was common, but many poor results were recognized early on. Reviews of Treatment Methods and Results Kleinberg and Buchman described the various treatments used in 1936 [43]. It was felt that conservative treatment could be effective but that it was advisable primarily in the pre-slipping or very mild slipping stages. In their opinion rest with prevention of all weight bearing in those instances would produce excellent results. Manipulative treatment was problematic since it was based on an incorrect assumption that the disorder was an incomplete fracture of the neck of the femur. Whitman brought about reduction by manipulation of the femur into extreme abduction and internal rotation, but results were felt to have been unsatisfactory. The authors pointed out that even in mild slip situations, reduction rarely resulted in perfect repositioning. They felt that even when the surgeon considered the manipulation to be gentle, extreme force was placed on the femoral head and neck causing extensive damage to the circulation to the femoral head. They indicated: “once the reduction is completed the possibilities for re-establishment of the circulation of the femoral head are rather meager.” At multiple arthrotomies the head was always firmly fixed to the neck, and displacement was possible only by means of osteotome and mallet. The only possible results of manipulations were further crushing of the head and increased damage to the blood supply. Manipulative treatments would be of value only in cases of acute slipping, and even then a manipulative reduc-

359

tion would have to be executed deliberately and with gentleness. For all other forms, the manipulative treatment was felt to be inadvisable because “(1) the reduction may be impossible; (2) the reduction is often incomplete and illusory; (3) the circulation of the head is likely to be disturbed as evidenced by late deformity of the head (avascular necrosis); (4) traumatic arthritis and even ankylosis may ensue; and (5) no change is effected in the pathologic epiphyseal plate.” While recommending operative treatment, they felt that open reduction and instrumentation were also inappropriate since there was no provision for the revascularization of the deranged epiphyseal plate tissue. There was potential value of drilling of the epiphyseal growth plate “for the purpose of re-establishing circulation and causing premature ossification of the part.” That approach was limited to the pre-slipping or mildly slipping stages but had been used in several cases with good results. Most of their attention surgically was directed to correction of moderate to severe deformity and to establishment of the circulation across the physis to produce early fusion between the head and neck. This was best accomplished by an operative resection of the epiphyseal plate and realignment to allow contact between the cancellous bone of the head and neck of the femur. Open reduction involved a wedge resection of the physis and the adjacent femoral neck, curettage of cartilage from the head region, and placement of the head onto the reshaped neck followed by cast immobilization. They actually dislocated the head, and clearly relatively little attention was paid to maintaining vascularity although comments on the need for it were made. Postoperative treatment was prolonged until revascularization of the head occurred. In summary, many of the principles of current treatment were recognized in the early decades of the twentieth century. Slipped epiphyses were not fractures and could not be treated as such. The treatment should be atraumatic so as to correct deformity and establish circulation between the head and neck of the femur. Treatment of the pre-slipping stage was by bed rest or by brace although transphyseal drilling of the head and neck could be effective. In a report on 44 cases published in 1945, Moore stressed the need to avoid trauma to the epiphysis by operation or manipulation as much as possible and strongly felt that the risk of damage to the blood supply to the epiphysis following surgery outweighed the advantages of earlier mobilization [160]. Virtually all his patients were treated conservatively either with bed rest alone or with hip spica immobilization in which the limb was positioned in wide abduction and internal rotation but without efforts at specific reduction of deformity. The limb was protected until transphyseal ossification of the growth plate in particular in its central regions was noted radiographically. AVN occurred in only 1 of 29 patients with minimal displacement, in 4 of 23 patients with m oderate displacement, and in 2 of 3 epiphyses following complete separation. He felt that his results were much better than those being reported following manipulative or surgical

360

reposition of the epiphysis. By the 1940s, there was good awareness that results of treatment in many cases of SCFE were poor leading to arthritis in early adult life. Howarth [47, 106] and Klein et al. [126, 147] also summarized the attempts of the previous decades recognizing in particular that correction of the deformity by forceful closed manipulation followed by fixation in plaster casts led to poor results as did subsequent attempts at open reduction followed by fixation in plaster. Open reduction and osteotomy at neck and peritrochanteric regions were resorted to frequently. Transphyseal Drilling Simple transphyseal drilling was shown to induce premature growth plate fusion in cases of slipped capital femoral epiphysis. Pomeranz and Sloane used this method in six cases in 1935 [161]. Mayer reported on the procedure in 20 patients diagnosed early when slipping was extremely slight [111]. Under radiographic control a drill bit was placed through the greater trochanter and neck across the physis and into the femoral head. The plate was penetrated at seven or eight different points. The limb was then immobilized in a hip spica cast for approximately 10 weeks until radiographic transphyseal fusion was identified. The average period was 3 months. In the 20 cases treated, there was normal function of the hip and an excellent gait. Kiaer [162] and Mathiesen [163] reported a similar approach. Mathieson reported that 36 cases of minimal slipped epiphysis (less than 1.5 cm) had been treated by transphyseal drilling between 1939 and 1954. Three to four drill holes were placed through the physis into the head following which the patient was treated either with traction or hip spica. Most patients had fused the physis within 3 months. There were no instances of avascular necrosis and results were considered excellent.

a

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

Transphyseal Bone Graft Howarth, in an article with Ferguson in 1931, first described the open transphyseal bone graft to enhance head-neck fusion in slipped capital femoral epiphysis [145]. Howarth was the chief proponent of the transphyseal bone graft or bone-pegging operation. He performed several hundred of these procedures for mild and moderate slips. No open reduction or wedge osteotomy of the neck was done. A small entry point was made in the anterior surface of the femoral neck after which a drill or curette was passed up the neck and across the physis centrally into the femoral head. Three small bone grafts were cut from the ilium that were largely cortical and averaged 1 1/4 inches long and 3/16 of an inch wide. The bone pegs were placed into the hole in the neck, through the physis, and into the bone of the secondary ossification center. After several days in bed while the soft tissues healed, the patient proceeded either to wheelchair or crutches without weight bearing. Transphyseal fusion occurred generally within 12 weeks. In a 1966 report on 200 bone-pegging operations, Howorth claimed near universal excellent results with no documented episodes of chondrolysis or avascular necrosis. He felt the operation to be vastly superior to nailing regardless of the method of fixation [49]. He also claimed that good or excellent results had been reported for the bone- pegging operation done by others in 151 out of 152 cases. Heyman and Herndon reviewed the literature in relation to experiences with pinning in the early phases [164]. As a response to the less than perfect results with other treatments, they reported that an iliac crest cancellous bone graft placed from the femoral neck across the physis into the center of the femoral head allowed for rapid fusion thus eliminating the concern about further slippage (Fig. 3.9a, b). Reduction was not done. In 19 operations the average time to

b

Fig. 3.9 (a) and (b) surgical technique for the transphyseal bone graft (bone-peg) operation is outlined. (Reprinted from Heyman and Herndon. JBJS Am. 1954;36A:539–54, by permission Wolters Kluwer Health, Inc.)

3.2 Slipped Capital Femoral Epiphysis

fusion of the epiphysis as defined by x-ray was 2.3 months. Clinical and radiologic results were excellent in all cases. There was no postoperative plaster immobilization, and patients were ambulatory on crutches within a few weeks of surgery. In no case did they note any increase in displacement, and there was no evidence of chondrolysis, avascular necrosis, or other degenerative changes. They exposed the anterior surface of the neck of the femur and removed a rectangular shaped section of bone cortex about 3 cm long and 1 cm wide along the long axis of the neck of the femur. They then used a drill or reamer 1 cm wide directed across the growth plate and into the center of the femoral head. Small pieces of cancellous bone obtained from the iliac crest were impacted into the depths of the defect enhancing fusion. Internal Fixation Without Reduction. Pinning In Situ The physis in SCFE is already damaged, and transphyseal fixation can lead to relatively rapid fusion. Utilization of the three-flanged Smith-Peterson nail was beginning to give improved results as in situ stabilization, and postoperative rehabilitation were simplified since hip spica immobilization was not needed [112, 126, 147]. Results with the relatively large Smith-Peterson nail, however, while improved from previous reports were far from perfect, and many observers reported that although the device stabilized the head on the neck, it did little to increase or enhance the rate of fusion. Fusion was still quite slow appearing to take anywhere from 4 to 18 months, while in many instances fusion never occurred, and the femoral head grew off the tip of the nail. Wilson commented on the value of pinning in situ in 1938 [112]. He stressed the importance of early diagnosis and early treatment and also indicated that his previous experiences using weight-bearing caliper braces and ambulatory plaster spicas as protective devices were highly ineffective. Recumbent treatment in plaster hip spica was also criticized as being uncertain and needed to continue until fusion was obtained which could be 1 year or more. Even after many months of immobilization, subsequent slippage was noted along with significant muscle and bone atrophy. There was controversy between those using the drilling method from the femoral neck through the epiphyseal plate into the head with insertion of small bone grafts and those who performed metallic stabilization. Wilson indicated that the open epiphysiodesis method “requires arthrotomy and seems unnecessarily complicated.” He favored insertion of the Smith-Peterson nail by transtrochanteric pinning under radiographic control without arthrotomy. Reduction was not attempted. He clearly documented the value of the nail that stabilized the epiphysis, preventing further displacement, and penetrated the cartilaginous growth plate, allowing the ingrowth of reparative elements and thus promoting early fusion of the epiphysis. Obliteration of the growth plate occurred in 4–6 months. Bilaterality was common. Serious shortening of the involved extremity did not

361

develop following pinning since most growth had occurred by the age when the slip occurred and the fact that relatively little femoral growth was at the proximal end. Cervical Osteoplasty Heyman et al. subsequently described a conservative operative procedure for severe slips of the femoral capital epiphysis in which transphyseal stabilization had been achieved, but the displacement leads to severely disabling limitation of motion [165]. Rather than proceeding with open reduction or cervical osteotomy, they simply shaved the prominent superolateral portions of the neck to enhance the range of motion, a procedure referred to as an osteoplasty (Fig. 3.10a, b). The procedure was recommended for extreme limitation of flexion as well as extensive loss of abduction and internal rotation. At operation, they noted severe downward and backward displacement of the head of the femur with solid union. The obstruction to motion was a large bony prominence at the anterosuperior aspect of the neck of the femur at its junction with the displaced epiphysis. This was seen to impinge against the rim of the acetabulum. The procedure is clearly recognizable today as a treatment for femoroacetabular impingement (Fig. 3.10c). Rather than performing an osteotomy with its inherent risk to the vascularity, the bony prominence was removed with an osteotome. This markedly improved the motion of the hip without affecting the position of the head and neck. In many cases, they noted that an almost complete range of motion was obtained by removing the bony prominence that had impinged against the acetabular rim. The patients were rehabilitated with early postoperative motion and weight bearing with crutches. It is the proximal portion of the femoral neck that appears as the angular or rounded ridge of bone at the anterosuperior aspect. The osteoplastic operation consists in removing the angular or rounded prominence of bone at the neck of the femur where it is adjacent to the femoral head. If complete fusion had not occurred, then bone graft epiphysiodesis was performed at the same time. They reviewed 28 patients having the procedure that was done on those with severe slippage and external rotation of the femur with marked limitation of internal rotation and abduction. Flexion of the hip was also limited and was most free only at the extreme range of external rotation. The coxa vara and shortening were unaffected by the operation although improvements in the range of motion, in particular, involving increased abduction and internal rotation, were felt to be highly beneficial. Pain was eliminated. The gait improved markedly with the knee directed straight forward and the residual limp slight. There was no evidence of avascular necrosis or early osteoarthritis. The procedure had been described several decades previously by Poland, by Whitman, and by Vulpius and Stoffel (Fig. 3.10d), but its relatively wide adoption was due to the work of Heyman et al. [106, 165].

362

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

a

b

d

c

e

A

B

C

D

3.2 Slipped Capital Femoral Epiphysis

A study by Wilson et al. on 300 hips treated from 1936 to 1960 summarized well the positive and negative features of the clinical approaches at that time [114]. The primary apparatus used for pinning in situ was the Smith-Peterson nail. This study began to focus on the problems with this nail, in particular instances of distraction of the epiphysis with insertion of the device that was seen in 12 of 222 fixations in situ. This complication, also noted by Wiberg, led to the use of thinner nails [166]. There were markedly better results with in situ fixation alone however than in those cases with moderate or severe deformity where efforts to correct the deformity were made. In hips treated with pinning in situ, most of which were only slightly displaced, 13% were rated radiographically fair or poor, and 9% were clinically fair or poor. However, when correction of deformity followed by fixation was used, and most of these were moderate to severe in extent, the radiographic fair and poor results were 45% and clinical fair or poor 38%. Each of the complications more clearly recognized in later decades were described including pin penetration, avascular necrosis, acute cartilage necrosis, and osteoarthritis. When hips were treated other than by fixation in situ, the operative approach used involved wedge osteotomy of the femoral neck and either open or closed reduction. Wedge osteotomy of the femoral neck in this era was highly problematic with the combined results from many published studies being good in 164, fair in 15, and poor in 72. The incidence of poor results, which involved primarily osteonecrosis or severe arthritis, was 29% in the combined reported series and 23% in their own series. Wiberg noted a 27% incidence of AVN (23 of 84 cases) from his institution [166]. Correction of marked displacement by either open or closed reduction also resulted in a large percentage of poor results with a report by Moore on open reduction in 87 hips with marked displacement showing the incidence of poor results at 50% [160]. Closed reduction either failed to correct the deformity or if successful was followed by a high incidence of necrosis. This was emphasized by Jerre who, in 117 hips treated by closed manipulation, noted an incidence of AVN of 41% in 24 hips that were shown to have the displacement

363

corrected and even in 7.5% of hips where no reduction followed [66]. The authors concluded that early diagnosis when the slipping was minimal led to the best results with treatment by fixation in situ; when efforts were made to correct deformity, the complication level rose dramatically.

3.2.10.3 C ontinuing Evolution from the 1960s to 1980s Treatment was well structured and described by the 1960s. The review by Howorth is the most comprehensive referring to virtually all papers published on treatment from the 1890s to 1960 [106]. Morscher reviewed the earlier European approaches to both nailing and transphyseal bone grafting for slipped epiphysis [167]. He reported generally good to excellent results with these techniques in mild to moderate slips, reserving cervical wedge osteotomy only for the most severe cases. Manipulation, spica immobilization, and open reduction without cervical shortening were all recognized to have a high incidence of avascular necrosis and poor results. Hall, in a study of 173 hips treated by many methods in the United Kingdom in the late 1940s and 1950s, documented similar findings in 1957 [168]. His study concluded that manipulation was a relatively safe and effective method of reducing deformity in patients seen soon after an acute episode but that it should be reserved for this acute group only. Straight longitudinal traction was of little value, and it was the medial rotation that was essential to gain reduction. It was recognized that the use of the term “acute episode” referred to the situation where the slip was “severe enough to prevent walking.” Hall reported that avascular necrosis was the commonest cause of a poor result. AVN was seen in 38% of the 42 cases of cervical osteotomy done in this series. AVN however was least common after low cervical wedge osteotomy (base of neck) rather than at the immediate subcapital physis level. The Smith-Peterson nailing led to many poor results, most of which were no longer seen when a switch was made to the much thinner Moore pins. Multiple Moore pins were used in 20 cases with no complications. Intertrochanteric osteotomies were done where the epiphysis was fused, did not cause AVN, and led to good results.

Fig. 3.10 (a) and (b) hip radiographs, anteroposterior (left) and lateral (right), show the prominent bone protuberances at the head-neck offset junction. Arrows indicate the region for the bony resection. The osteoplasty removes the bony obstacle to smooth flexion and abduction of the hip caused by the prominent superolateral portion of the femoral neck rather than attempting to reposition the head by more complicated osteotomy procedures. (c) The rounded but prominent superolateral region of the neck immediately inferior to the head is removed with a curved osteotome to eliminate abnormal femoroacetabular impingement with hip flexion and abduction. (Reprinted from Lavigne et al. Clin Orthop Relat Res. 2004;418:61–6, by permission Wolters Kluwer Health, Inc.) (d) Illustration from a book by Vulpius and Stoffel depicts the bony prominence of the superolateral femoral neck that was subsequently removed at surgery to improve hip motion. (Reprinted from Vulpius and Stoffel. Orthopädische Operationslehre. Stuttgart: F Enke; 1913.) (e) These images illustrate the maneuvers recommended by Parsch et al. to gently reduce an acute slipped capital femoral epiphysis at open operation. The manipulation is cautious, and the reduction is performed with the assistance of pressure from the finger on the displaced head. Stabilization is then obtained with K-wire fixation. (Reprinted from Parsch et al. J Pediatr Orthop. 2009;29:1–8, by permission Wolters Kluwer Health, Inc.)

364

Durbin, in a study of 81 hips also from Great Britain, presented a further view of treatment generally used at this time (1960) [118]. He concluded that there was no value to conservative non-operative treatment. If closed reduction was attempted, it needed to be “gently manipulated under general anesthesia” although this was usually ineffective. With slight displacement, pinning in the position of displacement with three or four small (narrow) partially threaded pins was recommended. When slip was 50% or more, the best approach was intertrochanteric (or subtrochanteric) wedge osteotomy stabilized by hip spica immobilization. Cervical osteotomy was contraindicated since 4/9 hips treated that way developed AVN. The approaches in use at mid-century were thus well defined [47, 118, 168]. In the second half of the twentieth century, in situ stabilization gained in popularity, and the Smith-Peterson nail was abandoned in favor of thinner pins. Multiple (3–4) Knowles, Moore or Hagie pins were used in many centers. Different varieties of screws were developed in several countries. As problems with screw penetration through the articular cartilage into the joint were recognized, a tendency to use a single cannulated 7.3 mm screw predominated. Open reduction fell into disfavor owing to the high numbers of cases with AVN, and repositioning or compensatory osteotomies for moderate and severe deformity at basicervical, intertrochanteric, and subtrochanteric regions were developed.

3.2.10.4 Late-Twentieth-Century and Early- Twenty-First-Century Approaches to Treatment Acute/Unstable Slipped Capital Femoral Epiphysis It is essential to differentiate whether a displaced femoral head of moderate to severe to complete magnitude represents an acute, acute-on-chronic, or chronic slip. Inherent in this assessment also is consideration using the stable/unstable classification. The truly acute slip with no previous symptoms is essentially an acute fracture-separation in an epidemiologically predisposed patient. Its treatment varies from the chronic slip – we will discuss the acute-on-chronic variants separately. Acute slippage is defined as presenting with some form of trauma (often very mild), discomfort less than 3 weeks, and instability characterized by inability to bear weight with or without crutches. The question of whether acute displacement of the femoral head should be reduced as well as the timing of any reduction needs to be considered. Acute slips are far less frequent than chronic slips. Loder et al. in their large multinational study of 1993 slips noted 85% to be chronic (greater than 3 weeks of symptoms) and 15% acute [120]. Rattey et al. found 12.5% (26/208) of slips to be acute [98]. Over the past few decades, management of the acute slip has concerned itself with the following issues: (i) should reduction be attempted or should the position be

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

accepted and pinned in situ?; (ii) if reduction is performed, what method should be used?; (iii) what is the appropriate time frame for closed or open treatment?; (iv) should the hip joint be decompressed to eliminate pressure against the adjacent vessels and hopefully minimize avascular necrosis, and, if so, by what method?; and (v) what method of stabilization is best? It is widely recognized that there is a high incidence of AVN with acute slips, and virtually no AVN in chronic slips pinned in situ without attempts at reduction. Some surgeons attempt “gentle” closed reduction in acute instances (in the temporal classification group) or in unstable hips (in the stable-unstable classification of Loder et al.) with moderate to severe deformity. The problem with attempting reduction in these instances is the inability to monitor vessel function or to know how much reduction force is safe and how much is dangerous in relation to the vascularity. Two types of closed reduction can be done. One approach is to place the patient at bed rest with skin traction pulling the lower extremity longitudinally and into medial/internal rotation. Once position has improved, pinning in situ is performed. A second approach is to perform “gentle” closed reduction by manipulation under general anesthesia followed by pinning in situ. It is also recognized that spontaneous partial or full reduction of an acute slip can occur in association with the unavoidable moving of a patient especially at the times of radiologic diagnosis, positioning the patient on the operating room table, and the actual performance of the surgical pinning operation. It came to be recognized that results were better if reduction was performed within 24 h of presentation (AVN, 7%) compared with those reduced after 24 h (AVN, 20%). A 14% AVN rate (5/35) was also reported by Casey et al. although they felt that reduction would be safer if done by skin traction with internal rotation followed by in situ fixation [169]. They treated 35 acutely slipped epiphyses with manipulation only or manipulation and traction (followed by pinning in situ) but the time from onset to manipulation ranged from 1 to 34 days. Fahey and O’Brien reviewed 75 cases of acute slip from the literature [170]. Closed reduction and cast (40 cases) led to 47.5% satisfactory results; closed reduction and internal fixation (23) led to 65.2% satisfactory results; and open reduction and internal fixation (12) led to 83.3% satisfactory results. In their own cases they had satisfactory results in 9 of 9 with closed manipulative reduction under general anesthesia with in situ pinning. Aadalen et al. showed the relative safety of closed manipulative reduction and epiphysiodesis or pin fixation in 50 hips with acute slipped capital femoral epiphysis [171]. There were 34 boys and 14 girls. The patients were placed into four groups based on treatment. The best results with no episodes of AVN were in patients treated by manipulative reduction within 24 h of the onset of acute symptoms although this only involved eight patients. There were no

3.2 Slipped Capital Femoral Epiphysis

cases of AVN in girls although this has not been found in other series. There was a 15% rate of AVN (7 of 47) where treatment was by manipulative reduction followed by open epiphysiodesis or pin fixation or both. In various additional subgroupings no significant difference between epiphyseal stabilization by pin fixation or by epiphysiodesis was noted. Rattey et al. treated 26 acute slips, 23 by in situ pinning alone, and 3 by reduction and pinning [98]. Four developed AVN (15%), but two of the three treated by closed reduction had the AVN complication. Aronson and Loder have taken a cautious, non- manipulative approach for the acute or unstable form of slipped epiphysis [172]. Their priorities are to avoid avascular necrosis, avoid chondrolysis, and prevent further slip. In relation to the correction of deformity they recognize that it is associated with a high incidence of complications “so manipulative reduction under anesthesia or an acute corrective osteotomy is not recommended.” They recommend preoperative bed rest to decrease the synovitis and intra-articular effusion followed by operative stabilization with a single pin with careful positioning of the patient on the operating room table with no attempt to do a manipulative reduction. Even with great care incidental reduction could occur on occasion but they stress that it was never actively sought. No preoperative traction was used since that alone could increase the intra-articular pressure or damage the posterior vasculature by effectively bringing about a relatively forceful reduction. Current Update 2000+ A large study from France continues to show considerable variety in treatments used along with continuing occurrence of AVN in both acute/unstable and, to a lesser extent, stable hips [173]. Amara et al. assessed the treatment of 186 cases of severe SCFE (>45° displacement) in a recent French multicenter review. The average displacement was 60° with an almost equal division between unstable and stable cases (92/94). Treatments were by in situ fixation, lateral Dunn, anterior Dunn, and (for unstable cases) reduction by traction or manipulation under anesthesia. The AVN occurrence in the unstable cases was 21.7% (after in situ fixation 1, 11%; after anterior Dunn 7, 19%; after lateral Dunn 3, 43%; and after preoperative reduction 8, 21%). Two large reviews of several papers each again confirmed the continuing high rate of AVN with acute slips. Zaltz et al. in a large review report an overall rate of osteonecrosis of 23.9% (15 papers published between 1993 and 2010) [102] and Loder a rate of 21% (88/417) [101]. Studies from individual centers however assessing various subgroups of approaches begin to show some promise in terms of limiting the frequency of this problem. Palocaren et al. assessed 27 patients who had treatment with in situ pin fixation; AVN occurred in 6/27 (22.2%) with female sex and slip magnitude, the only two identifiable factors statistically predispos-

365

ing to AVN [100]. Chen et al. showed a slightly lowered rate of AVN of 4/28 (18%) with their protocol involving urgent treatment, “positional” reduction, fixation with two cannulated screws, and hip joint arthrotomy by percutaneous aspiration (16) or open capsulotomy (5) [174]. Kalogrianitis et al. had a high rate of AVN 8/16 (50%) in a small series, but seven of the eight that developed AVN were treated between 24 and 72 h after symptom onset, and eight without AVN were treated either within 24 h (5) or at 8 days (3) [175]. Sankar et al. found 14/70 (20%) with AVN with in situ fixation (3/16, 19%) alone or closed reduction/screw fixation (10/38, 26%) at the high rates, while open reduction and internal fixation (1/16, 6%) had a much lower rate [99]. Younger patients and shorter duration of symptoms predisposed to a poorer result. Gordon et al. also identified timing of reduction and joint arthrotomy as diminishing the risk of AVN in unstable slips [176]. An assessment of 16 consecutive unstable hips showed no AVN in 10 when reduction was done within 24 h accompanied by hip joint arthrotomy and use of 2 cannulated screws. Overall, 2/16 developed AVN (12.5%). Rached et al. used reduction and fixation with a single screw or multiple wires with an AVN rate of 14.8% and chondrolysis 36% seen [177]. They found risk to increase with the reduction between the 2nd and 7th day after slip. It is the report of Parsch et al. however that shows the best results in a relatively large series with a low AVN rate of 4.7% (3/64) [178]. Their method, followed over a 19 year period, featured emergency surgery within 24 h, hip joint capsulotomy for evacuation of intra-articular hematoma or effusion, controlled gentle manual reduction of the head onto the neck under direct observation, and fixation of the reduced epiphysis by 3 K-wires. Two of the cases had presented after 24 h. The authors stress that the reduction is controlled under direct visualization at arthrotomy with gentle pressure of the fingertip of the surgeon repositioning the epiphysis. The severity of the slip was not a factor in leading to a poorer result. The technique is illustrated in Fig. 3.10e. In summary, the acute slip – strictly defined – is the only variant of slipped capital femoral epiphysis deemed capable of reasonably safe management by attempted gentle reduction prior to pinning in situ within 24 h of presentation. The risk of AVN still appears to be in the 15% range although some of this may be caused by the initial injury rather than the treatment. Even in this subgroup, many opt strictly for pinning in situ with no attempt at reduction. The best results in acute/unstable slips consistent with efforts to obtain the best alignment to minimize long-term FAI and osteoarthritis use each of the following: treatment considered to be urgently needed with intervention as an emergency procedure and certainly within 24 h of occurrence of acute onset, hip joint capsulotomy (open) to remove hematoma and joint effusion fluid, cautious gentle manipulation for reduction under direct observation with surgeon manually directing reduction of the

366

head onto the neck, and stabilization of the reduction by pinning with a single central cannulated screw or 3 smooth transphyseal wires. Chronic Slipped Capital Femoral Epiphysis Overview

Mild to moderate slippage is treated generally with pinning in situ. If the deformity is severe with posterior slip of the head onto the back of the neck in a chronic state, pinning in situ while preventing further displacement and alleviating pain cannot return the head-neck-shaft relationship to a normal position. These patients would then continue with considerable shortening, external rotation positioning of the lower extremity, and a Trendelenburg gait. Attention has been directed in some instances of moderate and in severe slippage to correcting the head-neck relationship with the acetabulum and shaft. The major danger in this approach remains the possible causation of femoral head avascular necrosis and an increased risk of chondrolysis. The slippage has occurred gradually and the periosteum and vessels on the posterior medial concavity of the head-neck junction are shortened with newly synthesized fibro-osseous tissue (callus) also deposited in the concave regions. There is a high risk of tearing this tissue by reduction efforts, closed or open, further damaging the femoral head blood supply. There are several methods used to reposition the femoral head in a chronic slip and differing levels and times at which correction can be done. Acute-on-Chronic Slipped Capital Femoral Epiphysis In some patients there is a history of mild symptoms of several weeks duration. Rather than call this acute-on-chronic, some groups, like Peterson et al., focus on the acute pain and usually moderate to severe displacement and suggest: “duration of prodromal symptoms is an unreliable subjective indicator of acute SCFE” [179]. In their series of 91 acute slips over a 40-year period, the incidence of AVN was 14% (13/91). They felt that manipulative reduction under general anesthesia did not increase the risk of AVN. Crawford has written excellent reviews of overall management [180, 181]. Induction of Growth Plate Fusion In Situ to Prevent Further Slippage

Transphyseal Pinning In Situ, Without Reduction The commonest approach to treatment for stable or chronic slips is to stabilize the displacement in situ without attempts at reduction to allow transphyseal fusion to occur (Figs. 3.11a–d and 3.12a–f). This prevents further slippage of the head-neck relationship on a mechanical basis owing initially to the sta-

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

bilizing effects of the pin fixation and shortly afterward to the induction of premature growth plate fusion. It is the induction of growth plate fusion that remains the primary goal of therapy. Perforation of the epiphyseal growth plate one or multiple times by transphyseal drilling in the course of pinning in situ allows the epiphyseal circulation with its osteoprogenitor cells to relate intimately to the metaphyseal circulation with its osteoprogenitor cells forming a series of small bone bridges which ultimately lead to full fusion. Physeal damage is also caused by the slippage itself, and the continued weight bearing prior to diagnosis such that the plate is already damaged when diagnosis is made. The greater the degree of slippage, the longer the time to diagnosis, and the more unstable the hip, the greater the growth plate damage will be. Once fusion has occurred, further slippage is prevented. Initially, the Smith-Peterson nail was used to treat this condition. The nail had been developed for femoral neck and intertrochanteric fractures in adults. The technique was soon abandoned in children because the extreme hardness of the epiphyseal and femoral neck bone and the relatively tenuous attachment of the head to the neck in slipped capital femoral epiphysis were such that the large, bulky nail itself frequently furthered the displacement of the head often with causation of AVN. Shortly thereafter, thinner screws were used to stabilize the head to the neck. A single screw was used at Children’s Hospital Boston for several years. After screw insertion the patients were protected on crutches for as long as 6 months. The desired central placement of the screw in both anteroposterior and lateral radiographic projections frequently required the passing of guide wires and drill holes several times through the growth plate to assure appropriate positioning. Use of the single screw, rest on crutches, and multiple perforations of the physis during the process of accurately positioning the screw all enhanced transphyseal fusion. Multiple thin transfixing screws began to be used in many centers to enhance the mechanical stabilizing effect although initially this was for the most part a clinical intuitive assumption [182]. Steinmann, Hagie, Moore, or Knowles pins provided an effective mechanical stabilization, shortening the period of crutch use and still allowing for premature fusion. The multiple pins led to some major complications, however, recognized only after a several year delay due to penetration of one or more screw tips through the articular cartilage into the joint. Many types of transphyseal fixation screws have been used over the past few decades depending on time and country, but the treatment principle remains the same. A single cannulated screw of 7.3 mm diameter is currently favored in many centers for reasons listed below.

3.2 Slipped Capital Femoral Epiphysis Fig. 3.11 (a–d) Slipped capital femoral epiphysis treated with pinning in situ is illustrated. An excellent result was achieved with pinning using three Knowles pins. (a) Slight medial head displacement is noted on the anteroposterior preoperative radiograph at left; the superior neck is in line with the lateral border of the bone at the femoral head rather passing within it. (b) There is slight widening of the physis and slight posterior displacement of the head relative to the neck on the lateral view (right). (c) Pinning in situ led to rapid transphyseal fusion and excellent appearance at skeletal maturity in anteroposterior projection (left). (d) Fusion is also noted on lateral view (right)

a

c

Danger of Pin Penetration Through Articular Cartilage (Walters and Simon) Walters and Simon performed an important study when their assessment of results following pinning in situ of slipped capital femoral epiphysis led to the observation that many patients suffered from unrecognized pin penetration following treatment [183]. Most of the procedures assessed had used multiple pins. They noted that the fixation nail or screw position may appear to be entirely within the bone of the femoral head on both the anteroposterior and lateral x-ray projections but in reality might be protruding through it on certain oblique projections into or beyond the articular cartilage since radiographs post-surgery were unable in every instance to evaluate adequately the position of the tip of the pin (Fig. 3.13). If the tip of the internal fixation device relative to the x-ray beam did not lie in one of the planes of assess-

367

b

d

ment, it might well be through the cartilage and in the joint, while its projection in the plane of the x-ray film showed it to be within the femoral head bone due to overlap. The accuracy of this observation was confirmed by producing experimental models on cadaver femurs in which some degree of protrusion of the pin through the head was obtained, and x-rays in varying positions were shown not to reveal this. Clinical study was then performed retrospectively on 102 patients treated between March 1971 and October 1977. The treatment was by a single screw in two with the others treated by multiple Knowles or Hagie pins. Their findings indicate how dramatically the concept of pinning in situ has changed since the 1970s. In 90% of the patients, evaluation of anteroposterior lateral and frog lateral radiographs showed no visual evidence of any pin protrusion beyond the subchondral bony surface of the femoral

368

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

head, while in 10% a single pin, usually noted on a single view, was actually found to be at the bony surface or protruding 1–2 mm beyond the head into the joint. Pins were frequently left in place even when they were shown on clinical radiographs to protrude if it was felt that they were a

not directed onto a “weight-bearing surface” and if they were asymptomatic. When the “true distance” of pin location was calculated, only 40 of 102 patients could be classified in category I where all pins were contained within the femoral head. A further 40 patients (40%) were found to be

b

d

Fig. 3.12 (a–d) These figures demonstrate how the entry point for a single screw, when used for in situ fixation, must be progressively more proximal and anterior to the greater the degree of posterior displacement, since the tip of the pin should be as close as possible to the center of the head in both projections. Arrows illustrate the approximate point of entry in all figures. (a) (top, anteroposterior view) and (c) (bottom, lateral view) show a relatively mild slip, while (b) (top, anteroposterior view) and (d) (bottom, lateral view) show a case with severe displace-

c

e

ment. (e) Drawing of CT scan shows position of posterior displacement of head relative to neck. (f(i–viii)) Anteroposterior (i) and lateral (ii) radiographs of a severe slip show single screw position. A series of CT images (iii–viii) shows progression of pin tip from anterolateral entry point on neck (iii) to eventual central position in the head (viii). Note head displacement relative to the neck (v–vi) compared to normal side. CT image itself shows the orientation

3.2 Slipped Capital Femoral Epiphysis

fi

369

fii

fiii

fiv

fv

fvi

fvii

fviii

Fig. 3.12 (continued)

370

Fig. 3.13 Pin protrusion through the femoral head articular cartilage can escape detection by standard radiographic views. Both anteroposterior and lateral radiographs at a 90° relationship to each other taken at a slight tilt from this photographic projection showed the tip of the pin (arrows) to lie within the bone contours of the head. (Reprinted from Nguyen and Morrissy. J Pediatr Orthop. 1990;10:341–46, by permission Wolters Kluwer Health, Inc.)

in category II where maximum pin protrusion beyond the bony femoral head was 5 mm or less. Twenty-two of the 102 patients were in category III where maximum pin protrusion beyond the femoral head was more than 5 mm. When postoperative films were assessed, no patients in the first group had evidence of joint cartilage damage. Sixtyeight percent of patients in the second group I showed some evidence of chondrolysis or subchondral bone changes, and all patients (100%) in the third group showed some evidence of chondrolysis or bone changes. All patients in this study had only mild or moderate deformity initially. Walters and Simon pointed out that, from a theoretical standpoint, devices close to the surface of the head may protrude through it and may not appear to be outside it on either anteroposterior, true lateral, or frog lateral radiographs. Their model in vitro demonstrated that unrecognized pin protrusions could occur and not be detected by standard radiographic projections. They established a template used to calculate the safety limits of pin placement in relation to proximity to the surface of the head. The closer to the periphery of the head the pin tip was positioned on either anterior-posterior or lateral projections, the less the margin of safety. Walters and Simon had identified a major problem with pinning in a retrospective review noting that many pins had been passed beyond the subchondral bone plate of the femoral head into and frequently through the femoral head cartilage causing damage to both femoral and acetabular hip joint cartilages. The deep position of the pins either was not recognized or, if recognized, was not felt to be of concern. Independent confirmation of this finding soon followed

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

from other centers. Biplanar radiographic views can fail to detect articular cartilage penetration owing to geometric configuration of the head but measures to minimize or eliminate misinterpretation have been suggested. The increased awareness has minimized one of the complications of treatment, cartilage necrosis or chondrolysis, although not all cartilage necrosis is due to pin penetration. Shortly after the work by Walters and Simon, retrospective reviews of pinning in situ from other centers confirmed the high incidence of unsuspected intra-articular pin penetration. Lehman et al. reviewed 63 hips that were pinned as treatment for SCFE and noted a 36.8% incidence of pin penetration not previously recognized [184]. Bennet et al. assessed 148 hips using a three-dimensional mathematical model of the femoral head and pin position and noted pin protrusion in 41 hips, an incidence of 27.7% [185]. Of these 22 had protrusion sufficiently great to have entered the joint space with the others showing the tip of the metallic implant to be in articular cartilage. These studies confirmed that standard radiographs taken at right angles to each other often missed evidence of pin protrusion. Several groups then generated models showing this occurrence (Fig. 3.14). Swiontkowski reviewed 66 pinning procedures and noted a 30% incidence of pin penetration [186]. These three studies, published shortly after the report by Walters and Simon and thus with surgery done prior to the time of awareness of their findings, confirmed the high incidence of pin-related problems. A retrospective study by Gonzalez-Moran et al. showed frequent pin penetration when multiple Kirschner wires (1–6, average number 3.8) were used for fixation; the rate was 34% (40 of 188 wires used) [187].

Fig. 3.14 These illustrations (Figs. 3.13 and 3.14) show the importance of appropriate pin placement. While it is important for the pin to be placed centrally within the head on both anteroposterior and lateral projections, it is also important that it not protrude through the posterior cortex of the neck (arrow) where the vascularity is most prominent or through the articular cartilage. (Reprinted from Nguyen and Morrissy. J Pediatr Orthop. 1990;10:341–46, by permission Wolters Kluwer Health, Inc.)

3.2 Slipped Capital Femoral Epiphysis

Continuation of Physeal Growth Following Pin Insertion An additional problem with pinning was occasional continuation of growth following insertion of the pin. Most instances of this resulted in no problem, but in some there was increased slippage owing to loss of the protective effect on the one hand and failure to induce physeal fusion on the other. Laplaza and Burke performed a study in 71 hips that had been treated either with Steinmann pins, Knowles pins, or cannulated screws [188]. Evidence of continuing epiphyseal growth was seen in 29% of hips treated with a Steinmann pin, 18% of hips treated with Knowles pins, but in no hips with treatment by cannulated screws. They recommended the use of one cannulated screw in the treatment of mild and moderate slipped capital femoral epiphysis. No cases of chondrolysis or AVN were noted in the entire series. Progression of the slippage did not happen in any of the hips although in some repositioning of the pins was performed. Time to Physeal Fusion Following Pinning Fusion of involved physes following pinning shows the effect of transphyseal perforation as well as the fact that the physes undergoing displacement are already damaged and predisposed to early fusion. Gonzalez-Moran et al. showed a mean time to closure of 7.86 months in 31 hips treated with Kirschner wire fixation and of 7.12 months in 31 hips treated with single cannulated screws [187]. Stanton and Shelton found the mean time to closure to be 12 months post-surgery in 26 hips compared to 22.2 months in the non-operated contralateral side [189]. Closure time is variable but most articles comment on a 6–9 month range. Difficulties with Removal of Pins after Physeal Fusion An additional problem with pinning in situ involves difficulties and complications related to removal of the pins once physeal fusion had occurred. The problems are (i) inability to remove the pin because of breakage with the deeper part becoming relatively inaccessible; (ii) removal of the pin with such great difficulty that excessive bone must be removed, and the patient requires lengthy protection with crutches; and (iii) fracture though the lateral pin entrance point. These problems are increased when multiple pins are used. Swiontkowski documented difficulty with pin removal [186]. In 11 of 18 cases where pin removal was performed, the removal took longer with a larger blood loss than the original procedure. Pin fracture or shearing of the ends occurred in five of these eight cases requiring the procedure to be halted. Large amounts of lateral cortical bone removed in three cases required extra protection postoperatively. Studies on the Number of Pins Needed for Effective Stabilization Studies have been performed in an effort to document the clinical impression that a single screw provided adequate treatment for a slipped capital epiphysis as

371

long as it was placed centrally in anteroposterior and lateral projections. Chung and Hirahata characterized the biomechanical features of multiple pin fixation [190]. Kruger et al. demonstrated experimentally in dogs that single pin fixation was 83% as strong and 78% as stiff as the normal femoral physis but then proceeded to recommend use of two pins [191]. Subsequent investigations however supported the use of single compared with double screw fixation. Biomechanical analysis by Karol et al. was done following creation of SCFE in bovine femurs [192]. One side was repaired with a single screw and the other with two screws. Specimens were reloaded to failure and double pin fixation yielded only a 33% increase in the stiffness compared with single pin fixation. The stiffness however of neither double or single screw fixation approximated that of the intact physis. Single screw fixation was recommended because the small gains in stiffness by the second screw did not offset the increased risk of complications. A similar study was performed by Kibiloski et al. in 12 pairs of bovine femora with single pinning on one side and double on the other [193]. The specimens were then subjected to physiological shear loads across the epiphysis. The rates of creep were decreased 23% with double screws compared to single screws at slow walking and 30% at fast walking. The results were not statistically significant, and the authors again recommended single screw fixation because the small gains in resistance to cyclic creep at physiological loading were not statistically different and did not offset the complications to be expected with multiple screws. Technique for Percutaneous Fixation Griffith pointed out that since the slipped epiphysis was positioned posteriorly directly behind the neck, it was important in performing internal fixation to introduce the pin through the anterolateral or even the anterior aspect of the proximal end of the femur to direct posteriorly [69]. Colton also illustrated this point [194]. Morrissy has pointed out that “in situ pinning is a radiographic technique” [195]. Muscular forces and the particular anatomy of a proximal femur determine the direction in which the displacement of the femoral head occurs. In SCFE, the femoral neck rotates externally and the femoral head slides posteriorly around the axis of the femoral neck but remains in the acetabulum. On plain radiographs the femoral head appears to have slipped inferiorly, but, in reality, inferior displacement of the head is prevented by the shape of the femoral neck. CT scans show that it is really the posterior displacement that is occurring. With severe slips, the femoral neck may begin to move proximal in relation to the femoral head so that the metaphysis comes to the lateral edge of the acetabulum. This occurs with complete slips. CT scan is quite helpful to assess the amount of posterior slip along with the often extensive remodeling of both the anterior and posterior femoral neck surfaces.

372

Morrissy clearly defined use of the single pin technique for percutaneous fixation [195, 196]. The pins must be placed more anteriorly in entrance position than most appreciate since a direct lateral approach can lead to difficulty with central positioning of the pin in the femoral head. The technique, which utilizes a cannulated screw, stresses the need for a more anterior point of entry into the proximal femur to allow for direction of the screw into the central aspect of the femoral head in its posteriorly displaced position. Morrissy emphasized the importance of radiographic control for accuracy of screw placement. It is essential for the screw to be in the central axis of the femoral head and that it not penetrate the joint. One screw is recommended and there is only one correct location for it, at or very close to the central axis of the femoral head (Fig. 3.12a–d, f). The operative procedure is done on a fracture table with the affected leg abducted 10–15° and internally rotated without force to bring the femoral neck as nearly as possible parallel to the floor. Fluoroscopic image intensification with a free range of motion is essential. A single screw is now recognized by most as adequate and appropriate for the chronic slip. Evidence continues to increase that a single screw is also effective for the acute and acute-on-chronic slip. If two screws are to be used in the latter situation, the first should be in the central axis of the femoral head with the second below it. Morrissy indicates that the second screw should be at least 8 mm from the subchondral bone to prevent penetration that can be difficult to detect radiographically. The screw in the central axis on all projections can be inserted as deeply as 2 mm from the subchondral bone on the true lateral. The more anterior entry point allows central perpendicular fixation in relation to the growth plate axis. Nguyen and Morrissy stressed the importance of the anterior femoral neck as the starting point for the in situ fixation screw/pin with the exact location depending on the amount of slippage [197]. The greater the posterior slippage the more anterior and proximal the pin entry point must be. The entry point is on the femoral neck rather than the lateral cortex. CT scans are particularly effective in showing the need for altering the position of entry with the changing degrees of posterior displacement. Another valuable concept in determining the point of placement of the screw is that not only should it be in the central axis of the femoral head, but it should also pass at right angles to the physis. It is extremely important not to place the screw into the superior and lateral segment of the femoral head where the lateral epiphyseal arteries are most prominent. It is also important not to have the entry site on the lateral cortex of the femur. The need to have the screw in the central axis of the femoral head and perpendicular to the epiphyseal plate mandates that the starting point for insertion would vary according to the degree of slipping. The more severe the slip, the further anterior the starting point. For those used to pinning a hip in an adult, a meaningful

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

change in orientation is needed in relation to the slipped capital femoral epiphysis, and this change in orientation is greater, the greater the degree of posterior slippage. The pinning is done from anterior to posterior since it is essential both to attain position in the central axis of the femoral head and to do so by passing at right angles to the epiphyseal line. If the screw winds up in the superolateral aspect of the femoral head the dangers of penetration are greater as are the dangers of damage to the lateral epiphyseal artery. If on the other hand the screw passes out of the posterior surface of the neck and then into the head, the danger of damage to the retinacular vessels on the posterior lateral aspect of the neck is also great (Fig. 3.14). Brodetti related the vascular anatomy of the femoral head to the possibility of bad results with pinning in the superolateral aspect [198]. Even if the screw or pin remains within the neck and head placement into the superolateral aspect can damage the lateral epiphyseal arteries that are immediate intraosseous branches of the lateral ascending cervical vessels. His experimental work recommended placement of screws in the central zone of the femoral head to minimize or completely prevent interference with the blood supply. More recent studies have shown that the lowest complication rates occur when the screws were in the varus position and below the superior quadrant of the head. Single screw fixation for acute and acute-on-chronic SCFE has been reported in 21 hips treated from 1990–1993 by Goodman et al. [199]. All hips were fixed with a single cannulated screw with no attempt made at reduction. Results were excellent with no cases of AVN or chondrolysis. There was no loss of position during the healing phase, and physeal closure occurred at a mean of 9.6 months. It was concluded that single screw fixation was adequate for uncomplicated acute and acute-on-chronic SCFE. In this series 9 of the hips were defined as acute and 12 acute-on-chronic. An acute slip was defined as one in which there was a sudden onset of usually severe symptoms with less than 3 weeks of symptom duration. There were no radiographic signs of remodeling or new bone formation at the epiphyseal-metaphyseal junction. The acute-on-chronic designation was made when there was a sudden worsening of symptoms with hip pain greater than 3 weeks, and no radiographic signs of recent remodeling or new bone formation. Single percutaneous pin fixation for chronic SCFE was found to be a valid technique by Samuelson and Olney [200]. In a review of 24 chronic slips treated with a single Knowles pin results were excellent with all patients experiencing complete closure of the growth plate within 12 months and no evidence of AVN, chondrolysis, slip progression, pin penetration, hardware failure, or intertrochanteric fracture. The first 7 patients in this series were treated with 2 pins, while the subsequent 17 were treated with single pins without problems. The single pin was placed directly at the center of the femoral head in all radiographic projections. Ward et al.

3.2 Slipped Capital Femoral Epiphysis

reported on 53 hips treated with a single screw [201]. After a mean follow-up of 32 months, 92% demonstrated physeal fusion with no cases of chondrolysis or AVN reported. They also found that single pin placement provided excellent stability in each of five hips with acute or acute-on-chronic slips. Aronson and Carlson reported excellent or good results in 36 of 38 hips with mild slips, 10 of 11 hips with moderate slips, and 8 of 9 hips with severe slips with single screw fixation [202]. AVN developed in only one patient with no evidence of chondrolysis. There was loss of position after single screw treatment in only one of eight patients with acute slips. Continuing efforts to assure safe pin placement include the use of a cannulated pin through which radiopaque dye is injected at completion of the procedure. If dye is seen within the hip joint, the pin penetration has been too deep [203]. Stambough et al. analyzed 80 patients with chronic SCFE in whom pinning was performed and found complications in 10 [204]. The severity increased as the number of pins increased. Fewest complications occurred with varus pin position with tip placement inferiorly in the head. Most problems occurred when the pin tip was within the superior and anterior quadrants. All their cases were chronic with symptoms greater than 3 weeks. Their group I patients were treated with three or more pins, and group II patients had two or one pin. Those in group I tended by necessity to have more valgus or lateral positioning of the pins and those in group II more horizontal or varus placement. Complications were highly concentrated in group I where there were three cases of AVN, five of chondrolysis, and one of subtrochanteric fracture; none of these complications occurred in group II with only one pin fracture seen. When complications were assessed in relation to pin position, most complications occurred when the pin was in the superior-anterior region of the head and when it was left in 2.5 mm from the subchondral bone surface. The authors clearly favor the use of two or one pin as complications markedly diminished in those situations. They felt it is mandatory to avoid the superior and anterior quadrant because of the higher incidence of AVN, pin penetration, and chondrolysis. Single or two screw placement also proved to have favorable results in a review by Herman et al. even in 23 hips with grade III slips of greater than 50% [205]. Four were acute, 11 acute-on-chronic, and 6 chronic. Stabilization without progression of slip was achieved in all patients, and screw placement was satisfactory by the criteria of Stambough in all patients. Complications in this group still occurred however with three cases of AVN and one with chondrolysis. The patient with chondrolysis had that present preoperatively. One of the potential problems was unintentional reduction of the displacement that occurred in 9 of 15 patients just based on their movements within the hospital since attempted reduction was not part of the management scheme. AVN developed in three of these nine hips. The hips were thus unstable.

373

Although earlier studies indicated that black children might be more susceptible to complications with treatment of SCFE [206], in particular chondrolysis and AVN, more recent studies have indicated this not to be the case. One of the earliest studies showing an absence of increased complications in the black patient was the study by Bishop et al. who noted a satisfactory result in 87% of 70 hips with slipped epiphyses in 50 black children [207]. Chondrolysis developed only in 6%, and only when there was persistent intra- articular protrusion of the pin fixation device. AVN was present in 7% but was always attributable to over reduction with an acute slip. Aronson et al. [139] concluded in a study of 55 children, 89% black, that black children were not more susceptible to chondrolysis, a conclusion also reached by Stambough et al. [204]. Aronson and Loder in a larger review of 74 black children with 97 slips reiterated that conclusion showing satisfactory results with only 3 cases of chondrolysis in multiple or single pinning in situ [208]. The study by Aronson et al. [139] reviewing patients treated from 1977 to 1983 showed good results in general with pinning in situ although the surgical approach was generally from the lateral aspect of the cortex in terms of pin placement and the slipped epiphysis was stabilized by 2 pins in 48 hips, 3 in 28, and 4 in 4. Eighty hips were assessed with excellent or good results in 86% of mild slips, 55% of moderate, and 27% in severe. Poor pin position was felt subsequently to be associated with 60% of the 20 poor results. The complication of chondrolysis developed in 3 hips (4%) with AVN in 2 hips (3%). The authors concluded that pinning in situ was still the advisable approach but that it would be improved by anterolateral pin entry points and also that one screw would be sufficient for stabilization and would minimize complications. This study also showed the value of early diagnosis since the results were progressively worse the greater the severity of displacement. By the late twentieth century, in situ fixation for chronic slips was well established as the primary treatment for mild and moderate slips and even for severe slips by some. Much of the clinical support for the in situ fixation approach was based on the extensive and long-term studies of all treatment modalities from the University of Iowa. Boyer et al. from that institution reported, in 1981, a follow-up of 121 patients treated between 1915 and 1952 [209]. They concluded: “the long-term results suggested that, even in patients who had a moderate and severe slip, the outcome was better after in situ fixation than after manipulative treatment”. Carney et al. extended that study in 1991 reviewing 155 hips at a mean follow-up of 41 years [210]. The acute, chronic, and acute- on-chronic categorizations were used; and the head-shaft angle defined 42% mild, 32% moderate, and 26% severe. Reduction was done in 39 hips and realignment in 65. For the chronic hips, treatment ran the spectrum from symptomatic only (25%), hip spica cast (30%), pinning (24%), and

374

o steotomy (20%). Both the Iowa hip-rating and the radiographic classification of degenerative joint disease worsened with increasing severity of the slip and when reduction or realignment had been done. In particular, osteonecrosis (12%) and chondrolysis (16%) were also more common with reduction or realignment and both led to a poor result. They again concluded: “regardless of the severity of the slip, pinning in situ provided the best long-term function and delay of degenerative arthritis with a low risk of complications.” Technique of realignment had a risk of “appreciable complications” whose negative implications were worse than the natural history of the disease itself. The natural history of deterioration with time was recognized but noted to be mild but worsened in relation to the severity of the slip and to complications of its treatment. Long-term results reported after 2000 continued to reach similar conclusions. Boero et al. (2003) assessed 48 hips treated by in situ pinning at a median follow-up of 17.9 years (8.8–29.2) [211]. Assessment was by mild, moderate, and severe slippage. Slipping degree was directly correlated with worsening of results, and the presence of chondrolysis or epiphyseal osteonecrosis (due to reduction maneuvers on chronic slips) always led to early osteoarthritis. Wensaas et al. (2011) assessed 76 hips at a mean follow-up of 38 years (21–57), 69 of which were chronic [212]. In 51 patients (mean slip angle 32°) treated with in situ fixation, the clinical outcome was good in 35 (69%) with no difference between screw fixation and the bone-peg epiphysiodesis. In eight patients with a large chronic slip (mean 53°) treated with bone-peg fusion and femoral osteotomy, six had a poor outcome. Larson et al. (2012) assessed 176 hips with in situ pinning for all degrees of severity at a mean follow-up of 16 years (2–43) to determine the deterioration of results with pinning in situ in a non-anatomic position [213]. They found that 21 hips (12%) had undergone reconstructive surgery for persistent symptoms: 8 total hip arthroplasty, 2 surgical hip dislocation, and 11 femoral osteotomy. Pain was reported in 33%. They also noted that the procedures were spread among those with mild, moderate, and severe slips. DeLullo et al. found that in situ pinning of 38 hips at a mean follow-up of 7.6 years (with 98%, 36/37 mild or moderate) led to a generally good clinical outcome reporting only slightly less function and more pain than age-matched healthy individuals [214]. They felt that residual femoral neck deformity in the form of metaphyseal prominence or head-shaft alignment “appears to exert a modest role in these outcomes.” Loder and Dietz (2012) reviewed results from the literature of the different methods of SCFE treatment concentrating in particular on comparing methods with or without surgical dislocation [215]. The best treatment for a stable hip is single screw fixation and for unstable hips urgent gentle reduction, hip joint decompression, and internal screw/pin fixation. For

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

both stable and unstable slips, the short-term small series comparisons do not currently demonstrate an advantage or improvement of outcomes using surgical hip dislocation methods compared with the conclusions in the previous sentence. With greater recognition and definition of the femoroacetabular impingement (FAI) syndrome and strongly interventionist bias of many in efforts to decrease the hip degeneration secondary to childhood and adolescent hip disease, the basic approaches to treating SCFE are undergoing some change. Millis and Novais comment on in situ fixation as follows: “femoroacetabular impingement causes some mechanical abnormality in every hip affected by SCFE, even when the slip is mild…” [216]. In situ fixation alone rarely relieves the femoroacetabular impingement in SCFE and FAI must at a minimum be assessed in every hip affected with SCFE even when the deformity is mild. Several treatment options exist for FAI in SCFE, but all are invoked under the premise that since cumulative injury to the articular cartilage can result from impingement, “it is better to prevent this type of injury than to treat it later.” Considerable differences of opinion exist however, again relating to concerns that complications of FAI oriented procedures can lead to problems (primarily but not exclusively osteonecrosis) significantly greater and significantly earlier than would occur from the natural history of the disorder as treated by in situ pinning. These matters will be considered in greater detail below. Limb Leg Discrepancies In the large majority of patients, the induced premature closure of the proximal femoral capital epiphysis leads to insignificant limb length discrepancy problems and no meaningful change in proximal femoral structural orientation at maturity. In the subgroup of patients, however, who develop the slip at the early end of the time spectrum in the juvenile age group, premature closure can lead to limb length discrepancy and perhaps more importantly to a relative coxa vara owing to overgrowth of the uninvolved greater trochanter. In a review of 33 hips that had pinning for juvenile SCFE, growth disturbances in 64% of the hips were noted including trochanteric overgrowth, coxa vara, and coxa breva. Segal et al. defined the juvenile or younger age grouping as those with onset of slip at least one year less than the reported mean age for the disorder which in boys was less than 12.5 years and in girls 10.5 years of age [217]. The growth changes were defined in this subset of patients. This group of patients was also felt to show a higher incidence of bilaterality and a higher incidence of endocrine disorder. In an effort to minimize growth problems, differing approaches have been suggested in this particular age group. One involves the use of a hip spica to allow for stabilization of the pathologic process without compromising growth.

3.2 Slipped Capital Femoral Epiphysis

Although immobilization in the spica itself can induce premature fusion, a high number of patients worsen the slip following discontinuation of immobilization. Surgical approaches to enable continuation of growth have also been devised. These involve stabilizing the periphyseal region and bone of the secondary ossification center with a central pin that is non-threaded, thus allowing growth to continue while providing some mechanical support. Segal et al. showed a schematic diagram of an apparatus they developed which used two non-threaded or smooth pins passed from trochanter through neck and physis into the secondary center [217]. The two smooth pins were then bent distally from the trochanter and were stabilized against the side wall of the cortex with a 1/3 semitubular plate with two screws which served to buttress the pins.

Stabilization In Situ Using Open Femoral Head-Neck Epiphysiodesis with Bone Graft

Open epiphysiodesis with iliac crest bone can be effective in causing head-neck union but is a more extensive procedure and is not used widely today. Ferguson and Howorth developed the open bone graft epiphysiodesis in 1931 [145]. In the bone graft epiphysiodesis procedure, the growth plate is drilled extensively from a window made in the femoral neck. Curettage is then done to damage further the physeal area. The core site is enlarged to accommodate at least three cortico-cancellous grafts taken from the adjacent ilium. Each graft is at least 0.5 cm wide and passes from the neck across the damaged physis into the head. Postoperative management consists of 1–2 weeks of bed rest followed by gradual walking with crutches and increased weight bearing as indicated by x-ray assessment. Howorth [47, 49, 106] was reporting on over 200 cases without major complication. At the time surgical stabilization of the head-neck junction was difficult to obtain, and use of the Smith-Peterson nail had many complications associated with it. The hip capsule was opened, and several holes were drilled through the neck and the epiphyseal line into the head. Bone strips were inserted into the drill holes. The operation led to relatively rapid growth plate obliteration and subsequent fusion, generally within 8–10 weeks, and gained considerable popularity with good results reported. Weiner et al. reported a 30-year experience with bone graft epiphysiodesis in 1984, assessing 207 hips with very few complications [218]. There was only one case of AVN in the chronic group and no cases of acute cartilage necrosis. The authors felt that the procedure was as least as good as multiple pin fixation for SCFE owing to rapid growth plate closure, the avoidance of pin penetration or hardware removal, and the extremely low incidence of overall complications. They also described the procedure in greater detail in an earlier report in which 106 hips were

375

available for examination [219]. There was no evidence of acute cartilage necrosis or of avascular necrosis and the physis closed following bone grafting in all but four hips. There was occasional graft resorption and further slipping in two hips, graft resorption alone in one hip, and one instance where the graft was inappropriately placed. At that time over 500 of these procedures had been reported. Many skeptics remained, however, since the procedure was relatively complicated, requiring opening of the hip joint as well as surgical intervention at a deformed head-neck region which was somewhat fragile. Recent reports have indicated a relatively high level of complication in these procedures although the number of cases done per surgical unit would have some bearing on the outcome. Rao et al. reviewed the records of 43 patients, who underwent 64 open bone peg procedures [220]. Healing occurred at an average of 17 weeks after surgery. At time of healing, however, 42% of the hips showed a change in the degree of slip. The average operating time was slightly over 2 h per hip that is considerably greater than most pinning in situ procedures. Complications involved 4 hips with AVN, 3 with chondrolysis, 3 with infections, 4 with delayed wound healing, and 44 with heterotopic ossification. They concluded: “because of the potential morbidity of this procedure, we no longer perform it as the primary operation for stable slipped capital femoral epiphysis.” Ward and Wood also disparaged routine use of open bone graft epiphysiodesis [221]. They noted physeal fusion in only 12 of 17 cases owing to resorption, movement or fracturing of the graft. Femoral head position changes occurred along with one case of chondrolysis, one of myositis ossificans, and ten of anterolateral thigh hypesthesia. Zahrawi et al. performed a comparative study of pinning in situ and open epiphysiodesis over a 12-year period from the same institution [222]. Pinning was performed in 61 hips and open arrest in 33. At follow-up evaluation, which averaged 6–7 years, 91.7% of patients treated by pinning in situ had good or excellent results compared with 71.6% of the patients treated by epiphysiodesis. Further analysis showed that in those treated by pinning, 3.3% were considered failures, while in the epiphysiodesis group, 25% were considered failures. Comparative results showed wound infection of 3% for pinning and 12% for open epiphysiodesis. Cartilage necrosis occurred in two patients in after pinning in situ. AVN occurred but only where manipulation was performed prior to pin insertion.

3.2.10.5 Immobilization Treatment with Hip Spica Casting Without Operative Stabilization Immobilization treatment either with hip spica casting or prolonged traction was a common treatment prior to routine use of transphyseal pinning. Overall results were fair to

376

the poor, but the method continued to be used by some practitioners and clinics for several decades. Griffith reported on 67 patients treated by immobilization with traction or a plaster hip spica [69]. Further slipping occurred in 24% of these hips, and in 10 of the hips, the slipping occurred even while immobilization was being performed. Chondrolysis developed in 19 hips (28%). In only 31 (46%) of the hips treated by immobilization was further slipping prevented without complications affecting the viability of the joint. Betz et al. have reported that immobilization in a bilateral hip spica without operative intervention can prevent further slippage in a majority of cases while allowing persistence of growth after stabilization of the pathologic process [223]. Some patients close to skeletal maturity will proceed to growth plate fusion in cast owing to the fact that the physis has been significantly damaged by the previous slippage and continued weight bearing. The size of the large majority of patients with slipped capital femoral epiphysis plus the awkwardness of hip spica treatment in general argues against this treatment, but it can be used on occasion where operation is contraindicated by systemic illness or the patient is so young that transphyseal fusion must be avoided. They reported on 32 patients with 37 slipped hips treated in hip spica. The mean time of immobilization was 12 weeks, and while there were no cases of AVN, slip progression occurred in two and chondrolysis in seven (19%).

3.2.10.6 T reatment Leading to Improvement of Femoral Head Position Closed Reduction At present, efforts at closed reduction are always considered to be unwise in chronic slips but can be effective for acute unstable slips with moderate to severe deformity where improvement of position is sought. Early attempts at improving head-neck position in all cases of SCFE used strong skeletal traction through a femoral K-wire to reduce the head physically as though the displacement had been caused by an acute fracture. In general this approach either did not succeed in reducing displacement and whether it did succeed or not served often to tear the posterior vessels such that femoral head avascular necrosis was a common occurrence. As noted above in the section on pathoanatomy, the slippage occurs slowly such that the capsule, retinacular vessels, and periosteum posteriorly tend to shorten and thicken, and epiphyseal-neck stability can be enhanced by associated periosteal new bone formation. The femoral neck periosteum would have to be stretched and torn in reducing any chronic/stable displacement, and in many instances AVN was caused. Attention then turned to open surgical intervention reducing and repositioning the head so as to

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

improve its relationship to the neck and acetabulum and hopefully protect the blood supply. At present, awareness of whether the slippage is stable or unstable is leading to more physiologic management. If a hip is categorized as stable, it implies that the head-neck bond, much of which is mediated by the periosteum, is relatively intact such that efforts at reduction would likely tear and damage the vascularity. If a hip is unstable, efforts at reduction are more likely to succeed in changing position but still carry some risk for further tearing the vascularity. Two definite conclusions reached by extensive studies over an extended period of time are that (i) closed reduction for chronic slips is not indicated and (ii) closed reduction for acute slips needs to be done within 24 h of presentation and in a very cautious fashion. Griffith has shown (in 1976) the limited effectiveness as well as dangers of manipulation for SCFE [69]. In his large series, a group of 15 hips with gradual slipping were manipulated, but none showed any improvement in the position of the epiphysis. Of the 44 hips with acute slipping, 29 were manipulated with improvement in the position of the epiphysis in 11 (38%). However avascular necrosis subsequently developed in 8 of these 11 reduced hips (73%). Of the 15 acute slipped epiphyses that were not manipulated, only one developed avascular necrosis and even that had been reduced by traction. He thus demonstrated a highly significant association between closed reduction and avascular necrosis. Traction also had negative sequelae with few benefits. Seventy-nine hips showing gradual slipping were treated with longitudinal traction and internal rotation of the leg from 2 weeks to 12 months. None of the hips showed any improvement in the position of the epiphysis, but chondrolysis developed in 20 hips. Twenty of the 44 acute slipped epiphyses were treated by longitudinal traction for at least 7 days. The one epiphysis that was effectively reduced subsequently developed avascular necrosis. Six of the hips were later reduced by manipulation. Chondrolysis developed in 8 of the 20 hips. Griffith also addressed the issue of “gentle” manipulation and indicated that each surgeon considered his manipulation to have been gentle and the fact that only 38% of the acute slipped epiphyses were reduced supports this contention. In spite of the care exercised, the blood supply was frequently disrupted. He concluded: “manipulation is too hazardous to be recommended.” In spite of this concern, there have been some who regard manipulation in chronic slips as worthy of attempt. Fairbank reported a small series of 16 concluding that the dangers of manipulative reduction properly performed may have been overestimated [224]. The value of correction was in obviating need for extensive operative procedures. His only contraindication was bony fusion of the epiphysis. He stressed the need for “gentle” manipulation that was followed by pinning.

3.2 Slipped Capital Femoral Epiphysis

Such concepts as “without undue force” and “firm but never forceful” indicate awareness of concerns revascularity; as noted previously, however, it is simply not possible to know by gross unmonitored reduction how the fragile vascular supply is responding.

377

393 hips in which the number with AVN was 83, giving a similar incidence of 21.1%. They felt in fact that the incidence of poor results would have been higher since many series did not comment either on chondrolysis or on late term degenerative arthritis. In the 71 cuneiform osteotomies done in their study prior to 1968, 60 were in the subcapital area, 3.2.10.7 Open Reduction Alone for 10 in the mid-neck region, and only 1 in the base of the neck. Repositioning All hips had moderate or severe chronic slipping, and the The next approach taken for moderate to severe deformity in average age at surgery was 13 years 7 months with a range chronic slips was open reduction, done initially with no from 10 years 9 months to 17 years 2 months. They defined removal of the metaphyseal femoral neck bone. Similar the displacement as moderate if it was between 25–50% and problems to those described with closed reduction occurred as severe if it was more than 50%. The amount of slipping in that stretching and tearing damage to the posterior capsule was determined by the amount of displacement of the epiphand vessels proved to be extensive and subsequent avascular ysis on the femoral neck expressed as a percent of the diamnecrosis unacceptably high. Open reduction alone for chronic eter of the neck as seen on the one radiograph. Cartilage stable slipped capital femoral epiphysis, without neck short- necrosis was seen in 29 of 77 hips for an incidence of 38%. ening via cuneiform osteotomy, is never warranted in current Thirteen of the 29 hips also had AVN leaving 16 with chonmanagement schemes. drolysis alone. Twenty-one of 29 showed some improvement of cartilage joint space with time. AVN developed in 22 hips 3.2.10.8 Open Reduction and Cuneiform (28.5%). The average time from surgery to diagnosis was Osteotomy 8.8 months with some evidence of necrosis as early as The next approach in treatment evolution was open reduction 2 months. The AVN was felt to be complete in 14 hips and with removal of a proximal wedge of femoral neck bone and partial in 8. Preoperative manipulation contributed meaningphyseal tissue (cuneiform osteotomy) to shorten the rela- fully to the rate of AVN. They had done a few procedures, tively prominent anterior, superior, and lateral portions of the six, with a basilar neck osteotomy, and in fact if these were neck and allow the head to be reduced onto the remaining excluded from the analysis, the rate of AVN went up to 31% neck with no tightening or tearing applied to the posterior and that of chondrolysis to 41%. They did not, however, feel thickened and shortened capsule and periosteum. As early as that there was any documented relationship between chon1898, Alsberg [225] and 1909 Whitman [158] described drolysis and postoperative immobilization. Although a chonopen reduction with cuneiform osteotomy. Whitman noted drolysis diagnosis was not desirable, there was some that it was sometimes possible to reduce the femoral head evidence for recovery presumably because some of the basal onto the neck without removing any bone, which represents chondrocytes had survived. The complications of necrosis of a pure open reduction, while in other instances a wedge- either bone or cartilage were related to the severity of the shaped section of the bone from the femoral neck needed to preoperative slipping or the amount of correction obtained. be removed before satisfactory position could be obtained To a great extent, it appeared that problems were related to which represents one of the earliest reports of a cuneiform the procedure itself. Their conclusion, not surprisingly, was osteotomy. This technique was referred to by Green in 1945 that cuneiform osteotomy of the femoral neck either in the [226] and Martin in 1948 [227] both stressing the importance subcapital or mid-neck region had such a high incidence of of removing a sufficient portion of the neck to allow the head severe complications that it should be abandoned. They indito be repositioned and to gently elevate the periosteum from cated their preference for either basal neck or subtrochanthe neck without damaging the inferior cervical vessels in teric osteotomy. the retinacula of Weitbrecht (Fig. 3.15a, b). The procedure The inherent attractiveness of the repositioning approach at continued to be done over the next few decades although this level, however, continued to encourage surgeons to perrelatively poor results were reported because of the high inci- form it while making great efforts to minimize the negative dence of AVN. sequelae. Dunn of England reported on his approach in 1964 Gage et al. reviewed their results with cuneiform osteot- [229]. He clearly described the six possible ways in which a omy for moderately or severely slipped femoral capital displaced slipped upper femoral epiphysis can present. These epiphysis done over a 35-year period beginning in 1938 and involved an acute traumatic displacement in a hip that was confirmed the highly problematic nature of the approach previously normal, an early chronic slip, an acute-on-chronic with AVN at 28.5% and cartilage necrosis or chondrolysis at slip, a severe chronic slip in which the epiphyseal line was still 37.6% [228]. They performed a detailed review of the litera- open, a severe chronic slip in which the epiphyseal line was ture in relation to AVN with cuneiform osteotomy assessing closed, and the finding of secondary arthritis with deformity in

378

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara Trochanter major

ai

M. vastus lat.

M. vastus lat.

aii

M. gluteus med.

M. gluteus med. SH ax.

M. piriformis M. vastus lat.

aiii

Fossa acetabuli

M. gluteus med.

Caput femoris

aiv

M. gluteus med.

M. quadratus fem. M. gemelius inf. M. obturatorius int. Retinaculum

av

Curette

avi

Retinaculum

avii

M. gluteus max.

3.2 Slipped Capital Femoral Epiphysis

bi

379

bii Anterior arm

Periost

Fig. 3.15 (continued)

Fig. 3.15 Open reduction and subcapital femoral neck osteotomy is increasingly used for direct re-alignment in cases of severe displacement. Although the method had been used for several years in the mid- twentieth century, increased awareness of the ability to protect the vascularity has led to its current wide adoption. (a) The method of Leunig et al. is illustrated here. Subcapital osteotomy along with removal of callus repair bone is done by means of surgical hip dislocation. (a(i)) Greater trochanteric osteotomy, capsule cut being careful not to cut in the posterolateral head-neck region where the blood supply of the head is concentrated. (a(ii)) head is dislocated posteriorly. (a(iii)) periosteum is cut along anterior border of neck, still protecting the posterolateral regions. (a(iv)) The head gently separated from the neck and callus. (a(v)) osteotomy removes anterior based wedge of bone including callus, subcapital neck bone, and repair bone spur at posteroinferior margin of neck. (a(vi)) Curette growth plate cartilage and repair tissue from underside of epiphysis. Reduce the head onto the neck and stabilize with K-wire from the head into the neck. (a(vii)) Final stabilization

with three Kirschner wires through the neck into the head. Remove other Kirschner wire. The head is then reduced and soft tissue layers and greater trochanter repaired. (Republished from Leunig et al. Oper Orthop Traumatol. 2007;19:424–32, by permission Springer.) (b) Some centers will perform the subcapital osteotomy and resection of repair callus without surgical hip dislocation. Maronna reports on the technique from his unit. The trochanter is not osteotomized. Hip joint capsule and then periosteum are carefully sectioned and retracted with great care taken to protect the associated blood supply. (b(i)) The saw cuts to remove the wedge of bone based anteriorly (pink) must not cut the posterior cortex; a wedge in the range of one-half to two-thirds of the bone thickness is removed with the remaining bone removed cautiously with a small rongeur. (b(ii)) The closure of the osteotomy site is then stabilized with three Kirschner wires. The periosteal flaps (triangular-shaped) and the capsular flaps (rectangular-shaped) are then closed. (Republished from Maronna. Oper Orthop Trauma. 2007;19:424–32, by permission Springer.)

380

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

C

Fig. 3.16 Physiologic considerations in performance of open reduction and cervical osteotomy, derived from the works of Dunn, are shown here. Dunn reintroduced the procedure in the 1960s, stressing the need to remove the callus and shorten the neck prior to reduction so that the shortened, thickened, and contracted posterior periosteum and its associated blood supply would not be stretched over the reactive callus and deformed neck with reduction. Improved appreciation of the posterior hip vascular anatomy by Ganz and colleagues in the 1980s further improved safety of the procedure. The two images at the top show the likely occurrence of avascularity of the femoral head when reduction without callus/femoral neck tissue resection is done. The shortened posteromedial periosteum with adjacent vascularity is damaged with stretching over the callus and neck causing the avascularity. Below, the vascular supply (dotted head) is preserved by safe reduction due to the shortening and thinning of the femoral neck with callus/neck tissue resection. C = callus. (Reproduced with modifications from Exner et al. Orthopäde. 2002;31:857–65, by permission Springer-Verlag.)

early adult life. The physiologic considerations are illustrated in a series of drawings in Fig. 3.16. His treatment plan followed in logical fashion with the acute slip which was essentially a type I epiphyseal fracture-separation treated as an emergency with reduction performed within a matter of hours and the femoral head fixed by closed pinning. In the early chronic slip with mild to moderate displacement, the femoral head could be stabilized without manipulation by pinning in situ. In the acute-on-chronic slip, the retinacular vessels will have been shortened so no attempt should be made to bring

about closed reduction with the deformities best treated by open operation with shortening of the neck by trapezoid osteotomy. If the epiphyseal line was open, Dunn felt that repositioning osteotomy was valuable and that open reduction and cervical osteotomy were preferable to a subtrochanteric osteotomy. He stressed the importance of careful technique, entering the hip by a lateral approach, elevating the trochanter, and protecting the retinacular vessels by careful elevation of the synovium from the back of the femoral neck and cervical shortening before replacing the head on the end of the neck. If the epiphyseal line was closed, cervical osteotomy and open reduction had no part to play with deformity to be corrected either by subtrochanteric osteotomy or cervical (femoral headneck) osteoplasty. Dunn thus proposed the open reduction and proximal cervical osteotomy for the acute-on-chronic slip in which neck shortening with the trapezoidal osteotomy was essential and for the severe chronic slip where the epiphyseal line was still open. For the late chronic slip where the epiphyseal line had closed, cervical osteotomy would clearly transect the intraosseous vessels such that even if the posterior retinaculum remained intact, the risk of AVN was far too high. The primary object of the open reduction operation was to place the femoral head on the end of the neck without stretching or compromising the retinacular vessels. This could be achieved in one of two ways: first, the neck can be shortened by a cut just distal to the epiphyseal line and at right angles to the long axis of the neck so as to include about 1/4″ of the posterior aspect of the neck with the segment removed trapezoidal rather than cuneiform as often indicated since the posterior bony beak was also removed. The second point involved care such that vessels on the posterior surface of the neck and head would not be damaged. He suggested an anterior approach to the neck with subperiosteal dissection maintaining the posterior vascularity such that the head could be eased off the neck and the synovium raised from the back of the neck right down to the base of the capsule. The back of the neck could now be seen, the bony beak trimmed, and the superior surface of the neck trimmed. It was then possible for the head to be replaced with the entire posterior synovium of the neck free of tension. Stabilization was performed with a nail and the patient kept at bed rest for 1 month at which time crutch walking began. Union of the epiphysis to the metaphysis generally occurred around 3 months. He reported on 23 open reductions of which 19 did very well going on to become clinically normal. Four hips developed complications involving one segmental necrosis of the femoral head, one complete AVN, and two instances of chondrolysis. He compared his approach to the intertrochanteric osteotomy and supported the former since it fully and directly corrected the anatomical abnormality rather than compensating for it. Dunn and Angel further reported their results in detail in 1978 [230]. They had also reviewed the literature in relation to the open reduction and cuneiform osteotomy for severe

3.2 Slipped Capital Femoral Epiphysis

slips. Their review table reported only those operations that were undertaken as a primary procedure and where particular attention was paid to the blood supply to the femoral head. They noted documentation of 94 cases in reports from 1945 to 1972 with an AVN rate of 14% (13 cases). They separated out the acute slip with a violent injury and the severe chronic slip treated after growth plate closure as those represented different entities. They stressed that open reduction and cuneiform osteotomy should not be performed in the severe slip with growth plate closure because of the unacceptably high risk of AVN owing to intraosseous vessel communication between the metaphyseal and femoral head vessels. They reviewed their assessments using the three- type categorization of an early chronic slip, an acute-on- chronic slip, and a chronic slip with the physis open. The term “acute-on-chronic” slip was used when there was evidence of a new bone on the posterior aspect of the metaphysis indicating that the acute episode had been preceded by a minor degree of chronic slip. Even though some minor trauma might have occurred, the severity of injury would not have been expected to fracture the neck of the femur or displace the epiphysis in an otherwise healthy person. They were thus reluctant to use the term “acute” even as others used it in which there was severe pain less than 3 weeks prior to treatment and inability to bear weight. They reserved the term acute for what is essentially a type I epiphyseal fracture- separation. The open reduction and cuneiform osteotomy were reserved for the acute-on-chronic slip with severe displacement or for the chronic slip with severe displacement with open physis. Seventy-three procedures done over a 23-year period were assessed. No cases of acute traumatic slips were included. There were 25 patients with acute-on-chronic slips and 48 with severe chronic slips. They reduced the femoral head anatomically. It was necessary to remove the posterior bony beak otherwise the retinacular vessels would be stretched over it. A lateral approach was preferred. The greater trochanter was elevated through its growth plate. The capsule was incised in the long axis of the neck with the incision extended around the anterior and posterior edges of the acetabulum. The subluxated femoral head was visible along with the reddish posterior surface of the femoral neck. The front of the femoral neck was pale and avascular. The synovial membrane on the neck was incised in front of the vascular area around the anterior margin of the head. The posterior soft tissues were then stripped subperiosteally to the margin of the head and down to the base of the neck, using extreme caution. The growth plate was visualized peripherally, and a wide gouge was inserted between the cleavage plane through the growth plate between the head and neck of the femur. Dissection was performed on the remains of the growth plate levering the head off the neck. Two osteotomies were performed. The first was along the

381

long axis of the neck to remove the bony beak from its posterior aspect. The second shortened the neck by a few millimeters and was made with a slightly curved sweep transverse to the top of the neck. This removed the remains of the growth plate from the neck and the adjacent metaphyseal neck bone. The remains of the growth plate were then removed from the femoral head. The head at this stage was reduced onto the neck without tension on the posterior structures. If there was any degree of tension, the neck was further shortened. Three pins were drilled up the neck to emerge at different positions. The deformity was reduced and the pins driven further into the head for stabilization. In the lateral view, the head was fully reduced. In the anteroposterior view, it was important not to over position the head into the varus but rather to leave it with a slight valgus position of some 20°. The trochanter was reduced and held with pins, and soft tissue closure followed. Skin traction was maintained for 4 weeks after which crutch walking began. Assessments were made by subjective, clinical, and radiological criteria with results expressed as good, fair, or poor. The major complications assessed were AVN, chondrolysis, and osteoarthritis. In 73 hip operations, the entire series, the clinical results showed 75.3% good, 8.2% fair, and 16.4% poor. Radiologically, the results were less effective but still registered 56.2% good, 16.4% fair, and 27.4%. In review by diagnostic category, there were 7 operations done with a severe chronic slip in which the growth plate was closed. It soon became evident that results in this category were poor and indeed 6 were listed as poor with only one good. This group subsequently was not felt appropriate for open reduction, and toward the latter part of the series these individuals were treated with intertrochanteric osteotomy. The two other groups had improved results. Analysis was then performed of 40 cases of severe chronic slips with the physes open in white children only and 23 acute-on-chronic slips in white children only. The relatively few black patients in the series were not assessed in the subgroupings because of the then widely held belief that blacks did less well with the SCFE disorder. Indeed, of the four black patients in the series, all did poorly. In the patients with severe chronic slip with plate open, the clinical assessment indicated 92% good, 3% fair, and 5% poor, while radiologically there were 75% good, 15% fair, and 10% poor. In the acute-on-chronic group, clinically there were 70% good results, 4% fair results, and 26% poor results, while radiologically there were only 43% good results, 26% fair results, and 30% poor. The patients that did best, however, were those with the severe chronic slip with physis open. AVN occurred in only one case of 40 with open growth plate and no acute displacement. With acute-on- chronic slips, vascular changes were much more frequent. Dunn and Angel felt this was not due to technical reasons since in this group the operation was technically easier but rather due to the fact there had been damage to the vascular

382

pedicle at the time of the acute slip or to kinking of the vessels before the lesion was repaired surgically. Chondrolysis was seen more throughout the spectrum of disorders; there were 13 cases with severe chronic, 3 with acute-on-chronic slip, 3 in 4 of the black patients and 4 in the 7 patients with closed growth plates. A review of the relatively few radiographs shown in the article, using the benefit of knowledge gained subsequently based on the work of Walters and Simon, shows that some of the pins were too long and might have compromised the articular cartilage. In addition, some patients had a wide, rigid, tri-flanged, single nail inserted which may have caused destruction of the vessels posteriorly at the time of passage of the nail from the relatively soft neck into the harder head. Broughton et al. reported a later series where cuneiform osteotomy had been performed in 115 hips [231]. At a mean follow-up of 12 years, there were three cases of AVN alone, two of AVN and chondrolysis, and eight of chondrolysis alone. Rey and Carlioz reported good results with the Dunn procedure [232]. One of the main proponents of cuneiform osteotomy of the femoral neck in the United States was Fish [233, 234]. He assessed 42 hips with sufficiently severe displacement to require surgical correction by means of a cuneiform osteotomy of the neck of the femur just distal to the physis [233]. The purpose, similar to that of Dunn, was to restore the normal anatomical relationship of the head to the neck. His results were excellent in 40 of 42 with only one instance of avascular necrosis and one of osteoarthritis. Fish noted that reviews from the earlier literature reporting the worst results were those in which manipulation was performed prior to surgery, and a Smith-Peterson nail was used for stabilization. The report of Fish involving 42 hips with cuneiform osteotomy indicated that all had slips of greater than 30° of displacement. Seven patients had an acute slip on a pre-existing chronic slip. Cuneiform osteotomy was considered to be the only procedure that could fully correct the deformity anatomically with the basicervical procedure allowing correction for a slip of 50° and the biplanar trochanteric osteotomy for one up to 70°. Surgery was done within 24–48 h after presentation. The patient had the limb elevated on pillows in flexion, abduction, and external rotation with no effort at traction. Sufficient bone had to be removed to allow easy anatomical reduction of the head on the neck. The base of the curved wedge must be in the plane of anticipated correction of the epiphysis. The wedge is removed so that the curved contour of the physis will match the corresponding curved cancellous surface of the neck of the femur. The bone in the posterior aspect is usually removed by a small curette or hand-held curved osteotome. When anatomical reduction of the epiphysis has been accomplished, fixation is obtained by using two to four pins. He concluded that it was possible to perform cuneiform osteotomy of the femoral neck with only slight danger of avascular necrosis of the head of the femur.

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

Patients whose physis had already closed were not candidates for cuneiform osteotomy because of increased damage to the intraosseous vessels. Gentle preoperative positioning without traction, the need for no manipulation, and surgery of a meticulous type done shortly after presentation and certainly within 24 h were essential. Fish reviewed his series 10 years later at which time a long-term clinical and x-ray assessment of cuneiform osteotomy was possible in 61 patients with a slip of greater than 30° [234]. The results were excellent in 55 hips, good in 6, fair in 2, and poor in 3. Pin penetration was noted in all six of the patients who had osteoarthritis and in one patient who had chondrolysis. Complete AVN developed in two patients with segmental AVN in one. Each of these three had an acute-on-chronic slip. Fish thus documented excellent results in 83.3% of 66 patients, good in 9.1%, fair in 3%, and poor in 4.5%. Using the grading categories of Dunn and Angel, good results (excellent + good) totaled 92.4%. The value of the open reduction and cuneiform osteotomy was great in the severe chronic slipped epiphysis, but the riskier group was the acute-on-chronic slip. Since the AVN only occurred in those patients, the precipitating factor appeared to be the injury to the capsular blood supply at the time of the slip and not the osteotomy. Clarke and Wilkinson used a different surgical approach to cervical osteotomy and reported improved results [235]. They utilized Muller’s anterolateral trochanter sparing approach. Following a T-shaped incision in the capsule the large amount of callus protruding from under the anterior edge of the acetabulum was seen. The attachment of the periosteum to the posterior edge of the epiphysis was preserved. The callus was removed with a curved osteotome followed by curettage. Once the repair callus had been removed with the original femoral neck remaining intact, the epiphysis was noted to slip forward into normal place. The femoral neck was rotated laterally to expose the posterior lip of repair at its proximal end, and this was carefully removed. The epiphysis was gently reduced by internal rotation of the shaft and neck following which a cannulated compression screw was inserted for stabilization. The patient was protected in a plaster spica for 8 weeks. Results with this approach were considered superior to the classic Dunn procedure. They reported excellent results in 13, good in 2, and only 1 poor result with avascular necrosis. In the 16 patients, there were two episodes of segmental AVN, two of chondrolysis, and one of osteoarthritis. Figure 3.17a, b show the type of deformity that may be amenable to open reduction with callus removal. DeRosa et al. supported the value of cuneiform o steotomy of the proximal neck of the femur in severe SCFE [236]. They reviewed 27 severe Grade III slips over a 10-year period treated this way. No hips were rated excellent since they felt that such a rating was warranted only for normal hips unaffected by disease or surgery. The results were 19 good, 4 fair, and 4 poor with each of the poor results associated with

3.2 Slipped Capital Femoral Epiphysis

AVN. The AVN rate was 15%. In percentages, the results were good in 70.4%, fair in 14.8% and poor in 14.8%. They concluded that the cuneiform osteotomy continues to have a place in the treatment of severe SCFE and in particular with severe Grade III slips of 60° or greater and open physes. The shortening resection of the neck in particular posteriorly is needed to remove reactive bone. Velasco et al. performed open reduction and cervical cuneiform wedge resection according to the technique of Dunn in 66 hips with moderate to severe slippage [237]. Avascular necrosis occurred in seven cases and chondrolysis in eight. In 48 hips followed more than 10 years, and for a mean of 20.6 years, the results were classified as good in 22, moderate in 16, and poor in 10. They felt that their results were better than studies from pinning in situ with moderate and severe slips. Fron et al. (2000) reported on 50 cases for severe slips using Dunn’s osteotomy at a mean follow-up of 4.5 years [238]. Clinical results were defined as good/excellent in 90% and fair/poor in 10%. There were seven patients (14%) with the major complications of avascular necrosis or chondrolysis. Lawane et al. reviewed 25 cases for severe slips using the Dunn procedure with removal of the greater trochanter [239]. The slips were chronic in 16 and acute-on-chronic in 9. There were 15 good results but also 10 (40%) immediate or late complications. The immediate complications involved four osteonecroses (16%), three chondrolyses, and one mechanical. Each of the AVN and chondrolysis cases evolved into a degenerative joint arthritis. A few other cases, initially considered to be doing well, had late deteriorations. In spite of the theoretical attractiveness of the procedures, technical problems with transtrochanteric detachment and posterior cervical periosteal attachment could lead to imperfect results. They reached the conclusion that the extent of AVN and chondrolysis causing immediate loss of joint function were unacceptable. Their approach was then changed to a direct anterior approach after arthrotomy leaving the trochanter intact. They felt that vascular risk was reduced by “an anterior approach, following anterior arthrotomy and….painstaking detachment of the anterior neck periosteum, simultaneously and progressively [performing] anterior cuneiform resection in the metaphyseal region….” The major point of contention between those favoring open reduction and cuneiform osteotomy and those recommending more distal procedures either at the base of the neck or the inter- or subtrochanteric region is the question of whether or not full anatomic restoration is needed to prevent future osteoarthritis. Although the basicervical and trochanteric procedures are to an extent compensatory, it has not been shown that the compensation leads to a change in the head-acetabular relationship sufficient in itself to induce the osteoarthritis. A detailed study in this regard would be of extreme importance.

383

3.2.10.9 Compensatory Osteotomies Early reports with open reduction going back to the 1930s to 1950s (prior to the work of Dunn and Fish) documented a 30% to 40% incidence of avascular necrosis. While this diminished in later reports with the Dunn procedure to the 15–20° range, it was not eliminated. At this stage in the evolution of the treatment of slipped epiphyses, compensatory osteotomy in safer vascular regions more distal to the a

Fig. 3.17 (a) Radiographic appearance of a severe chronic slip is shown. Competing management schools suggest either pinning in situ, to prevent further slippage, followed by redirectional osteotomy to restore anatomic alignment versus open reduction with excision of callus and cervical shortening. (b) These images highlight the factors with this deformity that need to be considered for the management chosen. At the upper left, in situ pinning with a single screw has stabilized the lesion. The pathologic specimen at upper right shows the extensive repair and remodeling that prevent a simple closed reduction that would have extensive complications of avascular necrosis, as well as being nonreducible. Callus repair is extensive, and a new bone has formed at the posteroinferior surface of the head-neck junction. The hypertrophic callus between head and neck must be excised at open reduction along with removal of some of the adjacent neck bone for shortening. The image at lower left highlights the extensive callus that needs to be resected as well as adjacent subcapital neck bone as part of any realignment by open reduction. At the lower right the anatomy of a severe displacement is shown; A, epiphysis; B, femoral neck bone; C, callus; X, prominent spur of repair bone at posteroinferior surface of neck at concavity of deformity; d, epiphyseal line (growth plate); e, triangular field of cartilage on underside of epiphysis, c to c’, uncovered part of growth plate; d’, epiphyseal line (growth plate), and a and b representing the points of stretching/rupture of the periosteum at its point of attachment to the epiphysis and physis

384

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

b

A b c

d

d’ e c’ a c

B

Fig. 3.17 (continued)

epiphyseal vascularity was resorted to. Osteotomy has also been done in the intertrochanteric and immediate subtrochanteric region to correct the head-neck deformity in a compensatory fashion while minimizing or avoiding the severe complication of avascular necrosis. Crawford reviewed the various levels and types of corrective procedures used [181]. The levels of osteotomy are illustrated in Fig. 3.18a(i). Subtrochanteric/Intertrochanteric Osteotomies Southwick [93] in North America and Imhäuser [107, 240] in Europe both developed similar triplanar osteotomies. The osteotomy removes an anterolateral wedge of the bone that allows for correction of varus, extension, and external rotation deformities. The principles are summarized in

Fig. 3.18a(ii) from several operative approaches. The Imhäuser, often performed as the Imhäuser-Weber osteotomy [241], is intertrochanteric. The final geometric change involves internal rotation of the distal fragment to correct the fixed external rotation positioning of the shaft, flexion of the distal segment to correct the proximal extension deformity position of the head, and valgus repositioning to correct for the varus position of the head. The Southwick procedure is defined as subtrochanteric. Basicervical and Intertrochanteric Osteotomies Subsequently surgeons have performed the triplanar intervention at more proximal levels. An intertrochanteric approach can give excellent correction and is performed closer to the site of the deformity than subtrochanteric

3.2 Slipped Capital Femoral Epiphysis

385

ai

aii

A

1 2

1 P

aiii

Triplanar Osteotomies

A

L

M B

1 3

2

closing anterior wedge

3

1. Anterior Based Wedge • to correct posterior head tilt (flexion osteotomy)

4

2. Lateral Based Wedge • to correct varus head position (valgus-abduction osteotomy) 3. Internal Rotation • to correct external rotation deformity (derotation osteotomy)

bi

bv

C

bii

biii

D

E

biv

bvi

Fig. 3.18 Even if corrective osteotomy is chosen, procedures at differing levels have been described. (a(i)) Osteotomy to regain alignment following moderate and severe slippage can be performed at four possible levels; 1, open reduction with resection of callus and adjacent neck bone (subcapital osteotomy); 2, basicervical osteotomy; 3, intertrochanteric osteotomy; and 4, subtrochanteric osteotomy. Each of these approaches addresses the posterior slippage or hyperextended position, the varus malformation, and the external rotation deformity. (Reprinted from Zilkins et al. Orthopäde. 2010;39:1009–1022, by permission Springer.) (a(ii)) The principles of compensatory osteotomy for moderate to severe slipped capital femoral epiphysis are illustrated. (a(iii)) The Imhaüser osteotomy is widely used for correction. It is an intertrochanteric procedure. The chisel is inserted into the femoral neck at a right angle to the shaft. Osteotomy removes an anterolateral based wedge of bone. The blade plate is placed and the distal fragment internally rotated (to correct retroversion). The blade plate is stabilized with screws and the slip is stabilized with three Kirschner wires. (Reprinted from Schatzker J, (ed). The intertrochanteric osteotomy. Berlin:Springer;

1984, by permission Springer – permission #12, Chap. 3.) (b) Clinical radiographs of corrective procedures are shown in figures (b(i)) to (b(vi). (b(i)) Anteroposterior radiograph shows bilateral slipped capital femoral epiphysis following initial treatment. A mild slip was pinned in situ with an Ace screw and healed uneventfully. The opposite side severe slip was pinned initially with two AO compression screws. (b(ii)) Frog lateral view shows posterior displacement of the head relative to the neck. The position of the two compression screws is seen. (b(iii)) Repositioning osteotomy was later performed on the more displaced hip. An intertrochanteric osteotomy fixed with an AO blade plate improved the varus, external rotation, and extension of the head and neck fragment. (b(iv)) Frog lateral view shows improved position of the head in particular in relation to the flexion component of the osteotomy. Note the changed orientation of the two persisting compression screws. (b(v) and b(vi)) Enlarged images of the more severely displaced hip region, showing anteroposterior (b(v) at left) and lateral (b(v) at right) projections, better demonstrate the extensiveness of correction with triplanar osteotomy

386

although still in the safe range in terms of the proximal femoral vasculature. Kramer et al. have performed a basicervical osteotomy [242] with a similar osteotomy performed by Barmada et al. [243]. In the Kramer procedure, referred to as a compensating osteotomy at the base of the femoral neck, a section of the bone at the base of the neck is removed through the anterior trochanteric region. The osteotomy is proximal to the greater trochanter of the femur such that the abductor musculature is restored to its physiologic position. The procedure serves to correct both the varus, on the basis of the laterally based wedge, and the extension deformity on the basis of the slightly anterior position of the wedge. His group selected for operation patients with 40° or more of deformity on either the anteroposterior or lateral radiograph. The capsule of the hip joint is opened anteriorly, and the amount of bone to be removed is determined primarily by visualization at time of surgery. The osteotomy is considered by some to be extracapsular technically since anteriorly the capsule of the hip joint and the retinacular blood supply extend to the intertrochanteric line but posteriorly the capsule of the hip joint and the blood supply extend only to the junction of the middle and distal thirds of the femoral neck. Compensatory osteotomy thus lies distal to the more important posterior blood supply. The more distal of the two osteotomy lines is made first and is perpendicular to the femoral neck along the anterior trochanteric line. The osteotomy is extended to the posterior cortex that is left intact. The second osteotomy is oblique. The wedge itself is anterolateral to correct both the varus and extension components. The osteotomy has been held with 2–3 pins between the trochanter and neck. The greater trochanter is then reattached. The initial report was on 56 compensating osteotomies although the results were not presented in a particularly detailed fashion. There was a need for additional surgery in two to either increase the valgus position of the head or to change the rotational component. There were four instances of avascular necrosis, but these were felt to be due to falls subsequent to the surgery and compression fracturing of the femoral head. A few instances of chondrolysis were also reported although in one case in particular, it was felt to be due to pin penetration. A detailed long-term review of an extracapsular base of neck osteotomy by Abraham et al. reviewed 36 hips with moderate-severe SCFE [244]. They felt that 90% had excellent or good results and documented no instances of AVN. The procedure involved a two-plane wedge osteotomy based anteriorly and superiorly on the anterior surface of the base of the neck. The distal osteotomy line starts from the lesser trochanter and passes through the growth plate of the greater trochanter. The extent of the wedge at its widest part is generally in the 15 mm range. The cuts converge posteriorly to make a single osteotomy on the posterior cortex. The external rotation component is also corrected as the osteotomy is closed with an internal rotation maneuver. The

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

maximum bone wedge removed is 20 mm. The osteotomy is then fixed with 3–4 cannulated screws. The postoperative care involves partial weight bearing with crutches for 6–8 weeks. Interpretation of their results indicated 21 hips excellent, 11 good, 2 fair, and 2 poor leading to 90% excellent and good results and 12% fair and poor results. There were no cases of AVN. Chondrolysis developed in five hips in three patients. The extracapsular base of neck osteotomy was a safe way to correct moderate to severe SCFE, and AVN had not been shown to develop in any of the 36 hips treated. The bony cuts are performed distal to the medial circumflex artery as it courses along the posterior edge of the hip capsule of the femoral neck. With a severe slip, the amount of correction of the femoral head varus and posterior tilt was somewhat limited with this procedure and with the most severe slips, a perfectly normally head shaft angle could not be restored with the procedure. Removal of a wedge greater than 20 mm in width at its greatest extent was not warranted. Southwick popularized osteotomy through the lesser trochanter for SCFE in which a triplanar correction was achieved [93]. He proposed use of the procedure when the head had slipped from 30° to 70° in any plane. His initial report was based on 28 hips. Southwick wrote on the detailed preoperative planning with templates to mark the wedge angle and size needed. The operation is essentially an anterolateral closing wedge with that part based laterally allowing for correction of the varus deformity into valgus and that part based anteriorly allowing for correction from extension of the head to a more flexed and normal position. The triplanar aspect is fully achieved with internal rotation of the distal fragment to correct the external rotation deformity. Fixation was by a relatively crude external fixation device rather than with the blade plate that came increasingly to be used by others with time. The distal fragment was internally rotated, flexed, and abducted to bring about the triplanar correction. Once the osteotomy healed and the limb was placed in the weight-bearing position, varus deformation would have been corrected to a valgus orientation, the posterior angulation or extension of the head would have been flexed into a normal position, and the external rotation deformity of the lower extremity would have been corrected by the internal rotation mechanism. Southwick noted that if posterior slipping of the head was greater than 60°, it was incompletely corrected by the wedges because tilting of the osteotomy site of more than 60° would produce excessive shortening of the femur. The internal rotation of the distal shaft partially rectified the positioning of the head. Southwick did not specifically pin the physis and noted that by the time the osteotomy had healed, in virtually all instances, the physis had undergone fusion as well. A hip spica was used post-surgery for 8 weeks primarily to protect the epiphysis that had not been stabilized by pinning in situ. In rating his results in the 28 hips, there were

3.2 Slipped Capital Femoral Epiphysis

21 excellent, 5 good, and 2 fair. There were no poor results and no cases of avascular necrosis in the 28 hips. In an additional note, he added that 55 patients had been treated by this method (the later ones not being included in the report) and avascular necrosis still had not been documented. The study assessed patients who had been followed for 5 years or more. There were a few instances of joint space narrowing which would be referred to as chondrolysis although considerable improvement with time was noted consistent with other reports. Supporters of the intertrochanteric and subtrochanteric osteotomy point to its value in not causing avascular necrosis and indicate that the compensatory nature of the correction leaves the head and neck in an appropriate anatomic relationship to the acetabulum. The intertrochanteric procedure became the standard procedure for severe and in some instances of moderate SCFE in many centers over the next three decades. Southwick discussed the level of osteotomy for severe slippage in an editorial in the Journal of Bone and Joint Surgery in 1984 in response to two papers, one of which described excellent results following the cervical osteotomy and the other which described excellent results following the biplanar osteotomy at the lesser trochanter [245]. Southwick reminded readers: “high femoral neck osteotomy has proved to be very dangerous in the hands of most surgeons.” He repeated the rationale for the triplanar approach stressing again the absence of avascular necrosis in the procedure, felt that chondrolysis was not inherent to the procedure, and supported the value of the basicervical procedure of Kramer. El-Mowafi et al. compared extracapsular base of neck osteotomies with the Southwick osteotomyin treatment of moderate to severe chronic slips [246]. Both types of osteotomy were shown to be equally safe and effective procedures with minimal risks of AVN and chondrolysis. Follow-up averaged 3.5 years (1–6). In the base of neck osteotomies (15), there were 86.9% satisfactory results, and in the Southwick osteotomies (18), 90% were satisfactory. Ireland and Newman reviewed 35 intertrochanteric compensatory osteotomies from their unit [70]. They pointed out that Perkins [247] in 1932 and Newman [248] in 1956 had done the procedure as well. The operative corrections achieved were intended to be approximate being based on two plane radiographic appearances of varus and extension. The osteotomy itself, therefore, combined valgus and flexion with a third component being a final derotation to neutral of the distal femoral shaft on the proximal fragment giving the triplanar correction. They did not pin the involved epiphysis noting no tendency to redisplacement after osteotomy. Correction was held with a blade plate. The surgical description indicated only that an appropriate anterolateral wedge was removed after which the femoral shaft was flexed and slightly abducted to close the wedge at which time the leg was rotated to the neutral position and the side plate s tabilized

387

to the femoral shaft. Hip spica was used for 6 weeks. Their results included both clinical and radiological assessment. The clinical results were good 28, fair 5, and poor 2, and the radiological results were a good 21, fair 10, and poor 4. There were no instances of avascular necrosis although chondrolysis was noted in four all of which went on to a radiologically poor result. There was no evidence of further epiphyseal slip post-surgery. They concluded that although 3/4 of the hips had clinically good results, some of these were downgraded because of the radiologic appearance that was based on inadequate correction of deformity. In all cases, this was felt to be either an undercorrection as seen on the lateral radiograph with not enough flexion built into the procedure or a tendency to overcorrection on the anterior posterior radiograph with too much valgus created. Imhäuser has written extensively on his intertrochanteric osteotomy [107, 240]. A report of 55 patients followed 11–22 years post three-dimensional intertrochanteric osteotomy indicated that 40 of the 55 hips examined clinically and radiologically had full ranges of motion, 10 hips showing minimal, and 5 hips significant limitation of motion. Radiologically 73% of the 55 hips were rated excellent or good with 27% showing early degenerative arthritis. Imhäuser felt that the intertrochanteric osteotomy prevented or at least delayed the development of degenerative changes. Weber supported the value of the Imhäuser osteotomy for major slips between 20° and 50° [241]. The procedure was combined with transphyseal pinning. The Imhauser procedure is illustrated in Fig. 3.18a(iii). It is increasingly favored at present in many centers. Kartenbender et al. reviewed 39 hips with severe SCFE treated by the intertrochanteric Imhäuser osteotomy [249]. The reviews were done at a mean of 23.4 years post-surgery (19–27). Good to excellent results clinically were seen in 77% and radiologically in 67% with only two cases of AVN (5%). Schai and Exner reviewed 51 Imhäuser intertrochanteric osteotomies done for moderate to severe slips at a mean of 24 years (20–29) post-surgery [250]. There was only partial femoral head necrosis in one hip joint (2%). In those cases, the epiphyses were stabilized with two 3 mm Steinmann pins and the osteotomies with 90° AO adolescent plates. At final assessment 28 (55%) were clinically asymptomatic and radiologically free of degenerative changes, moderate changes for 14 (28%), and advanced changes in 9 (17%). Saisu et al. reported on 32 intertrochanteric flexion osteotomies for symptomatic patients with FAI after SCFE [251]. The mean follow-up was 5 years (2–9). At last followup, only 2 of 32 patients continued to complain of pain. Radiologic evaluation using a modified alpha angle (beta angle) on a Lauenstein view showed reduced angles by an average of 39°. Fineschi and Guzzanti described a linear intertrochanteric osteotomy for severe slips in which deformity was corrected

388

by manipulation of the proximal fragment to position the head in a normal centered position in the acetabulum, rather than by removal of specific wedges, followed by insertion of a blade plate [252]. They followed 21 cases from 4 to 12 years post-surgery and noted no episodes of avascular necrosis. The purpose of the procedure was to align the head appropriately with the acetabulum with no concern about the compensatory deformity of the neck and trochanteric regions. They reported excellent functional results in 18 of 21 cases. Of the other three, two had slight restriction of movement, and one was graded as poor because of stiffness. Their favorable impression of the inter- or transtrochanteric osteotomy was based on its ability to re-center the epiphysis efficiently with no to minimal risk of avascular necrosis. In centers doing relatively large numbers of realignment procedures and having experience in several types of operative approach, good to excellent results are frequently demonstrated. Ballmer et al. reported on 63 slipped capital femoral epiphyses treated either by femoral neck osteotomy or by an intertrochanteric osteotomy [253]. At an average, follow-up of 10 years, 90% were rated good to excellent clinically. In 33 hips with a follow-up of more than 10 years, mild degenerative arthritis was present in 36%. They felt that slipping of 60° or less was best treated by the intertrochanteric osteotomy with the femoral neck osteotomy, which had a higher complication rate, reserved for slipping of 60° or more. Szypryt et al. also compared two operative approaches for moderate or severe slips [254]. They assessed 23 patients who had undergone a Dunn’s open reduction and 30 hips treated by epiphyseal arrest and osteoplasty as advocated by Heyman and Herndon. The 11 hips with moderate slip (30– 50°) treated by the Heyman-Herndon procedure did significantly better than 18 hips with severe slip (greater than 50°) treated by the same method. The Dunn procedure was more effective in those with severe slips displaced greater than 50°. The Kramer osteotomy provides correction of 50° and the Southwick procedure correction of 60°. Carlioz et al. reviewed operative intervention for SCFE in 80 cases [255]. They detailed their guidelines from earlier studies which involved in situ fixation for displacement less than 30°; in situ fixation followed by corrective osteotomy at the intertrochanteric level for those in which there was 30–60° displacement; osteotomy through the femoral neck (cervical osteotomy, Dunn procedure) for displacement between 60–90°, and closed reduction and screw fixation for acute slipping. In no instances was casting used. When the open reduction and cuneiform osteotomy were used, they reported 20 good results, 3 fair, and 4 poor. There was relatively little use of the intertrochanteric osteotomy, but results in five cases were four good and one poor. In the open reduction cases, of the four poor results, three involved cases of chondrolysis. They concluded that the open reduction with cervical osteotomy was “difficult and dangerous” and that it

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

should be used rarely and only in those with extreme displacement. They ultimately recommended the triplanar trochanteric procedures as being much safer. They revised their earlier protocol recommending the open reduction only for slipping of 90° of greater with the growth plate open. With such a degree of severe deformity with the growth plate closed, intertrochanteric osteotomy was mandatory. They supported the belief that AVN never occurred spontaneously in the natural history of slipped epiphysis and that it was essentially always a complication of intervention. Chondrolysis could occur spontaneously but was generally associated with management involving such criteria as closed reduction, penetration of the screw into the joint cavity, excessive immobilization in cast, or excessive degrees of valgus or flexion during corrective osteotomy. There are two possible times of intervention for any compensatory osteotomy. One approach is to do both head-neck stabilizing and repositioning procedures at the same time. The other is to perform an in situ stabilization pinning and then wait several months before performing the corrective osteotomy. This latter approach is preferable in the view of some for two reasons. Most importantly, the complication of hip stiffness has been reported when osteotomy is done at the same time as pinning. In addition, some patients do not find the slight external rotation position of the lower extremity either cosmetically or functionally troubling and elect not to proceed with re-positioning osteotomy. It will be important for clinical studies to determine the long-term sequelae of leaving moderate and even severe slips without corrective osteotomy. Examples of osteotomies to correct severe deformity are shown in Fig. 3.18b.

3.2.10.10 C urrent Conflicting Approaches to Treatment in Moderate and Severe Slipped Capital Epiphysis Studies and Arguments Supporting In Situ Pinning as the Initial Treatment for All Mild, Moderate, and Even Severe Slips Although it was widely accepted that pinning in situ was warranted for mild slips, universal agreement has not been reached concerning treatment of moderate and severe slipped capital femoral epiphysis. Many still recommend pinning in situ of a slipped epiphysis regardless of its extent followed by observation of the patient and performance of compensatory osteotomy at a later date and sometimes only with meaningful symptoms as an indication for intervention. Many patients handled under this approach therefore would have the pinning alone with no osteotomy ever performed. This is particularly true in mild and moderate slips. The tendency to need surgical realignment has generally been greater with the severe slips, but even here some centers have stabilized the slip in situ, allowing patients to go without correction and noting relatively minimal long-term symptoms.

3.2 Slipped Capital Femoral Epiphysis

Post-pinning Remodeling The early effectiveness of pinning in situ alone even in the presence of some deformity relies on the fact that patients can compensate for the mild to moderate deformity. They can often do so symptom-free or with only minimal symptoms for several years. The radiologic appearance of the femur and hip ranges of motion become somewhat improved over months to a few years by remodeling of the head-neck region in association with repair and continued use. Walking in a comfortable fashion is certainly possible with slightly increased external rotation. O’Brien and Fahey showed how femoral neck remodeling improved the radiographic appearance of the proximal femur several years after pinning in situ [256]. In a subgroup of 12 patients with moderate and severe displacement assessed 2–17 years post pinning all, but two had satisfactory remodeling of the head and neck and were asymptomatic. These are patients with a clear pistol-grip deformity of the proximal femur seen on plain radiographs. Even the two with minimal remodeling were asymptomatic. They also reported some spontaneous correction of the external rotation deformities. Cervical Osteoplasty of Heyman et al. to Improve Hip Motion In those where failure of remodeling has become a problem, they recommended the cervical osteoplasty of Heyman et al. to improve motion and relieve discomfort [165]. The bony prominence at the superolateral aspect of the femoral neck at the head-neck junction that formed due to repeated cam impingement at the anterior rim of the acetabulum was surgically removed and the area smoothened by the Heyman cervical osteoplasty (Fig. 3.10a, b). This was often done in an effort to protect hip joint comfort and function without the need for osteotomy. This approach continues to be used. Longer-Term Studies Following In Situ Pinning Bellemans et al. reviewed 59 hips in 44 children with SCFE, all treated by pinning in situ [257]. The average clinical and radiologic follow-up was 11.4 years, and they noted 53 hips (90%) to be excellent or good. Their study assessed postoperative remodeling which was accomplished by local resorption and apposition of the bone, and also, they felt, by correction of the disturbed anatomic axes in proportion to the severity of the slip along with global thickening of the femoral neck. Resorption of the superolateral prominent portion of the metaphysis of the femoral neck was noted in 54% of cases, and apposition of a new bone at the posteroinferior aspect of the neck was seen in 59%. The average frog-leg head-shaft angle of Southwick decreased an average of 13.5° from 25° immediately postoperatively to 12.5° at latest follow-up. The AP head-shaft angle decreased an average of 7°, from 16° on the first postoperative radiograph to 9° at latest follow-up. The average thickness of the neck was also significantly increased with time compared with the

389

c ontralateral normal side. Favorable results could be obtained with pinning in situ owing to the global remodeling process since the remodeling was more extensive than had been reported previously. Owing to the fact that pinning was felt to have fewer complications than redirectional osteotomies, the simpler procedure provided very satisfactory long-term results owing to the remodeling processes. The Southwick lateral head-shaft angle was calculated on the frog-leg lateral radiographs by subtracting the head-shaft angle on the normal side from that on the affected side. The average slip in this study according to the Southwick criteria was 25° with 56% of the slips mild (less than 30°), 39% moderate (30– 60°), and 3% severe (greater than 60°). Jerre studied hip motion at an average of 32.7 years after SCFE in 128 hips without signs of osteoarthritis [258]. They concluded that hips with no treatment or those treated with fixation in situ only showed no clinically significant loss of hip motion as compared with normal hips. The greatest loss of motion of the hips in those treated with fixation in situ was diminution of internal rotation. They concluded that the loss of hip motion after fixation of the epiphysis in situ over the long-term was “very slight and hardly clinically relevant” and that there was no indication for early surgical intervention with osteotomy. Siegel et al. studied 39 patients 2 years after pinning in situ for slipped capital femoral epiphysis [259]. The study was done to assess range of motion of the hip and femoral remodeling. Although there was considerable increase of motion from the preoperative state, they felt that this motion returned despite minimum bony remodeling. The greatest percentage of motion of the hip returned within 6 months after treatment. This served in particular to increase the amounts of flexion, abduction, and internal rotation. They attributed this increased motion to relief of pain, spasm, and synovitis and to subsequent soft tissue stretching. Plain radiograph and CT imaging studies assessed bone remodeling itself. In spite of smoothing of the contours of the head- neck axis with resorption of bone from the superior surface of the neck and synthesis of bone inferiorly and posteriorly, there was only minimal change in the relationship of the femoral head to the shaft and no change in the angle between the femoral neck and the shaft after fixation in situ. Remodeling and resorption thus led to a smoothing of the contours of the proximal femur without change in the axes of the deformed bones. DeLullo et al. noted some femoral remodeling with time that they felt contributed to slightly improved outcomes [214]. An evaluation of 38 hips pinned in situ at a mean follow-up of 7.6 years (1.4–26) showed slight improvements in head-shaft angle from 143°+/−13.9° to 140°+/−13.3°. Angulation decreased in 22/38, 58% of hips. Aronson et al. strongly support the value of pinning in situ for the stable slip regardless of the severity of displacement

390

[260]. They stress that stabilization of the slip and prevention of AVN and chondrolysis are paramount and feel that there is no role for early realignment procedures. They stressed that a stable slip will not reduce with intraoperative manipulation and that open reductions are still associated with too many complications. This is particularly the case if osteotomy is done at the same time that the slip is stabilized.

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

however, in which slippage exceeded 40°, osteoarthrosis was present in 15 of 16 even though correction by osteotomy was adequate. They concluded that the intertrochanteric approach did not prevent degeneration in cases with the most severe slips and recommended only fixation in situ without realignment with correction only of the rotational component during the adolescent years. Follow-up was longer than in most studies being a mean of 9 years with a range from 3 to Complications of Results from In Situ Pinning 26 years but still far from indicating even mid-adult results. to Realignment Osteotomies; Relationship of AVN The surgical approach had involved either a valgus derotato Poor Results tion osteotomy or a formal Southwick procedure. They used Hall found AVN to be the commonest cause of a poor result a relatively rigorous grading system for osteoarthrosis of the [168]. The percentage with AVN was calculated for each of hip that was based on radiographic rather than clinical criteseveral treatment groups in a major study of 173 hips with ria. The inference was clear that many of these patients SCFE. The pattern of AVN worsened with the incidence of would go on to symptomatic hips although at the time of manipulation and the more proximal positioning of osteot- assessment, the problem was primarily radiographic. This omy in relation to the physis. When pinning was performed report is somewhat disconcerting about the long-term results with the narrow Moore pin with or without manipulation, with 40° slippage being the border below which good results there was no AVN. It increased to 5% with a Smith-Petersen could be achieved and above which osteotomy resulting in nail without manipulation and was 9.1% with non-operative good correction did not seem to prevent some degree of treatment that did however include manipulation. degenerative change. The report also provided an excellent Subtrochanteric osteotomy had a 10.9% incidence of AVN, overview of results in other studies following in situ fixation non-operative treatment without manipulation 12.5%, Smith- alone without reorientation of the head and neck region using Petersen nail plus manipulation 37.5%, and all types of cer- either metal pins or transphyseal bone grafting. Although a vical osteotomies 38.1%. The results of these treatment large number of papers are referred to, the gradation involves profiles regarding the incidence of AVN set a reflected pat- mild, moderate, and severe cases with the analysis involving tern that, to a great extent, was reinforced by findings over only the percentage classified as excellent or good. Since in the next 50–60 years. all series those with in situ pinning tend to be extensively Realignment operations in various forms, while widely populated by mild to moderate slips, this analysis may not be practiced for moderate and severe slipped capital femoral valid for the entire spectrum of the disorder. They interpreted epiphysis, still carried the possibility of short-term complica- the results, however, to indicate that internal fixation by pintions although it was widely accepted that femoral head-neck ning or by bone graft epiphysiodesis without realignment position closer to the anatomic norm would minimize osteo- produced long-term results that were good, even in cases arthritis later. Jerre et al. assessed realignment procedures in with moderate to severe slippage. Review of the intertro37 hips at an average follow-up of 33.8 years [261]. The chanteric osteotomy assessments from the literature, howrealignment procedures had been performed from 1946 to ever, indicated that the question of subsequent arthrosis had 1959. They noted serious short-term complications in 7 of 22 not been assessed in detail. The cervical osteotomy did have hips treated by subcapital osteotomy, 3 of 11 hips treated by many instances of poor results documented owing to arthrointertrochanteric osteotomy, and 3 of 4 hips treated several sis and AVN. They present the argument that even moderate years previously by manipulative reduction. They concluded to severe cases of SCFE in the adolescent period should be that the natural history of slipped epiphysis was “probably treated without realignment using only pinning or bone graft not improved by any of the treatments used in our study. We epiphysiodesis. Subsequent correction would be performed therefore discourage the use of subcapital and intertrochan- only for symptomatic states rather than attempting realignteric osteotomy as well as manipulative reduction for the pri- ment for all to prevent such problems. mary treatment of chronic slipped upper femoral epiphysis.” Another long-term study by Schai et al. assessed 51 Results following severe slippage, even with intertro- patients with unilateral severe SCFE of 30–60° treated by chanteric osteotomy, were often far from perfect as demon- intertrochanteric osteotomy and examined an average of strated in a long-term follow-up study of 26 patients by 24 years post-surgery [263]. They concluded that 55% Maussen et al. in 1990 [262]. They reviewed 26 of their own showed neither radiographic nor clinical signs of degeneracases of moderate to severe SCFE treated by the intertro- tive hip disease with 28% having moderate disorders and chanteric approach. Where slippage was less than 40° in 10 17% severe osteoarthritis. Intertrochanteric osteotomies hips, subsequent arthritis was present only in 1. In 16 cases, were performed with stabilization by either AO blade plates

3.2 Slipped Capital Femoral Epiphysis

or AO condylar plates. Stabilization of the epiphysis pre- osteotomy with two to three Steinmann pins was important. The principle of re-establishing hip anatomy indirectly, by compensatory means through the intertrochanteric region, had been introduced by both Imhäuser and Southwick. They concluded that the results were superior to pinning in situ alone. They felt that angles greater than 30° warranted corrective osteotomy. Prophylactic Pinning of Contralateral Side at Initial Presentation The relatively high incidence of bilaterality noted in patients followed to skeletal maturity plus the fact that many patients at presentation have only a unilateral slip has led to the practice in many centers of pinning the contralateral normal hip at the same time that treatment is undertaken for the slipped epiphysis so as to eliminate immediately any chance of contralateral slipping. The advantages of the approach are the ability of the patient to resume full activity at all levels once both physes have fused with no risk of subsequent slippage. Some of the second-side slips are asymptomatic, and the possibility exists that the second slip will not be recognized until late in its evolution such that a moderate or severe deformity is presented for treatment. Since the results are best in those with minimal to no slippage, there is value in pinning in situ. Disadvantages have resulted from prophylactic pinning however for two reasons. In most series, although approximately 50% of patients had bilateral involvement, 25% of them were bilateral at time of presentation such that only 25% experienced the second-side slip over the ensuing period prior to skeletal closure. If all patients in such a group had been pinned prophylactically, three of four operations would have proven to be unnecessary. Jerre et al. recommended that prophylactic pinning of the contralateral hip should not be standard [125]. In reviewing 61 patients treated for unilateral slipped upper femoral epiphysis, there were 14 (23%) who had evidence of bilateral slipping at initial primary review, while subsequently 11 (18%) slipped prior to skeletal maturity. They concluded that if all 61 contralateral hips had been pinned prophylactically at primary admission, 36 of the operations (59%) would have been unnecessary. They thus recommended that radiographs be done every 3–4 months until growth plate closure with only those hips where a slip occurred to be pinned. The reason for x-rays was that in only 2 of 25 patients with bilateral involvement (8%) was slipping of the contralateral hip symptomatic. Of greater concern however was the finding in some retrospective series that the contralateral asymptomatic hip had a complicated pinning such that damage was caused even though none was present initially. With greater awareness of the dangers of inappropriate pin placement, these complications have minimized.

391

In many centers the patients and families are presented with two options. One simply involves pinning in situ at the same time that the primary slip is treated. The other is a recommendation to follow the child closely throughout the remaining years of growth and to intervene surgically only if a slip develops. Instruction is given into the importance for immediate orthopedic assessment for warning signs of early slippage with the development either of limp, however slight, or hip, thigh, or knee discomfort. Most will also take the further precaution of recommending radiographs every 3–4 months since some slips on the opposite side are asymptomatic and limiting sporting activities. Those clinics showing the highest extent of bilaterality are most supportive of the value of the prophylactic approach. Jensen et al. recommended bilateral pinning at initial treatment in all patients with a SCFE [128]. Engelhardt strongly recommends prophylactic treatment and quotes many papers from the past few decades showing the incidence of bilaterality ranging from a low of 19% to a high of 80% in the Billing and Severin study [264]. He prefers CT assessment when the child is approaching the age of skeletal maturity since growth plate closure is seen more clearly and thus earlier by CT than by plain radiography. Hagglund recommends prophylactic pinning of the contralateral hip in all cases of SCFE, with the proviso that the technique used should have a low complication rate [265]. In an effort to be predictive of which patients will develop a second-side slip, information continues to accumulate allowing for some selectivity. Riad et al. found that chronological age was the only significant predictor for the development of a second-side slip (among the clinical and epidemiologic factors they studied) [133]. All girls younger than 10 years of age and all boys younger than 12 years of age presenting unilaterally developed a second-side slip. Of potentially greater value is the radiologic index found by Phillips et al. whereby 83.3% of contralateral hips with a posterior sloping angle of greater than 14° on the frog lateral radiograph developed a secondside slip [151] (see Sect. 3.2.9.1).

3.2.10.11 Relationship of Femoroacetabular Impingement (FAI) to SCFE Management and Post-2000 Trends in Treatment of SCFE Beginning around 2000, the concept of femoroacetabular impingement (FAI) was becoming increasingly well defined, and surgical management approaches were developed and, in some centers, aggressively applied. FAI is particularly relevant to SCFE where displacement of the femoral head (capital epiphysis) posteriorly and external rotation of the increasingly uncovered neck favor impingement of the superolateral neck against the acetabular labrum and articular cartilage. [The Femoroacetabular Impingement (FAI) entity is also discussed in detail in Chap. 4.] With greater

392

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

recognition and definition of the femoroacetabular impingement (FAI) syndrome and strongly interventionist bias of many in efforts to decrease the hip degeneration secondary to childhood and adolescent hip disease, the basic approaches to treating SCFE are undergoing some change. Millis and Novais comment on in situ fixation as follows: “femoroacetabular impingement causes some mechanical abnormality in every hip affected by SCFE, even when the slip is mild…” [216]. These issues have raised concerns about the approaches to therapy for SCFE, contrasting in situ pinning with immediate restoration of anatomic femoral head-neck anatomy based on the premise that this will eliminate femoroacetabular impingement. These two approaches can be introduced by addressing the following matters: (a) Immediate realignment procedures to restore normal femoral head-neck anatomy with mild, moderate, and severe slips versus in situ pinning regardless of degree of displacement followed by use of osteoplasty/osteotomy only when based on symptom development (b) Long-term value of initial realignment procedure versus complications associated with them Fig. 3.19 The concepts of femoroacetabular impingement (FAI) are outlined in these illustrations. A represents an anteroposterior view of the hip, while B represents the view with flexion as the head-neck junction is brought into close relationship to the anterior acetabular rim. Cam impingement is shown at left in A and B (this is the common deformity in slipped capital femoral epiphysis with reduced femoral head-neck offset); pincer impingement is shown at the right in A and B (this occurs with excessive overcoverage of the femoral head by the acetabulum). The views in flexion (B) show the neck impingement against the labrum. (Reprinted from Beck et al. Clin Orthop Relat Res. 2004;418:67–73, by permission Wolters Kluwer Health, Inc.)

a

b

Femoroacetabular Impingement Ganz and colleagues have played a major role in outlining the FAI entity and defining it as the primary underlying causes of osteoarthritis (OA) of the hip [266]. They feel that FAI is the mechanism responsible for the development of early OA for most non-dysplastic hips. Abnormal contact between the proximal femur and the acetabular rim during the terminal flexion motion of the hip leads to lesions of the acetabular labrum and adjacent acetabular cartilage. When this occurs in the adolescent, continuing activity causes the lesions to progress leading to degenerative joint disease. The main mechanism in slipped capital femoral epiphysis is caused by the cam impingement mechanism based on the non-spherical head and disturbed head-neck junction (Fig. 3.19). The superior labrum is damaged along with the anterosuperior acetabular articular cartilage with eventual separation between the labrum and cartilage [266]. This mechanism is activated in SCFE as even mild slippage leads to an aspherical junction between the head and neck causing the pistol-grip deformity. This offset is measured by the α (alpha) angle as described by Notzli et al. [267]. As the femoral head slips off the neck at the physeal level, the anterior superolateral neck (metaphysis)

3.2 Slipped Capital Femoral Epiphysis

becomes progressively uncovered and prominent, and camtype impingement occurs as the abnormal head-neck junction is driven, with flexion, into the acetabulum producing damage to the labrum and cartilage at the anterosuperior rim. As explained by Beck et al.: “……..a cam is an eccentric part added to a rotating device. During flexion the eccentric part (supero-lateral neck) slides into the anterosuperior acetabulum and induces compression and shear stresses at the junction between the labrum and the cartilage and at the subchondral tidemark” [59]. Femoroacetabular Impingement Specifically Related to SCFE In relation to his impingement concern, Leunig et al. (2000) explored 14 consecutive hips with SCFE: 6 early chronic, 3 acute-on-chronic, and 5 chronic with slip severity graded as mild (60°) in 1 [55]. A lateral approach was used to each hip, followed by osteotomy of the greater trochanter and capsulotomy. This allowed for inspection of the femoral head and acetabular cartilage, labrum, and the superolateral-anterior metaphysis (femoral neck) by either hip subluxation or dislocation. Treatment was either by in situ pin fixation with femoral neck osteoplasty, intertrochanteric Imhäuser osteotomy, or Dunn subcapital osteotomy. In all hips inspection was done with impingement visually noted with hip flexion. In all hips similar pathologic findings were noted. The metaphysis (superior-anterior neck) impinged against the acetabulum and rough edges with bleeding ulcerations of the metaphyseal surface were demonstrated in seven hips. Flexion and internal rotation at the hip caused impingement of this metaphyseal area against the adjacent labrum and acetabulum. In mild to moderate slips, the metaphyseal region with flexion actually entered the acetabulum but in severe slips, it was so prominent it could not pass beyond the rim. The labrum invariably showed damage in terms of deformation, erosions, scars, or inner surface tears. Even in mild slips, there was chondromalacia of the acetabular cartilage proceeding to cleavage from the underlying bone and occasional flaps or full-thickness loss. Damage from the rim inward extended to 1.0–1.5 cm and a length of 3 cm. The authors concluded: “mechanical jamming seems to be the major factor causing direct and early mechanical acetabular rim and cartilage damage, triggering arthrosis in SCFE.” Over the next 15 years, many cases of SCFE at all ages and degrees of severity were assessed in variable ways using the FAI framework. Fraitzl et al. studied 16 patients who had in situ pinning for mild slips (mean angle 17°, range 12–31°) at a mean period of 14.4 years (11.3–21.2) prior to assessment [268]. The study was designed to see if those hips with mild slips were developing evidence of the FAI syndrome in early adulthood. Measurement of the α angle of Notzli et al. was one of the main parameters studied. The angle was

393

abnormal on the slip side (compared to the unaffected side) by statistically significant margins. The normal α angle is considered to be 42°. The mean α angle on the anteroposterior radiographs was 86° (55–89°) in the affected hips and 62° on the other side. On the lateral, the mean values for SCFE/non-affected sides were 55°(40–94°)/46°. The patients had minimal clinical problems, other than decreased internal rotation and abduction, although flexion was normal, but the radiological findings were significantly altered into the FAI range. They interpreted the findings as supportive of restoring the anatomy of the proximal femur (by osteochondroplasty at the femoral head-neck junction) to avoid or delay development of FAI and osteoarthritis following SCFE. Dodds et al. also studied a group of patients (49 hips) having in situ pinning using a single cannulated screw at a mean of 6.1 years (2.2–13.1) previously [269]. All patients had reached skeletal maturity. Hip pain and/or stiffness were experienced by 15/49 of the patients (31%) while 16/49 (32%) had clinical signs of impingement. The anterior offset angle measured the severity of deformity and was the most influential variable. The mean value was 62.1° (+/−11.9°) for the slip side and 56° (+/−10.1°) for non-impingement. There was no pattern between slip grade and pain. Castaneda et al. followed 121 patients with a stable SCFE treated with in situ fixation reviewed at a follow-up of 22.3 years (20.1–32.5). The slip was grade 1 (34 hips), grade 2 (65), and grade 3 (22) [270]. Clinical and radiographic signs of FAI were seen in 96/121 and radiographic signs of OA in all 121 (Tönnis grade 1, 14; grade 2, 32; and grade 3, 75). The occurrence of FAI (or a pistal-grip deformity) was common even after a low-grade slip. Sink et al. assessed acetabular articular cartilage and labral damage in 39 hips undergoing open surgical hip dislocation for symptomatic stable SCFE [58]. The radiographic categorization of deformity was mild 8, moderate 20, and severe 11. Labral injury was observed in 34/39 (87%) hips and cartilage injury in 33/39 (85%) hips with a range in both throughout the Beck categorizations of severity. Significant acetabular chondromalacia and labral injury were noted in hips with SCFE with open surgical dislocation allowing for direct confirmation of the mechanism of impingement of the anterior superolateral metaphysis on the acetabular cartilage and labrum with hip flexion. In a further effort to assess acetabular cartilage and labral damage after mild SCFE, Lee et al. presented five case descriptions [60]. Each patient was undergoing arthroscopy and femoral osteoplasty within 18 months of in situ stabilization, and all were symptomatic with groin pain or limping or both. Slip angles were 28°, 21°, 15°, 31°, and 29°. All patients had acetabular cartilage or labral damage or both. Descriptive terms used for the impingement caused irregularities including synovitis, labral hyperemia, anterior acetabular chondral fissure with flap tear, chondral fissure

394

(partial thickness), labral fraying, acetabular cartilage fibrillation, chondral surface abrasive changes, partial thickness labral tear, and roughened metaphyseal neck contact region. The authors concluded that mild and moderate slips can lead to joint damage within weeks even in the absence of symptoms and pointed out the need to determine whether immediate osteoplasty soon after or at the same time as in situ fixation for these hips was warranted. At the other end of the age spectrum, Abraham et al. compared femoral head and neck specimens removed at time of total hip arthroplasty from patients with SCFE (16) and primary OA (84) [57]. There was premature development of arthritis in the SCFE patients who were on average 11 years younger than the other group. The pistol-grip deformity characterized the SCFE specimens. These differed from the general OA group having: (i) loss of head-neck offset, (ii) acetabular neck impingement, and (iii) loss of superior peripheral articular cartilage adjacent to the superolateral neck. Each of these three findings was different from observations of the general OA specimens. They felt that the SCFE specimens were clear examples of cam impingement. In addition, where acetabular observations were made, the superolateral acetabulum was often denuded of articular cartilage and the labrum absent or destroyed. The offending factor was the loss of femoral head sphericity at the superolateral and anterolateral neck left exposed after slippage of the head. Therapeutic Responses to the FAI Observations in SCFE The mechanisms of femoroacetabular impingement playing a major role in causing degenerative hip osteoarthritis have helped to explain why all variants of childhood hip disease can eventually lead to adult hip problems. The therapeutic implications of this relationship for developmental dysplasia of the hip and Legg-Calvé-Perthes disease are indirect since the goal in treating these two disorders beginning at the time of diagnosis is to enhance development toward a normal hip. The relationship of impingement to OA is much more direct in SCFE, however, since the established initial treatment profiles come into question even in mild and certainly moderate degrees of slippage. Those fully supportive of the observations that even mild slips predispose from the beginning to acetabular cartilage and labral degeneration express two interpretations: (i) that these changes will progress inevitably toward a degenerative arthritis of the hip if the proximal femoral anatomy is not immediately restored completely to normal and (ii) that anatomical restoration to normal should be the goal of treatment in all cases of SCFE regardless of the degree of slippage or, at a minimum, should be the goal with all moderate and severe slips. Two technical procedures have been adapted and modified to relate to treatment of FAI in SCFE to bring about anatomical head-neck restoration. These are (i) hip arthroscopy to perform an

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

osteochondroplasty of the proximal femur to trim the prominent neck in mild slips and restore femoral head-neck (offset) anatomy and (ii) open surgical hip dislocation with improved preservation of proximal femoral vascularity to allow for direct inspection of the acetabulum, femoral head epiphysis, and femoral neck and intervention to restore normal proximal femoral relationships with open reduction followed by in situ pinning in unstable/acute slips or the modified Dunn procedure in stable/chronic slips. SCFE is one of the best-studied disorders in pediatric orthopedics and has been for several decades. There are two major immediate complications of treatment in SCFE, avascular necrosis of the femoral head and hip joint chondrolysis. There is then a longer-term degenerative hip osteoarthritis that develops either a few to several decades later in most but not all patients (as long as the immediate complications have not occurred) but which develops much earlier (in late adolescence or early adulthood) if avascular necrosis and/or chondrolysis following treatment have occurred. Of the two major complications, avascular necrosis is invariably iatrogenic and chondrolysis frequently follows treatments of all types. Both are extremely damaging to clinical function from the moment they occur, essentially destroying the hip and requiring arthroplasty in late teenaged or early adult years; whereas degenerative joint disease occurring solely on the basis of cam impingement can evolve slowly over several decades before total hip arthroplasty in some is warranted. Stated another way, AVN and chondrolysis cause more problems to the patient at a much younger time period than would the natural history of the disorder, regardless of severity, where in situ stabilization alone was done. The primary treatment for SCFE in the chronic variant, pinning in situ, has virtually no AVN and an extremely low incidence of chondrolysis associated with it, although long-term (several decades) OA can occur even in mild cases. It thus becomes incumbent on those using aggressive interventions at time of slip or shortly afterward to show not only that these yield perfectly restored anatomy but also do not cause the iatrogenic complications which meaningfully worsen the clinical status of a patient even in late adolescence. Studies on treatments based on the FAI concerns in SCFE follow. In Situ Pinning Accompanied by Arthroscopy with Osteochondroplasty This technique is generally used for mild slips although some have also applied it to moderate variants. Two approaches are being used: the first is to perform an in situ pinning and the arthroscopic osteochondroplasty at the same time; and the second is to stabilize the slip with in situ pinning and then do the arthroscopic osteochondroplasty at a later date either electively as an essential part of management or as a response when symptoms of FAI start becoming troublesome. Leunig et al. reported briefly on three cases of in

3.2 Slipped Capital Femoral Epiphysis

situ pinning for mild chronic slips with immediate arthroscopic osteochondroplasty at the same time [61]. Slip angles were between 15° and 30°. Arthroscopic examination showed labral fraying, acetabular chondromalacia, and a prominent metaphyseal ridge. After these procedures patients reported being pain-free with improved hip motion and normalized epiphyseal-metaphyseal offsets and α angles. Wylie et al. reported on nine hips undergoing arthroscopic treatment of mild to moderate deformity at a mean time of 58.6 months (18–169) after in situ pinning [271]. The arthroscopic osteochondroplasty was done in response to symptoms limiting physical activity. The α angle improved from a mean 75° preoperatively to 46° postoperatively. Symptoms were improved to variable degrees in eight and remained the same in one. The mean follow-up was 28.6 months (12.6–55.6). All patients had some degree acetabular cartilage or labral pathology at time of surgery. The femoral osteochondroplasty reshaped the femoral head-neck offset, removing the impingement lesion. Range of motion into flexion and abduction demonstrated improved joint congruity during and after tissue resection. Treatment of Moderate and Severe Chronic/Stable SCFE Using Modified Dunn Approach (with Open Surgical Hip Dislocation) Studies on the vascularity of the femoral head [272] and early reports on its protection during surgical dislocation of the femoral head [273] indicated that the modified Dunn procedure could be done safely and also showed its applicability to adolescent SCFE. It became evident to observers that not only severe cases but also moderate and even mild slips had a clear tendency to be associated very early on with the femoroacetabular impingement syndrome, damaging the femoral head and acetabular articular cartilage and the interposed labrum with cam-type lesions. The possibility was then raised for minimizing long-term hip joint damage by immediate or at least very early anatomic re-positioning of the femoral head-neck anatomy rather than simply pinning in situ and only responding to symptoms several months or even years later by proximal femoral osteotomies at variable levels. Early reports of reasonably large series showed excellent results with what is now referred to as the modified Dunn procedure. This refers to the repositioning of the head at open operation (open subcapital reorientation of the epiphysis or wedge osteotomy of the femoral neck) using the anterior trochanter-sparing approach where the hip joint is dislocated (surgical hip dislocation) and a soft-tissue flap is derived subperiosteally from the retinaculum and external rotators to protect the blood supply to the epiphysis. Dislocation of the hip allows for completing the detachment of the epiphysis under direct vision, complete callus resection from the femoral neck without causing tension in the

395

retinaculum, and curettage of the epiphyseal plate. Leunig et al. (2007) reported on 30 hips with moderate to severe slips with posterior displacement 30–70° [274]. Most hips were categorized as chronic/stable with six acute-on-chronic. Three hips needed revision due to screw/wire failure but femoral head necrosis did not occur. Ziebarth et al. (2009) reported on 40 moderate to severe slips reorienting the epiphysis through a surgical dislocation approach, also with no cases of osteonecrosis or chondrolysis developing [275]. The addition of the surgical dislocation approach that allowed for the development of an extended retinacular soft tissue flap to the capital realignment procedure of Dunn was supported as a meaningful way to prevent or at least minimize the effects of femoroacetabular impingement on early and mid-adult hip degeneration. Of the 40 patients, 12 were unstable. Surgical technique was outlined in great detail. The slip angles, α angles, and hip ranges of motion were all considerably improved. The authors also observed that while clinically unstable hips had physes that were grossly unstable or easily separated at surgery, half of the cases classified clinically as being stable (13 of 26 hips) had physes that were also grossly unstable or easily separated by direct inspection at time of surgery. The modified Dunn approach “allows safe reduction by removal of the posterior callus and thinning of the femoral neck”. They felt that the unstable slip would benefit most from this procedure where the rate of osteonecrosis was 10–40% with in situ pinning, in particular those with moderate to severe SCFE. Huber et al. (2010) studied an additional 30 hips treated for moderate-severe SCFE using the modified Dunn procedure with surgical hip dislocation and also reported highly favorable results [276]. Anatomic and near-anatomic reduction was achieved in all cases with only one case of osteonecrosis (1/30, 3.3%) with failure of implants and revision surgery needed in four. The slip angle was mild in 3 (less than 30°), moderate in 17, and severe in 10. Only three patients were felt to have an unstable hip. The procedure was described in detail. Slongo et al. (2010) related their experiences with the modified Dunn procedure in 23 unstable cases [277]. Two patients developed osteonecrosis (2/21, 9.5%). Preoperatively 20 were considered to be stable and 3 unstable. This study also found that six hips considered stable preoperatively were observed to be unstable at surgery. Reports from other centers continued to show good results although some complications were increasingly found. Madan et al. (2013) reported on 28 hips having the procedure with 2 developing AVN postoperatively (2/28, 7%) [278]. There was excellent correction of deformity of the severe slips with the lateral slip angle corrected a mean of 50.9°. No other major complications occurred, and the authors considered the procedure to be safe and reliable. Alves et al. used the procedure for unstable slips comparing six patients having the open surgical dislocation technique with six closed

396

reductions with percutaneous pinning [279]. AVN occurred in both groups with the surgical dislocation technique, in the opinion of the authors, not decreasing the rate of AVN (although the number of patients treated was small). The open reduction by surgical dislocation group had 4/6 (67%) cases develop AVN while 2/6 (33%) with closed reduction had AVN. Sankar et al. (2013) reviewed outcomes of the modified Dunn procedure in unstable slips in a multicenter series. Assessment involved 27 hips at a mean follow-up of 22.3 months (12–48) [280]. AVN occurred in 7/27 (26%), and implant assessment showed 4 (15%) with broken devices needing surgical revision. AVN led to a poorer clinical result. Souder et al. (2014) compared in situ fixation of stable and unstable hips (71) with a smaller group, undergoing capital realignment by the modified Dunn/surgical dislocation technique, of 17 hips, stable (10) and unstable (7) [281]. In the stable hips undergoing in situ stabilization, there were no cases of AVN even in severe cases, but in stable hips treated by the modified Dunn method, there were 2/10 (20%) developing AVN. AVN was frequent in unstable hips in both groups: 3/7 (43%) with in situ management and 2/7 (29%) with the modified Dunn method. The modified Dunn procedure in their opinion needed to be “considered with caution.” In another large study of the modified Dunn procedure for SCFE, Upasani et al. (2014) assessed a consecutive series of 43 hips from one institution including both stable and unstable cases [282]. There were 26 unstable and 17 stable hips, and, overall, slips were severe in 37/43, 86%. When all patients were grouped, 22 complications were noted in 16 patients (16/43, 37%) including such events as AVN, fixation failure with deformity progression, and postoperative hip dislocation. The group indicated a narrowing of indications for the modified Dunn procedure to acute/unstable and severe slips (>50°) that could be operated within 24 h of occurrence. At least two patients had AVN and degenerative joint disease already severe enough to need referral for total hip arthroplasty. The recent revelation that hip instability can occur post- surgery with the surgical dislocation of the hip procedure adds another factor that must be considered with use of this procedure in SCFE. Upasani et al. and an International SCFE Study Group, from centers where the procedure has been actively done, reported that in a total of 406 cases, 17 (4.2%) developed hip subluxation or dislocation at a median of 3 weeks post-surgery (1 day–2 months) [283]. The procedure was generally performed for patients with severe deformity with a slip angle greater than 40°. The complication had extremely serious implications since, in addition to the displacement needing further management, the incidence of AVN was extremely high involving 14 of the 17 and 3 had already had total hip arthroplasty. These patients, if seen in clinics favoring the in situ approach, would have been treated with in situ fixation with or without basicervical or

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

intertrochanteric osteotomy where AVN for chronic slips is rare and hip displacement never reported. AVN occurring near skeletal maturity has virtually no healing potential meaning that adolescent age patients developing AVN are sufficiently incapacitated that even total hip arthroplasty is considered or done at extremely young ages. Modified Dunn Procedure (Subcapital Correction Osteotomy) for Healed SCFE Some groups have performed the modified Dunn procedure for early FAI after the proximal femoral epiphyseal growth plate has fused. Bali et al. reported on use of the technique (including surgical dislocation of the hip) in 8 healed hips at a mean age of 17.8 years (13–29) [284]. They reported excellent results for the symptomatic FAI and noted no AVN. In another study, however, Anderson et al. treated 12 healed hips for SCFE using subcapital correction following surgical hip dislocation at an average age of 15 years with complications in 4/12, 33%, failure of fixation requiring further surgery, a lower hip score than pre-surgery, and 2 cases of AVN (2/12, 17%) [285].

3.2.11 More Detailed Review of Complications of Treatments for Slipped Capital Femoral Epiphysis Five major complications can be associated with SCFE. Complications of treatment can often be worse than the natural history of the disorder left untreated. Howorth [47, 106] and Hansson et al. [119], among others, observed that avascular necrosis does not occur in the untreated hip regardless of the degree of displacement and that hips allowed to heal with moderate and in some instances severe deformity can continue to do well into mid-adult or late adult life.

3.2.11.1 Avascular Necrosis Avascular necrosis in chronic stable hips is almost always a complication of treatment rather than a complication of the disease itself; it is rarely seen in association with these slips in continuity even if the femoral head has been completely displaced posteriorly. AVN is also rarely seen with in situ pinning in SCFE, regardless of the severity of the slip. Closed reduction of the chronic slip is never warranted because any correction gained risks excessive trauma that tears the posterior periosteum and callus and their enclosed vessels. Although there has been a definite change in the relationship of the femoral head to the femoral neck in chronic SCFE, this process has been occurring for several weeks to several months, and spontaneous attempts at stabilization posteriorly and medially by fibrous, fibrocartilaginous, or osseus tissue have led to thickening and shortening of the posterior periosteum within which the retinacular vessels are situated.

3.2 Slipped Capital Femoral Epiphysis Fig. 3.20 (a) Radiograph illustrates an example of avascular necrosis. (b) An example of chondrolysis following SCFE treatment is shown. Joint space narrowing indicates the destruction and lysis of the articular cartilage. Both of these complications will lead to early adult osteoarthritis

a

Many will also preclude use of closed reduction in an acute- on-chronic situation since it is not possible to know when only the acute component has been reduced (Fig. 3.20a). Acute slips, on the other hand, have a relatively high incidence of AVN with or without closed reduction and in situ pinning. In the acute disorder, specific steps to minimize AVN include urgent gentle reduction within 24 h of the event (either by closed manipulation under careful fluoroscopic control or at open capsulotomy – not open hip dislocation – with the surgeon manually directing the repositioning of the head), arthrotomy/capsulotomy of the hip joint (to decompress any hip joint pressure by removing joint fluid or hematoma), and in situ fixation. In a large study of 240 patients by Tokmakova et al., 21 patients (21/240, 8.8%) developed AVN, but all 21 (of 36) had presented with an unstable acute SCFE. None of the 204 patients who had a stable slip developed AVN regardless of the grade of severity [286]. Treatment in all used either a hip spica cast or stabilization with one to four pins or screws. In those with an unstable slip, AVN risk was increased with attempts at reduction, severity of the slip, and use of multiple pins. Pinning in situ without reduction with stabilization by a single cannulated screw was defined as the method of choice for treatment of a slipped capital femoral epiphysis. Avascular necrosis as a complication of SCFE began to gain wide recognition in the late 1920s. Axhausen recognized the entity referred to as “aseptic” necrosis initially to distinguish it from avascular necrosis problems known to occur after infection. Although SCFE is clearly a precursor to osteoarthritis of the hip, it is avascular necrosis that represents the most serious sequel of the disorder. Hall recognized this point specifically in his detailed report in 1957 [168]. Furthermore, avascular necrosis does not occur based on

397

b

natural history studies of the disorder, even with severe deformation, but is recognized as occurring only as a complication of active treatment. Since the AVN occurs near the end of the growth period when the femoral head has reached adult proportions, there is essentially no capacity for effective repair. The hip is generally destroyed, and management requires (relatively ineffective) repositioning osteotomies or even joint resurfacing or total hip arthroplasty in early adulthood. The collapse of the femoral head articular surface during the AVN episode limits the value of actual repair of the underlying bone tissue. AVN rarely occurs with in situ pinning, never in chronic slips, and occasionally with acute slips. The observations that moderate/severe (and even mild) slips predispose to femoroacetabular impingement (FAI) and then early to mid adult osteoarthritis has led some to treat such slips initially with repositioning surgical procedures to restore completely anatomic proximal femoral anatomy. While there is detailed recognition of the problem of AVN, it continues to occur with many interventions beyond the in situ approach. Each more complicated therapy, while improving the anatomy, predisposes to a certain amount of AVN. Gautier and Ganz detailed the vascular anatomy, the basic outlines of which had been detailed previously (see Chap. 1), and outlined the modified Dunn osteotomy for anatomic realignment based on these vascular features [272, 273]. The term “modified” refers to the use of surgical dislocation of the hip with the technique designed to better preserve the vascularity than in previous approaches. Efforts have also been described to detect postoperative ischemia by MRI which can lead to immediate cast repositioning with closed or open reduction cases of DDH but has little capability form effective response after capital realignment/pinning repositioning procedures for SCFE.

398

Assessment of Femoral Head Vascularity Intraoperatively Efforts have been made to assess femoral head vascularity intraoperatively. Two methods are used as outlined by Ganz et al. [273] in 2001 and Slongo et al. [277] in 2010. These involve (i) intraoperative drilling of a 1.5–2 mm hole through the femoral head cartilage into the underlying bone marrow and observing for active bleeding and (ii) assessing perfusion of the femoral head using laser Doppler flowmetry. Novais et al. applied these methods in the course of modified Dunn procedures for unstable SCFE where 7/27, 26%, developed AVN [287]. They accumulated much information but reached no definitive conclusions as to effectiveness. Hall reviewed 173 hips with SCFE noting 27 cases of AVN (15.6%) and 3 of chondrolysis although the incidence of AVN in 42 cases of cervical osteotomy was 38% [168]. The complication is recognized as occurring due to damage to the blood vessels that are concentrated on the superolateral and posterior aspects of the neck. Since displacement occurs in a posterior and medial direction, there is a tendency for the vessels to the head to be stretched. Spontaneous episodes of AVN in the absence of therapy, however, are virtually unknown. The vasculature while slightly at risk appears to stretch gradually in relation to the slowly evolving chronic slip. When treatment is done too vigorously, however, damage and subsequent AVN follow. The factors relevant in consideration of AVN involved delay in diagnosis, the amount of displacement at time of diagnosis, and the type of treatment. Factors predisposing to AVN include moderate to severe displacement requiring relatively more extensive efforts at improving position, relatively longer time prior to a diagnosis, and treatments which are characterized by manipulation and efforts at reduction be they closed or open. Owing to the age of the patient and the relatively small time of growth remaining for remodeling avascular necrosis in SCFE is a major problem and almost invariably leads to osteoarthritis sometimes in young adult life. AVN is always associated with development of deformation of the femoral head and articular surface. Avascular necrosis can also occur following a pinning of the slip. The problem relates to pins that are placed in the anterosuperior and posterosuperior quadrants in the area where the lateral epiphyseal vessels enter the femoral head. While there can be some damage with a pin remaining totally within the head and neck bone, major problems could occur if the pin exits the neck in superolaterally or posteriorly and then reenters the head thus causing extensive damage to the vascular leash which lies on the surface of the neck prior to entering the head between the physeal and articular cartilage regions. This region is very difficult to assess radiographically and thus the recommendation stressed recently by Aronson and Loder is to place internal fixation strictly in the central or neutral zones of the femoral head and neck to

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

inimize chances of injuring the blood supply [172]. More m limited damage can also occur if vessels traverse the neck in the anteroinferior quadrant where the inferior metaphyseal vessels enter the head and neck and supply some regions. Lowe summarized both his own large experience with early AVN and chondrolysis and other previous assessments [288]. In a study of 100 cases of SCFE, he noted 21 hips that developed necrosis of the bony epiphysis or chondrolysis following treatment. There were 6 cases of AVN and 15 of chondrolysis. Although the feeling was widespread that AVN was strictly a complication of treatment, there were some instances, for example, reports by Moore, that epiphyseal necrosis could occur naturally even in cases where there was minimal displacement. The AVN could be variable and in some instances recovery was possible. All hips had had major displacement and had been treated. The incidence of AVN was much higher after “successful” reduction that implied movement of the head-neck junction with its implication of tearing of the associated vessels.

3.2.11.2 Chondrolysis Chondrolysis refers to a destruction of the articular cartilage of the femoral head that is demonstrated radiologically with joint space narrowing and clinically with discomfort, decreased range of motion, and on occasion actual fusion of the joint [67, 289] (Fig. 3.20b). It can also involve destruction of acetabular articular cartilage. Waldenstrom pointed to chondrolysis as a complication of SCFE in 1931 [62]. He reported on three cases of necrosis of the joint cartilage after a slipped epiphysis. All were chronic having showed symptoms from 6 months to 1 year and were subsequently treated with closed reduction under anesthesia followed by hip spica immobilization with the involved lower extremity in abduction and internal rotation. Mobilization was continued for 2 months prior to cast removal and rehabilitation. After some months the joints became more and more stiff and finally all mobility ceased. The earliest finding radiographically after a few months was thinning of the joint cartilage. The joint continued with increased stiffness and often a formal ankylosis. Waldenstrom clearly differentiated the disorder from avascular necrosis of the bone that involved variable parts of the femoral head bone but never the joint surface alone. He felt that cartilage necrosis was not due to avascularity of the bone especially since the cartilage on the acetabulum was also affected. He felt that the necrosis in his described cases involved only the joint cartilage and that damage to the capsule and synovium was the cause of the joint surface necrosis since that was the source of their nutrition. Since all his cases occurred in patients who had been reduced which meant manual reduction under anesthesia in a fairly forceful fashion by then current standards, he felt it was the tearing of the capsule that induced fibrosis and subsequent poor nutrition. He felt that it was the reduction of the epiphyseal slip by

3.2 Slipped Capital Femoral Epiphysis

relatively vigorous treatments that caused the cartilage necrosis. He then treated 24 subsequent cases without vigorous reduction but rather by slow traction with the patient in bed. Chondrolysis can occur in patients with SCFE who have not undergone treatment. The large majority of instances, however, are associated with treatment with most of those following internal fixation in which the fixation pin is either into or through the articular cartilage leading to mechanical destruction with movement. The chondrolysis entity, however, has also been reported after immobilization in a hip spica cast and after intertrochanteric osteotomy. While early papers documented a higher incidence in black patients, more recent and detailed studies have not shown any predisposition in blacks to the disorder. The first report by Lowe was unusual in that more cases of chondrolysis in his series were present than of AVN [288]. The result with chondrolysis was almost invariably poor owing to stiffness of the hip and malposition. Diminution of the joint space on radiographs remained the primary radiologic sign and clinically there was discomfort and usually markedly diminished ranges of motion. The highest positive associations were with immobilization in plaster spica or prolonged traction greater than 7 weeks. In a second report of six cases of chondrolysis, he observed that some potential for repair existed from cells deep in the cartilage adjacent to the tidemark [290]. Multiple types of treatment were associated with the finding including closed reduction, immobilization by hip spica, traction, pinning, and osteotomy. The prolonged immobilization was also a feature in other series notably those of Moore [160], Jerre [66], Hall [168], and Waldenstrom [62]. Although chondrolysis and AVN can coexist on occasion, in general they are separate entities. Heppenstall et al. noted a 26% incidence of chondrolysis in a series of 65 patients (17 involved), but only 3 of 21 hips with chondrolysis had associated AVN [291]. Ingram et al. reviewed the literature on chondrolysis from Waldenstrom’s initial report of 3 cases in 1930 to their own review of 79 cases in 329 hips with SCFE in 1981 [67]. Many series were showing very high levels of chondrolysis the greatest being a 55% incidence in 116 patients by Orofino et al. [206] with other high incidences being Boyd et al. (16%) [292], Howorth (41%) [see 67], Maurer and Larsen (28%) [293], Tillema and Golding (40%) [294], Hartman and Gates (16%) [295], Heppenstall et al. (26%) [291], Gage et al. (38%) [228], and Ingram et al. (24%) [67]. When all series, including their own, were averaged, the rate of chondrolysis was 19% (332 of 1746). A review by Lubicky in 1996 however reported a much lower global influence of 7% with all forms of treatment [289]. It can occur after any type of treatment, whether operative or non-operative. More recently, Pinheiro (2011) reported on 106 hips treated non-operatively by plaster casts (69% hip spicas and 31% bilateral short leg/long leg casts in abduction

399

and internal rotation with stabilizing bars) in which the incidence of chondrolysis was 12/106, 11.3% [296]. Treatment for chondrolysis has been non-specific and symptomatic involving bed rest, crutches, traction, and various physical therapy modalities. Anti-inflammatory drugs were helpful. The feeling was prevalent that there was a predisposition to the disorder that involved some immunologic processes, never clearly defined, which led to a synovitis and subsequent degenerative cartilage changes. There was increasing evidence that the disorder was higher with certain types of therapy in particular when the joint was penetrated by internal fixation devices (51%), after open reduction (55%), after cervical osteotomy (37%), after trochanteric osteotomy (59%), and with increased immobilization in cast posttreatment. The disorder was felt to be less frequent with mild slips and acute slips reflecting the more benign treatment methods used and the lack of complications associated with them. Most chondrolysis complications are still related to treatment rather than to an inherent predisposition of the patient to the disorder. Ingram et al. reported in detail on radiologic and biopsy assessments of the hip joint from 16 previous cases from the literature plus 23 of their own [67]. The radiographic changes involved progressive joint space narrowing, usually superiorly, which often proceeded to entire concentric diminution. There tended to be a generalized demineralization of adjacent femoral and acetabular bone. The situation frequently resolved and stabilized but on occasion proceeded within a few years to a formal osteoarthritis with marked joint space narrowing, osteophytes, subchondral cysts, and subchondral sclerosis. Persistence of synovitis greater than 6 weeks post-surgery should lead to concern about a developing chondrolysis [297]. In one of the larger series, 41 cases of chondrolysis were reported by Lance et al. [65]. A wide variety of treatments had been used to treat the primary slip including plaster casts alone 7, closed reduction and plaster 3, intertrochanteric osteotomies 10, open reduction 6, wedge osteotomies of the neck 5, transphyseal bone graft 3, extra articular epiphyseal arrest 4, and pedicle graft 1. Two patients had not been treated but still developed the disorder. One of the major principles of treatment for chondrolysis is to diminish weight bearing using either crutches, or if possible, bed rest with traction. It is important to keep the hip region extended and positioned along the neutral axis to prevent abduction, adduction, flexion, or rotation deformities from developing. Gentle range of motion exercises and occasional use of anti-inflammatory drugs are usually helpful. Repair tends to be very slow, and symptoms can last several months to years before full or at least clinically useful motion is regained. Many patients however do recover to a functional hip. In those cases not healing well, there is radiographically evident progressive diminution of joint space, osteophyte formation, and a triangular appearance of the head in association with adjacent bone and

400

c artilage collapse. The authors felt that the best treatment was suspension-traction to enhance articular motion and joint lubrication. In those that went on to further joint destruction, a hip arthroplasty was often needed preceded in many instances by osteotomies. The authors felt that only 44% of those affected obtained a good or excellent result with absence of pain, a normal gait, and reasonably good motion.

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

most patients with minimal to moderate slip who had the transphyseal bone-pegging procedure almost always had shortening limited to 1/4 to ½ inch, and often less [47]. The fact that most patients either have only mild to moderate slippage or are anatomically corrected if they have greater displacement, minimizes the extent of shortening. A contralateral distal femoral epiphyseal arrest is needed on occasion.

3.2.11.5 Adult Osteoarthritis Some patients with slipped capital femoral epiphysis will develop osteoarthritis of the hip in middle to late adult years. 1. Mechanical. Chondrolysis is often associated with pin Those who suffer either chondrolysis or avascular necrosis, penetration into the joint. With resumption of walking, however, in the course of treatment will generally develop the pin tip scarifies the adjacent articular cartilage. moderate to severe arthritic symptoms much earlier either in 2. Nutritional. Inability to receive synovial nutrition was late adolescence or early adult life. The importance of minioften a cause in particular in those patients immobilized mizing and if possible completely eliminating the osteoarin cast for several months. thritis associated with SCFE is evident. What remains 3. Ischemia. Ischemia of the bone did not necessarily lead to uncertain is the amount of deformation of the head-neck avascular necrosis alone but also to chondrolysis on occa- region that contributes to eventual adult degenerative change. sion. It was noted that 33% of chondrolysis cases occurred The longer-range studies over several decades indicated that after osteotomy of the neck. mild slippage, for the most part, is not causative in this regard 4. Intra-articular pressure. It was felt that osteotomies of the and that even moderate slippage left untreated other than by neck and intertrochanteric region often tightened the hip fixation in situ has a low incidence of osteoarthritic change joint capsule in the process of the valgus, flexion, and well into the fourth and fifth decades. Most but not all agree internal rotation repositioning which increased the intra- that with severe and complete slips corrective osteotomy or articular pressure and secondarily limited the ability of even primary open reduction and internal fixation are warnutrients to diffuse into the cartilage. ranted because of gait and functional considerations. Some will stabilize any slip, regardless of degree, in situ and perform osteotomy later only for troublesome symptoms of 3.2.11.3 Stiffness Hip stiffness occurs only as a complication of AVN or chon- pain, limited hip flexion, and internal rotation and the drolysis. External rotation left untreated diminishes the func- Trendelenburg gait abnormality. A matter of further importional range of motion although the actual arc of motion is tance, however, is whether correction must be done at the head-neck junction for optimal long-term results or whether numerically unchanged. the compensatory osteotomies at the basicervical, intertrochanteric, and even subtrochanteric regions provide suffi3.2.11.4 Shortening on the Involved Side Shortening on the involved side must be assessed in a patient cient correction to render treatment at those sites preferable. Krahn et al. reviewed 36 patients out of 264 with SCFE with SCFE. If the disorder turns out to be bilateral, clinically significant limb length discrepancy rarely occurs. Even in who had developed AVN [298]. Twenty-four hips were those patients who have unilateral involvement, there is infre- assessed at an average follow-up of 31 years. AVN indeed was quent clinically significant limb length discrepancy. Any a complication with many negative long-term sequelae. Nine shortening is due to a combination of factors including treat- of 22 patients had already undergone reconstructive surgery, 4 ment that induces a premature fusion of the proximal femoral during adolescence and 5 during adulthood. In the remaining capital growth plate, removal of bone in association with 13 patients (15 hips), there had not yet been any operative compensatory osteotomy, and loss of length caused by the intervention, but all showed degenerative changes on radiograslippage. Lower extremity length discrepancy, however, is phy. This report did not give the treatment methods in the rarely a problem because of the age at which the disorder patients without the AVN complication, but certain characteroccurs and the fact that most patients have only mild to mod- istics were seen in those who had the AVN complication. In erate deformation. Since only 30% of the growth of the femur the four hips that had early deformation requiring early reconand 15% of the growth of the lower extremity occurs at the structive surgery, two had closed reduction with pinning and proximal femoral capital epiphysis and since most patients two had open reduction in one of whom c uneiform osteotomy are 11 years of age or greater at the time of the disorder, there was also done. In five patients with progressive changes is rarely sufficient growth remaining to account for a discrep- requiring surgery in adulthood, two had closed reduction with ancy of greater than ½–3/4 of an inch. Howorth noted that pinning, two had closed reduction with casting, and one also Pathogenesis of Chondrolysis There are four contributing factors leading to the disorder:

3.2 Slipped Capital Femoral Epiphysis

had cuneiform osteotomy. In those patients with progressive degenerative changes which had not at that time required surgery, cuneiform osteotomy had been performed in 5 of the 13, closed reduction and pinning in 4, and open reduction in 1. In the series, therefore, most patients had either closed or open reduction, and many had cuneiform osteotomy as well. These three techniques in particular have almost always been implicated in the patient with AVN. Larson et al. studied hip arthroplasties done in patients originally having a SCFE from a hip registry database at their institution [299]. In the 38 hips, 28 had a total hip arthroplasty, 8 hip resurfacing, and 2 hemiarthroplasty. The specific diagnoses included 25 hips with AVN or chondrolysis and 13 with degenerative changes or femoroacetabular impingement. When listed, slip displacement was severe in 20, moderate in 4, and mild 7. Avascular necrosis was associated with severe slips (14/20, 70%) and an acute or acute-on-chronic diagnosis. The mean time from slip to arthroplasty was only 7.4 years in those with AVN or chondrolysis (age at surgery 20 years) and 23.6 years in those with degenerative change (age at surgery 38 years).

3.2.12 Long-Term Follow-Up Studies Boyer et al. performed a detailed long-term study of 149 hips with SCFE assessed 21–47 years post-diagnosis at a mean of 31 years [209]. They assessed treatment methods performed between 1915 and 1952 when the surgical interventions were relatively crude by current standards. They confirmed the previous belief that “the mild slip has an excellent prognosis when pinned in situ and if no realignment procedure is attempted.” This report, written in 1981, appeared before the work of Walters and Simon [183] that indicated that some cases of pinning were problematic even if not so recognized by the surgeon. In other words, the long-term results of mild slips pinned in situ without pin penetration would remain excellent, perhaps even somewhat better than the assessment in the Boyer paper. Their conclusion remains intact today, namely, that pinning in situ is safer than any type of reduction or realignment since it presents fewer technical problems and also requires a minimum of immobilization. Their work also concluded that malposition of the moderate slip should be accepted and pinned in situ. They noted a considerable number of technical complications with the realignment operative procedures then in use. They concluded also that pinning in situ was a safe and reliable method of treating the moderately slipped capital femoral epiphysis. They recognized that severe slips would benefit from realignment procedures although results in their series were not good, again because of technical surgical problems. They had a high incidence of chondrolysis or AVN. They were also able to study seven patients with severe uncorrected slips and noted the long-term results to be “remarkable,” meaning remarkably good. Six of the seven had good clinical results,

401

marred by some limping and diminution of abduction and internal rotation but overall showing good, painless hip function. The seventh patient had a poor result. They pointed to the concern about AVN and chondrolysis after a closed manipulation of chronic SCFE and AVN after femoral neck osteotomy. In assessing their cases, they used the Southwick method of measurement in some but where they only had radiographs of the affected hip a more general measurement was used defining a mild slip as one with displacement of the head on the neck of less than 1/3 of the diameter of the femoral neck; moderate, displacement of 1/3 to ½ the diameter; and severe, displacement of more than ½ the diameter. Carney et al. continued a long-term follow-up study of SCFE from the Iowa group in 1991 [210]. In this study, 155 hips were assessed at a mean follow-up of 41 years after onset of symptoms. Of these, 42% were mild, 32% moderate, and 26% severe. With chronic slips, symptomatic treatment only was used in 25%, a spica cast in 30%, pinning in 24%, and osteotomy in 20%. The study indicated that degenerative joint disease, classified on hip radiographs, worsened with increasing severity of the slip but also when reduction or realignment had been done. AVN 12% and chondrolysis 16% were also common with increasing severity of the slip and when reduction or realignment had been performed; their presence almost always indicated a poor result. Deterioration with time was noted to be more frequent with increasing severity of the slip. Their study confirmed that the natural history of a malunited slip was mild deterioration, with poorer results related to increasing severity of the slip and presence of complications. Realignment procedures were felt to be associated with a risk of appreciable complications and thus adversely affected the natural history of the disease. Their conclusion was that “pinning in situ provided the best long-term function and delay of degenerative arthritis with low risk of complications.” Their findings continued to support those of Boyer et al. even though this series was reviewed 12–15 years after the former. A long-term study by Ordeberg et al. (1984) assessed 49 cases of SCFE without primary treatment 20–60 years after diagnosis [300]. Assessment was by questionnaire in all and clinical and radiographic examination in 44 of the 49. Patients seen originally between 1910 and 1960 were assessed, involving 57 hips with a mean observation time of 37 years (20– 60 years). As a general conclusion they indicated that “2 of these 49 (still living) cases have required surgery because of secondary arthrosis, far fewer than were found in a comparable group of cases treated with closed reduction and hip spica.” A positive Trendelenburg test was noted in approximately 1/3 of the patients, and even fewer noted some limping after walking considerable distances. Limb length discrepancy of 2–5 cm was noted in 15. The index of pain correlated with the degree of displacement, and marked functional restrictions were noted mainly in cases with severe slip. Deterioration in the later years of life was “comparatively slight and can in part be explained by

402

age.” They concluded that clinical observations regarding pain, walking capacity, and range of motion showed much better function than expected. The cases with severe clinical problems were almost invariably among those with severe slipping, but some even in this group were doing well. Studies in the patients at a 35-year interval did not support the expected finding of deterioration with time. They concluded: “with few exceptions, coxarthrosis developed only in hips with severe displacement.” Studies reported by Hagglund et al. [301] and Hansson et al. [119] in a Southern Sweden SCFE population noted that in 57 cases with no primary treatment, no AVN (segmental collapse) or chondrolysis was seen. At long-term follow-up, 12/53 had “severe arthrosis,” but only 1 had hip replacement surgery, and clinically “most patients had a good hip function with at least tolerable pain and a good walking capacity.” Symptomatic treatment or pinning in situ resulted in high clinical ratings with only 2% of hips needing a secondary reconstruction procedure. When closed reduction and spica casting had been used, the combined rate of AVN and chondrolysis was 13% with reconstructive procedures sought in 35% of the hips. Cervical osteotomy had a combined rate of AVN and chondrolysis of 30% with reconstructive procedures needed in 15% [302]. The group recommended pinning in situ for treatment since either open or closed reduction clearly increased problems. In their series, of 39 hips in which the slips had been reduced, osteonecrosis developed in 12 (31%) and chondrolysis in 11 (28%). In 116 hips that had not been reduced, AVN occurred in 7 (6%) and chondrolysis in 14 (12%). They concluded that the natural history of the malunited slip was one of mild deterioration related both to the severity of the slip and complications of treatment. Realignment, however, risked “substantial complications and adversely affects the natural course of the disease.” They supported pinning in situ regardless of the severity of the slip as providing the best long-term function associated with a low risk of complications and thus the most effective method of delaying the development of degenerative arthritis. Hagglund et al. studied long-term results after nailing in 204 slipped epiphyses evaluated at an average of 28 years post procedure [301]. The only early complication noted was segmental collapse of the femoral head seen in 4 of 179 hips nailed in situ and 4 of 25 hips operated after reduction. Subsequent osteoarthritis was twice as frequent after reduction was part of the treatment than after fixation in situ. They concluded: “nailing or pinning in situ is the method of choice when possible, regardless of the degree of slipping.” Prophylactic pinning of the contralateral hip was indicated because of the high incidence of bilaterality. The second long-term study by this group of femoral neck osteotomy assessed 33 patients with a severe slipped capital femoral epiphysis treated primarily with wedge osteotomy of the femoral neck [302]. The patients were assessed at an average of 28 years post-surgery. Segmental collapse and or chondrolysis developed in 10 of the 33 patients. In nine of these available for reassessment, all

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

had severe arthritis with poor function. Their conclusion was that value of realignment by wedge osteotomy of the femoral neck was questionable. These long-term studies are essential to help determine which adolescent deformities today would best be treated by compensatory osteotomy or open reduction with cuneiform osteotomy, using the modified Dunn procedure with surgical dislocation, and which can be treated effectively by pinning in situ alone. The long-term studies referred to in this section, which almost unanimously support in situ fixation as the primary treatment, cannot be disregarded. They appear to be extremely well done. One criticism of the findings mentioned by some is that the criteria for total joint arthroplasty have changed since these patients were treated (with in situ fixation alone) with more being done now due to improved techniques and a population more desirous of remaining physically active at older ages, stressing their hips more and being more willing to undergo replacement surgery. In addition to this observation, there is the belief that realignment procedures have also improved compared to several decades ago and the concept of femoroacetabular impingement in SCFE has pinpointed and quantified the specific findings predisposing to early hip osteoarthritis. The use of aggressive interventional realignment based on the principles of FAI and designed to immediately restore anatomic femoral head-neck structure has shown excellent early results; however, the complications of AVN, to a lesser extent chondrolysis, and now post-procedure hip instability when they occur damage the hip so extensively that degenerative hip disease leading to early arthroplasty occurs in late adolescence or early adulthood several years to a few decades before the natural history of osteoarthritis developing following uncomplicated in situ fixation. The next decade will help determine not so much whether one or the other approach will be recommended for all slips but, rather, which methods are best in specific cases. Coxa Vara Due to Other Acquired Causes Some entities are discussed in their more specific sections in this volume in Chap. 1, Developmental Dysplasia of the Hip, and Chap. 2, Legg-Calvé -Perthes Disease, and in volume I in Chap. 3, Skeletal Dysplasias, and Chap. 4, Bone and Joint Deformity in Metabolic, Inflammatory, Neoplastic, Infectious, and Hematologic Disorders.

3.3

Developmental Abnormalities of the Femur: Including Proximal Femoral Focal Deficiency (PFFD), Congenital Short Femur, and Infantile Coxa Vara

Developmental abnormalities of the femur comprise an extremely wide spectrum of disorders from complete absence of the femur to its presence with normal structure and only

3.3 Developmental Abnormalities of the Femur: Including Proximal Femoral Focal Deficiency (PFFD) , Congenital Short Femur…

mild shortening. Many classifications have been described, but the disorders are so variable that no single universal approach has been adopted. The overall pattern of categorization, however, is reasonably defined although overlap in categories does exist.

3.3.1 Terminology One approach is to divide the structural developmental abnormalities of the femur into two groups: (1) congenital/developmental femoral abnormalities, developing in utero and present at birth, and (2) infantile coxa vara, generally considered not to be present at birth and thus to be either an acquired deformity or, if developmental, one worsened by postnatal mechanical effects. Disorders recognizable for the congenital/developmental categorizations include proximal femoral focal deficiencies (PFFD) (most of which have a coxa vara component), congenital short femur (some of which have a coxa vara component), and distal femoral abnormalities. These groupings concentrate on the primary sites of abnormality within the femur and direct considerations to specific treatment approaches. Infantile coxa vara has been defined as a reasonably distinct entity over the past few decades. Part of the confusion and overlap in terminology comes from the fact that these categorizations can occur as isolated deformities in some, while in others two or even three of the deformities are present in the same femur. Some studies have concentrated

403

primarily on describing the proximal femoral abnormalities many of which have a coxa vara component; some have focused on coxa vara itself which can be a feature of both the severe and the milder deformities but also exists in isolated fashion as infantile coxa vara and in systemic disorders like skeletal dysplasias and osteogenesis imperfecta; and some have considered congenital short femur as a specific entity although many femurs with this abnormality also have a mild coxa vara. Distal femoral developmental abnormalities are described infrequently but do occur and can lead to clinically symptomatic deformities.

3.3.2 Proximal Femoral Focal Deficiency Virtually all of the severe developmental abnormalities of the femur are concentrated at its proximal end [303–310]. It took several decades before reasonably accurate classifications of these variable deformities were outlined. These truly congenital disorders were often discussed with and confused with infantile coxa vara that is now recognized as an isolated disorder of postnatal onset in the large majority of cases. Drehmann [2] (Fig. 3.21) and Golding [311] showed examples of coxa vara that included but did not differentiate congenital and infantile varieties. Although the overall entity is now known as proximal femoral focal deficiency (PFFD), the several variants were often discussed several decades ago under the term congenital coxa

Fig. 3.21 Illustrations from the work of Drehmann [2] show good awareness of the underlying pathoanatomy of proximal femoral focal deficiency

404

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

vara [2, 3, 143]. The deformities are quite variable in extent, and several newer classifications have been presented to better understand the entity and categorize it for treatment. Several classifications are presented below since each provides information on the types of deformity that have been observed. Some are strictly pathoanatomic while others categorize disorders needing varying treatment approaches.

Class C. The acetabulum is severely dysplastic, and there is neither a bone nor cartilage model of the femoral head. The shaft of the femur is short with an ossified tuft at the proximal end. Class D. Both the acetabulum and the femoral head are completely absent, there is a deformed shortened femoral shaft, and there is no proximal tufting of the shaft of the femur. This class of deformity is frequently bilateral.

3.3.2.1 Classifications of PFFD (Many of Which Include Congenital Short Femur and Distal Femoral Abnormalities)

Amstutz Classification Amstutz defined proximal femoral focal deficiency as “the absence of some quality or characteristic of completeness of the proximal femur, including stunting or shortening of the entire femur” [304]. In his study, and that of Aitken, a portion of the distal femur was always present, even if only represented by a misshapen ossicle. Amstutz also brought attention to the fact that a coxa vara deformity in addition to shortening was characteristic of many cases of proximal femoral focal deficiency. His classification defined five morphologic groups identifiable radiographically at birth and also included developmental changes with time. Congenital bowed femur with coxa vara, which had not previously been included with PFFD entities, was represented (Fig, 3.22a, b). Type 1. Congenital bowed femur with coxa vara. The anterolateral bowing of the femoral shaft, most apparent in the proximal half, is associated with medial femoral cortical sclerosis. The capital femoral epiphysis ossifies and, since it

Aitken Classification Aitken divided the entity into four classes, A through D [303]. Class A. The head of the femur is present along with an adequate acetabulum and a very short femoral segment. Initially there is no bony connection noted between the femoral segment and the head of the femur, but at skeletal maturity, bone continuity is seen although in most instances there is a subtrochanteric pseudarthrosis. Class B. The femoral head is present and the acetabulum is adequate, but the femoral shaft is short and deformed with a small bony tuft on its proximal end and no bone or cartilage continuity between the shaft and head and neck segment at any time.

a

PFFD TYPE I

TYPE II

TYPE III

Fig. 3.22 (a) The basic proximal femoral focal deficiency classification of Amstutz [304] is shown. This continues to provide an excellent overview of the variable defects seen. There are five categories. (b(i–v)) Amstutz further illustrated the variability of changes and compared their radiographic appearances within the first year of life

TYPE IV

TYPE V

and adolescence in a second series of images. The five basic types are again shown, but the variability is further demonstrated for each type. In paired drawings the radiographic appearances in the first year of life are shown at left and the appearances of the same hip in adolescence at the right

3.3 Developmental Abnormalities of the Femur: Including Proximal Femoral Focal Deficiency (PFFD) , Congenital Short Femur…

bi

405

bii

TYPE I EARLY

TYPE IA LATE

TYPE II EARLY

TYPE IB LATE

TYPE II LATE

biii

TYPE III EARLY TYPE IIIA LATE

TYPE IIIB LATE

biv

TYPE IIIC LATE

TYPE IIID LATE

bv

TYPE IV EARLY

TYPE IV LATE

TYPE V EARLY

TYPE V LATE

Fig. 3.22 (continued)

is well positioned in the acetabulum, is not associated with acetabular dysplasia. There may be a delay in appearance of the secondary ossification center. Type 2. There is a subtrochanteric pseudarthrosis with lack of bone continuity between the head-neck-trochanteric region and the rest of the shaft. This is characterized c linically

by a progressive varus of the proximal femur and delayed development of the head. Two possible patterns of development follow; in one there is progressive varus and lack of union of the two fragments, and in the other there is either complete bone repair or a rigid pseudarthrosis with close apposition of the bone fragments.

406

Type 3. The hip region is formed with a cartilaginous femoral head present and the acetabulum shows no evidence of dysplasia. There is no initial radiologically defined bone continuity between the head, neck, and trochanteric region of the femur and the shaft. The shaft is shortened and has a variable proximal bulbous shape. Ossification of the femoral capital epiphysis is often markedly delayed. Subsequent development can be variable leading to four subgroups: type 3A, union between the proximal and distal fragments with a coxa vara persisting; type 3B, union between the proximal and distal fragments, but the coxa vara is much more marked, and the greater trochanteric epiphysis greatly overgrows the superior surface of the head; type 3C, there is only a fibrocartilaginous union between the two fragments with marked proximal displacement of the distal fragment; and type 3D, there is no continuity present. Type 4. The hip joint components are formed with an acetabulum and capital femoral epiphysis present in all so that acetabular dysplasia is not seen or is only minimal although ossification of the capital femoral epiphysis may occur as late as 2½ years of age. The proximal end of the distal femoral shaft tapers sharply, almost to a point, which differentiates it from type 3 that has a bulbous proximal end. The tapering represents an unfavorable prognostic sign as union of the two fragments never occurs and is followed by proximal migration of the distal fragment. The acetabulum eventually becomes dysplastic because of this failure of union. Type 5. Dysgenesis is severe, and none of the normal precursor hip components involving either the capital femoral epiphysis or acetabulum develop. The classification of Amstutz has been widely used. Panting and Williams reviewed their cases in relation to his approach [312]. They also pointed out that radiographic evidence of an acetabulum in the first year of life indicated the presence of a well located femoral head and neck even if they were not seen owing to delayed ossification of the secondary center. Additional Classifications of Proximal Femoral Focal Deficiency Over the course of the next several years, the PFFD syndrome became better recognized, and treatment was intensive. Efforts were made to refine the classifications of Aitken and Amstutz to define disorders into a more practical and clinically oriented framework in relation to specific diagnosis, treatment approaches, and projected outcomes. These multiple approaches also indicate that virtually no two cases are identical; the disorder represents a spectrum in terms of involvement, and, with time, the radiographic appearance changes in ways that also differ depending on whether there is worsening or improvement which can be spontaneous or based on treatment.

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

Fixsen and Lloyd-Roberts Radiologic criteria for assessing proximal femoral dysplasia were described by Fixsen and Lloyd-Roberts [305]. These criteria helped to determine whether a case of proximal femoral focal dysplasia was stabilizing or worsening with growth. In many children with this disorder, radiographs showed a short femur with a seemingly absent proximal third of the femoral shaft, head, neck, and greater trochanter. Clinical findings, however, often demonstrated a stable hip in association with the shortening and the proximal femur flexed and externally rotated. The stability implied continuity between the femoral head and the proximal end of the shortened femoral shaft where the intervening radiolucent area would be occupied by a cartilage model in which ossification was delayed. Two possible outcomes were seen clinically. In the favorable state, the cartilage model of the proximal femur gradually ossified with maintenance of stability, and eventually an entire femur was seen although the shaft remained short. The unfavorable outcome was associated with the development of one or more pseudarthroses at the osteocartilaginous junction or within the cartilaginous model so the continuity between the hip and femoral shaft was lost. The resulting instability lead to proximal migration of the femoral shaft in relation to the head and neck since a breakdown of the pseudarthrosis predisposed to further upward displacement of the femoral shaft although the proximal segment remained in the acetabulum. Their criteria for distinguishing those that would proceed to spontaneous healing from those that would not were based on a study of 30 hips in 25 patients. All were observed until final definition of a stable or unstable state had been established. (i) In the stable hips, the shaft length was greater than one-half of the normal side in five of six cases, whereas the shaft length was less than one-half the normal side in six stable and eight unstable cases. When the acetabulum resembled the normal, the femoral head was always present although ossification might be delayed. If there was no acetabulum, the head was absent. If there was acetabular dysplasia, then the head might dislocate with time. In general, the shorter the ossified part of the shaft, the less was the likelihood of spontaneous healing. The distance of the proximal end of the ossified shaft from the acetabulum was an important factor. When the distance was greater than the normal side, the hip was ultimately stable in ten and unstable in five, and when the distance was equal to or less than the normal side, the hip was stable in one and unstable in four. (ii) All unstable hips showed progressive migration of the femoral shaft upward which indicated an impending dissolution of the pseudarthrosis. Where the proximal end of the ossified shaft was bulbous, all 12 hips were stable; where there was a tuft or cap, only 3 were stable with 10 unstable, and when there was a tapered point, none were stable, and 5 were unstable. (iii) The proximal end of the ossified shaft was

3.3 Developmental Abnormalities of the Femur: Including Proximal Femoral Focal Deficiency (PFFD) , Congenital Short Femur…

either blunt and irregular or pointed. Sclerosis of the proximal shaft was related either to the site of angulation or to a pseudarthrosis. In those hips that became stable, the sclerosis was essentially in the mid-shaft region, well below the proximal end and associated with angulation. In the unstable defects, the sclerosis was almost always immediately distal to the site of the pseudarthrosis or angulation, had the appearance of an inverted “V,” and was more proximal. In a retrospective study, several of the classifications in this section were reviewed, and it was the feeling of Sanpera et al. [313] that the radiologic parameters described by Fixsen and Lloyd-Roberts were the most reliable factors for predicting future outcome of the femur from the time of birth onward. Lange, Schoenecker, and Baker Lange et al. classified their 42 patients into four categories [314]. They often had difficulty assigning some patients to the specified Aitken or Amstutz group and formulated their categorization to conform to treatment approaches. In the less severely involved cases, the proximal shaft signs and stable-unstable concepts of Fixsen and Lloyd-Roberts were incorporated. Class 1. There is a coxa vara with the apex of angulation at the subtrochanteric region and a mild to moderate shortening of the femur associated with anterolateral bowing. Class 2. The acetabulum is present on the earliest radiographs but there is delayed (8–18 months) ossification of the femoral head. The proximal femoral shaft is displaced laterally but the femoral head, neck, and diaphysis are united by continuous cartilage tissue. This bridge tissue will eventually ossify, although a pseudarthrosis might be present. The continuity is evident by stability with passive motion on clinical examination. Class 3. The acetabulum is intact, but there is also late ossification (12–18 months) of the femoral head. The femoral shaft is separated from the head radiographically, but neither bone nor cartilage bridges are seen either initially or with time, and there is detectable instability with independent motion between the shaft and the head. Class 4. There is severe proximal bone and cartilage deficiency with absence of the acetabulum, femoral head, and most of the shaft. At the distal end (knee), there is often only a small segment of the bone. On occasion, this bone fragment may be completely missing, or the distal femoral bone is fused to the proximal tibial epiphyseal secondary ossification center forming one tissue mass without a knee joint. Gillespie and Torode Gillespie and Torode divided patients into two groups that could be differentiated on clinical grounds and led to markedly different treatment options [307]. Group 1. The congenital hypoplastic femur has sufficient development that the hip and knee could be made functional

407

and lower extremity length equalization in many would be possible. Group 2. There is a proximal femoral focal deficiency where the hip joint is never normal and the knee joint always nonfunctional. The lower extremity length discrepancies in group 1 are rarely as great as those in the more severely involved group 2. In addition, the flexion and abduction deformities at the hip in group 1 are less marked. In group 1, the radiographic characteristics show the femur to be 40–60% of normal length with proximal to distal continuity, coxa vara usually in the subtrochanteric region, lateral bowing of the shaft, and a hypoplastic knee; in group 2, the femur is markedly shortened and a deficiency in bone always noted. The head and neck are often absent, the shaft markedly deficient, and the knee hypoplastic. Kalamchi, Cowell, and Kim Kalamchi et al. outlined five types [308]. Type 1. There is congenital shortening of the femur with a normal hip joint and no other femoral defects. Type 2. There is a congenital short femur and coxa vara, generally with bowing and medial proximal femoral shaft sclerosis. The acetabulum is normal and the femoral head well positioned although the secondary ossification center often forms late. Type 3. The proximal femur is deficient, but the acetabulum is normal indicating the presence of the femoral head. This also tends to ossify late. There is shortening of the shaft and sclerosis of the proximal and middle thirds. If the proximal metaphysis of the affected femur is dysplastic, showing marked broadening and irregularity, the limb is designated type 3. Two patterns occur with growth. In one, the defect goes on to ossify in various degrees of varus, while in the other pattern, the defect do not ossify, leading to a pseudarthrosis and lack of continuity between the two segments. Type 4. There are limbs with no acetabulum and no femoral head and thus with an unstable distal segment. The distal segment is short and in the form of a tapered spike. Type 5. There are limbs with no hip joint and no evidence of a separate femoral segment except perhaps a small distal fragment of the bone that is usually adjacent to the tibia. Haminishi Haminishi reviewed a large series of 70 patients with 91 congenital short femurs encompassing the entire spectrum of femoral developmental abnormalities [315]. He compared the structural differences caused by the drug thalidomide with those of spontaneous occurrence and noted no essential anatomic difference between the two groups, although the whole complex of abnormalities differed in that the thalidomide group tended to show radius anomalies while the non-thalidomide group had femur-fibula-ulna anomalies.

408

In relation to the spontaneous type, there were 67 affected femurs in 56 patients. His all-inclusive classification divides the entity into five types with varying subtypes. Type I. Simple hypoplasia of the femur: (a) normal shape, (b) slightly angulated shaft and cortical thickening. Type II. Short femur with angular shaft: (c) marked lateral angulation and cortical thickening resulting from transverse subtrochanteric ossification defect, and (d) decreased neck- shaft angle. Type III. Short femur with coxa vara: (e) type IIIa (straight shaft), stable coxa vara with marked cortical thickening at the lesser trochanter; (f) type IIIb (angulated shaft), progressive coxa vara with thickened cortex. Type IV. Absent or defective proximal femur: (g) absent or fibrous neck and trochanter; migration of the upper shaft, short shaft-head distance, and diaphysial transverse ossification defect; (h) absent neck and trochanter and small femoral head connecting directly to the tapered shaft; and (i) all the proximal femur is absent. Type V. Absent or rudimentary femur: (j) rudimentary distal femur that is ossified later. Pappas Pappas developed a 9-class categorization based primarily on patients followed longitudinally [309]. Pappas defined the percent of femoral shortening in each of the 9 classes; detailed the femoral and pelvic abnormalities; assessed associated abnormalities of the tibia, fibula, patella, and feet; and defined treatment objectives. The large number of patients available for this study demonstrated a continuum of abnormalities. Class I. The femur is entirely absent and the acetabular region of the pelvis markedly hypoplastic. Class II. The proximal 75% of the femur is absent. Class III. There is no bony connection between the femoral shaft and head although the femoral head with delayed ossification is present in the acetabulum. Class IV. The femur is present to approximately one-half its length but the proximal abnormalities show the femoral head in the acetabulum with the head and shaft joined by irregular calcification in a fibrocartilaginous matrix. [It is these 4 disorders that are generally referred to as proximal femoral focal deficiency.] Class V. The femur diaphysis and distal end are incompletely ossified and hypoplastic. Class VI. The proximal two-thirds of the femur is perfectly normal and the hypoplasia is in the distal third with an irregular distal femoral region and no evident distal epiphysis. [Class V and VI are examples of what could be described as distal femoral focal deficiency.] Class VII. Congenital coxa vara with a hypoplastic femur that is shortened and somewhat bowed and also demonstrates lateral femoral condylar deficiency.

3 Slipped Capital Femoral Epiphysis: Developmental Coxa Vara

Class VIII. Proximal femoral coxa valga, a hypoplastic femur, and abnormality of the distal femoral condyles with the lateral condyle being somewhat flattened. [Most would include congenital short femur in this category that perhaps most represents Class VIII, although it characteristically has anterolateral bowing which Pappas does not demonstrate.] Class IX. The femur is essentially normal and might be defined by others as having only shortness referred to as hemiatrophy or anisomelia. Pappas also demonstrates the frequently seen underdevelopment of the lateral femoral condyle predisposing to both a valgus deformity at the knee and a tendency toward lateral patellar subluxation. Paley Paley has divided the various deformities into four types helping define treatment approaches [310]. Type 1. Intact femur with normal hip and knee. (a) Normal ossification (normal appearing femur which may be slightly short); (b) delayed ossification, subtrochanteric type (cartilage tissue with continuity in subtrochanteric region, coxa vara); and (c) delayed ossification, neck type (cartilage tissue with continuity in neck and intertrochanteric region, coxa vara). Type 2. Mobile pseudarthrosis with mobile knee. (a) Femoral head present in acetabulum but proximal femur mobile (unstable) due to loss of tissue continuity between head-neck and shaft (pseudarthrosis), middle and distal femur and knee normal; (b) femoral head absent or rudimentary in acetabulum and proximal femur mobile (unstable), middle and distal femur and knee normal Type 3. Diaphyseal deficiency of femur. (a) Proximal ½ of femur completely absent, distal femur hypoplastic but knee present with >45° motion; (b) proximal ¾ + of femur completely absent, distal femur abnormal at knee with 45°, (b) distal physis present, knee motion 2 lengthening or amputation]. The type II fibular deficiency is a foot unsuitable for salvage irrespective of the extent of limb shortening which referred to those with two or fewer rays; these are then subclassified into type IIA with a functional upper extremity and IIB with a nonfunctional upper extremity (which had some bearing on lower extremity management). [2A, early amputation; 2B consider salvage to lower extremity]. (vi) Paley. The Paley classification defines four types [108] (Table 6.5). It is considerably different from the preceding types since it primarily addresses ankle and foot position first, followed by tibial shortening, for the express purpose of outlining surgical interventions correcting these and leading to a plantigrade foot and lengthened tibia for weight-bearing function. As Paley indicates, “the best prognostic factor is the foot deformity itself.” The four types of deformity, all of which have tibial shortening of variable amounts, include: • Type I: Stable ankle. The ankle is stable with no significant deformity and the foot has no deformity. The fibula may only show proximal shortening of no clinical signifiTable 6.5 Paley Classification of Fibular Hemimelia Type I: Stable ankle. The ankle is stable with no significant deformity and the foot has no deformity. The fibula may only show proximal shortening or it may be completely atrophic. The tibia is shortened but not otherwise deformed Type II: Dynamic valgus. There is a dynamic valgus foot deformity that can usually be passively corrected; often a ball and socket ankle deformity; limited ankle dorsiflexion but no fixed equinus deformity and fibula is shortened distally at the ankle further predisposing to valgus positioning. Type III: Fixed equinovalgus deformity. Fixed equinovalgus foot deformity with distal tibial valgus/procurvatum deformity and/or a malunited subtalar (talocalcaneal) coalition Subtype (3A) ankle type: distal talar valgus malorientation with procurvatum Subtype (3B) subtalar type: deformity characterized by a malunited subtalar coalition Subtype (3C) both ankle and subtalar types combined Type IV: Fixed equinovarus deformity. Deformity is centered at subtalar coalition but with varus malposition. There is still distal tibia valga with procurvatum The classification is based on the position of the ankle and foot and serves as a guide to the surgical correction needed to obtain a plantigrade functional foot rather than early amputation with prosthetic fitting

cance, or it may be completely atrophic. The tibia is shortened but not otherwise deformed. • Type II: Dynamic valgus. There is a dynamic valgus foot deformity that can usually be passively corrected. There is often a ball and socket ankle deformity. There can be limited ankle dorsiflexion, but there is no fixed equinus deformity. The fibula is shortened distally at the ankle further predisposing to valgus positioning with gait. • Type III: Fixed equinovalgus deformity. There may be a fixed equinovalgus foot deformity with distal tibial valgus/procurvatum deformity and/or a malunited subtalar (talocalcaneal) coalition. Subtype 3A) ankle type. Characterize by a distal talar valgus malorientation with procurvatum Subtype 3B) subtalar type. Deformity characterized by a malunited subtalar coalition Subtype 3C) both ankle and subtalar types combined • Type IV: Fixed equinovarus deformity. Deformity is centered at subtalar coalition but with varus malposition. There is still distal tibia valga with procurvatum. Paley is a strong advocate for recognition of these foot deformity patterns so as to manage each surgically, along with tibial lengthening, and thus minimize or eliminate the need for amputation. While soft tissue procedures are needed to a certain extent, combination of distal tibial osteotomy with shortening (bone resection) allows for ankle joint alignment without leading to extensive soft tissue corrections (such as heel cord lengthening) and long-term recurrent soft tissue problems. The supramalleolar distal tibial osteotomies remove appropriate bone wedges to correct valgus and procurvatum (anterior angulation) and resect tibial bone to relax the adjacent soft tissues (muscle/tendon) limiting the need to lengthen them; subtalar osteotomy further corrects valgus or varus deformation of the foot; and limb lengthening manages overall tibial shortening. Paley has derived two terms to describe the two basic surgical corrective procedures used, depending on the classification, in any particular case. The SHORDT procedure refers to a SHortening Osteotomy Realignment of the Distal Tibia. The SUPERankle procedure refers to the Systematic Utilitarian Procedure for Extremity Reconstruction at the ankle (others being described at the hip and knee). The ankle procedure involves a shortening corrective wedge distal tibia osteotomy that essentially always corrects distal tibial/ankle valgus, soft issue releases as needed, and +/− subtalar osteotomy for coalition malalignment. Brief overview of procedures used in the various types. The type I deformity requires a lengthening of the tibia only (no foot surgery needed). Type II is managed with a SHORDT procedure where the distal tibial osteotomy involves shortening, varus tilt to correct the valgus deformity, and slight extension to correct any procurvatum (held with a bent surgical plate). Tibial lengthening (Ilizarov) is performed either at

6.5 Congenital Lower Extremity Limb Deficiencies

a second stage or at the same time as the SHORDT procedure. Types III and IV require the SUPERankle approach with a distal tibial shortening – varus – extension osteotomy, subtalar correction (subtalar talocalcaneal osteotomy/fusion) depending on the direction of deformity that can produce varus or valgus deformation and considerable equinus, and tibial lengthening. Paley recommends the initial SUPERankle alignment procedure at 18–24 months. If lengthening is not done at that time, the lengthening procedures generally begin at 4 years of age and may need to be repeated at 8 and 12 years depending on the extent of the discrepancy.

6.5.3.5 Clinical Studies From 1960 to 1990 (i) Farmer and Laurin. Farmer and Laurin recommended early Syme amputation when the length discrepancy was projected to be more than 7.6 cm (3 inches) at maturity, especially when severe foot deformity was present [100]. (ii) Westin et al. Westin et al. made a similar recommendation in their review of 32 patients with 37 fibular deficiencies [109]. Many of their patients underwent Syme’s amputation; the two indications were (i) a foot deformity so severe that any surgery to make the foot plantigrade and functional was likely to fail and (ii) a lower extremity length discrepancy of 7.5 cm or more would be present at skeletal maturity in the absence of any management. Amputation was performed in 29 of 37 cases. They considered the results of the Syme’s amputations to be uniformly good. Growth inhibition was constant with time. The growth inhibition in those treated without amputation ranged from 7% to 12% in the tibia and from 0 to 14% in the femur. In the amputee group, the inhibition was 22–42% in the tibia and 0–22% in the femur. In a listing of nine patients with unilateral Syme amputations who were followed to skeletal maturity and who had growth data to skeletal maturity, the final femoral and tibial discrepancy ranged between 7.8 and 24.1 cm with a mean of 14.4 cm with the tibial discrepancies themselves ranging from 6.7 cm to 14.8 cm with a mean of 11.0 cm. (iii) Hootnick et al. Hootnik et al. studied 43 patients with partial or complete absence of the fibula and a congenital short tibia [110]. They also determined that the relative difference in growth between the two limbs remained remarkably constant and thus adhered to the Shapiro type I pattern of length discrepancy development. The patients studied had a strictly unilateral variant, and all measurements were determined radiographically by scanograms or from films showing both tibias on the same x-ray plate in the youngest children. The serial radiographic measurements of leg length were available in 14 patients covering an average observation period of 9.3 years. Those with sequen-

651

tial radiographs were in the more severe end of the spectrum with the fibula absent in 11 patients and present but abnormal in 3. The amount of limb shortening was greater as the number of rays (metatarsal bones) diminished. There were 36 patients on whom assessments could be made in terms of the number of metatarsal bones and the amount of lower extremity shortening. In 12 patients with 5 metatarsal bones, the average shortening was 8.7 cm (range 3.6–12.7); in 11 patients with 4 metatarsals, the average shortening was greater at 9.5 cm (range 3.8–13.5); in 11 patients with 3 metatarsals, the average shortening was 11.8 cm (range 4.8–16.5); and in 2 patients with only 2 metatarsals, the average shortening was 14.6 cm (range 11.9– 17.3). The average age reached in the first three groups was 11 years and in the final group 9.5 years of age. The femur was only minimally affected in these patients. In the 14 patients followed radiographically, there is excellent documentation that the percent inhibition of growth in the affected limb compared to the normal remained unchanged from the earliest documentation to skeletal maturity. Femoral involvement at skeletal maturity was relatively small ranging from 86% to 96% length compared to the normal side, while tibial involvement was somewhat greater ranging from 73% to 82% length of the normal side. In patients followed for several years, although not quite to skeletal maturity, the same pattern persisted with femoral shortening in all patients except one being only 92–99% of normal with associated tibial shortening 61–90% of the normal side. The ability to project accurately the final discrepancy is of great clinical value. They felt that if the predicted shortening was less than 8.7 cm, efforts at limb equalization were warranted, while if projected discrepancies were between 8.7 and 15.0 cm, early amputation by a modified Syme’s procedure was in order. In those discrepancies projected to be greater than 15.0 cm retention of the foot and its adaptation to a prosthesis by the rotation Van Ness procedure was warranted. (iv) Choi et al. Choi et al. assessed the extent of growth discrepancy as well as management considerations [111]. They evaluated 48 extremities in 43 patients with the disorders skewed to the more severe types in their series. There were 7 fibulas of the type IA categorization, 2 type IB, and 39 type IIB (complete absence of the fibula or presence of only a distal vestigial fragment according to the classification of Achterman and Kalamchi). Treatment varied between those having amputation and those having surgical lengthening and reconstruction procedures. They subclassified their patients according to the amount of inequality projected for the lower limbs. In group I the percentage of

652

6 Torsional, Angular, and Deficiency Disorders of the Lower Extremity

shortening was 15% or less with the foot of the shorter extremity at the distal third of the contralateral normal limb; group II, between 16% and 25% shortening with the foot of the shorter extremity at the level of the middle third of the contralateral normal limb; and group III, greater than 26% shortening with the foot of the shorter extremity at the level of the proximal third of the contralateral normal limb. They concluded that lengthening was best suited only for group I patients who had stable hips, knees and ankles, and a plantigrade foot, while patients in groups II and III were best served by ablation of the foot and a prosthetic fitting. Either the Syme or Boyd amputation was used with the latter increasingly favored. The data provided indicated the extensiveness of the shortening in these disorders. The group projected the limb length discrepancy at maturity that, owing to the almost invariable Shapiro type I pattern in these deformities, was clinically accurate. In 15 patients in the group I category, the mean discrepancy projected to skeletal maturity was 8.85 cm with a range from 5.0 to 12.07 cm. Group II, 20 involved limbs had a mean projected discrepancy of 16.29 cm with a range between 12.5 cm and 22.5 cm. Deformities were even greater in terms of extent and bony deformity for group III patients, but numbers were not provided.

6.5.3.6 Treatment by Early Amputation Versus Limb Preservation with Surgical Lengthening and Reconstruction Virtually every article reviewing treatment for moderate and severe limb length discrepancies and deformation with fibular hemimelia discusses the question of whether early amputation and prosthetic fitting or a prolonged series of surgical reconstructive procedures, limb lengthening, and opposite limb epiphysiodesis is the preferred method to recommend and employ. Improvements in lower extremity limb lengthening and deformity management, using the circular Ilizarov and Taylor spatial frames, have led to surgical correction of the more severe variants with improved outcomes. The need for several operative procedures in each patient and the increasing complication rate as the amount of lengthening increases have further stimulated the debate regarding choice of management. McCarthy et al. did a comparison of outcomes between 15 patients having early amputation and 10 undergoing tibial lengthening [112]. All procedures were from the same unit by two senior surgeons. All the preoperative clinical parameters were actually more severe in the amputation group. The mean age at amputation was 1.2 years and at initial lengthening was 9.7 years. At follow-up the study “demonstrated that children who undergo early amputation are more active, have less pain, are more satisfied, have fewer complications, undergo fewer procedures, and incur less cost than those who undergo lengthening.” Similar

views were described in other reports, many of which sought to define the parameters for deciding which approach was warranted based primarily on expectations of foot function in association with ray deletions and other deformities as well as the expected limb discrepancy at skeletal maturation. As noted above, at this time the Ilizarov (and Taylor) technique was achieving greater bone and soft tissue correction than in earlier decades. Patel et al. outlined the reasons for proceeding with lengthening, reconstruction, and limb preservation in a well-reasoned discussion [113], and large studies, discussed in the next section, began to appear where surgical methods were applied to increasingly severe type II patients.

6.5.3.7 More Recent Clinical Studies, 2010 and After (i) Distribution of types and associated abnormalities. In a study of 45 patients by Rodriguez-Ramirez et al., type IA (Atcherman-Kalamchi) was seen in 29 (64.4%), type IB in 3 (6.7%), and type II in 13 (28.9%) [114]. Lateral femoral condylar hypoplasia was seen in 42 (93%), along with tarsal coalition in 23 (51%), congenital short femur in 22 (49%), ball and socket ankle in 21 (47%), and forefoot ray deletion in 20 (44%). Nine patients had 3 rays, 11 had 4 rays, and 25 had 5 rays. There was a marked, but not absolute, tendency for the type IA patients to have the least discrepancies (generally 50% of the length of the ipsilateral tibia had preservation of the foot. However, an absent or vestigial fibula still had foot preservation in 41.5% of cases. Oberc and Sulko studied 31 patients with fibular hemimelia [115]. Four had early Syme’s amputation in the first year of life based on a severely affected two- or three-ray foot and a projected length discrepancy of more than 50%. The others underwent either no treatment (mild discrepancies) or combinations of lengthening and epiphyseal arrest. Most lengthenings were done with the Ilizarov technique. The tarsal coalitions commonly involved the talocalcaneal joint as well as navicular and cuboid bones. El-Sayed et al. reviewed their extensive experience treating 157 consecutive patients (180 limb segments) with severe type II hemimelia between 1986 and 2009 [116]. They sought to lengthen and preserve the limb in each instance, and amputation was not done in any of the 157 patients. Only 12.1% did not reach the treatment goal length.

6.5 Congenital Lower Extremity Limb Deficiencies

The Ilizarov technique was used for the elongations, including some of the soft tissue contractures, and, although many surgeries were needed, they considered their overall results to be “favorable.” Of the 157 patients, 23 were bilateral (14.6%). In the 180 affected tibial-fibular segments, 166 (92%) had ankle instability, 117 (65%) an associated tarsal coalition, and 57 (31.7%) an absent anterior cruciate ligament. In the 180 feet, none had 5 rays, 3 (1.7%) had 4 rays, 81 (45%) had 3 rays, 85 (47.2%) had 2 rays, and 11 (6.1%) had only 1 ray. In all the unilateral cases, the projected length discrepancy was more than 12 cm at skeletal maturity. Management at the end of the first year of life (where the patient presented early enough) involved extensive surgery to align the lower extremity with centralization of the foot under the tibia. Procedures included (where needed) posterolateral foot/ankle release, tendon releases and lengthenings, tendo Achilles lengthening, excision of the fibular anlage, correction of tibial angulation with a wedge osteotomy, and centralization of the foot under the tibia with a transcalcaneal-tibial K-wire followed by bracing once the corrective procedures were healed. Bone lengthening procedures were then concentrated at preschool age (around 5 years of age) and again around 14 years of age. The Ilizarov construct was usually extended above the knee because of knee instability and to prevent contractures with lengthening and sometimes also to perform simultaneous femoral lengthening. The mean lengthening achieved, in total, was 13.6 cm (9.9–20.1). Complications were frequent but manageable. Independent walking was achieved in all cases, the foot was plantigrade in 141/180 (78.3%) with mild equinus or valgus in 39 (21.6%), and eventual non-union in the tibia was not seen. Residual shortening or deformity was managed by orthoses. The authors stressed the value of limb retention and pointed out the importance of initial soft tissue and bony correction of contractures/angulation and then a general expectation for lengthening at two age frames around 5 and 14 years. Catagni et al. also reviewed their recent experience with the most severe forms of hemimelia attempting to lengthen, reconstruct, and preserve the limb with multiple surgical procedures rather than choosing early amputation [117]. They assessed 32 patients who had reached skeletal maturity with type III deformity based on their categorization of those with complete absence of the fibula, severe tibial deformity and shortening, and an equinovalgus foot (type II patients in the Achterman-Kalamchi classification). The management plan involved initial surgical correction of the foot deformities as soon as possible followed by limb lengthening at the beginning of school age and a final set of bone lengthening and correction toward the end of skeletal growth. Foot deformities were corrected with soft tissue releases, lengthening of the peroneal and Achilles tendons, and insertion of a transcalcaneal- tibial pin. The Ilizarov technique was used for tibial lengthening with foot and knee extensions as needed to control

653

contractures (Fig. 6.23). The mean lengthening achieved was 13 cm (8–17). Complications were frequent but manageable. Two patients (6%) eventually required a Syme amputation. Valgus angulations of the ankle were noted to be a predictor of a poor result. They were satisfied that many of the most severely limbs could be retained with an extensive surgical program and reserved early amputation for the extensively involved children with projected discrepancies of 20–30 cm and marked foot deformities with only one or two rays. Changulani et al. reported a small series of eight limbs successfully treated with the Taylor or Ilizarov lengthening frames but did use selective early Syme amputation in their unit based on more widely used selection criteria [118]. Patients with only one or two rays were considered to have a nonfunctional foot and were offered Syme amputations that were done below 2 years of age. Those with four or five rays were considered to have a functional foot and were offered limb reconstruction, while those with an intermedi-

a

b

Fig. 6.23 Illustration shows surgical correction of fibular hemimelia with Ilizarov apparatus. Osteotomy corrects angular deformity of mid- shaft (a) and allows for elongation and healing by mechanism of distraction osteogenesis (b). Foot deformity is corrected by tendo Achilles lengthening, peroneal tendon lengthening (if needed) and soft tissue releases before application of the Ilizarov foot extension. (Reprinted with permission from Catagni et al., Clin Orthop Rel Res 2011;469:1175–1180, Springer Publishing)

654

6 Torsional, Angular, and Deficiency Disorders of the Lower Extremity

ate three rays were assessed in detail for either approach. Prior to bone lengthening, soft tissue surgery was done to align the foot and leg with such procedures as excision of the fibular anlage, posterolateral soft tissue releases, and centralization of the foot. They supported the value of the circular frame for limb reconstruction in children with less severe forms of fibular hemimelia. An overview of management considerations for fibular hemimelia is presented in Table 6.6.

6.5.4 Congenital Developmental Abnormalities (Limb Deficiencies) of the Tibia; Tibial Hemimelia 6.5.4.1 Jones et al. Classification Partial or complete absence of the tibia is referred to as tibial hemimelia. These disorders are rare and markedly less frequent than the fibular variant. There is marked shortening and bowing of the involved leg, a flexion contracture of the

Table 6.6 Treatment Considerations for Fibular Hemimelia While the disorder is referred to as fibular hemimelia, treatment considerations relate primarily not to the fibula but to: (i) Tibial deformity that can include distal valgus with tibial articular surface obliquity and mid-diaphyseal procurvatum (sagittal plane angulation with apex anterior) (ii) Tibial shortening (iii) Ankle instability (due to a shortened/proximally displaced to absent distal fibula and oblique tibial articular surface) (iv) Equinovalgus or equinovarus foot deformity (due to tendon shortening and contracture of the tendo Achilles and lateral or medial tendons, respectively, and to subtalar/talocalcaneal coalition with the two bones in a deformed position) (v) Number of rays in involved foot (vi) Extent of femoral shortening on involved side Tibial deformity needs to be corrected to provide a transverse or horizontal distal tibial articular surface for optimal weight bearing. The distal osteotomy removes a wedge of bone that corrects valgus deformation (varus osteotomy) and persisting sagittal procurvatum and also provides shortening to minimize or eliminate the need for tendon lengthening. The osteotomy is fixed internally with an angled side plate. Mid-diaphyseal angulation is generally corrected with osteotomy, aligned and stabilized by external fixation Tibial shortening is corrected by distraction lengthening after ankle alignment is corrected or in conjunction with its correction. The lengthening procedure may need to be repeated one or more times until skeletal maturity. Tibial lengthening by distraction osteogenesis techniques when done with open growth plates also causes growth slowdown that must be factored into length projections and surgical timing Ankle instability is due to several factors that need to be addressed by surgical treatments. The lateral malleolus (of the fibula) may be completely missing, more proximally positioned than normal, or represented only by a firm fibrocartilaginous anlage that tethers and resists distal tibial physeal growth. Each of these findings is associated with limited lateral ankle support and favors lateral tilt and displacement of the ankle joint and foot. Distal tibial osteotomy makes the articular surface horizontal thus increasing ankle stability. Additional measures to minimize or correct this deformity include release of the distal tibial-fibular ligamentous attachments or even removal of the distal anlage if it still constitutes a deforming force Equinovalgus foot deformity is the commonest foot position with fibular hemimelia. While tendo Achilles lengthening can overcome equinus positioning, removal of sufficient bone from distal tibial osteotomies relaxes the tendon and may eliminate any need for lengthening it. Bone removal may also eliminate the need for lateral (peroneal) or medial (tibialis posterior and tibialis anterior) tendon lengthening procedures. A tarsal coalition with the talus in severe equinus and the calcaneus displaced to the lateral side, and further tilted into valgus, is corrected with subtalar osteotomy to align the long axis of these tarsal bones along the midline. On occasion the coalition predisposes to an equinovarus position that is also corrected with a subtalar osteotomy. The distal tibial articular surface, however, is still tilted into valgus even if the foot itself tilts varus Number of rays in involved foot can influence the ability to achieve a good long-term result. Loss of the lateral two rays makes it even more difficult to achieve an acceptable result, while those with four or five rays can respond well to intervention. Some groups determine prognosis based on more or less than three rays (with approximately half of those with only three rays responding well to interventions) Extent of femoral shortening on involved side must be taken into consideration since virtually all patients with fibular hemimelia have some shortening of the femur as well Multiple surgeries may be needed in the childhood years to provide a plantigrade, balanced, weight-bearing foot with the length of the involved lower extremity the same as the contralateral side. Optimal management is based on improved surgical techniques with fuller understanding of the pathoanatomy of the deformity More refined surgical techniques and better understanding of the pathoanatomy help reach the goal of a functional, pain-free, plantigrade, weight-bearing foot with minimal to no lower extremity length discrepancy Many centers still favor an early and initial plan of amputation and prosthetic fitting for severe variants as walking begins, but many surgeons and families work for limb retention and many cultures strongly disfavor amputation References 1. Paley D. J Child Orthop 2016;10;557–583 2. McCarthy et al. J Bone Joint Surg Am 2000;82:1732–1735 3. Patel et al. J Bone Joint Surg Am 2002;84:317–319 4. Catagni et al. Clin Orthop Relat Res 2011:469;1175–1180 5. Shapiro F. J Bone Joint Surg Am 1987;69:684–690 6. Birch et al, J Bone Joint Surg Am 2011;93:1144–1151

6.5 Congenital Lower Extremity Limb Deficiencies

knee, and a rigid varus foot. Four basic patterns were identified and classified as defined by Jones et al. [119] (Fig. 6.24). In type IA, the tibia is completely absent, and there is a markedly hypoplastic lower femoral epiphysis. In type IB, the tibia is also completely absent, but there is a normal lower femoral epiphysis. This distinction is important since a normal distal femur implies that there may be a proximal tibial epiphysis that is not apparent in the newborn and first year of life since it is cartilaginous with delayed ossification. Recognition is important since the presence of the proximal tibia allows for preservation of the knee that is highly positive for prosthetic fitting. This also points to the current value of early MR imaging to identify cartilage models. In both

1a

655

instances (types IA and IB), the fibula rides high laterally in relation to its normal position had the proximal tibia been present. In type II, the upper proximal tibial epiphysis is present as is a small portion of the metaphysis, but the rest of the tibia distally does not form. This is important since the knee joint is preserved. In types I and II, the foot is in a varus position that tends to be very rigid and difficult to correct even with repeated surgeries. In type III, the proximal tibia is absent but the distal one-third is present. This implies some ankle joint preservation and function. In type IV, there is a marked diastasis between the tibia and the fibula, the talus may be displaced upward into the gap, and the distal one- third of the tibia is absent. The most distal tibial segment

1b

2

I

3

II

III

4

Fig. 6.24 The Jones et al. and Kalamchi-Dawe classifications provide a basic outline of the range of tibia hemimelia deformities. Although more detailed classifications of Weber and Paley have been presented recently, the older classifications need to be understood since treatment interventions and assessments of results used those two groupings prior to approximately 2005 when the newer ones were described. (a) Jones et al. classification of tibial hemimelia (left) has four types. At the upper left, type I has complete absence of the tibia, with a less developed hypoplastic distal femoral epiphysis referred to as Jones type Ia and a better-developed distal femur referred to as Jones type Ib. At the upper right, there is partial presence of the tibia involving the proximal portion referred to as Jones

type II. At the lower left, there is also a partial presence of the tibia, but it is the distal segment present referred to as Jones type III. At the lower right, there is separation (diastasis) of the tibia and fibula distally and aplasia of the distal tibia and proximal displacement of the talus between the lower ends of tibia and fibula. (b) Kalamchi and Dawe classification is shown at right. Type I (left) has a completely absent tibia with proximal/ lateral displacement of the fibula, type II (middle) has the proximal tibia only of varying length with distal tibial aplasia with the fibula closer to its normal position, and type III (right) has a dysplasia of the distal tibia with separation of tibia and fibula distally at the syndesmosis (diastasis) and proximal positioning of the talus between the distal ends

656

6 Torsional, Angular, and Deficiency Disorders of the Lower Extremity

tends to be curved medially. In type IV deformity, the knee also tends to be well formed, but the talus has subluxed proximally between the now separated tibia and fibula. Treatment is directed toward early disarticulation of the knee and prosthetic fitting for type I lesions. Prostheses are also helpful and more effective if a functioning knee can be retained. Tibiofibular fusions can be helpful where the knee is retained to increase effective single bone mass. Fibulacalcaneal fusions for type IV deformities may help salvage an ankle. Efforts are made to save the foot, but the difficulty of correcting the varus deformity even with frequent surgery and the lack of an ankle joint in some variants can lead to Syme amputation for stable prosthetic fitting. Syme and Boyd amputations enable the distal extremity to fit comfortably into weight-bearing prostheses.

6.5.4.2 Schoenecker et al. Review of 71 Limbs Schoenecker et al. studied 71 limbs in 57 patients with congenital tibial hemimelia, using the Jones et al. classification [120]. There were 33 type IA limbs, 6 type IB, 15 type II, 7 type III, and 10 type IV. In the 54 limbs with the type I or II deficiency, there were 22 that had knee disarticulation, 25 Syme amputation, and 1 Chopart amputation; the foot was retained only in six with these two variants. In the 17 extremities with a type III or IV deficiency, there were nine that had Syme amputation and four a Chopart amputation; the foot was retained in four in these two variants. They report that 56 of 57 patients walked independently. The severity of the tibial hemimelia disorder was great making even attempts at reconstruction daunting in terms of the degree of shortening, insufficient proximal or distal tibia to allow for meaningful knee or ankle reconstruction, the absence of most or all of the major leg bone (tibia) leaving only the narrow fibula to work with, and a rigid varus foot deformity as well. 6.5.4.3 Kalamchi and Dawe Classification Kalamchi and Dawe simplified the classification defining three types [121]. In type I there was total absence of the tibia; type II, distal absence of the tibia; and type III, distal deficiency with tibiofibular diastasis. The fibula was always present but in type I was subluxed proximal and lateral to the distal femur, with a similar position sometimes occurring also in the type II variant. In type I the disorder was also invariably associated with a marked flexion contracture of the knee, variable rotation of the leg, and marked inversion and adduction deformities of the foot. The distal femur was usually hypoplastic with marked retardation of the ossification center of the distal epiphysis. The functional status of

the quadriceps muscle, the severity of the flexion contracture of the knee, and the position and function of the foot all had to be considered in planning surgical straightening or ablation. Jones et al. stressed that in all three, the fibula is relatively normal in form and development although it often is situated (in types I and II) proximal to its normal position at the knee [121]. In some the tibial segment is greater than it appears at birth since it is present as cartilage manifesting delayed ossification. Careful clinical examination and other forms of imaging are important to clearly define the anatomic structure in the newborn. They assessed management in 24 limbs: 10 type I, 11 type II, and 3 type III. Their eventual management profile included early knee disarticulation for type I limbs; for type II limbs, tibiofibular synostosis with maintenance of the knee for improved function and either maintenance of the foot or conversion of the foot to a Boyd amputation fusing the os calcis to the distal fibula; and, for type III limbs, calcaneofibular fusion to maintain foot stability, fusion of the distal tibial- talar-fibular diastasis if warranted, and then conversion to a Boyd amputation if necessary for type III limbs. Each of these methods is designed to adapt the limb to optimize prosthetic manufacture and maintenance for independent walking.

6.5.4.4 Weber Classification Weber has provided an extensive classification of tibial hemimelia designed to facilitate surgical intervention by stressing the wide variability of cases seen [122] (Fig. 6.25). The classification is based on 95 limbs in 63 patients. He defines seven types of the tibial disorder with types III to VII having two subgroups each (for a total of 12 groups). Previous classifications were based on radiographic findings while stresses the need to include sonographic and MR imaging studies to define the presence or absence of cartilaginous anlage as well as the tibial and fibular bone components. Detailed classification is important not only to assess the pathoanatomy but also to guide surgical management. He defines hypoplasia as a complete but underdeveloped tibial bone, aplasia as a partial loss of the bony (osseous) component of the tibia, and agenesia (agenesis) as a total lack of tibial bone. The assessment always includes determination of the presence or absence of a cartilaginous anlage for the tibia; this represents the subgrouping with a = presence of a cartilaginous anlage and b = absence of a cartilaginous anlage. The classification is as follows: type I, tibial hypoplasia; type II, diastasis (of distal tibia and fibula, with talus relatively proximally displaced into the ankle gap); type III, distal aplasia, partial loss of the tibial bone distally at ankle

6.5 Congenital Lower Extremity Limb Deficiencies

Type-I

Type-II

657

a

Type-III

b

a

Type-IV

b

a

Type-VII

b

Femur

Tibia

Fibula

Talus

a

Type-V

b

a

Type-VI

b

Fig. 6.25 Weber classification of tibial hemimelia is shown. This is designed to provide greater detail than previous classifications for the purpose of directing surgical correction of the involved limb. (Reprinted with permission from Weber M. J Child Orthop 2008;2:169–175, Springer)

region (a = cartilage anlage, b = absent cartilaginous anlage); type IV, proximal aplasia, partial loss of tibial bone at knee region (a/b); type V, bifocal aplasia, partial loss of tibial bone both distally and proximally (a/b); type VI, agenesia with double fibula (a/b); and type VII, agenesia with single fibula (a/b). Weber documented the classification diagnosis in the group he studied as type VII, 61%; type III, 15%; type I, 6%; type V, 6%; type II, 5%; type IV, 3%; and type IV, 3%. The first two commonest groups accounted for 76% of all patients. Weber also developed an even wider classification to include the hip, femur, patella, fibula, foot, as well as the tibia. The fibula was graded as normal, hypoplasia, agenesia, and the foot as being I, normal with all 5 rays; II, with rays 3 and 4 present; and III, with only rays 1 and 2 present.

6.5.4.5 Courvoisier et al., Ilizarov Treatment Courvoisier et al. reviewed nine cases of congenital longitudinal tibial deficiency (types I or II – Jones) treated by the Ilizarov technique [123]. Patients were assessed at a mean follow-up of 18 years (4–32 years). The mean maximum knee flexion was 35° (0–90°) in type I deficiencies and 118° (90–140°) in type II deficiencies. Only one patient underwent amputation and one had knee fusion. The Ilizarov technique provided satisfactory progressive corrections. An example of tibial hemimelia is shown in Fig. 6.26. There is also value in fusing the small proximal tibial bone to the hypertrophied fibula for closer to normal limb length and enhancement of weight-bearing capability (Fig. 6.27). A detailed overview of tibial hemimelia is presented in Table 6.7.

658

6 Torsional, Angular, and Deficiency Disorders of the Lower Extremity

a

bi

bii

a

biii

b c

d

c a

b

Fig. 6.26 The range of forms that tibial hemimelia can take are shown. (a) Four of the commonly referred to classifications of tibial hemimelia are compared. Studies prior to 2000 assessed management and results using the Jones et al. and Kalamchi-Dawe categorizations. With an increasing emphasis on surgical correction of some of the variants, classifications of Weber and Paley have been developed to better define the spectrum of changes. (Reprinted with permission from Kaplan-List K et al. Pediatr Radiol 2017;47:473–483, Springer.) (b) Illustrations from the early reports of absent tibia (tibia hemimelia) highlight the severity of shortening and associated deformities. (b(i)) Illustrates (a) knee flexion- adduction deformity, (b) shortened leg segment, and (c) marked varus deformation of the foot. (b(ii)) Illustrates (a) short leg, (b) varus foot deformity, (c) knee flexion-adduction deformity, and (d) shortened thigh. (b(i, ii)) Reprinted with modifications from Joachimsthal G, Zeitschr f Orthop Chir 1894;3:140–173. (b(iii)) Illustrates knee flexion deformity, short leg, prominent external malleolus, and varus foot deformity. (Reprinted from Billroth Th, Arch Klin Chir 1861;1:251–258.) (c) Radiographs highlight findings in various types of tibia hemimelia. In these radiographs, the severe knee and foot contractures and deformities

and the variable radiographic projections create the appearance of extremely “non-anatomic” relationships. (c(i)) shows lower extremity with absent tibia straight fibula and severe varus foot deformity. (Reprinted with permission from Paley D, J Child Orthop 2016;10:529–555.) (c(ii)) shows absent tibia, proximal tibia displaced posteriorly and proximally, straight fibula with normal appearing epiphyses at either end, severe varus deformity of foot, and talocalcaneal coalition. Foot deformity is so great that the dorsum of foot would rest against the floor. (Reprinted with permission from Schwarzweller F, Arch Orthop Unfall-Chir 1939;39:400– 419, Springer.) (c(iii)) shows absent tibia with (interrupted arrows) curved fibula, varus foot deformity, talocalcaneal coalition, and forefoot abnormalities. (c(iv)) shows clinical appearance at the right; at the left, presence of small proximal tibia only with straight fibula displaced to outer side of lateral femoral condyle. The distal femoral and proximal tibial epiphyses are well developed however. (c(v)) shows diastasis of the distal ends of the tibia and fibula. The distal fibular epiphysis is well developed but there is distal aplasia of the tibia. The talus is displaced proximally between the two long bone ends. (c(iv, v)) (Reprinted with permission from Spiegel DA et al., Int Orthop 2003;27:38–342, Springer)

6.5 Congenital Lower Extremity Limb Deficiencies

ci

civ

Fig. 6.26 (continued)

cii

659

ciii

cv

660

6 Torsional, Angular, and Deficiency Disorders of the Lower Extremity

Fig. 6.27 A common surgical treatment of tibial hemimelia with an extremely short proximal tibia usually includes fusion of the proximal tibial fragment to the hypertrophied fibula. (Reprinted with permission from Courvoisier et al. Orthop Traumatol Surg Res 2009;95:431–46, Elsevier Masson)

Table 6.7 Overview of Tibial Hemimelia (a) Background information Major clinical findings are (i) extensive shortening of lower extremity including some femoral shortening, (ii) flexion and adduction deformity at the knee, and (iii) inversion of foot (varus) Disorder often part of an inherited genetic syndrome such as the Langer-Giedion syndrome, approximately 60% of cases with associated congenital abnormalities Often bilateral, opposite side of variable severity, frequent associated upper extremity involvement (split-hand), usually with the same side femoral shortening, right side affected much more commonly, male and female involvement Foot may have fewer rays (deficiency) but more often has excessive bones (duplication, supernumerary) such as greater than five metatarsals and greater than five toes; synostoses common Tibial hemimelia is less common that fibular hemimelia (the most common long bone deficiency) but when present is much more severe (b) Pathoanatomy (i) Bones: Fibula – always present and structurally intact from proximal to distal ends; epiphyses at both ends; usually displaced proximally, laterally, and posteriorly at knee where tibia completely absent; often less displacement when proximal tibia present; may be bent/curved with convexity pointing laterally or may be straight; often wider than normal before walking begins and then hypertrophies more with walking Tibia – may be completely absent and replaced either by a dense fibrous band opposite fibula attached to cartilage at either end or by no evident tissue at all; when partially absent the proximal part is present, often very small, with distal aplasia; occasionally central or distal component of tibia is only part present (short bone/cartilage segment); proximal tibial cartilage anlage may not ossify until several years after birth Foot – inverted at ankle due to lack of distal tibia, medial malleolus; associated equinovarus position with tendon tightness; medial side concave, lateral side convex; knee, foot deformities plus shortening may position inner foot against medial thigh; often with excessive number of metatarsals, extra toes (supernumerary digits, duplication); less often with absent rays/phalanges (deficiency); absent rays are from medial side of foot; medial tarsal bones may be absent (navicular, cuneiforms); tarsal coalitions common (talocalcaneal, talocalcanealnavicular); localized phalangeal synostoses common Patella – may be present or absent; if present it may be malpositioned and/or misshapen Femur – slightly/moderately shortened on involved side; distal condyles may be underdeveloped with small secondary ossification center (distal hypoplasia) (ii) Joints: Knee – flexion-adduction contracture; extensor mechanism may be intact with patella and quadriceps in place with partial absence of tibia; absent patella usually indicates defective extensor mechanism; proximal fibula displaced laterally/proximally and posteriorly with complete absence of tibia but can form a femoro-fibular joint with capsule; this joint may be beneath or at lateral wall of lateral femoral condyle; menisci and cruciate ligaments absent in most

6.5 Congenital Lower Extremity Limb Deficiencies

661

Ankle – inversion of foot; equinovarus deformity (club foot positioning); distal fibula forms true joint with a capsule at lateral talus or, if missing, with superolateral calcaneus; completely rigid in variant with distal tibiofibular widening (diastasis) and proximal subluxation of talus between the two long bones (iii) Soft tissues: Completely absent tibia – may not be represented by any tissue or may be replaced by rigid fibrous band adjacent to fibula ending in cartilage at either the end and attached by persisting “interosseous” membrane to fibula Partially absent tibia – has almost always a proximal presence (variable size) and distal aplasia; the proximal segment may be bone (ossified at the proper time) but often remains as a non-ossified cartilage anlage for several years; rarely it is the central or distal part of the tibia that is present in partial cases Muscle groups – are a combination of (i) absent, (ii) present and normal appearing (with origins-insertions limited by altered bony structures), and (iii) abnormal muscles seen only in lesser species [Muscles with normal origins above knee or from fibula may insert, in altered bone regions, into whatever adjacent tendons, membranes or bones are available; muscles normally originating from absent bones are themselves absent] (iv) Blood vessels and nerves: normal; “the arteries showed a comparatively normal arrangement”; major vessel/nerve components identifiable and only limited/changed by surrounding bone and muscle; some motor function; sensation intact (v) Lower extremity length discrepancy – can be extreme approaching 18–20 cm at skeletal maturity; primarily due to tibial component but almost always some femoral shortening as well (c) Approach to patient with tibial hemimelia (i) Classifications: awareness of four basic presentations desirable (Kalamchi-Dawe, Jones, Weber, Paley); older (KW, Jones) simpler but used for many retrospective assessments; newer (Weber, Paley) complicated but include the many variations, designed to direct more active surgical interventions (ii) Matters to consider in planning therapy Patella – present or absent Knee joint – quadriceps mechanism present or absent; assess for range of motion, contractures Tibia – present as bone or as cartilage anlage (assess with ultrasound or MRI); extent of shortening Fibula – position at knee regarding relation to femur; position at ankle Ankle – assess range of motion, position of talus Foot – inversion deformity (varus), equinovarus positioning similar appearing to clubfoot, assess degree of rigidity Lower extremity length discrepancy – include femur length assessment; determine entire lower extremity current discrepancy and discrepancy projected at skeletal maturity [discrepancy will follow a type I (Shapiro) developmental pattern] (iii) Management based on considerations listed in b above with aid of various classifications: Initial management approaches, decided if possible in the first year of life, involve decision between early amputations versus efforts to maximize limb availability by staged surgical interventions throughout the growing years Early amputation: generally a disarticulation at the knee with prosthetic fitting; the associated femoral shortening still leaves the desired prosthetic articulation at the level of the opposite knee Surgical maximization of available structures present: procedures include centralization of the fibula under the femur (femoral-fibular arthroplasty) to serve as main weight-bearing structure [Brown procedure, 1965; described also by Myers in 1905, 1910, and Simmons et al. in 1996]; Tibial-fibular synostosis to convert into a single leg bone; Lengthening of fibula (or tibia/fibula) by distraction osteogenesis; Ttransverse widening of fibula by application of distraction osteogenesis with transverse distraction; Correction of distal tibial-fibular diastasis or incorporation of diastasis into ankle fusion; Improve knee and or ankle joint position by overcoming contractures with soft tissue release/lengthening or application of external fixation for correction (Ilizarov); Equinovarus (“clubfoot”) correction with soft tissue release/lengthening or application of external fixation for correction (Ilizarov); Assure plantigrade foot with shoe-wearing capability via extra ray/extra toe excision (de-bulking) Surgical maximization of leg structure followed by foot ablation: procedures listed above provide maximization of leg structure but ankle/ foot problems may not be sufficiently correctable, leading to use of an ankle disarticulation Syme (or Boyd/modified Boyd) procedure; Syme, ankle disarticulation with removal of both malleoli References 1. Evans EL, Smith NR. Congenital absence of tibia. Arch Dis Child 1926;1:194–229 2. Ollerenshaw R. Congenital defects of the long bones of the lower limb. J Bone Joint Surg 1925;7:528–552 3. Schoenecker P et al. Congenital longitudinal deficiency of the tibia. J Bone Joint Surg Am 1989;71:278–287 4. Spiegel et al. Congenital longitudinal deficiency of tibia. Int Orthop 2003;27:338–342 5. Myers TH. Congenital absence of tibia: transplantation of head of fibula: arthrodesis of ankle. Am J Orthop Surg 1905;3:72–85, [follow-up, Am J Orthop Surg 1910;8:398–400] 6. Brown FW. Construction of a knee joint in congenital absence of the tibia. J Bone Joint Surg Am 1965;47:695–704 7. Simmons ED Jr et al. Brown’s procedure for congenital absence of the tibia revisited. J Pediatr Orthop 1996;16:85–89 8. Kaplan-List K et al. Systematic radiographic evaluation of tibial hemimelia with orthopedic implications. Pediatr Radiol 2017;47:473–483 9. Paley D. Tibial hemimelia: new classification and reconstructive options. J Child Orthop 2016;10:529–555

662

6 Torsional, Angular, and Deficiency Disorders of the Lower Extremity

References 1. Staheli LT, Corbett M, Wyss C, King H. Lower-extremity rotational problems in children. Normal values to guide management. J Bone Joint Surg Am. 1985;67-A:39–47. 2. Heinrich SD, Sharps CH. Lower extremity torsional deformities in children: a prospective comparison of two treatment modalities. Orthopedics. 1991;14:655–9. 3. Seber S, Hazer B, Köse N, Göktürk E, Günal I, Turgut A. Rotational profile of the lower extremity and foot progression angle: computerized tomographic examination of 50 male adults. Arch Orthop Trauma Surg. 2000;120:255–8. 4. Losel S, Burgess-Milliron MJ, Micheli LJ, Edington CJ. A simplified technique for determining foot progression angle in children 4 to 16 years of age. J Pediatr Orthop. 1996;16:570–4. 5. Davids JR, Davis RB, Jameson LC, Westberry DE, Hardin JW. Surgical management of persistent intoeing gait due to increased internal tibial torsion in children. J Pediatr Orthop. 2014;34:467–73. 6. Staheli LT. Rotational problems in children. J Bone Joint Surg Am. 1993;75-A:939–49. 7. Engel GM, Staheli LT. The natural history of torsion and other factors influencing gait in childhood. Clin Orthop Relat Res. 1974;99:12–7. 8. Staheli LT, Engel GM. Tibial torsion. A method of assessment and a survey of normal children. Clin Orthop Relat Res. 1972;86:183–6. 9. Dunlap K, Shands AR Jr, Hollister LC Jr, Gaul JS Jr, Streit HA. A new method for determination of torsion of the femur. J Bone Joint Surg Am. 1953;35-A:289–311. 10. Crane L. Femoral torsion and its relation to toeing-in and toeing- out. J Bone Joint Surg Am. 1959;41-A:421–8. 11. Fabry G, MacEwen GD, Shands AR Jr. Torsion of the femur. A follow-up study in normal and abnormal conditions. J Bone Joint Surg Am. 1973;55-A:1726–38. 12. Svenningsen S, Terjesen T, Auflem M, Berg V. Hip rotation and in-toeing gait. A study of normal subjects from four years until adult age. Clin Orthop Relat Res. 1990;251:177–82. 13. Kingsley PC, Olmsted KL. A study to determine the angle of anteversion of the neck of the femur. J Bone Joint Surg Am. 1948;30-A:745–51. 14. McSweeny A. A study of femoral torsion in children. J Bone Joint Surg Br. 1971;53-B:90–5. 15. Ligier JN. Anomalies rotationnelles des membres inférieurs. Orthop Traumatol. 1992;2:17–20. 16. Cahuzac J-P. Vices des torsion des membres inférieurs. Rev Chir Orthop. 2006;92:395–7. 17. Hubbard DD, Staheli LT, Chew DE, Mosca VS. Medial femoral torsion and osteoarthritis. J Pediatr Orthop. 1988;8:540–2. 18. Wedge JH, Munkacsi I, Loback D. Anteversion of the femur and idiopathic osteoarthrosis of the hip. J Bone Joint Surg Am. 1989;71-A:1040–3. 19. Kitaoka HB, Weiner DS, Cook AJ, Hoyt WA Jr, Askew MJ. Relationship between femoral anteversion and osteoarthritis of the hip. J Pediatr Orthop. 1989;9:306–404. 20. Weinberg DS, Park PJ, Morris WZ, Liu RW. Femoral version and tibial torsion are not associated with hip or knee arthritis in a large osteological collection. J Pediatr Orthop. 2017;37:e120–8. 21. Tönnis D, Heinecke A. Diminished femoral antetorsion syndrome: a cause of pain and osteoarthritis. J Pediatr Orthop. 1991;11:419–31. 22. Tönnis D, Heinecke A. Acetabular and femoral anteversion: relationship with osteoarthritis of the hip. J Bone Joint Surg Am. 1999;81-A:1747–70.

23. Weseley MS, Barenfeld PA, Eisenstein AL. Thoughts on in-toeing and out-toeing: twenty years’ experience with over 5000 cases and a review of the literature. Foot Ankle. 1981;2:49–57. 24. Staheli LT, Clawson DK, Hubbard DD. Medial femoral torsion: experience with operative treatment. Clin Orthop Relat Res. 1980;146:222–5. 25. Svenningsen S, Apalset K, Terjesen T, Anda S. Osteotomy for femoral anteversion: complications in 95 children. Acta Orthop Scand. 1989;60:401–5. 26. Svenningsen S, Terjesen T, Apalset K, Anda S. Osteotomy for femoral anteversion: a prospective 9-year study of 52 children. Acta Orthop Scand. 1990;61:360–3. 27. Payne LZ, DeLuca PA. Intertrochanteric versus supracondylar osteotomy for severe femoral anteversion. J Pediatr Orthop. 1994;14:39–44. 28. Gordon JE, Pappademos PC, Schoenecker PL, Dobbs MB, Luhmann SJ. Diaphyseal derotational osteotomy with intramedullary fixation for correction of excessive femoral anteversion in children. J Pediatr Orthop. 2005;25:548–53. 29. Le Damany P. La torsion du tibia. Normale, pathologique, expérimentale. J Anat Physiol. 1909;45:598–615. 30. Badelon O, Bensahel H, Folinais D, Lassale B. Tibiofibular torsion from the fetal period until birth. J Pediatr Orthop. 1989;9:169–73. 31. Hutchins PM, Rambicki D, Comacchio L, Paterson DC. Tibiofibular torsion in normal and treated clubfoot populations. J Pediatr Orthop. 1986;6:452–5. 32. Bruce WD, Stevens PM. Surgical correction of miserable malalignment syndrome. J Pediatr Orthop. 2004;24:392–6. 33. Dodgin DA, De Swart RJ, Stefko RM, Wenger DR, Ko J-Y. Distal tibial/fibular derotation osteotomy for correction of tibial torsion: review of technique and results in 63 cases. J Pediatr Orthop. 1998;18:95–101. 34. Delgado ED, Schoenecker PL, Rich MM, Capelli AM. Treatment of severe torsional malalignment syndrome. J Pediatr Orthop. 1996;16:484–8. 35. Paget J. Ununited fractures in children. Chapter 13. London: Longmans, Green and Company; 1891. p. 130–5. 36. Hefti F, Bollini G, Dungl P, Fixsen J, Grill F, Ippolito E, Romanus B, Tudisco C, Wientroub S. Congenital pseudarthrosis of the tibia: history, etiology, classification and epidemiologic data. J Pediatr Orthop B. 2000;9:11–5. 37. Andersen KS. Radiological classification of congenital pseudarthrosis of the tibia. Acta Orthop Scand. 1973;44:719–27. 38. Crawford AH. Neurofibromatosis in children. Acta Orthop Scand. 1986;57(Suppl 218):8–60. 39. Boyd HB. Pathology and natural history of congenital pseudarthrosis of the tibia. Clin Orthop Relat Res. 1982;166:5–13. 40. Keret D, Bollini G, Dungl P, Fixsen J, Grill F, Hefti F, Ippolito E, Romanus B, Tudisco C, Wientroub S. The fibula in congenital pseudarthrosis of the tibia: the EPOS multicenter study. J Pediatr Orthop B. 2000;9:69–74. 41. Ducroquet R. A propos des pseudarthroses et enflexion congénitale du tibia. Mem Acad de Chir. 1937;63:863–8. 42. Barber CG. Congenital bowing and pseudarthrosis of the lower leg: manifestations of von Recklinghausen’s neurofibromatosis. Surg Gynecol Obstet. 1939;69:618–26. 43. Green WT, Rudo N. Pseudarthrosis and neurofibromatosis. Arch Surg. 1943;46:639–51. 44. Aegerter EE. The possible relationship of neurofibromatosis, congenital pseudarthrosis and fibrous dysplasia. J Bone Joint Surg Am. 1950;32-A:618–26. 45. McElvenny RT. Congenital pseudo-arthrosis of the tibia. The findings in one case and a suggestion as to possible etiology and treatment. Q Bull Northwest Univ Med Sch. 1949;23:413–23. 46. Boyd HB, Sage RP. Congenital pseudarthrosis of the tibia. J Bone Joint Surg Am. 1958;40A:1245–70.

References 47. Blauth M, Harms D, Schmidt D, Blauth W. Light- and electron- microscopic studies in congenital pseudarthrosis. Arch Orthop Trauma Surg. 1984;103:269–77. 48. Ippolito E, Corsi A, Grill F, Wientroub S, Bianco P. Pathology of bone lesions associated with congenital pseudarthrosis of the leg. J Pediatr Orthop B. 2000;9:3–10. 49. Van Nes CP. Congenital pseudarthrosis of the leg. J Bone Joint Surg Am. 1966;48-A:1467–83. 50. Grill F, Bollini G, Dungl P, Fixsen J, Hefti F, Ippolito E, Romanus B, Tudisco C, Wientroub S. Treatment approaches for congenital pseudarthrosis of tibia: results of the EPOS multicenter study. J Pediatr Orthop B. 2000;9:75–89. 51. Morrissy RT, Riseborough EJ, Hall JE. Congenital pseudarthrosis of the tibia. J Bone Joint Surg Br. 1981;63-B:367–75. 52. Farmer AW. The use of a composite pedicle graft for pseudarthrosis of the tibia. J Bone Joint Surg Am. 1952;34-A:591–600. 53. Hardinge K. Congenital anterior bowing of the tibia: the significance of different types in relation to prognosis. Ann Roy Coll Surg Eng. 1972;51:17–30. 54. McFarland B. Pseudarthrosis of the tibia in childhood. J Bone Joint Surg Br. 1951;33-B:36–46. 55. Boyd HB. Congenital pseudarthrosis. Treated by dual bone grafts. J Bone Joint Surg. 1941;23:497–515. 56. Charnley J. Congenital pseudarthrosis of the tibia treated by the intramedullary nail. J Bone Joint Surg Am. 1956;38-A:283–90. 57. Strong ML, Wong-Chung J. Prophylactic bypass grafting of the prepseudarthrotic tibia in neurofibromatosis. J Pediatr Orthop. 1991;11:757–64. 58. Ofluoglu O, Davidson PS, Dormans JP. Prophylactic bypass grafting and long-term bracing in the management of anterolateral bowing of the tibia and neurofibromatosis-1. J Bone Joint Surg Am. 2008;90-A:2126–34. 59. Bassett CA, Caulo N, Kort J. Congenital pseudarthrosis of the tibia: treatment with pulsing electromagnetic fields. Clin Orthop Relat Res. 1981;154:136–49. 60. Ohnishi I, Sato W, Matsuyama J, Yajima H, Haga N, Kamegaya M, et al. J Pediatr Orthop. 2005;25:219–24. 61. Pannier S. Congenital pseudarthrosis of the tibia. Orthop Traumatol Surg Res. 2011;97:750–61. 62. Shah H, Rousset M, Canavese F. Congenital pseudarthrosis of the tibia: management and complications. Indian J Orthop. 2012;46:616–26. 63. Khan T, Joseph B. Controversies in the management of congenital pseudarthrosis of the tibia and fibula. Bone Joint J. 2013;95:1027–34. 64. Elefteriou F, Kolanczyk M, Schindeler A, Viskochil DH, Hock JM, Schorry EK, et al. Skeletal abnormalities in neurofibromatosis type I: approaches to therapeutic options. Am J Med Genet A. 2009;149A:2327–38. 65. Sofield HA. Congenital pseudarthrosis of the tibia. Clin Orthop Relat Res. 1971;76:33–42. 66. Anderson DJ, Schoenecker PL, Sheridan JJ, Rich MM. Use of an intramedullary rod for the treatment of congenital pseudarthrosis of the tibia. J Bone Joint Surg Am. 1992;74-A:161–8. 67. Dobbs MB, Rich MM, Gordon JE, Szymanski DA, Schoenecker PL. Use of an intramedullary rod for treatment of congenital pseudarthrosis of the tibia. J Bone Joint Surg Am. 2004;86-A:1186–97. 68. Joseph B, Mathew G. Management of congenital pseudarthrosis of the tibia by excision of the pseudarthrosis, onlay grafting and intramedullary nailing. J Pediatr Orthop B. 2000;9:16–23. 69. Johnston CE II. Congenital pseudarthrosis of the tibia. Results of technical variation in the Charnley–Williams procedure. J Bone Joint Surg Am. 2002;84-A:1799–810. 70. Weiland AJ, Weiss A-PC, Moore JR, Tolo VT. Vascularized fibular grafts in the treatment of congenital pseudarthrosis of the tibia. J Bone Joint Surg Am. 1990;72-A:654–62.

663 71. Gilbert A, Brockman R. Congenital pseudarthrosis of the tibia. Long-term followup of 29 cases treated by microvascular bone transfer. Clin Orthop Relat Res. 1995;314:37–44. 72. Dormans JP, Krajbich JI, Zuker R, Demuynk M. Congenital pseudarthrosis of the tibia: treatment with free vascularized fibular grafts. J Pediatr Orthop. 1990;10:623–8. 73. Sakamoto A, Yoshida T, Uchida Y, Kojima T, Kubota H, Iwamoto Y. Long-term follow-up on the use of vascularized fibular graft for the treatment of congenital pseudarthrosis of the tibia. J Orthop Surg Res. 2008;3:13. [BioMed Central https://doi. org/10.1186/1749-799X-3-13]. 74. Paley D, Catagni M, Argnani F, Prevot J, Bell D, Armstrong P. Treatment of congenital pseudarthrosis of the tibia using the Ilizarov technique. J Pediatr Orthop. 1997;17:668–74. 75. Grill F. Treatment of congenital pseudarthrosis of the tibia with the circular frame technique. J Pediatr Orthop B. 1996;5:6–16. 76. Guidera KG, Raney EM, Ganey T, Albani W, Pugh L, Ogden JA. Ilizarov treatment of congenital pseudarthrosis of the tibia. J Pediatr Orthop. 1997;17:668–74. 77. Richards BS, Oetgen BE, Johnston CE. The use of rhBMP-2 for the treatment of congenital pseudarthrosis of the tibia. A case series. J Bone Joint Surg Am. 2010;92-A:177–85. 78. Lee FY-I, Sinicropi SM, Lee FS, Vitale MG, Roye DR Jr, Choi IH. Treatment of congenital pseudarthrosis of the tibia with recombinant human bone morphogenetic – 7 (rhBMP-7). A report of five cases. J Bone Joint Surg Am. 2006;88:627–33. 79. Fabeck L, Ghafi D, Gerroudj M, Baillon R, Delincé PH. Bone morphogenetic protein 7 in the treatment of congenital pseudarthrosis of the tibia. J Bone Joint Surg Br. 2006;88-B:116–8. 80. Birke O, Schindler A, Ramachandran M, Cowell CT, Munns CF, Bellemore M, Little DG. Preliminary experience with the combined use of recombinant bone morphogenetic protein and bisphosphonates in the treatment of congenital pseudarthrosis of the tibia. J Child Orthop. 2010;4:507–17. 81. Masserman RL, Peterson HA, Bianco AJ Jr. Congenital pseudarthrosis of the tibia. A review of the literature and 52 cases from the Mayo Clinic. Clin Orthop Relat Res. 1974;99:140–5. 82. Pappas AM. Congenital posteromedial bowing of the tibia and fibula. J Pediatr Orthop. 1984;4:525–31. 83. Hofmann A, Wenger DR. Posteromedial bowing of the tibia. Progression of discrepancy in leg lengths. J Bone Joint Surg Am. 1981;63-A:384–8. 84. Carlioz H, Langlais J. Les courbures congénitales de jambe à concavité antérieure: 18 observations chez l’enfant. Rev Chir Orthop. 1986;72:259–66. 85. Napiontek M, Shadi M. Congenital posteromedial bowing of tibia and fibula: treatment option by multi-level osteotomy. J Pediatr Orthop B. 2014;23:130–4. 86. Frantz CH, O’Rahilly R. Congenital skeletal limb deficiencies. J Bone Joint Surg Am. 1961;43-A:1202–24. 87. Henkel L, Willert HG. Dysmelia. A classification and a pattern of malformation in a group of congenital defects of the limbs. J Bone Joint Surg Br. 1969;51-B:399–414. 88. Kay HW, Day HJ, Henkel HL, et al. The proposed international terminology for the classification of congenital limb deficiencies. Dev Med Child Neurol Suppl. 1975;34:1–12. 89. Freund A. Congenital defects of the femur, fibula and tibia. Arch Surg. 1936;33:349–91. 90. Shapiro F. Developmental patterns in lower-extremity length discrepancies. J Bone Joint Surg Am. 1982;64-A:639–51. 91. Aitken G. Proximal femoral focal deficiency. In: Swinyard CA, editor. Limb development and deformity: problems of evaluation and rehabilitation. Springfield: Charles C Thomas; 1969. p. 456–76. 92. Amstutz HC, Wilson PD Jr. Dysgenesis of the proximal femur (coxa vara) and its surgical management. J Bone Joint Surg Am. 1962;44-A:1–24.

664

6 Torsional, Angular, and Deficiency Disorders of the Lower Extremity

93. Ring PA. Congenital short femur; simple femoral hypoplasia. J Bone Joint Surg Br. 1959;41-B:73–9. 94. Vlachos D, Carlioz H. Les malformations du femur. Leur évolution spontanée. Rev Chir Orthop. 1973;59:626–40. 95. Pappas AM. Congenital abnormalities of the femur and related lower extremity malformations: classification and treatment. J Pediatr Orthop. 1983;3:45–60. 96. Paley D, Chong DY, Prince DE. Congenital femoral deficiency reconstruction and lengthening surgery. In: Sabharwal S, editor. Pediatric lower limb deformities. Heidelberg: Springer; 2016. p. 361–425. 97. Kostuik JP, Gillespie R, Hall JE, Hubbard S. Van Nes rotational osteotomy for treatment of proximal femoral focal deficiency and congenital short femur. J Bone Joint Surg Am. 1975;57-A:1039–46. 98. Lange DR, Schoenecker PL, Baker CL. Proximal femoral focal deficiency: treatment and classification in forty-two cases. Clin Orthop Relat Res. 1978;135:15–25. 99. Coventry MB, Johnson EW Jr. Congenital absence of the fibula. J Bone Joint Surg Am. 1952;34-A:941–55. 100. Farmer AW, Laurin CA. Congenital absence of the fibula. J Bone Joint Surg Am. 1960;42-A:1–12. 101. Epps CHJ, Schneider PL. Treatment of hemimelias of the lower extremity. Long-term results. J Bone Joint Surg Am. 1989;71:273–7. 102. Kruger LM. Recent advances in surgery of lower limb deficiencies. Clin Orthop Relat Res. 1980;148:97–105. 103. Hiroshima K, Kurata Y, Nakamura M, Ono K. Ball-and-socket ankle joint: anatomical and kinematic analysis of the hindfoot. J Pediatr Orthop. 1984;4:564–8. 104. Catagni MA, Cattaneo R, DeRosa V. Le traitement de l’hemimelia externe avec la methode d’Ilizarov. In: Dimeglio A, Caton J, Herisson C, Simon L, editors. Les inegalites de longueur des membres. Paris: Masson; 1994. p. 177–81. 105. Pappas AM, Hanawalt BJ, Anderson M. Congenital defects of the fibula. Orthop Clin North Am. 1972;3:187–99. 106. Achterman C, Kalamchi A. Congenital deficiency of the fibula. J Bone Joint Surg Br. 1979;61-B:133–7. 107. Birch JG, Lincoln TL, Mack PW, Birch CM. Congenital fibular deficiency: a review of thirty years’ experience at one institution and a proposed classification system based, on clinical deformity. J Bone Joint Surg Am. 2011;93-A:1144–51. 108. Paley D. Surgical reconstruction for fibular hemimelia. J Child Orthop. 2016;10:557–83. 109. Westin GW, Sakai DN, Wood WL. Congenital longitudinal deficiency of the fibula: follow-up of treatment by Syme amputation. J Bone Joint Surg Am. 1976;58:492–6.

110. Hootnick D, Boyd NA, Fixsen JA, Lloyd-Roberts GC. The natural history and management of congenital short tibia with dysplasia or absence of the fibula. J Bone Joint Surg Br. 1977;59:267–71. 111. Choi IH, Kumar SJ, Bowen JR. Amputation or limb-lengthening for partial or total absence of the fibula. J Bone Joint Surg Am. 1990;72:1391–9. 112. McCarthy JJ, Glancy GL, Chang FM, Eilert RE. Fibular hemimelia: comparison of outcome measurements after amputation and lengthening. J Bone Joint Surg Am. 2000;82-A:1732–5. 113. Patel M, Paley D, Herzenberg JE. Limb-lengthening versus amputation for fibular hemimelia. J Bone Joint Surg Am. 2002;84-A:317–8. 114. Rodriguez –Ramirez A, Thacker MM, Becerra LC, Riddle EC, Mackenzie WG. Limb length discrepancy and congenital limb anomalies in fibular hemimelia. J Pediatr Orthop B. 2010;19:436–40. 115. Oberc A, Sulko J. Fibular hemimelia – diagnostic management, principles, and results of treatment. J Pediatric Orthop B. 2013;22:450–6. 116. El-Sayed MM, Correll J, Pohlig K. Limb sparing reconstructive surgery and Ilizarov lengthening in fibular hemimelia of Achterman-Kalamchi type II patients. J Pediatr Orthop B. 2010;19:55–60. 117. Catagni MA, Radwan M, Lovisetti L, Guerreschi F, Elmoghazy NA. Limb lengthening and deformity correction by the Ilizarov technique in type II fibular hemimelia. An alternative to amputation. Clin Orthop Relat Res. 2011;469:1175–80. 118. Changulani M, Ali F, Mulgrew E, Day JB, Zenios M. Outcome of limb lengthening in fibular hemimelia and a functional foot. J Child Orthop. 2010;4:519–24. 119. Jones D, Barnes J, Lloyd-Roberts GC. Congenital aplasia and dysplasia of the tibia with intact fibula. Classification and management. J Bone Joint Surg Br. 1978;60:31–9. 120. Schoenecker PL, Capelli AM, Millar EA, et al. Congenital longitudinal deficiency of the tibia. J Bone Joint Surg Am. 1989;71:278–87. 121. Kalamchi A, Dawe RV. Congenital deficiency of the tibia. J Bone Joint Surg Br. 1985;67:581–4. 122. Weber M. New classification and score for tibial hemimelia. J Child Orthop. 2008;2:169–75. 123. Courvoisier A, Sailhan F, Thevenin-Lemoine C, Vialle R, Damsin JP. Congenital tibial deficiencies: treatment using the Ilizarov’s external fixator. Orthop Traumatol Surg Res. 2009;95:431–6.

7

Developmental Disorders of the Foot and Ankle

7.1

evelopment of the Foot and Ankle; D Embryologic, Fetal, and Postnatal

7.1.1 Early Development Embryological differentiation of the foot is noticeable by the fifth intrauterine week. By the end of the fifth week, condensation of the mesenchymal tissue has resulted in the formation of tarsal bone anlage (earliest developmental models). Tissue differentiation to precartilage and cartilage follows and all skeletal elements of the foot have begun to chondrify by 7 weeks with the exception of the distal phalanx of the little toe. By the end of the second month, the outline of each bone is becoming distinct [1–3]. Homogenous interzones develop at the ankle and between all tarsal, tarsal-metatarsal, metatarsal-phalangeal, and interphalangeal joints with the three-layered interzone developing earliest at the ankle and metatarsal-phalangeal joints at the end of the embryonic period (8 weeks). Cavitation with vascularized synovial tissue is seen in most joints of the foot at 9–11 weeks (early fetal period). A proximodistal sequence of development unfolds. By the end of the embryonic period (8 weeks), the foot resembles that of the adult in most details. All elements of the foot destined to become bone have begun to chondrify (except the distal phalanx of the little toe) by 7 weeks of embryonic life. Anomalies of the skeletal elements (such as tarsal coalition) form very early. The articular surfaces of the ankle joint and of the other joints of the foot reach a high degree of differentiation before the resorption phase of the interzone begins around 8 weeks of age. Hesser noted that the articular surfaces of the ankle joint reach a high degree of differentiation even before joint cavitation is complete [4]. The talonavicular and calcaneocuboid joints also reach close to their final shapes very early. Straus also stated that the joints of the foot are laid down in their definitive form by the ninth intrauterine week [5]. Cavitation of the interzone to form the synovial joints begins first at the ankle and progresses distally, forming last in the interphalangeal joints. O’Rahilly et al. further reviewed the skeletal development of the foot [6]. The prenatal development of the human foot has

been well studied and described by Gardner et al. who reviewed the extensive developmental literature as well as assessing 184 human embryos and fetuses.

7.1.2 Vascularization and Ossification of the Foot Bones Cartilage canals are seen first in the talus and calcaneus before appearing in the other tarsals. Some observers note the talus to have cartilage canals first, originating from the sinus tarsi. The calcaneus is the first tarsal to ossify. The main calcaneus center formed by endochondral bone is seen in all specimens from 6 months on. Periosteal bone formation at the inferolateral part of the calcaneus is often the first bone site but usually central endochondral bone formation is first. A posterior calcaneal apophysis forms between 6 and 10 years of age and fuses by 17 years. Intrauterine ossification of the talus via the endochondral mechanism begins in males at 7 months, about 4 weeks later than in the calcaneus. The ossification center forms near the geometric center of the talus and is referred to as a centric ossification center. In the female the centers appear a few weeks earlier than in the male. In most specimens the endochondral center joins a layer of periosteal bone in the roof of the tarsal canal. The cuboid is the next tarsal bone to ossify doing so from a single nucleus just before birth in females and just after birth in males even though it is actually the first tarsal bone to chondrify. The other tarsals ossify postnatally: lateral cuneiforms (third) in first year, medial (first) in third year, and middle (second) in fourth year and navicular in the fourth year. Metatarsals: centers for the shaft appear at the 8th–10th week (prenatal) with the 2nd and 3rd ossifying first followed in succession by the 4th and 5th, with the 1st last to ossify. The metatarsal epiphyses ossify in the 3rd to 5th years. Phalanges: shaft centers appear in fetal period – first in the distal phalanges (end of 2nd month), then in the proximal row (3rd to 4th month), and then in the middle row (4th month). Epiphyseal centers appear in 3rd to 5th years.

© Springer Nature Switzerland AG 2019 F. Shapiro, Pediatric Orthopedic Deformities, Volume 2, https://doi.org/10.1007/978-3-030-02021-7_7

665

666

The metatarsals are long bones that grow at both ends from cartilaginous physes but have epiphyses and secondary ossification centers only at one end. The epiphysis of the first metatarsal is at the proximal end, while those of the second to fifth metatarsals are distal. The phalanges also have a single epiphysis and secondary ossification center but all are proximal.

7.1.3 A dditional Observations on Ankle and Foot Embryonic and Fetal Development Joints As embryologic development proceeds, homogenous interzones become three-layered, cavities appear in the middle layers, and synovial tissue later lines the cavities. Cavitation in the larger ankle and foot joints usually begins between 6 and 7 weeks but the small joints may not show a cavity until the early fetal period. Sesamoid Bones The sesamoid bones at the plantar surface of the metatarsophalangeal joint of the big toe begin to chondrify shortly after the beginning of the fetal period. They are rarely seen at other metatarsophalangeal joints. Cartilage Canals Cartilage canals appear during the fetal period in all the skeletal elements of the foot. The appearance of cartilage canals is not an indication either of immediate following of ossification or the order of chondrification or ossification. Abnormal Fusions Abnormal fusions between foot bones seen first as cartilage continuities are common. The fusion rate between the middle and distal phalanges of the little (5th) toes is as high as 30–45% of people in some North American and European populations. Talocalcaneal and calcaneonavicular fusions (referred to as coalitions) with cartilage model continuity form early in embryogenesis. They are much less common but, when present, often have postnatal clinical significance causing rigid flatfeet with discomfort.

7.1.4 Descriptive Divisions of the Foot Regions The foot is divided for descriptive purposes into three regions. This is valuable clinically to describe sites of deformity and approaches to treatment. The hindfoot consists of the talus and calcaneus; the midfoot of the navicular, cuboid, and three cuneiforms; and the forefoot of the metatarsals and phalanges. The hindfoot is separated from the midfoot along the transverse plane referred to as Chopart’s joint encompassing the talonavicular and calcaneocuboid articula-

7 Developmental Disorders of the Foot and Ankle

tions and the midfoot is separated from the forefoot along the transverse plane referred to as Lisfranc’s joint encompassing cuneiform-metatarsal (first to third) and cuboid-metatarsal (fourth and fifth) joints (Fig. 7.1a, b). As well as the transverse plane separations listed above, some refer to a longitudinal plane separation as having clinical significance; this divides the foot along the longitudinal plane between the third and fourth metatarsals, the third cuneiform and the cuboid, and the talus and cuboid with the calcaneus associated with both planes (Fig. 7.1c). The medial longitudinal plane (or arch) component is composed of the first to third metatarsals: all three cuneiforms, navicular, talus, and calcaneus, while the lateral longitudinal plane (or arch) involves the fourth and fifth metatarsals, cuboid, and calcaneus (Fig. 7.1d). Cavus and inversion deformities can be considered to have their apex or most prominent deviation in relation to one or other of these arches. The bones of the foot and ankle as seen from medial and lateral sides are shown in Fig. 7.1e. Joint Motion Hicks has detailed the ranges of joint motion at the ankle and foot in a mechanical study of the region [7]. The axis of the ankle joint passes through the talus just above the talocalcaneal articulation. The movement is one of flexion (plantarflexion) and extension (dorsiflexion) with the respective ranges of 27° and 13°. There is slight side-to-side gliding, rotation, and adduction/abduction in dorsiflexion. The shape of the talar dome articular surface indicates two axes; when the joint is in dorsiflexion (two-thirds of its range), the slope is downward and medially, and when in plantarflexion the slope is upward and medially. Other joint complexes are the talocalcaneal-navicular, the midtarsal (Chopart’s) composed of the talonavicular and calcaneocuboid articulations, and the anterior tarsometatarsal joints. According to Hicks “all movements and changes in shape of the foot are rotations about axes.” Rotation about a transverse axis is flexion-extension, about a vertical axis abduction adduction, and about an anterior-posterior axis pronation supination. When the leg is the moving part, there is medial rotation lateral rotation. Rotations about oblique axes have compound names (e.g., pronation-abduction-flexion/supination-adduction-extension). Ankle Joint Axes The dorsiflexion axis is the one where rotation occurs when the joint is in the extension part of its range and the plantarflexion axis is the one where rotation occurs in the flexion part of its range. Midtarsal Axes The two midtarsal axes can be used simultaneously. These involve the talocalcaneal-navicular axis and the talonavicular and calcaneocuboid axes. The midfoot axis passes near the center of a sphere constituted by the head of the talus. The midtarsal axes are crucial for the side-toside foot movements (pronation, supination, abduction, adduction).

7.1 Development of the Foot and Ankle; Embryologic, Fetal, and Postnatal

a

b

Phalanx III.

667

Phalanx III. Phalanx II.

Phalanx II.

Phalanx I.

Phalanx I.

F Os metatarsale I.

Os metatarsale I.

L Os cuneiforme I.

Tuberositas ossis metatarsalis v.

Os cuneiforme II. Os cuneiforme III.

Os cuboideum

Os naviculare C

Os cuneiforme I.

Tuberositas ossis metatarsalis V.

Os cuneiforme II. Os cuneiforme III.

Os cuboideum

Os naviculare

M

Calcaneus Talus

Calcaneus

Talus H

c

d

Metatarsalia I

II

III

IV

V

Cuneiformia

I

II

III Guboideum

Naviculare

Talus

Calcaneus

Fig. 7.1 The foot can be divided into regions for descriptive purposes and to indicate localization of specific deformities. (a) Anteroposterior view of the foot highlights the forefoot (metatarsals and phalanges), midfoot (navicular, three cuneiforms, cuboid), and hindfoot (talus, calcaneus) regions. Arrows indicate Lisfranc and Chopart joints between the regions. (Reprinted with permission and modified from Praktische Anatomie, von Lanz and Wachsmuth, Berlin, Springer Verlag, 1938.) (b) The foot regions are separated in this figure: H, hindfoot; M, midfoot; and F, forefoot. The forefoot can be subdivided at the metatarsophalangeal joints (arrow) into a proximal (p) and distal (d) segment. (Reprinted with permission and modified from Praktische Anatomie, von Lanz and Wachsmuth, Berlin, Springer Verlag, 1938.) (c) Anteroposterior view of the foot highlights the medial longitudinal arch (first to third metatarsals, three cuneiforms, navicular, talus, and calcaneus) and the lateral

longitudinal arch (fourth and fifth metatarsals, cuboid, and calcaneus). The calcaneus is considered to belong to both medial and lateral arch segments. The arch is considered (anatomically and functionally) to pass from the metatarsal heads to the sole of the calcaneus, thus excluding the phalanges. (Reprinted from Rauber’s Lehrbuch der Anatomie des Menschen, volume 2, Leipzig, Verlag von Georg Thieme, 1906.) (d) Medial (top) and lateral (bottom) views of foot illustrate the longitudinal arches (curved black lines) that pass from the plantar surface of the calcaneus to the first (medial) and fifth (lateral) metatarsal heads. (Reprinted with permission and modified from Praktische Anatomie, von Lanz and Wachsmuth, Berlin, Springer Verlag, 1938.) (e) The bones of the ankle and foot are shown from medial (top) and lateral (bottom) sides. (Reprinted with permission from Praktische Anatomie, von Lanz and Wachsmuth, Berlin, Springer Verlag, 1938)

668

7 Developmental Disorders of the Foot and Ankle

e

Malleolus tibiae Sulcus mi. tibialis posterioris

Trochlea Tali

Collum Caput

Tuberc. fib. Tuberc. fib.

Os cuneiforme II

Sulcus mi. flexoris hallucis longi

Os cuneiforme I Basis Ossis metatarsi I

Tuber

Corpus

Corpus

Caput

Os scsphoideum metatarsophalangoum

Malleolus fibulae Sulcus malleolus fibulae

Processus posterioris tali

Calcaneal

Sustentaculum tali Tuberositas ossis Tuberositas metatarsi I ossis navicularis Facies malleolaris fibularis Collum

Tali

Caput Os cuboideum

Tuberculum fibulare processus posterioris tali

Os naviculare

Processus fibularis tali

Os cuneiforme III Os cuneiforme II Os cuneiforme I Caput

Tuber calcanei

Corpus Sinus tarsi Spina tuberis calcanei

Basis

Ossis metatarsi I

Processus Os fibulare Tuberositas Sulcus tendinis mi, fibularis trochlearis (sonderfall) ossis metatarsi V longi

Fig. 7.1 (continued)

Tarsal-Metatarsal Axes These anterior axes involve the five rays, defined as a toe unit anterior to the midfoot (e.g., cuneiform plus metatarsal [medial 3] or metatarsal [lateral 2]). All rays have a flexion-extension component.

in oblique directions, always produce only one of two movements, clockwise or anticlockwise rotation about the axis. Ball-and-socket joints such as the hip have planes of movement determined by the direction of forces acting across them. The unique anatomy and function of the talonavicular joint Hicks has studied the combination of movements at each is stressed. Although it appears as a ball-and-socket shape in axis and estimates the primary range at the ankle (dorsiflexion- isolation, in reality it is part of two different joint complexes plantarflexion) totaling 50°, talocalcaneal-navicular 24°, mid- and its motion remains a simple hinge movement in each tarsal oblique 22o and anteroposterior 8°, first ray 22°, and case. It participates in both the talocalcaneal-navicular comfifth ray 10° (with intermediate values between first and fifth). plex and, with the calcaneocuboid articulation, the midtarsal The foot joints are all considered to be hinge joints with complex. The navicular therefore rotates on the talus about each joint having its plane of movement predetermined by three different axes: talocalcaneal-navicular and one or other its own intrinsic structure. Different tendons, even if they run of the midtarsal axes, oblique and anteroposterior.

7.1 Development of the Foot and Ankle; Embryologic, Fetal, and Postnatal

7.1.5 A ngular Radiographic Measurements of Talus and Calcaneus Talocalcaneal Angle On anteroposterior and lateral radiographs, the normal talocalcaneal angle between the longitudinal bone axes of the two bones has been measured in several studies, but the variability between studies and even between feet in individual studies is great. The ranges of values are wide. Normal positioning on anteroposterior standing foot x-rays shows the longitudinal axis of the talus aligned with the first metatarsal and that of the calcaneus with the fifth metatarsal. If the angle is less on the anteroposterior projection, we infer a varus heel deformity with the calcaneus inverted under the talus. Equinus deformity on the lateral projection also lessens the angle between longitudinal axes of the two bones as the anterior part of the calcaneus tilts downward. The talocalcaneal angle can also be said to decrease with supination and increase with pronation. The angle is quite variable in different studies even in normal feet and in clubfeet can be so variable that it is considered by many to be of no significant help overall, except perhaps in following changes with treatment and growth

over time in individual patients. There are several reasons for the widely ranging measurements even considering biologic variability. The measurement on lateral projections is well standardized involving the angle between the plantar surface of the calcaneus and the central longitudinal axis of the talus (with the radiograph generally taken in the standing position or with dorsiflexion to neutral). Different observers however use different landmarks for the anteroposterior film that, while similar, are not exactly the same. Most measure the angle between the central long axes of the calcaneus and talus, but some use the medial border of the talus and the lateral border of the calcaneus, neither of which are necessarily straight or parallel to the central long axis. The combination, on anteroposterior views, of variable foot and radiographic projections, incomplete ossification in very young children, asymmetric talocalcaneal bone development and positioning with clubfoot (and other) deformities, and differing landmarks for measuring account for the difficulty in interpretation. The measurement axes most commonly used for the medial and lateral talocalcaneal angles are shown in Fig. 7.2a, b. The tibial-calcaneal angle is also commonly measured for pediatric age disorders (Fig. 7.2c).

a

b

Talocalcaneal

Talocalcaneal

Fig. 7.2 Many measurements have been used to quantitate foot position from plain radiographs in both anteroposterior and lateral projections. Measurements that help assess normal, clubfoot, metatarsus adductus, and congenital vertical talus positioning are the talocalcaneal angles in both projections. Since some studies measure these angles using differing landmarks for each axis, care must be taken in comparing data from various studies (see text). (a) The anteroposterior projection for measurement of the talocalcaneal angle is shown. (b) The lateral projection for measurement of the talocalcaneal angle is shown. (c) The tibial-calcaneal angle is also frequently used. (Figures (a), (b), and (c) reprinted with permission from Petterson and Ringertz, Angle Measurements of the foot/age [radiography], pages 80–82 in Measurements in Pediatric Radiology, Springer, London, 1991)

669

c

Tibiocalcaneal

670

With clubfeet or congenital vertical talus, difficulty in stabilizing the foot in uniform fashion especially in children under 3–4 years of age, the variable deformities, and asymmetric bone deposition in deformed bones all contribute to making interpretation difficult. The best results are obtained with standing radiographs in normal children or by using the contralateral normal foot for comparison in children with a unilateral clubfoot or vertical talus. Studies in infancy need to be as standardized as possible with continuation of studies throughout childhood until skeletal maturity to document normal changes with time and to assess responses to clubfoot treatment. Joseph et al. studied the talocalcaneal angles on anteroposterior and lateral radiographs in 75 normal feet, 145 clubfeet, and 15 normal fetal limbs [8]. The talocalcaneal angles were lower in clubfeet than in normal feet, but there was considerable overlap in the two groups. A decrease of the normal talocalcaneal angle by up to 10° occurs as the child grows because of the alteration of shape and relative position of the ossific nucleus of the talus with maturation. Vanderwilde et al. studied 74 normal infants and children (148 feet) from 6 to 127 months of age and established measurement profiles for the following angles: anterior talocalcaneal, calcaneus-fifth metatarsal, and talus-first metatarsal; lateral talocalcaneal, tibiocalcaneal, tibiotalar, talus-first metatarsal, and talo-horizontal angles; and on a lateral maximum dorsiflexion radiograph the talocalcaneal and tibial c alcaneal angles [9]. They also critically reviewed results from several other papers. Their data showed the anteroposterior talocalcaneal angle to gradually decrease from an approximate mean of 42° (27–56°) at 6 months to 25° (13–37°) at 10 years while the lateral angle stayed relatively the same at 40° ranging from 25° to 55°. The decrease with time in the anteroposterior plane is confirmed by other studies. Ippolito et al. documented a mean normal anteroposterior talocalcaneal angle of 23° +/− 4° at skeletal maturity [10]. Gentili et al. indicated that the anteroposterior talocalcaneal angle at skeletal maturity ranged between 17° and 21° [11]. The lateral angle ranged between 35° and 50°. Garceau and Palmer listed the mean anterior talocalcaneal angle as 22.6° [12]. This is sometimes referred to as Kite’s angle, although he never indicated its measurement. Cooper and Dietz measured the normal anteroposterior talocalcaneal angle at skeletal maturity at 23° +/− 8° and the lateral talocalcaneal at 44° +/− 9° [13] while Dobbs et al. had the anterior angle (normal, skeletal maturity) at 14° +/− 2° and the lateral at 42° +/− 2° [14]. Ippolito et al. concluded that measurement of the anteroposterior talocalcaneal angle on radiography was misleading for assessing the degree of hindfoot correction in 75% of treated congenital clubfeet [10]. CT scanning was felt to be more accurate, although not generally necessary in routine clinical practice. Talar Head-Neck Deviation In the normal fetus the talar head and neck region is more medially deviated than in the adult. Normal deviations have been documented of 35° in the

7 Developmental Disorders of the Foot and Ankle

fetus [15, 16], 30° in the newborn [17], and 10–12° [2, 16], 15° (35 specimens) [18], and 19° (100 specimens; range, 18–25°) [17] in the adult.

7.1.6 Anatomy of the Talocalcaneal Joint Knowledge of the characteristics of the subtalar (talocalcaneal) joint is important in identifying and understanding foot pathoanatomy and its treatment. There are three separate articulations that make up the talocalcaneal relationship: the posterior, middle, and anterior subtalar joints [19] (Fig. 7.3a– d). The largest articulation is the posterior subtalar joint. The middle joint is the most medially positioned and is much smaller than the posterior joint. The middle facet of the talus is oval and slightly convex and rests on the concave middle facet on the superior medial aspect of the calcaneus. The calcaneal middle facet is supported by the strong prominent projection from the medial side of the calcaneus referred to as the sustentaculum tali. The anterior joint is the smallest and is anterior and slightly lateral to the middle joint but still positioned at the anteromedial aspect of the undersurface of the talar head. The anterior calcaneal facet may or may not be continuous with the middle facet, varying from specimen to specimen. Bunning and Barnet described the three varieties of talocalcaneal facet patterns as type A, three separate facets; type B, combined anterior and middle facets with a separate posterior facet; and type C, combined anterior, middle, and posterior facets [20] (Fig. 7.3a). Shereff and Johnson identified extra-articular osseous landmarks of the two bones. On the talus, there are medial and lateral (posterior) tubercles and medial and lateral facets (more anteriorly positioned) of the body which relate to the medial tibial malleolus and lateral fibular malleolus, respectively. On the calcaneus, there is the posterior-inferior calcaneal tuberosity, the lateral calcaneofibular ligament insertion and, just inferior to it, the peroneal trochlea, and medially the sustentaculum tali and the anterior tubercle of the calcaneus. Martus et al. recently described an accessory anterolateral talus facet which was present in 34% (27/79) of pediatric skeletal specimens (average age 13.4 years) [21]. This had previously also been identified by Sewell [22] and by Sarrafian [23]. It was present adjacent to the inferior angle of the lateral calcaneus formed by the calcaneal posterior facet and the dorsolateral calcaneal neck. (They discussed its possible cause of adult flatfoot pain.)

7.1.7 Tibiofibular Torsion Tibial torsion, or more accurately tibiofibular torsion, plays a major role in determining foot position since it orients the alignment of the talus (talar dome) that moves almost exclusively in the dorsiflexion/plantarflexion plane. Early studies by

7.1 Development of the Foot and Ankle; Embryologic, Fetal, and Postnatal a

a

c

i

b

671 c

ii

Caput tail

b

Facies articularis navicularis Facies articularis calcanea anterior

Collum tali Facies articularis calcanea media Facies malleolaris medialis Facies articularis calcanea post.

Facies superior

Sulcus tali

Facies malleolaris lateralis

Processus lateralis tali Sulcus m. flexoris hallucis longi

Processus posterior tali

iii Facies articularis anterior

Facies articularis media Sulcus calcanei Sustentaculum tali Facies articularis posterior

Corpus calcanei

Fig. 7.3 The talocalcaneal joint has a unique structure. There are usually three separate articulations (facets) – anterior, middle and posterior – with the posterior the largest. In some calcanei, two articulations (usually anterior and middle) are continuous as a single mass. Occasionally, all three articulations are continuous. (a) Classification of the subtalar superior surface calcaneal facets is shown. (Reprinted with permission from Bunning and Barnett, J Anat (London) 1965; 99: 71–76, John Wiley and Sons Ltd/Anatomical Society.) (b) The superior surface of the calcaneus shows each of the three facets for articulation with the talus. a anterior; m middle (medial); p posterior. (Reprinted with permission from Uygur et al., Arch Orthop Trauma Surg 2009; 129: 909–914, Springer.) (c(i,ii,iii)) Anatomic drawings illustrate (i) the right talus from above

(left), (ii) the inferior surface of the right talus (right) and (iii) the right calcaneus from above (lower image). (Reprinted from Rauber’s Lehrbuch der Anatomie des Menschen, volume 2, Leipzig, Verlag von Georg Thieme, 1906.) (d) Drawings clarify anatomy of the subtalar joints. At left, talus and calcaneus are seen from the lateral side with the three talocalcaneal articulations (facets) indicated. At right, the talus has been removed to show a superior view of the calcaneus and its three articular facets (posterior, middle, and anterior). The navicular articular surface (for the talar head) is visible as is the calcaneonavicular ligament. (Reprinted with permission from Herzenberg et al., Foot and Ankle 1986; 6: 273–288, Sage Publications)

672

7 Developmental Disorders of the Foot and Ankle

d

Talus

Anterior facet

Navicular Plantar calcaneonavicular ligament

Anterior facet

Middle facet Posterior facet Middle facet

Posterior facet

Calcaneus

Fig. 7.3 (continued)

LeDamany on the tibia alone showed tibial torsion to be medial in the fetus, neutral at birth, and then becoming lateral during the first years of life reaching a stable mean value of 20° by age 5 years [24]. It is now accepted that the angle between the transverse axis of the proximal tibia and the distal tibiofibular transmalleolar axis is the more important rotational measurement. This can differ from the tibial measurement alone. Badelon et al. studied 100 limbs from 50 fetuses free of malformation and found the torsion in 83 cases lateral, 11 neutral, and only 6 medial [25]. The average value increased progressively from a clear lateral torsion by fetal age 5 months to a mean of 20° at the end of gestation. Hutchins et al. measured tibiofibular torsion postnatally from age 5 to age 25 years [26]. In 352 normal subjects, lateral torsion progressively increased from 10° at 5 years of age to 17° in men and 14° in women at 25 years of age. In mature clubfoot patients, the values were slightly less. In 252 treated clubfeet greater than 14 years of age, the male value was 16° and female 10°. In a small clubfoot subset with intoeing, there was still lateral torsion but the value was only 6.7°. Regardless of the methods used to assess rotation (clinical, anatomic, or multiple imaging modalities), the values vary from study to study because of the difficulty in assessing in particular the proximal tibial transverse axis and also the transmalleolar axis. Measurement of tibial torsion is shown in Fig. 7.4 and has been discussed in detail in Chap. 6.

7.2

erminology of Foot and Ankle T Deformities

7.2.1 Deformities of the Foot Alone Metatarsus Adductus Midfoot and forefoot deformity characterized by adduction in the transverse plane relative to a normally positioned hindfoot.

Metatarsus Varus The midfoot and forefoot are tilted into varus such that the sole of the foot points inward. Metatarsus Adductovarus The midfoot and forefoot are both adducted in the transverse plane and tilted into varus. Skewfoot A relatively severe deformity, sometimes referred to as an S-shaped foot, with severe midfoot and forefoot adduction and excessive and fixed valgus of the hindfoot.

7.2.2 Deformities of Both Foot and Ankle Most foot deformities also involve the ankle since they are associated with over- or underactivity of the musculotendinous units that span the ankle from leg to foot. Terminology for these deformities was defined in English by Little in the mid-1800s in his detailed writings on clubfoot and other entities [27, 28]. He used the term “talipes” derived from talus and pes (foot) as a generic expression and he grouped the deformities of the foot into four principal divisions. He described the four basic directions of deformity as equinus, varus, valgus, and calcaneus and noted that biplanar deformities were common. Adams also described the multiple varieties of foot deformities in his writings on clubfoot [29]. The most common deformity was talipes equinovarus or clubfoot. Unidirectional (uniplanar) deformities are defined as: • Talipes equinus: elevation of the heel, with plantarflexion of the foot. • Talipes varus: inversion of the foot. • Talipes valgus: eversion of the foot. • Talipes calcaneus: depression of the heel, with dorsiflexion of the foot.

7.3 Metatarsus Adductus

673

Fig. 7.4 Examples measuring for tibial torsion of the left lower extremity using magnetic resonance or computerized tomography images are shown. Thick arrow indicates the tibial torsion and direction of the foot from the top image in each set (transmalleolar axis; lateral malleolus, distal tibial articular surface, medial malleolus) as it relates to the proximal tibia and fibula in the bottom image (posterior tibial condylar axis; small round bone = fibula). Upper left is normal; upper right, increased outward tibial torsion = outtoeing; lower left, neutral torsion; and lower right, torsion inward less than neutral = in-toeing. Tibial torsion measurement with additional images has also been described in Chap. 6, Fig. 6.4a (i–iii), b

There are also four compound varieties that include: • Talipes equino-varus (TEV): the foot is inverted into varus as well as plantarflexed. This is the clubfoot deformity. • Talipes equino-valgus: the foot is everted as well as plantarflexed. • Talipes calcaneo-varus: the foot is inverted with depression of the heel. • Talipes calcaneo-valgus: the foot is everted with depression of the heel. Other deformities include: • Talipes cavus: elevation of the midfoot arch relative to both the hindfoot and forefoot.

• Talipes cavovarus: midfoot arch elevation and inversion of the anterior portion of the foot. • Talipes calcaneocavus: depression of the heel and elevation of the midfoot arch.

7.3

Metatarsus Adductus

7.3.1 Terminology and Deformity Metatarsus adductus is a foot deformity in which the forefoot is deviated medially into adduction in relation to the midfoot and hindfoot. The metatarsals and toes may be adducted only in the transverse plane, but the adduction may also be associated with variable forefoot inversion and varus. There is no equinus deformity and the heel is actually in slight valgus.

674

The inner border of the foot is concave and the lateral border is convex with the base of the fifth metatarsal prominent. Some internal tibial torsion may be present. The disorder is congenital but may not be recognized by the family for several months after birth and occasionally not until walking starts. The term metatarsus varus is also used for this deformity. Some refer to more severe variants of the deformity as skewfoot or serpentine foot and the terms metatarsus adductovarus and pes adductus have also been used. McCormick and Blount applied the term skewfoot to all variants including metatarsus adductus, metatarsus varus, metarsus adductovarus, and even metatarsus adductocavovarus [30]. The disorder was recognized as a distinct separate entity by Henke in 1863 describing both the forefoot adduction and the heel valgus [31], but it was not until the first two–three decades of the 1900s that it was better defined and more widely recognized, primarily in the German literature. Duncker (1912) [32] and Kauffmann (1929) [33] provided detailed overviews along with Madier and Massart (1923) in France [34]. Detailed awareness of the disorder came later in the United States and England with discussions by Peabody and Munro (1930) [35] and McCormick and Blount (1949) [30] prominent. The term “skew” means oblique or slanting. Kite identified 12 of 2818 feet with forefoot adduction that were resistant to treatment and subclassified them as serpentine metatarsus adductus [36]. Kite assessed his extensive group showing an equal male/female distribution with 45% unilateral and 55% bilateral. The position of the forefoot in either the adducted transverse or the adducted inverted plane initially represents a dynamic variant where muscle pull or the presence or absence of weight-bearing appear to predispose to one plane or the other (Fig. 7.5a). Often the transverse adducted position is seen with weight bearing and the inverted plane with nonweight bearing. When some rigidity develops in the uncorrected metatarsus adductus foot, one plane or the other is favored. The heel in metatarsus adductus is in valgus and the heel cord (tendo Achilles) is not tight. The standing position or active inversion of the foot may worsen the adduction component of the medial foot with overactivity of the abductor hallucis that also preferentially increases the space between the big toe and the second toe. Internal tibial torsion may be present. The deformity is initially flexible and the child can actively correct the foot with eversion or the examiner can readily straighten and even overcorrect the adducted foot position with gentle manipulation. With time, usually beyond 9–12 months of age, a persisting deformity can become rigid and bone growth of the adducted metatarsals and of the tarsalmetatarsal joints may increasingly conform to the deformed position. This initially is an adducted orientation of the metatarsals whose long axes are straight although more severe cases not treated or detected for several years may actually develop curvature of the metatarsal shafts. Peabody and Muro

7 Developmental Disorders of the Foot and Ankle

felt that more severe variants resistant to correction were due to more distal insertion of the tibialis anterior tendon at the base of the first metatarsal [35]. The hindfoot valgus may be seen radiographically with an increased talocalcaneal angle on anteroposterior foot radiographs to greater than 35°. Early references to the entity described adductus as simply an adduction deviation of the metatarsus on the tarsus, reserving the varus terminology for the more deformed variants where the cuneiform bones rotate in a varus direction almost overlapping one another and carrying the metatarsals along the same plane. The torsion of the forefoot makes the patient walk on the outer border of the foot with the sole pointing inward. The much more common adduction only variant is milder, has all metatarsals bearing weight, without varus tilt, and is characterized by a fairly sharp medial angulation with apex at the medial cuneiform [30, 35].

7.3.2 Developmental Pattern of Deformity The deformity is commonly seen in the newborn and is widely attributed to positioning in utero. Many cases correct spontaneously in the weeks following birth with some observers indicating that as many as 85–90% of cases resolve without intervention. In the series of Ponseti and Becker, 90% are spontaneously corrected [37]. Of those needing serial casts, the average age at start of treatment was 6 months and the average number of casts per foot needed for correction was 4. Rushforth found that 86% of 130 affected feet in 83 children developed normal or only mildly deformed feet without treatment and all were fully mobile, 10% remained moderately deformed but asymptomatic, and only 4% remained deformed and stiff [38]. The resistant cases only became apparent after 3 years of age. Expectant treatment only with close follow-up of the children was recommended.

7.3.3 Non-operative Treatment If the disorder has not resolved or is not resolving by 4–6 months of age, many feel that treatment is warranted. Some physicians will begin treatment earlier. Application of corrective forces with the heel held into mild varus (supinated) can be maintained by (i) an orthotic support designed specifically to stabilize the hindfoot and abduct the forefoot or (ii) serial short- or long leg casting. A reverse or straight last shoe has been used in the past but is not recommended now since the approach is nonphysiologic and fails to control heel position. A short leg corrective cast is applied with the heel stabilized and held into varus as well as some inversion (supination), while the forefoot is abducted but not excessively everted or pronated with counterpressure placed laterally at

7.3 Metatarsus Adductus

675

a

b

Fig. 7.5 This figure illustrates aspects of metatarsus adductus. (a) At right side of the photograph the forefoot is positioned into adduction in the transverse plane; at left the forefoot moves into the adducted position primarily with an inversion tilt. (b) Surgical correction is infrequently needed for metatarsus adductus. In cases resistant to cast correction, procedures such as releasing the abductor hallucis, releasing the ligaments and capsule of the first cuneometatarsal joint medially, superiorly, and inferiorly (leaving the lateral portions intact), and performing osteotomies of the base of the second, third, and fourth metatarsals have been useful. In the example shown, the first ray is stabilized with a K-wire from the tip of the big toe through the phalanges, first metatarsal, navicular, and into the talus. (Reprinted with permission from Cahuzac et al., J Pediatr Orthop B 1993; 2: 176–181, Wolters Kluwer Health Inc.) (c) The anteroposterior radiographic profile of a

normal foot (left), metatarsus adductus foot (middle), and clubfoot (right) are shown. The normal separation of the long axes of the talus and calcaneus is 30°. In metatarsus adductus, this angle is increased to 50° (in this example) since the calcaneus in metatarsus adductus is actually in valgus. In a clubfoot the talocalcaneal angle is decreased (to 5° in this example) as the calcaneus is inverted under the talus. [The arrows were placed to indicate the pressure points for molding the corrective casts; that for metatarsus adductus remains correct since efforts are made to actually invert the heel but pressure is no longer applied to the outer calcaneus and cuboid with clubfoot correction since that is now recognized as preventing the calcaneus and cuboid from moving into the corrected position with forefoot manipulation.] (Reprinted with permission from Ponseti and Becker, J Bone Joint Surg Am 1966; 48A: 702–711, Wolters Kluwer Health Inc.)

676

7 Developmental Disorders of the Foot and Ankle 5˚

c 25˚ 50˚

60˚

30˚

Fig. 7.5 (continued)

the cuboid and the base of the fifth metarsal. The abduction pressure is placed medially on the metatarsals and not on the phalanges. Casting is done in an outpatient setting without need for anesthesia. The ankle is positioned at neutral with no dorsiflexion or plantarflexion since there is no heel cord tightness. After the short leg cast has hardened, it can be converted to a long leg cast with the knee flexed about 30–45° and the leg externally rotated to encourage growth correction by further stabilizing the ankle. Each cast is worn for 2 weeks prior to change. Once correction has been achieved, daily or twice-daily active eversion exercises and maintenance of position with orthotics are helpful to allow for continuing forefoot remodeling. It is best to aim for correction by the age of walking. Patients are occasionally seen with persisting metatarsus adductus after walking has started. The orthotic method listed above for the non-ambulatory age group can be tried if not previously used, but serial casting in this age group is favored.

7.3.4 Surgical Treatment Surgical correction may be needed in a very small percentage of patients if deformity persists in a child who walks. Several approaches have proven to be effective.

Release/Lengthening of Abductor Hallucis Tendon At 1–2 years of age, one can lengthen the abductor hallucis muscle intramuscularly or divide the tendon and release it from its capsular attachment and cast the foot in the corrected straight position for 3–4 weeks [39]. Contraction or shortening of the abductor hallucis muscle/tendon is a primary deforming factor especially in resistant cases where the adduction is combined with a degree of supination (varus) [36, 40]. Medial capsulotomy (of the 1st metatarsal- phalangeal joint) has been added effectively to abductor hallucis lengthening for correction of severe cases [41]. Tarsal-Metatarsal Capsulotomies By 2 years of age, the deformity is either sufficiently rigid or the cartilage and bone models have grown in angulated fashion such that more extensive surgical intervention is needed. Between the ages of 2–5 years, tarsal-metatarsal releases (capsulotomies) can be done as described by Heyman et al. [42]. A subsequent report detailed their longer-term results [43]: in 80 feet with metatarsus adductus, there were 92% with good or excellent results, and in 14 feet with clubfeet with persisting adduction of the forefoot, there were 88% good or excellent results. Minimal dissection, avoidance of damage to articular surfaces, division of only the intermetatarsal and tarsometatarsal ligaments, and maintenance of the lateral capsule at the

7.4 Clubfoot

fifth metatarsal-cuboid joint are needed. All dorsal capsules are released along with the medial two-thirds of the plantar capsules. The time frame for surgery was strictly defined as between 3 and 8 years of age. The average age at surgery was 5.5 years. Several months of cast immobilization followed by orthotic splinting are needed to allow for continuing metatarsal growth along the normal planes. Concern about possible development of early degenerative arthritis with this surgical method has been expressed by some since the reoriented metatarsals don’t fit anatomically into the tarsal components especially at the proximal second metatarsal which is recessed into the adjacent tarsals and is not in the same plane as the adjacent proximal first and third metatarsals. In a study by Stark et al. of 48 tarsometatarsal capsulotomies done for both idiopathic metatarsus adductus and residual clubfoot adductus, there was a very high 41% failure rate with 63% having difficulty with proper shoe fitting [44]. Proximal Metatarsal Osteotomies Proximal metatarsal osteotomies of all five bones have been done to circumvent this concern. Correction of the midfoot deformity is brought about by careful abduction of the forefoot post-osteotomies, making certain that there is no lateral metatarsal translation in the process. Once correction of the adduction has been achieved, stabilizing K-wire or wires must be used to prevent lateral metatarsal translation during the healing process by specifically stabilizing the fifth ray. The foot is then held in the correct position by an appropriately molded short leg cast while healing occurs. Good results with this approach have been reported by Berman and Gartland [45] and Steytler and van der Walt [46]. Tendon Insertion Releases/Transfers (Tibialis Anterior, Tibialis Posterior) Tönnis has mentioned the occasional need for proximal transfer of the insertion of the tibialis anterior tendon to the dorsum of the first cuneiform due to finding the insertion more distal than usual (in metatarsus adductus) at the base of the first metatarsal [47]. He provides an excellent overview of the disorder including all management variables. In resistant cases eventually explored surgically by Browne and Paton, an anomalous insertion of the tibialis posterior tendon was seen in 14 of 15 cases where the insertion bypassed the medial navicular and inserted into the medial cuneiform and probably the adjacent metatarsal. Repair was by transferring the tendon insertion to the navicular bone [48]. Ghali et al. reported on dorsal, medial, and plantar soft tissue releases along the first ray of the naviculocuneiform and first metatarsal-cuneiform joint capsules and ligaments along with the deeper band of the tibialis anterior tendon attachment adjacent to those joints followed by manipulation into abduction and postoperative casting [49]. Tönnis has also described atypical insertion sites for the

677

t ibialis posterior tendon and treatment by transfer of a part of the tendon to the navicular [47], similar to the Browne and Paton approach [48]. Combined Soft Tissue Release and Second, Third, Fourth Metatarsal Osteotomies (Cahuzac Procedure) Cahuzac et al. reported a combined procedure with partial capsulotomy of the first metatarsal-cuneiform joint, sectioning of the abductor hallucis brevis muscle and a dome-shaped osteotomy of the proximal metaphysis of the second, third, and fourth metatarsals [50] (Fig. 7.5b). Release of the insertion of the tibialis anterior tendon was also sometimes added. The procedure was used in 31 cases with good results in 27. Knorr et al. modified the procedure using a minimally invasive percutaneous approach aided by intraoperative fluoroscopy making two 2 mm portals to perform the metatarsal osteotomies with a 2 mm surgical burr, sectioning of the metatarsal-cuneiform joint capsule and the distal tibialis anterior insertion, and K-wire fixation after manipulation. Casting was used for 6 weeks [51]. Thirty-four cases of metatarsus adductus (most of which were uncorrected components of clubfeet) were all graded severe preoperatively with correction leaving all radiologic parameters normalized. The management profile for metatarsus adductus is outlined in Table 7.1. Figure 7.5c illustrates radiographic differences on anteroposterior radiographs in normal, metatarsus adductus and clubfoot.

7.4

Clubfoot

7.4.1 Terminology Congenital clubfoot is a rigid deformity present at birth characterized by ankle equinus, heel tilt into varus, and midfoot and forefoot adduction with varus tilt. There is usually additional equinus at the midfoot level referred to by some as cavus and associated underdevelopment of the calf musculature. The calf muscle underdevelopment usually persists into adulthood even with full correction and function of the former clubfoot; this indicates it to be a part of the pathologic process rather than a secondary disuse phenomenon. The clubfoot deformity is often referred to as talipes equinovarus (TEV). Passive correction of a clubfoot deformity in the newborn is not possible. Clinical examples of the newborn clubfoot are shown in Fig. 7.6a–e.

7.4.2 Incidence The clubfoot deformity is common. It varies in occurrence from 0.5 to 1.25 per 1000 births. A recent study indicated

678

7 Developmental Disorders of the Foot and Ankle

Table 7.1 Outline of management approach to metatarsus adductus →→ Shortly after birth to 6–8 months of life: approximately 90% of cases of metatarsus adductus will resolve spontaneously →→Toward the end of the first year and prior to the onset of walking: if resolution is slow conservative measures can be taken such as finger stimulation of the outer sole of the foot to encourage active foot eversion (2–3 times per day for a few minutes); passive stretching of the forefoot into abduction while stabilizing the hindfoot (2–3 times a day for a few minutes); and use of a nighttime soft abduction shoe/ boot designed to correct metatarsus adductus →→ After walking has begun and for several months: continuation of conservative measures →→ 18 months–2.5 years of age: if still present serial short leg/long leg cast correction is warranted →→3 years plus: graded surgical intervention depending on severity and rigidity of disorder with short leg/long leg casting post-surgery (i) Release of abductor hallucis muscle/tendon (if foot still passively correctable) (ii) Add first metatarsal-first cuneiform joint capsular release (dorsal, medial, volar) (iii) Add proximal second, third, fourth metatarsal osteotomies to aid correction →→4–5 years plus, especially with metatarsal bone curvatures: first MC joint capsulotomy plus second–fifth proximal metatarsal osteotomies with K-wire stabilization to prevent lateral displacement with subsequent repositioning, especially stabilizing the fifth metatarsal osteotomy site. Can perform proximal metatarsal osteotomies on all five metatarsals with care taken not to damage first metatarsal growth plate that is at proximal end →→Note: metatarsal-cuneiform capsulotomies at first to third joints and metatarsal-cuboid capsulotomies at fourth and fifth joints can give good correction, but problems with discomfort have been identified after several years since the post-release articulations are not anatomic, especially at the recessed second metatarsal-cuneiform joint. If done, they should be limited to the very young around 3 years of age to allow time for remodeling with growth →→Proximal transfer of insertion of tibialis anterior or tibialis posterior tendons, if these are found to be prominent causative factors by having considerable insertions onto the first metatarsal. Proximal tendon insertion transfer of tibialis anterior to 1st cuneiform and tibialis posterior to navicular can be done This discussion refers to isolated metatarsus adductus, not the residual forefoot adduction of a partially corrected clubfoot. Many of these procedures however, especially the metatarsal osteotomies, have been used successfully to correct residual clubfeet

clubfoot incidence per year in the United States from 1996 to 2006 averaged 0.53 cases per one thousand births [52]. Many studies comment on a 1–1.25 case per one thousand live births range [53, 54]. Older studies from relatively isolated societies show even larger frequencies [55, 56]. It can occur with other deformities; in one study from Australia, the incidence of isolated idiopathic TEV was 1.25 per 1000 births and that of TEV with other birth defects was 0.9/1000 [56]. There is a male predominance with most series in the 2:1 range. Approximately 50% of cases are bilateral [52].

7.4.3 Etiology 7.4.3.1 General Overview of Causation Genetic factors in causation appear to be present; the incidence of the deformity among first-degree relatives is 20–30 times higher than that in the normal population (2.9% to 0.1%) [53]. The disorder however does not adhere to a classic recessive or dominant pattern. A specific cause of the deformity has not been identified. The term idiopathic is used to refer to clubfoot where no other congenital abnormalities are seen. The large majority of patients have idiopathic or isolated clubfeet. The deformity has been considered to occur secondary to tarsal developmental anomalies or arrests [57, 58], subclinical neuromuscular abnormalities with muscular imbalances [59, 60], tendinous malpositions [61], or extrinsic mechanical pressures [62, 63]. Some have considered clubfoot to be a composite deformity involving all

tissues of the foot [64, 65]. The clubfoot deformity may be the end-stage result of a complexity of factors with interaction of abnormal environmental and hereditary factors either of a multifactorial sequence or of individual, differing initiating factors [53, 66]. It is highly unlikely that all cases of clubfoot have the same, single cause. Etiology can be considered theoretically to have a considerable bearing on the choice and effectiveness of therapy. If a cartilage anlage deformity has occurred, whether primarily or secondarily, cast correction with or without Achilles tenotomy and even open reduction of the talo-calcaneonavicular articulation both reposition the navicular onto an abnormal talus. Although this goes a long way toward improving the situation, it might lead to an imperfect result; the navicular would still be relating to an imperfectly shaped talus and with further growth recurrent deformity would be likely. This observation is one reason explaining the need for continuing splinting post-correction for several months to as long as 2–3 years with such devices as the Denis Browne bar, night casting, or an orthosis to allow the talonavicular joint time to undergo bone and cartilage remodeling. The navicular and the rest of the foot are placed into adduction, varus, and equinus deformity not due to dislocation but by the necessity for the navicular to relate to a deformed and malpositioned talar head. If the abnormality in certain clubfeet is secondary to extrinsic causes and a true, primary dislocation has occurred (subluxation would seem a more appropriate term), the possibility of remodeling following repositioning of the talo-calcaneonavicular articulation by closed or surgical

7.4 Clubfoot

a

c

679

Hindfoot

b

d

e

Fig. 7.6 Examples of clubfeet are shown. (a) The four major deformities with a newborn clubfoot are shown: (a) equinus (left) and equinus, varus and adductus (right). (Reprinted with permission and modifications from Goriainov et al., J Child Orthop 2010; 4: 439–444, Springer). Additional views show (b) adduction; (c) varus; (d) equinus and adduction from posterior view, normal on left; and (e) equinus from posterior

view with black lines marking varus/adduction angulation and arrows showing posterior heel crease (black) and medial forefoot/midfoot crease (gray). (Figures (b) and (c) reprinted with permission from Steinman et al., J Bone Joint Surg Am 2009; 91 Suppl 2 (Part 2): 299– 312, Wolters Kluwer Health Inc.)

680

means is greater. One belief regarding the cause of idiopathic clubfoot is that it is due to a developmental abnormality in early intrauterine development of the talus with imperfect formation of the talar model before 7 weeks. All subsequent changes can be considered secondary to that event. Another school of thought considers the soft tissue ligamentous contractures and/or relative muscle imbalance to be primary with bone subluxation and talar deformity secondary.

7.4.3.2 More Detailed Assessment of Etiologic Considerations Developmental Delay (Arrested Fetal Development) This view was expressed as early as 1833 by Walther and was subsequently supported by Bohm but is not currently accepted since the angle of talar head-neck deviation from the long axis of the body of the talus in clubfoot is greater than that occurring in the normal fetus (which progressively decreases with time). Also, the navicular is never as medially displaced in normal intrauterine development as it is in a clubfoot. Intrinsic Intrauterine Pressure The intrinsic pressure acts against the developing foot. The deformity occurs too early to attribute it to intrauterine pressure. This theory could still be invoked in those cases that correct quickly without evidence of talar head-neck deformity. Browne remained a strong proponent of this cause [66]. Nerve or Muscle (Neuromuscular) Abnormalities Neuromuscular abnormalities alter musculotendinous position and pressure on the tarsal bones relative to one another. These were considered to be unappreciated abnormalities of either the central nervous system or the peripheral neuromuscular system. This theory has been prominent beginning as early as the 1830s in the works by Little in England [27, 28] and Guérin in France [67]. Either excess neuromuscular activity (spasm) or diminished activity (paralysis or paresis) can be invoked. This can also be considered as transient during crucial stages of intrauterine development or sufficiently subclinical to be undetectable postnatally. This view of causation is an extension of the well-recognized fact that patients with evident central cerebral, peripheral neuropathic, or myopathic abnormalities have a relatively high incidence of newborn or postnatal onset equinovarus (clubfoot) deformity.

Abnormal Insertion of Tendons The tendon insertions appear normal in dissected specimens so this possibility is widely discounted today. Intrauterine Environmental Causes A strong positive correlation with maternal smoking has been noted. Clubfoot

7 Developmental Disorders of the Foot and Ankle

risk was 1.5× increased for light smokers during pregnancy to 3.9× for heaviest smokers [68]. Another study showed a high correlation of interaction favoring a clubfoot in a child whose mother smoked in the first trimester and who had a close family member with clubfoot [69]. Heritability of Clubfoot Multiple studies over several decades from several countries support the concept that there is a genetic predisposition to clubfeet that is multifactorial with strong evidence that both genetic and environmental factors are involved. Multiple comments on clubfoot inheritance from the Online Mendelian Inheritance in Man gene database (OMIM # 119800) show similar conclusions and include “autosomal dominant condition with incomplete penetrance”; “recurrence risk of about 10% and probable dominant inheritance with about 40% penetrance”; “risk of recurrence in subsequently born children in Caucasians between 3 and 8% if one child is affected and about 10% if one child and one parent are affected”; “if the index patient was female the chance of subsequent children being affected was 4%”; and “the best genetic model for clubfoot………..is a single dominant gene with a penetrance of 33%” [70]. Idelberger (1939) showed genetic and environmental factors in a large twins study [71]. Engell et al. (2014) assessing clubfeet in 46,418 twin individuals born in Denmark from 1931 to 1982 could fit their data into a model with genetic and environmental effects [72]. Wynne-Davies (1965) reported multifactorial, genetic, and environmental influences based on finding overall incidence in parents and siblings of patients with clubfeet in 12/560 (2.14%) that was about 17 times the population incidence of 0.124% [53]. Gene and Molecular Abnormalities Gene and molecular studies are beginning to define abnormalities associated with clubfeet [73]. Gurnett et al. identified a single missense mutation in the PITX1 bicoid homeodomain transcription factor critical for hind limb development in a family with multiple cases of clubfoot and other lower limb malformations [73]. Alvarado et al. subsequently bred a Pitx1 knockout mouse and observed a clubfoot-like finding in several newborns [74].

7.4.4 Types of Clubfeet The clubfoot entity has suffered in terms of clarity of understanding and specification of treatment by a poor body of knowledge concerning its cause and the actual deformities of the cartilage models of the tarsal bones and their three- dimensional relationships. Since the clinical appearance of the foot is the same, the tendency has been to simply bypass etiologic and pathoanatomic concerns. Efforts at categoriza-

7.4 Clubfoot

tion however do define at least four broad categories of clubfoot.

7.4.4.1 Idiopathic Clubfoot The child has an isolated unilateral or bilateral deformity but is otherwise normal. There is no clinical or pathophysiologic evidence of an underlying neuromuscular disorder nor does any develop or become apparent with time. 7.4.4.2 Newborn Neuromuscular Clubfoot Some cases of newborn clubfoot are clearly associated with underlying neuromuscular causes present in utero such as myelomeningocele or congenital myopathy. 7.4.4.3 Clubfoot with Arthrogryposis Arthrogryposis refers to multiple joint contractures present at birth. Workup in some cases shows an underlying neuromuscular disorder, but the absence of such a finding even with detailed biopsy studies characterizes many cases. At least one other joint should be affected with contractures for a diagnosis of arthrogryposis to be given although some refer to isolated rigid clubfeet that are nonresponsive to early treatment as cases of “localized” arthrogryposis. 7.4.4.4 Syndromal Clubfoot Congenital clubfoot is fairly common in genetic disorders characterized by multiple connective tissue abnormalities, such as diastrophic dysplasia. 7.4.4.5 Postnatal Neuromuscular Clubfoot A clubfoot deformity that develops after the newborn period is almost always due to an underlying neurogenic or myopathic cause. Deformity results from muscle imbalance where the plantarflexors (gastrocsoleus) and invertors (tibialis anterior and tibialis posterior) are stronger than the dorsiflexors (extensor hallucis and digitorum longus) and evertors (peroneus longus and brevis). It is most commonly seen with spasticity disorders of cerebral palsy or moderate to severe peripheral neuropathies (CMT or HMSN type I) but also occurs with myopathies, diastematomyelia, and tethered cord syndromes.

7.4.5 O verview of Responses to Therapy and Possible Relationship to Etiology Most cases of congenital clubfoot are considered to be idiopathic – isolated deformities with no known cause. The consequence of this clinically has been that few further investigations regarding cause of the deformity or specific structural abnormalities of the deformed foot are performed in individual cases. Treatment is carried out and a repetitive

681

scenario occurs. In virtually all treatment assessments over a century and a half, a few newborn feet respond quickly to manipulation and some form of immobilization, approximately two-thirds of other patients do well with varying types of treatment, and some 15–20% are recalcitrant to varying types of treatment, either not responding or recurring soon after apparent correction. It is possible to consider cases of clubfeet that respond well are due to different causes than those that do not; whereas there is a tendency to consider some treatments as superior/inferior to others or performed in correct/incorrect ways. A relatively small number of clubfeet respond well, over a brief few weeks, to manipulation and serial casting. One can estimate this from reports at around 20%. Approximately 60% or more respond reasonably well to more prolonged casting with a surgical intervention limited to a posterior release varying from a percutaneous Achilles tenotomy to an open Z-lengthening with or without posterior capsulotomy of the tibiotalar and talocalcaneal joints. Approximately 20% (or a range of from 10% to 25%) of patients then respond with delayed correction, partial correction, or recurrence requiring repeated casting or even fairly extensive surgery. It is in these latter cases that considerable uncertainty persists about quality of management. Some feel that closed or manipulative/casting treatment has not been skillfully or appropriately done and others feel that surgical intervention has been delayed too long or imperfectly performed. Missing from this dialogue throughout the almost 200 years of documentation of clubfoot management has been the possibility that those cases which respond poorly may have a different etiology and/or a more serious underlying structural abnormality than those that respond well. Until such time as causes of clubfoot are better understood on a case-by-case basis, this dichotomy appears unlikely to change. Little progress has been made in defining the etiology of congenital clubfoot in particular in relation to a specific individual case. While many patients with known neuromuscular disorders such as cerebral palsy, myelomeningocele, and the peripheral neuropathies develop clubfeet in the early years of life, the large majority of newborn clubfoot patients are considered to be otherwise normal. Motor strength of the involved foot remains strong even with the persistent calf atrophy that characterizes the deformity and sensation is intact. There appear to be either different etiologies or different intrauterine times of occurrence (or both) with the clubfoot entity. This statement is based on two findings – one clinical and one pathoanatomic. Clinically, some feet which appear at birth to be typical cases respond to manipulation/casting therapy within a few weeks to a few months with an excellent result which persists with time. These cases go a long way in explaining why various types

682

of non-operative therapy will initially show a good number of excellent results. It seems reasonable to conclude that those feet that do not respond well early on have a more resistant variant of the disorder warranting more detailed investigation.

7 Developmental Disorders of the Foot and Ankle

internally….It, therefore, leaves uncovered on the outer margin of the foot, a position of the articular surface of the anterior tuberosity of the os calcis with which in the healthy state the os cuboides is in exact continuity.” Furthermore, “the three cuneiform bones, the metatarsal, and those of the toes necessarily follow the diseased turning round their axis of the naviculare, cuboides, and os calcis; and, there7.5 Pathogenesis and Pathoanatomy fore, it follows of mechanical necessity that the toes of the congenital club foot do not rest horizontally on the ground, of Congenital Clubfoot but in a line nearly vertical to it.” 7.5.1 Scarpa Scarpa goes on to describe the subtalar deformity of the calcaneus with its plantarflexed position, posterior deviation Scarpa (1752–1832), an Italian anatomist and surgeon, pro- toward the fibula, and its body twisted under the sole of the vided the first detailed description of the pathoanatomy of foot. Calcaneal deformity is also due to turning around its congenital clubfeet and its treatment in 1803 [75]. He con- smaller axis. sidered the talus, and other tarsal bones, to be of normal The talus is the least malformed and malpositioned at the shape and attributed the correctability of infantile clubfoot tarsal bones (of both the ankle joint and even distally where to this fact. He identified the key pathology as talocalcaneal- it is “very little or almost not at all inclined internally or navicular dislocation. Scarpa observed that even in severe towards the internal malleolus…..” “The prominence on the deformity with clubfeet, the astragalus (talus) “is very little dorsum of the foot does not depend on the wrong position of or almost not at all inclined internally, or toward the inter- the articular head of the astragalus but on the morbid rotation nal malleolus, and that its anterior tuberosity, or the articu- of the navicular round its smaller axis.” Correction therefore lar head of the bone, which ----- is received into the os is not due to repositioning the articular head of the talus but, naviculare, retains very nearly its natural direction and rather, “because the naviculare is brought back to cover the position with the tibia and the malleoli.” Scarpa’s assess- articular head of the astragalus….” [Subsequent study has ments of clubfeet involved both anatomic dissections and shown that some clubfeet, but not all, show medial deviation clinical treatment observations. In his cases he did not of the head and neck of the talus on pathoanatomic studies.] appreciate medial and plantar angulation of the talar head Scarpa was credited by Little as the first to attempt “the sciand neck (possibly because they are not present in all entific adaptation of an apparatus to the anatomical condition cases). The relationship of the body of the talus and proxi- of the bones of the distorted foot” [76]. He also developed a mal articular surface at the ankle relating to the tibia and corrective shoe used in some regions in Europe into the fibula is little disturbed. He clearly recognized the marked twentieth century. change in position of the navicular (os naviculare), cuboid (cuboids), and calcaneus (os calcis) in relation to the head of the talus. “On dissecting the feet of such unfortunate 7.5.2 Little children, we find that the bones of the tarsus are not, properly speaking, dislocated but merely removed in part from Little, in agreement with Scarpa, felt that malformation of their contact, and twisted around their smaller axis. This the bones of the ankle and tarsus was not the cause of the displacement and twisting around the smaller axis are more clubfoot disorder. His examination of more than 30 pathoremarkable in the os naviculare, the cuboids, and os calcis, logic specimens confirmed the opinion that “of the several and much less in the astragalus (talus) without however bones of the foot the astragalus is the least displaced.” He either of these bones quitting entirely the cavity or acetabu- observed in cases examined anatomically, however great the lum in which they are contained. The os naviculare, “the clubfoot deformity, “some portion of each of the articular deep elliptical cavity of which receives the smooth articular surfaces of the trochlea of the astragalus (talus) was in conhead of the astragalus, is found turned around its smaller tact with an equal proportion of the three articular surfaces axis so that its apex or internal tuberosity ….is turned presented by the tibia and fibula in the ankle-joint and that obliquely upwards towards the internal malleolus (of the the essential peculiar characteristic of varus was the dragtibia).” “The os cuboides likewise is found turned round its ging inwardly of the navicular and of the remaining tarsal smaller axis in the direction from the upper part of the foot bones away from the astragalus and os calcis, by the conby the outer side to the sole of the foot. At the place where tracted anterior and posterior tibial muscles aided by the long the cuboides is in contact with the anterior tuberosity of the flexor of the great toe.” Furthermore, the os calcis is drawn os calcis, it makes an angle obtuse externally and acute upward; the tibia articular facets of the astragalus and its

7.5 Pathogenesis and Pathoanatomy of Congenital Clubfoot

round head are exposed upon the dorsum of the foot; but the scaphoid (navicular), cuboid, cuneiform, and metatarsal bones are not merely drawn toward the sole but also inward and upward, so that the innermost point of the navicular bone touches the internal malleolus.” The tendon of the tibialis anticus (anterior) is deflected so internally that its activity further increases the deformity. Little was of the opinion that congenital distortions which appeared so similar to non-congenital acquired deformities of the neurologic system had essentially a similar cause [77]. The phenomena of spasm (excess muscle activity) and paralysis (limited muscle activity) destroyed the equilibrium in the power of opposing sets of muscles leading to deformation of the parts on which they acted. As far as clubfoot was concerned, he considered its origin “from a lesion of the nervous system” [78]. He felt that the variable severity of the clubfoot deformity depended on “the earliness of the period of intra-uterine existence at which the deformity arises”; the later the time the fetus is affected, the less severe the deformity. This arrangement still cannot be definitively discounted, although one might need to consider a transient neurologic event in the absence of any noted structural problems. This led to his assertion that “the muscles are the parts primarily involved, and that the displacement of the bones is entirely secondary.”

7.5.3 Guérin Guérin of Paris considered clubfoot to be the result “of inequality in the antagonizing muscular forces, and of the permanent retraction of certain muscles” [67]. Guérin was an early and apparently overly zealous proponent of the use of tenotomy to correct childhood deformities. He did however outline his theory of deformity development in clubfoot and throughout the body, which revealed some good understanding. He articulated the belief, held by many, that congenital distortions of the trunk and limbs were the result of muscular contraction originally induced by an affection of the nervous center (brain) or its branches. He termed the original affection due to the spasmodic action of the muscle as “a contraction.” Its persistence led to the deformity becoming permanent when it was referred to as a “retraction” with the partial or complete change of muscle into a fibrous tissue. As further explained by Bigelow [79], “the duration of the state of simple contraction is indefinite; and during this period the soft parts may be elongated by proper means. But the fibrous change is attended with rigidity, unyielding in proportion to the extent of the transformation.” Most cases of clubfoot dated either from the fetal period or from some convulsive affection of early life. Their leading, distinctive feature is a tenseness of certain tendons, made evident subcutaneously

683

when an attempt is made to correct the deviation. They are then rigid and prominent and clearly interfere with the normal position of the limb. Retracted muscles are pale, atrophied, and partially converted into fibrous tissue. Permanent contraction, according to Guérin, affected the muscle two ways: (i) the muscle became shorter and assumed a straight line between points of insertion and (ii) the transformation of muscles was fatty (if they were left to themselves) or fibrous (when they were subjected to exaggerated traction). Also “simple contraction permits us to hope for the immediate elongation of the muscles by means proper to effect it—extension, kneading (massage), frictions, (etc.)—while veritable retraction, or shortening with fibrous degeneration, implies either the impossibility of the return of the muscle to a normal length or a sufficient mechanical elongation, and demands in consequence the aid of a cutting instrument.”

7.5.4 Adams Adams was not only an orthopedic surgeon specializing in childhood disorders but also a pathologist with specific interest in pathoanatomy of orthopedic deformity. Since the work of Adams, postnatal dissections of non-treated clubfeet have shown an invariable medial positioning of the navicular in relation to the head of the talus. Figure 7.7a–i illustrates pathoanatomic changes in the talus and in the positioning of the adjacent foot bones in the clubfoot disorder. Adams observed the dissections of numerous fetuses in different museums and collections and performed postmortem examinations in several children who had died of infantile diseases while under treatment for the deformity at his hospital. Adams dissected several clubfeet, including newborns, observing talar abnormalities in all. He compared infantile tali with those in uncorrected adult clubfeet and found the same deformity. His description of the morbid anatomy of talipes varus in the infant includes the following observations [29]: • Astragalus (talus). This is found at birth to have considerable deviations of both position and shape. It is tilted obliquely forward and downward, and the greatest alteration is of the head and neck which are deviated medially and plantarward. The articular surface of the head articulates with the medially displaced navicular. Even the body of the talus at the ankle joint is slightly altered at the superior articular surfaces being obliquely positioned into equinus. The anterior part is displaced from the joint onto the dorsum of the foot and the posterior part is wedge shaped. • Os calcis. This is altered in position to an extreme degree but its deviation in form (shape) is slight. It is drawn into

684

7 Developmental Disorders of the Foot and Ankle

a

b

c

256 432

1

e 307

1

378

186 345

**

90 *

210

th

OC

OC

Normal talus

d

Club foot talus

eov

th

Normal talus

Clubfoot talus

Normal talus

Club foot talus

f Superior Normal talus

Anterior Club foot talus

Normal talus

Fig. 7.7 Histologic assessments are shown in Fig. 7.7a–f in normal and abnormal tali from a 9-day-old child with a unilateral clubfoot who died with multiple congenital anomalies. (Reprinted with permission from Shapiro and Glimcher, J Bone Joint Surg Am 1979; 61: 522–530, Wolters Kluwer Health Inc.) (a) Normal talus pattern shows uniform physis surrounding and forming the ossification center of the talus in a midsagittal section. Arrows demonstrate the uniformity of formation and the lack of vessels breaching the physis. (b) Abnormal physeal/ ossification pattern in clubfoot talus is shown in midsagittal section. The posterior part of the ossification center forms normally at left and at lower right (between solid gray arrows), while the superior and anterior parts of the ossification center at top and upper right form in irregular fashion (between interrupted gray arrows) and are frequently breached by vessels. TB talar body, TH talar head. (c) The talar ossification center is central and oval in the normal talus but eccentric (not central) and misshapen in the abnormal clubfoot talus. The ossification centers are outlined by dotted lines; the straight lines represent the regions from which serial histologic sections were made. (d) The extraosseous blood supply entering the bone (arrows) is normal in both normal and clubfoot tali. eov extraosseous vessels, th talar head articular surface. (e) Breaching of the physis of the ossification center (OC) rarely occurs in the normal talus (*) but is quite common in the abnormal clubfoot talus (**). th talar head articular surface. (f) Normal talus and abnormal clubfoot talus are compared from superior, anterior, and medial views. (g) Drawing of bone structure in clubfoot shows talar

Medial Club foot talus

Normal talus

Club foot talus

head and neck (asterisk), the markedly displaced navicular (talonavicular subluxation), and the displaced navicular articulating with the medial malleolus (arrow) and thus positioning the rest of the midfoot and forefoot into varus and adduction. [In many cases of clubfoot, the head and neck of the talus are much more deformed and markedly medially and plantar deviated as well.] (Reprinted with permission from Carroll, Oper Tech Orthop 1993; 3: 115–120, Elsevier.) (h) The dorsal surfaces of a normal (at right) and abnormal (at left) talus are pictured from a micro-CT study; both specimens illustrate a talus from a left foot. The medial deviation of the affected clubfoot talus head (articular surface) and neck are evident along with the smaller size. (Reprinted with permission from Windisch et al., J Anat (London) 2007; 210: 761–766, John Wiley and Sons Ltd/Anatomical Society). (i) Reconstructed CT model shows corresponding clubfoot at left and normal foot at right. In clubfoot side, medial deviation of talar neck and head are seen along with the markedly displaced navicular that still maintains a relationship to the talar head articular surface. The red circle outlines the abnormal positions of the talus, navicular, and medial malleolus. The navicular directs the malposition of the midfoot and forefoot and lies immediately adjacent to the medial malleolus (arrow). The normal foot is shown in comparison; the two arrows point to the normal wide positional separation of medial malleolus and medial navicular bones. (Reprinted with permission from Windisch et al., J Anat (London) 2007; 210: 761–766, John Wiley and Sons Ltd/ Anatomical Society)

7.5 Pathogenesis and Pathoanatomy of Congenital Clubfoot

g

h

a

685

b

i

Fig. 7.7 (continued)

•

•

•

•

an oblique vertical position by the tightened gastrocsoleus muscle. Navicular. The navicular is altered in position to an extreme degree but with no major alteration in shape. It is positioned inward and upward by the tibialis posterior muscle directly and by the tibialis anterior and extensor hallucis longus muscles indirectly “so as to bring its inner border or tuberosity immediately under and in contact with the inner (medial) malleolus” (Fig. 7.7g). The long axis of the navicular bone lies parallel to the leg rather than being at right angles to it. Its position is due to its continuing relationship to the markedly abnormal head and neck of the talus. Cuboid. There is no material alteration in either position or shape in the newborn but alterations develop over the next several years. Cuneiforms and metatarsals. These bones are not altered in their relationship to one another or to the navicular and cuboid bones. Their abnormal positioning is due to the altered positions of the navicular, talus, and os calcis. Medial malleolus. The medial malleolus is always normal in size, position, and shape in the infants but abnormal due to secondary changes in the adult.

As far as etiology is concerned, Adams felt that the malformations in the newborn talus were the most marked of any of the tarsal bones, but there was little reason to attribute that malformation to a primary developmental abnormality of the talus itself. He felt that “the malformed condition of the astragalus (talus) is determined by the malposition of two of the bones with which it articulates viz., the navicular bone and the os calcis; and that its altered form is in evident adaptation to the altered position of these bones.” The altered shape was thus the result, not the cause of the deformity. Soft tissue abnormalities in congenital clubfoot The rigidity of the clubfoot deformity at birth was due not only to abnormally directed muscle pull but also to adaptive shortening of the ligaments. Dissections in a fetus with spina bifida and a clubfoot showed the most shortened ligaments to be the medial deltoid (anterior portion) that held the navicular to the medial malleolus and the posterior ankle ligaments holding the os calcis into equinus. The ligaments in the sole of the foot (along with the plantar fascia) were also shortened. The opposite lateral side ligaments were lengthened. In many cases, even with severe deformity, the ligaments were not

686

significantly tightened and thus ligament involvement appeared to be a secondary adaptive occurrence. Muscle function, gross appearance, and histologic appearance are unremarkable in cases of isolated clubfoot. Eventual successful treatment also supports a normal musculature. Some cases are clearly of neuromuscular origin. Tendons are invariably displaced in adaptation to the altered position of the bones of the foot. Those most affected are the tendo Achilles, tibialis posterior, and tibialis anterior. The tibialis posterior tendon may sublux anteriorly. The tendo Achilles tends to deviate somewhat laterally toward the fibula. Adams stresses the complexity of the clubfoot deformity associated with the primary deformity of the astragalus (talus) and the markedly altered relationship of the navicular to it. Malformation and all the adaptations secondary to it also involve the ligament, tendon, and muscle. It must be understood that any single-plane deformity must of necessity change the relationship of all adjacent bones, tendons, and ligaments. The primary abnormality in the congenital clubfoot is the misshapen and abnormally positioned head and neck of the talus, these being medially deviated and plantarflexed in relation to the talar body. The navicular maintains its relationship to the articular surface of the head of the talus but must itself deviate medially and plantarward to do so carrying the rest of the foot with it. The midfoot and forefoot structures deviate medially and plantarward as well. The calcaneus is not just plantarflexed into equinus but is also inverted into varus almost directly under the head of the talus. The anterior part is adducted, but the posterior part sits and moves laterally so that the superior lateral surface reaches the lateral malleolus. The calcaneus is held in this position posteriorly by the tight tendo Achilles and tight talocalcaneal capsule and posterolaterally by contracture of the peroneal retinaculum and the calcaneofibular ligament. The calcaneus is in equinus and varus and its anterior portion is medially angulated and rotated. The ligaments and capsules of the posterior, medial, and plantar parts of the foot are shortened and thickened. These observations have major implications for treatment. If the normal relative position of the bones is restored “at an early period,” the talus “during the subsequent period of its growth and ossification will gradually assume its natural form. The difficulty of restoration of normal function would increase in direct proportion to the extent to which ossification of this bone has been allowed to proceed in the deformed position.” Adams’ treatment implications from his observations warrant direct quotation. “In order that the astragalus (talus) may have the best chance of assuming its natural form during the period of active growth and ossification in early infancy, the cause of its malformation, viz., the abnormal position of the navicular bone and os calcis should be removed at an early period. With this object in view ------ it is advisable that the operation be performed as soon after the first month from

7 Developmental Disorders of the Foot and Ankle

the period of birth as the circumstances of the case permit having special reference to the healthy condition of the child and mother. Practically I find the most favorable time for operation is when the child is about two months old.” The operation he was referring to was Achilles tenotomy.

7.5.5 Evans Deformity in Midtarsal Joint Evans also was of the opinion that the essential lesion in a clubfoot is a congenital dislocation of the navicular bone on the talus. The navicular bone carries with it the cuboid bone and the calcaneum, changes of shape in the skeletal and soft tissues are secondary and adaptive, and the essential element in manipulative reduction must be the replacement of the navicular bone on the end of the talus, so restoring the medial column of the foot [80]. Based on the pathoanatomy, the aim of treatment was reduction of subluxation of the talonavicular joint and correction of the secondary adaptive changes resisting correction. In relapsed clubfoot these secondary changes were both the tight medial/posterior structures and also the relative overgrowth in length of the bones of the lateral column and the misshapen calcaneal-cuboid joint, described as conical and oriented slightly medially rather than forward. He felt that this opinion of the pathoanatomy was in line with observations by Scarpa [75], Adams [29], Elmslie [81], and also Brockman [82] that the essential deformity in clubfoot is in the midtarsal joint with the other elements of deformity, including varus of the heel, being secondary and adaptive. Scarpa had defined the deformity as a dislocation of the tarsus on the talus with the talus itself not that abnormal when compared to the other bones of the foot [75]. Adams also felt that the talus was not primarily at fault [29]. Elmslie defined a displacement of the navicular and cuboid inward at the midtarsal joint with rotation of the os calcis which brings its anterior end downward and inward [81]. The cuboid is then subluxated inward on the os calcis, the facet for articulation with it lies to the inner side of the anterior extremity of the os calcis, and the anterior end of the os calcis becomes conical in shape.

7.5.6 Irani and Sherman Irani and Sherman dissected 11 extremities with clubfeet and compared them to 14 normals [83]. All samples were from stillbirths (22–36 weeks) or neonatal deaths. Even with all muscles removed from the clubfoot, the position of the foot remained deformed and manually uncorrectable. Once the ligaments were divided between the navicular and the talus, the navicular and the calcaneus, and the capsule of the poste-

7.5 Pathogenesis and Pathoanatomy of Congenital Clubfoot

rior talocalcaneal joint, the deformity could be completely corrected. The insertions of the tibialis anterior tendon (to first cuneiform and base of first metatarsal) and tendo Achilles (to medial side of calcaneus) were always normal. The anterior subtalar joint was most abnormal owing to the medial position of the talus leaving the anterior surface of the calcaneus uncovered. They also found that “the most conspicuous and only constant abnormality is found in the anterior part of the talus.” The neck was always short. The head-neck talar angle in relation to that of the body was always significantly decreased. The normal angle was 150– 155° and the clubfoot value was 115–135°. The calcaneus was only mildly distorted, the navicular was of normal shape but slightly smaller, and the other tarsal bones were entirely normal. The changes in the calcaneus were believed to be secondary. There were no anatomic or histologic changes in vessels, nerves, muscles, or tendon insertions.

7.5.7 Settle Settle dissected 16 clubfoot specimens from the late fetal period and noted uniform abnormalities of the head and neck of the talus [65]. In a literature review covering the previous century, 44 of 52 cases of detailed pathoanatomic dissections had severe deformities of the head and neck of the talus. The 16 clubfeet studied were not associated with major soft tissue abnormalities such as absent muscle or anomalous tendon insertions. The tibia was essentially normal with only a small degree of internal tibial torsion in 4/16. All of the tali were severely distorted both in shape and orientation. While the talus was abnormal, the body and its relation to the ankle joint were close to normal. Size of the involved talus was diminished to about 75% normal. The neck was markedly deviated medially and plantarward. The foot deformity was increased by displacement of the talar articular surface for the navicular still further medially and plantarward upon the deviated neck. The navicular rested against the anterior portion of the medial malleolus. Also, the posterolateral portion of the calcaneus, because of its equinus and rotation, rested against the posterior aspect of the lateral malleolus. The subtalar surfaces were also deviated. Calcaneal changes were milder, but it was displaced into equinus, varus, and internal rotation to relate to the talus. The talonavicular and calcaneocuboid joints were beneath one another rather than side by side. No abnormalities were found in histologic sections of muscle, nerve, or lumbosacral portions of the spinal cord. No abnormal tendon insertions were found. These two detailed studies confirmed both the very high incidence of abnormalities of the head and neck of the talus, much milder structural changes of the other tarsal bones, and no reproducible abnormalities of the blood vessels, nerves, muscles, or tendon attachments in the involved extremities. Irani and Sherman considered all soft tissue changes to be

687

secondary, conforming to the position of the skeleton of the foot [83]. No abnormalities were found in biopsies of muscle, nerve, or lumbosacral regions of the spinal cord.

7.5.8 G ross and Histological Abnormalities of the Talus in Clubfoot There has been fairly consistent recognition of abnormalities of the talus, in particular at the head and neck region, in clubfeet. These have been documented frequently over an extended period of time [16, 58, 64, 65, 83–91]. The talar abnormalities have been found in all categories of clubfoot, raising the question as to whether they are primary or secondary. A difference of opinion exists even among those who postulate a primary talar abnormality. Some have thought that there is defective formation of the cartilage anlage. Others thought that the foot normally assumes a so-called physiological clubfoot position of equinus, supination, and adduction during the second embryonic month and that the eventual clubfoot deformity represents a developmental arrest with the early position persisting. The pathoanatomical findings in clubfoot are not invariable, however, nor are the responses to therapy. In some dissected specimens, the talus has been described as normal [59–61, 90]. Clinically, some feet are recalcitrant to manipulative and cast therapy while others yield readily. A working classification has evolved referring to the talar abnormality, or the recalcitrant clubfoot, as structural or teratologic, and to the normal talus or the foot that responds to cast treatment as postural or nonteratologic. Gross and histological abnormalities in the clubfoot talus have been reported in a boy who died at 9 days of age [92]. One foot was normal and one was a clubfoot that had not been treated. Multiple anomalies were present including congenital scoliosis, hypoplastic lungs, short ribs, genitourinary anomalies (renal hypoplasia with hydronephrosis and hydroureter), and congenital heart disease (patent ductus arteriosus and double aortic arch). There was no myelodysplasia or diastematomyelia. The neuromuscular examination was clinically normal and the central nervous system and spinal cord were normal to gross and histological assessments at autopsy. The comparative findings in the talus (normal foot versus clubfoot) are shown in Fig. 7.7a–f. The changes in the clubfoot talus have been demonstrated in other studies (Fig. 7.7h). These included: Size and Shape The normal talus was 15 mm long and 13 mm wide (at its widest part), and the body was 10 mm high. The clubfoot talus was smaller: 13 × 12 × 7 mm. The normal talus had a smooth, curvilinear head and articular surface, while that of the clubfoot talus was misshapen, stunted, and deviated medially with the neck. The deviation of the talar neck from the long axis of the body was 25° (155°) in the normal foot and 50° in the clubfoot. The body of the clubfoot

688

talus was not as high as that of the normal talus, its superior articular surface was flattened, the corners were rounded, and its posterior surface was oblique. Comparative size and shape differences are shown in Fig. 7.7f. Histology (i) Normal talus. The outline of the head, neck, and body was normal. The ossification center was developing at the central and inferior region of the neck with extension into of the posterior region of the head and the anterior region of the body. Bone development extended to the inferior surface where cortical bone was forming. The ossification center was in the approximate geometric center of the bone. The endochondral sequence with hypertrophic chondrocytes fanned outward in a virtually uninterrupted hemispheric arcade from the inferior surface (Fig. 7.7a). The bone center had not extended to the superior surface of the neck. Many vascular canals were present in the cartilage of the head and body. A prominent blood supply entered from the dorsal superior surface of the neck. The chondrocytes of the cartilage model, endochondral ossification sequence, resorption and marrow development, osteoblasts, osteocytes, osteoclasts, and chondroclasts were histologically normal and safranin-O staining showed normal glycosaminoglycan deposition. The hemispheric endochondral sequence was breached only occasionally by vascular ingrowths which were in either the anterior or the superior region; the posterior region (adjacent to the body) was never interrupted. (ii) Clubfoot talus. The talus was smaller and its shape was less regular than the normal. The ossification center was smaller. It was present at the inferior surface of the neck but was more anteriorly positioned, owing to the shortness and maldevelopment of the head-neck region (Fig. 7.7b). The body was the most normal part of this talus. Vascular channels were seen in the head and body and a prominent blood supply entered via the dorsal surface of the neck. Bone development extended to the inferior surface, with cortical bone present. Chondrocytes, osteoblasts, osteocytes, osteoclasts, and chondroclasts were normal in appearance and safranin-O staining indicated appropriate glycosaminoglycan synthesis. Endochondral ossification was occurring but marked abnormalities were present compared with the normal talus. The hemispheric pattern of the hypertrophic cells was only occasionally interrupted in the normal talus, but it was frequently and widely breached in the clubfoot talus by vascular channels, many of which were laying down intramembranous bone. Vascular breaching refers to the passage of a vessel from the cartilaginous region of the developing talus through the hypertrophied chondrocyte area to the ossification center. The interruptions occupied considerable

7 Developmental Disorders of the Foot and Ankle

space and clearly interfered with the smooth sequential development of cartilage into endochondral bone. As many as five major interruptions were seen on a single section and interruptions were present in each of the serial sections. The vascular- osseous invasions were always in the anterior and superior regions; the posterior region was never interrupted (Fig. 7.7e).

Quantitative Morphology: Size of Ossification Center (i) Normal talus. Eight hundred and ten sections comprised the normal talus. Endochondral ossification (considered from the beginnings of the hypertrophic cells) was present in 70% of the width and was in the approximate geometric center of the bone. The normal center was always surrounded by cartilage on its anterior, superior, posterior, medial, and lateral aspects. (ii) Clubfoot talus. Five hundred and fifty-five sections encompassed the width of the clubfoot talus with 50% of the width undergoing the ossification sequence. When the extent of ossification in the clubfoot was related to the extent in the normal talus, the ossification sequence of the clubfoot talus was 34% of the extent of normal bone. The ossification area was absolutely and relatively diminished in the clubfoot talus. The clubfoot talus was approximately two-thirds the size of the normal talus. Position of Ossification Center Coronal plane. In the anterolateral region of the neck of the clubfoot talus, several sections showed the bone of the ossification center immediately adjacent to fibrous tissue without the interposition of cartilage. Sagittal plane: In the normal talus, 70% of the body width was undergoing the ossification sequence and 30% was still cartilaginous. The ossification center was positioned more toward the medial side. In the clubfoot talus, the ossification center was positioned more laterally (Fig. 7.7c). Vascular Patterns of the Talus (i) Vascular penetration of talus. In the normal talus, the main blood supply entered from the dorsal surface of the neck and ramified in a posterior, inferior, and anterior direction (Fig. 7.7d). Most of the vessels entered at the body-neck region, but occasionally vessels entered toward the head-neck region. Vessels did not enter via the articular surfaces of the head, body, or talocalcaneal joint but occasionally entered from the posteroinferior surface of the body, from the inferior surface of the head adjacent to the neck, and from the inferior portion of the body. The vascular penetrations were spread from the medial to the lateral surface. Fifty-eight vessel entrances (62%) were along the superior surface of the neck via branches of the dorsalis pedis artery, 7 entrances (7%)

7.5 Pathogenesis and Pathoanatomy of Congenital Clubfoot

were via the posteroinferior surface of the body (posterior tubercle branches), 9 (9%) were via the inferior portion of the neck (tarsal sinus branches), and 21 (22%) were via the inferior part of the body (artery of the tarsal canal, from the posterior tibial artery). The vascular penetrance pattern was the same in the clubfoot talus as in the normal talus. Most of the vessels (26, 56%) entered via the dorsal surface of the neck adjacent to the body. Vessels never traversed the articular surfaces of the talus (tibiotalar, talonavicular, and talocalcaneal). There were some vessels at the inferior surface of the posterior aspect of the body (3, 6%) as well as along the inferior surface of the neck (8, 17%) and the inferior surface of the body (10, 21%). There were occasional vessels entering on both the dorsal and inferior surfaces of the neck adjacent to the head. (ii) Vascular pattern within the talus. Vessels were distributed throughout the body, neck, and head in the cartilage model of the normal talus. There were relatively few vessels in the superior region of the body. The vascular pattern of the clubfoot talus was not demonstrably different from that in the normal one (Fig. 7.7d). There did not appear to be either an abnormal pattern of vascularity within the cartilage model or a significantly changed number of vessels. Summary of Talar Abnormalities in Clubfoot Study The detailed study of a congenital clubfoot confirmed an abnormally shaped talus with clear histological abnormalities. The unilateral clubfoot allowed for comparison between a normal and an abnormal talus from the same individual. The gross talar abnormalities described previously have been r emarkably uniform and the clubfoot talus studied adhered to that pattern. The anterior portion, with the head and neck shortened, medially deviated, and malformed, was far more abnormal than the body. The clubfoot talus was smaller than the normal talus. Its ossification center was absolutely and relatively smaller and was positioned eccentrically in a more anterior and lateral position, with a portion of it immediately adjacent to dense fibrous tissue without the interposition of cartilage. Marked abnormalities of the anterior and superior regions of the endochondral sequence of the ossification center were demonstrated. The 50° head-neck/body angle in the clubfoot talus and similar large deviations previously reported in clubfoot tali [16, 63, 86, 90] fail to support a developmental arrest concept since the clubfoot medial deviations are far greater than the normal. Investigations have also been made of the talar neck angle in other mammals, many who normally have medial deviations greater than that seen in the human. No animal had normal deviation as great as that seen in the congenital human clubfoot. In a separate study, the talar neck angle in 40 species was never greater than 30°. Talar ossification via the endochondral mechanism begins at the 24th to 26th week of intrauterine life. The ossification center normally forms near the

689

geometric center of the talus and is referred to as a centric ossification center [93]. In this study, the ossification center of the normal talus was positioned centrally, surrounded by cartilage, except inferiorly, while the eccentric position of the ossification center in the clubfoot talus was markedly abnormal. The extraosseous blood supply was normal in both tali in the study and conformed well to the previously described pattern. The major supply entered the dorsal surface of the neck via branches of the dorsalis pedis artery and entered the inferior surface via the artery of the tarsal canal, a branch of the tibialis posterior artery [94, 95]. The intraosseous blood supply in both the normal and the clubfoot talus was similar in distribution and extent. The one conspicuous vascular abnormality not recognized previously was the highly increased susceptibility of the endochondral sequence of the ossification center in the clubfoot talus to breaching by vessels. This breaching was exclusively in the anterior and superior regions, as in the normal talus, but was markedly excessive and premature. The breaching interrupted the proliferating chondrocyte region and was associated with the marked growth retardation of the head and neck. It is not possible to assess whether this caused or was a result of the growth retardation, but the smooth sequence of growth and endochondral ossification is markedly disturbed in association with the premature invasion. The talar abnormality in clubfoot could represent a primary developmental defect in the cartilage model but could also be secondary to prolonged extrinsic mechanical pressure. A developmental arrest appears unlikely, based on the histological picture. The eccentric position of the ossification center is markedly abnormal and the changes of endochondral sequence irregularity are more characteristic of prematurely advanced rather than retarded development. Retardation or arrested development of an otherwise physiologically normal talus would feature a smaller, but normal ossification center, appropriately positioned and showing a regular vascular pattern with a minimum amount of breaching. The vascular pattern of normal and abnormal tali in a child with a unilateral clubfoot is shown (Fig. 7.7).

7.5.9 M uscle, Nerve, and Connective Tissue Abnormalities in Clubfoot Studies Muscle, Nerve Gray and Katz studied 193 muscle biopsies from 62 patients with idiopathic clubfeet [96]. Comparisons were made with biopsies from 13 normal legs. No significant difference was found between the diameter of the muscle fibers from normal and affected legs from children less than 6 months of age. Thinness of the clubfoot calf musculature was due to a reduction of the number of fibers rather than their size. The muscle structure was normal, a finding that excludes denervation and reinnervation as causative features. The only traces of abnormality in the affected soleus muscles

690

were 61% type I fibers versus 44.3% in normal (a finding also seen in the other leg muscles) leading to speculation of “a defective neural influence” on limb development. Isaacs et al. also suggested an underlying neurogenic origin based on their histopathologic studies [97]. Bill and Versfeld were unable to detect changes on electromyography [98]. Other studies, primarily histologic, have also failed to find a definitive neuromuscular cause for idiopathic clubfoot in otherwise normal patients with clinically normal neuromuscular examinations. Ipsilateral calf atrophy is seen but with normal gastrocsoleus strength. In addition, once an idiopathic clubfoot is corrected, there is no subsequent neuromuscular weakness that worsens with time. Ippolito and Ponseti also reviewed opinions in the literature concerning the subtle possible neurogenic origins of clubfoot, also finding that most found no evidence for it [99].

7 Developmental Disorders of the Foot and Ankle

third of the muscles of the posterior and medial leg; increased fibrous connective tissue in these muscles, their tendon sheaths, and the adjacent fascia; and shortening of the triceps surae. They noted thickening of the distal parts of the tendo Achilles and tibialis posterior tendon. They felt that the posterior-medial ligaments pulled the foot into the deformed position and noted marked shortening in particular of the tibionavicular and plantar calcaneonavicular ligaments. They proposed a retracting fibrosis as the primary cause of the clubfoot deformity. A morphologic histologic study using light and transmission electron microscopy of connective tissue obtained during surgical corrections failed to identify any myofibroblasts or myofibroblast-like cells from capsular, fascial, tendon sheath or ligamentous tissue specimens in nine clubfeet [103]. Fifty specimens were examined by light microscopy and 26 by electron Connective Tissue Some have concluded that connective microscopy assessing soft tissue from medial and lateral tissue abnormalities are the primary cause of the clubfoot capsule and fascia, spring ligament, and multiple tendon deformity with the skeletal changes secondary. Attention has sheaths from tightened muscles including tendo Achilles concentrated on the medial portion of the foot at the main and posterior tibial. The study was specifically underregion of concavity and soft tissue tightness. taken to assess the theories of retracting fibrosis or myofibroblast activity as causative mechanisms. The absence of myofibroblast-like cells or typical fibroblasts in (a) Histology. Fried studied 56 patients with clubfoot and clubfoot connective tissue structures did not support the noted that the insertion of the tibialis posterior tendon theory of abnormal fibroblasts causing contracture in was always the site of a thick hard fibrous mass [100]. idiopathic clubfoot. The adjacent plantar fascia was found to be contracted by Zimny et al. and their studies identified myofibroblast- ( b) Magnetic resonance imaging studies of leg muscles in clubfeet. It has been recognized that the calf and overall like cells as the possible cause [101]. Fukuhara et al. leg musculature is thinner on the clubfoot side than on studied 16 clubfeet and 27 normal feet from spontanethe normal side. MR imaging studies have assessed the ously aborted human fetuses ranging in age from 14.5 to leg musculature in clubfeet legs compared to the contra22.5 weeks [102]. Histologic studies did not find talar lateral normal leg in terms of volume of muscle tissue, deformity to be the primary lesion. They felt that “the its composition (the relation of muscle, fat, tissue), and cells and collagen fibers of the medial ankle ligaments of change with time. Ippolito et al. demonstrated leg musclub feet appeared to be the site of the earliest changes in cle atrophy to be part of the clubfoot disorder already that they had lost their spatial orientation and had contracted.” They felt that myofibroblast-like cells seemed to present in the fetal stages of a developing clubfoot and in create a disorder of the ligaments resembling fibromatonewborns prior to starting treatment [104]. Cross- sis. Fetuses with obvious teratological abnormalities sectional MR images midway between knee and ankle were excluded. The medial ligament complex studied were done in three groups of eight comparing normal to included the deltoid and spring ligaments and the inserclubfoot sides: untreated newborns 10 days–2 weeks of tion of the tibialis posterior tendon. These were conspicuage, children 2–4 years of age treated by the Ponseti ously thickened and firmly bound together the tuberosity method, and adults 19–23 years of age treated at of the navicular, the neck of the talus, and the sustentacu2–3 months of age including use of limited posterior lum tali. The navicular was always subluxated medially releases. Measurements were made of cross-sectional at the talonavicular joint. Ligament histology showed total leg volume, total muscle tissue volume, and adirounded nuclei in the clubfoot and fragmented bundles of pose tissue volume. As noted, marked atrophy (or undercollagen fibers. Some of the cells resembled fibroblasts. development) of the leg muscles on the clubfoot side Changes were more marked in severe deformities. They was found in fetuses and untreated newborns. The leg felt that the talar changes were secondary. Ippolito and muscle atrophy increased with growth along with the Ponseti studied 5 clubfeet from fetuses aborted at relative increase of the adipose tissue. Leg muscle atro16–20 weeks gestation using serial histologic sections phy was therefore an integral pathologic component of [99]. Fibrosis was the main finding in the normal feet/ the congenital clubfoot entity. Another MRI study by legs compared with three normal feet. There was a Ippolito et al. assessed normal and clubfoot legs in three decrease in the size and number of fibers in the distal groups of seven each at mean ages of 4.8 months,

7.5 Pathogenesis and Pathoanatomy of Congenital Clubfoot

11.1 months, and 4.7 years. All patients had been treated with the Ponseti technique including percutaneous Achilles tenotomy. This study confirmed that the muscles were thinner and shorter on the involved clubfoot side. This was seen in all three compartments (anterior, lateral, and posteromedial) accompanied by relatively longer distal tendon components (tibialis anterior, peroneus longus, and tendo Achilles) [105]. Moon et al. performed MR imaging on the legs of treatment- responsive and treatment-resistant clubfeet with comparisons to the normal side. Parameters measured were muscle area, subcutaneous fat, intracompartmental fat, and total area. Treatment-resistant clubfeet had more marked MR imaging leg changes than treatment-responsive clubfeet. The resistant clubfeet had excess intramuscular fat replacement and unique patterns of muscle hypoplasia. In treatment- resistant legs, the muscle area was decreased by minus (−) 47.8% compared to minus (−) 26.6% for the treatment- responsive side, while intracompartmental fat was 402.6% in the treatment-resistant group and only 9% in the responsive group. These values indicate more fatty t issue and less muscle in the intracompartmental muscles in poorly responsive legs. Fat is accompanied by fibrous tissue both of which respond less well than normal muscle to manipulations [106].

7.5.10 Rotational Abnormalities in the Clubfoot Deformity There is almost universal recognition of medial displacement of the navicular in relationship to the head of the talus and of the medial and plantar deviation of the head and neck of the talus. Most also recognize a medial shift of the cuboid in relation to the anterior calcaneus at the calcaneocuboid joint and an inversion of the calcaneus under the talus with the posterior portion laterally displaced toward the fibula. Several investigators have pointed out milder abnormalities of other tarsal bones. McKay felt that the talus was in neutral alignment in the ankle mortice but that the calcaneus was rotated horizontally around the interosseous ligament as well as being in equinus. This caused the calcaneus posteriorly to lie more laterally toward the fibula [107]. Carroll et al. recognized the talonavicular subluxation but also described external rotation in the long axis of the body of the talus at the ankle and the medial rotation of the calcaneus [108, 109]. Swann et al. [110] point out that rotational deformity at the ankle and hindfoot has been for the most part underappreciated and incorrectly understood in clubfeet. While many authors refer to medial tibial torsion in clubfeet, they note that the ankle mortice and hindfoot are laterally rotated with no medial torsion of the tibia. This places the lateral malleolus in the transmalleolar plane more lateral than in the normal. Lateral rotation of the hindfoot and ankle mortice upon the tibia occurs, while the forefoot remains medially rotated in

691

relation to the hindfoot. In lateral foot/ankle radiographs of an uncorrected clubfoot with the knee pointing forward, the calcaneus is in lateral profile but the ankle is seen as in anteroposterior projection. The talus appears “flat-topped.” In reality, they indicate that these findings are due to the lateral rotation; if the same projection is used with the leg medially (internally) rotated 30–45°, the hindfoot appears normal with the two malleoli superimposed and the talar dome round.

7.5.11 Structural Abnormalities of the Calcaneus in Clubfoot Some observers have studied the calcaneus in detail and found considerable abnormality although it is hard to understand how the changes could primarily lead to the clubfoot entity. Windisch et al. studied seven clubfoot tarsal bones in detail from fetuses at 25–37 weeks [111]. Characteristic talar changes were seen and measured in multiple planes, as were calcaneal changes. Previously, Gilbert et al. performed histological studies on the calcanei from several clubfeet and noted changes consistent with the hypothesis that an intrinsic primary growth disorder causes the formation of a small hypoplastic bone and a smaller foot [112].

7.5.12 Initial Clinical Assessment of Clubfoot Deformities Clubfoot is diagnosed by clinical assessment at birth and casting or splinting treatment is started. If the deformity is associated with an evident systemic neuromuscular or connective tissue disorder, note is made of that fact but treatment is also started early in similar fashion generally with the understanding that non-operative treatment will be less effective than in otherwise isolated or idiopathic congenital clubfoot. In the recent past except in isolated instances, few efforts were made to search for an etiology, document the deformity even in semiquantitative fashion, or perform imaging other than plain radiographs. Some centers even discontinued plain radiographs owing to difficulty in standardization and interpretation. There have been several detailed efforts more recently to better define and classify the disorder. Several of the classifications used to help treat and assess clubfoot deformities are outlined in Table 7.2.

7.5.12.1 Clinical Index of Severity (i) Early clinical grading systems: In the nineteenth century, clinical descriptions of the clubfoot and other congenital and non-congenital foot deformities were very detailed. Gradations of deformity were commonly recognized. (a) Brodhurst, 1856. Brodhurst, discussing congenital talipes (from talus and pes), referred to four princi-

692

7 Developmental Disorders of the Foot and Ankle

Table 7.2 Classifications of clubfoot (equinovarus) deformity → Brodhurst 1856. Brodhurst defined four degrees of deformity going from mild to severe First degree. In the mildest form the anterior part of the foot is drawn inward by the retraction of the tibialis anterior. No other deformity is super added. The description is suggestive of what we now refer to as metatarsus adductus Second degree. The inner edge of the foot is raised, the toes are inverted, and the heel is elevated. A line extending along the long axis of the leg in front of the ankle will fall along the axis of the little toe Third degree. This represents the common, ordinary form of clubfoot. The foot is inverted with the inner edge raised so that the plantar surface is at right angles (to the floor in a weight-bearing position) with the dorsum pointing outward, the heel is raised, the sole of the foot is shortened, concave and irregular in shape, and the extensor muscles of the foot and the adductors are more retracted than in the first two degrees Fourth degree. This is an exaggerated form of the previous third degree. The toes lie upward and inward, the dorsum of the foot lies outward and downward, the inner edge of the foot is so inverted that it approaches the inner side of the leg, and the heel is drawn upward and inward. The flexors of the toes are rigidly retracted, the plantar surface of the foot is abnormally concave, and the length of the foot is diminished. The tendo Achilles is not in the midline but has been pulled medially adjacent to the posterior tibial vessels on the posteromedial aspect of the leg → Bigelow 1900. Bigelow described the equinus and varus deformities separately but clearly recognized that congenital clubfoot was the combined equinovarus deformity. Each component of the deformity was defined into three degrees of severity. Descriptions go beyond the early months to include uncorrected deformity during walking years Equinus. First degree. There is direct elevation of the heel from the floor (action of gastrocnemius). Patient walks on toes that are hyperextended to a right angle Second degree. There is more severe involvement with the calcaneus sometimes touching the tibia. The talus is dislocated forward; plantar surface is arched (pulling heel and toes closer together) and with gait, the great toe is dorsiflexed and lateral four toes are plantarflexed Third degree. Further plantarflexion (equinus) directs toes backward until dorsal surface of the foot/toes becomes weight-bearing, acting as the sole of the foot. The metatarsals curve backward, and the gastrocnemius, flexors of toes, and plantar aponeurosis are all maximally contracted Varus. First degree. Inner edge of foot is raised from ground Second degree. Patient walks on outer sole of foot Third degree. Sole is directed upward and the dorsum of foot functions as the weight-bearing surface. Tibialis anterior and tibialis posterior muscles cause and worsen the varus/adduction deformity During the late 1800s and early 1900s, as shown above, descriptions of the deformities were frequently categorized in semiquantitative terms varying from three to five for each dimension to denote mildest to most severe forms → Harrold and Walker. 1983. Harrold and Walker classified clubfoot deformity into three grades of severity at birth Grade 1. Mild. The foot could be held at or beyond the neutral position Grade 2. Moderate. The foot could not be manually reduced to neutral, but the fixed equinus or angle of varus was 20° or less Grade 3. Severe. The fixed deformity was of more than 20° of equinus or varus → Ponseti and Smoley. 1963. Assessment of ankle dorsiflexion, heel varus, forefoot adduction, and tibial torsion (all expressed in angular degrees [°]) were done with the result expressed as good, acceptable, or poor Ankle dorsiflexion Heel varus Forefoot adduction Tibial torsion Result >10° 0° 0° to 10° 0° good 0° to 10° 0 to 10° 10° to 20° moderate acceptable 0° or < >10° >20° severe poor →Catterall. 1991. Assessment was made of the hindfoot regarding lateral malleolus position, equinus and visible creases, and the forefoot regarding lateral border (straight or curved), mobility, cavus deformity, and extent of supination Foot Resolving pattern Tendon contracture Joint contracture False correction Hindfoot Lat malleolus Mobile Posterior Posterior Posterior Equinus No Yes Yes Yes Medial crease No No Yes No Posterior crease No Yes Yes Yes Anterior crease Yes No No Yes Forefoot Lateral border Straight Straight Curved Straight Mobile Yes Yes No Yes Cavus No +/− +/− +/− Supination No No Yes No → Carroll et al., 1978, Carroll 2012. Classification was based on etiology defining four basic types of clubfoot Postural – benign, resolves with stretching, casts Idiopathic – the congenital clubfoot of variable severity Neurogenic – actually neuromuscular, associated with an underlying neurologic or muscle disorder Syndromal – occurs in many skeletal dysplasias and connective tissue disorders, tends to be very rigid

7.5 Pathogenesis and Pathoanatomy of Congenital Clubfoot

693

Table 7.2 (continued) → Dimeglio et al., 1995. Classification of clubfoot severity determined by angular measurements in the four planes of deformation Grades I – IV (benign to very severe) determined by a scoring system in which there is sagittal plane evaluation of equinus, sagittal plane evaluation of varus, horizontal plane evaluation of rotation, and horizontal plane evaluation of forefoot relative to hindfoot. For each plane of assessment, 1–4 points (normal to severe deformation) are given based on the angle of deformity measured. Additional scoring involves (0 or 1 point) for posterior crease, medial crease, cavus deformity, and muscle condition. The maximum deformity score would be 20 Sagittal plane, equinus: 1 point (normal) 0° to −20° (by which is meant dorsiflexion above neutral) 2 points 0° to 20° (mild equinus) 3 points 20° to 45° (moderate equinus) 4 points 45° to 90° (severe equinus) Sagittal plane, varus: 1 point (normal) 0° to −20° (=valgus) 2 points 0° to 20° (mild varus) 3 points 20° to 45° (moderate varus) 4 points 45° to 90° (severe varus) Horizontal plane, 1 point (normal) 0° to −20° derotation: 2 points 0° to 20° (mild) 3 points 20° to 45° (moderate) 4 points 45° to 90° (severe) Horizontal plane, forefoot 1 point (normal) 0° to −20° (=abduction) relative to hindfoot: 2 points 0° to 20° (mild adduction) 3 points 20° to 45° (moderate adduction) 4 points 45° to 90° (severe adduction) Additional scoring involves 0 or 1 point for posterior crease, medial crease, cavus deformity, and muscle condition Dimeglio et al. summary Grade Type Score Overview of reducibility I Benign 90% resolving II Moderate 5–9 >50% reducible, some resistance III Severe 10–14 15 50˚, soft-stiff, reducible, partially stiff >50˚, stiff-soft, stiff, partially reducible –20˚ Aggravating elements Posterior crease Medial crease Cavus Fibrous musculature Total score possible

Points 4 3 2 1 0 1 1 1 1 0 to 20

Fig. 7. 8 (continued)

clinical classification focuses attention on all components of the deformity. Its strength lies in quantitating response to treatment for an individual practitioner. The classification is outlined clinically in Fig. 7.8a. This classification, which does not consider etiology, provides a semiquantitative indication of the extent of deformity to help define the progress of treatment. (b) Pirani et al. Pirani et al. [117] described a 6-point clinical classification system based essentially on the contractures present (Fig. 7.8c). The scoring is based on three signs related to the midfoot (curvature of the lateral border, severity of the medial crease, position of the lateral part of the head of the talus) and three signs related to the hindfoot (severity of the posterior crease, emptiness of the heel, and rigidity of equinus). The principle of scoring for each parameter is 0, no abnormality; 0.5, moderate abnormality; and 1.0, severe abnormality. Each foot receives a total score between 0 and 6 (Fig. 7.8c). (c) Independent validation of current systems. Several studies have found the newer grading systems to be of value in projecting the number of casts needed for correction and the likelihood of the need for tenotomy. Dyer and Davis found the Pirani system useful in projecting the need for number of casts and tendo Achilles tenotomy [118]. With the

Ponseti method of treatment and the Pirani scoring system, in 70 feet a score of 4 or more is likely to need at least 4 casts, 1 scoring less than 4 will need 3 or fewer casts, and 1 with hindfoot score of 2.5 or 3 has a 72% chance of needing tenotomy. It is the hindfoot score that more accurately predicts the need for tenotomy. Scher et al. also found a good correlation with 85% of feet with a score above 5 requiring tenotomy [119]. Wainwright et al. [54] outlined and compared four classification systems for clubfoot: Dimeglio et al. [116], Ponseti and Smoley [114], Harrold and Walker [91], and Catterall [115]. That of Dimeglio et al. had the greatest reliability but all were helpful. Aydin et al. studied 108 feet undergoing Ponseti method treatment and found Pirani scores at the start and percentage change in scores during treatment projected need for Achilles tenotomy [120]. The higher the scores, the greater was the likelihood for needing tenotomy. Goldstein et al. also found that the higher the Dimeglio/Bensahel scores were at presentation, the higher was the likelihood of the need for surgical intervention [121]. A good correlation of the Pirani grades with subsequent relapse after the Ponseti method was found in a study of 80 clubfeet in which there were 17 relapses (21%) [122]. The median total score in those with no relapse was 3.5 and with relapse 5.0. In other

7.5 Pathogenesis and Pathoanatomy of Congenital Clubfoot

words, the stiffer the deformity of the beginning, the higher was the likelihood for relapse. While most reports find good value in the current classification systems, there are some reports showing little correlation and pointing out the need for continuing the search for earlier clues to prognosis. Better understanding of etiology and pathoanatomy at the start of management would allow for improvement in prognosis. Chu et al., using initial scores, found that both the Dimeglio/Bensahel and Catteral/Pirani scores had a low correlation with the number of Ponseti approach casts needed and also a low or absent correlation with need for tenotomy [123]. They concluded that a better classification system was still needed. As noted above, however, a subsequent report from the same institution 5 years later found some value (expressed as “increased risk for needing surgical intervention”) with the Dimeglio/Bensahel scores [121] and a report by Goriainov et al. [122] also found a good correlation of the Pirani method with subsequent relapse. These findings point out that the clubfeet scoring systems are semiquantitative at best with considerable subjective grading input. Gao et al. concluded that the prognostic value of the Dimeglio and Pirani scores “remains questionable……in the early treatment stages” with a low (Dimeglio) or absent (Pirani) correlation for the number of casts [124]. Several of the many classification systems proposed are summarized in Table 7.2. They show a common reference to the main positions of deformity with variable documentation of the degree (extensiveness) and rigidity of that deformity.

7.5.13 Imaging for Clubfoot Deformities 7.5.13.1 Plain Radiographic Indices Anteroposterior and lateral plain radiographic projections have been most commonly used to assess degree of deformity and correction in clubfeet. The anteroposterior talocalcaneal axis (defined by a line through the long axis of each bone) should be in the range of 35–50° in the infant and young child up to 2 years of age and then decreasing later to the 17–25° range in the AP plane and 35–50° in the lateral plane [9]. Ippolito et al. documented 23° ± 4° for the anteroposterior plane [10]. Joseph et al. have done a detailed study of this angle in 75 normal feet and 145 clubfeet assessing anteroposterior views and lateral stress dorsiflexion and stress plantarflexion views [8]. The other bone centers of the tarsals appear later but the metatarsals all ossify the diaphyses before birth. The multiplanar nature of the deformity, its relative rigidity, and the difficulty of holding the foot of a squirming newborn in any position resembling a standardized plane all combined to limit the value of the

697

radiographic information. The need for objective criteria remains great. Zimmerman et al. have done a controlled standardized study of various methods to document changing hindfoot position pre- and post-tenotomy on lateral radiographs [125]. They concentrated on the lateral talocalcaneal, tibiocalcaneal, and dorsiflexion (plantar surface of the foot and long axis of tibia) angles with all radiographs taken with the foot in forced dorsiflexion. Their study (in relation to Achilles tenotomy) outlines the standardized technique. Each measurement was found to be valuable and statistically reliable between observers. The mean differences pre- and post- tenotomy were dorsiflexion angle increase of 17° (from 9° to 26°), tibiocalcaneal angle increase of 19° (from 1° to 20°), and talocalcaneal angle increase of 9° (from 30° to 39°).

7.5.13.2 Ultrasound Assessments (i) Prenatal assessments. Ultrasound is capable of detecting up to 80% of clubfeet with prenatal studies. The earliest time of detection has been 12 weeks. In one study clubfoot was identified in 60% of cases [126]. Diagnosis was an early event in gestation (45% between 12 and 17 weeks), a late event (45% detected between the 18th and 24th weeks), and a very late event (10% detected between the 25th and 32nd weeks). Clubfoot ultrasound diagnoses were between 12 and 23 weeks in 86%. There was no relationship between prenatal diagnosis and degree of rigidity of the foot. Another large study pointed out that some cases of clubfeet were associated with other growth abnormalities such as abnormal karyotypes, hip or other limb abnormalities, and other organ abnormalities [127]. They recommended karyotypic or other studies based on the clubfoot finding. (ii) Postnatal studies. Studies in the mid-1990s demonstrated that ultrasound could assess cartilage model bone relationships in clubfeet. Suda et al. have provided a sonographic classification [128]. They used measured angles comparing ultrasonography of 24 newborns with 32 clubfeet and 13 newborns with 22 normal feet. Angle differentiation in the two groups showed clubfeet with higher dispersion in 95% confidence intervals for all angles. They classified patients sonographically as IIa, slight clubfoot; IIb, moderate; IIc, severe; and IId, very severe. Aurell et al. felt that ultrasound could be incorporated into clinical management profiles showing ultrasound to be valuable in determining the position of the navicular in relation to the head of the talus and the medial malleolus, a relationship visualized in each of the 30 untreated clubfeet studied [129]. The mean distance between the medial malleolus and the navicular

698

was significantly shorter in clubfeet compared to normal feet. Kuhns et al. showed the potential value of ultrasound of the navicular by showing the talonavicular relationship in the Ponseti approach [130]. Gigante et al. compared sonograms in 42 congenital clubfeet with 42 normal feet [131]. The most valuable projections were sagittal posterior where progressive gain in dorsiflexion could be assessed and coronal lateral sonograms where relationships between calcaneus and cuboid could be measured. The relationship of the navicular to the talar head was not well demonstrated however on either transverse of sagittal anterior views because of overlap or displacement. Shiels et al. used both static and dynamic sonography to evaluate cartilaginous structures in 13 clubfeet and 335 normal feet from 0 to 24 weeks of age [132]. Examination involved coronal oblique assessment of the medial malleolar- navicular (MMN) distance and the calcaneal-cuboid relationship, sagittal assessment of the talonavicular relationship, and transverse assessment of navicular subluxation and deformation. The dynamic aspect was based on changes with abduction/adduction stress maneuvers. Valuable comparative differences in the two groups were found. Coley et al., from the same center, extended the sonographic study throughout the first year of life to 127 feet [133]. Differences primarily in MMN relationship were initially diminished in clubfeet compared to normal feet and that parameter appeared valuable in assessing responses to therapy. Cash et al. showed the higher-resolution value of three-dimensional ultrasound for neonatal clubfeet [134]. One drawback for clinical use was the need for the child to be perfectly still. The pilot studies were done on children who were asleep. Accuracy to 0.5 mm was obtained however. Ossified structures cannot be assessed by any ultrasound technique. The method in the neonatal normal foot however clearly delineated the position of the navicular in relation to the medial malleolus and the anterior pole of the talar head. ( iii) Ultrasonography to assess tendon healing following percutaneous Achilles tenotomy. Two studies have shown the feasibility of assessing post-tenotomy healing by ultrasound. Barker and Lavy showed tendon continuity at 6 weeks post-release when their study ended [135]. Mangat et al. studied 27 Achilles tendons following percutaneous release for clubfoot treatment using the Ponseti approach [136]. Assessments were done in each at 3, 6, and 12 weeks post-release. The transition to normal structure was generally demonstrated at 12 weeks but not at 6, although tissue continuity was seen at 6 weeks. Full healing was considered to have occurred with homogeneous tendon fibers bridging the gap zone with obliteration of the distinct cut ends of the tendon. The repair tissue did not appear narrower than

7 Developmental Disorders of the Foot and Ankle

the normal tendon but was if anything somewhat bulbous at the repair site.

7.5.13.3 Computerized Axial Tomography Computerized tomography (CT) scans provide accurate quantifiable information about the three-dimensional aspects of a clubfoot deformity. Johnston et al. presented an early CT overview [137]. CT is increasingly valuable with progressive bone development since cartilage models are not well defined. CT is particularly valuable in assessing the navicular- talar head relationship. A study by Farsetti et al. has clarified the site of persistent in-toeing when seen in patients with residual clubfoot deformity. While many considered this to be due to internal tibial torsion, CT study has revealed that the tibia and fibula in clubfeet were actually externally rotated in comparison with the normal leg. In a study of 90 clubfeet, using the normal side as control, three groups of clubfoot patients had values of external tibiofibular torsion of 32.2° (posteromedial release), 23.9° (modified Ponseti treatment), and 21.1° (Ponseti treatment) with the normal side torsion angle 21.4° [138]. The tibiofibular torsion angle seemed to be related to the treatment technique. The study concluded: “in the treated congenital clubfoot, persistent intoeing is not related to the angle of tibial torsion but rather to the amount of correction of calcaneal inversion and residual forefoot adduction.” 7.5.13.4 Magnetic Resonance Imaging Magnetic resonance imaging (MRI) is particularly effective in imaging abnormal cartilage and bone structure. Pirani et al. devised an approach to the chondro-osseous abnormalities of the newborn untreated clubfoot deformity and then studied patients as they underwent the Ponseti treatment [139]. The pretreatment images revealed the expected findings as described in earlier pathoanatomic postmortem studies and all the major pathology could be visualized in vivo. With treatment all the abnormalities improved markedly or corrected completely. Positional relationships corrected but abnormal shapes of individual tarsal bones also corrected with the application of corrective mechanical force into anatomic planes. Cahuzac et al. studied 12 patients with unilateral clubfeet by MRI under anesthesia 15 days before surgery at a mean age of 11 months (9–16) [140]. The volume of structures in the clubfeet was about 20% smaller than in the normal feet. The reduction of the volume of the ossification center of the talus was 40% and of the calcaneus 20%. The cartilage model of the calcaneus was significantly medially rotated relative to the bimalleolar axis in the clubfeet compared to the normal (15° versus 3°). This clearly implies the need to externally rotate the calcaneus surgically if extensive open reduction is chosen. Kamegaya et al. concentrate on the talar neck angle and the talonavicular relationship [141]. They assessed 36 clubfeet at a mean age of 9 months (4–12) by MRI. The mean talar neck angle was much higher in the

7.5 Pathogenesis and Pathoanatomy of Congenital Clubfoot

clubfeet (44°+/− 8°) than in normal feet (31° +/− 6°). They were also able to assess the position of the navicular in relation to the malaligned talar head. In those eventually requiring operation, 18 had a medial shift of the navicular and none a lateral shift. All nine normal feet had a slight lateral shift of the navicular. MRI has also been used for longer-term assessment. Not surprisingly intertarsal relationships can be less than normal after closed or surgical treatment owing to the high resolution of the technique, but the clinical significance of MRI findings is still unclear [142, 143]. Moritani et al. evaluated the talonavicular relationship in 14 clubfeet by MRI. They found a significant difference in the talonavicular angle between patients who had a posteromedial release and those who had posterior release or casting only. The angle formed by the long axis of the talus and the short axis of the navicular was measured. Roche et al. studied seven adolescent subjects at a mean age of 13 years who had a unilateral clubfoot treated several years earlier by serial casting and posterior-lateral surgical [144] releases. Comparison was made to the normal side. 3D reconstructed MR images were done focusing on the size, shape, and articulating surface morphology of the distal tibia, talus, calcaneus, navicular, and cuboid. Multiple parameters were measured but the clinical usefulness of such information will only be determined over time. The specific morphological parameters assessed were tarsal volume, tarsal surface area, articular surface area, articular surface length, joint malalignment, the degree of talar flattening, and the degree of navicular flattening. These parameters contrast with radiographic assessments that concentrate on angular measurements. The sequelae of even successfully treated clubfeet are clearly revealed including a flattened talar dome, a wedge-shaped talar head, a deformed posterior navicular articular surface, and distorted subtalar joints especially including a deformed anterior calcaneal articular surface.

7.5.14 Historic Overview of Treatment, Ancient Times to Present: Manipulation (Kite, French Functional, Ponseti); Entire Range of Surgical Procedures; Current Approaches to Management 7.5.14.1 Ancient Times to the Late 1700s Clubfoot deformity was so common and so obvious to clinical observation that it attracted considerable attention and treatment from ancient times onward. This was well-outlined by Brodhurst of London in his 1856 book On the Nature and Treatment of Club-Foot [113]. Hippocrates wrote with great clarity on the disorder: “Most cases of congenital club-foot are remediable, unless the declination be very great, or when the affection occurs at an advanced period of youth. The best plan, then, is to treat such cases at as early a period as possible, before the deficiency of the bones of the foot is very

699

great, and before there is any great wasting of the flesh of the leg.” He commented on the use of bandages as a mode of cure followed by corrective shoes and splints – “for such cases yield sooner to treatment than one would believe” [113]. Although little was added over the next 2000 years, the deformity and its management were discussed in major medical works: Abroise Paré (Lyon 1641); Severinus (1643); Arcaeus (1658), illustrating a splint and boot he used to treat the deformity; Andry (Paris 1741); and duVerney (Paris, 1751), who referred to muscular retraction as the cause, “these distortions depend entirely on the unequal tension of the muscles and ligaments; for those muscles that are extremely tense draw the parts towards them, while their antagonists yield, being relaxed”. By the late 1700s, surgical treatment of the tight tendo Achilles was being tried in isolated instances and the value of mechanical manipulation therapy was understood and described with considerable insight.

7.5.14.2 Sheldrake Sheldrake published a book on clubfoot in 1798 containing descriptions and management principles accurate even today [145]. Sheldrake was a bracemaker who at that time also practiced repetitive manipulation therapy and bandaging for clubfeet and other lower extremity deformities. Several dozen cases are described in detail with the following observations and principles well outlined. • Clubfeet are correctable by bandaging/taping 1–2× per week for several weeks. • A major concern is the danger of recurrence (relapse) until the child is walking so bandaging must be continued even after correction is achieved. • The involved foot and leg will always be smaller than the non-involved side. • Even when the deformity is straightened, the muscles need to be worked on to strengthen them. • “Though such deformities may certainly be cured in a short time, during early infancy it will often be necessary and therefore always prudent to keep such feet restrained to their natural position till the patient is able to walk alone.” • The clubfoot is distorted but also initially rigid. • The manipulations and bandaging in infancy could correct deformity in as little as 3–6 weeks. • More rigid devices to attempt correction should not be used; “…the application of the things commonly called leg-irons is of no use in these cases, except to amuse the parents, by seeming to do something….” • “….When the foot of a young child has been distorted in this manner, it should not be left at liberty till it is able to walk; as the inequality in the action of the muscles of the foot will always give a strong tendency to resume the distorted form, unless prevented by proper bandages till the child walks when only it may be said to be secure from a relapse.”

700

Although not a physician, Sheldrake quoted extensively from the works of John Bell, an anatomist, assessing the causes and effects of the deformity on all tissues (bones, muscles, ligaments) of the limb. He refers extensively to the process whereby the cartilage model of each bone ossifies, a process well known even at that time, for the purpose of trying to determine when manipulative/splinting therapy would no longer be expected to work. Sheldrake also described and illustrated the specific bone deformities noting “the os navicular turning downwards and inwards from it (the astragalustalus) at a very acute angle.” Cure after 1 year of age was possible but only with more prolonged splinting. Even later however “when all the bones are completely ossified…when the patient has walked and thus perhaps increased the original deformity and distortion we are no longer justified in saying all such cases may be cured.” He also commented on how pressure differentially applied on the outer cartilage model with growth of the tarsal bones could both worsen deformation in the uncorrected state but also improve deformity with maintenance of splinting and progressive straightening. The deformity is also associated with contraction of the capsules and ligaments. These are slowly stretchable but should not be torn (“sprained”) with rapid forceful manipulation. In other words, the treatment process must not be painful. The best time to treat is within the first 2–3 months of life. After that, cure takes progressively longer. Consideration is also given to the muscles (invertors and plantarflexors) working more powerfully than their antagonists. Much of the two stages of treatment is designed to stretch contracted muscles and then continue to stabilize the foot until there is relative equality of muscle function in each direction to maintain correction. Sheldrake outlined the expected results of treatment based on the age of the child: 1. That species of club-foot, with which children are frequently born, may be perfectly cured, provided the cure is undertaken before the child begins to walk. 2. It is not impossible, that many cases may be perfectly cured, if undertaken after the patient has walked: though this must depend upon circumstances in particular cases which cannot be foreseen, and therefore was not advanced as a general fact. 3. If the cure is not attempted till the bones are completely ossified, it cannot be effected; and in many cases the deformity cannot be alleviated. The first group was aged less than 1 year, the second group in the 2–3-year-old range, and the third group older (although some could be cured even then to 10, 11, or 12 years of age). “Three distinct operations [not surgical] are requisite to cure this deformity: first, to reduce the bones to their natural position and natural form, if the patient’s age

7 Developmental Disorders of the Foot and Ankle

Fig. 7.9 An example of a clubfoot correction achieved with manipulation and taping is shown from Sheldrake’s book originally published in 1798. (Reprinted from Sheldrake, An essay on the Clubfoot, in A Practical Essay on Distortion of the Legs and Feet of Children, London, 1816)

has occasioned any malformation to take place; secondly, to produce extension of any muscle that has been actually contracted or seems to be from the position and consequent inactivity of the foot; and, thirdly, to keep the foot bound in its natural position till those muscles which have from the circumstances of the disease been weak and inactive perfectly recover their tone and power when, and when only, the cure will be complete.” Examples of good corrections obtained were illustrated in Sheldrake’s book published in 1816; one case is shown in Fig. 7.9.

7.5.14.3 Tendo Achilles Tenotomy Tendo Achilles tenotomy was done as early as 1784 by Thilenius of Frankfort, Germany, to correct talipes equinovarus. The procedure was done in a 17-year-old girl “afflicted from her earliest childhood” in whom “every kind of bandage boot, and ointment had proved no avail” such that “the foot was bent so much forward, and the tendo Achilles was so shortened that she walked almost entirely on the dorsum of the foot….” At surgery (done by Mr. Lorenz) “the tendo Achilles was divided. Immediately the heel descended 2 inches and the foot could be placed flat on the ground.” The foot was bandaged and “the healing process proceeded so favorably that [7 weeks later] the large wound had completely cicatrized without a single unfavorable symptom” leading to the observation “and now the girl can walk properly again, and like other people.” Additional cases were done by Sartorius (1806) in a 13-year-old and Michaelis (1809) in a 16-year-old, but it was not until 1816 when an intervention by Delpech of Montpelier, France, further advanced the surgical approach.

7.5 Pathogenesis and Pathoanatomy of Congenital Clubfoot

He divided the tendo Achilles in a 9-year-old boy. Infection followed with a prolonged course, but he eventually recovered being able to “walk and run without other deviation of the foot than a slight outward inclination.” The tendon had gained 2 inches in length. Brodhurst provides a detailed translation of Delpech’s discussion of the case including his physiologic reasoning about the approach [113]. The surgery represented a major advance since it was a subcutaneous operation (by which was meant a procedure not fully exposing the entire tendon). He considered that Delpech, based on his case description and rules for the subcutaneous division of tendons (written in his book De l’Orthomorphie, 1828), was the true originator of subcutaneous tenotomy even though he did no further cases. Case report. The boy was 9 years of age and had the foot deformity from birth. “When the foot was placed on the ground, the toes and the heads of the metatarsal bones were in contact with the ground; but in walking, the foot inclined inwards, so that the fourth and fifth metatarsal bones formed the basis of support. The tendo Achilles was very tense.” Delpech reasoned that the tendon, if released, would heal but that it would be better to keep the foot in the extended (plantarflexed) position for several weeks post-release to allow “intermediate substance” to partially heal prior to dorsiflexing the foot to the neutral position for final healing. By dividing the tendon, the muscle itself would not be affected. Operation involved two 1-inch incisions on either side of the tight tendon. The tendon was then divided “transversely from before backward without injury to the covering skin. Having satisfied myself that the foot could now be flexed on the left, I proceeded to fix the limb in the apparatus, at the same angle at which it was held before operation.” In spite of some local infection, by 28 days, “the re-union appeared to be sufficiently complete to permit extension (dorsiflexion) to be made, without risk of breaking the new substance.” The intermediate substance had gained one and one-half inches in length. In a few more days, the foot was brought to a right angle with the leg and maintained for an additional month in this position until full healing (at one-half the normal thickness) and 2 inches in lengthening had resulted. Delpech subsequently outlined his rules for subcutaneous tenotomy in a very clear and detailed fashion [146]. 1. The tendon to be divided should not be exposed. It should be divided a distance away from the point of entry of the knife through the skin. 2. After dividing the tendon, the cut ends should be brought into contact and held there by a suitable apparatus until early reunion is accomplished. 3. Reunion occurs by an intermediate fibrous substance, and gradual and careful extension (dorsiflexion) should be made to give required length to the shortened muscles until solidification takes place.

701

4. With extension complete, the limb should be held in this position until the new substance has acquired full firmness. Fifteen years later (1831), Stromeyer of Hanover, Germany, sectioned the tendo Achilles of a 19-year-old with clubfoot (acquired) using a percutaneous approach (to avoid suppuration and tendon necrosis) with the incision only as broad as the knife blade. Stromeyer applied Delpech’s principles, originally holding the foot in plantarflexion and then gradually dorsiflexing it a few degrees every few days reaching a right angle at 8 weeks. A walking boot was then applied. “At the end of 2 months, the foot was at a right angle and its outer edge horizontal. At the end of the 6 months, the position of the knee was entirely restored; and at the expiration of a year, his patient was enabled to lay aside his boot and iron and use an ordinary faced boot, in which he would walk securely without a stick.” Stromeyer followed with several more procedures in progressively younger patients and taught the procedure to other surgeons (including Little from England). The procedure was quickly adopted and by 1835– 1840 use of percutaneous tenotomy for correction of clubfoot, and other equinus and equinovarus deformities, was widespread in Europe.

7.5.14.4 W illiam Little and His Approach to Clubfoot Therapy The treatment of clubfoot by percutaneous heel cord lengthening was the procedure that effectively launched the operative component of pediatric orthopedic surgery. The first documented repeated and effective use of the procedure was by Stromeyer in Germany in 1831. The English orthopedic physician William Little who suffered from an equinovarus foot deformity due to infantile paralysis at 4 years of age visited Stromeyer and had the procedure performed on his own leg with significant improvement. He subsequently used the Achilles tenotomy procedure for clubfoot in England starting in 1837. Guérin popularized the technique in France and also expanded the use of tenotomy to deformities throughout the body [67]. Since that time there has been a series of pendulum swings with seeming regularity between operative and non-operative approaches to clubfeet. The patterns are remarkably similar. A relatively large number of patients respond well to a defined treatment, but some feet show partial correction or recurrent deformity and the treatment approaches shift. Little published his Treatise on Club-Foot and Analogous Distortions in 1839 [27]. Little treated clubfeet and other orthopedic disorders for several years and continued to write on, refine, and study his technique. He published a series of 18 lectures in Lancet in 1843–1844, many of which were devoted to clubfeet and published these in book form in 1853 Lectures on Deformities of the Human Frame [28].

702

He reviewed the treatment of infantile congenital talipes varus (ordinary clubfoot) in 1857 in a series of four articles in Lancet titled On Unnecessary Orthopaedic Operations [147]. The first three dealt almost exclusively with his management approach for clubfeet, summarizing his views on the treatment of congenital clubfoot after 20 years of experience with subcutaneous tenotomy. He was concerned that the procedure was being abused by too frequent use not only for clubfoot but also for deformities throughout the body. While the procedure was extremely valuable in the appropriate situations, he was concerned that it would fall into disrepute by inappropriate use. He stressed that surgical operation (tenotomy) for clubfoot was only a part of the treatment and that physiological, mechanical, and manipulative measures were invariably needed before and after tenotomy. In his lectures of 1843–1844, he advised that tenotomy should be used only after a trial of non-operative treatments. He defined the core of treatment however as being tenotomy of the tendo Achilles (gastrocnemii) and of the anterior and posterior tibial muscles, followed by mechanical treatment. He commented: “that operation in every case of congenital club-foot is neither necessary nor desirable.” In London, he said, operation was at that date (1857) invariably done with many strongly discouraging treatments without surgery. Little pointed out however that while tenotomy removed the deformity quickly, associated instrumental, manipulatory, and physiological treatment was “indispensable in every case.” “Where contraction is slight and curable without operation, a high development of the muscles of the calf, and consequently a more symmetrical limb may be obtained by cure without tenotomy than with its assistance. On the other hand, in severe cases, all the conditions-anatomical, pathological, and therapeutic-being different, the contrary obtains. In such cases, a more perfect limb results from the more speedy and effectual relief afforded by the operation.” Little felt that tenotomy was so helpful in clubfoot management since it addressed the major cause of deformity. His opinion of its origin was that “the muscles are the parts primarily involved, and that the displacement of the bones is entirely secondary” [77]. The offending antagonists were the adductors (tibialis anterior and particularly the tibialis posterior) and the gastrocnemii. Management was by mechanical extension (splinting) or tenotomy. Little described three groups of clubfoot patients [148]. “The whole of the third group, which is not the most numerous group, absolutely require operation” to correct deformity, “while the majority of the first grade and some of the second may be cured without operation if mechanical treatment and manipulations are done as carefully as they are when associated with surgery.” The removal of deformity is not the only object in view and “…those which have been cured without operation present a nearer approach to the anatomico-physiological standard….”. In those undergoing operation, “the

7 Developmental Disorders of the Foot and Ankle

state of the divided muscles and tendons is less satisfactory than when no operation has taken place, the divided tendons (as readily perceived in the calf) often remaining disproportionately long, the muscular bellies less developed, and consequent disposition to excessive walking on the heels (talipes calcaneus) being induced.” Manipulative treatment with splinting began on the first day or two after birth. When operation was needed in newborn cases, it was generally during the second or third months of age, often as early as 6 weeks of age. Non-operative treatment was often performed by parents, nurses, or attendants. Manipulation was done followed by splinting and bandages reapplied a few times a week. Skin care and the absence of pain were essential. Little clearly described the need to correct the varus before dorsiflexing the foot to correct equinus. “The more decided is the incurvation of the toes, the more necessary it is to direct attention to reduction of the varus to equinus (pointed toe) before attempting to bend the foot.” Stretching and manipulation of contracted parts maintains the joint free from rigidity, minimizes muscle atrophy, and prevents contractures of those muscles not being elongated. Stabilizing apparatuses are used “almost less as a means of forcing the lapsed part into a good or better position, than as a means of preventing its relapse into a bad position after use of manipulations.” “In the treatment of infantile varus without operation, as with its assistance, it is, I repeat, of first importance to obtain eversion of point of foot before attempting depression of heel-in short to correct the varus into the equinus.” Gentle manipulation and splinting toward neutral dorsiflexion is used. If persistent resistance is met, tenotomy is then done. Little emphasized early on the treatment was in two phases. First, the varus was corrected to an equinus only position. If necessary, tenotomy of the tibialis posterior and sometimes tibialis anterior tendons was needed to complete correction of the varus deformity. Only then was equinus addressed. Recurrence of deformity can occur from different causes. Little reports: “the great majority of the cases [in his book] have remained ‘cured’ but he discusses those that recurred. Many cases recurred because of failure to continue aftertreatment that involved continuing manipulation, active range of motion exercises and use of a “retentive apparatus.” Such cases could be straightened either by repeat surgery or by diligent resumption of “instrumental, manipulative and physiological treatment.” The causes of recurrent deformity were (i) failure of complete initial correction due to an insufficient number of tendons divided, (ii) insufficient straightening of non-operated tight structures by splints or manipulations, or (iii) premature discontinuation of the three aspects of aftercare. In a pathological sense, the tissues of the noncontracted region often grew more quickly than the operated contracted

7.5 Pathogenesis and Pathoanatomy of Congenital Clubfoot

tissues. He stressed the importance of not regarding a case as cured “so long as a trace of contraction remains.”

7.5.14.5 Late 1800s, Early 1900s In the later decades of the nineteenth century, many operative and non-operative methods were in use. Forcible correction of the clubfoot with wrenching apparatuses referred to as “machines” along with braces was widely used. In some centers patients were anesthetized for cast changes and forcible foot manipulation. Tenotomy was frequently resorted to balance the foot. Midfoot and hindfoot osteotomies were done as were full and partial tarsal bone resections. From the 1840s to 1900s, multiple tenotomies were the primary approach to clubfoot deformity surgery when manipulation and splinting/casting failed. Bigelow, summarizing clubfoot management in 1900 [79], refers to sectioning of the tendo Achilles for elevation of the heel (equinus), the tibialis anterior for the foot turned upon its outer edge (inverted), the tibialis posterior for adduction, the peroneus tertius and all extensors of the toes for the foot turned on its inner edge (everted), the peroneus longus and brevis for curvature of the internal border, and the corresponding long and short muscles for permanent flexion or extension of the toes. The plantar aponeurosis also might need to be released. Bigelow indicated that Guérin in France had approached the deformities by tenotomy as follows: for equinus–tendo Achilles, +/− flexor digitorum longus; for equinovarus– tendo Achilles, tibialis anterior and tibialis posterior, extensor digitorum longus and adductor of the great toe, +/− peroneus longus; and for valgus–peroneus tertius and peronei longus and brevis. Bigelow defined machines “as consisting of a series of pieces, each adapted to a corresponding detached portion of the skeleton, and united by joint the movements of which represent those of the articulations.” Screws or other mechanical contrivances would provide forcible restoration of the parts to a normal position. These methods were often used in the manipulative stage of management to induce correction by stretching prior to any surgery. Combined approaches were being used. Sometimes an apparatus was used (e.g., to correct varus) at which time the tendo Achilles tenotomy was done; at other times tenotomy was followed by application of appliances either immediately after surgery or a few days later when early healing was underway. More aggressive approaches were developed including fasciotomy, tendon releases, tendon transfers, and bony correction with enucleation of tarsal bone centers, wedge resection, and osteotomy. The feet were often straightened with these procedures, but persisting rigidity followed by discomfort was an increasing problem. Phelps of New York developed an open operative approach to club foot that gained extensive popularity in the late 1800s–early 1900s [149]. The approach is instructive about

703

the treatment philosophy of the era. His original description was in 1881 but his detailed article in 1890 reported on his first 161 cases. He indicated: “the distortion of the soft part has been out of all proportion to the deformity of bone.” Manipulation and non-operative treatment was warranted and often effective in the first year of life. His operative procedure was for persisting or recurrent (relapsed) deformity. He basically described the first detailed open surgical approach to clubfoot concentrating on soft tissue releases. The procedures were done under anesthesia. He indicated that one should “cut the contracted parts as they first offer resistance, cutting in the order of those parts which first contracted when the deformity was produced.” This was to be preceded by a vigorous firm manual reduction. The tendo Achilles was sectioned. Incision was then made from the internal malleolus one-quarter across the medial foot toward the inner level of the navicular and onto the sole of the foot. The next tenotomy was of the tibialis posterior tendon at its insertion, followed by releasing the abductor pollicis muscle/ tendon, plantar fascia, long toe flexors, and complete deltoid ligament of the inner malleolus and adjacent tarsals. That approach is understandable even today, but the forceful manipulation that followed is now considered to be excessive. Both manual forceful reduction and machine apparatus applied force were used. An example from Phelps’ article indicated the need to “use strong force after each tissue cut.” He stressed that “any amount of force can be applied to the heel and instep… turns up the screws applying any amount of force required.” In effect, the operator should not cease operating until the foot is completely corrected. The force can be sufficient for “breaking ligaments that he would find difficult or impossible to cut.” If the soft tissue releases and staged manipulations did not work for complete correction, then he proceeded with bony surgery as follows: linear osteotomy of the neck of the talus and resection of a bone wedge from the calcaneus (meeting with the talar osteotomy). The correction was then held with plaster of Paris. The summarizing terminology was remarkably aggressive: the surgeon “advances step by step in a proper order and need not stop or retreat until the deformity is overcome beginning with manipulation and ending with osteotomy if necessary.” Phelps indicated that he had done 161 cases and listed a further 181, mostly from Europe, reported using his technique totaling 342 cases. He reviewed all of his 93 patients (161 cases) with a case-by-case linear summary. The average age at surgery was 6.5 years and by the 4th month all were straight. Of all the cases reported, 171 required osteotomy. During that era, there were some who favored an even more vigorous use of bone surgery. In 1905 M’Kenzie of Toronto, Canada, reviewed numerous bony procedures proposed by others including osteotomies (navicular, tibia just above the ankle joint), enucleations of a single bone (cuboid, talus with removal of a wedge from the anterior os calcis),

704

enucleations of multiple bones (talus and cuboid, talus/ cuboid/navicular, navicular/cuboid), and even resections (of the head of the talus, a wedge of the outer half of the neck of the talus, two wedges from median tarsal and talocalcaneal joint, or a large midfoot trans-tarsus) [150]. M’Kenzie himself favored subcutaneous section of tendons, ligaments, and fascia with manipulation and splinting. He considered “removal of bone contraindicated and harmful.” Open incision was considered to “interfere with normal development of the foot.” Further overviews of the management principles used for clubfoot treatment of that era were provided by Bradford (1889) [151] and Hoke (1912) [152] followed by Elmslie (1920) [81] and Ombredanne (1927) [153] slightly later outline well the wide variety of approaches. Elmslie [81] demonstrated good understanding of the pathoanatomy and stressed that any surgery needs to be “based on the known pathoanatomy of the deformity.” There was clear recognition of the increased obliquity of the head and neck of the talus pointing downward and inward. This led to medial and plantar displacement of the navicular (scaphoid) and also pulled the cuboid medially at the midtarsal joint. The os calcis was in equinus and also oblique with its anterior end downward and inwards, its posterior end tilted upward and outward (toward the lateral malleolus), and the whole bone also twisted (inverted) with the outer surface coming to lie underneath. The major deformity was the astragaloscaphoid (talonavicular) subluxation with the capsule “converted into a dense fibro-cartilagenous mass binding the malleolus and the sustentaculum very closely to the tuberosity of the scaphoid (navicular).” He felt that “this mass resisted eversion to a corrected position more than the tibialis posterior tendon. The resistance to correction is formed largely by the astragolo-scaphoid (talo-navicular) capsule, the plantar fascia and the tendo-Achillis.” Manipulation was done under 1 year of age and as early as 1 month but under anesthesia. “It is perfectly easy to correct or overcorrect the deformity in a child under 1 year of age by simple manipulation under an anesthetic, repeated if necessary on two, three, or four occasions, with retention of the foot in plaster of Paris between manipulations.” The use of anesthesia certainly implies application of some degree of force to induce stretching. Also, in a resistant case (under anesthesia), “the foot is laid upon its outer side across and downward pressure made upon the heel and anterior part of the sole.” The initial manipulations are designed to correct the forefoot adduction returning the navicular and cuboid to their normal positions by stretching the talonavicular capsule, the medial fibrocartilaginous mass, and the medial plantar fascia. Only when this is corrected are efforts made to manipulate into dorsiflexion to correct the equinus. Achilles tenotomy was done only if manipulation into dorsiflexion failed since there was concern about leaving the tendo Achilles weakened. He sought also to externally rotate

7 Developmental Disorders of the Foot and Ankle

the foot (re-calcaneal position) that required a long leg cast with the knee flexed to a right angle. Response to relapse was relatively radical even by the standards of those using open reduction today. It needs to be recalled however that many relapses were treated at ages 1–5 years. As an example of the wide array of treatments used, Elmslie lists several approaches. These include: 1. Forcible correction at one or two sittings with fixation in plaster of Paris. 2. Repeated manipulations every few days with fixation in plaster of Paris. 3. Forcible correction with a Thomas’ wrench or other mechanical appliance. 4. Subcutaneous division of resisting structures with immediate reduction of deformity and plaster fixation. These include tenotomy or release of the tibialis anterior and tibialis posterior tendons, tendo Achilles, plantar fascia, and talonavicular capsule. 5. Subcutaneous division of structures as listed in 4 above without immediate correction and casting; once skin has healed corrective shoes/splints and muscle strengthening are used to encourage slow progressive correction while (presumably) limiting stiffness. 6. Division of “all the structures of the sole and inner side of the foot into the mid-tarsal joint,” a procedure that could easily damage blood vessels and nerves in the process. 7. Bony procedures including removal of the cuboid, removal of a wedge of bone from the outer side of the foot, removal of the talus, transverse division of the tarsus, transverse division of the calcaneus, and even the talar neck osteotomy to correct the oblique anterior deformity. 8. Lateral transfer of the tibialis anterior tendon. It became evident even then and, certainly to our current perspective, that most of these approaches addressed gross correction of deformity rather than pain-free, flexible function. In the slow learning process regarding clubfoot therapy, approaches in numbers 1, 3, and 6 are no longer considered and manipulation (as in number 2) is not done under anesthesia or with any forceful maneuvers such as bending the foot over wedges. It is essential to minimize scarring and preserve joint function without adhesions, transarticular fusions, or severe collapsing deformity of joint surfaces. Tenotomies as in numbers 4 and 5 are selectively helpful and tendon transfer, as in number 8, remains a valuable tool. Bony procedures are occasionally needed as in number 7 after 5 or 6 years of age. Removal of tarsal bones for clubfoot correction (aside from severe arthrogryposis) is contraindicated regarding any semblance of normal function. Calcaneal osteotomy is safe and does not affect growth centers. Selective midtarsal wedge procedures are occasionally needed, but

7.5 Pathogenesis and Pathoanatomy of Congenital Clubfoot

talar neck osteotomy is not done, since it almost invariably leads to avascular necrosis. Ombredanne [153] recognized that complete manipulative correction could be gained in the first year of life. He stressed obtaining manipulative overcorrection into a position of pronation to allow for a good long-term result. “It is possible to obtain the complete reduction practically always in the early days of life.” He also stressed necessity of retaining the correction once achieved. This was done by “mobilization” (active and passive stretching) “four times daily” and holding the overcorrection in a light brace or splint. This splinting was used night and day for 3 months after correction was gained and then at night only up to 1 year of age and even for 15 days per month after that. If there was incomplete correction or recurrent deformity before the time of bony malformation, soft tissue tendon/ ligament surgery was resorted to. Varus-adduction deformity could be corrected with medial plantar releases (cutting muscles and ligaments) and internal ligament release at the tibiotarsal joint. Once the varus was corrected, tendo Achilles tenotomy would be helpful. This approach could be effective in the 1–4-year-old range. After 4 years of age, bony deformity had developed to such an extent that soft tissue surgery alone was felt to be ineffective. He condemned the use of forceful manipulation using mechanical devices like wedges, levers, or the clubfoot wrench of Thomas as being far too traumatizing. He referred to this as “tarsal crushing” which was destructive of normal bone and cartilage development. It clearly induced bony malformation, cartilage joint surface destruction with fibrosis, and growth arrest. He also advised against enucleation procedures where central tarsal bone was removed to allow for easier manipulative correction. Any correction achieved would be due to cartilage model and cartilage surface collapse with growth arrest and degenerative arthritis soon to follow. The final bone surgery recommendation in this era was the use of wedge resections that required midfoot- hindfoot osteotomies to correct deformity to a reasonably straight foot, although there were consequences on growth if done in the first decade. He advised against the use of talectomy.

7.5.14.6 Return to Manipulation and Casting (i) Kite. In response to this treatment situation characterized by relatively forceful manipulation and excessive surgery including many bony interventions in young children, correction by serial plaster casting was strongly advocated by Hiram Kite of Atlanta, USA, beginning in the late 1920s. Kite outlined the principles and methods of what he referred to as the “non-operative treatment of congenital club-feet” in a series of papers beginning in1930 [154–158] and continuing as late as 1970 [159].

705

His first major article outlining and illustrating the treatment and reviewing the first 100 cases was published in 1930 [154]. The non-operative method involved no surgery and no manipulation under anesthesia. Over the previous years at his hospital, 176 clubfoot cases were treated by operation, 15 by manipulations under ether anesthesia, and then his first 100 non-operative cases. The non-operative treatment was best in infants but older children also benefited; 90% of clubfoot cases were treated successfully by this non-operative method. In decrying forceful manipulation he stressed: “since flexibility is as desirable as anatomical correction, it is most important that the feet be treated by a method that will not injure the articulation surfaces.” First phase: The first step was correction of the varus deformity. This involved not only manipulation and casting but also wedging of the cast on two or three occasions to further correct position, after which a new cast was applied and the wedging continued. The hindfoot was stabilized with one hand that everted the varus heel (calcaneus) posteriorly, stabilized the anterior-lateral calcaneocuboid region, and abducted the forefoot (metatarsals) in relation to the hindfoot. While the toes and the metatarsals were flattened between the opposite hand thumb and index finger, “at the same time the foot is everted slightly in order to turn down the great toe and to set the sole more nearly flat on the ground.” Final molding was done as the cast set. The cast was then converted to a long leg cast with the knee flexed to a right angle. This was done to prevent slippage and continue stability after wedging. Kite stressed that correction of varus and adduction was done with the forefoot in equinus in relation to the posterior foot and to the ankle joint that was also held in its original equinus position. If an attempt is made to flex the ankle joint dorsally before the varus deformity is corrected, the foot will yield in the transverse tarsal joint and not in the ankle joint, thus producing “rocker bottom” which is very undesirable. By keeping the forefoot in equinus as it was abducted, the scaphoid (navicular) was brought around from the medial side of the head of the astragalus (talus) until it was directly in front of it. The varus deformity can never be corrected unless the scaphoid can be brought around in front of the head of the astragalus. The restoration of the normal alignment between these two bones was considered to be essential to correction of any deformity of the foot. If a clubfoot was dorsally flexed while being abducted, the scaphoid would ride up on the medial side of the head of the astragalus and never get around in front of it. Cast wedging. Cast wedging to improve the extent of forefoot abduction was done every few days to further correction. After two or three wedgings, the cast was changed. The lateral abduction wedge was cut circumferentially from the level of the anterior end of the os calcis laterally and the

706

scaphoid medially and the second cut from the cuboid to the medial apex. Once the wedge was removed, the forefoot was abducted and the new position held with a figure of eight plaster bandage. Abduction increased over several weeks with the forefoot eventually abducted 30–40° beyond the midline. Once the varus heel corrected to neutral under the tibia, it was stabilized to prevent valgus. Second phase: Correction of equinus deformity was then done. Manipulation and casting were done and wedging of the cast into dorsiflexion was also done on two or three occasions, after which the cast was changed and the process continued. A similar cast was applied “with the forefoot held in abduction while pressure is made upward and inward under the calcaneo-cuboid joint.” With subsequent casts the forefoot was returned to the midline. Wedging in dorsiflexion. The circumferential proximal cut was at the level of the malleoli with the distal dorsal wedge over the dorsum of the foot. With subsequent wedging the whole foot was brought increasingly into dorsiflexion and the cast held with plaster. As the gastrocnemius stretched, the body of the astragalus tilted back into the ankle mortise. Dorsiflexion above neutral to the normal range was sought. Third phase: The third phase involved retention in long leg cast in an overcorrected position. The foot was held in extreme dorsiflexion for 8–12 weeks. Fourth phase: The fourth phase concerned the aftertreatment. Kite stressed two aspects for the fourth phase. One was that “braces are never needed if the architectural deformity has been corrected.” The other was attention to the type of shoes. His method required (for the most part) use of straight last high top shoes for several years. Reverse last shoes were used where some forefoot adduction persisted but “a normal shoe should not be worn for several years.” For 6 months the shoes were worn for 24 h with once-daily series of dorsiflexion stretching. Recurrences (usually of varus) warranted recasting. The average time of treatment was 25 weeks (6 months). Recurrences were corrected in 6 months (on average). A good or excellent result was achieved in 85, including 12 recurrences. [It is important to recognize that several decades after Kite initially described his technique, those who still made efforts to correct clubfeet initially by manipulation and casting had often abandoned (at least in North America) the wedging component of management and relied on multiple cast changes alone.] In 1939, based on over 400 patients with clubfeet personally treated, Kite made “a plea for conservative treatment of congenital club-foot” by calling attention to certain mechanical and pathological principles involved [157]. He concluded: “most club-feet can be successfully corrected by a series of plaster casts and wedging without the use of anesthetics, forcible manipulations or operative procedures and with far better results.” He called his approach “non-operative

7 Developmental Disorders of the Foot and Ankle

treatment of clubfeet.” He felt that much of the eventual stiffness reported after forcible manipulations sometimes done under anesthesia was due to ankylosis of the tarsal joints with fibrous and even osseous adhesions crossing from the cartilage of one bone to another. Examples of transphyseal bone bridges of the distal tibial epiphysis, caused by forceful foot manipulation, were also described. His program was based on the use of early casting, beginning within a few days of birth. Long leg casts, applied without anesthesia or forcible manipulations, were used. Kite recognized that a clubfoot deformity was divided into three parts: adduction of the forepart of the foot, inversion of the heel, and equinus. The forefoot is adducted in relation to the hindfoot. The calcaneus is rotated inward under the talus causing the entire foot to assume an inverted position. The equinus deformity is present at two sites. The forefoot is plantarflexed in relation to the posterior foot (forefoot equinus) and the entire foot is plantarflexed in the ankle joint (ankle equinus). “To untwist this foot, the adduction of the forepart of the foot must be first corrected, and then the inversion of the heel can be corrected. Both deformities must be completely corrected before bringing the foot into dorsiflexion.” Kite indicated that “each of these three deformities must be completely corrected, and in the order mentioned.” (a) Correction of adduction deformity. Forefoot adduction is corrected with equinus and inversion persisting. Kite indicates that this manipulation is designed to move the medially displaced navicular laterally into its normal position in front of the head of the talus. With this maneuver the cuboid also slides laterally to normalize its relationship to the calcaneus. “The forepart of the foot must be brought out into abduction until the cuboid is in front of the calcaneus and the navicular in front of the head of the talus; then the foot must be abducted a little further until it is in a mild flat-foot position.” To get true forefoot correction, as the plaster cast sets “the foot must be molded pushing the fore part of the foot laterally and the lateral border of the foot in the direction of the calcaneocuboid joint medially.” He cautions that “if an attempt is made to dorsiflex the foot before the navicular is in proper positon, the latter will be forced up on the medial side of the head of the astragalus (talus).” He warns that forefoot adduction recurs if not completely corrected. He also warns about overcorrection into a flatfoot position. (b) Correction of inversion deformity. The inversion deformity is recognized on the anteroposterior radiograph where the anterior end of the calcaneus is under the head of the talus. In the normal the two diverge with the calcaneus rolled outward with the talar head and neck in line with the first and second rays and the long axis of the calcaneus in line with the fourth and fifth

7.5 Pathogenesis and Pathoanatomy of Congenital Clubfoot

rays. The inversion is corrected either by manipulating the calcaneus and holding it into valgus by the cast or by wedging the cast with a lateral wedge-circumferential cut and eversion positioning held by a plaster overwrap. Once the anterior ends of the calcaneus and talus are separated, it is time to begin correcting the equinus. If dorsiflexion is started before correction of adduction and inversion, a “rocker-bottom” midfoot deformity can be created. An anteroposterior x-ray of the foot helped confirm when the first two deformities had been corrected. (c) Correction of equinus deformity. Once the two other deformities are fully corrected, the equinus is addressed by gradually bringing the foot into dorsiflexion. As progressive dorsiflexion occurs, the body of the talus is brought into a normal relationship under the tibia at the ankle joint. “The foot must be brought into a mild flatfoot position and then brought up into dorsiflexion in this flat-foot position for a week or two. After, the foot must be held in dorsiflexion in the mid-line to stretch the Achilles tendon.” “If the dorsiflexion is done with the calcaneus not fully corrected from its inverted position, it fails to slide in relation to the talus, a rockerbottom/false correction occurs, and the talocalcaneal joint becomes even more rigid.” A paper by Kite reviewing his first 200 patients outlined the patient population concerning age, sex distribution, and length of time in casts [155]. Kite’s papers lacked even semiquantitative summation of results. In a 1939 paper on “over 400” patients he had treated, he indicates that 90% were corrected by plaster casts and wedgings while the remaining 10% (chiefly older patients or those treated previously by operative procedures) needed operative Hoke clubfoot stabilization. In 1963 he continued to support his method but again produced no breakdown of cases by age, length of treatment needed, etc. By that time over 1500 patients had been treated in his clinic and the method had been widely adopted elsewhere. Most prolonged treatments or treatments needing additional work were due to failure to correct adduction and in particular calcaneal inversion before attempting dorsiflexion. His papers provide little exact information on the timing of the various phrases of treatment, but he did indicate that the first 4 weeks of treatment were the most important. Overcorrection in valgus was not desirable. He cautions against correcting the adduction deformity with the foot in any position other than plantarflexion. Abduction is continued until the forefoot is slightly beyond the midline. He repeatedly comments in his papers to “be sure that the adduction and inversion deformities are thoroughly corrected before beginning dorsiflexion.” When the foot is fully corrected, it is held in full dorsiflexion for about 8 weeks, changing the cast every 2 weeks.

707

After the feet have been corrected and the casts discontinued, no special shoes or braces are used for walking. If there is still enough deformity to warrant use of special shoes then “the deformity has not been satisfactorily corrected.” Kite specifically noted that “it is the responsibility of the orthopaedist- not the shoemaker or the bracemakerto correct the foot.” An early report on the Kite conservative treatment regimen by Sell was supportive of the method [160]. In 70 treated cases of clubfeet less than 1 year of age, the rate of recurrence was 11%. The average age of patients at beginning of treatment was 4.4 months and the average length of time under treatment was 6 months with casts changed completely every 2 weeks. Sell also strongly disparaged treatment by “forcible manipulations under anesthesia or open operation.” He stressed that restoration of normal or near-normal foot alignment was important but that “restoration and maintenance of foot function” were also essential for a good result. Forcible manipulation and operative surgery often led to stiff, inelastic, painful feet even if anatomically they were straight. Short leg casts with a skin adhesive (benzoin) were used. Adduction (varus) and inversion were corrected first. Adduction was considered to be the hardest component to correct. Dorsiflexion correction was made with a wedge removed anteriorly, closed, and held with plaster. Once corrected a longer period of casting was maintained. “Club foot shoes cannot and will not correct deformities, nor prevent their recurrence.” (ii) Ponseti. Ponseti of Iowa, USA, was subsequently a strong advocate of early casting and only minimal surgery if needed to correct deformity. His investigations led to an understanding of the deformity stressing the presence of a medial column plantarflexion producing an early cavus deformity that required forefoot correction by supination as well as abduction. He initially supinates the foot to align it and the calcaneus with the navicular. This supination is maintained as talonavicular position is improved by abduction toward and then beyond the neutral plane. [It is Ponseti’s recognition of medial foot column plantarflexion and the need to invert the entire foot to align all metatarsals in the same plane in the initial phase of manipulation that differentiates his method of manipulation from that of Kite.] Once adduction and inversion have been corrected, however, percutaneous Achilles tenotomy is done to correct equinus if it does not respond quickly to serial manipulation into dorsiflexion rather than the much more prolonged stretching with repetitive casts and wedging as used in the Kite technique. Once the foot has been corrected, great emphasis is placed on maintaining the correction with full-time bracing for 3 months and nighttime bracing for 2–4 years.

708

Ponseti began developing his approach to treating clubfeet in 1948. The protocol remained remarkably steady from his initial detailed report in 1963 [114] to those in 1980 [161] and 1992 [162]. In spite of the very good results reported, the approach became widely accepted and practiced throughout North America and elsewhere only in the mid-1990s although many groups had continued to use the Kite manipulative approach, although often with little enthusiasm and imperfect attention to detail, until opting for more extensive surgery. Ponseti, similar to Kite, stressed that forceful manipulation; medial, posterior, and lateral soft tissue surgical releases; and osteotomy and arthrodesis had met with “limited success” with relapses, overcorrections, and in particular “considerable stiffness” occurring. Surgery was often designed to obtain anatomical correction of the clubfoot deformity, but the sequelae listed above were troublesome. In 1980, with 30 years of experience using the technique, Laaveg and Ponseti commented that “our goal was then to improve the functional results and to obtain plantigrade, mobile feet in the shortest time possible [161]. Once we no longer tried to attain a perfect anatomical result, few soft- tissue releases were done.” First Review, 1963 The first review assessed 94 clubfeet in 67 patients treated from 1948 to 1956 with a 5–12 year follow-up [114]. Eliminated were complex cases with other deformities on the severe side and mild cases corrected by simple manipulations using 1–3 plaster casts. The deformities assessed were defined as “severe, although many variations in the degree of rigidity of the feet were present.” Method Gentle manipulations of the feet are done without use of anesthesia. Well-molded long leg casts are applied every 4–7 days. The short leg cast is applied first from toes to just below the knee and this is then converted to a long leg cast to upper thigh with the knee flexed to a right ankle and the leg externally rotated to correct tibial torsion. Application of the manipulative technique is based on Ponseti’s understanding of the multi-planar deformity. “The cavus deformity must be corrected with the first cast. Since the cavus deformity is related to the pronation of the forepart of the foot with respect to the hindpart, the cavus is corrected by placing the fore part of the foot in supination in proper alignment with the hind part” [114]. The forefoot is thus supinated first and then is moved laterally into abduction. To correct the inversion, the thumb is placed on the lateral aspect of the head of the talus as the fulcrum with lateral direction (outward) pressure then placed on the first metatarsal and first cuneiform. When the navicular and cuboid are displaced laterally, the anterior part of the calcaneus is displaced outward and upward to correct the varus of the heel. Care is taken not to pronate the foot or place direct pressure on the calcaneus. Four to five cast changes generally correct the

7 Developmental Disorders of the Foot and Ankle

adduction and inversion components. The equinus deformity is corrected next by dorsiflexing the foot with the heel in neutral or slight valgus. If progress is slow after 2–3 casts, percutaneous tendo Achilles tenotomy is done. Ponseti originally used general anesthesia for the tenotomy but later described use of local anesthesia in the clinic setting. A long leg cast is then applied for 3 weeks with the foot in maximum dorsiflexion. From five to ten casts were worn for an average time of 9.5 weeks for full correction. Percutaneous tenotomy was done in 74 of 94 feet (79%). Denis Browne splints were used full time for 3 months and at night for an average of 21 months. Recurrences were seen in 53 feet (56%), usually at an average of 2 1/2 years after treatment. These were attributed to premature discontinuation of Denis Browne splinting by the family and the finding of more severe initial involvement. Some patients eventually needed a tibialis anterior tendon transfer to the third cuneiform since supination was a greater problem with recurrence than equinus deformity. A second recurrence was noted in 18% at an average of 3 years of age. A third recurrence was seen in 10%. Assessment included anteroposterior foot x-rays measuring the talocalcaneal long axis angle (30° considered normal). On lateral foot x-rays, varus was measured by the angle between the long axis of the calcaneus and that of the first metatarsal. Heel varus, metatarsus adductus, cavus deformity, equinus, and tibial torsion were almost always corrected. The long leg cast with the knee flexed at a right angle was essential to correct the tibial torsion. Bone surgery was never needed. Other than percutaneous Achilles tenotomy and tibialis anterior tendon transfer, other procedures only infrequently needed were open tendo Achilles lengthening, medial release operations, and capsulotomy of the midfoot Lisfranc joint. The results were 71% good, 28% slight residual deformity, and 1% poor. Second Review, 1980 A subset of clubfoot patients treated between 1950 and 1967 were assessed long term involving 189 clubfeet less than 6 months old when first seen with longer-term analysis done in 104 feet (70 patients) [161]. Mean age at follow-up was 18.8 years. Manipulation began as soon as possible after birth. It was again stressed that abduction distal to the talus needed to be accompanied by eliminating pronation by inverting (supinating) the forefoot. Casts were changed weekly. The anterior part of the foot was never everted (pronated) while being abducted. Percutaneous tendo Achilles tenotomy (under local anesthesia) was the final corrective approach before applying the last cast with the ankle in full dorsiflexion for 3–4 weeks. A Denis Browne bar with the shoes/feet in 70° external rotation was worn constantly for 3 months and at night only for 2–6 years. Relapses were treated with manipulation and 2–4 long leg

7.5 Pathogenesis and Pathoanatomy of Congenital Clubfoot

casts changed every 2 weeks plus occasional tendo Achilles lengthenings. Tibialis anterior tendon transfers were done at 2.5 years of age for continuing inversion deformity. Of the 104 clubfeet, 13 had manipulation and plaster casts only, 42 had plaster casts and percutaneous tendo Achilles lengthenings, 48 had tibialis anterior tendon transfer to the third cuneiform (sliding it laterally under the flexor retinaculum), and 1 had a tibialis posterior tendon transfer to the dorsum of the foot through the interosseous membrane. Seventeen had additional soft tissue procedures. The mean age at beginning of treatment was 6.9 weeks, mean number of casts for initial treatment was 7 over a mean duration of 8.6 weeks, and mean time of Denis Browne bar use was 50 months. Relapses: Fifty-five patients (53%) had no relapse, 49 (47%) had one relapse (mean age 39 months), 25 a 2nd relapse (53 months), 10 a 3rd relapse (63 months) and 3 a 4th relapse (77 months). Functional rating (mean) was 87.5 for all patients with 93.1 for casts alone, 92.4 for casts and tenotomy of tendo Achilles, and 80.5 for tibialis anterior tendon transfer. Results were excellent in 54% (90–100 points), good 20% (80–89), fair 14% (70–79), and poor 12% (