Secondary Logo

Journal Logo

Operative Treatment

Labral Support Shelf Arthroplasty for Containment in Early Stages of Legg-Calve-Perthes Disease

Bowen, J. Richard MD*; Guille, James T. MD*,†; Jeong, Changhoon MD*; Worananarat, Patarawan MD*; Oh, Chang-Wug MD*; Rodriquez, Andres MD*; Holmes, Larry PhD, DrPH; J. Rogers, Kenneth PhD, ATC*

Author Information
Journal of Pediatric Orthopaedics: September 2011 - Volume 31 - Issue - p S206-S211
doi: 10.1097/BPO.0b013e31822910ba
  • Free


The pathogenesis of Legg-Calve-Perthes disease1–3 involves necrosis of varying extents in the proximal femoral epiphysis and physis. The disease progresses through 4 stages as described by Waldenstrom,4–6 which include necrosis, fragmentation (resorption), reossification, and remodeling (remaining growth).7 During the stages of necrosis and fragmentation, the necrotic femoral epiphysis may collapse resulting in subluxation and impingement (hinged subluxation8–10) of the hip.11 After the lateral third of the femoral epiphysis reossifies, further collapse of the femoral epiphysis is minimal.12 Children who are older than 6 years of age and who have an area of necrosis greater than half of the femoral epiphysis are at risk of femoral head deformity.13,14 The goals of treatment in the early stages of Legg-Calve-Perthes disease (necrosis and fragmentation) are to prevent deformity of the femoral head and hip impingement. Containment of the femoral epiphysis within the acetabulum has been a popular strategy of treatment and several methods to achieve containment have been postulated, which include Petri casting,15 varus femoral osteotomy,14,16,17 innominate osteotomy of Salter,12,18–21 and recently the labral support shelf arthroplasty has been proposed.22–30

This study presents a variation in the technique of performing the labral support shelf arthroplasty, which uses a minimal incision of approximately 2.5 cm, uses an arthroscope for visualization, and forms the labral support shelf with a mixture of autograft and allograft buttress bone.28 The outcome in a group of hips undergoing this minimal-incision variation is compared by the Stulberg classification31 with a group treated by the traditional anterolateral open operative approach, which uses the bikini incision of Salter, exposes the anterior iliac wing, and uses autograph bone for the labral support shelf. In addition, the Stulberg classification outcome of both groups of labral support shelf arthroplasties are then compared with other containment techniques used at the Alfred I. duPont Hospital of the Nemours Foundation consisting of Petric cast, varus femoral osteotomy, and innominate osteotomy of Salter.

Our hypothesis was that the labral support shelf arthroplasty by either the traditional method or the minimal-incision variation offers similar Waldenstrom class outcomes when compared with a Petric cast, a varus femoral osteotomy, or an innominate osteotomy of Salter.


The labral support shelf arthroplasty study population consisted of 42 patients who fulfilled the following criteria: (1) unilateral hip disease; (2) age equal to or older than 6 years at diagnosis; (3) Catterall group III or IV9,16; and (4) Herring B or C disease.16,17 The data for this study was retrospectively collected and the project was approved by our Institutional Review Board. The patients were divided into 2 groups based on the operative approach and the type of buttress bone graph: group I, n=20, shelf acetabuloplasty was constructed using a standard anterolateral approach and autogenous iliac bone was used for the entire shelf28,32 and group II, n=22, shelf acetabuloplasty was constructed using a 2.5 cm incision (arthroscope assistance for visualization) and the shelf was constructed with autogenous iliac bone graft placed above the hip capsule and crushed corticocancellous freeze-dried allograft was placed as a buttress on top of the autograft shelf. The average amount of corticocancellous freeze-dried allograft was 45 cc per procedure. The mean age at diagnosis was 7.0 years (range, 6 to 9.9 y) in group I and 8.2 years (range, 6 to 14 y) in group II.

Postoperative care consisted of a single-leg spica cast, which holds the involved extremity in 45 degrees of abduction, that is maintained for 6 weeks and the patient were allowed to walk with crutches in the cast as soon as postoperative pain resolves. When the cast was removed, the patient was allowed to walk with “toe touch” crutch weight bearing for an additional 6 weeks. An abduction contracture of the hip was expected to persist approximately 6 weeks after cast removal. During the 6 weeks after cast removal, an abduction hip pillow was used at night. Exercises of the hip were encouraged to maintain flexion, extension, and abduction (adduction is not encouraged for at least 6 to 8 weeks after cast removal). Then full ambulation without support was allowed. Abduction exercises of the hip to maintain at least 45 degrees abduction was continued until reossification of the lateral column of the femoral head.

The spica cast, subsequent abduction contracture, and abduction hip pillow at night offered initial containment; then the labral support (shelf) acted as a buttress against the labrum to prevent resubluxation and hinging of the hip until the lateral column of femoral epiphysis reossifies.

All radiographs were accessed with the Catterall classification,14 Waldenstrom classification,4–6,33 and Herring lateral pillar classification34,35 at the time of operation. Radiographic measurements were made that included the following: center-edge angle (CEA), acetabular index (AI), acetabular coverage (AC), and lateral subluxation (LS) preoperatively and at the last follow-up. LS was defined as a ratio of the width of medial joint space of the affected hip to that of the medial joint space of the unaffected hip. AC is the percentage of femoral head that is covered by the acetabulum and is calculated as the horizonal distance from the medial margin of the femoral head to the outer osseous edge of the acetubulum divided by the horozonal width of the femoral head multiplied by 100.25,28 To evaluate the extent of the extra-articular shelf compared with the whole acetabulum, we measured the shelf/acetabular (S/A) ratio,23 which was the ratio of the distance of the shelf to the distance of the total acetabulum on anteroposterior radiographs taken immediately postoperatively and at the time of follow-up. The outcome results were evaluated by using the classification of Stulberg et al.31

Statistical Analysis of Groups I and II

Data from groups I and II was analyzed with the independent sample t test to determine significance of age at diagnosis and age at operation. Fisher exact test was used to evaluate the difference in preoperative radiographic classification and the Stulberg outcome between groups. Independent sample t test was used to assess the significant difference in CEA, AI, and S/A ratios between the 2 groups. Paired t test was used to determine the significant difference of CEA, AI, and S/A ratios between preoperative and follow-up. To find the pattern of changes of S/A ratio over time after the operation, we used regression analysis. A significant level was 0.05, and all analyses were performed using STATA 10.0 (Statacorp, College Station, TX).

Statistical Analysis of Multiple Containment Methods

The Stulberg classification outcomes of the 2 groups of patients who were treated by the labral support shelf artroplasty were compared with Stulberg classification outcome data of patients treated by other containment methods in the author's hospital. The senior author17 (J.R.B.) previously published the Stulberg classification outcomes of patients treated by either Petri cast, varus femoral osteotomy, and/or inominate osteotomy of Salter. We examined the effect of type of treatment in improving Stulberg class outcomes. There were 101 treatments performed, with the most involving the “shelf” technique (n=42), Petri cast (n=29), and the least being Varus (n=15) and Salter (n=15). To determine if outcomes by Stulberg classification differed by the type of treatment performed, we used the χ2 statistic and Fisher exact test to test this hypothesis. First, we examined the distribution of Stulberg classification by treatment type and computed the χ2 statistic test. Using 0.05 (5%) type I error tolerance or level of significance, 12 degrees of freedom (n-1 for treatment class-4)×(n-1 for Stulberg class-5), implying 3×4=12, we obtained the χ2 test and the P value for the χ2 value. The same approach was used for the comparison of technique based on “good/fair” or “poor” outcome. With this method, we obtained 3 degrees of freedom, χ2 value, and the significance level for the χ2 value. The significance level was 0.05, and all analyses were performed using STATA, 11.0 (Statacorp, College Station, TX).


Preoperative Assessment

There were 20 patients (20 hips) in group I and 22 patients (22 hips) in group II (Table 1). The mean age at operation was 8.5 years (range, 6.4 to 12 y) in group I and 9.0 years (range, 6.1 to 14.1) in group II. Mean follow-up period after the labral support shelf arthroplasty was 4.3 years (range, 2 to 10 y) in group I and 3.6 (range, 2.6 to13 y) in group II. There was no statistical difference in age at diagnosis, age at labral support shelf arthroplasy, age at last follow-up, and duration of follow-up between groups (P>0.05) (Table 1). Preoperative radiographic classification including the Catterall classification, Waldenstrom classification, Herring lateral pillar classification did not show a statistical difference between the 2 groups (P>0.05) (Table 1).

Comparisons of Demography and Preoperative Roentgenographic Classifications of the Patients Who had a Shelf Acetabuloplasty

Group I

The mean CEA and AC at follow-up was increased significantly compared with the preoperative measurement (P<0.0001, Table 2). The mean AI and LS at follow-up were reduced significantly compared with the preoperative value (P<0.05, Table 2). The S/A ratio at follow-up was decreased significantly compared with the immediate postoperative S/A ratio (P<0.05, Table 3). According to the Stulberg et al31 classification, 19 hips were good (Stulberg I or II) and 11 hips had fair results (Stulberg III).

Results of Roentgenographic Measurement
Result of Shelf Ratio

Group II

The average CEA, AC, and AI were improved significantly compared with the preoperative CEA and AI (P<0.05, Table 2). The mean value of LS at follow-up in this group was significantly improved compared with the preoperative measurement. The S/A ratio decreased gradually with time after the operation in a similar pattern as group I. According to the Stulberg et al31 classification, 9 hips had a good result (Stulberg I or II), 11 had a fair result (Stulberg III), and 2 had poor results (Stulberg IV).

Intergroup Comparisons

The mean values of preoperative and follow-up CEA, AI, and LS between the 2 groups did show a significant difference (P>0.05) (Table 3). The mean value of immediate postoperative S/A ratio in group II was greater than that of group I (P=0.034). However, there was no difference in measurement at follow-up between the 2 groups (P=0.822). The results between the 2 groups according to the Stulberg et al31 classification did not show any significant difference (P=0.4). There was no significant complication such as infection and nerve injury after surgery in either group.

Comparison of Multiple Containment Methods

We compared the 4 methods that is, Varus, Salter, labral support shelf arthroplasty (groups I and II), and Petri; and obtained a χ2 value of 20.60, critical value=21.03 (greater than the evidence from our data). With this critical value, we failed to reject the null hypothesis of no difference with respect to the type of surgery and Stulberg classification, χ2, 20.60 (12), P>0.05 (Table 4).

Distribution of Stulberg Class by the Technique of Treatment

Table 5 presents the reclassification of Stulberg classification into good/fair and poor outcomes, with good/fair as Stulberg I, II, and III, whereas poor as Stulberg IV and V. With this reclassification, we obtained 3 degrees of freedom (n-1 for reclassified=1×n-1 for type of surgery=3), and using 5% as the significance level, we found no statistically significant difference with respect to the surgical technique or management type, and outcomes, χ2=5.93 (3), P>0.05. The critical χ2 value with 5% significance level and 3 degrees of freedom is 7.81; and as the evidence from our data again was less than the critical value, we accepted the null hypothesis of no difference in outcome with respect to the type of treatment technique.

Outcomes of Treatment by Treatment Technique


Legg-Calve-Perthes disease may result in a deformity of the femoral head36,37 and acetabulum,38 which may cause incongruity of the hip that leads to impingement.39 Prognostic factors are the age at onset of symptoms,17,40,41 degree of necrosis of the capital femoral epiphysis,14,16 premature capital physeal growth disturbance,7,42 and deformity of the femoral head.34,37,43 During the Waldenstrom stages of necrosis and fragmentation, the femoral head has the potential to deform by collapse and lateral extrusion of the epiphysis, widening of the femoral neck, hinged subluxation, and hip joint incongruity. Containment of the femoral head within the acetabulum is currently the preferred method of treatment during these early stages of Legg-Calve-Perthes disease; however, the best method for containment remains controversial.

In this study, we show the outcome of containment by a labral support shelf arthroplasty technique and its Stulberg classification outcome compares equally with other containment methods used in the author's hospital such as Petri cast, varus femoral osteotomy, and innominate osteotomy of Salter. We believe the labral support shelf arthroplasty technique is simple to perform and does not induce a permanent deformity in the proximal femur or acetabulum.23,38 The femur is not shortened as with the femoral varus osteotomy or the acetabulum is not retroverted as in the innominate osteotomy of Salter. In addition, no metallic implants are required for fixation, therefore additional procedures for their removal are not necessary. Moreover, the shelf component of the labral support shelf arthroplasty reabsorbs,28 which prevents impingement of the hip. We believe the selection of treatment in Perthes disease involves a “preference sensitive decision” in which the choice of treatment is shared between the clinician, patient, and family.

We believe the indications of the labral support shelf arthroplasty include the following: (1) necrosis of more than 50% of the proximal capital femoral epiphysis; (2) age 6 to 11 years; (3) Waldenstrom stages of necrosis or fragmentation; and (4) reducable femoral head within the acetabulum. Ideally, the operation should be performed before substantial deformity occurs to the capital femoral epiphysis; however, mild subluxation of the hip with hinged impingement (acetabular-femoral impingement) is not a contraindication as long as the deformed femoral can be reduced within the acetabulum and under a nondeformed labrum. Children, younger than 6 years of age, often heal well without operations and a containment procedure is probably needed in only a few cases in which a labral shelf arthroplasty may be used. Contraindicatons may include the following: subluxation that cannot be reduced into the acetabulum, children younger than 6 years of age that heal well without treatment, necrosis less than 50% of the proximal capital femoral epiphysis that typically heal well without treatment. The 2 cases of poor outcomes in our series were in children older than 11 years of age and we believe that these older children may not benefit from the labral support shelf acetabuloplasty.

The technique of containment with a labral support shelf arthroplasty require special attention to several critical steps. First, the femoral epiphysis must be reduced into the acetabulum and under a nondeformed labrum before the procedure. Typically, the involved extremity requires approximately 45 degrees of abduction to achieve this containment. We recommend that the extremity be maintained in this abducted position during the procedure so that the shelf will be placed against the capsule and directly above the labrum. In addition, the indirect head of the rectus femoris muscle is retracted laterally during the capsular exposure and replaced over the graft to secure the graft onto the capsule as described by Staheli.32 Second, the shelf must be of adequate size to cover the labrum to its fibrous lateral lip; however, the shelf should not be extra large and cause hip impingement. We believe the benefit of the shelf is to stabalize the labrum for months after the procedure and until the lateral column of the femoral epiphysis reossifies. Third, the involved extremity is maintained in a single-leg spica cast for 6 weeks. This period of time allows some maturation of the shelf, which we believe prevents resubluzation of the femoral head. Fourth, after cast removal, an abduction splint is used during sleep, and maintains an abduction contracture of the hip. The abduction contracture is considered to be good because it facilitates continued containment. We use therapy to maintain 45 degrees of abduction and strongly discourage adduction of the hip. Again, we believe this persistent abduction contracture is helpful in maintaing the hip containment. The abduction contracture of the hip had spontaneouly resolved in our cases at follow-up and we believe that most contractures resolve in approximately 5 months postoperatively (Fig. 1). A weakness of this study is that we have no patient satisfaction data or hip function data.

A, Arthrogram of a boy of age 7 years with Perthes disease that shows epiphyseal deformity, extrusion, and hinged subluxation. B, Radiograph after an adductor tenotomy, containment by abduction of the hip, a labral support shelf arthroplasty, and spica casting. C, Radiograph at age 11 years that shows remodeling of the epiphysis, increased growth of the acetabulum, and almost complete reabsorbing of the shelf.

The operative approach and technique has been modified from the standard operation consisting of a bikini incision19 and exposure of the anterior iliac wing to a very limited incision of 2.5 cm in which visualization is facilitated with an arthroscope (dry, without saline). In these cases, the shelf was constructed by applying autogenous bone graft, which is obtained from the outer cortex of the iliac wing, against the capsula as described by Staheli32 and the buttress for the shelf uses allograph. We believe the use of autogenous graft directly above the capsule is critical in producing an adequate shelf that will require several years to reabsorb ( an average of 3.5 y23). The senior author used only allograph for the shelf in a very few cases, which have not been reported, and was very disapponited; in that the majority of the graft reabsorbed rapidly, almost before reossification of the lateral collumn of the femoral head.

In summary, the labral support shelf arthroplasty by either the traditional method or the minimal-incision variation offers similar Waldenstrom class outcomes when compared with a Petric cast, a varus femoral osteotomy, and/or innominate osteotomy of Salter.


1. Calve J. On a particular form of pseudo coxalgia grafted onto the characteristic deformation of proximal femur. Rev Chir. 1910;42:54
2. Legg AT. An obscure affection of the hip joint. Bost Med Surg J. 1910;162:202
3. Perthes G. About juvenile arthritis deformation. Dtsch Z Chir. 1910;107:11
4. Waldenstrom H. The first stage of coxa plana. Acta Orthop Scand. 1934;5:1
5. Waldenstrom H. The first stages of coxa plana. J Bone Joint Surg. 1938;20:559
6. Waldenstrom H. Coxa plana, osteochondritis deformans coxae, Legg-Calve-Perthes disease. Zen Tralblatt fur Chirug. 1920;47:539
7. Bowen JR, Schreiber FC, Foster BK, et al. Premature femoral neck physeal closure in Perthes disease. Clin Orthop Relat Res. 1982;171:24
8. Bowen JR, Foster BK, Hartzell CR. Legg-Calve-Perthes disease. Clin Orthop Relat Res. 1984;185:97
9. Catterall A. Adolescent hip pain after Perthes' disease. Clin Orthop Relat Res. 1986;209:65
10. Guille JY, Lipton GE, Tsirikos AI, et al. Bilateral Legg-Calve-Perthes' Disease: presentation and outcome. J Pediatr Orthop. 2002;22:458
11. Richards BS, Coleman SS. Subluxation of the femoral head in coxa plana. J Bone Joint Surg Am. 1987;69:1312
12. Thompson G, Westin GW. Legg-Calve-Perthes disease: results of discontinuing treatment in the early reossification stage. Clin Orthop Relat Res. 1979;139:70
13. Catterall A Legg-Calve-Perthes Disease. 1982 New York Churchill Livingstone
14. Catterall A. The natural history of Perthes disease. J Bone Joint Surg Am. 1971;53B:37
15. Petrie JG, Bitenc I. The abduction weight-bearing treatment in Legg-Calve-Perthes disease. J Bone Joint Surg Am. 1967;49:1483
16. Catterall A. Legg-Calve-Perthes syndrome. Clin Orthop Relat Res. 1981;158:41
17. Wang L, Bowen JR, Puniak MA, et al. An evaluation of various methods of treatment for Legg-Calve-Perthes disease. Clin Orthop Relat Res. 1995;314:225
18. Kalamchi A. Modified Salter osteotomy. J Bone Joint Surg. 1982;64A:183
19. Salter RB. Legg-Perthes Disease: treatment by innominate osteotomy. AAOS Instr Course Lect. 1973;22:309
20. Salter RB. The present status of surgical treatment of Legg-Calve-Perthes disease: current concepts review. J Bone Joint Surg Am. 1984;66:961
21. Salter RB, Thompson GH. Legg-Calve-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
22. Daly K, Bruce C, Catterall A. Lateral shelf acetabuloplasty in Perthes' disease. J Bone Joint Surg Br. 1999;81B:380
23. Domzalski ME, Glutting J, Bowen JR, et al. Lateral acetabular growth stimulation following a labral support procedure in Legg-Calve-Perthes' disease. J Bone Joint Surg Am. 2006;88A:1458
24. Gill AB. Plastic construction of an acetabulum in congenital dislocation of the hip: the shelf operation. J Bone Joint Surg. 1935;17:48
25. Heyman CH. Long-term results following a bone-shelf operation for congenital and some other dislocations of the hip in children. J Bone Joint Surg Am. 1963;45:1113
26. Kruse MRW, Guille JT, Bowen JR. Shelf arthroplasty in patients who have Legg-Calve-Perthes disease. J Bone Joint Surg Am. 1991;73:1338
27. Love BRT, Stevens PM, Williams PF. A long-term review of shelf arthroplasty. J Bone Joint Surg Br. 1980;64:321
28. Oh CW, Rodriguez A, Guille JT, et al. Labral support shelf arthroplasty for the early stages of severe Legg-Calvé-Perthes disease. Am J Orthop. 2010;39:26
29. Willett K, Hudson I, Cattarall A. Lateral shelf acetabuloplasty: an operation for older children with Perthes' disease. J Pediatr Othop. 1991;12:563
30. van de Heyden AM, van Tongerloo RB. Shelf operation in Perthes disease. J Bone Joint Surg Br. 1981;63:282
31. Stulberg SD, Cooperman DR, Wallensten R. The natural history of Legg-Calve-Perthes disease. J Bone Joint Surg Am. 1981;63:1095
32. Staheli LT. Slotted acetabular augmentation. J Pediatr Orthop. 1981;1:321
33. Waldenstrom H. The first stages of coxa plana. J Bone Joint Surg. 1938;20:559
34. Herring JA, Neustadt JB, Williams JJ, et al. The lateral pillar classification of Legg-Calve-Perthes disease. J Pediatr Orthop. 1992;12:143
35. Herring JA, Kim HT, Browne R. Legg-Calve-Perthes' disease: classification of radiographs with use of the modified lateral pillar and Stulberg classifications. J Bone Joint Surg Am. 2004;86:2103
36. Mose K, Hjorth J, Ulfeldt M, et al. Legg-Calve-Perthes disease: the late occurrence of coxarthrosis. Acta Orthop Scand. 1977;169(Suppl):1
37. Mose K. Methods of measuring in Legg-Calve-Perthes with special regard to the prognosis. Clin Orthop Relat Res. 1980;150:103
38. Grzegorzewski A, Synder M, Koztowski P, et al. The role of the acetabulum in Perthes disease. J Pediatr Orthop Am. 2006;26:316
39. McAndrew MP, Weinstein SL. A long-term follow-up of Legg-Calve-Perthes disease. J Bone Joint Surg Am. 1984;66:860
40. Bowen JR, Foster BK, Wein BK, et al. Legg-Calve-Perthes disease in patients under six years of age. Orthop Trans. 1981;5:446
41. Guille JT, Lipton GE, Szxke G, et al. Legg-Calve-Perthes disease in girls: a comparison of the results with those seen in boys. J Bone and Joint Surg. 1998;80A:1256
42. Grzegorzewski A, Snyder M, Kozlowski P, et al. Leg length discrepancy in Legg-Calve-Perthes disease. J Pediatr Orthop. 2005;25:206
43. Quain S, Catterall A. Hinge abduction of the hip: diagnosis and treatment. J Bone Joint Surg Br. 1986;68:61

shelf arthroplasty; Legg-Calve-Perthes disease, hip

© 2011 Lippincott Williams & Wilkins, Inc.