The final study group included 172 patients (152 boys and twenty girls) with 188 affected hips. Two hundred and one patients were excluded from the study. Ninety-seven were excluded because medical records could not be located; sixty-three, because of inadequate radiographs; thirty-one, because the femoral head was not seen to be fully reossified on the final radiograph; six, because the earliest radiographs had been made when the disease was past the fragmentation stage; three, because the femoral head involvement was too slight for it to be classified as Legg-Calvé-Perthes disease; and one, because radiographs of other joints revealed multiple joint abnormalities.
Fifty-four patients (31%) were treated with a brace or cast. An Atlanta brace was used for twenty-six patients; a Craig splint, for ten; an A-frame brace, for four; an unspecified abduction brace, for six; and a Petrie cast, for eight. Seven hips treated with a Petrie cast also had an adductor tenotomy, and three of them also had an iliopsoas tenotomy at the time of cast application. One hundred and fourteen patients (66%) were observed or had treatment of symptoms only. Charts could not be located for four patients.
The Fisher exact test was used to compare rates, and multiple logistic regression methods were used to predict the probability of a particular outcome. Significance was defined as a p value of <0.05.
The entire group included 188 hips diagnosed in 172 patients before they were six years old. The average age at the onset of the Legg-Calvé-Perthes disease was 4.6 years (range, 2.0 to 5.9 years). On the final radiograph, eighty-six patients were seen to have an open triradiate cartilage and eighty-six had a closed triradiate cartilage. According to the lateral pillar system, seven hips (4%) were in group A, 108 (57%) were in group B, thirty (16%) were in the B/C border group, and forty-three (23%) were in group C. Since there were so few group-A hips and all had a good outcome, they were included with the group-B hips (for a total of 115 group-A or B hips [61%]) for statistical analysis (Table I). There were 156 patients with unilateral disease and sixteen with bilateral disease. One hundred and fifty-two results (81%) were good (Stulberg Class I or II), seventeen (9%) were fair (Stulberg Class III), and nineteen (10%) were poor (Stulberg Class IV). There was a significant correlation between the lateral pillar classification and the outcome (p < 0.001).
The 172 patients were divided into two groups on the basis of the radiographic appearance and progression of the disease. The first group consisted of patients with unilateral or bilateral Legg-Calvé-Perthes disease with typical characteristics. The second group comprised those with concurrent onset and progression of bilateral Legg-Calvé-Perthes disease.
Typical Legg-Calvé-Perthes Disease (Group I)
This group consisted of 160 patients with 164 affected hips. According to the lateral pillar classification, there were seven group-A hips (4%), 101 group-B hips (62%), twenty-seven group-B/C-border hips (16%), and twenty-nine group-C hips (18%) (Table II). One hundred and thirty-one results (80%) were good (Stulberg Class I or II), fourteen (9%) were fair (Stulberg Class III), and nineteen (12%) were poor (Stulberg Class IV). The Fisher exact test showed a significant correlation between the lateral pillar classification and the final outcome (p < 0.001).
When the patients were divided into two groups according to their age at onset (newborn to three years and eleven months or four years to five years and eleven months), the combination of the younger age group and a lateral pillar group-A or B classification was found to be significantly correlated with a better outcome (p = 0.002) with use of logistic regression methods. The resulting logistic regression model correctly classified the hips in terms of outcome 83.7% of the time and also estimated the probability of a good outcome as shown in Table III.
A logistic regression model was used to determine whether treatment (bracing with or without adductor tenotomy) affected the Stulberg outcome (good, fair, or poor) after adjustment for the lateral pillar classification (A or B, B/C, or C). The lateral pillar classification significantly affected outcome (p < 0.0001), but brace treatment did not regardless of the lateral pillar classification (p = 0.75) (Table IV). We also could not identify a correlation between gender and outcome.
Bilateral Legg-Calvé-Perthes Disease with Concurrent Onset and Progression (Group II)
Twelve patients were diagnosed with bilateral Legg-Calvé-Perthes disease with a concurrent onset prior to the age of six years. Each patient had also had concurrent bilateral progression through the stages of fragmentation and reossification. Ten of these twelve patients had radiographs of the spine and other joints, which did not reveal any abnormalities. Of the twenty-four hips, none were in group A, according to the lateral pillar classification, seven were in group B, three were in the B/C border group, and fourteen were in group C. Final radiographs revealed twenty-one good results (Stulberg class I or II), three fair results (Stulberg Class III), and no poor results (Stulberg Class IV). All three fair results occurred in group-C hips. With the numbers studied, we could not identify a correlation between the lateral pillar classification and outcome.
Comparison of Groups I and II
Ten of the twenty-nine lateral pillar group-C hips in the group with typical Legg-Calvé-Perthes disease (Group I) and none of the fourteen group-C hips in the group with concurrent bilateral disease (Group II) had a poor outcome. This difference was significant (p = 0.038).
Typical Features of Group II
There were radiographic changes that could be interpreted as indicating Meyer dysplasia in two patients (four hips) in Group II. In six patients (twelve hips), there was a distinct increased density phase not seen in Meyer dysplasia (Figs. 2-A through 2-E). Two patients (four hips) had metaphyseal changes, and another two had radiographic changes in the femoral head that did not resemble those of Meyer dysplasia.
No strong conclusions can be drawn from the studies published to date regarding the treatment and outcomes of Legg-Calvé-Perthes disease in patients with an onset prior to the age of six years5-8,10,18. Small sample sizes and variable results have made interpretation of these data difficult, and to our knowledge there has been no large multicenter prospective study of this age group. With use of the large database of young patients with Legg-Calvé-Perthes disease at our institution, we attempted to answer three questions. First, do patients who have the onset of Legg-Calvé-Perthes disease prior to the age of six years have good outcomes? Second, is the lateral pillar classification predictive of outcome in this younger population? Third, does treatment affect the outcome in this group of patients?
Fabry et al. performed a meta-analysis of the literature in an attempt to support or refute the belief that a young age is predictive of a good outcome18. They concluded that the reported results were too variable to support any correlation between a young age and a good outcome. Furthermore, their own study of thirty-six patients who had had the onset of Legg-Calvé-Perthes disease prior to the age of five years suggested that the extent of femoral head involvement was of greater prognostic importance than age. One-third of the patients in their series had a poor result. Of nineteen hips that were classified as being in lateral pillar group C, only four had a good result and nine had a poor result. Fabry et al. concluded that, as evidenced by the high percentage of hips in lateral pillar group C, patients diagnosed prior to the age of five years were not protected from severe femoral head involvement and that greater involvement was predictive of a poor outcome. Schoenecker et al. reported similar results in their review of the cases of 109 young patients; they concluded that the degree of femoral head involvement was the single most important variable predicting outcome6.
Similar conclusions can be drawn from the results of the present study. The results for the entire group of patients were very good, with 81% having a good outcome and only 10% having a poor outcome. Further analysis showed that, among these younger children, age was not a significant independent variable. We found that the severity of femoral head involvement, as assessed with the lateral pillar classification, had a stronger independent correlation with outcome than did the age at the time of the disease onset. While 95% of the lateral pillar group-A and B hips had a good outcome, only 58% of the lateral pillar group-C hips had a good outcome in this early-onset population (p < 0.001). Excluding the bilateral cases with simultaneous onset, 48% of the group-C hips had a good result. Furthermore, we found that the combination of age and the lateral pillar classification was an even stronger predictor of outcome. Patients diagnosed with a lateral pillar group-A or B hip when they were between the ages of zero and three years and eleven months had a 99% probability of having a good outcome, whereas those diagnosed with a lateral pillar group-B/C or C hip when they were between the ages of four years and six years and eleven months had only a 43% probability of having a good outcome.
While analyzing the radiographs, we noted that there was a distinct group of patients whose disease pattern and progression did not fit that of “typical” Legg-Calvé-Perthes disease. Initially, we identified thirteen patients who had bilateral Legg-Calvé-Perthes disease with concurrent onset and simultaneous progression through the radiographic stages. Typically, in patients with bilateral Legg-Calvé-Perthes disease, the onset of disease in the two hips occurs at different ages and there is asymmetric progression. Analysis of radiographs of the spine and other joints revealed only one patient who had features consistent with multiple epiphyseal dysplasia, and this patient was then excluded from the study. The remaining twelve patients had normal findings on radiographs of the spine and the other extremities and had no other disorders. Fourteen of these twenty-four hips were classified as being in lateral pillar group C, and many would have been expected to have a poor radiographic outcome5,10. However, none had a poor result, and twenty-one (88%) of the twenty-four hips had a good result. Since these results were much better than expected, one may question the diagnosis of Legg-Calvé-Perthes disease in those patients. All patients who had radiographs of the spine or other joints showed none of the features of an epiphyseal dysplasia. As described by Meyer in 1964, dysplasia epiphysealis capitis femoris is characterized as a delayed appearance of bilateral capital femoral epiphyseal ossification20. Once the nucleus does appear, the epiphysis is fragmented but then it demonstrates constant improvement to the end point of a normal, round femoral head, and increased density is never seen on radiographs. This disorder is also characterized by a normal femoral neck without metaphyseal reaction or widening, and there is never flattening of the femoral head. As described by Rowe et al., the average age of onset of Meyer dysplasia (2.5 years) is earlier than that of Legg-Calvé-Perthes disease21. Taking these factors into consideration, we do not believe that these twelve patients should be classified as having Meyer dysplasia. Our theory is that this group with atypical Legg-Calvé-Perthes disease may have had an as yet undescribed disorder of the femoral head, which will require further study.
The third question that we hoped to answer with this study was whether treatment affected the outcome in this young population. In a multicenter prospective study, Herring et al. demonstrated that patients who are over eight years of age at the time of the onset of Legg-Calvé-Perthes disease and have a lateral pillar classification of B or B/C have a better outcome with surgical treatment than they do with nonoperative treatment11. These authors noted that nonoperative treatments such as bracing or casts had no influence on outcomes. Since the predominant approach to treatment for the patients in our study was observation and treatment of symptoms, all were treated essentially nonoperatively. Those that did receive some form of treatment underwent one of several forms of abduction bracing. The Craig brace, used in the early years, is an abduction bar attached to the shoes with which the patient can flex the knees and adduct the hip and hyperabduct the unaffected hip. Of the 172 patients studied, fifty-four were treated with bracing or splinting. Of these, seven had an adductor and/or iliopsoas tenotomy prior to bracing. One hundred and fourteen patients had only observation or treatment of symptoms. We could not detect a significant effect of bracing on outcome, regardless of the lateral pillar classification, compared with the outcomes in the untreated hips. We found that hips in lateral pillar groups B/C and C had a higher risk of a poor outcome. We cannot comment on the use of surgical containment of the femoral head for these lateral pillar groups since none of the hips in our series were treated surgically. On the basis of the very good results in this age group with lateral pillar group-A or B hips, it seems that more aggressive treatment is not indicated for these hips.
In conclusion, most patients with the onset of Legg-Calvé-Perthes disease before the age of six years have a good result. The lateral pillar classification is a strong predictor of outcome in this group of patients, and only those with a lateral pillar group-B/C border or C hip are at risk for a poor outcome. The age at onset is not an independent predictor of outcome in this group of younger patients. However, the combination of the age of onset and the lateral pillar classification is strongly correlated with outcome.
Disclosure: The authors did not receive any outside funding or grants in support of their research for or preparation of this work. Neither they nor a member of their immediate families received payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, division, center, clinical practice, or other charitable or nonprofit organization with which the authors, or a member of their immediate families, are affiliated or associated.
Investigation performed at the Texas Scottish Rite Hospital for Children, Dallas, Texas
1. 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;66:479-89.
2. Ratliff AH. Perthes' disease. A study of thirty-four hips observed for thirty years. J Bone Joint Surg Br. 1967;49:102-7.
3. O'Hara JP, Davis ND, Gage JR, Sundberg AB, Winter RB. Long-term follow-up of Perthes' disease treated nonoperatively. Clin Orthop Relat Res. 1977;125:49-56.
4. Ippolito E, Tudisco C, Farsetti P. The long-term prognosis of unilateral Perthes' disease. J Bone Joint Surg Br. 1987;69:243-50.
5. Snyder CR. Legg-Perthes disease in the young hip–does it necessarily do well? J Bone Joint Surg Am. 1975;57:751-9.
6. Schoenecker PL, Stone JW, Capelli AM. Legg-Perthes disease in children under 6 years old. Orthop Rev. 1993;22:201-8.
7. Ismail AM, Macnicol MF. Prognosis in Perthes' disease: a comparison of radiological predictors. J Bone Joint Surg Br. 1998;80:310-4.
8. Kelly FB Jr, Canale ST, Jones RR. Legg-Calvé-Perthes disease. Long-term evaluation of non-containment treatment. J Bone Joint Surg Am. 1980;62:400-7.
9. Dekelver L, Fabry G. Legg-Calve-Perthes follow-up study. Arch Orthop Trauma Surg. 1982;100:143-9.
10. Mukherjee A, Fabry G. Evaluation of the prognostic indices in Legg-Calve-Perthes disease: statistical analysis of 116 hips. J Pediatr Orthop. 1990;10:153-8.
11. Herring JA, Kim HT, Browne R. Legg-Calvé-Perthes disease. Part II: prospective multicenter study of the effect of treatment on outcome. J Bone Joint Surg Am. 2004;86:2121-34.
12. Pietrzak S, Napiontek M, Tomaszewski M. [Inter-observer variation of the Caterall and Herring classification in Perthes disease]. Chir Narzadow Ruchu Ortop Pol. 2000;65:33-8. Polish.
13. Podeszwa DA, Stanitski CL, Stanitski DF, Woo R, Mendelow MJ. The effect of pediatric orthopaedic experience on interobserver and intraobserver reliability of the Herring lateral pillar classification of Perthes disease. J Pediatr Orthop. 2000;20:562-5.
14. Specchiulli F, Cofano RE. Long-term follow-up of Perthes' disease in adolescence. Chir Organi Mov. 2001;86:7-13.
15. Ritterbusch JF, Shantharam SS, Gelinas C. Comparison of lateral pillar classification and Catterall classification of Legg-Calvé-Perthes' disease. J Pediatr Orthop. 1993;13:200-2.
16. Grzegorzewski A, Bowen JR, Guille JT, Glutting J. Treatment of the collapsed femoral head by containment in Legg-Calvé-Perthes disease. J Pediatr Orthop. 2003;23:15-9.
17. Specchiulli F, Scialpi L. Catterall versus Herring classification in Perthes' disease. Chir Organi Mov. 1997;82:289-93.
18. 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;12:222-7.
19. Herring JA, Kim HT, Browne R. Legg-Calvé-Perthes disease. Part I: classification of radiographs with use of the modified lateral pillar and Stulberg classifications. J Bone Joint Surg Am. 2004;86:2103-20.
20. Meyer J. Dysplasia epiphysealis capitis femoris. A clinical-radiological syndrome and its relationship to Legg-Calvé-Perthes disease. Acta Orthop Scand. 1964;34:183-97.
Copyright 2007 by The Journal of Bone and Joint Surgery, Incorporated
21. Rowe SM, Chung JY, Moon ES, Yoon TR, Jung ST, Kim SS. Dysplasia epiphysealis capitis femoris: Meyer dysplasia. J Pediatr Orthop. 2005;25:18-21.