25.1.1 Pseudarthrosis

Biologic pseudarthrosis in the spine is a lack of bony fusion with formation of a false joint. A pseudarthrosis can result from several causes including insufficient stability or infection and often presents in the form of pain, deformity progression, or failed instrumentation. If there is no evidence of implant failure or obvious lucency in the fusion mass, workup often proceeds to computed tomography (CT) scan for more definitive diagnosis. Magnetic resonance imaging (MRI) can be helpful to assess the central canal and neuroforamina.

Unfortunately, a pseudarthrosis can occur in these patients even after a technically well-performed spinal surgery. The lever arm is often long in the neuromuscular patient and places a tremendous amount of stress at the end of fusion, especially at the lumbosacral junction. Prevention lies in being meticulous in surgery with facetectomies, use of robust bone graft, and stable fixation. Overall incidence of pseudarthrosis after spinal deformity surgery in the neuromuscular population is 1.88% with higher rates in myelomeningocele (12.63%) compared to cerebral palsy (0.05%) and Duchenne’s muscular dystrophy (2.97%). ⁶ In cerebral palsy, the incidence has declined over the last 50 years with the advancement of spinal implants. Pseudarthrosis rates from 11 to 40% were reported with use of Harrington rods and improved to 13% with Luque instrumentation. More recently Tsirikos et al reported on 45 consecutive patients with cerebral palsy who underwent posterior spinal fusion with a pedicle screw construct and reported no cases of pseudarthrosis. ⁸

Despite its relatively low incidence, pseudarthrosis is the most common reason for revision surgery in the neuromuscular population outside the immediate postoperative period. ⁹ Although not all pseudarthroses necessitate revision surgery, in the setting of pain or implant failure with loss of correction, surgical revision may be required. Infection should be ruled out as the cause of the pseudarthrosis prior to planning reoperation. Revision surgery for pseudarthrosis typically involves assessment and debridement of the pseudarthrosis with deformity correction and rigid stabilization of the segment.

Fortunately, if appropriately managed, pseudarthrosis can be treated successfully. Dias et al reported on four children with cerebral palsy who underwent revision spine surgery at their intuition for symptomatic pseudarthrosis with progressive deformity and implant failure. ¹⁰ Resolution of symptoms and correction of deformity were successfully achieved in three patients. The fourth child underwent revision with takedown and bone grafting of the pseudarthrosis without instrumentation, resulting in persistent and symptomatic pseudarthrosis. They concluded that rigid instrumentation combined with pseudarthrosis debridement and bone grafting is imperative for successful deformity correction and resolution of symptoms.

Yagi et al had similar findings on 50 cases of pediatric revision spine surgeries, of which 13 involved neuromuscular scoliosis. ¹¹ Nine of 13 patients (69%) underwent revision for pain or progressive deformity from pseudarthrosis. Eight patients had resolution of symptoms after undergoing revision and one patient had residual symptoms with recurrent pseudarthrosis, which was treated successfully with a second revision. This is consistent with previous reports in that pseudarthrosis is the most common postoperative complication requiring revision in the neuromuscular scoliosis population, and with clear indications, bone grafting, osteotomies if necessary, and appropriate instrumentation, the risk of pseudarthrosis recurrence is minimized and successful outcome is likely.

25.1.2 Biomechanical Failure

Other sources of increased stress on the bone–implant interface and potential biomechanical failure beside pseudarthrosis include obesity, poor bone quality, and significant preoperative spinal deformity, especially in the sagittal plane and at the lumbopelvic junction. The larger the curve, the stiffer the curve, and the softer the bone, the more important the bone–implant interface. Sink et al found a high rate of proximal and distal instrumentation pullout and failure in the management of cerebral palsy spinal deformity using the Luque–Galveston instrumentation and identified hyperkyphosis as a significant risk factor for implant failure. ¹² With a 54% failure rate, they concluded that the Luque–Galveston instrumentation was not ideal given the significant deforming forces in hyperkyphotic spinal deformities.

Although once the most popular instrumentation for neuromuscular spinal deformities, the Luque–Galveston instrumentation has fallen out of favor in recent years due to biomechanical advantages of newer screw-based constructs. ¹³ This is especially true in treating the significant pelvic obliquity often found in neuromuscular scoliosis. Options have expanded significantly including iliosacral screw fixation, iliac screw fixation, and sacral–alar–iliac (SAI) screw fixation. Although these new techniques have been shown to be powerful in the correction and maintenance of pelvic obliquity, they also have introduced new modes of failure.

Iliac screws became popular with proponents citing diminished implant failure and lower pseudarthrosis rates; however, complications related to rod disengagement from screws and connectors were reported. ^{14 ,​ 15} More recently, S2-iliac screw fixation has been gaining popularity as a powerful technique to control pelvic obliquity; however, the course of S2-iliac screws crosses the sacroiliac joint in the majority of cases and the long-term clinical significance of this is unknown. ^{16 ,​ 17}

Although uncommon, revision for failed pelvic fixation in neuromuscular scoliosis is generally reserved for symptoms or significant loss of deformity correction. In the case of implant failure in Galveston rods, revision to modern screw-based constructs is commonly used if fusion is not present. Revision of screw-based constructs commonly consists of additional or longer screw placement with deformity correction and debridement and bone grafting of pseudarthrosis, if present. Longer screws have been shown to improve implant stability if they reach anterior to the caudal extension of the middle osteoligamentous column. ¹⁸

Sponseller et al compared 32 patients who underwent S2-iliac fixation to 27 patients who had traditional sacral or iliac screws in patients with significant pelvic obliquity from cerebral palsy. ¹⁶ Revision rates were comparable as each group had one revision for implant failure. One patient in the S2-alar group required revision for sacral joint pain, which improved with longer screw placement, and one patient in the traditional group required revision for failure of fixation and pain at implant site.

Myung et al retrospectively looked at 41 patients with neuromuscular scoliosis who underwent posterior spinal fusion to the pelvis with iliac screws in 31 patients and S2-iliac screws in 10 patients. ¹⁹ They reported an overall implant complication rate of 29% with 9 occurring in the iliac screw group and 1 in the S2-iliac screw group. Despite this, only 2 patients required revision surgery, both in the iliac screw group for failed pelvic anchors. They noted that no failure occurred in patients in whom there were 6 or more screws in L5 and below. They concluded that more robust distal pelvic anchorage was protective against implant failure and this was easier to achieve with S2-iliac screws.

In neuromuscular spinal deformity, there are considerable forces the implant must withstand to prevent complications and achieve a successfully outcome. Each technique has cited advantages and disadvantages, and understanding several is necessary to manage a revision surgery for failed spinal or sacropelvic instrumentation.