Secondary Logo

Journal Logo

Evaluation and Classification

Magnetic Resonance Imaging in Legg-Calvè-Perthes Disease: Review of Literature

de Sanctis, Nando MD

Author Information
Journal of Pediatric Orthopaedics: September 2011 - Volume 31 - Issue - p S163-S167
doi: 10.1097/BPO.0b013e318223b575
  • Free



Catterall, in 1971, proposed his radiographic classification1 on the basis of the extension of the necrosis in the femur head, and in 1980 he added other signs that defined head at risk. This classification is still used, although other studies have demonstrated the poor intraobserver and interobserver reliability, and Catterall2 himself in 1995 entitled an European instructional lecture: “Perthes disease: eighty years of an obscure condition.”

In 1982, Catterall in his book3 wrote that 58% of the patients get better without any treatment, whereas the remaining one-third improves with time.

In 1971, Gower and Johnston,4 reviewing 36 patients (average age, 45 y) found that 86% of the hips were still functioning and only 3 out of their 36 patients (8%) had come for arthroplasty; Catterall seems to be right.

However 13 years later, McAndrew and Weinstein5 reviewed the same 36 patients (average age, 53 y) finding that 40% of them underwent an arthroplasty. Only 40% had maintained a good level of function. This means that 60% of the hips had bad results.

In 1977, Brotherton and McKibbin reported the results of a long-term review of 102 hips in 87 patients with Legg-Calvè-Perthes disease (LCPD) treated with long bed rest, the mean follow-up interval being 17 years. The results were very satisfactory, with only 2% poor results and 10% fair results. The remaining 88% were good results.

The radiological results at the end of the treatment have also been compared with the control series described by Catterall (1972) and with the osteotomy series of Lloyd-Roberts et al (1976). From this study, it seems that the described regimen offers no benefit compared with the natural history in Catterall's groups I and II, and in group III the results were only marginally better than those after osteotomy. In group IV, however, where the femoral head was totally involved, the benefit was important.6


In 1994 in his current concept review, Herring studying the literature and analyzing several remarks (only patients younger than 6 years, age not reported, number of patients too low, bias in Catterall group, Catterall group changed during treatment, classification made on a single radiogram, patients excluded without explanation, criteria of exclusion to surgery not reported, only 14% of the patients studied, and other remarks such as incorrect Rx criteria, absence of control group, and improvement of the disease owing to natural history rather than treatment) reported that “the message of the current review is that most of the reported studies have simply lacked the controls necessary to allow valid scientific conclusion to be drawn” and “there is a critical lack of scientific rigor, because patients have been included or excluded from the studies on the basis of the Catterall classification, which itself lacks interobserved reliability.”7

Moreover the Salter Thompson classification, based on the subchondral fracture, cannot always be used because, in later stages of the disease, it is rarely encountered.

The Herring classification8 based on the height of lateral pillar is difficult to evaluate in very young patients and in bilateral cases. Moreover, there are no long follow-ups.

However until today, it seems to be the best approach.

Arthrography, taking into account the flattening and the extrusion of the cartilaginous epiphysis, yields useful information even in early stages. However, we must consider that it is an invasive procedure and is difficult to repeat.9 Moreover, with this method, although we can provide evidence that the labrum horizontalization seems to be an important predicting sign, it is not possible to show the metaphyseal and physeal damages.

Earlier, bone scan was used to quantify the extent of the ischemia and to predict the necrosis severity,10,11 but now it is not routinely used because of the radiation risk and because it does not allow appreciation of the shape of the femur head and its relationship with acetabular fibrocartilage.

Other investigators have reported about abnormal growth of the proximal femoral plate with growth acceleration or retardation. Growth arrest with bone bridge formation influences the proximal femur shape. Because of the difficulty in investigating the physis, the question regarding the responsibility of its impairment in the deformity of the hip still remains controversial.12–17

At this moment, some investigators have decided that there is no sufficient reliability of prognostics criteria and it must be useful to explore a new way.


Twenty-six years ago, first studies about magnetic resonance imaging (MRI) in LCPD were published where the ability of the nuclear magnetic resonance scanner to reconstruct images of the hip in both coronal and axial planes was proved. This permitted the study of the congruity of the joint and the distribution of infarction in which it was previously difficult to orientate. The demonstrated ability of the scanner to detect infarction of the femoral head and to image articular surfaces without ionizing or injection of contrast agents are major advantages over conventional imaging methods.18

In 1989, Pinto and Hamlet19 affirmed that MRI is valuable in the diagnosis of vascular changes in bone; it is better than other techniques in demonstrating the full extent and precise location of the necrotic region.20,21 In addition, the physeal line, which is not appreciated on isotopic scans, is visualized. MRI avoids ionizing radiation and is noninvasive. It is important to underline the advantages of this procedure in the early diagnosis of LCPD, before the development of typical radiologic changes. An early diagnosis and treatment of LCPD may result in a more favorable outcome.


In 1989, Mitchell et al., studying the metaphyseal cyst, affirmed that thecavity with a low-intensity signal in T1 and a high-intensity signal in T2 contains a watery fluid.22

In 1993, Hoffinger et al23 said that although metaphyseal cysts in LCPD have gained much attention in orthopedic literature, yet their etiology, pathophysiology, and even location remain unclear.

With MRI, it is possible to show that physeal irregularity and step off are common on MRI scan of hips with LCPD and metaphyseal changes on plain radiographs. In addition, these findings may also exist in hips without any radiographic metaphyseal changes. MRI seems to be a more sensitive tool than plain radiographs in detecting this physeal irregularity.

In 1991, Bos and others wrote: Catterall assessed the extent of epiphyseal involvement in radiographs and therefore the precise degree of involvement could not be determined before the process of resorption was completed. On MRIs, the repair tissue interface can indicate whether and to what extent the infarcted bone include the growth plate. The second set of MRIs, 6 months after the onset of the disease, precisely delineates the extent of the infarcted zone below the subchondral fracture line. This means that MRI shows the true extent of necrosis sooner than x-rays. In the later stages of the disease, the MRI findings and radiographs correlated more closely.24

In 1995, Jaramillo and others25 proved that epiphyseal abnormality seen in 74% of cases was not related to growth disturbances, whereas cartilaginous physeal and metaphyseal abnormalities in LCP disease are common and frequently result in growth arrest.

In 2000, Song and others26 affirmed that the metaphyseal changes in LCPD have been recognized as a radiologic risk factor and a contributor to poor outcome. Katz and Siffert27 reported that the hips with poor results had a higher (49%) incidence of metaphyseal cyst compared with that (26%) of hips with acceptable results.

In 2004, Song wrote again that the hips without metaphyseal change were less involved in epiphysis compared with the hips having a metaphyseal change. Metaphyseal cysts disappeared on radiographs and MRI scans during the healing or remodeling stage.28

In 2000, two papers were published about an MRI prospective study based on 4 parameters: necrosis extension, lateral extrusion, and physeal and metaphyseal involvement in 2 different series. Twenty-eight hips from Santobono Hospital, Naples, and 31 hips from Montpellier University were studied.29

A statistical study was carried out correlating the necrosis extension of the physeal damage, the lateral extrusion, and the metaphyseal involvement. For these parameters, simple and multiple linear regressions were used. The interobserver variation analysis showed an agreement on 80% of the parameters and a correlation study between MRI and clinical and radiographic parameters was carried out at an average of 4 years follow-up. A statistical correlation with Stulberg class and with clinical radiographic results (TOT) at follow-up was carried out.

In this study, it was found that all the 4 MRI parameters were significantly correlated with the results at follow-up and with the Stulberg class. Physeal involvement and lateral extrusion are reciprocally correlated. In LCPD when abnormal growth of the proximal femoral physis occured,11–16 physeal involvement showed to be the strongest predictor accounting by itself for 61.7% of STU variation and 55.5% of TOT variation, and played a key role in determining the ultimate shape of the proximal femur.


  1. MRI represents the only method that indicated very early the extension of the necrosis and the damages of the growth cartilage and metaphysis (Figs. 1–3).30
    • This is an important step forward in obtaining easily an early diagnosis in several uncertain cases. Moreover, it yields useful data to evaluate precociously the risk cases.
  2. This review shows the early involvement of the growth cartilage in severe cases.
    • In the last few years, the concept of the weight of the physeal and metaphyseal alterations has become increasingly important. They can represent early signs of the seriousness of the disease, thereby, improving the correct timing of the effective therapy.
  3. If the growth cartilage is not involved or less involved, the final results are always good, even if the necrosis extension is wide.
    • After Catterall classification, many articles have underlined the correlation between necrosis extension and the severity of the outcomes. If this is true, it is also true that, if the epiphysis is unloaded,3,14,16,30 the risk of having bad results decreases.
  4. The horizontalization of the labrum seems to be an important prognostic predictor (Fig. 4).
    • This point, put in evidence by an accurate arthrographic study carried out by Laredo (UNIFESP of Sao Paolo, Brazil), seemed immediately as a very important sign that can demonstrate the extrusion of the head until hinge abduction (when the limbus is inverted).
  5. The presence of the cyst is an important sign frequently correlated with growth plate damages25,26,29 (Fig. 5).
    • The metaphyseal cyst described by Catterall among the risk signs has become more important, in the last years, because some articles have demonstrated that the contest of the cyst is fluid and there exists a correlation between the necrotic fluid in the head and the fluid in the cyst that can be transferred through to growth plate.
  6. In the necrotic stage, weight bearing is harmful. It is possible to correlate this concept to the pressure on the head owing to the load. This pressure can push the necrotic fluid into the metaphysis disrupting the growth cartilage.5,31
  • • The problem of the weight bearing as its consequence on the necrotic head is always under discussion.
  • • Catterall said in the beginning that most of the cases did not need any treatment.
  • • McAndrew and Weinstein, in contrast, demonstrated that 60% of the cases at a very large maturity review had bad results.
  • • Brotherton and McKibbin also demonstrated that the unload yields the best results.
  • • The effectiveness of the bracing in abduction to obtain the containment of the head in maintaining sphericity between head and acetabulum is not demonstrated.
  • • At this moment, it is possible to reduce the risk of having the hinge abduction, but it is not the same for the flattening and sphericity of the head.
  • • It seems useful to accept that the load is dangerous in the necrotic stage and obviously it is more dangerous when the necrosis is wide and when the ossific nucleus is bigger, as in older patients, because the amount of the fluid is greater.
A 3-year-old child with recurrent transient synovitis.
An MRI showing full necrosis of the nucleus. MRI indicates magnetic resonance imaging.
One year after, Perthes disease with metaphyseal cyst.
Normal head at right, Perthes disease with full head necrosis at left. Flattening and extrusion of the head with labrum orizontalization.
Epiphyseal necrosis and large metaphyseal cyst in the femoral neck with sequential MRI scan that shows the presence of large necrotic fluid in the cyst. MRI indicates magnetic resonance imaging.

The Future

After this revision, MRI seems to be the best method, which is able to provide evidence the different components as bone as cartilaginous of the neck, the head and acetabulum in children. The availability of new more sophisticated MRI techniques such as fat suppression, intravenous contrast administration, and gadolinium enhances MRI, showed multiple cartilaginous vascular structure of the growing skeleton,32 parallel imaging that can enhance the depiction abnormalities, and increased speed of imaging and improves overall quality of the study.33

Recently, diffusion-weighted imaging applied in conjunction with balanced steady state free precession sequences in the assessment of cartilage repair such as autologous chondrocyte have had good sensitivity.

Delayed gadolinium-enhanced MRI of cartilage is based on the observation that ions within the interstitial fluid in hyaline cartilage are distributed in relation to the concentration of negatively charged glycosaminoglycan molecules, which in turn are related to the amount.

The term delayed in delayed gadolinium-enhanced MRI of cartilage reflects the time needed to allow the penetration of Gd-DTPA2 through the full cartilage thickness.

The majority of the studies about this new technique have been correlated to the articular cartilage surface. Very few studies have been published until now about the growth cartilage and LCPD. However, in conclusion, we can affirm with Jaramillo and others34 that gadoteridol-enhanced T1-weighted images showed multiple anatomic structures within the epiphysis and physis, some of which were not visible through images obtained with other pulse sequences. The appearance of the growing structures changed after injection as the gadolinium diffused through the cartilage and marrow. Enhancement was greatly influenced by the skeleton maturation.

It seems very important to have more studies using the new MRI techniques in LCPD to increase our knowledge and to improve our chances to reach a reliable classification to obtain best results in the treatments.


1. Catterall A. The natural history of Perthes disease J Bone Joint Surg Br. 1971;53:37–53
2. Catterall A. Perthes disease: eighty years of an obscure condition European Instructional Course Lectures. Vol. 2.. 1995 London The British Editorial Society of Bone and Joint Surgery
3. Catterall A Legg-Calvè-Perthes disease. 1982 Edinburg Churchill Livingstone
4. Gower WE, Johnston RC. Legg-Perthes disease. Long term follow-up of 36 patients J Bone Joint Surg.. 1971;53-A:759
5. McAndrew HP, Weinstein SL. A long term follow up of Legg-Calvè-Perthes disease J Bone Joint Surg Am.. 1984;66:860–869
6. Brotherton BJ, McKibbin B. Perthes' disease treated by prolonged recumbency and femoral head containment: a long-term appraisal J Bone Joint Surg Br.. 1977;59-B:8–14
7. Herring JA. Current concepts review. The treatment of Legg-Calvè-Perthes disease. A critical review of the literature J Bone Joint Surg.. 1994;76:448–458
8. Herring JA, Neustad JB, Williams JJ, et al. The lateral pillar classification of Legg-Calvè-Perthes disease J Pediatr Orthop. 1992;12:143–150
9. Laredo JF. Doença de Legg-Calvè-Perthes. Parte II. Classificaçao artrogràfica Rev Bras Ortop.. 1992;27:7–10
10. Bensahel H, Bok B, Cavailloles F, et al. Bone scintigraphy in Perthes disease J Pediatr Orthop.. 1983;3:302–305
11. Conway JJ. A scintigraphic classification of Legg-Calvé-Perthes disease Semin Nucl Med.. 1993;23:274–295
12. Katz JF. Legg-Calvè-Perthes disease: the role of distorsion of normal growth mechanism in the production of deformity Clin Orthop.. 1970;71:193–198
13. Barnes JM. Premature epiphyseal closure in Perthes disease J Bone Joint Surg Br.. 1980;62:432–437
14. Bowen JR, Schreiber FC, Foster BK, et al. Premature femoral neck physeal closure in Perthes disease Clin Orthop.. 1982;171:24–29
15. Clarke MP, Harrison HN. Painful sequelae of coxa plana J Bone Joint Surg Am.. 1983;65:13–18
16. Keret D, Harrison MH, Clarke NM. Coxa plana: the fate of the physis J Bone Joint Surg Am.. 1984;66:870–877
17. Allen BL. Graphic analysis of femoral growth in young children with Perthes disease J Pediatr Orthop.. 1997;17:255–263
18. Scoles PV, Yoon YS, Makley JT, et al. Cleveland, Ohio Nuclear. Magnetic Resonance Imaging in Legg-Calvé-Perthes Disease J Bone Joint Surg Am.. 1984;66-A:1357–1363
19. Pinto MR, Hamlet A. Peterson and Thomas H. Berquist. Magnetic Resonance Imaging in Early Diagnosis of Legg-Calvé-Perthes Disease J Pediatric Orthop.. 1989;9:19–22
20. Markisz JA, Knowles RJ, Altchek DW, et al. Segmental patterns of avascular necrosis of the femoral heads: early detection with MR imaging Radiology.. 1987;162:717–720
21. Mitchell DG, Rao VM, Dalinka MK, et al. Femoral head avascular necrosis: correlation of MR imaging, radiographic staging, radionuclide imaging, and clinical findings Radiology.. 1987;162:709–715
22. Mitchell DG, Steinberg ME, Dalinka MK, et al. Magnetic resonance imaging of the ischemic hip: alterations within the osteonecrotic viable and reactive zones Clin Orthop.. 1989;244:60–67
23. Hoffinger SA, Henderson RC, Renner JB, et al. Magnetic Resonance Evaluation of “Metaphyseal” Changes in Legg-Calvé-Perthes Disease J Paediatr Orthopaed.. 1993;13:602–606
24. Bos CFA, Bloem JL, Bloem RM. Sequential Magnetic resonance imaging in Perthes' disease J Bone Joint Surg.. 1991;73-B:2
25. Jaramillo D, Kasser JR, Villegas-Medina OL, et al. Cartilaginous abnormalities and growth disturbances in Legg-Calvé-Perthes disease: evaluation with MR imaging Radiology.. 1995;197:767–773
26. Song HR, Dhar S, Na JB, et al. Classification of Metaphyseal Change with Magnetic Resonance. Imaging in Legg-Calve´-Perthes Disease J Pediatr Orthopaed. 2000;20:557–561
27. Katz JF, Siffert RS. Capital necrosis, metaphyseal cyst and subluxation in coxa plana Clin Orthop.. 1975;106:75–85
28. Song HR. Evaluation of the hip joint space and involvement of the epiphysis and metaphysis with magnetic resonance imaging in legg-calve'-perthes disease J Bone Joint Surg [Br]. 2004;86-B:148
29. De Sanctis N, Nunziata Rega A, Rondinella F. Prognostic evaluation of legg-calvè-perthes disease by MRI Part I: the role of physeal Involvement J Paed Orthopaed.. 2000;20:455–462
30. Ecklund K, Jaramillo D. Imaging of growth disturbance in children Radiol Clin North Am.. 2001;39:823–841
31. De Sanctis N, Rondinella F. Prognostic Evaluation of Legg-Calvè-Perthes Disease by MRI Part II: pathomorphogenesis and new classification J Paed Orthopaed.. 2000;20:463–470
32. Sebak G, Ducou Le Point H, Klein I. Dinamic gadolinium ehonced-subtraction MR imaging: a simple technique for the early diagnosis of the Legg-Calvè-Perthes disease-preliminary results Paed. Radiol.. 1997:27,216–27,220
33. Jaramillo D, Laor T. Pediatric musculoskeletal MRI: basic principles to optimize success Pediatr Radiol.. 2008;38:379–391 Epub 2007 November 29.
34. Jaramillo D, Villegas-Medina OL, Doty DK, et al. Age-Related vascular changes in the epiphysis, physis, and metaphysis: normal findings on gadolinium-enhanced MRI of piglets AJR.. 2004;182:353–360

LCPD; MRI; growth plate; early diagnosis and prognosis

© 2011 Lippincott Williams & Wilkins, Inc.