30 The Oxford Experience
Spinal deformity frequently arises in growing children with neuromuscular conditions, such as cerebral palsy, myelodysplasia, spinal muscular atrophy, Friedreich ataxia, Duchenne muscular dystrophy, traumatic paralysis, and, in the past, poliomyelitis. The incidence of deformity varies directly with the severity of neuromuscular disease and ranges from 25% in children with diplegic cerebral palsy to 100% in those with thoracic myelodysplasia or traumatic paralysis who are younger than 10 years of age. Children classified under the Gross Motor Function Classification System (GMFCS) as level 4 or 5—those who require wheeled mobility or lack head and trunk control—are most frequently found to have spinal deformity. In a 2011 review of the pediatric and young adult spinal deformity service of Scotland’s National Health Service, it was estimated that 44 patients with neuromuscular scoliosis would require corrective surgery yearly, extrapolated to an annual incidence of 0.07% of the pediatric population undergoing scoliosis correction. 1 A national database study from the United States reported an incidence of neuromuscular spinal deformity of 2.5% in a cohort of nearly 18,000 patients admitted for scoliosis surgery. 2
Deformity is often characterized by kyphosis and pelvic obliquity, rapid progression at a young age because of the early onset of the conditions, subsequent curve rigidity, and limited patient mobility. There is a pattern of long, c-shaped thoracolumbar curves related to the early onset of deformity or of lumbar curves related to a later onset. The risk for progression of deformity is extremely high because of neuromuscular imbalance and is modified by growth. Natural history studies have reported ranges of curve progression from 7 to 40 degrees. 3 , 4 , 5 A Japanese study of the natural history of untreated scoliosis in patients with spastic cerebral palsy found that risk factors for progression included a curve with a magnitude of 40 degrees before the age of 15 years, total-body-involvement cerebral palsy, bedridden status, and a thoracolumbar curve. 6 Other factors related to curve progression include early onset and evolution of neuromuscular disease, skeletal immaturity, and severe weakness. Peak skeletal growth is associated with the greatest progression of deformity, as in idiopathic scoliosis. Unlike the deformity in idiopathic scoliosis, however, a neuromuscular deformity is likely to progress after skeletal maturity.
30.1 Management and Complications
Nonoperative management of neuromuscular spinal deformity is indicated as soon as deformity is detected in an attempt to prevent progression. Braces may be used to slow curve progression and improve sitting posture in skeletally immature patients with flexible curves. 7 , 8 When a brace is chosen, it is imperative to consider the neuromuscular patient’s poor motor control, skin sensation, respiratory function, and nutritional status. Because these patients are often mobilized in manually operated or power-assisted chairs, moulded or modular seating within the chair is indicated. The seat and chair are frequently reassessed for modification as the spinal deformity progresses.
Although the traditional indication for operative management in neuromuscular scoliosis is deformity greater than 50 degrees with rapid progression, it is much more important to consider how the patient’s overall function, or lack thereof, is affected by progressive deformity. Indeed, Cobb angles become increasingly difficult to measure in weight-bearing films as the magnitude of deformity increases and some curves collapse. It is critical to consider factors such as comfortable sitting with independence and balance, positioning for upper extremity use, and cardiopulmonary compromise. Pain and discomfort secondary to spinal deformity are also factors that persuade families to proceed to surgery. For some, cosmesis is an issue that encourages surgical correction of deformity.
The means to achieve spinal fusion in neuromuscular deformity have evolved over recent decades. Segmental spinal instrumentation with pedicle screws from the upper thoracic spine to the pelvis appears to be the currently accepted method of fixation and has replaced traditional alternatives, such as Harrington rods, Luque–Galveston instrumentation, and Unit rod instrumentation. The outcomes of these fusion techniques are discussed below, with an emphasis on complications rather than the benefits of specific instrumentation systems.
Lonstein and Akbarnia reported on a series of 107 patients who had neuromuscular deformities treated either with cast correction and posterior fusion followed by prolonged supine immobilization, posterior fusion, and Harrington instrumentation, or with combined anterior and posterior fusion and instrumentation followed by short supine immobilization. 9 Functional outcome was worse in one patient, unchanged in 82 patients, and improved in 24 patients. The complication rate was exceedingly high, at 81%; the most frequent complications included pressure sores, wounds, and instrumentation problems. The pseudarthrosis rate was 17%, and the infection rate was 5%. There were three deaths, and one patient was rendered paraplegic. Despite these outcomes, the authors concluded that surgery was still of benefit in these patients.
Comstock et al. reported the results of a series of 79 patients with total-body-involvement spastic cerebral palsy who underwent posterior Luque instrumentation, anterior spinal fusion, or both. 10 Late progression of deformity, pelvic obliquity, or decompensation occurred in more than 30% of the patients, and 21% of these patients required revision procedures to address the progression. Most of the patients were skeletally immature at the time of surgery and underwent posterior fusion only, whereas those without progression underwent combined anterior and posterior fusion. The results of questionnaires administered to caregivers and parents revealed that 85% were satisfied with results of surgery, including improvements in sitting ability, physical appearance, ease of care, and comfort.
Tsirikos et al retrospectively reviewed 241 patients with cerebral palsy who underwent posterior or combined anterior and posterior fusion with Unit rod instrumentation at a single institution performed by one of two surgeons. 11 Satisfactory corrections of coronal deformity and pelvic obliquity were achieved. In posterior-only procedures, the average blood loss was 2.8 L, the intensive care unit (ICU) stay was 4.9 days, and the hospital stay was 19.6 days. In combined procedures, the average blood loss was 3.4 L, the ICU stay was 6.7 days, and the hospital stay was 24.5 days. Major complications included three perioperative deaths and 18 deep wound infections (7.5%). A survey of caretakers reported a 96% satisfaction rate.
Lonstein et al reported an early complication rate of 58% in a series of 93 patients with cerebral palsy who underwent spinal fusion for neuromuscular deformity with Luque–Galveston instrumentation. 12 The infection rate was 1.1%, and the pseudarthosis rate was 7.5%. The length of stay was statistically significantly increased, from 7 to 9 days, in patients who experienced one complication.
A review of the Scoliosis Research Society morbidity and mortality database by Reames et al revealed 1,971 complications in 19,360 patients with pediatric scoliosis of all etiologies. 13 The complication rate for the correction of neuromuscular scoliosis was significantly higher, at 17.9%, than the complication rate for the correction of scoliosis of all other etiologies. The rate of neurologic deficit in this population was 1.1%, and mortality was 0.3%.
Piazzolla et al reported the results of Cotrel–Dubousset instrumentation and posterior spinal fusion in 24 consecutive patients with neuromuscular scoliosis. 14 Satisfactory corrections of coronal deformity, sagittal imbalance, and pelvic obliquity were achieved. The mean intraoperative blood loss was 2.1 L, major complications affected 8.3% of patients, and one postoperative death occurred.
Modi et al reported the results of all pedicle screw construct posterior spinal fusion procedures in a series of 18 patients with paralytic neuromuscular scoliosis and poor pulmonary function, defined as preoperative forced vital capacity (FVC) of less than 30%. 15 Satisfactory corrections of coronal scoliosis and pelvic obliquity were achieved. There was a subtle decrease in the FVC at 6 weeks postoperatively and a significant decrease at the mean final follow-up of 32 months (from 25.2% preoperatively to 20.6% at follow-up). The forced expiratory volume in 1 second (FEV1) also significantly decreased in a similar manner. The perioperative complication rate was 44.4%, including five pulmonary complications and one intraoperative death. Granata et al reported similar results in a series of 30 patients with Duchenne muscular dystrophy. 16 Head control was lost in 14 patients who developed severe extension contractures of the neck. There was one death due to cardiac arrest. The mean preoperative FVC was 57%, and this decreased to 34% at 4-year follow-up. Still, more than 90% of parents reported that they would again give their consent for operative intervention.
30.2 Intraoperative Blood Loss
Evidence on the application of blood-sparing techniques in spine surgery has historically been derived from the field of cardiac surgery, with the bulk of the literature considering hemodynamic methods, such as hypotensive anesthesia and planned autologous donation. 17 However, the efficacy of these methods has been reported rarely in spinal surgery, and not at all in pediatric spinal deformity surgery. An extensive review of the literature on blood loss in pediatric spine surgery revealed significantly higher rates of blood loss in neuromuscular scoliosis correction than in adolescent idiopathic scoliosis correction. 18 Mean ranges of blood loss were 1,300 to 2,200 mL in patients with cerebral palsy treated with posterior approaches, 2,500 to 4,000 mL in patients with Duchenne muscular dystrophy, and 2,000 to 3,500 mL in studies that pooled data for all patients with neuromuscular disorders. Blood loss was also shown to be progressively greater in patients having larger numbers of vertebral levels incorporated into the fusion, having posterior fusions vs. anterior fusions, and having both anterior and posterior fusions.
Modern blood-sparing techniques in spinal surgery have been increasingly applied to pediatric spinal deformity surgery. The use of antifibrinolytics was recently reported in 12% of 1,547 neuromuscular scoliosis correction procedures over nearly 4 years at 37 U.S. children’s hospitals. 19 However, the median hospital-specific rate of red cell transfusions was 43%, and antifibrinolytic use was not associated with a decrease in the odds of a red cell transfusion requirement. Nevertheless, a matched-cohort therapeutic comparative study reported by Dhawale et al demonstrated significantly less intraoperative blood loss in a group of patients with scoliosis secondary to cerebral palsy who received antifibrinolytics and consequently less cell salvage transfusion. 20 Although no significant differences were found in total transfusion requirements, a trend toward decreased hospital stay was found in the group that received antifibrinolytics.
As demonstrated above, a high complication rate for neuromuscular scoliosis correction has been reported; the range in rates is likely attributable to how complication is defined in each study. Nevertheless, the rates of infection, pseudarthrosis, late progression of deformity, intraoperative blood loss, and perioperative mortality are far higher than what is considered acceptable for an intervention that is meant to improve quality of life. Furthermore, the finding of further compromise in respiratory function postoperatively is interesting and contradictory to the classic indication of deformity correction to preserve cardiopulmonary function. Although the reported rate of satisfaction with scoliosis correction among caregivers and parents is high, presumably an inverse relationship between satisfaction rate and complication rate exists.
30.3 Patient Comorbidity
The typical patient with spastic cerebral palsy has a host of comorbidities involving multiple organ systems. 21 Neurologic conditions often include epilepsy, visual and hearing impairments, cognitive dysfunction, and sleeping and pain disorders. Patients often have gastrointestinal motility dysfunction leading to gastroesophageal reflux, delayed gastric emptying, and constipation. Oropharyngeal dysphagia makes oral feeding difficult, and impaired swallowing is associated with a risk for aspiration, so that many children are fed via an endoscopically placed percutaneous gastrostomy tube. Nutrition and growth are impaired by decreased caloric intake and must be monitored by a medical practitioner or nutritionist. Chronic pulmonary conditions include recurrent pneumonia, atelectasis, bronchiectasis, and restrictive lung disease. Frequent urologic conditions include voiding dysfunction, retention, incontinence, and urinary tract infection. These multiple medical comorbidities have obvious effects on life expectancy and must be appropriately addressed preoperatively in the patient with neuromuscular scoliosis. The studies summarized below have explored these issues in greater detail.
Tsirikos et al studied the life expectancy of 288 patients with severe spastic cerebral palsy after neuromuscular scoliosis correction and found the mean predicted survival after surgery to be 11 years. 22 Factors negatively affecting the survival rate were a high number of days in intensive care postoperatively and thoracic hyperkyphosis. Factors that did not affect survival included sex, age at surgery, level of ambulation, cognitive ability, degree of coronal plane spinal deformity, amount of intraoperative blood loss, surgical time, and number of days in the hospital. Cardiopulmonary function and postoperative complications apparently were not studied as risk factors.
Erickson and Baulesh conducted a literature review on the topic of preoperative optimization by identifying relevant comorbidities and placing patients with neuromuscular scoliosis on appropriate care pathways. 23 Risk factors associated with perioperative and postoperative complications included the following: seizure disorders, decreased cognitive ability, poor pulmonary status, restrictive lung disease, history of frequent pneumonias, sleep apnea, malnutrition, cardiac disease, immunocompromise, low social status, poor ambulatory status, and a complex procedure. Along the same theme, Miller et al analyzed a treatment protocol emphasizing perioperative work-up to compare surgical outcomes before and after its implementation. 24 Significant reductions in the overall length of stay, ICU length of stay, and perioperative complication rate were observed in the patient group managed according to the protocol.
Almenrader and Patel conducted a study to determine risk factors for postoperative ventilation in patients with nonidiopathic scoliosis. 25 The rate of postoperative ventilation was 23.8%, with postoperative ventilation having been planned preoperatively for half of the patients, and 40% of the patients with Duchenne muscular dystrophy required postoperative ventilation. An increased tendency for children with Duchenne muscular dystrophy and those with a preoperative FVC below 30% to require postoperative ventilation was observed. Abu-Kishk et al performed a similar study to determine possible associations with prolonged mechanical ventilation postoperatively and found an odds ratio of 31.25 when postoperative mechanical ventilation was correlated with neuromuscular scoliosis correction. 26
It is critical to discuss the factors mentioned here with patients’ families during the preoperative assessment because the perioperative expectations must be guided realistically. Multidisciplinary team involvement is accepted as the standard of care for these complex patients in order to anticipate potential complications and ensure that the patient is in an optimal physiologic state before undergoing surgery.
30.4 Outcomes Measurement
The assessment of surgical outcomes in patients with neuromuscular scoliosis has traditionally been dominated by radiographic measures of curve correction and pelvic obliquity, or the documentation of postoperative complications. Recently, interest has been shown in studying the effect of deformity correction on quality of life; however, no accepted, reliable, and validated outcomes instrument is available to capture this entity. A systematic review of measures of activities of daily living for children with cerebral palsy revealed the Pediatric Evaluation of Disability Inventory (PEDI) to have the strongest psychometric properties, and the Assessment of Motor and Process Skills (AMPS) to be the most comprehensive evaluation of underlying motor and cognitive abilities. 27 However, these tools have not been used to assess outcomes of neuromuscular deformity correction, perhaps in part because further research is required to determine their reliability and stability in this particular patient population. Bagó et al published a literature review of health-related quality of life and disease-specific instruments used to assess scoliosis surgery. 28 The SF-36 (Short Form [36] Health Survey) and the EuroQol EQ-5D have been used to assess generic health-related quality of life, and the SRS-22 (Scoliosis Research Society-22) Patient Questionnaire and QLPSD (Quality of Life Profile for Spinal Deformities) have been used as disease-specific instruments. These have been complemented more recently with the use of instruments to assess a single aspect of spinal deformity, such as perception of trunk deformity and body image or effects of brace use. However, these tools have been used and partially validated only in patients with adolescent idiopathic scoliosis, and none has been sufficiently validated and analyzed in the population of patients with nonidiopathic scoliosis.
Ersberg and Gerdhem used EQ-5D and SRS-22r outcome instruments to report the effect of spinal fusion for neuromuscular scoliosis in 32 patients. 29 At 2-year follow-up, both tools demonstrated statistically significant improvements (EQ-5D score increased 0.15 points, SRS-22r score increased 0.5 points) in quality of life for the cohort. To our knowledge, this is the only published paper that has used these robust measures for quality of life in this patient population. However, it is imperative to note that both the EQ-5D and SRS-22-r indices have not been validated in patients with neuromuscular scoliosis.
Watanabe et al conducted a retrospective study to assess the satisfaction of patients with spastic cerebral palsy and their parents with the results of spinal fusion. 30 Questionnaires were designed to assess expectations, cosmesis, function, patient care, quality of life, pulmonary function, pain, health status, self-image, and satisfaction. Responses were obtained from 84 patients and families with an average follow-up of just over 6 years. The overall satisfaction rate was 92%, and a similar proportion of respondents reported improved sitting balance and cosmesis, whereas 71% reported improved quality of life. Functional improvements were relatively limited, but 8 to 40% still perceived the surgical results as improvement. When the respondents were divided into “satisfied” and “less satisfied” groups, the less satisfied group was observed to have a higher rate of late complications, less major curve correction, and hyperlordosis of the lumbar spine postoperatively. Lumbar hyperlordosis is known to be associated with severe back pain, superior mesenteric artery syndrome, malnourishment, loss of bowel and bladder control, and difficulty with sitting balance. One of the chief reasons for satisfaction was the achievement of expectations; in this case, expectations most commonly were to “prevent progression of deformity” and “prevent cardiopulmonary problems.” This finding is in contrast to those of previous studies, which reported improved appearance and improved sitting balance as common expectations.
Obid et al performed a similar retrospective questionnaire study of 32 patients with an average of 3 years of follow-up. 31 Patients and/or their caregivers strongly agreed with statements that quality of life was improved, they were satisfied with surgery, and that expectations were fulfilled. Cassidy et al used a health care worker questionnaire to assess the impact of spinal fusion on patients with severe cerebral palsy with regard to comfort, function, health, and ease of nursing care. 32 When 17 patients who had scoliosis of 35 degrees and underwent fusion were compared with 20 patients who had scoliosis of 76 degrees and did not undergo fusion, no differences were found in any of the parameters except for comfort; the majority of health care workers believed that the patients with fusion were more comfortable. These findings must be tempered by the lack of reported studies on the correlation of caregiver perceptions of discomfort with patients’ actual perceptions of pain.
Tsirikos et al compared 190 parents’ and 122 professional caretakers’ perceptions of functional outcome after spinal fusion in patients with spastic cerebral palsy. 33 Both groups reported a positive impact of surgery on patients’ overall function, quality of life, and ease of care. Parents appreciated improvement in appearance, whereas educators and therapists acknowledged improvement in gross and oral motor function. Most parents (95.8%) and caretakers (84.3%) said they would recommend spine surgery.
The majority of reports of caregiver satisfaction after spinal fusion for neuromuscular scoliosis have been retrospective in nature, hence introducing bias. Jones et al reported the results of POSNA (Pediatric Orthopaedic Society of North America) outcomes questionnaires prospectively administered to 20 parents preoperatively, 6 months, and 1 year after spinal fusion. 34 No significant changes were found between preoperative and postoperative assessments of physical function, school absence, co-morbidities, and parental health. Patient pain and frequency of feeling sick and tired decreased, patient happiness increased, and parental satisfaction improved significantly by 1 year postoperatively. The presence of complications did not significantly affect results. The authors concluded that subjective gains noted in previous retrospective studies were substantiated by these prospective results.
A standardized, reliable, and validated outcomes instrument for patients with neuromuscular scoliosis would enhance our understanding of the effect of spinal deformity correction. Ideally, such a tool would encompass surgical outcomes and effects on patient quality of life simultaneously, allowing data to be communicated and extrapolated on an international scale.