Although the age of onset can vary widely from 3 years old to 20 years old, neuromuscular scoliotic curves generally develop at a younger age than adolescent idiopathic scoliosis [7]. Some of the patients present with significant curves prior to the prepubescent growth spurt as a result of earlier curve onset. With the growth spurt, which is typical delayed, the flexible, postural curve tends to develop into a torsional structural deformity. Finally, a stiff curve of considerable magnitude develops before the growth complete.

The single most important factor that affects the magnitude of the curve is the severity of CP. There appears to be proportional relation between the severity of involvement of CP and the curve severity; 67 % of the quadriplegic and 18 % of the non-quadriplegic spastic CP in this study had curves that exceeded 60° [7].

While the rate of curve progression is highly variable, the average progression cited in one report is 0.8° a year in curves less than 50° and 1.4° a year in curves more than 50° [8]. During periods of rapid growth, much more severe progression can occur. Eighty-five percent of patients who had curve of more than 40° by age 15 years progressed to 60°, while only 13 % of those with a curve of less than 40° by age 15 progressed to 60° [7]. Some authors reported curve progression after intrathecal baclofen therapy was instituted to control severe spasticity even in skeletally mature patients [9]. In an adequately powered, case–control study, it was found that the rate of progression was not affected by the use of intrathecal baclofen therapy and that curve incidence and progression were rather related to neurologic involvement [10].

Curve progression increases the magnitude of deforming forces and leads to subsequent deformity, truncal imbalance, and pelvic decompensation. The pelvis is often the end vertebra – the most tilted vertebra with residual axial rotation of the C-curve. This was described as the pelvic vertebra by Dubousset. Less commonly, pelvic obliquity presents as a compensatory fractional curve to the C-curve. Pelvic obliquity alters the sitting position and the pressure at the typically well-distributed sitting tripod at both ischial tuberosities and pubic symphysis. The undue increased pressure at the ipsilateral ischial tuberosity is further exacerbated in patients who have increased pelvic tilt and results in pressure sores.

Depending on the dominant deforming forces and the interplay of spasticity, patients may present with kyphoscoliosis or lordoscoliosis. In kyphoscolosis, progressive deformity with associated pelvic obliquity and retroversion may compromise the often-limited ambulatory function particularly standing to transfer. Significant pelvic obliquity makes sitting adaptation difficult and sometimes impossible. In lordoscoliosis, patients may present with extensor posturing clinically. The progressive deformity renders sitting impossible. Patients may need to be nursed in a semi-reclined position in a wheelchair. This group of patients may present with acute pain that is not alleviated with any sitting adaptation.

Significant spinal deformity is known to compromise cardiopulmonary function, affect gastrointestinal motility, and result in rib-pelvis impingement. The morbidity associated with these are difficult to quantify in this vulnerable and low-demand population especially those with pre-existing difficulties in swallowing, dependence on G or J-tube, and multiple medical comorbidities and are unable to articulate their symptoms.

Given the universal progressive nature of neuromuscular scoliosis, early diagnosis of the deformity is essential. Initial evaluation should consist of clinical monitoring by physical examination. During physical examination, the patient is examined in a sitting position for a curve and pelvic obliquity. When a curve is identified, the crux of the examination is to assess the flexibility of the curve clinically and the remaining growth potential by serial height and weight measurements and radiographic markers.

Curve flexibility is assessed by holding the patient up at the axillary areas in a sitting position. In a smaller framed patient, a clinical fulcrum bend test over the examiner’s knee is possible. Pelvic obliquity is assessed by lying the patient in a prone position with the hips and knees hanging free. Infrapelvic causes of pelvic obliquity such as hip subluxation/dislocation or adductors contracture are evaluated and managed appropriately. Suprapelvic causes of pelvic obliquity arise from the scoliosis and are assessed clinically for flexibility and reducibility.

When a significant curve is identified, standing position (when possible) 36-in. posterior–anterior (PA) and lateral radiographs of the spine should be obtained. Radiographs in sitting position may be obtained if the patient is unable to stand; it may be necessary to support the head and trunk in severely affected children with poor truncal control. At our center, we use a standardized sitting frame with lateral support straps to obtain films in the sitting position with minimal external support.

Radiographically, the curve characteristics (curve type, magnitude, and progression), spinal balance (sagittal and coronal), pelvic balance (pelvic obliquity and tilt), and the growth remaining indicators (status of the tri-radiate cartilage and Risser sign) are documented (see Fig. 12.2a, b). Vertebral rotation with rib deformity and wedging suggest that the deformity is structural rather than positional. Among the various techniques of pelvic obliquity measurement, horizontal pelvic obliquity has the least intra-observer and inter-observer variability [11]. The patient with established scoliosis due to CP requires at least yearly follow-up examination to assess curve progression, but with severe curves or during periods of rapid growth, biannual follow-up is desirable.

A magnetic resonance imaging (MRI) should be obtained if there is any suspicion of intraspinal pathology, such as very rapid progression at a young age, increasing lumbar hyperlordosis, or a change in neurologic status, which could be harbingers of a tethered cord.

In the global planning of disease management, several factors need to be considered. Of paramount importance are the alleviation of pain, preservation of function, and facilitation of daily care. Non-operative management of patients with neuromuscular spinal deformities should be directed at maximizing sitting ability and postural control to facilitate interaction with the surrounding environment. A normalized eyebrow-chin angle allows visual and cognitive stimulations with motor response.

Initial close observation of curves that are 20° or less is reasonable; if progression occurs, initial intervention with a brace may be an option. The role of bracing in CP scoliosis is dependent on the severity of the curve and neurological involvement. In a patient with spastic quadriplegic CP, it is generally accepted that bracing is ineffective but may slow the rate of progression. Miller et al. found no impact of a rigid thoracolumbosacral orthosis (TLSO) on scoliosis curve, shape, or rate of progression in spastic quadriplegic patients that were braced 23 h per day over a mean period of 67 months compared to a similar cohort that were not braced and were followed to spinal fusion [12]. Terjesen et al. [13] retrospectively examined a cohort of 86 patients with spastic quadriplegic CP and found a mean rate of progression per year of 4.2° with a custom-molded polypropylene TLSO. Interestingly, 25 % of the patients had no progression or progression of less than 1° per year. The degree of curve correction in the orthosis appeared to correlate with non-progression of the curve. Of note, Terjesen et al.’s study had a mean initial Cobb angle of 68.4°.

Although it may not alter the final disposition, a soft (polypropylene foam) TLSO can provide seating support and augment function. Improved sitting in a child may correlate with attentiveness in class, ease of care, improved self-image, and decreased rate of decubitus ulcers.

Another option for patients with flexible curves in need of seating support is the adjustment of offset lateral chest supports and modular seating systems on the wheelchair. This three-point control of the coronal deformity will prop up the child and address sitting balance. The wheelchair should be the primary seating device. In an ambulatory patient (GMFCS level I-II), it is believed that a hard brace may slow the progression of the curve similar to the patients with adolescent idiopathic scoliosis. The brace is indicated beyond 25° in immature patients with significant growth remaining. The brace should be worn for a minimum of 12 h. An optimal brace time is 16–18 h per day. Therapeutic stretching, electrical stimulation, or botulinum toxin is lacking scientific validity and should have no role in the management of deformity.

Goldberg argued that the potential gains from interventions should be assessed by the following components: functional health gain, patient satisfaction, and technical success [14]. Given the wide spectrum of disease presentation and progression as well as the concomitant variability in functional status of the patient, the decision to proceed with operative correction and stabilization is based, in large part, on patient-specific factors with the broad aim of maintaining the functional health against the progressive deformity and its associated morbidity, achieving reasonable patient/caregiver satisfaction, and minimizing the complications associated with the surgical intervention.

For higher functioning patients, operative intervention aims to provide a more normal spinal balance and alter the progression of disease with the goal to preserve function with respect to ambulatory potential. The parents or caretakers can make an informed decision weighing the risks and benefits for their children.

For patients with no ambulatory potential (GMFCS 5), the aim is to maintain independence in sitting and facilitate care. As expected, the burden of care in this group of patients with severe learning disability may change significantly [15, 16]. As observed by Madigan and Wallace [4], the severity of scoliosis is directly proportional to the severity of involvement of CP. Concern has been raised regarding the risk of an extensive surgical procedure in a medically compromised patient. Surgical treatment in this group represents a palliative measure that allows the family to provide maximal medical treatment with the intent of caring for the child at home and keeping the child involved in school and other outside, community activities.

A prospective study by Larsson et al. in a cohort of neuromuscular scoliosis with a varied spectrum of learning disabilities found that the overall care burden decreased with improved sitting position and lung function (vital capacity) on follow-up [15]. Comstock et al. [17] assessed both patient and caregiver satisfaction in a cohort of 100 patients with total-body-involvement spastic CP who underwent spinal fusion. The satisfaction of both caregivers and patients was assessed via interview responses to standardized questions, and physical examination was used to assess functional status. Eighty-five percent of the parents interviewed indicated that they were satisfied with the results and would repeat the surgery again. There was an impression by caregivers that the patients had an improved self-image, and patients who were able to respond to questions confirmed this. Both parents and caregivers felt that the surgery had a positive impact on the patient’s sitting ability, physical appearance, comfort, and ease of care. Multiple authors including Bulman et al. [18], Sussman et al. [19], and Watanabe et al. [20] found similar satisfaction rates in their studies.

In general, surgical intervention is considered for curve magnitude greater than 50° with significant deterioration in function [17, 22, 27]. There is sufficient evidence that these curves will progress, even if the child has completed his growth. For curves 60–90°, surgery is indicated when the deformity becomes stiff by physical examination, even if substantial growth remains. If the spine displays continued flexibility on physical examination during growth, surgery can be delayed until 90° and can still be performed with a posterior-only procedure. In a flexible curve of greater than 90°, sitting may be a challenge and further exacerbated by the associated pelvic obliquity.

In planning for surgery, specific considerations should be given to the level of instrumentation, early-onset scoliosis, sagittal plane deformity correction, pelvic and infrapelvic coronal deformities, intraoperative neuromonitoring, the necessity of anterior release, intraoperative femoral traction, and intrathecal baclofen therapy.

Spinal instrumentation and fusion are indicated for collapsing deformities and painful sitting when no other alternatives exist [43]. Historically, fusions with Harrington instrumentation had an unacceptably high rate of pseudarthrosis in 18–27 % of cases [5, 17, 44, 45]. The advent of segmental instrumentation with Luque rod and sublaminar wiring yielded improved results over the Harrington system [19, 21, 46–48] and obviated the need for prolonged postoperative casting. Comstock et al. [17] found a mean correction of 51 % in a posterior-only instrumentation cohort 57 % and in an anterior–posterior cohort.

Multiple authors have noted progression of pelvic obliquity if the fusion was not extended to the pelvis [17, 19, 23]. The Galveston technique to extend the fusion across the pelvis by placing each Luque rod between the pelvic tables [49] demonstrated acceptable fusion rates across the L5–S1 segment and provided good control of pelvic obliquity. It was associated with a high incidence of loosening secondary to micromotion at the ilium at the sacroiliac joints, which was described radiographically as the “windshield-wiper” effect. While the impaction of two Luque rods into the pelvis with associated segmental fusion via sublaminar wires provides a strong construct in the sagittal plane, there exists a moment arm of rotation about the two rods allowing for rod translation with respect to one another, loss of torsional control, and subsequent progression of pelvic obliquity, pseudarthrosis, and implant failure [50]. The use of Luque rods smaller than one-fourth-inch diameter may increase the incidence of implant failure [21, 23, 51], but the intraoperative bending of one-fourth-inch diameter steel rods to the optimal geometry for pelvic implantation presents a technical challenge. Lonstein et al. found in a cohort of 93 patients a 50 % correction of the major scoliotic curve with a mean preoperative scoliosis of 72° and 40 % correction of pelvic obliquity at a mean follow-up of 3.8 years using a dual Luque-Galverston instrumentation technique [52]. With a similar construct, Sanders et al. [23] found that postoperative residual curve greater than 35°, preoperative curves greater than 60°, crankshaft deformity, and not fusing to the pelvis were the factors associated with postoperative curve progression. It is clear that rigid fixation is essential for surgical success.