Legg Calve Perthes disease (LCPD) is a self-limiting condition caused by temporary interruption of blood supply to the growing proximal femoral epiphysis leading to necrosis, collapse and revascularization. Though the cause of vascular occlusion is not yet known, it leads to morphological and developmental changes in the proximal femoral head, femoral neck and acetabulum. The resultant irreversible deformation of femoral head, growth retardation, shortening of femoral neck, coxa vara and joint subluxation results in articular incongruity, altered joint mechanics predisposing the hip to premature osteoarthritis.

History

1905-         Kohler described radiographic changes of a patient similar to LCPD

1909-         Henning Waldenström described it as a benign form of tuberculosis.

1909-         Independently described by Arthur Thornton Legg from USA, Jacques Calve from France and George Perthes from Germany. 

1921-         Phemister described histologic findings suggestive of osteonecrosis in the specimens of a patient treated by curettage. He described creeping substitution during reconstitution phase.

1926-         Konjetzny first showed interruption of vascular supply.

1952-         Varus containment osteotomy described by Soeure of Belgium.

1962-         Salter described innominate osteotomy.

Epidemiology

Reported incidence range from 0.2 to 19.8 per 100000 population per year. Age of presentation of active disease range from 2 years to 11 years, but most commonly occur between 4-8 years of age. Boys are 4-5 times more commonly affected than girls. 10-15% of cases have involvement of opposite hip either simultaneously or at a later age. Boys 5 times more commonly affected than girls. Bilateral involvement is more common in girls.

Aetiology

Vascular occlusion leading to osteonecrosis in the growing epiphysis is the pathophysiologic pathway of LCPD. LCPD results from repeated episodes of vascular occlusion and hypercoagulable states may play a role in the pathogenesis. Vascular occlusion may be extrinsic or intraluminal. Vascular occlusion may be arterial or venous. Veins of proximal femur are susceptible to occlusion because of very thin walls, low flow rates, tortuous course around arteries and low pressure. Venous occlusion can lead to increased intraosseous pressure, reduced arterial blood flow, hypoxia and ischemic necrosis.

The exact cause of occlusion is not yet identified. The proposed mechanisms include trauma, genetic mutations, hypercoagulable states and transient synovitis. 

Higher incidence of factor V Leiden mutation, protein S deficiency, elevated factor VIII, and prothrombin G20210A mutation are reported especially in males. These conditions may result in thrombophilia or hypofibrinolysis. Multifactorial origin acting through a single shared common pathway is the currently favored model. Recently a single missense mutation of COL2A1 gene for type II collagen leading to replacement of glycine with serine at the codon 1170 have been reported in Asian families with multiple affected members.

A recent study reported coagulation anomalies in 75% of cases. The anomalies were protein C deficiency, protein S deficiency, elevated lipoprotein A and hypofibrinolysis. Other studies failed to demonstrate a similar finding.

Pathogenesis

Vascular occlusion leads to ischaemic necrosis. Ischaemic necrosis has biological and mechanical effects. Biologic consequence is a period of softness of head making it vulnerable for deformation and cessation of epiphyseal growth. Growth arrest can lead to femoral neck shortening and coxa vara. Greater trochanteric overgrowth of varying severity is seen in over 90% of patients with LCPD which can lead to Trendelenburg limp. Mechanical weakening leads to collapse due to repeated mechanical stresses. Mechanical effects are due to lose of sphericity of femoral head.

Biological effects are due to vascular compromise leading to cell death, followed by vascular ingrowth and bone resorption, finally ending with new bone deposition. Mechanical effects are due to softening, subchondral fracture, collapse, loss of sphericity, lateral extrusion, superolateral hinging, labral tear, stress concentration, articular cartilage degeneration and osteoarthritis. 

Ischaemic damage is followed by invasion of dead trabeculae by neovascularization, removal of dead trabeculae by osteoclasts followed by reossification by creeping substitution or repair by fibrovascular tissue which leads to weakening of femoral Head. Weakened femoral head collapses leading to flattening and deformation of head. Superolateral part of deformed head gets extruded outside the confines of acetabulum. Progressive deformation leads to lateral subluxation, articular incongruity and impingement. Altered joint forces leads to early arthritis in adulthood. 

The disease process evolves through the stages of avascular necrosis, fragmentation, reconstitution and healing. Extrusion of femoral head is the most important factor in deformation of femoral head. It occurs during the latter part of fragmentation stage. The final outcome is coxa magna (enlarged head and neck), coxa breva (shortened neck) and coxa plana (flattened head) with femoroacetabular impingement.

Associations

LCPD is associated with delayed skeletal age and low birth weight. Cranial structures are relatively spared from growth retardation than caudal structures with the foot showing greatest growth retardation. Low socioeconomic status is a strong association pointing towards environmental factors. Low concordance in identical twins is an evidence against genetic factors. Inguinal hernia, genitourinary malformations, undescended testes are reported associations. 

Clinical Features

Patients may present in childhood, adolescence or in adulthood. Usually asymptomatic initially till collapse or subchondral fracture develops. Child presents with painless limp and later complains groin pain or knee pain. Groin pain and lateral hip pain are the most common presentations. Lateral pain may be due to abductor insufficiency, trochanteric bursitis or trochanteric impingement. Groin pain may be due to impingement, instability or arthritic changes. Posterior pain is usually due to impingement.  

Gait may be antalgic or Trendelenburg type. Lumbar lordosis may be exaggerated due to flexion deformity. Anterior superior iliac spine may be at a higher level in presence of adduction deformity. There may be wasting of thigh and gluteal muscles. Tenderness over the anterior and posterior joint line may be present. Greater trochanter is usually thickened and elevated. Prominence of greater trochanter is due to lateral subluxation, muscle wasting, coxa vara and relative overgrowth of greater trochanter. Limitation of range of movements especially abduction and internal rotation seen. Fixed flexion deformity is seen. In the normal hip, rotations are more in 90 degree flexion than extension, called differential limitation of rotations. In LCPD, rotations are more limited in flexion than extension. Those children on skin traction may show near normal range of movements.

In adolescence, the hip is in healed stage and the physical findings depends on the amount of incongruity, impingement and femoral neck shortening. Residual deformity can cause intra-articular and extra-articular impingement. Many patients present with pain due to impingement. Shortening of variable degree is present. Adults present with pain and limitation of movement due to early onset osteoarthritis. Limitation of movements affects abduction and internal rotation, but other movements are affected as per severity of degenerative changes. External rotation deformity is seen in those with advanced degenerative changes. Flexion deformity and adduction deformity also is seen in those with severe incongruity or osteoarthritic changes.

Imaging

Diagnosis is usually made with the conventional x-rays, but radiological changes usually appears late. Hence recent trend is to rely on advanced imaging modalities such as perfusion MRI using gadolinium.

A correctly positioned AP view, false profile view of acetabulum and frog leg lateral view (Lowenstein view) should be taken in all patients suspected to have LCPD. AP view is taken in 15 degrees of internal rotation with the beam centred over a point midway between the superior margin of pubic symphysis and a line connecting both ASIS. If properly positioned the ilium, tear drop, obturator foramen should be symmetrical, the tip of coccyx should be in the midline in line with pubic symphysis and the distance from tip of coccyx to symphysis pubis should be within 2cm. 

The AP radiograph is used to measure the following radiographic parameters: lateral center edge angle, acetabular index, Shenton’s line and femoral head extrusion index. Look for the head at risk signs. Initial x-rays may be normal. Radiographic changes appear about 6 months after the first infarction has occurred. Earliest radiologic sign is a relatively smaller size of ossific nucleus of the involved head of femur.

Widening of medial joint space is found early seen (Waldenström sign). Widening of joint space due relative smaller size of ossific nucleus, lateral subluxation and thickening of articular cartilage.

Later the ossific nucleus becomes progressively sclerotic. Subchondral fracture may be seen. Subchondral fracture is called Caffey’s sign. Fragmentation and collapse leads to progressive deformation of femoral head. Lateral extrusion of head occurs with flattening of epiphysis. Shenton’s line shows superolateral migration of femoral head. Shenton’s line is considered to be broken if there is a step off of more than 5mm. Although changes are more noticeable in the proximal femur, acetabular changes are frequently present. Osteoporosis of acetabular roof, alterations in contour and change in dimensions is often seen. Acetabulum may show bicompartmentalization, ischium varum and early closure of triradiate cartilage. Acetabulum becomes shallow and misshapen. 

False profile view is taken with the patient standing and the body tilted 65 degrees to the x-ray beam, with the uninvolved side forward. On the false profile view of acetabulum measure the vertical centre edge angle. 

Catterall has described several radiographic findings as head-at-risk signs as indicators of poor prognosis. These are lateral extrusion, calcification lateral to the epiphysis, poorly ossified lateral part of epiphysis (Gage sign- Described by Courtney Gage in 1933), diffuse metaphyseal reaction (described by Smith in 1982) and horizontal growth plate. 

Measure the Heyman Herdon acetabular head index, Dickens and Menelaus femoral head extrusion index and Reimer migration index.

If containment surgery is planned, x-rays are repeated in hip abduction and in abduction-internal rotation to assess containment. Arthrography through an anterolateral or subadductorl access is an important adjunct to preoperative assessment as it visualizes the contour of articular cartilage and allows dynamic assessment of hip joint congruity. On the arthrogram assess the presence or absence of impingement, amount of subluxation manifested by medial clear space, presence or absence of hinged abduction and containment of femoral head within the confinements of acetabulum.

Bone scan findings precedes radiographic signs by 3 months. It is rarely used due to risk of radiation. CT is almost never used in children, but can be of use in adults to study impingement or to assess adequacy of columns before total hip replacement. 

Conventional radiography and x-ray based classifications have the limitation of inability to visualize the shape and deformation of cartilage. In addition, radiographic changes are present only in later stages of the evolution of LCPD. MRI has the advantage of visualization of cartilage and allows early detection of osteonecrosis. It is useful in detection of full extent and exact location of osteonecrosis, visualization of physis, chondral pathology and labral lesions. Epiphyseal changes are not associated with growth arrest in 76% of cases, but physeal and metaphyseal changes are associated with growth arrest.

MRI avoids ionizing radiation and is noninvasive. However, unenhanced MRI may fail to show changes in early stages of disease. Perfusion of the head can be studied using dynamic gadolinium enhanced subtraction MRI. MR perfusion index has been described as a measure of epiphyseal perfusion using digital image analysis of gadolinium enhanced subtraction MR images. It has been shown to correlate with femoral head deformity measured using conventional follow up x-rays. Necrosis extension, lateral extrusion, and the extent of physeal and metaphyseal involvement on MRI are important predictors of outcome. Horizontalization of labrum can also be identified on MRI which is indicative of significant femoral head deformation. Metaphyseal changes have been shown to be important predictors of physeal involvement and prognosis.

Classification

Classification can describe the extent and severity of disease at the time of presentation or its outcome. Classification help in predicting prognosis, deciding on treatment, comparison of treatment.

Waldenström Staging (1938)

Necrosis stage– Small sclerotic head with increased medial joint space.

Fragmentation stage– Epiphysis sclerotic, fragmented and collapsed.

Re-ossification stage– Reossification proceeding from lateral to medial and posterior to anterior

Healed stage– Density returned to normal with residual changes in shape and size of head.

Elizabethtown Classification

4 stages 

1. Sclerotic 
2. 1. Without loss of height
3. 1. With loss of height
4. Fragmentation
5. 1. Early
6. 1. Late
7. Healing
8. 1. Peripheral – <1/3^rd>1/3rd
9. Healed

Catterall Classification (1971)

Group 1– <25% involvement. Anterior epiphysis only involved. No collapse, no metaphyseal involvement. Usually revascularised.

Group 2– Up to 50% involved. Collapse present. Central segment fragmentation and collapse. Necrotic portion appears sclerotic. Epiphyseal height maintained. Necrotic portion separated from viable portion on the lateral view in a characteristic ‘V’.

Group 3– >50% but not total involvement. Anterior, central and lateral involvement. Posterior part of head remains viable. Head-within-head appearance. Collapse present. Extensive metaphyseal changes with broadening of neck.

Group 4– Whole epiphysis affected. Collapse present. Mushroom shaped head. Extensive metaphyseal changes.

Described 4 “head-at-risk signs”

         1, Lateral subluxation

         2, Calcification lateral to the epiphysis

         3, Gage sign- Inverted V shaped defect laterally)

         4, Horizontal growth plate

         5, Diffuse metaphyseal reaction described by Smith in 1982

Salter Thomson Classification (1984)

It can be applied only in presence of subchondral fracture. Extent of subchondral fracture correlates with extent of subsequent collapse.

Group A– Subchondral fracture extent <50% of epiphysis.

Group B– Extent of subchondral fracture >50% of head.

Herring Lateral Pillar Classification (1992)

Done in the fragmentation stage. Epiphysis divided into lateral, central and medial pillars on the true AP view. Lateral pillar is the lateral 5-30% of epiphysis depending on the location of lucent line that separates it from the central necrotic area. If lucent line is absent, take the lateral 25% as lateral pillar. Maximum reduction in the height of lateral pillar measured in relation to normal side.  It is difficult to apply in bilateral cases and in the very young.

Group A– lateral pillar height fully maintained

Group B– Lateral pillar height >50% of normal. Good outcome if age is <9 years.

Group B/C– Lateral pillar height 50%, but poorly ossified. Added in 2004.

                            B/C1– Only 2-3mm width

                            B/C2– Minimum ossification

                            B/C3– Lateral pillar more depressed that central pillar

Group C– Lateral pillar height <50% of normal

All group A have good result. 2/3 of group B have good results. Only 25% of group B/C have good result. Only 1 in 8 with group C have a good result. 

Difficult to classify very young children. Needs about 7 months for proper classification. About 30% needs upgrading and only 4% remain in group A on subsequent follow up. Difficult to classify in bilateral cases.

Mose Classification

Done to assess outcome. Done on x-rays taken after 16 years of age. Classified into 3 types by placing Mose template with concentric circles at 2mm increments. Assess the sphericity of head.

Spherical

Nonspherical

Spherical but crescent shaped

Fasting scintigraphy classification(1980)

Grade 1– Decreased activity in <25% of head

Grade 2– Decreased activity in 50% of head

Grade 3– Decreased activity in 75% of head

Group 4– Decreased activity in whole of head

Conway Scintigraphy Classification (1992)

2 tracks. Track A of recanalization and Track B of neovascularization.

4 stages in each track.

Track A- 

         Whole head

         Lateral column

         Anterior and medial extension

         Complete

Track B

         Whole head

         Base filling

         Mushrooming

         Complete

Stulberg Classification 1982

Head shape divided into spherical (I or II), ovoid (III) or flat (IV or V). Subdivided according to coxa magna, steep acetabulum, short neck, 

I-      Normal

II-      Spherical – Within the circle by <2mm in both AP and lateral views. Coxa magna, short neck, acetabulum steep.

III-    Ovoid head. Out of shape by >2mm on either AP or lateral views

IV-    Aspherical congruency– Aspherical. Congruent.

V-      Aspherical incongruency– >1cm flattening on the superolateral weight bearing area. Incongruent.

I and II are spherical congruency, III and IV are aspherical congruency and V is aspherical incongruency.

Laredo Arthrographic Classification

Nature and extent of femoral head deformation and the severity of lateral extrusion are the most important factors that determine the prognosis. Radiographs show only the ossific portion proximal femoral epiphysis and may not represent the anatomical reality of the femoral head acetabular congruence. The understanding of the status of the cartilaginous portion is important especially in presence of persistent restriction of motion. However, it is an invasive procedure and difficult to repeat. Horizontalization of labrum is an important finding suggestive of deformation of head and it can be quantified by measuring the labral angle.

Group I – Normal hip

Group II

Femoral head larger than normal but spherical

Extrusion present at the neutral position and absent at 30 degrees’ abduction and slight internal rotation.

Group III

Femoral head larger than normal and ovoid

Extrusion present at neutral position and in 30 degree abduction and slight internal rotation.

Group IV 

Femoral head larger than normal and flattened

Extrusion is present at 30 degrees’ abduction and slight internal rotation 

Labrum loses its concavity and becomes elevated and straightened Hinged abduction present.

Group V

Femoral head larger than normal and saddle shaped

Extrusion is present at neutral position and in 30 degrees’ abduction and slight internal rotation.

Labrum is elevated and sometimes everted, with abnormal pooling of contrast medium at the saddle deformity area. 

Sphericity Deviation Score

Mark the medial and lateral edge of primal femoral growth plate on the AP and lateral views. Draw the maximum inscribed circle (MIC) touching these points without extending outside the femoral head. Mark the centre of the circle. Draw a second concentric circle as the minimum circumscribed circle (MCC) without extending inside the femoral head. Measure the difference between the radii of MIC and MCC (roundness error) on both AP and lateral views. Measure the difference between radii of femoral head on the AP and lateral views (ellipsoid deformation). The sum of roundness error on the AP and lateral views and the ellipsoid deformation is the Sphericity Deviation Score. SDS of normal hip is 0-3.8. If SDS at healing is less than 10, then the chance of hip having Stulberg I or II is high. If SDS is more than 20, then the hip is likely to have higher Stulberg grades at maturity.

Prognosis

60-80% of patients with LCPD have good outcome but it deteriorates on long term follow up. At a mean age of 45 years, 86% percentage of hips were found to be functioning and only 8% were found to have arthroplasty. But another report published later on the same group of patients at an average age of 53 years found that 40% of hips were replaced, 40% had functioning hips and 20% had osteoarthritic symptoms.  

Long term sequelae of LCPD are due to nonspherical head, short and broad neck, greater trochanteric overgrowth and labro-acetabular changes. Severity of these changes depends on the age of onset, gender, extent of femoral head necrosis, severity of femoral head deformation and method of treatment.

Age at onset between 5-7 years do better than those with age at onset more than 8-9 years. Catterall group III and IV patients with age at onset younger than 6 years may have poor outcome. In the multicenter study by Legg Perthes study group 59% of those aged less than 8 years at onset had a Stulberg II outcome while only 39% of those aged more than 8 years had a Stulberg II outcome. According to a study by Wiig from Norway 59% of those aged less than 6 years had a Stulberg I or II outcome and 38% of those above 6 years had a similar outcome.  

Age at the time of healing considered to be more important than age at onset. Age at follow up is also important with 86% having osteoarthritis at 65 years of age. About 50% of patients will ultimately need a total hip arthroplasty.

Lateral pillar classification also is helpful in prognosis determination. Stulberg I to II outcome found in 70-100% of lateral pillar A hips, 51-62% of lateral pillar B hips, 28% in B/C borderline hips and 13-30% of C hips.

Caterall grading also is significant in prognosis. 84% of Catterall 1 and 2 have a Stulberg I or II outcome and only 44% of Catterall 3 or 4 have a Stulberg I or II outcome.

Girls have poor outcome when compared to boys of same age due to more advanced skeletal age. Outcome is poor especially in girls aged more than 8 years. Overweight, longer duration for healing, persistent restriction of motion found to be associated with poor outcome.

Radiologically the following are important; extent of epiphyseal involvement, extent of metaphyseal changes, extent of lateral extrusion >20% and height of lateral pillar <50%. 

Method of treatment has a bearing on outcome in those aged more than 8 years with lateral pillar B or B/C with 73% of lateral pillar B treated by surgery having Stulberg I or II outcome while only 44% of those treated nonoperatively have a similar outcome. Greater trochanteric overgrowth is more common in lateral pillar C (44%) than in type B and B/C groups.41

 3 points from the Perthes Study Group regarding prognosis

1, Children with onset at less than 8 years are likely to have better outcomes. Generally lower age at onset have a better prognosis, but only half of those with age less than 5 years at onset with lateral pillar C have a good outcome.

2, Children more than 8 years at onset, if at least 50% of lateral pillar height maintained have better results if treated by surgery.

3, Outcome in Type C hips with less than 50% lateral pillar height is generally poor irrespective of treatment method.

Treatment

”TO BE CONTINUED”