13.18 Legg–Calve–Perthes disease

The radiological changes of Legg–Calve–Perthes disease were identified soon after the development of x-ray machines for clinical use. In 1909, Waldenström was the first to publish a description of the radiological changes seen in the condition but his name is not credited to the disease because he considered the appearances to be a form of tuberculosis of the hip.

In 1910, Legg, Calve, and Perthes secured their place in history when each published observations distinguishing the condition from tuberculosis, reflecting a more benign clinical course and dissimilar radiological progression. Perthes original description (1913) is still valid today:

a self-limiting, non-inflammatory condition, affecting the capital femoral epiphysis with stages of degeneration and regeneration, leading to a restoration of the bone nucleus

Currently it is accepted that the series of changes described by Perthes are initiated by events that lead to avascular necrosis (AVN) of the ossific nucleus followed by gradual restoration of the blood supply and subsequent regeneration. The underlying aetiology of these events remains elusive and so the condition cannot be prevented. Management is focused on trying to limit progressive femoral head deformity after the AVN has occurred.

Perthes disease occurs worldwide. The incidence is relatively low in black populations (0.45/100 000), intermediate in Asians (3.8/100 000), and high in Caucasians (15.4/100 000). Within populations, the incidence varies from region to region and reports from the United Kingdom and South India suggest an association with deprivation. Most clinicians will be aware of patients with a family history of the disease but the evidence for a genetic predisposition is unclear. Some studies show little support for a genetic aetiology whilst others suggest a multifactorial inheritance pattern.

The condition is four to five times more common in boys than girls. Eighty per cent of children present between 4–9 years of age. Patients may share some typical characteristics and the notion of the ‘susceptible child’ has emerged (Box 13.18.1).

The blood supply of the child’s femoral epiphysis is vulnerable. No blood reaches the epiphysis from the metaphysis because the physeal plate presents a complete barrier and the blood supply from the ligamentum teres is negligible. The primary supply to the femoral epiphysis is derived from the medial femoral circumflex artery which pierces the capsule of the hip joint in the posterior trochanteric fossa and becomes the lateral ascending cervical artery (Figure 13.18.1).

Effectively, a single trunk vessel supplies the epiphysis which makes it vulnerable to ischaemic necrosis. The anteromedial part of the femoral epiphysis is furthest from the main trunk and this may explain why it is the first area to show evidence of AVN in Perthes disease and why the posterolateral portion, closer to the trunk vessel, is sometimes preserved.

The nature and cause of the disturbance in the blood supply in Perthes disease is uncertain. A single infarct of the femoral epiphysis does not necessarily lead to the radiological changes seen in the disease. Most reports to date support the notion that recurrent infarcts are necessary to cause the typical radiological features. Once AVN is established, progressive changes may occur. The current working hypothesis suggests that a fracture develops in the weakened avascular bone which is unable to unite until it is revascularized. As revascularization ensues, the repair process results in the characteristic radiological changes, as dead bone trabeculae are slowly removed and replaced. This process can take years to complete and during this time the cartilaginous femoral head, devoid of its internal supporting structure, can become deformed.

Why an infarct should occur is still unknown. Reports have implicated disturbances on both the arterial and the venous side of the femoral head circulation. Trauma has been postulated as a potential explanation as has transient synovitis: both may tamponade the circulation but although symptoms of synovitis are often the first presentation of the disease, most studies do not support a causal relationship between transient synovitis and Perthes disease. More recent reports implicated various abnormalities of the plasma proteins involved in the coagulation cascade. Some are associated with a predisposition to clot formation (thrombophilia) and others are associated with a limited potential to ‘lyse’ clots once formed (hypofibrinolysis). To date no single explanation for the aetiology has been identified and it may be that Perthes disease is simply the final common pathway of a variety of circumstances, environmental and/or genetic, that leads to AVN of the femoral epiphysis.

Perthes disease occurs in healthy children and therefore there are few accounts where the sequential radiological changes have been correlated with the histological morphology (Figure 13.18.2).

In the early stages following infarction, there is a growth disturbance. The bony epiphysis, devoid of its blood supply, stops growing but the cartilage does not and this relative overgrowth, particularly on the medial and lateral aspects of the femoral head, results in a cartilaginous coxa magna. Thus, in the early stages of Perthes disease, radiographs sometimes reveal a smaller bony epiphysis on the affected side although the apparent joint space is wider (Figure 13.18.2A).

It is postulated that recurrent infarction within the epiphysis eventually weakens the bony epiphysis and trabecular fractures and collapse occur. Radiographically, this leads to loss of bony epiphyseal height and increased density (Figure 13.18.2B). Catterall has shown that the density noted on fine-detail radiography during the stage of sclerosis is, in part, due to mechanical compression of the fractured necrotic trabeculae but, in the main, is due to calcification of the necrotic marrow within the epiphysis. During this stage, the growth plate becomes abnormal with distortion of the cell columns and an increase in the quantity of calcified cartilage in the primary spongiosa.

Once infarction has occurred, a process of repair becomes established to revascularize the avascular area. A process of creeping substitution slowly removes the necrotic bone and replaces it with fibrocartilage (Figure 13.18.2C). This repair process produces the radiological appearance of fragmentation. Islands of new bone formation in the thickened anterior and lateral articular cartilage gradually enlarge and coalesce. These islands appear radiologically as areas of calcification lateral to the epiphysis and reveal the extent to which the cartilaginous femoral epiphysis has been extruded outside the confines of the acetabulum. In the growth plate, areas of unossified cartilage extend down into the metaphyseal region producing the radiological appearance of a metaphyseal cyst. If these are large, the normal architecture of the growth plate is lost. It has been postulated that such an area might act as a tether to further growth in the anterolateral part of the neck, leading to tilting of the femoral head on the long axis of the neck. In the long term, this process might explain the radiological transcervical line described as the sagging rope sign. This radiological appearance is the result of superimposition of the anterolateral part of the femoral head overhanging the femoral neck, like the head of a mushroom overhanging its stalk (see Figure 13.18.14).

In the healing phase of the disease the necrotic trabeculae have been removed and replaced by fibrocartilage which is progressively reossified. The last portion to reform is the anterosuperior portion of the epiphysis and this is seen as a residual central lucency on the anteroposterior (AP) radiograph (Figure 13.18.2D). As the structural function of the epiphysis to maintain the femoral head shape is lost during the stages of sclerosis and fragmentation, the cartilaginous femoral head becomes flatter and broader under load, and when the healing process is complete the femoral head is ovoid rather than spherical. Clearly the extent to which the shape changes, depends on how much or how little of the epiphyseal support is affected by the disease.

Early Perthes disease typically presents with groin, thigh, or referred knee pain and an antalgic limp. This is the clinical picture of an irritable hip. On examination the most sensitive sign of hip irritability is discomfort with internal rotation, but as irritability increases abduction becomes limited and eventually an external rotation, adduction, flexion posture develops. When a child presents with an irritable hip, ultrasound identification of a joint effusion is the key investigation to confirm the clinical suspicion that the hip joint is the source of the problem. Perthes disease is but one cause of hip irritability and other differential diagnoses must be considered depending on the age of the child at presentation (Box 13.18.2). The most common explanation for an irritable hip is transient synovitis. Septic arthritis is an important differential diagnosis in any child and must always be considered. In older children a slipped upper femoral epiphysis is possible.

At initial presentation there may be no radiological evidence of Perthes disease and the condition is difficult to differentiate from transient synovitis. If a patient’s symptoms do not resolve within a week or two of presentation then a diagnosis of transient synovitis should be questioned.

A radionuclide bone scan is a useful investigation because it will show a ‘cold’ area of decreased uptake in Perthes disease which distinguishes the condition from most other differential diagnoses (Figure 13.18.3).

Magnetic resonance imaging (MRI) scanning may help identify altered signal from the epiphyseal marrow in Perthes disease but, in the author’s opinion, offers little advantage over bone scanning (Figure 13.18.4).

Eventually radiological changes will be apparent. While Perthes disease is the most common explanation for the radiological appearance of epiphyseal AVN of the hip in children, other known causes of AVN will give a similar appearance. Conditions including systemic corticosteroid therapy, sickle cell disease, other haemoglobinopathies, and storage diseases (Gaucher’s) should be considered. Bilateral Perthes disease is unusual but when present it is asymmetrical with each affected hip at a different stage. Bilateral symmetrical disease should raise suspicion of bony dysplasias such as multiple epiphyseal dysplasia.

Premature degeneration of the hip joint due to joint deformity is the potential consequence of Perthes disease. Eighty-six per cent of patients will develop osteoarthritis by the age of 65, but in the majority symptoms will not become a problem until the fifth or sixth decade (Figure 13.18.5).

It is important to understand why some hips have a better long-term prognosis than others and Stulberg’s work gives some insight into the explanation (Box 13.18.3). If the femoral head remains spherical and the acetabulum matches, the hip is termed spherical and congruent. Although this outcome is relatively uncommon such hips show no propensity to long-term degeneration. If the femoral head becomes ovoid and the acetabulum grows to match then the hip is aspherical and congruent. Such hips develop mild to moderate arthritis in late adult life. Finally, if the femoral head becomes ovoid or flattened and the acetabulum does not grow to match then the hip is aspherical and incongruent. Such hips develop severe arthritis before the age of 50. Current treatment methods focus on preventing the development of an aspherical and incongruent hip.

An early report on Perthes disease described a favourable outcome when only the anterior half of the epiphysis was involved and Catterall later published a milestone paper introducing a classification of Perthes disease which separated patient groups based on how much of the epiphysis was involved by the condition (Figure 13.18.6). The findings of this study reflect the findings of the majority of published works both past and present. Children with limited involvement of the epiphysis (group I) were shown to have a favourable prognosis irrespective of treatment. Children with onset below the age of 4 years also had a favourable prognosis irrespective of treatment and girls had a worse prognosis than boys. The recognition that a subset of patients might have a favourable outcome without treatment spared many children from the blanket application of the treatment regimens such as prolonged recumbency, weight relief, and bracing that were popular at the time.

Catterall further described five head at risk signs that might give an early indication of a potentially unfavourable outcome (Figure 13.18.7).

The lateral pillar classification recognizes the importance of the integrity of the lateral part of the femoral epiphysis (Figure 13.18.8). If the lateral portion is intact it functions like a strut, column, or pillar, bearing the load transmitted to the femoral head from the acetabular roof. The remaining medial and central parts of the epiphysis are thus protected from load bearing and epiphyseal collapse is prevented. Increasing fragmentation and collapse of the lateral pillar is associated with a worsening prognosis. This concept was first observed by Ferguson but recognition that changes in the lateral epiphysis are a bad prognostic sign was also embodied in some of Catterall’s head at risk signs.

Herring’s classification divides the femoral epiphysis into a lateral third, a central third and a medial third. Changes seen in the lateral pillar at the end of the fragmentation stage determine the group (Table 13.18.1). The Herring classification is reported to have lower intra- and interobserver variability compared to the Catterall classification.

Classification of the extent of femoral epiphyseal involvement using either the Catterall or the Herring system can only be established at the end of the fragmentation stage. Consequently neither system is able to predict how much collapse will develop until the collapse has actually occurred. This limits the usefulness of the two systems for determining who might do badly and therefore who might benefit from early treatment aimed at prevention or limitation of femoral head deformation. Salter and Thompson observed a subchondral crescentic radiolucent line early in the disease process. They believed that the line represented a pathological fracture and that the fracture was responsible for the onset of pain and clinical symptoms. It was observed that only the portion of the epiphysis underlying the subchondral fracture was affected by the avascular process. A simple classification that would predict the extent of the collapse was proposed. In Group A the fracture extended for less than half of the head and the prognosis was relatively good and in Group B (Figure 13.18.9) it extended over more than half of the head and the prognosis was relatively poor. Unfortunately the subchondral fracture was apparent in less than 33% of the patients studied.

Historically, patients have been enrolled into a variety of treatment programmes as soon as Perthes disease was recognized and it has therefore been difficult to unravel the natural history of the condition from the effect of treatment. Nevertheless most studies suggest that age, extent of epiphyseal involvement, and gender each influence the prognosis.

The prognosis is better for the younger child. The remaining growth potential is at least part of the explanation for this phenomenon. Even if a young patient’s femoral head becomes deformed and aspherical there is still time for the immature acetabulum to grow to match the femoral head and the long-term result is an aspherical congruent joint with a fair prognosis. In an older patient the more mature acetabulum has already grown to match a spherical femoral head: if the femoral head then deforms, the mature acetabulum is unable to remodel because remaining growth is limited and the result is an aspherical and incongruent joint with a relatively poor prognosis. Nevertheless, being young at disease onset does not guarantee a good outcome and up to 20% of cases in young children can ultimately reach a poor outcome, even with adequate treatment. Nonetheless most children who develop long-term poor results have disease starting over the age of 8 years.

The extent of epiphyseal involvement is defined by the Catterall, Salter and Thompson, and Herring classification systems. The prognosis for an individual case is proportional to the degree of the radiological involvement. It is also accepted that early treatment, before significant deformity has occurred, is likely to deliver a better outcome.

Most series reveal that girls have a worse outcome than boys. The reason for this is not entirely clear but some reports show that, although the condition is much less frequent in girls, a greater proportion of girls present with more extensive epiphyseal involvement than boys.

If there is extensive epiphyseal involvement during the early stages of Perthes disease, the cartilaginous femoral head loses its internal bony support and its spherical shape may not be maintained because living cartilage is a plastic material. During this period of plasticity the ultimate shape of the femoral will be influenced by the external mechanical forces applied to it. Later in the disease process when the femoral head is healing or healed, a new bony epiphysis is in place and the cartilaginous femoral head is again supported by an internal bony structure, albeit often a different shape from the original. At this stage, the femoral head is no longer plastic and the shape is fixed. Thus, once radiological healing is established there will be no further deterioration in femoral head shape and its shape cannot be influenced by the external mechanical environment. The management approach to these two periods is thus quite different. The approach to the early plastic stage requires strategies that attempt to keep the cartilaginous femoral head as spherical as possible by altering its external mechanical environment; in contrast, the approach to the late non-plastic stage must accommodate the new shape by orientating it into a congruent position, relative to the acetabulum, whilst maintaining a functional arc of movement.

The femoral head remains plastic over a prolonged period (1–2 years) until there is radiological evidence of healing. Consequently, unless the physiological processes at the root of fragmentation can be stopped or reversed, any treatment directed at changing the external mechanical environment of the cartilaginous femoral head, in order to influence shape, must be equally prolonged.

When Perthes disease was discovered, the initial management approach was similar to that used for tuberculosis of the hip joint. Programmes of prolonged non-weight bearing, including hospitalization and enforced recumbency or via the use of crutches or weight-relieving calipers were adopted. From the outset the efficacy of such treatment methods was questioned and gradually the approach was superseded by the development of newer strategies. Currently, it is accepted that non-weight bearing alone is ineffective: it does not succeed in unloading the femoral head. Indeed biomechanical studies suggest that an ischial bearing caliper might actually increase hip joint loading because of compressive muscle forces acting across the joint.

When the hip is in the anatomical position, the lateral part of the epiphysis is directly below, or just outside, the lateral rim of the bony acetabulum. In the absence of hip movement, compression forces will tend to flatten the superior part of a plastic femoral head while the uncovered anterolateral part is pushed outwards. Hip abduction brings the lateral epiphysis under the cover of the bony acetabulum redirecting the forces applied to the femoral head. The ‘containment’ hypothesis suggests that the femoral head should then become the same shape as the surrounding acetabulum, like jelly poured into a mould. This hypothesis has been adopted widely and is the principle behind most current treatment approaches.

Containment can be achieved by a number of means. The hip can be abducted, the epiphysis can be realigned surgically to face into the acetabulum by means of a femoral osteotomy, or the acetabulum can be redirected or extended over the femoral epiphysis by various pelvic osteotomies. At presentation, many patients have significant limitation of abduction usually as a consequence of synovitis and muscle spasm but occasionally due to a true mechanical limitation of abduction because of established femoral head deformity. It is self-evident that abduction must be restored before any containment approach can be applied. Methods for restoration of abduction include bed rest, traction, serial abduction casting, or even tenotomy, but if abduction cannot be restored then the epiphysis cannot be contained and treatment following the containment concept is futile.

Initially, good results were reported using bilateral above knee cylinder casts with broomsticks applied to hold each leg abducted to 45 degrees and internally rotated to 5–10 degrees for around 18 months. The Petrie cast, like many subsequent braces, is a restrictive device not conducive to ‘a normal life’. Such treatment methods have become unacceptable and a plethora of ambulatory braces such as the Atlanta Scottish Rite brace were developed to maintain abduction whilst preserving mobility. Although bracing is still used, its efficacy compared to the surgical alternatives is questioned.

A femoral osteotomy aims to redirect the lateral or sometimes the anterolateral epiphysis inside the confines of the lateral rim of the bony acetabulum. An intertrochanteric osteotomy is performed and then fixed in varus angulation. On occasion, some internal rotation or extension of the proximal fragment is included and some advocate simultaneous trochanteric growth arrest to limit the high trochanter that commonly occurs in this condition. Before embarking on surgery, it must be documented that the affected hip will abduct sufficiently to contain the epiphysis and that the patient will be left with some residual abduction range after the osteotomy has been performed. These prerequisites are usually confirmed by examination under anaesthesia and arthrography. If they are not met a varus osteotomy will only serve to put the affected leg into an adducted posture and the epiphysis will not be contained. Children who are treated between the ages of 5–8 years, when the disease is in the early or fragmentation stages and before significant femoral head deformity has developed, have the best outcome. Femoral osteotomy is not without its side effects and reported problems include failure of the varus to remodel, especially in the older patient, limb shortening, increased abductor lurch, trochanteric overgrowth, and the need to remove the fixation device.

While the femoral osteotomy moves the uncovered epiphysis under the acetabulum, the Salter innominate osteotomy does the reverse. The prerequisites for the procedure with regard to range of movement and epiphyseal containment are thus the same and again are often confirmed with an examination under anaesthesia and arthrography. Advantages of the Salter osteotomy include the absence of the iatrogenic varus and consequent limb shortening that is associated with femoral osteotomy. The Salter osteotomy typically leads to modest limb lengthening which may be advantageous given that Perthes disease usually leads to limb shortening. However, relative overlengthening tends to offset the containment effect by adducting the limb slightly when the patient bears weight. Nevertheless comparative studies show little difference between the outcome of Salter versus femoral varus osteotomy.

The Shelf osteotomy is a form of acetabular augmentation which extends the acetabular roof to cover the uncovered anterolateral femoral epiphysis. Once again the procedure has the advantage of avoiding iatrogenic varus and limb shortening from a femoral osteotomy. This can be important in the older patient who may not have the growth potential to remodel a varus alignment. This procedure is used less frequently than the femoral or Salter osteotomies but encouraging reports of its use, especially in the older patient, are growing.

If the structural integrity of the femoral epiphysis could be maintained during the revascularization of the epiphysis then femoral head collapse and deformity might be prevented. It is on these grounds that some groups are attempting to treat Perthes disease with bisphosphonates

Some authors postulate that joint distraction in early Perthes disease may prevent collapse and enhance recovery by unloading the femoral epiphysis. A few studies report the use of external fixators to distract the joint in this way but again the efficacy of this approach remains uncertain.

The clinician faces the challenge of identifying the patient who is likely to benefit from intervention and then choosing the appropriate intervention. In 1971, Catterall wrote ‘the reported results of treatment are so variable that it is difficult to be sure that the (published) series are strictly comparable’. More than two decades later Herring wrote ‘it is difficult if not impossible to compare the results of these (published) studies’. To date, no adequate prospective randomized controlled trial has compared the outcome of untreated cases of Perthes disease against similar cases treated by various means.

Herring and associates studied prospectively the outcome of Perthes disease in children over 6 years of age who had been managed by various means. The outcome of physiotherapy and several containment methods (surgical and non-surgical) were compared. The study identified a group of patients who benefit from containment treatment and showed that surgical containment methods (both femoral varus osteotomy and Salter innominate osteotomy) led to a better outcome that non surgical containment by abduction bracing but only some patients benefited (Table 13.18.2).

When the lateral pillar was completely preserved (Herring A) all patients did well and treatment, by any means, made little difference. Therefore if a Group A patient could be identified accurately, treatment could be withheld safely. When the lateral pillar collapsed completely (Herring C) the outcome was poor and treatment by any means made little difference. This implies that it is futile to offer such patients any of the containment methods studied. Unfortunately the clinician managing a patient who presents with early disease is unable to foretell whether the presenting hip will remain a Herring A or become a Herring C and it is therefore difficult (or impossible) to decide to withhold treatment for a particular patient on radiological grounds alone. Moreover it is accepted that treatment before deformation occurs is preferable and it is therefore difficult to deliberately delay treatment, and risk compromising the outcome, while waiting for the hip to declare itself.

It is apparent that although the Catterall and Herring radiological classification systems are useful as research tools to ensure that studies can compare like patients at final review, they cannot be used to determine which patients should be treated in the early stages of the disease. With this backdrop how can the clinician make rational management choices?

How can current knowledge regarding Perthes disease be resolved into a useable clinical management approach? The clinician manages patients not radiographs and the author’s philosophy is a clinical rather than a radiological approach (Figure 13.18.10).

All patients who develop Perthes disease must contain the vulnerable epiphysis. Some patients achieve this containment for themselves by maintaining a full range of movement, especially abduction. The author refers to this circumstance as passive containment and the patient simply requires observation not intervention.

Clinical evaluation of abduction is therefore the most important decision making assessment. During the clinical examination it is important to abduct the unaffected hip maximally and flex the knee over the edge of the examination couch to lock the pelvis before measuring abduction on the affected side otherwise unconsciously rolling or tilting of the pelvis means that abduction of the affected hip can appear spuriously good.

Passive containment is most apparent in the young patient (under age 4). Clinical experience shows that many young patients maintain an excellent range of movement with modest symptoms. These favourable clinical findings do not always match the sometimes unfavourable extent of radiological epiphyseal involvement but may explain why the outlook is often good for younger patients. The bony nucleus of the young epiphysis only represents a small proportion of the whole cartilaginous head and its contribution to maintenance of femoral head shape is modest in the younger (and lighter) patient. In older patients extensive femoral epiphyseal involvement is typically matched by significant clinical symptoms and loss of movement, especially abduction.

If a patient begins to lose range of movement, particularly with abduction less than 20 degrees, then the lateral femoral epiphysis will not move in and out of the acetabulum during normal activity. Such patients are unable to contain their own epiphysis adequately. This loss of clinical abduction is the cue that the patient needs help in containing the epiphysis and the clinician must intervene. The author refers to this circumstance as active or interventional containment. This approach is inclusive to all patients, even for the rare case of a young patient who presents with such loss of abduction. It is possible that the small proportion of young patients who present in this way are likely to represent the young patients whose outcome is unexpectedly poor when managed passively. Given current evidence it seems reasonable to expect interventional containment in such children to improve the outcome, although there is no study to date to support this notion.

Based on current evidence, the author’s preferred interventional containment procedure for children under 8 years of age is a femoral varus osteotomy (Figure 13.18.11).

Following the surgical procedure the patient is immobilized in a spica cast for 6–8 weeks to maintain abduction. Experience has shown that without the cast, movement is lost and the hip adducts.

Femoral varus osteotomy in the child over age 8 years has some distinct disadvantages. There may not be enough growth remaining for the iatrogenic varus deformity to remodel with time and if the deformity persists it leads to limb shortening and a Trendelenburg lurch. The author’s preferred method of interventional containment for the older patient is the shelf osteotomy (Figure 13.18.12).

This procedure has none of the disadvantages of the varus osteotomy and increasingly, reports show its efficacy as a containment procedure. The author has found that postoperative immobilization with a spica cast is unnecessary. Partial weight bearing is allowed with increase of the load allowed over a period of 12 weeks.

Before embarking on an interventional containment approach the prerequisites of the chosen method must be met. Examination under anaesthetic and arthrography confirms that containment is possible and sufficient abduction is available. Early flattening of the cartilaginous femoral head is acceptable provided the lateral part of the epiphysis comes under the cover of the bony acetabulum (Figure 13.18.13A).

According to the author, interventional containment is indicated because the patient has lost abduction in the outpatient setting. This loss of abduction is present only when the patient is conscious: it is usually the result of muscle spasm secondary to synovitis and pain. Once the patient is under general anaesthetic the spasm is overcome and most patients then have an adequate range of abduction and a containable epiphysis. If abduction is limited, even under anaesthetic, it may sometimes be restored by serial casting, or tenotomy, but if the restriction is mechanical because of hinge abduction (Figure 13.18.13B) then the epiphysis is not containable and interventional containment is contraindicated.

The author believes that patient follow-up following the management of Perthes disease should be continued until they reach skeletal maturity. Patients rarely develop symptoms that require attention in their childhood or early adult life but most do develop a modest shortening on the affected side of about 10–15mm. Although this rarely requires intervention, limb length monitoring is the primary reason for follow-up. If the discrepancy is greater than expected or if symptoms develop, then monitoring during growth will facilitate limb length equilibration using a well timed epiphysiodesis.

Many childhood hip disorders leave a deformed femoral head in their wake. If a patient develops an aspherical and incongruent joint then symptoms of discomfort can present as early as the late teens or early twenties. The typical anatomy of the deformed hip includes a short broad femoral neck with a high greater trochanter and a large broad aspherical (flattened) femoral head (Figure 13.18.14). The acetabulum may match the femoral head shape or not and the severely deformed femoral head develops a saddle-shaped depression under the region of the bony acetabular rim.

These anatomical deformities present clinically as a short limb and a Trendelenburg lurch with abductor fatigue and discomfort in the gluteal muscles after activity. Superior flattening and increasing width of the femoral head makes the head ovoid rather than spherical. While a spherical joint allows easy movement in all directions the ovoid head behaves more like a roller with its long axis in the coronal plane. There is usually good flexion but limited rotation and abduction. Although patients often have good flexion they cannot always achieve full flexion with the leg in the sagittal plane. As flexion proceeds the bulge of the deformed anterolateral part of the femoral head will not fit under the anterior lip of the acetabulum. To achieve further flexion the patient’s leg (and femur) must externally rotate to direct the bulge around the anterolateral rim of the acetabulum: the classical sign of a deformed femoral head.

In the worst cases the superior surface of the femoral head is flat or even concave and abduction is severely limited because the deformed femoral head will not fit under the lateral rim of the acetabulum when abducted. If the hip is forced into abduction, the lateral part of the head impinges on the lateral rim of the acetabulum. Further abduction causes the hip to be levered from the medial floor of the acetabulum with the fulcrum or ‘hinge’ of the lever at the point of impingement. This is so called ‘hinge abduction’. A patient whose hip hinges will not allow any abduction when conscious because it is too painful. Under anaesthetic the hip can be forced into abduction and the hinging demonstrated on an arthrogram (Figure 13.18.13B).

Once the femoral head has reached the healing stage its shape is fixed and cannot be further influenced by containment. The objective now is to accept the shape and align the femoral head with the acetabulum in a congruent position that also offers the patient a functional arc of movement. The management principle is the ‘position of best fit’ and can be applied to any deformed femoral head.

Dynamic arthrography is the key investigation for evaluation of the deformed hip. In the typical case, following Perthes disease the arthrogram often demonstrates a congruent fit when the hip is adducted. In these circumstances a proximal femoral valgus osteotomy will reproduce the congruent alignment when the patient is standing (Figure 13.18.15).

The valgus osteotomy will allow some clinical abduction before the point of impingement is reached. Furthermore the osteotomy can sometimes lengthen the short leg and the greater trochanter will be brought to a lower and more lateral position, lengthening the abductor lever arm and improving the Trendelenburg lurch and abductor fatigue. The procedure reduces symptoms and can extend the lifespan of the patient’s hip joint before interventions such as joint replacement become necessary.