Serial Casting for Infantile Idiopathic Scoliosis: Radiographic Outcomes and Factors Associated With Response to Treatment.

BACKGROUND: Serial casting for early-onset scoliosis has been shown to improve curve deformity. Our goal was to define clinical and radiographic features that determine response to treatment. METHODS: We retrospectively reviewed patients with idiopathic infantile scoliosis with a minimum of 2-year follow-up. Inclusion criteria were: progressive idiopathic infantile scoliosis and initial casting before 6 years of age. Two groups were analyzed and compared: group 1 (≥10-degree improvement in Cobb angle from baseline) and group 2 (no improvement). RESULTS: Twenty-one patients with an average Cobb angle of 48 degrees (range, 24 to 72 degrees) underwent initial casting at an average age of 2.1 years (range, 0.7 to 5.4 y). Average follow-up was 3.5 years (range, 2 to 6.9 y). Sex, age at initial casting, magnitude of spinal deformity, and curve flexibility (defined as change in Cobb angle from pretreatment to first in-cast radiograph) were not significantly different between groups (P>0.05). Group 1 had a significantly higher body mass index (BMI) than group 2 at the onset of treatment (17.6 vs. 14.8, P<0.05). Univariate analysis of demographic, radiographic, and treatment factors revealed that only BMI was predictive of Cobb improvement (P=0.04; odds ratio=2.38). Group 1 (n=15) had a significantly lower Cobb angle (21 vs. 56 degrees) and rib vertebral angle difference (13 vs. 25 degrees) compared with group 2 at latest follow-up (P<0.05). A significantly larger proportion of children who were casted at less than 1.8 years of age had a Cobb angle <20 degrees at latest follow-up (P=0.03). Group 2 maintained stable clinical and radiograph parameters from pretreatment to most recent follow-up. CONCLUSIONS: To maintain a homogeneous cohort, we excluded patients with syndromes and developmental delays. We believe that analyzing a homogeneous group provides more meaningful results than if we studied a heterogeneous sample. BMI was significantly associated with outcome such that for each unit increase in BMI, there is a 2.38× increase in the chance of improvement. Curve flexibility was similar between groups, which suggest that the amount of correction obtained at initial casting does not confirm treatment success. Key aspects of treatment that may determine success include age of less than 1.8 years at initiation of casting and derotation of the spine to correct rib vertebral angle difference of <20 degrees. LEVEL OF EVIDENCE: Level IV-Therapeutic.