Operative treatment for hip subluxation in spinal muscular atrophy

Patients and Methods

We retrospectively analysed the records of patients with Evans type II or type III spinal muscular atrophy and recorded the current level of pain, function, and satisfaction. All patients were assessed on a six-monthly or yearly basis at a tertiary-level specialist clinic for children with neuromuscular disorders. The diagnosis was confirmed in all cases by a muscle biopsy. Radiographs of the hips (antero-posterior pelvis) and spine (posteroanterior view entire spine) were performed at the first review and when clinically indicated thereafter. During growth, children with subluxed hips were offered surgery to relocate the hip, improve acetabular cover and prevent future functional problems. The preferred operation was a single-stage combined procedure including appropriate soft-tissue releases (rectus femoris, psoas, adductor longus, gracilis and anterior portion of adductor magnus, open reduction, femoral varusrotational osteotomy with shortening where indicated, combined in some cases with a Chiari pelvic osteotomy. Bilateral operations were usually performed six weeks apart. Early in the study the care of the hips and spine was undertaken by the same surgeon. Later, a specialist spinal surgeon managed the scoliosis. Surgery to stabilise the scoliosis was indicated for progression of the curve and/or pain when sitting in a wheelchair.

The patients were assessed by a specialist registrar in orthopaedics (MZ), a neuromuscular specialist nurse (CC) and a senior physiotherapist who worked closely with the children and their families. Clinical assessment included hip pain and ability to sit. Two observers reviewed the radiographs independently, measuring the Cobb angle, fixed pelvic obliquity, and the percentage migration of the femoral head.

Fixed pelvic obliquity was measured as the angle between the transiliac crest line and a line perpendicular to the tips of the spinous processes of L4 and L5.⁸ A pelvic tilt of more than 10° was considered significant. The migration percentage as described by Reimers,⁹ is the fraction of the femoral head outside the acetabular rim on an anteroposterior radiograph. Measurements were made along the transischial line or the line joining the medial edges of the triradiate cartilages. Where the measurements varied significantly between observers, an average of the two was used. A migration percentage of 33% or more indicated that the hip was subluxed, and 100% signified dislocation.

Results

Fifty-four children with spinal muscular atrophy were seen or referred for treatment between 1982 and 1996. Of these, four were severe type I. They all died in the first few months of life and no orthopaedic intervention was required. Ten children were type IV. Five of these had scoliosis, but only one needed surgery. Another child with type IV had a triple arthrodesis for pes cavus. Children with types I and IV were excluded from the study.

Forty children had types II or III. Ten were lost to follow-up, and therefore only 30 children participated in the study. Of these, 16 had type II and 14 type III (Table I).

Nine children underwent surgery to reduce a dislocated hip (Table II). The mean post-operative follow-up period was nine years (3 to 15). The mean age at the latest follow-up was 15.8 years (8 to 21). There were eight bilateral procedures and one unilateral. Four of the 17 operated hips (23.5%) had subluxed at the latest follow-up. Three of these had a significant pelvic tilt. Only one child reported an improvement in hip pain and function after surgery. A family with two children with intermediate spinal muscular atrophy refused surgical treatment for subluxation of the hip in one child, as they were not satisfied with the result of the operation on the other child.

All hip procedures included soft-tissue releases (adductor, rectus femoris and psoas), open reduction and femoral varus-rotational osteotomy (with shortening, where indicated). Ten operations included a Chiari as well as a femoral osteotomy.

One of the seven open reductions that only had a femoral osteotomy as a bony procedure was revised because of re-subluxation, and a Chiari osteotomy was performed at that time. Unfortunately, the hip subluxed even after revision surgery. At the latest follow-up, of the six other hips that had only a femoral osteotomy, one was found to be subluxed with a migration percentage > 33%.

Of the ten open reductions combining femoral and Chiari osteotomies, one was revised because of re-subluxation and a shelf procedure was performed at that time. Two of the other nine hips were subluxed at the latest follow-up.

Three post-operative complications followed these 17 open reductions. One child had a superficial wound infection, and another a chest infection. A third developed numbness over the toes and dorsum of the foot, and weakness of dorsiflexion of the toes and ankle. These signs persisted at the latest follow-up. One child died three years after hip surgery from unrelated causes.

Twenty-one children (12 girls, 9 boys) with type II or III spinal muscular atrophy did not have a hip operation. Their mean age at latest follow-up was 16 years (7 to 23). Of the 42 hips, 28 (67%) hips were subluxed in 18 patients. Of those 28, 21 (75%) had a significant pelvic tilt. Only one child with bilateral hip subluxation complained of bilateral hip pain.

Discussion

The propensity for dislocation or subluxation of the hip in spinal muscular atrophy is well recognised.^5,7,10 Dislocation seems to be secondary to global symmetrical proximal weakness, with resulting coxa valga and lateralisation of the hip. Diminished weight-bearing with profound gluteal weakness probably depresses the stimulus for trochanteric apophyseal growth. This provokes coxa valga, which contributes to subsequent subluxation. Type-IV patients are treated as unaffected, and type I do not require surgical treatment because the disease is fatal in early life.

Controversy surrounds the management of dislocation of the hip in types II and III spinal muscular atrophy. Advocates of surgical treatment believe that a located hip is required to improve balance, maintain pelvic alignment, minimise pain and reduce the incidence of ischial pressure sores. For example, Shapiro and Bresnan⁵ recommended aggressive soft-tissue procedures and varus osteotomy of the proximal femur in these patients. Others have questioned the benefits of operative treatment, because of recurrent subluxation requiring further surgery.¹¹ In our study four of the nine children who had hip surgery had subluxation of the hip at the latest follow-up. Sporer and Smith¹⁰ reported minimal pain and functional difficulties when reviewing 41 patients with types II and III spinal muscular atrophy who were treated conservatively with 18 years of follow-up.

Pelvic tilt in neuromuscular disorders is part of a complex deformity that involves the hip and the spine, thereby reducing the potential for sitting. Our data show that pelvic obliquity in the intermediate forms of spinal muscular atrophy is strongly associated with but not always present in subluxation or dislocation of the hip. Frischhut et al¹² reported that pelvic tilt in neuromuscular disorders can be effectively corrected by spinal surgery, but no effect was demonstrated with hip surgery. Our data suggest that there may be an improvement in pelvic tilt after hip surgery. Of the children who had open reduction four of nine (44%) had pelvic obliquity (> 10°), compared with 14 of 21 (67%) who had pelvic obliquity (> 10°) and did not have hip surgery.

Our study showed that patients who had surgery for dislocation of the hip had better femoral head cover at the latest follow-up. However, this was not associated with a better functional outcome. Only one of the nine operated patients who had surgery reported an improvement. Conversely, only one of the 21 patients who were treated conservatively reported pain related to subluxation or dislocation. Clinical review of our patients failed to demonstrate better perineal care or sitting balance following open reduction of the hip. Our results agreed with those of others^7,10,11 who reported minimal or no symptoms following hip subluxation or dislocation in patients treated conservatively or operatively. This may be due to profound muscle weakness which, unlike spasticity, produces minimal tension or pressure between the dislocated femoral head and soft tissues or the ilium.

Of the 17 hips treated surgically, ten had both a Chiari and a femoral osteotomy, seven had only a femoral osteotomy. Although these numbers are too small for statistical analysis, it seems that the addition of a Chiari osteotomy was not related to a better outcome. Two children, one in each group, had a revision procedure for subluxation. At the latest follow-up one hip was subluxed in the group that only had a femoral osteotomy, and two in the group that had a Chiari osteotomy as well.

The four children who had further subluxation of the hip after surgery were followed up for more than nine years. It is possible that more hips would have subluxed with longer follow-up. There were two children with respectively 10 and 12 years of follow-up who did not have subluxation when last seen. These children were classified as Evans type III, but the small numbers do not allow any definite conclusions.

Considering the complexity of hip surgery and anaesthesia in patients with spinal muscular atrophy, their frequent hospitalisation for other medical reasons, and the high rate of recurrent subluxation, we found no justification to operate on hips which are at risk of dislocation. We no longer offer surgery to patients with intermediate spinal muscular atrophy and subluxation of the hip in the absence of specific problems related to subluxation.