Legg-Calve-Perthes Disease

Key Points:

  • Legg-Calve-Perthes disease is a childhood femoral head ischemia of unknown etiology that may lead to permanent hip deformity.
  • Presence of synchronous, relatively symmetric bilateral head involvement should prompt the search for other etiologies.
  • The Waldenstrom radiographic classification stages the disease and can be used to guide timing of treatment as early treatment leads to best results.
  • The Catterall and Lateral Pillar radiographic classifications estimate severity of involvement and offer general guidelines for prognosis, with better outcomes typically occurring in those with less severe involvement.
  • Treatment is aimed at maintaining a spherical femoral head within a reduced and concentric joint.
  • The most common treatment is containment which can be achieved non-operatively through casting and bracing or operatively through femoral or pelvic osteotomy.

Description:

Originally independently described by Waldenstrom who thought this to be “tuberculosis of the hip”, this disorder was better defined in 1910 by the three researchers for whom it is now named, Legg-Calve-Perthes Disease (LCPD). This is a childhood ischemic disease of the femoral head that can lead to a variable degree of permanent femoral head deformity.  The disease follows distinct radiographic stages described by Waldenström: 1) initial - with ischemia-induced sclerosis; 2) fragmentation - when absorption, fibrovascular invasion, physeal changes, and cartilaginous metaplasia lend a lytic appearance to the epiphysis and metaphysis, with variable collapse; 3) re-ossification - when the fibrovascular and cartilage islands are replaced by bone (Waldenström, 1922; Catterall, 1982); and 4) healed/residual - when woven bone is replaced by normal bone.  On average, the stages last five, nine, and 19 months, respectively (Perry, 2011).

Epidemiology:

LCPD incidence ranges from three to 15 cases per year per 100,000 children age 0-14 years old (Joseph, 2003) .  While peak age of first presentation in northern countries occurs around five years old, it can average as late as 10 years old in more equatorial countries such as India (Joseph,1988; Wiig, 2006; Perry, 2011).  Prevalence in several studies varies according to socioeconomic status, geography, race, and education level, with increased numbers seen in those who are less affluent, live in cities or at higher latitude, are Caucasian, or have less educated parents (Joseph, 1988; Wiig, 2006; Perry, 2011; Johansson, 2006; Perry, 2012).  However, health care access can confound findings.  No matter the social variables, prevalence in males outnumber females at a rate of 3-5:1.  There is some association with history of smoke exposure. 

Clinical Findings:

The onset is most frequently between 4 to 7 years of age but can occur as late as 12 years old. The child usually presents with hip pain, although thigh pain and referred knee pain can also be the presenting complaint. This is often accompanied by a limp. Typically, symptoms wax and wane, worsening with activity and improving with rest and non-steroidal anti-inflammatories (NSAIDs).  Physical examination in the early stages reveals a hip flexion contracture (Thomas’ test) and limitation of abduction and internal rotation. As the disorder progresses, an adduction contracture may develop. An antalgic gait is frequently seen with a shortened stance phase with early weight shifting off on the affected side. With more severe involvement, deformed hip joints may develop symptoms of impingement and restricted range of motion, and go on to develop acetabular chondral lesions, femoral head osteochondral defects, and labral tears.  Eventually, severe deformities may lead to early arthritis.

Imaging Studies:

Clinical severity, prognosis, and outcome are largely based on radiologic classification systems that employ standard AP and frog-leg lateral pelvis radiographs. 
 
The Salter-Thompson classification is a radiographic classification based on the extent of the subchondral fracture that occurs in the late initial stage. Class A hips (<50% femoral head involved in the fracture) having a better prognosis than class B (>50%).  

The Catterall classification is based on the amount of femoral head involved during fragmentation (Catterall,1971):
            I – anterior epiphysis only area involved
            II – anterior and central epiphysis involved
            III – everything but the most medial and lateral epiphysis involved
            IV – entire epiphysis involved.
The Catterall classification additionally includes radiographic signs that place the femoral head at particular risk for worsened outcome: lateral head subluxation, Gage sign (lateral V-shaped lucency), a horizontal physis, and presence of calcification lateral to the epiphysis.  

A third prognostic system is the lateral pillar classification, which is based on the condition of the lateral 1/3 of the femoral head during the late fragmentation stage (Herring 1992, Herring 2004):
            A – no loss of lateral pillar height
            B – lateral pillar with <50% loss of height
            B/C – thin pillar, loss of lateral pillar height at 50%
    C – >50% collapse of lateral pillar.
Outcomes for inter- and intra-observer reliability between these three classifications vary depending on disease stage timing and level of expertise of reviewers (Park, 2012; Sambandam, 2006).

The modified Elizabethtown classification is a modification of the Waldenstrom classification and is useful in determining the stage of the disease (Joseph, 2003). 
Stage Ia - Sclerosis of the epiphysis with no loss of height
Stage Ib - Sclerosis of the epiphysis with loss of height but no fragmentation
Stage IIa - Early fragmentation with just one or two vertical fissures in the epiphysis on the AP or frog leg lateral view
Stage IIb - Advanced fragmentation with no new bone lateral to the fragmented epiphysis
Stage IIIa - early “porotic” new bone formation at the periphery of the epiphysis covering less than a third of the epiphysis
Stage IIIb - New bone formation of “normal” texture and covers more than a third of the epiphysis.
Stage IV - Complete healing with no radiographically identifiable avascular bone. 

Prognosis is judged by the risk of long-term hip arthritis, which is correlated to the Stulberg classification for radiographic morphology of the residual LCPD hip (Stulberg, 1981)16:    
      I – normal shaped femoral head and acetabulum
II – spherical femoral head, but with at least one type of deformation including coxa magna, shortened neck, or steep acetabulum
III – ovoid femoral head with congruently ovoid acetabulum and neck
IV – flat femoral head with congruently flat acetabulum and neck
V – flat femoral head with normal acetabulum and neck; incongruently incongruous.
The Stulberg classes correlate with a lifetime risk of 0%, 16%, 58%, 75%, and 78% for arthritis, respectively.

Recently, perfusion MRI has been described as a useful study to determine the extent of loss of perfusion to the femoral epiphysis.  The percent perfusion is correlated to radiographic lateral pillar such that in patients with less than 40% perfusion, the odds ratio of developing into lateral pillar C involvement was 23.8 (Kim, 2014).  This study may be used early in the disease process before deformity occurs to assist in the decision on whether or not to pursue surgical treatment.

Etiology:

The cause of LCPD is not completely understood, but several factors likely play a role.  Population studies demonstrate that the ischemia could be caused by various coagulopathies (Stulberg,1981) which may decrease arterial supply or slow venous drainage, which is seen in early stages of disease.  There may also be a hereditary aspect, as small familial subsets have been found to have type-2 collagen gene defects.  Affected children often share a set of qualities, including younger bone age, shorter stature, relatively shorter distal limb segments (Vosmaer, 2010), and attention deficit hyperactivity disorder (ADHD) (Burwell, 1988), that suggest a correlation to a more global process.

Treatment:

The main goal in LCPD treatment is to achieve a spherical femoral head and concentric joint at the conclusion of the disease process, thereby remaining within Stulberg class I and II.  The type of treatment recommended is based on patient’s age, stage of disease, and severity of involvement (radiographic class).  Standard treatment options include symptomatic treatment and containment treatment.  Symptomatic treatment typically includes activity modifications, protected weightbearing, and physical therapy focused on maintaining hip range of motion, specifically abduction.  Such treatment is typically considered for patients with less severe involvement (lateral pillar A or B, without extrusion, with onset younger than 8[l1] years old, or in patients with lateral pillar A or Catterall class I or II, without extrusion, with onset up to 11 years old).  There have been no demonstrated changes in outcomes based on symptomatic treatment (Loder, 2003; Herring, 2004).  

Containment treatment, which means to position the anterior-lateral portion of the femoral epiphysis within the acetabulum in order to prevent deformation, can be performed through either operative or non-operative means (Joseph, 2005).  Non-operative containment can be achieved through Petrie casting and abduction bracing.  Rich et al reported that treatment of hips with maintenance of range of motion and use of an A-frame orthosis resulted in congruency in 93% of hips (Rich, 2013). Surgical containment can be achieved through either varus-producing intertrochanteric femoral osteotomy or pelvic osteotomy.  The most common pelvic osteotomies performed in containment surgery are a shelf osteotomy or a triple osteotomy (Kruse, 1991; Mosow, 2017).  

The type of treatment recommended is based on patient’s age, stage of disease, and radiographic class.  Patients with onset after 11 years or lateral pillar class C have been considered to have a poor prognosis no matter the treatment type.  Onset prior to six years old has been considered a universally good prognosis factor.  However, evidence shows that radiographic stage is a stronger determinant of prognosis than age, with poor outcomes still occurring at this young age in Catterall class III/IV and lateral pillar class C hips (Herring, 1992; Loder, 2003; Kruse, 1991; Canavese, 2008; Rosenfeld, 2007). The classic studies by Herring et al suggested that surgical containment resulted in best outcomes in children over 8 years old with Catterall class III or IV, and lateral pillar class B or B/C border hips (Herring 2004).  However, recently it has been appreciated that containment treatment is most effective when initiated in the early stages of disease (prior to when the lateral pillar classification can be made in late fragmentation) or prior to hip extrusion regardless of age (Joseph, 2011). 

Valgus intertrochanteric osteotomy has also been described as a treatment option in LCPD.  This is typically undertaken in the later stages of disease after there has been deformation of the epiphysis resulting in hinge abduction.  Valgus osteotomy results in rotation of the flattened lateral portion of the head away from the acetabulum and allows for a greater range of impingement free hip abduction and improved articular congruency (Gheldere, 2011).    

Patients with residual deformity after healed Perthes disease may have anatomic deformities including femoral-acetabular impingement (FAI) and acetabular dysplasia. These may lead to chondral defects on both sides of the joint and labral tears.  Treatment is aimed at repairing or minimizing these residual deformities.  Femoroacetabular impingement and labral tears may be treated by either arthroscopy or surgical hip dislocation with labral repair and femoral head-neck osteochondroplasty.  Surgical hip dislocation can also include relative femoral neck lengthening by greater trochanteric advancement in the case of trochanteric overgrowth.  Acetabular dysplasia may be treated with a periacetabular osteotomy[l2].  These residual phase surgical options have been shown to improve patient outcomes via the Harris Hip Score (Schoenecker 1993, Clohisy 2015).  When these treatment options fail, total hip replacement may be required.

Complications:

Complications are specific to treatment, but in general the most concerning complication is iatrogenic avascular necrosis which compounds the femoral head ischemia.  There is risk for this with femoral osteotomy and surgical hip dislocations by direct injury to the blood supply via the medial femoral circumflex artery.  Excessive femoral head pressure from an innominate osteotomy applied to an incongruent joint may also lead to further impairment of femoral head blood flow.  Progressive ischemic collapse of the femoral head may also occur in the early stages, particularly with large subchondral fractures, from inappropriate activity and weight-bearing guidelines. 

References:

  1. Anderson LA, Erickson JA, Severson EP, et al.  Sequelae of Perthes disease: treatment with surgical hip dislocation and relative femoral neck lengthening.  J Pediatr Orthop.  2010; 30(8): 758-66.
  2. Burwell RG.  Perthes’ disease: growth and aetiology.  Arch Dis Child.  1988; 63(11): 1408-12.
  3. Canavese F, Dimeglio A.  Perthes’ disease: prognosis in children under six years of age.  J Bone Joint Surg Br.  2008; 90(7): 940-5.
  4. Catterall A.  The natural history of Perthes’ disease.  J Bone Joint Surg Br.  1971; 53(1): 37-53.
  5. Catterall A, Pringle J, Byers PD, et al.  A review of the morphology of Perthes’ disease.  J Bone Joint Surg Br.  1982; 64: 269-
  6. Clohisy JC, Nepple JJ, Ross JR, et al.  Does surgical hip dislocation and periacetabular osteotomy improve pain in patients with Perthes-like deformities and acetabular dysplasia?  Clin Orthop Relat Res.  2015; 473(4): 1370-7.
  7. de Gheldere A, Eastwood DM. Valgus Osteotomy for Hinge Abduction. Orthopedic Clinics of North America. 2011;42:349-354.
  8. Herring JA, Kim HT, Browne R.  Legg-Calve-Perthes disease.  Part I: classification of radiographs with the use of the modified lateral pillar and Stulberg classifications.  J Bone Joint Surg Am.  2004; 86: 2103-20.
  9. Herring JA, Kim HT, Browne R.  Legg-Calve-Perthes disease. Part II: Prospective multicenter study of the effect of treatment on outcome.  J Bone Joint Surg Am.  2004; 86-A (10): 2121-34.
  10. Herring JA, Neustadt JB, Williams JJ, Early JS, Brown RH.  The lateral pillar classification of Legg-Calve-Perthes disease.  J Pediatr Orthop.  1992; 12: 143-50.
  11. Johansson T, Linblad M, Bladh M, et al.  Incidence of Perthes’ disease in children born between 1973 and 1993.  Acta Orthop.  2006; 2: 1-5.
  12. Joseph B, Chacko V, Rao BS, et al. The epidemiology of Perthes’ disease in south India. Int J Epidemiol. 1988;17(3): 603–607.
  13. Joseph B, Price CT. Principles of Containment Treatment Aimed at Preventing Femoral Head Deformation in Perthes Disease. Orthopedic Clinics of North America. 2011;42:317-327.
  14. Joseph B, Rao N, Mulpuri K, et al.  How does a femoral varus osteotomy alter the natural evolution of Perthes’ disease?. J Pediatr Orthop B. 14:10
  15. Joseph B, Varghese G, Mulpuri K, Rao NKL, and Nair NS. Natural Evolution of Perthes Disease: A Study of 610 Children Under 12 Years of Age at Disease Onset. Journal of Pediatric Orthopaedics. 2003.23:590–600.
  16. Kim HK, Wiesman KD, Kulkarni V, et al. Perfusion MRI in Early Stage of Legg-Calvé-Perthes Disease to Predict Lateral Pillar Involvement: A Preliminary Study. The Journal of Bone and Joint Surgery-American Volume. 2014;96:1152-1160.
  17. Kruse RW, Guille JT, Bowen JR.  Shelf arthroplasty in patients who have Legg-Calve-Perthes disease.  A study of long-term results.  J Bone Joint Surg Am.  1991; 73(9): 1338-47.
  18. Loder RT, Schwartz EM, Hensinger RN.  Behavioral characteristics of children with Legg-Calve-Perthes disease.  J Pediatr Orthop.  2003; 13(5): 598-601.
  19. Park MS, Chung CY, Lee KM, et al.  Reliability and stability of three common classifications for Legg-Calve-Perthes disease.  Clin Orthop Relat Res.  2012; 470(9): 2376-82.
  20. Perry DC, Hall AJ.  The epidemiology and etiology of Perthes disease.  Orthop Clin North Am.  2011; 42(3): 279-83.
  21. Perry DC, Machin DM, Pope D, et al.: Racial and geographic factors in the incidence of Legg-Calvé-Perthes’ disease: a systematic review. Am J Epidemiol. 175:159 2012
  22. Rich MM, Schoenecker PL. Management of Legg-Calvé-Perthes Disease Using an A-Frame Orthosis and Hip Range of Motion: A 25-Year Experience. Journal of Pediatric Orthopaedics. 2013;33:112-119.
  23. Rosenfeld SB, Herring JA, Chao JC.  Legg-calve-perthes disease: a review of cases with onset before six years of age.  J Bone Joint Surg Am.  2007; 89(12): 2712-22.
  24. Sambandam S N, Gul A, Shankar R, et al.  Reliability of radiological classifications used in Legg-Calve-Perthes disease.  J Pediatr Orthop B.  2006; 15(4): 267-70.
  25. Schoenecker PL, Stone JW, Capelli AM.  Legg-Perthes disease in children under 6 years old.  Orthop Rev.  1993; 22(2): 201-8.
  26. Stulberg SD, Cooperman DR, Wallensten R.  The natural history of Legg-Calve-Perthes disease.  J Bone Joint Surg Am.  1981; 63: 1095-Waldenström H. The definitive forms of coxa plana. Acta Radiol. 1922; 1:384–394. 
  27. Vosmaer A, Pereira RR, Koenderman JS, et al.  Coagulation abnormalities in Legg-Calve-Perthes disease.  J Bone Joint Surg Am.  2010 Jan; 92(1): 121-8.
  28. Wiig O, Terjesen T, Svenningsen S, et al. The epidemiology and aetiology of Perthes’ disease in Norway: a nationwide study of 425 patients. J Bone Joint Surg Br. 88:1217 2006.
  29. Wiig O, Terjesen T, Svenningsen S.  Prognostic factors and outcome of treatment in Perthes’ disease: a prospective study of 368 patients with five-year follow-up.  J Bone Joint Surg Br.  2008; 90(10): 1364-71.

Top Contributors:

Jennifer M. Bauer, MD
Mihir M. Thacker, MD