Intrathecal Baclofen Pumps do not Accelerate Progression of Scoliosis in Quadriplegic Spastic Cerebral Plasy

Abstract

Introduction Intrathecal baclofen pumps are an effective treatment for spasticity in patients with cerebral palsy. Recently, concerns have been raised regarding the potential for intrathecal baclofen pumps to accelerate scoliosis progression and increase the need for spinal fusion surgery. To date the literature remains unclear, with studies often failing to account for the natural history of scoliosis progression in this population when assessing the influence of baclofen pumps. We aim to clarify matters by comparing progression of scoliosis in quadriplegic spastic cerebral palsy with and without intrathecal baclofen pumps. Material and Methods: Design: Retrospective matched cohort . Patients: Inclusion criteria: Quadripegic spastic cerebral palsy, GMFCS level 5, follow up > 1 year, adequate radiographic/clinical data . Methods: Suitable patients who underwent intrathecal baclofen pump insertion were matched to cases with the same diagnosis and function level with comparable age and baseline Cobb angle but without intrathecal baclofen pumps. Outcomes: Primary: Sequential radiographs measured to assess annual coronal curve progression and peak curve progression . Secondary: Need for spinal fusion for curve progression . Analysis: Between group comparisons made with paired t-tests and chi square tests. Results Baclofen group: 17 patients (7 female), mean age at insertion of pump 9.4 (3.6), 10 cases with a scoliosis > 10° at baseline. Initial Cobb angle 21° (21) and Risser grade 0.8. Twelve cases had hip subluxation/dislocation at baseline. Mean follow up 4.7 (1.3–8.8) years. Mean Cobb angle at final follow up was 77° (27). Non-baclofen group: 17 patients (10 female), mean age at baseline 9.5 (3.7), 12 cases with a scoliosis at baseline. Initial Cobb angle 25° (20) and Risser grade 0.9. Three cases had hip subluxation/dislocation at baseline. Mean follow up 3.6 (1–7.5) years. Mean Cobb angle at final follow up was 67° (30). The two groups were statistically similar in terms of age, initial Cobb angle, Risser grade and presence/absence of scoliosis at baseline. The baclofen group had a higher proportion of hip subluxation/dislocation at baseline ( p = 0.006). Mean annual curve progression was 13.4° (7.2) for the baclofen group vs 13.6° (9.1) for the non-baclofen group ( p = 0.8). Mean peak curve progression was 31.7 (34) for the baclofen group and 27.1 (18) for the non-baclofen group ( p = 0.95). 4 patients with baclofen pumps and 5 without required spinal fusion for curve progression during follow up. All patients, aside one non-baclofen case, had developed a scoliosis by follow up; mainly thoracolumbar/lumbar collapsing C curves. Conclusion Patients with quadriplegic spastic cerebral palsy with and without baclofen pumps showed significant curve progression over time. Intrathecal baclofen pumps do not appear to alter the natural history of curve progression in this population.