Abstract
Hip dysplasia, subluxation, and eventual hip dislocation are commonly encountered in the cerebral palsy population secondary to spasticity and loss of motor control, especially in those patients with more severe neurologic involvement. The treatment of hip disorders in these patients should take into account the degree of limb and hip involvement, pain severity, and overall functioning. Conservative management focuses on mitigating spasticity and preserving range of motion in order to provide an environment in which the femoral head remains concentrically reduced in the acetabulum. However, operative management, consisting of soft tissue or tendon releases, femoral or pelvic osteotomies, or hip salvage procedures, is sometimes necessary to treat the painful, subluxated, or dislocated hip. Radiographic hip surveillance in the pediatric cerebral palsy population is used to guide operative treatment. Long term hip containment is generally improved when surgical intervention is performed in the earlier stages of dysplasia. Younger patients who demonstrate progressive hip subluxation despite conservative measures may be carefully selected to undergo soft tissue procedures. Bony reconstruction, with adjunctive soft tissue procedures, is often necessary to better contain the proximal femur in patients above the age of four years.
Keywords
Introduction
Radiographic hip surveillance in the pediatric cerebral palsy population is critical in guiding early and proactive operative treatment [1–4] There is a higher incidence and severity of hip dysplasia in patients with declining gross motor function. Hip dysplasia is much more prevalent and often necessitates operative treatment in nonambulators (i.e., Gross Motor Function Classification System [GMFCS] levels IV and V), in whom increasing spasticity, muscle imbalance, and the lack of joint loading can hinder normal acetabular development [5].
Standard pelvis and frog lateral hip radiographs may be obtained in the cerebral palsy population as early as 18 months of age [6]. Radiographs should definitely be obtained by 30 months of age in all patients unless difficulty with ambulation arises earlier [7] or patients begin to manifest groin pain, limited hip abduction or adductor spasticity, or a positive Galeazzi sign prior to 30 months of age. Monitoring for any serial hip subluxation, as evidenced by Reimers’ migration index percentage (MP) [8], is typically performed semiannually or annually with pelvis and frog lateral hip radiographs. It is advisable to involve the pediatric orthopedist early when treating children who manifest dysplastic hip changes or femoral head uncovering, persistent groin pain, or abnormal hip motion.
Soft tissue procedures
Adductor and hip flexor spasticity drives the femoral head against the posterosuperior acetabulum, which can become deficient due to these abnormal muscular forces [9]. This makes hip containment progressively more difficult as the resultant acetabular dysplasia worsens through growth. It is postulated that performing hip adductor release, with or without iliopsoas lengthening and hamstring lengthening, at a young age allows time for proper acetabular development up to skeletal maturity [10]. Soft tissue releases alone are typically considered in patients under the age of four to eight years with the presence of any serial hip subluxation, if the MP is not more than 25–50% [10, 11]. Children with minimal to no hip dysplasia who are experiencing groin pain or difficulty with perineal care, along with scissoring and limited hip abduction (typically less than 30 degrees), also may be indicated for soft tissue procedures [10].
There is little consensus on when additional iliopsoas or hamstring lengthening should be performed, since Reimers concluded that the addition of iliopsoas lengthening to adductor tenotomy had no clear effect on mitigating future hip subluxation [8]. Most studies demonstrate satisfactory outcomes in anywhere from 64 to 72% of patients undergoing adductor tenotomy alone, independent of other muscle groups being addressed [10, 13]. However, most surgeons will opt to perform iliopsoas lengthening when a Thomas test demonstrates a hip flexion contracture in ambulators. Additionally, hamstring recession or lengthening may be helpful if the popliteal angle is greater than 45 degrees to decrease displacement forces across the hip joint [10]. However, combined medial and lateral distal hamstring lengthening has been previously reported to result in increased anterior pelvic tilt, particularly in GMFCS III patients [14, 15].
It is recommended to begin with adductor longus tenotomy and, if intraoperative hip abduction (with the hip flexed) remains below 45 degrees, then release of the gracilis and adductor brevis may be necessary [16]. This procedure can be performed through an open incision abutting the pubic tubercle or percutaneously, with one study demonstrating similar efficacy in preventing further subluxation and no significant differences in complication rates or neurovascular injury [17]. Iliopsoas lengthening is done when there is a significant hip flexion contracture, either by releasing the tendon entirely from the lesser trochanter in nonambulatory patients or dividing the myotendinous junction at the pelvic brim in those patients who ambulate [18]. To date, there are no large scale comparative studies to support percutaneous myofascial lengthening for all muscle groups addressed in the cerebral palsy hip, while open tendon lengthening remains well-studied and reproducible.
Once femoral head subluxation becomes more severe (MP above 50–60%), soft tissue procedures are less effective alone and bony reconstruction becomes necessary [10, 19]. Presedo et al. found that adductor tenotomy with iliopsoas lengthening provided satisfactory outcomes when MP was 23% (range, 0–40%) at one year postoperatively, but unsatisfactory outcomes were seen with a higher average MP of 34% (range, 10–60%) one year postoperatively [10]. All hips that were severely subluxated (MP > 60%) one year postoperatively required later bony reconstruction [10]. Additionally, soft tissue releases tend to be more successful in children who ambulate, in whom improved motor and neurologic function aids proper acetabular development. A recent study demonstrated no evidence for further hip migration after adductor tenotomy alone in 94% of children at GMFCS level II, as opposed to 49% of children at GMFCS level III, 27% at GMFCS level IV, and only 14% at GMFCS level V [16].
Bony reconstruction
Proximal femoral varus derotational osteotomy (VDRO) is performed in patients above the age of four who demonstrate a worsening Reimers’ migration index past 40–60% [20]. This provides derotation, shortening, and varus reorientation, thereby correcting increased valgus and anteversion which contribute to hip instability. The osteotomy is typically fixed with a side plate and screws, aiming for a neck-shaft angle (NSA) of closer to 100 degrees in nonambulatory patients and 110–120 degrees in ambulatory patients [21]. Greater varus reorientation (i.e., decreased NSA) improves femoral head coverage in nonambulant patients, but excessive varus angulation may limit hip abduction and negatively alter gait mechanics in those patients who ambulate.
Pelvic osteotomy is indicated when acetabular deficiency and hip instability are present or the hip remains laterally subluxated following completion of the proximal femoral osteotomy [22]. Song et al. also recommended pelvic osteotomy when preoperative MP is above 70% [23]. The reverse Dega or San Diego osteotomy, an incomplete transiliac osteotomy, allows the surgeon to insert a bone graft wedge to reshape the dysplastic acetabulum [24, 25]. This provides increased posterosuperior acetabular coverage, which is typically deficient in the cerebral palsy population [24, 25]. This has been used with good outcomes, with a study by McNerney demonstrating 95% hip stability at 7 years with a low rate of avascular necrosis [25]. In skeletally mature patients with dysplasia, surgeons have also used modifications of the Bernese-Ganz periacetabular osteotomy [26].
Combined pelvic and femoral osteotomies have been shown to be long-lasting and durable when severe hip subluxation is present. Song reported long-term re-subluxation rates of only 12% when pelvic osteotomy was combined with VDRO, as opposed to 26% in patients receiving only VDRO [23]. Addition of a pelvic osteotomy to a femoral osteotomy improved pain and radiographic parameters, including acetabular index and center edge angle, at follow-up of four years in a group of 52 hips [27]. Twenty-five percent of these patients undergoing VDRO alone eventually required repeat bony reconstruction, while no revision reconstruction was required following combined osteotomies. Another study elected to defer pelvic osteotomy when intraoperative hip arthrogram demonstrated a concentrically reduced hip following VDRO; they found that 21% of patients with severe preoperative hip subluxation (MP > 50%) in which arthrogram indicated VDRO alone eventually required repeat bony reconstruction [28]. There should be a low threshold to include a pelvic osteotomy when severe preoperative hip migration is seen, even if the hip appears reduced following VDRO. In these cases, an eroded and dysplastic acetabulum from progressive hip migration may not provide sufficient long term hip containment. The addition of a pelvic osteotomy also appears to lower the risk of later repeat bony reconstruction when compared with proximal femoral osteotomy alone.
Remodeling following bony reconstruction in the cerebral palsy hip is not always predictable due to persistent abnormal neuromuscular forces. Recurrent subluxation requiring a second reconstruction may be necessary in patients who receive surgical intervention at a very young age [29]. Patients undergoing VDRO under four years of age lost up to 96% of the correction of the NSA and 42% of the anteversion angle at a mean follow-up of 15.4 years [30]. In contrast, Rutz et al. performed combined pelvic and femoral osteotomies on a group of older patients (average 11 years, n = 168) and found that re-subluxation requiring repeat hip reconstruction occurred in only two patients at an average 7 year follow-up [31]. However, each one year decrease in age at time of surgery was associated with worsening of joint congruency, alignment, and femoral head contour through follow-up. Noonan et al. actually found less postoperative hip migration in patients undergoing VDRO prior to 6 years of age (18% of patients < 6 yrs. versus 33% of patients > 6 yrs.) [32], while Mallet et al. demonstrated that patients both below and above 8 years of age at combined femoral and pelvic osteotomy had similar outcomes [33]. This suggests an optimal age range for VDRO closer to 4 to 6 years, while age appears to have less of an effect on the long-term success of combined pelvic and femoral osteotomy.
Further studies are needed to elucidate the long term results and complications of bony reconstruction, which should be optimally timed to avoid the need for recurrent surgery. VDRO should be considered as young as four to eight years of age if soft tissue procedures do not fully ameliorate hip subluxation or adductor spasticity. As patients get closer to the age of eight and older, pelvis osteotomy may be necessary when severe hip migration and altered acetabular morphology is seen. Delaying bony reconstruction until this is indicated by serial progressive hip migration allows for the surgeon to operate when the patient is older with greater physiologic reserve, but this must be balanced with the decreased potential for acetabular remodeling as patients approach adolescence.
Salvage procedures
Ideally, hip salvage procedures will be unnecessary when proper hip surveillance is performed. However, salvage procedures may be required in the persistently painful hip in which bony reconstruction has failed or is not possible due to end-stage hip dysplasia or degeneration. Hip salvage surgery is considered when conservative measures, including spasticity reduction, medical therapy, and bracing, have not been effective in alleviating pain secondary to degenerative joint disease. Salvage is more commonly utilized in nonambulators in whom severe femoroacetabular dysplasia and poor bone stock often coexist, and extensive bony reconstruction may not be attainable. The primary goals of hip salvage are to provide pain relief while maintaining reasonable hip motion to allow for continued function, perineal care, and daily hygiene.
Proximal femoral resection with interposition of soft tissue has been successful [34]. This shortens the femur in an effort to avoid proximal femoral migration due to spasticity. Proximal femoral valgus osteotomy at the subtrochanteric level is another option, which shifts the femoral head laterally away from the acetabulum to avoid friction and subsequent pain. McHale demonstrated that the proximal femoral valgus osteotomy successfully avoided superior head migration, resulting in less stiffness and heterotopic bone formation compared to proximal femoral resection alone [35]. Total hip arthroplasty or hip resurfacing is rarely performed for children or adolescents but may be considered in select cases, and hip arthrodesis may also be performed in adolescents or young adults. A recent systematic review found better pain relief in patients who underwent proximal femoral resection, proximal femoral valgus osteotomy, or total hip arthroplasty as opposed to hip arthrodesis [36].
Ambulators requiring hip salvage will typically gain better function from total hip arthroplasty, since this more reliably restores proper joint function and gait mechanics. This must be balanced with the increased risk for dislocation and complications following total hip arthroplasty in the young ambulant cerebral palsy population [36]. Overall, the literature appears to suggest that both ambulators and nonambulators who undergo salvage procedures experience long-lasting pain relief, even though postoperative gait and functioning may be less predictable.
Conflict of interest
No funding or conflicts of interest to disclose.