Should Children with Myelomeningocele and Scoliosis Undergo Spinal Surgery?

In children with myelomeningocele, scoliosis is a common and complicated problem. Most of the published literature on this topic addresses the technical correction of spinal deformities. Perhaps more important, however, are the functional consequences of such a serious surgery in this patient population.

Studies have clearly shown the ability to achieve significant reduction of the scoliotic curvature with surgical treatment in patients with myelomeningocele.^1–3 Progressive, untreated scoliosis often results in loss of truncal stability, which may endanger sitting balance. Thus, the main goals of surgical intervention should be prevention of progressive spinal deformity and improvement of sitting balance.^1,4

It is important to note that a corresponding improvement in ambulation, motor skills, and activities of daily living (ADL) has not been demonstrated in the literature. In fact, studies evaluating ambulatory ability after spinal fusion have suggested ambulation may be more difficult after surgery.^{1,2, 4} Multiple studies have also shown no significant difference in the ability to perform ADL after surgical intervention.^{1,2, 4}

Evidence

One study by Schoenmakers and colleagues¹ assessed 10 children with myelomeningocele who underwent spinal surgery. They found that ambulation became more difficult for three of the four patients who had been able to ambulate before surgery. They also note no long-term effects on ability to perform ADL after surgery (Level IV).

Another study by Müller and coworkers² examined the influence of surgical treatment for scoliosis on both ambulation and motor skills. This study included 14 patients from different levels of dysraphism. They found that in the eight patients with preoperative hip flexion contractures, averaging 15.2 degrees, there was a significant increase after surgery to 38.4 degrees. No patients experienced improvement in hip flexion contracture after surgery. In addition, seven patients lost the ability to ambulate with or without assistive devices after surgery. These authors did note three patients who gained better sitting balance. They also found no significant difference in the ability to manage ADL from before to after surgery. Of note, the authors found no significant difference in the postoperative changes with regard to motor skills, ambulation, and ADL between the different levels of dysraphism (Level IV).

A study by Mazur and researchers⁴ examined the effect of spinal fusion on sitting balance and ambulatory ability in 49 patients. They grouped results according to whether patients underwent staged anterior and posterior fusion, posterior fusion alone, or anterior fusion alone. They found improved sitting balance in 70% of patients after anterior and posterior fusions, 67% after posterior only, and 28% after anterior only. The authors also report an adverse effect on ambulation in 67% of patients with combined fusions, 27% of posterior-only fusions, and 57% of anterior-only fusions. No patients in this study showed improvement in ambulatory status after surgery. The authors also note little change in ability to perform ADL after surgery (Level III).

In a study by Wai and investigators,³ 80 children with myelomeningocele were assessed to determine the relation of spinal deformity to physical function and self-perception. They found no significant relation between spinal deformity and overall physical function or self-perception. The only aspect of spinal deformity that showed an effect on one aspect of physical function was coronal imbalance on sitting. The authors conclude that simple interventions such as chair modifications should be explored as a means to improve coronal balance and sitting function (Level IV).

It is important to keep in mind that the presence of infrapelvic pelvic obliquity can also contribute to the overall imbalance seen in patients with myelomeningocele with scoliosis. As an example, this could result from the combination of a unilateral hip adduction contracture with a contralateral hip abduction contracture. Patients should be assessed for infrapelvic obliquity, and if present, this should be addressed at the same time as correction of any spinal deformity.

In Children Who Undergo Surgery, Should Fusion Extend to the Sacrum?

In patients with spinal deformity and pelvic obliquity, the procedure of choice typically involves combined anterior and posterior surgery with fusion extending to the sacrum to achieve better curve correction and a lower pseudoarthrosis rate.⁵ Indications for extension of fusion to the sacrum include progressive scoliosis with lumbar spine involvement, neuromuscular scoliosis, pelvic obliquity greater than 15 degrees, poor sitting balance, and posterior lumbar and sacral dysraphism.⁶

Including the sacrum in fusion results in better correction of pelvic obliquity and improves sitting balance. However, it is important to note that fusion of the spine to the pelvis creates a rigid trunk, making walking more difficult.⁴ This is especially true in patients with myelomeningocele accustomed to ambulating with a swinging gait. In addition, extension of fusion to the sacrum with the resulting loss of lumbosacral mobility can adversely affect ability to perform wheelchair transfer.⁷ Another potential complication associated with fusion to the sacrum is the development of ischial pressure sores. Various studies have reported on this complication with a frequency varying from 3%⁵ to 33%.⁸

Wild and coworkers⁷ performed a prospective study on 11 patients with myelomeningocele who underwent two-stage anterior and posterior fusion without fixation to the sacrum to evaluate their functional outcome. With an average of 4 years 11 months of follow-up, they found that pelvic obliquity spontaneously corrected when scoliosis was adequately treated with instrumented fusion not including the sacrum. In their study, the achieved correction remained stable at follow-up. The authors believe that sparing the lumbosacral segment from primary fusion offered the patients better freedom of mobility. They also state that these patients were spared the risk for complications associated with lumbosacral fixation including hardware failure, loss of correction, pressure sores, high infection rate, and high pseudoarthrosis rate (Level IV).

Should the Technique Used for Fusion Involve Anterior Surgery Only, Posterior Surgery Only, or Combined Staged Procedures?

Multiple different surgical techniques have been used to treat children with myelomeningocele and spinal deformity who are deemed appropriate candidates for operative treatment. These techniques typically fall within three broad categories: single-stage anterior fusion, single-stage posterior fusion, and two-stage anterior and posterior fusion. Multiple studies exist with the goal of determining which of these techniques provides the best results whereas minimizing risk for complications.

In a classic study, Osebold and investigators⁸ examined a group of 40 patients with myelomeningocele treated with either posterior fusion alone, anterior fusion alone, or combined anterior and posterior fusion. They found a high rate of pseudoarthrosis (46%) with posterior fusion alone and a loss of initial correction for an average final correction of scoliosis of only 12 degrees. In contrast, when anterior fusion with instrumentation was used in combination with instrumented posterior fusion, they report an average final correction of 45 degrees with a decrease in the pseudoarthrosis rate to 23% (Level III).

Banta⁵ also reports on a series of 50 children treated with combined anterior and posterior fusion. He notes correction of scoliosis from an average of 73 degrees to average of 34 degrees, with a pseudoarthrosis rate of only 12%. He concludes that combining anterior fusion with posterior fusion leads to increased correction of both spinal deformity and pelvic obliquity whereas also contributing significant strength to the fusion mass (Level IV).

Ward and colleagues⁹ performed a retrospective review of 38 patients with myelomeningocele who underwent surgical correction of scoliosis with at least 2 years of follow-up. Patients were treated with either one-stage anterior fusion, one-stage posterior fusion, or two-stage anterior and posterior fusion. The determination of success or failure was made on the basis of radiographic results only, assessing radiographic fusion, lack of curve progression, or both. The authors report a 50% failure rate with anterior or posterior fusion alone, compared with only an 8% failure rate with two-stage procedures (Level III).

Banit and coauthors¹⁰ report on a series of 50 patients with myelomeningocele treated with posterior-only spinal fusion for scoliosis. The authors examined whether the use of modern segmental instrumentation systems could obviate the need for two-stage procedures. Using historical controls, they found an equivalent degree of correction of scoliosis, truncal decompensation, and pelvic obliquity in their study group. They also report a pseudoarthrosis incidence rate of 16% in their study group. The authors conclude that using segmental instrumentation decreases the pseudoarthrosis rate compared with that reported with nonsegmental instrumentation. However, regardless of the instrumentation used, they did not believe posterior-only techniques could match the fusion rates obtained with combined anterior and posterior surgery (Level IV).

In another study, Basobas and coworkers¹¹ assessed the results of selective anterior fusion with instrumentation for the treatment of neuromuscular scoliosis in 21 patients. The authors propose that certain carefully selected patients may benefit from preservation of motion segments whereas still achieving good curve correction. They report an average curve correction of 60.3 degrees with correction of pelvic obliquity from an average of 15.1 to 3.2 degrees. Of note, they also found no change in the ambulatory status of any patient. They had only one patient (4.7%) with pseudoarthrosis. The authors conclude that anterior fusion alone is a valuable alternative to posterior fusion alone or combined anterior and posterior techniques based on their demonstrated success with curve correction and low complication rate. However, they did not comment on selection criteria for patients to be treated appropriately with anterior fusion alone (Level IV).

In contrast, Sponseller and coauthors¹² report on a series of 14 patients treated with anterior fusion alone for myelomeningocele scoliosis to define appropriate selection criteria. They conclude that anterior-only fusion may offer significant advantages including less extensive surgery with lower infection rate than posterior fusion. However, they advocated this technique only in patients with thoracolumbar curves less than 75 degrees, compensatory curves less than 40 degrees, no increased kyphosis, and no syrinx (Level IV).

In summary, many authors have reported on the challenges, complications, and poor results associated with techniques of anterior fusion alone and posterior fusion alone.^8,9 The current literature widely supports the use of staged combined anterior and posterior fusion to adequately treat patients with myelomeningocele with scoliosis and achieve the highest rates of fusion (Grade C). Only in carefully selected patients should the option of anterior-only fusion be considered. However, in the appropriate population, anterior-only fusion can provide the advantage of preserving motion segments and maintaining the ability to perform ADL with a lower rate of infection.^11,12

Does Hip Surgery in Patients with Myelomeningocele Provide Improved Functional Results?

Transfer of iliopsoas is a technique that has been used in patients with myelomeningocele to maintain reduction of paralytic hip dislocations. Numerous studies exist reporting on the results of iliopsoas tendon transfer; however, many of these studies classify success or failure on the basis of radiologic results only. Critical review of the literature is necessary to determine the functional implications of hip surgery in the myelomeningocele population. Concerns exist regarding whether stable hip reduction leads to restricted range of motion and pathologic fractures, which would compromise the functional result.¹³

Evidence

Sherk and Ames¹³ reviewed a series of 36 patients with myelomeningocele with an average follow-up of 7 years who underwent iliopsoas transfers with open reduction and capsular plication for dislocated hips. They found that 47% maintained reduction of the hip, but 17.6% of those patients demonstrated loss of hip motion and difficulty with sitting. In all patients, the transferred muscle did not function as a strong abductor. In their series, 11% experienced a worsening of their neurologic deficit. They also report a high rate of pathologic fractures after surgery that they related to disuse osteoporosis. The authors conclude that other factors, such as level of spinal lesion, lower extremity alignment, and presence of scoliosis and pelvic obliquity, were more important in determining function rather than maintenance of hip reduction (Level IV).

In another study, Duffy and investigators¹⁴ examined gait patterns in 28 children with myelomeningocele using three-dimensional gait analysis to determine whether ambulation was improved by tendon transfers. All of the patients who had undergone posterolateral iliopsoas tendon transfer had concentrically reduced hips at the time of the study. They found no significant difference in range of pelvic obliquity in those patients who had iliopsoas transfer as compared with those who had not. They also reported worse pelvic rotation and significantly worse range of hip abduction/adduction in those who had undergone psoas transfer. The authors conclude that gait was not improved by posterolateral iliopsoas transfer (Level III).

Feiwell and colleagues¹⁵ reviewed a series of 76 patients with myelomeningocele to compare the functional results in those who had undergone surgical treatment to reduce the hip with those who had not had surgical treatment. They found that the presence of a concentric reduction did not lead to improved hip range of motion or ability to ambulate. They also did not find a decrease in need for bracing or decrease in pain with a stable reduction of the hip. Rather, the authors did report a high rate of complications in the surgically treated group including loss of motion (29%) and pathologic fractures (17%). They conclude that the most important factor in determining ability to walk is level of neural involvement and not the status of the hip (Level III).

In a retrospective, comparative review, Sherk and coworkers¹⁶ compared 30 patients with no surgical treatment of hip dislocation with 11 patients who had surgical treatment. In agreement with the previously reported studies, they found the ability to walk was independent of hip reduction and instead depended on neurologic level. In the surgically treated group, four patients (36%) had worsened ambulatory capacity as a result of surgical complications. Of note, the authors report that patients in the group treated without surgery had no difficulty in sitting with either unilateral or bilateral dislocated hips (Level III).

Gabrieli and researchers¹⁷ report on a series of 20 patients with low lumbar myelomeningocele who walked with crutches that underwent three-dimensional gait analysis to determine the influence of unilateral hip dislocation on gait. They report the walking speed of patients with dislocated hips was 60% of normal, a value that corresponds to the walking speed in low-lumbar patients without hip dislocation observed in previous studies from the same center. The authors also note that gait symmetry corresponded to either absence of hip contractures or bilateral symmetric contractures. They found no relation between gait symmetry and hip dislocation, and conclude that there is no indication for surgical relocation of the unilaterally unstable hip. Rather, they recommend correcting unilateral soft-tissue contractures to restore gait symmetry (Level IV).