Secondary Logo

Journal Logo

Evaluation and Classification

Classification of Legg-Calvé-Perthes Disease

Kuo, Ken N. MD*,†,‡; Wu, Kuan-Wen MD; Smith, Peter A. MD§; Shih, Shu-Fang PhD; Altiok, Haluk MD§

Author Information
Journal of Pediatric Orthopaedics: September 2011 - Volume 31 - Issue - p S168-S173
doi: 10.1097/BPO.0b013e318223b50e
  • Free


Legg-Calvé-Perthes Disease (LCPD) is one of the least understood pediatric orthopedic problems in the last 100 years. The condition was originally reported by Henning Waldenstrom of Stockholm and Arthur T. Legg of Boston in 1909. Then, later, in 1910, it was reported in journals by Legg1 in English, Perthes in German, and Calvé in French. The name has evolved from coxa plana to the present Legg-Calvé-Perthes Disease.2,3

There are reasons to have a standard classification. They are of prognostic value and can be used for study comparison and for learning natural history. Most classifications aim to gauge the severity of the disease, with the aim of predicting future outcome.

As it has an obscure pathoetiology, classifications vary depending on the focus of importance. A number of researchers have attempted to classify the disease; however, it is imperative that the classification can predict or relate to the final outcome. Despite some previous attempts to classify the LCPD, it was not until 1971 when Catterall4 published a classification that became widely used in the literature.

The purpose of this article is to review the popular classifications since 1971 and to look into their use in the literature.


From the PubMed database, we identified 1216 publications of studies in English literature using the keywords “Legg Calvé Perthes disease” or “Perthes disease” or “LCPD” or “LCP disease” from 1970 to 2009. We identified the most commonly quoted classifications. All abstracts of these 1216 publications were reviewed and screened. The criteria were articles with good delineation of classifications, good outline of the treatment, and well-defined treatment results. The most widely used classifications are the Catterall classification, the lateral pillar classification, and the Salter-Thompson classification. Sixty-one articles fit the original criteria. A second screening of these 61 full articles, limiting to prospective and retrospective study articles with an outcome evaluation using Mose or Stulberg classification, and excluding articles with missing information on classification or treatment information, yielded 20 articles for final analysis.


The first widely used classification was a study published in 1971 by Catterall.4 He based it on a series of 121 patients with 133 hips with a diagnosis of LCPD at the Hospital for Sick Children, Great Ormond Street, in London. Among them, 46 hips were untreated. The treated cases were carefully matched with controls and it was found that there was little difference in the overall results between the groups. He noted different radiographic courses among the different degrees of involvement of the disease. He used Sundt5 outcome criteria (good, fair, and poor) in his study. The description of the radiographic pathology was very detailed. He classified the LCPD into 4 groups on the basis of anteroposterior and lateral radiographs of the hip with additional arthrogram in 59 cases:

  1. Group 1 (Fig. 1): Only the anterior part of the epiphysis is involved without collapse, and complete absorption of the involved segment occurs without sequestrum formation. Metaphyseal changes are unusual in early stage. The course of the disease seems to be absorption of the involved segment, followed by regeneration.
  2. Group 2 (Fig. 2): Rather more anterior part of the epiphysis is involved with collapse of the segment after the initial phase of absorption. In the anteroposterior radiograph, the sequestrum appears as a dense oval mass with viable fragments on both medial and lateral sides; therefore, the epiphyseal height is maintained. On the lateral radiograph, the sequestrum is separated posteriorly from the viable fragments by a characteristic “V”.
  3. Group 3 (Fig. 3): In this group, only a small part of the epiphysis is not sequestrated. The anteroposterior radiograph showed “head in head” in the early stage. Central sequestrum collapse appears in later stage, with small normal segments on the medial and lateral sides. The lateral fragment is frequently small and osteoporotic. Only a very small portion of the posterior part is not involved. Metaphyseal changes are more generalized and extensive with broadening neck.
  4. Group 4 (Fig. 4): The whole epiphysis is completely sequestrated and collapsed. Early loss of the height between the growth plate and the acetabular roof indicates head flattening. With displacement of the epiphyseal fragment anteriorly and posteriorly, it produces a mushroom-like appearance of the head. There is no posterior viable portion on the lateral radiograph. Metaphyseal changes are also extensive.
Catterall group 1, anteroposterior and lateral view of the right hip showed that only anterior part of epiphysis is involved without collapse and sequestrum formation.
Catterall group 2, anteroposterior and lateral view of the right hip showed rather more anterior part of epiphysis involvement with collapse of the segment. The sequestrum appears as a dense oval mass with viable fragments on both medial and lateral sides.
Catterall group 3, anteroposterior and lateral view of the right hip showed that only a small part of epiphysis is not sequestrated. Central sequestrum collapse appears in the later stage. The lateral fragment is small and osteoporotic. Only a small portion of posterior part is not involved.
Catterall Group 4, anteroposterior and lateral view of the right hip showed that the whole epiphysis is sequestrated.

He also mentioned that the age and sex of the children have a great impact on the prognosis. The Sundt5 outcome classification showed positive relation to the Catterall classification in the article.

In 1984, Salter and Thompson,6 after investigating 1057 children with 1264 involved hips from 4 hospitals (Hospital of Sick Children in Toronto, 603 patients; Shriner's Hospital, Mountain Unit, 164 patients; Shriner's Hospital, LA Unit, 169 patients; University Hospitals of Cleveland, 121 patients) discovered that the extent of subchondral fracture correlated precisely with the subsequent extent of maximum resorption of the femoral head. They determined that subchondral fracture (Fig. 5) is of prognostic significance in predicting the eventual involvement of the femoral head. After classification using Catterall grouping and checking the final outcome, they proposed a 2-group simple classification on the basis of the extent of femoral head involvement of subchondral fracture.

The lateral view of this right hip showed the extent of subchondral fracture in anterior part of the femoral head.
  1. Group A: less than half of the head involved.
  2. Group B: more than half of the head involved.

In 1992, Herring et al7 published a new classification called the lateral pillar classification. He based the study on 93 hips in 86 patients with radiographic follow-up to skeletal maturity. The final outcome was determined according to Stulberg et al8 classification of the shape of the femoral head. He divided the femoral head into 3 parts lateral, central, and medial pillars. The determination is based on the true anteroposterior view of the hip. The greatest involvement of the lateral pillar in radiograph was used for classification.

  1. Group A (Fig. 6): There is no involvement of the lateral pillar with the full height maintained. The medial and central pillars may have lucency and collapse.
  2. Group B (Fig. 7): There is >50% of lateral pillar height maintained. The lateral pillar may have some lucency with maintenance of the bone density at the height between 50% and 100% of the original height of the lateral head. The height of central and medial pillars may be partially or completely decreased.
  3. Group C (Fig. 8): There is <50% of lateral pillar height remaining. The lateral pillar is more lucent, and the remaining bone is <50% of the original height. The degree of decreasing height of medial and central pillars may vary.
Lateral pillar group-A, this anteroposterior view of the right hip showed central lesion without involvement of the lateral pillar. There were no density changes and no loss of the height of the lateral pillar.
Lateral pillar group-B, this anteroposterior view of the right hip showed lucency in the lateral pillar with some loss of height, but not exceeding 50% of the original height.
Lateral pillar group-C, anteroposterior view of right hip showed more lucency in the lateral pillar and >50% loss of height.

The final outcome of this group of patients using the Stulberg classification showed that Group A had a uniformly good outcome. The Group B had a good outcome if the patients were <9 years old, but was less favorable if patients were older than 9 years at the time of onset of disease. In the group C, majority of the femoral head became aspherical in both age groups.

Herring et al9 later modified their classification and published an additional article in 2004 adding an intermediate group B/C border (Fig. 9), which includes a femoral head with a thin or poorly ossified lateral pillar and with a loss of lateral pillar height at 50%. With the addition of this group, there was greater interobserver and intraobserver reliability than previous classification systems.

Lateral pillar group-B/C, this anteroposterior view of the right hip showed a very narrow lateral pillar (2 to 3 mm wide) that was 50% of the original height with some lucency and depressed relative to the central pillar.

For the early sign of LCPD and its progress, scintigraphic evaluation is another method of studying LCPD. It is cited that radiographic appearance of the avascular lesion is already late in diagnosis of LCPD. Fasting et al10 from Norway first described scintigraphic evaluation of the severity of LCPD in 1980 after evaluation of 20 hips in 18 patients. They classified LCPD into 4 grades:

  1. Grade 1: decreased activity involving quarter of the epiphysis.
  2. Grade 2: decreased activity involving half of the epiphysis.
  3. Grade 3: decreased activity involving 3-quarter of the epiphysis.
  4. Grade 4: decreased activity involving whole epiphysis.

In 1993, Conway11 published a more elaborate study by scintigraphy. He studied 31 patients with various durations of symptoms from 2 weeks to 18 months. He divided the group into 2 tracks: A track with recanalization and B track with neovascularization. With this, Conway has observed 4 stages of healing process of each track. In A track, he staged the healing process in sequence as whole head, lateral column, anterior and medial extension, and complete revascularization. In B track, the 4 stages are whole head, base filling, mushrooming, and complete revascularization.

For the outcome study, Moss12 of Denmark published the femoral head spherical measurement system in 1980. One year later, Stulberg et al8 proposed an outcome evaluation that has become the most commonly used standard. They presented a long-term follow-up study of 2 groups of patients: a group of 88 patients (99 hips) followed for 40 years from 3 hospitals and a group of 68 (72 hips) patients followed for 30 years from 1 hospital. They classified the outcomes into 5 classes according to the radiographic appearance at maturity. This can be summarized according to the congruency of femoral head.

  1. Spherical congruency (class I and II hips): arthritis does not develop.
  2. Aspherical congruency (class III and IV hips): mild-to-moderate arthritis develops in late adulthood.
  3. Aspherical incongruency (class V hips): severe arthritis develops before 50 years of age.

The shape of the femoral head determined symptoms and age of development of degenerative arthritis.

In 61 articles reviewed with good delineation of classifications using Catterall, Salter-Thompson, or lateral pillar classification, with good outline of the treatment and well-defined treatment results, 2 were prospective, 45 were retrospective, 2 were preliminary, and 12 were not well stated. It showed greater use of the Catterall classification because it superseded others by 20 years. Before 1993, there were 34 articles using the Catterall classification and 8 after 1993. There were 7 articles using only the lateral pillar classification since its publication in 1992, whereas 10 articles used both lateral pillar and Catterall classifications at the same period. There was 1 using the Salter-Thompson classification and 1 using the Waldenstrom classification.

In those 207,13–31 articles reviewed, which were only limited to prospective and retrospective studies with the outcome evaluation limited to Mose or Stulberg classification and which excluded missing information on classification or treatment information, the Stulberg outcome classification versus Catterall classification (Fig. 10) or lateral pillar classification (Fig. 11) showed comparative results except for the lateral pillar classification, which showed more precise differentiation. If the patients were divided by age groups of 9 years and above and below 9 years against Mose rating (Fig. 12) and Stulberg (Fig. 13), it showed a general trend of the older children at onset having worse results. There are a smaller number of poor results in Stulberg rating compared with Mose rating due to the different criteria measuring the enlargement of femoral head.

The outcome of Catterall classification by Stulberg outcome classification showed a positive relationship.
The outcome of lateral pillar classification by Stulberg outcome classification showed more precise differentiation of a positive relationship.
Mose outcome rating versus age groups showed better results in the group below 9 years old.
Stulberg outcome rating versus age group showed a general trend in older age group at onset to have worse results.


After publication, the Catterall4 classification gained wide acceptance for study of LCPD. There is a shortcoming as it came under some criticism for low interobserver reliability. However so far, it is still a popular method used in the literature. The Salter and Thompson6 classification is simple and easily understandable. They rely on the early sign of subchondral fracture in early radiographs, which is not always available. As the latest classification scheme, the Herring et al7,9 lateral pillar classification has the highest reputed reliability among observers. The Herring classification is gaining popularity among later studies in the literature.

There are problems with the radiographic classification. The most significant problem is inconsistent x-ray position and the lack of standard in x-ray technique in many hospitals. The radiographic classification is often based on the picture of fragmentation stage that may cause external rotation of the hip when radiograph is obtained. Moreover in many studies, there was no mention of how radiographs were obtained, whether in a standing or supine position.

In the radiograph of the fragmentation stage of a skeletally immature hip, it is often difficult to measure or to visualize the pathology. The progressing and timing of each stage may vary in different patients. The hip may be in a late stage when the disease is discovered. Scintigraphy is more sensitive to early discovery, but it is not practical in clinical practice. The plain radiograph is still the best convenient way for the classification of the LCPD.

For the outcome evaluation, Herring et al have modified the Stulberg et al8 classification description to obtain a better interobserver reliability. They clarified the definition of Stulberg to define Stulberg I as a normal-appearing hip, whereas Stulberg II as round and fitting within 2 mm of a circle on both anteroposterior and lateral radiographs. A Stulberg class III hip is out of the circle by >2 mm on either anteroposterior or lateral radiographic views, and a class IV Stulberg hip has at least 1cm of flattening of the weight-bearing articular surface. The Stulberg classification remains the best predictor of outcome in terms of subsequent hip arthrosis, pain, and function. As such, it is used to gauge results of various treatments of LCPD.


A classification system is necessary to describe the variability of severity of Perthes disease, and an appropriate outcome rating system for the radiographic appearance of the femoral head at maturity is essential in understanding the results of treatment. We reviewed the literature to determine which classification systems were most widely used and useful. The Catterall rating has historically been most widely used, although later published studies have used the most recent modification of the lateral pillar classification proposed by Herring et al,7,9 because it has the best reported interobserver reliability. The Stulberg classification system has been most useful in describing the end result shape of the femoral head in adulthood and has been shown to correlate with age of onset of arthritis. It has recently been modified by Herring et al9 to improve interobserver reliability.


1. Legg AT. An obscure affection of the hip joint Boston Med Surg J.. 1910;162:202
2. Legg AT. The end results of coxa plana J Bone Joint Surg Am. 1927;9:26–36
3. Waldenström H. The first stages of coxa plana J Bone Joint Surg Am. 1938;20:559–566
4. Catterall A. The natural history of Perthes' disease J Bone Joint Surg Br.. 1971;53:37–53
5. Sundt H. Further investigations respecting malum coxal Calvé-Legg-Perthes with special regard to prognosis and treatment Acta Chir Scand (suppl). 1949;148:1–21
6. Salter RB, Thompson GH. Legg-Calvé-Perthes disease: the prognostic significance of the subchondral fracture and a two-group classification of the femoral head involvement J Bone Joint Surg Am. 1984;66:479–489
7. Herring JA, Neustadt JB, Williams JJ, et al. The lateral pillar classification of Lee-Calvé-Perthes disease J Pediatr Orthop.. 1992;12:143–150
8. Stulberg SD, Cooperman DR, Wallensten R. The natural history of Legg-Calvé-Perthes disease J Bone Joint Surg Am. 1981;63:1095–1108
9. Herring JA, Kim HT, Browne RH. Legg-Calvé-Perthes disease. Part 1: classification of radiographs with use the modified lateral pillar and Stulberg classifications J Bone Joint Surg Am. 2004;86:2103–2119
10. Fasting OJ, Bierkreim I, Langeland N, et al. Scintigraphic evaluation of the severity of Perthes' disease in the initial stage Acta Orthop Scand.. 1980;51:655–660
11. Conway JJ. A scintigraphic classification of Legg-Calvé-Perthes disease Semin Nucl Med.. 1993;13:274–295
12. Moss K. Method of measuring in Legg-Calvé-Perthes disease with special regard to the prognosis Clin Orthop. 1980;150:103–109
13. Aksoy MC, Cankus MC, Alanay A, et al. Radiological outcome of proximal femoral varus osteotomy for the treatment of lateral pillar group-C Legg-Calvé-Perthes disease J Pediatr Orthop B. 2005;14:88–91
14. Crutcher JP, Staheli LT. Combined osteotomy as a salvage procedure for severe Legg-Calvé-Perthes disease J Pediatr Orthop.. 1992;12:151–156
15. Denton JR. Experience with Legg-Calvé-Perthes disease (LCPD) 1968 to 1974 at the New York Orthopaedic hospital Clin Orthop.. 1980;150:36–42
16. Fabry K, Fabry G, Moens P. Legg-Calvé-Perthes disease in patients under 5 years of age does not always result in a good outcome: personal experience and meta-analysis of the literature J Pediatr Orthop B. 2003;12:222–227
17. Freeman RT, Wainwright AM, Theologis TN, et al. The outcome of patients with hinge abduction in severe Perthes disease treated by shelf acetabuloplasty J Pediatr Orthop.. 2008;28:619–625
18. Gigante C, Frizziero P, Turra S. Prognostic value of Catterall and Herring classification in Legg-Calvé-Perthes disease: follow-up to skeletal maturity of 32 patients J Pediatr Orthop.. 2002;22:345–349
19. Heikkinen E, Puranen J. Evaluation of femoral osteotomy in the treatment of Legg-Calvé-Perthes disease Clin Orthop.. 1980;150:60–68
20. Hoikka V, Lindholm TS, Poussa M. Intertrochanteric varus osteotomy in Legg-Calvé-Perthes disease: a report of 112 hips J Pediatr Orthop.. 1986;6:600–604
21. Hoikka V, Poussa M, Yrjönen T, et al. Intertrochanteric varus osteotomy for Perthes' disease: radiographic changes after 2 to 16-year follow-up of 126 hips Acta Orthop Scand.. 1991;62:549–553
22. Sponseller PD, Desai SS, Millis MB. Comparison of femoral and innominate osteotomies for the treatment of Legg-Calvé-Perthes disease J Bone Joint Surg Am. 1988;70:1131–1139
23. Kamhi E, MacEwen GD. Treatment of Legg-Calvé-Perthes disease: prognostic value of Catterall's classification J Bone Joint Surg Am. 1975;57:651–654
24. Kim WC, Hosokawa M, Tsuchida Y, et al. Outcomes of new pogo-stick brace for Legg-Calvé-Perthes' disease J Pediatr Orthop B.. 2006;15:98–103
25. Nomura T, Terayama K, Watanabe S. Perthes' disease: a comparison between two methods of treatment, Thomas' splint and femoral osteotomy Arch Orthop Trauma Surg.. 1980;97:135–140
26. Noonan KJ, Price CT, Kupiszewski SJ, et al. Results of femoral varus osteotomy in children older than 9 years of age with Perthes disease J Pediatr Orthop.. 2001;21:198–204
27. Olney BW, Asher MA. Combined innominate and femoral osteotomy for the treatment of severe Legg-Calvé-Perthes disease J Pediatr Orthop.. 1985;5:645–651
28. Reddy RR, Morin C. Chiari osteotomy in Legg-Calvé-Perthes disease J Pediatr Orthop B. 2005;14:1–9
29. Robinson HJ Jr, Putter H, Sigmond MB, et al. Innominate osteotomy in Perthes disease J Pediatr Orthop. 1988;8:426–435
30. Schmid OA, Hemmer S, Wünsche P, et al. The adult hip after femoral varus osteotomy in patients with unilateral Legg-Calvé-Perthes disease J Pediatr Orthop. B.. 2003;12:33–37
31. Trias A. Femoral osteotomy in Perthes disease Clin Orthop.. 1978;137:195–207

Legg-Calve-Perthes Disease; LCPD; Catterall classification; Salter-Thompson classification; Lateral Pillar classification; Stulberg classification; Mose measurement

© 2011 Lippincott Williams & Wilkins, Inc.