JBJS | Legg-Calvé-Perthes Disease in Girls. A Comparison of the Results with Those Seen in Boys*

The Journal of Bone & Joint Surgery, Volume 80, Issue 9

Articles | September 01, 1998

Legg-Calvé-Perthes Disease in Girls. A Comparison of the Results with Those Seen in Boys*

JAMES T. GUILLE, M.D.†; GLENN E. LIPTON, B.A.‡; GEORGE SZÖKE, M.D.§; J. RICHARD BOWEN, M.D.§; H. THEODORE HARCKE, M.D.§; JOSEPH J. GLUTTING, PH.D.#, WILMINGTON, DELAWARE

View Disclosures and Other Information

Investigation performed at the Alfred I. duPont Institute, Wilmington

The Journal of Bone & Joint Surgery. 1998; 80:1256-1263

5 Recommendations (Recommend) | 3 Comments | Saved by 3 Users Save Case

Article

Tables

References

text A A A

Abstract

We reviewed the records and roentgenograms of all patients with Legg-Calvé-Perthes disease who had been seen at our institution between 1940 and 1996. One hundred and five girls (122 hips) and 470 boys (531 hips) were identified. Thus, 18 per cent of the 575 patients in the present series were girls. Seventeen (16 per cent) of the girls and sixty-one (13 per cent) of the boys had bilateral involvement. Although more girls than boys had severe involvement of the femoral head and the lateral pillar, we could not detect a significant difference between the two groups with respect to the distribution of the involvement of the hips according to the system of Catterall or the lateral pillar classification (p > 0.05, beta = 0.99). Serial roentgenograms that showed all four stages of the disease according to the system of Waldenström were available for fifty-two hips in girls and 184 hips in boys. A review of these roentgenograms revealed that the average ages of the girls at the stages of necrosis, fragmentation, reossification, and remodeling were 6.8, 7.3, 7.9, and 9.5 years, respectively, whereas the average ages of the boys were 6.8, 7.3, 7.9, and 9.9 years, respectively. Girls, however, had closure of the affected proximal femoral physis at an average age of 12.9 years, whereas boys had closure at an average age of 15.8 years. Therefore, girls had a shorter potential period for remodeling of the femoral head (average, 3.4 years) compared with boys (average, 5.9 years). Sixty-four girls (seventy-eight hips) and 363 boys (416 hips) had reached skeletal maturity by the time of the latest follow-up and were evaluated according to the system of Stulberg et al.; we could not detect a significant difference between boys and girls with respect to the distribution of the hips according to this system (p > 0.05, beta = 0.99). Although the numbers were too small for statistical analysis, our findings suggest that boys and girls who have the same Catterall or lateral pillar classification at the time of the initial evaluation can be expected to have similar outcomes according to the classification system of Stulberg et al.

Figures in this Article

Introduction

Introduction | Materials and Methods | Results | Discussion | References

Legg-Calvé-Perthes disease is a disorder involving the growth of the immature proximal femoral epiphysis. The etiology of the condition remains unknown, but various hypotheses have been proposed^11,29. Regardless of the cause, Legg-Calvé-Perthes disease primarily affects boys; more specifically, the condition is four to five times more common in boys than in girls. The prevalence of the condition in different regions of the world has ranged from one in 1200 in Massachusetts in the United States²⁰ to one in 12,500 in England¹. The frequency of involvement of the right and left sides is approximately equal²⁸. The frequency of bilateral involvement has ranged from 8 to 13 per cent, with various authors having reported bilaterality in seven (8 per cent) of eighty-seven patients²⁰, seventeen (9 per cent) of 185 patients⁴, twelve (10 per cent) of 121 patients⁵, eighteen (10 per cent) of 185 patients²⁷, thirty-five (11 per cent) of 310 patients²⁹, and twenty-five (13 per cent) of 188 patients⁸.

Numerous articles have been written on the subject of Legg-Calvé-Perthes disease and its treatment^1,3-9,11-29. Most patients in these reports have been boys, with girls mentioned only occasionally. Previous investigators have believed that girls who have Legg-Calvé-Perthes disease have worse outcomes than boys^{5,7,14,17,19,22}. This belief has been attributed to the fact that girls reach skeletal maturity earlier than boys and therefore have a shorter potential period for remodeling of the femoral head. One investigator asserted that the disease is more severe in girls and that this increased severity could be the cause of the unsatisfactory results⁶. Other authors have attributed the poor outcomes to the age at the onset of the disease^8,19. To resolve these issues, we reviewed the records and roentgenograms of all patients with Legg-Calvé-Perthes disease who had been seen at our institution. The purpose of the present study was to characterize the disease in girls and to prove or disprove the impression that girls have more severe involvement and worse outcomes than boys.

*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.

†Department of Orthopaedic Surgery, Allegheny University Hospitals, Broad and Vine Streets, Philadelphia, Pennsylvania 19102.

‡Temple University School of Medicine, Broad and Ontario Streets, Philadelphia, Pennsylvania 19140.

§Departments of Orthopaedics (G. S. and J. R. B.) and Medical Imaging (H. T. H.), Alfred I. duPont Institute, P.O. Box 269, Wilmington, Delaware 19899.

#University of Delaware, School of Education, Newark, Delaware 19716.

*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.

†Department of Orthopaedic Surgery, Allegheny University Hospitals, Broad and Vine Streets, Philadelphia, Pennsylvania 19102.

‡Temple University School of Medicine, Broad and Ontario Streets, Philadelphia, Pennsylvania 19140.

§Departments of Orthopaedics (G. S. and J. R. B.) and Medical Imaging (H. T. H.), Alfred I. duPont Institute, P.O. Box 269, Wilmington, Delaware 19899.

#University of Delaware, School of Education, Newark, Delaware 19716.

View Larger

TABLE I DATA ON THE PATIENTS

*Adequate roentgenograms were available for fifty-two hips in girls and 184 hips in boys. †Roentgenograms that had been made at the approximate time of closure of the capital femoral physis were available for thirty-two hips in girls and seventy-eight hips in boys.

	Girls	Boys
No. of patients	105	470
No. of hips	122	531
Average age at onset of disease (yrs.)	6.4	6.7
Average age at diagnosis (yrs.)	6.7	6.9
Unilateral involvement (no. of hips)	88 (84%)	409 (87%)
Bilateral involvement (no. of hips)	17 (16%)	61 (13%)
Roentgenographic classification (no. of hips)
Catterall⁵
Group 1	8 (7%)	29 (6%)
Group 2	20 (17%)	100 (20%)
Group 3	51 (43%)	177 (35%)
Group 4	40 (34%)	193 (39%)
Total	119 (100%)	499 (100%)
Lateral pillar¹⁵
Group A	23 (21%)	132 (29%)
Group B	41 (37%)	163 (36%)
Group C	48 (43%)	156 (35%)
Total	112 (100%)	451 (100%)
Stulberg et al.
Class I	23 (29%)	169 (41%)
Class II	21 (27%)	117 (28%)
Class III	21 (27%)	87 (21%)
Class IV	12 (15%)	42 (10%)
Class V	1 (1%)	1 (0.2%)
Total	78 (100%)	416 (100%)
Average age at each stage of disease* (yrs.)
Necrosis	6.8	6.8
Fragmentation	7.3	7.3
Reossification	7.9	7.9
Remodeling	9.5	9.9
Average age at skeletal maturity† (yrs.)	12.9	15.8

Add to My JBJS

TABLE I DATA ON THE PATIENTS

	Girls	Boys
No. of patients	105	470
No. of hips	122	531
Average age at onset of disease (yrs.)	6.4	6.7
Average age at diagnosis (yrs.)	6.7	6.9
Unilateral involvement (no. of hips)	88 (84%)	409 (87%)
Bilateral involvement (no. of hips)	17 (16%)	61 (13%)
Roentgenographic classification (no. of hips)
Catterall⁵
Group 1	8 (7%)	29 (6%)
Group 2	20 (17%)	100 (20%)
Group 3	51 (43%)	177 (35%)
Group 4	40 (34%)	193 (39%)
Total	119 (100%)	499 (100%)
Lateral pillar¹⁵
Group A	23 (21%)	132 (29%)
Group B	41 (37%)	163 (36%)
Group C	48 (43%)	156 (35%)
Total	112 (100%)	451 (100%)
Stulberg et al.
Class I	23 (29%)	169 (41%)
Class II	21 (27%)	117 (28%)
Class III	21 (27%)	87 (21%)
Class IV	12 (15%)	42 (10%)
Class V	1 (1%)	1 (0.2%)
Total	78 (100%)	416 (100%)
Average age at each stage of disease* (yrs.)
Necrosis	6.8	6.8
Fragmentation	7.3	7.3
Reossification	7.9	7.9
Remodeling	9.5	9.9
Average age at skeletal maturity† (yrs.)	12.9	15.8

View Larger

TABLE II OUTCOME ACCORDING TO THE SYSTEM OF STULBERG ET AL. AT THE LATEST FOLLOW-UP EXAMINATION

	Girls (no. of hips)						Boys (no. of hips)
Total	Class I	Class II	Class III	Class IV	Class V	Total	Class I	Class II	Class III	Class IV	Class V
Outcome with respect to Catterall classification⁵
Group 1	5	5	0	0	0	0	21	20 (95%)	0	1 (5%)	0	0
Group 2	11	5	5	0	1	0	85	63 (74%)	17 (20%)	3 (4%)	2 (2%)	0
Group 3	36	8 (22%)	12 (33%)	9 (25%)	6 (17%)	1 (3%)	137	48 (35%)	47 (34%)	32 (23%)	9 (7%)	1 (1%)
Group 4	26	5 (19%)	4 (15%)	12 (46%)	5 (19%)	0	164	36 (22%)	52 (32%)	48 (29%)	28 (17%)	0
Outcome with respect to lateral pillar classification¹⁵
Group A	15	10	5	0	0	0	119	94 (79%)	19 (16%)	4 (3%)	2 (2%)	0
Group B	25	9 (36%)	9 (36%)	4 (16%)	3 (12%)	0	145	58 (40%)	56 (39%)	22 (15%)	9 (6%)	0
Group C	33	3 (9%)	6 (18%)	16 (49%)	8 (24%)	0	134	12 (9%)	38 (28%)	56 (42%)	27 (20%)	1 (1%)

Add to My JBJS

TABLE II OUTCOME ACCORDING TO THE SYSTEM OF STULBERG ET AL. AT THE LATEST FOLLOW-UP EXAMINATION

	Girls (no. of hips)						Boys (no. of hips)
Total	Class I	Class II	Class III	Class IV	Class V	Total	Class I	Class II	Class III	Class IV	Class V
Outcome with respect to Catterall classification⁵
Group 1	5	5	0	0	0	0	21	20 (95%)	0	1 (5%)	0	0
Group 2	11	5	5	0	1	0	85	63 (74%)	17 (20%)	3 (4%)	2 (2%)	0
Group 3	36	8 (22%)	12 (33%)	9 (25%)	6 (17%)	1 (3%)	137	48 (35%)	47 (34%)	32 (23%)	9 (7%)	1 (1%)
Group 4	26	5 (19%)	4 (15%)	12 (46%)	5 (19%)	0	164	36 (22%)	52 (32%)	48 (29%)	28 (17%)	0
Outcome with respect to lateral pillar classification¹⁵
Group A	15	10	5	0	0	0	119	94 (79%)	19 (16%)	4 (3%)	2 (2%)	0
Group B	25	9 (36%)	9 (36%)	4 (16%)	3 (12%)	0	145	58 (40%)	56 (39%)	22 (15%)	9 (6%)	0
Group C	33	3 (9%)	6 (18%)	16 (49%)	8 (24%)	0	134	12 (9%)	38 (28%)	56 (42%)	27 (20%)	1 (1%)

Materials and Methods

Introduction | Materials and Methods | Results | Discussion | References

We compiled a list of every patient with Legg-Calvé-Perthes disease who had been seen at the Alfred I. duPont Institute between its opening on July 1, 1940, and December 31, 1996. A total of 575 such patients were identified, and 105 (18 per cent) of them were girls.

The study was divided into two parts. The first part included every girl in the series and concerned the demographic features of Legg-Calvé-Perthes disease in the female population. Data on the boys were included for comparison. The second part included only the girls who had reached skeletal maturity and for whom adequate documentation and roentgenograms were available. Again, data on the boys who met these criteria were used for comparison (Table I). The hips were graded by three of us (J. T. G., G. S., and J. R. B.) according to the system of Catterall⁵, the lateral pillar classification¹⁵, and the system of Stulberg et al. Every available roentgenogram for each patient was reviewed chronologically and concurrently until a unanimous decision was made with regard to these parameters. Copies of the original articles describing these classification schemes were posted during this process for reference. Chi-square analyses and power studies were performed to determine whether there were significant differences between girls and boys with respect to the classification of the hips according to the three systems. In addition, studies in the English-language literature in which the results for male and female patients had been presented separately were evaluated with chi-square analysis for global comparison.

According to the system of Catterall⁵, group 1 is characterized by involvement of approximately 25 per cent of the anterior part of the epiphysis, no collapse of the epiphysis, and no formation of a sequestrum; group 2, by involvement of approximately 50 per cent of the anterolateral part of the epiphysis, collapse of the involved segment, preservation of the medial and lateral pillars, and formation of a sequestrum; group 3, by involvement of approximately 75 per cent of the epiphysis, formation of a sequestrum, so-called head-within-a-head appearance, and preservation of the medial pillar; and group 4, by total epiphyseal collapse and sequestration. According to the lateral pillar classification¹⁵, group A indicates that the lateral pillar has a normal height; group B, that the lateral pillar has more than 50 per cent of its normal height; and group C, that the lateral pillar has less than 50 per cent of its normal height. According to the system of Stulberg et al., class I indicates a normal hip; class II, a spherical femoral head, coxa breva, coxa magna, and a steep acetabulum; class III, a non-spherical oval femoral head, coxa breva, coxa magna, and a steep acetabulum; class IV, a flat femoral head and abnormalities of the femoral head, femoral neck, and acetabulum; and class V, a flat femoral head, a normal femoral neck, and a normal acetabulum.

Roentgenograms of the hand were evaluated by one of us (H. T. H.), a pediatric skeletal radiologist, to determine bone age. Roentgenograms of the hand and wrist were available for thirty-two (30 per cent) of the 105 girls; the roentgenograms had been made on the left side for twenty-nine patients and, for reasons unknown, on the right side for three. The pictorial method of comparison with use of the atlas of Greulich and Pyle was applied with two guiding principles. First, the development of the distal phalanges was used as the determining indicator when there was a discrepancy between carpal and phalangeal epiphyseal maturity. Second, if the development seen on a roentgenogram was judged to be equally placed between two standards, the bone age was placed halfway between the two. For example, a bone age of six years and four months was assigned when the roentgenogram was judged to reflect development halfway between the standards of five years and nine months and six years and ten months.

Results

Introduction | Materials and Methods | Results | Discussion | References

Data at the Time of Presentation

Girls

One hundred and five girls (122 hips) who had Legg-Calvé-Perthes disease were seen at our institution during a fifty-seven-year period. Seventeen girls (16 per cent) had bilateral involvement, and eighty-eight (84 per cent) had unilateral involvement. Of the eighty-eight girls who had unilateral involvement, forty-four had involvement of the right hip and forty-four had involvement of the left hip. The average age at the onset of symptoms was 6.4 years (range, 2.0 to 14.1 years), and the average age at the time of diagnosis was 6.7 years (range, 2.3 to 14.7 years). Twenty-eight girls were first seen because of a limp without pain; fifty-one, because of pain in the hip or groin; and eleven, because of pain in the knee. The presenting symptom was unclear from the review of the charts of the other fifteen patients. Anteroposterior and lateral roentgenograms that were adequate for the determination of the Catterall classification⁵ were available for 119 hips: eight hips (7 per cent) were in group 1, twenty (17 per cent) were in group 2, fifty-one (43 per cent) were in group 3, and forty (34 per cent) were in group 4. Anteroposterior and frog-leg lateral roentgenograms that were adequate for the determination of the lateral pillar classification¹⁵ were available for 112 hips: twenty-three hips (21 per cent) were in group A, forty-one (37 per cent) were in group B, and forty-eight hips (43 per cent) were in group C.

Serial roentgenograms that showed all four stages of the disease according to the system of Waldenström were available for fifty-two hips. The average age of the patients was 6.8 years (range, 2.3 to 13.1 years) at the necrosis stage, 7.3 years (range, 2.5 to 13.5 years) at the fragmentation stage, 7.9 years (range, 2.9 to 13.9 years) at the reossification stage, and 9.5 years (range, 3.6 to 14.6 years) at the beginning of remodeling. Roentgenograms that had been made during adolescence at the approximate time of closure of the affected capital femoral epiphysis were available for thirty-two hips. The average age of the patients at the time of closure was 12.9 years (range, 11.0 to 15.0 years).

Boys

Data on all 470 boys (531 hips) were included for comparison. Sixty-one boys (13 per cent) had bilateral involvement, and 409 (87 per cent) had unilateral involvement. Of the 409 boys who had unilateral involvement, 207 had involvement of the right hip and 202 had involvement of the left hip. The average age at the onset of symptoms was 6.7 years (range, 2.1 to 14.7 years), and the average age at the time of diagnosis was 6.9 years (range, 2.2 to 14.7 years). One hundred and eighty-five boys were first seen because of a limp without pain; seventy-four, because of pain in the knee; and 173, because of pain in the hip or groin. The presenting symptom was unclear for thirty-eight patients. Roentgenograms that were adequate for the determination of the Catterall classification⁵ were available for 499 hips: twenty-nine hips (6 per cent) were in group 1, 100 (20 per cent) were in group 2, 177 (35 per cent) were in group 3, and 193 (39 per cent) were in group 4. Roentgenograms that were adequate for the determination of the lateral pillar classification¹⁵ were available for 451 hips: 132 hips (29 per cent) were in group A, 163 (36 per cent) were in group B, and 156 (35 per cent) were in group C.

Serial roentgenograms that showed all four stages of the disease according to the system of Waldenström were available for 184 hips. The average age of the patients was 6.8 years (range, 2.2 to 13.1 years) at the necrosis stage, 7.3 years (range, 2.9 to 13.7 years) at the fragmentation stage, 7.9 years (range, 3.5 to 14.7 years) at the reossification stage, and 9.9 years (range, 3.9 to 15.9 years) at the beginning of remodeling. Roentgenograms that had been made around the time of closure of the affected capital femoral epiphysis were available for seventy-eight hips. The average age of the patients at the time of closure was 15.8 years (range, 12.8 to 17.6 years).

Comparison of Girls and Boys

We could detect no significant difference between girls and boys with respect to the distribution of hips according to the classification of Catterall⁵ (chi square = 2.72, degrees of freedom = 2, p < 0.5618). A power analysis, performed with the level of significance set at p < 0.05, an estimated medium effect-size, and a sample size of 563, showed a beta of 0.99. Therefore, there was a 99 per cent chance that a significant difference would have been found if such a difference had been present.

We also could detect no significant difference between girls and boys with respect to the distribution of hips according to the lateral pillar classification¹⁵ (chi square = 4.18, degrees of freedom = 2, p < 0.1213). A power analysis, performed with the level of significance set at p < 0.05, an estimated medium effect-size, and a sample size of 563, again showed a beta of 0.99.

Results at the Latest Follow-up Examination

Female Patients

Sixty-four female patients (seventy-eight hips) who had reached skeletal maturity had adequate roentgenograms and clinical follow-up documentation. Skeletal maturity was evidenced by bilateral closure of the capital femoral physis and bilateral fusion of the iliac apophysis. The exact age at which skeletal maturity had been reached was not available for all patients as several patients had not been evaluated with roentgenograms during adolescence. Fourteen patients had bilateral involvement, and fifty had unilateral involvement. Of the fifty patients who had unilateral involvement, twenty-five had involvement of the right hip and twenty-five had involvement of the left hip. The average age at the latest follow-up examination was 20.2 years (range, 11.0 to 52.0 years), and the average duration of follow-up was 13.4 years (range, 3.1 to 44.6 years).

Roentgenograms that were adequate for the determination of the Catterall classification⁵ were available for all seventy-eight hips: five hips (6 per cent) were in group 1, eleven (14 per cent) were in group 2, thirty-six (46 per cent) were in group 3, and twenty-six (33 per cent) were in group 4. All five hips that had group-1 involvement were rated as class I according to the system of Stulberg et al. Of the eleven hips that had group-2 involvement, five were class I, five were class II, and one was class IV according to the system of Stulberg et al. Of the thirty-six hips that had group-3 involvement, eight (22 per cent) were class I, twelve (33 per cent) were class II, nine (25 per cent) were class III, six (17 per cent) were class IV, and one (3 per cent) was class V according to the system of Stulberg et al. Of the twenty-six hips that had Catterall group-4 involvement, five (19 per cent) were class I, four (15 per cent) were class II, twelve (46 per cent) were class III, and five (19 per cent) were class IV according to the system of Stulberg et al. (Table II).

Roentgenograms that were adequate for the determination of the lateral pillar classification¹⁵ were available for seventy-three of the seventy-eight hips: fifteen hips (21 per cent) were in group A, twenty-five (34 per cent) were in group B, and thirty-three (45 per cent) were in group C. Of the fifteen hips in group A, ten were class I and five were class II according to the system of Stulberg et al. Of the twenty-five hips in group B, nine (36 per cent) were class I, nine (36 per cent) were class II, four (16 per cent) were class III, and three (12 per cent) were class IV according to the system of Stulberg et al. Of the thirty-three hips in group C, three (9 per cent) were class I, six (18 per cent) were class II, sixteen (48 per cent) were class III, and eight (24 per cent) were class IV according to the system of Stulberg et al. (Table II).

The average chronological age of the thirty-two girls for whom roentgenograms of the hand and wrist were available was 7.6 years (range, 2.7 to 12.3 years) at the time of the initial visit. The average bone age was 6.8 years (range, 1.5 to 13.0 years). Therefore, the average difference was 0.8 year, which is not a marked difference when the standard deviations for normal bone age are considered. For example, a seven-year-old girl with a normal bone age has one standard deviation in bone age of 8.3 months.

Male Patients

Three hundred and sixty-three male patients (416 hips) who had reached skeletal maturity had adequate roentgenograms and clinical follow-up documentation. Skeletal maturity was evidenced by bilateral closure of the capital femoral physis and bilateral fusion of the iliac apophysis. The exact age at which skeletal maturity had been reached was not available for all patients as several patients had not been evaluated with roentgenograms during their teenage years. Fifty-three patients had bilateral involvement, and 310 had unilateral involvement. Of the 310 patients who had unilateral involvement, 157 had involvement of the right hip and 153 had involvement of the left hip. The average age at the time of the latest follow-up visit was 17.4 years (range, 11.9 to 45.0 years), and the average duration of follow-up was 10.8 years (range, 2.1 to 38.2 years).

Roentgenograms that were adequate for the determination of the Catterall classification⁵ were available for 407 hips: twenty-one hips (5 per cent) were in group 1, eighty-five (21 per cent) were in group 2, 137 (34 per cent) were in group 3, and 164 (40 per cent) were in group 4. Of the twenty-one hips that had group-1 involvement, twenty (95 per cent) were class I and one (5 per cent) was class III according to the system of Stulberg et al. Of the eighty-five hips that had group-2 involvement, sixty-three (74 per cent) were class I, seventeen (20 per cent) were class II, three (4 per cent) were class III, and two (2 per cent) were class IV according to the system of Stulberg et al. Of the 137 hips that had group-3 involvement, forty-eight (35 per cent) were class I, forty-seven (34 per cent) were class II, thirty-two (23 per cent) were class III, nine (7 per cent) were class IV, and one (1 per cent) was class V according to the system of Stulberg et al. Of the 164 hips that had group-4 involvement, thirty-six (22 per cent) were class I, fifty-two (32 per cent) were class II, forty-eight (29 per cent) were class III, and twenty-eight (17 per cent) were class IV according to the system of Stulberg et al. (Table II).

Roentgenograms that were adequate for the determination of the lateral pillar classification¹⁵ were available for 398 hips: 119 hips (30 per cent) were in group A, 145 (36 per cent) were in group B, and 134 (34 per cent) were in group C. Of the 119 hips in group A, ninety-four (79 per cent) were class I, nineteen (16 per cent) were class II, four (3 per cent) were class III, and two (2 per cent) were class IV according to the system of Stulberg et al. Of the 145 hips in group B, fifty-eight (40 per cent) were class I, fifty-six (39 per cent) were class II, twenty-two (15 per cent) were class III, and nine (6 per cent) were class IV according to the system of Stulberg et al. Of the 134 hips in group C, twelve (9 per cent) were class I, thirty-eight (28 per cent) were class II, fifty-six (42 per cent) were class III, twenty-seven (20 per cent) were class IV, and one (1 per cent) was class V according to the system of Stulberg et al. (Table II).

Comparison of Female and Male Patients

We could detect no significant difference between the female patients and the male patients with respect to the distribution of hips according to the classification of Stulberg et al. (chi square = 6.66, degrees of freedom = 4, p < 0.1540). A power analysis, performed with the level of significance set at p < 0.05, an estimated medium effect-size, and a sample size of 563, showed a beta of 0.99.

Discussion

Introduction | Materials and Methods | Results | Discussion | References

Much of the literature on Legg-Calvé-Perthes disease is based on studies that have included a preponderance of boys. Therefore, statements that have been made in the literature concerning the demographic features and the results of treatment may not be applicable to girls. In a recent report on boys who had congenital dislocation of the hip, a condition that is usually seen in girls, Borges et al. showed that boys had a more difficult course and did not always respond well to standard modes of treatment. These findings are one reason that we studied the characteristics and outcome of Legg-Calvé-Perthes disease in girls. To our knowledge, we are reporting on the largest series to date of patients with Legg-Calvé-Perthes disease who were seen at one institution.

We found that most patients (84 per cent of the girls and 87 per cent of the boys) had unilateral involvement. The frequency of involvement of the right and left sides was approximately equal in both groups. The girls and boys had a similar average age at onset (6.4 and 6.7 years, respectively). This finding agrees with that of Cameron and Izatt, but it differs from those of other studies, in which the onset was found to occur earlier in girls than in boys^16,18. Lovell et al. stated that girls had a different age at onset than boys but did not mention if the onset in girls was earlier or later than that in boys. Fisher noted a higher proportion of girls and a higher frequency of bilateral involvement in younger age-groups; he believed that increased skeletal maturity somehow confers resistance to development of the disease. In our series, the average ages at the stages of necrosis, fragmentation, and reossification were the same for boys and girls. The remodeling stage began approximately 0.4 year earlier in girls than it did in boys. Closure of the affected proximal femoral epiphysis occurred an average of 2.9 years earlier in girls than it did in boys, which implies that girls have a shorter potential period for remodeling of the femoral head.

Although it should be noted that roentgenograms of the hand and wrist were available for only thirty-two (30 per cent) of the 105 girls, we found only a slight delay (0.8 year) in bone age. This finding agrees with that of Ralston. Perhaps this finding would have been different if the size of the population had been larger. Burwell et al. reported that girls display a pattern of impaired growth that is similar to that seen in boys. Girdany and Osman noted that the delay in skeletal maturation was "not as conspicuous" in girls as it was in boys. Weiner and O'Dell showed that most children (106 of 176 boys and girls) had an average or above-average body weight when the disease occurred. They also noted that, of twenty-two girls for whom the height was recorded, fourteen were below the forty-first percentile and eight were below the sixteenth percentile. Harrison et al. reported that the phenomenon of so-called skeletal standstill, in which the bone age does not change for several years, occurred only in boys.

We also found that the percentage of involvement of the femoral head and the lateral pillar is not significantly higher in girls than in boys. The converse commonly has been reported in the literature^5,14,19, but it has not been documented in a large study, as far as we know. We could detect no significant difference between girls and boys with respect to the distribution of hips in Catterall groups 1, 2, 3, and 4 or lateral pillar groups A, B, and C. In both groups of patients, there was a preponderance of hips (76 per cent for girls and 74 per cent for boys) in Catterall groups 3 and 4.

We believe that a comparison of the results in our female patients with those in our male patients was appropriate as both groups were seen during the same time-period and by the same orthopaedic surgeons. In addition, treatment regimens were rendered regardless of gender, and a review of the medical records showed no bias in the management of our patients. The fundamental principles of the treatment of Legg-Calvé-Perthes disease were applied to all patients throughout the period of study¹⁸. Maintenance of range of motion and prevention of stiffness were paramount. Physical therapy, traction, and casts were used as necessary. Containment was achieved by both non-operative and operative methods, including casts, braces, and osteotomies^26,28. Salvage procedures were performed as needed.

Catterall⁵, in his classic report on the natural history of the disease, noted that girls have a poor overall prognosis compared with boys. The average age at the onset of the disease was similar for both genders, and most of the girls had group-3 or 4 involvement. We believe that Catterall may have misinterpreted the report by Evans when he stated that Evans explained that the poor results in his series⁷ were due to the fact that most of the girls were in the older age-groups. In fact, approximately half of the girls and boys in the series of Evans were less than six years old. Catterall attributed the poor prognosis for girls to the fact that most of the girls in his series⁵ had group-3 or 4 involvement. However, a review of Catterall's Table IV, which details the results according to gender, reveals that three girls had group-1 involvement, four had group-2 involvement, five had group-3 involvement, and three had group-4 involvement. In comparison, twenty-eight boys had group-1 involvement, twenty-seven had group-2 involvement, seventeen had group-3 involvement, and seven had group-4 involvement. We could detect no significant difference between girls and boys with respect to the distribution of hips according to the Catterall classification when we applied chi-square analysis to those data (chi square = 3.35, p > 0.05). According to Catterall, the result was considered to be good if the hip had no symptoms and a full range of motion. Roentgenographically, the femoral head was round and well contained in the acetabulum. There could be no adaptive changes in the acetabulum and no increase in the medial joint space, but there could be some loss of epiphyseal height. The result was considered to be fair if the hip had no symptoms and a small decrease in the range of motion. Roentgenographically, the femoral head was round but slightly broadened and was contained in the acetabulum. There could be adaptive acetabular changes, and there was loss of epiphyseal height. The result was considered to be poor if there was a decreased range of motion, with or without pain in the hip. Roentgenographically, the femoral head was flattened, broad, and irregular, with as much as 20 per cent of the head uncovered by the acetabulum. The acetabulum showed adaptive changes, and the joint space was widened medially. In the series of Catterall⁵, six girls had a good result; five, a fair result; and four, a poor result. In comparison, fifty-one boys had a good result; nineteen, a fair result; and nine, a poor result. We could detect no significant difference between girls and boys with respect to the results when we applied chi-square analysis to those data (chi square = 3.81, p > 0.05). Catterall stated that the prognosis was approximately the same for a boy and a girl who had similar involvement of the femoral head, although no statistical data were provided.

Evans reported on a series of fifty-two children (thirteen girls and thirty-nine boys) who had Legg-Calvé-Perthes disease. Approximately half of the children were first seen before the age of six years. The result was considered to be good when anteroposterior and lateral roentgenograms showed a circular femoral head, a normal acetabulum, and no osteoarthrosis and the affected side would appear normal if it was not compared with the contralateral side. The result was considered to be fair when roentgenograms showed an elliptical yet smooth and regular femoral head on one or both projections, no osteoarthrosis, and a well adapted acetabulum that covered the femoral head. The result was considered to be poor when both projections showed a deformity of the femoral head and varying degrees of osteoarthrosis. Only six of the thirteen girls had a good or fair result, whereas thirty of the thirty-nine boys had a good or fair result. We could not detect a significant difference between the results for girls and those for boys when chi-square analysis was applied to those data (chi square = 4.49, p > 0.05). Evans concluded that the prognosis was worse for girls, most likely because they reach skeletal maturity earlier than boys.

Lovell et al. reported on a series of 136 girls (155 hips) who had been seen at three tertiary pediatric orthopaedic centers. They found that the age at the onset of the disease and the Catterall classification⁵ were consistently associated with fair and poor results. Those authors concluded that if girls had a worse prognosis than a comparable series of boys, this finding would be attributable to the age at onset and the Catterall grade rather than to gender. As only an abstract of the report was published, the criteria for classifying the results were not included.

Mukherjee and Fabry reported that girls had markedly worse results than boys. A complex system for the evaluation of roentgenographic, clinical, subjective, and functional data was used to classify the results. Roentgenographic parameters included the sphericity of the femoral head (as assessed with Mose circles), the epiphyseal quotient, and the center-edge angle. The result that was obtained with use of Mose circles was rated as good when the measurements on anteroposterior and lateral roentgenograms were the same, as fair when the measurements varied by two millimeters or less, and as poor when the measurements varied by more than two millimeters. The epiphyseal quotient was rated as good when it was 0.6 or more, as fair when it was between 0.4 and 0.6, and as poor when it was less than 0.4. The center-edge angle was rated as good when it was more than 25 degrees, as fair when it was 20 to 25 degrees, and as poor when it was less than 20 degrees. Collectively, three good ratings were considered to be an excellent roentgenographic result; two good ratings and one fair or poor rating, a good roentgenographic result; one good rating and two fair or poor ratings, a fair roentgenographic result; and three fair or poor ratings, a poor roentgenographic result. Clinical evaluation was based on range of motion. A full range of motion was considered to be a good result; a slight restriction of motion, a fair result; and gross limitation, a poor result. Subjective and functional evaluation was based on a modified Harris hip score, with a score of 90 points or more indicating a good result; 80 to 89 points, a fair result; and 79 points or less, a poor result. The findings on gross examination at the end of treatment revealed that three girls had an excellent result; eleven, a good result; none, a moderate result; and thirteen, a poor result. In comparison, twenty boys had an excellent result; forty-two, a good result; nine, a moderate result; and eighteen, a poor result. Overall, thirteen (48 per cent) of twenty-seven girls and eighteen (20 per cent) of eighty-nine boys had a poor result. A significant difference was found between girls and boys when chi-square analysis was applied to those data (chi square = 10.31, p < 0.05). The distribution of the hips according to the Catterall classification⁵ was not given for the girls, and no reasons were given for the poor results.

In the series of Kelly et al., three of eleven girls in whom the disease had developed at or after the age of six years had a poor result compared with four (11 per cent) of thirty-eight such boys. Those authors used a combination of roentgenographic and clinical data to evaluate the results. The roentgenographic result, determined with use of Mose circles, was classified as good if the femoral head had the same radius on both the anteroposterior and the lateral roentgenograms as that on the contralateral side, as fair if the measurements varied by two millimeters or less, and as poor if the measurements varied by more than two millimeters. Collectively, good and fair roentgenographic results were rated as acceptable and poor results were rated as unacceptable. The clinical result was rated as satisfactory if the patient had no pain, no limitation of activity, a full range of motion or a decrease of motion of less than 10 per cent, and no limp or a slight limp secondary to shortening. Overall, a satisfactory clinical rating and an acceptable roentgenographic rating were considered to be a good result, a satisfactory clinical rating and an unacceptable roentgenographic rating were considered to be a fair result, and an unsatisfactory clinical rating and an unacceptable roentgenographic rating were considered to be a poor result.

Stulberg et al., in their study on the natural history of the disease, rated nineteen hips (three in girls and sixteen in boys) as class I, eighteen hips (two in girls and sixteen in boys) as class II, eleven hips (two in girls and nine in boys) as class III, ten hips (four in girls and six in boys) as class IV, and fourteen hips (one in a girl and thirteen in boys) as class V. We detected no significant difference between girls and boys when chi-square analysis was applied to those data (chi square = 5.26, p > 0.05).

In summary, although a slightly higher percentage of girls than boys had Catterall group-3 or 4 involvement as well as lateral pillar group-C involvement, the difference was not shown to be significant. The frequency of involvement of the right and left sides as well as the frequency of bilateral involvement was similar for girls and boys, and the stages of the condition occurred at approximately the same ages in girls and boys. Girls, however, had a shorter potential period for remodeling of the femoral head because they reached skeletal maturity earlier than boys. We could detect no significant difference between female patients and male patients with respect to the distribution of hips according to the classification of Stulberg et al. at the latest follow-up examination (p > 0.05, beta = 0.99). Although we were unable to perform statistical analysis on these data because of the small number of patients in some cells, it appears that girls and boys who have the same Catterall or lateral pillar classification at the initial examination can be expected to have similar results according to the classification system of Stulberg et al.

References

Introduction | Materials and Methods | Results | Discussion | References

Barker, D. J. P.; Dixon, E.; and Taylor, J. F.: Perthes' disease of the hip in three regions of England. J. Bone and Joint Surg.,60-B(4): 478-480, 1978.60-B(4)478 1978

Borges, J. L. P.; Kumar, S. J.; and Guille, J. T.: Congenital dislocation of the hip in boys. J. Bone and Joint Surg.,77-A: 975-984, July 1995.77-A975 1995

Burwell, R. G.; Dangerfield, P. H.; Hall, D. J.; Vernon, C. L.; and Harrison, M. H. M.: Perthes' disease. An anthropometric study revealing impaired and disproportionate growth. J. Bone and Joint Surg.,60-B(4): 461-477, 1978.60-B(4)461 1978

Cameron, J. M., and Izatt, M. M.: Legg-Calvé-Perthes disease. Scottish Med. J.,5: 148-154, 1960.5148 1960

Catterall, A.: The natural history of Perthes' disease. J. Bone and Joint Surg.,53-B(1): 37-53, 1971.53-B(1)37 1971

Catterall, A.: Legg-Calvé-Perthes. Part A. Morbid anatomy and natural history. In Instructional Course Lectures, American Academy of Orthopaedic Surgeons. Vol. 38, pp. 297-301. Park Ridge, Illinois, American Academy of Orthopaedic Surgeons, 1989.

Evans, D. L.: Legg-Calvé-Perthes' disease. A study of late results. J. Bone and Joint Surg.,40-B(2): 168-181, 1958.40-B(2)168 1958

Fisher, R. L.: An epidemiological study of Legg-Perthes disease. J. Bone and Joint Surg.,54-A: 769-778, June 1972.54-A769 1972

Girdany, B. R., and Osman, M. Z.: Longitudinal growth and skeletal maturation in Perthes' disease. Radiol. Clin. North America,6: 245-251, 1968.6245 1968

Greulich, W. W., and Pyle, S. I.: Radiographic Atlas of Skeletal Development of the Hand and Wrist. Ed. 2, p. 57. Stanford, California, Stanford University Press, 1959.

Guille, J. T., and Bowen, J. R.: Legg-Calvé-Perthes disease. In Principles of Orthopaedic Practice, edited by R. Dee, L. C. Hurst, M. A. Gruber, and S. A. Kottmeier. Ed. 2, pp. 723-734. New York, McGraw-Hill Health Professions Division, 1997.

Harrison, M. H. M.; Turner, M. H.; and Jacobs, P.: Skeletal immaturity in Perthes' disease. J. Bone and Joint Surg.,58-B(1): 37-40, 1976.58-B(1)37 1976

Herring, J. A.: Current concepts review. The treatment of Legg-Calvé-Perthes disease. A critical review of the literature. J. Bone and Joint Surg.,76-A: 448-458, March 1994.76-A448 1994

Herring, J. A.: Management of Perthes' disease [editorial]. J. Pediat. Orthop.,16: 1-2, 1996.161 1996

Herring, J. A.; Neustadt, J. B.; Williams, J. J.; Early, J. S.; and Browne, R. H.: The lateral pillar classification of Legg-Calvé-Perthes disease. J. Pediat. Orthop.,12: 143-150, 1992.12143 1992

Katz, J. F.: Conservative treatment of Legg-Calvé-Perthes disease. J. Bone and Joint Surg.,,49-A: 1043-1051, Sept. 1967.49-A1043 1967

Kelly, F. B.; Canale, S. T.; and Jones, R. R.: Legg-Calvé-Perthes disease. Long-term evaluation of non-containment treatment. J. Bone and Joint Surg.,62-A: 400-407, April 1980.62-A400 1980

King, E. W.; Fisher, R. L.; Gage, J. R.; and Gossling, H. R.: Ambulation-abduction treatment in Legg-Calvé-Perthes disease (LCPD). Clin. Orthop.,150: 43-48, 1980.15043 1980 [PubMed]

Lovell, W. W.; MacEwen, G. D.; Stewart, W. R.; Hopper, W. C., Jr.; Riddle, C. D., Jr.; and Collins, W. C.: Legg-Perthes disease in girls. In Proceedings of the Seventh Combined Meeting of the Orthopaedic Associations of the English-Speaking World. J. Bone and Joint Surg.,64-B(5): 637, 1982.64-B(5)637 1982

Molloy, M. K., and MacMahon, B.: Incidence of Legg-Perthes disease (osteochondritis deformans). New England J. Med.,275: 988-990, 1966.275988 1966

Mose, K.: Methods of measuring in Legg-Calvé Perthes disease with special regard to the prognosis. Clin. Orthop.,150: 103-109, 1980.150103 1980 [PubMed]

Mukherjee, A., and Fabry, G.: Evaluation of the prognostic indices in Legg-Calvé-Perthes disease: statistical analysis of 116 hips. J. Pediat. Orthop.,10: 153-158, 1990.10153 1990

Ralston, E. L.: Legg-Perthes disease and physical development. In Proceedings of the American Academy of Orthopaedic Surgeons. J. Bone and Joint Surg., 37-A: 647, June 1955.

Stulberg, S. D.; Cooperman, D. R.; and Wallensten, R.: The natural history of Legg-Calvé-Perthes disease. J. Bone and Joint Surg.,63-A: 1095-1108, Sept. 1981.63-A1095 1981

Waldenström, H.: On coxa plana. Osteochondritis deformans coxae juvenilis. Leggs disease, maladie de Calveé, Perthes krankheit. Acta Chir. Scandinavica,55: 577-590, 1923.55577 1923

Wang, L.; Bowen, J. R.; Puniak, M. A.; Guille, J. T.; and Glutting, J.: An evaluation of various methods of treatment for Legg-Calvé-Perthes disease. Clin. Orthop.,314: 225-233, 1995.314225 1995 [PubMed]

Weiner, D. S., and O'Dell, H. W.: Legg-Calvé-Perthes disease. Observations on skeletal maturation. Clin. Orthop.,68: 44-49, 1970.6844 1970 [PubMed]

Wenger, D. R.; Ward, W. T.; and Herring, J. A.: Current concepts review. Legg-Calvé-Perthes disease. J. Bone and Joint Surg.,,73-A: 778-788, June 1991.73-A778 1991

Wynne-Davies, R., and Gormley, J.: The aetiology of Perthes' disease. Genetic, epidemiological and growth factors in 310 Edinburgh and Glasgow patients. J. Bone and Joint Surg.,60-B(1): 6-14, 1978.60-B(1)6 1978

Topics

View Larger

TABLE I DATA ON THE PATIENTS

	Girls	Boys
No. of patients	105	470
No. of hips	122	531
Average age at onset of disease (yrs.)	6.4	6.7
Average age at diagnosis (yrs.)	6.7	6.9
Unilateral involvement (no. of hips)	88 (84%)	409 (87%)
Bilateral involvement (no. of hips)	17 (16%)	61 (13%)
Roentgenographic classification (no. of hips)
Catterall⁵
Group 1	8 (7%)	29 (6%)
Group 2	20 (17%)	100 (20%)
Group 3	51 (43%)	177 (35%)
Group 4	40 (34%)	193 (39%)
Total	119 (100%)	499 (100%)
Lateral pillar¹⁵
Group A	23 (21%)	132 (29%)
Group B	41 (37%)	163 (36%)
Group C	48 (43%)	156 (35%)
Total	112 (100%)	451 (100%)
Stulberg et al.
Class I	23 (29%)	169 (41%)
Class II	21 (27%)	117 (28%)
Class III	21 (27%)	87 (21%)
Class IV	12 (15%)	42 (10%)
Class V	1 (1%)	1 (0.2%)
Total	78 (100%)	416 (100%)
Average age at each stage of disease* (yrs.)
Necrosis	6.8	6.8
Fragmentation	7.3	7.3
Reossification	7.9	7.9
Remodeling	9.5	9.9
Average age at skeletal maturity† (yrs.)	12.9	15.8

Add to My JBJS

TABLE I DATA ON THE PATIENTS

	Girls	Boys
No. of patients	105	470
No. of hips	122	531
Average age at onset of disease (yrs.)	6.4	6.7
Average age at diagnosis (yrs.)	6.7	6.9
Unilateral involvement (no. of hips)	88 (84%)	409 (87%)
Bilateral involvement (no. of hips)	17 (16%)	61 (13%)
Roentgenographic classification (no. of hips)
Catterall⁵
Group 1	8 (7%)	29 (6%)
Group 2	20 (17%)	100 (20%)
Group 3	51 (43%)	177 (35%)
Group 4	40 (34%)	193 (39%)
Total	119 (100%)	499 (100%)
Lateral pillar¹⁵
Group A	23 (21%)	132 (29%)
Group B	41 (37%)	163 (36%)
Group C	48 (43%)	156 (35%)
Total	112 (100%)	451 (100%)
Stulberg et al.
Class I	23 (29%)	169 (41%)
Class II	21 (27%)	117 (28%)
Class III	21 (27%)	87 (21%)
Class IV	12 (15%)	42 (10%)
Class V	1 (1%)	1 (0.2%)
Total	78 (100%)	416 (100%)
Average age at each stage of disease* (yrs.)
Necrosis	6.8	6.8
Fragmentation	7.3	7.3
Reossification	7.9	7.9
Remodeling	9.5	9.9
Average age at skeletal maturity† (yrs.)	12.9	15.8

View Larger

TABLE II OUTCOME ACCORDING TO THE SYSTEM OF STULBERG ET AL. AT THE LATEST FOLLOW-UP EXAMINATION

	Girls (no. of hips)						Boys (no. of hips)
Total	Class I	Class II	Class III	Class IV	Class V	Total	Class I	Class II	Class III	Class IV	Class V
Outcome with respect to Catterall classification⁵
Group 1	5	5	0	0	0	0	21	20 (95%)	0	1 (5%)	0	0
Group 2	11	5	5	0	1	0	85	63 (74%)	17 (20%)	3 (4%)	2 (2%)	0
Group 3	36	8 (22%)	12 (33%)	9 (25%)	6 (17%)	1 (3%)	137	48 (35%)	47 (34%)	32 (23%)	9 (7%)	1 (1%)
Group 4	26	5 (19%)	4 (15%)	12 (46%)	5 (19%)	0	164	36 (22%)	52 (32%)	48 (29%)	28 (17%)	0
Outcome with respect to lateral pillar classification¹⁵
Group A	15	10	5	0	0	0	119	94 (79%)	19 (16%)	4 (3%)	2 (2%)	0
Group B	25	9 (36%)	9 (36%)	4 (16%)	3 (12%)	0	145	58 (40%)	56 (39%)	22 (15%)	9 (6%)	0
Group C	33	3 (9%)	6 (18%)	16 (49%)	8 (24%)	0	134	12 (9%)	38 (28%)	56 (42%)	27 (20%)	1 (1%)

Add to My JBJS

TABLE II OUTCOME ACCORDING TO THE SYSTEM OF STULBERG ET AL. AT THE LATEST FOLLOW-UP EXAMINATION

	Girls (no. of hips)						Boys (no. of hips)
Total	Class I	Class II	Class III	Class IV	Class V	Total	Class I	Class II	Class III	Class IV	Class V
Outcome with respect to Catterall classification⁵
Group 1	5	5	0	0	0	0	21	20 (95%)	0	1 (5%)	0	0
Group 2	11	5	5	0	1	0	85	63 (74%)	17 (20%)	3 (4%)	2 (2%)	0
Group 3	36	8 (22%)	12 (33%)	9 (25%)	6 (17%)	1 (3%)	137	48 (35%)	47 (34%)	32 (23%)	9 (7%)	1 (1%)
Group 4	26	5 (19%)	4 (15%)	12 (46%)	5 (19%)	0	164	36 (22%)	52 (32%)	48 (29%)	28 (17%)	0
Outcome with respect to lateral pillar classification¹⁵
Group A	15	10	5	0	0	0	119	94 (79%)	19 (16%)	4 (3%)	2 (2%)	0
Group B	25	9 (36%)	9 (36%)	4 (16%)	3 (12%)	0	145	58 (40%)	56 (39%)	22 (15%)	9 (6%)	0
Group C	33	3 (9%)	6 (18%)	16 (49%)	8 (24%)	0	134	12 (9%)	38 (28%)	56 (42%)	27 (20%)	1 (1%)