Serial derotational casting in idiopathic and non-idiopathic progressive early-onset scoliosis

Introduction Serial derotational casting has been used as a definitive treatment or as delaying strategy in progressive idiopathic (IS) and non-idiopathic (NIS) early-onset scoliosis (EOS). Methods Retrospective chart and radiographic review of patients who underwent serial casting for progressive EOS between 2005 and 2012 at a single institution. Results A total of 74 consecutive patients entered serial cast treatment. Twenty-eight were currently being casted, 30 completed cast treatment and were converted to thoracolumbosacral orthosis (TLSO), 9 were treated surgically, 6 were lost to follow-up, and 1 had no further treatment. The researchers diagnosed IS in 41 patients; 33 had NIS. At presentation the IS group had an average Cobb angle (CA) of 49° and a rib vertebral angle difference (RVAD) of 37°. The NIS group had a CA of 51° (p =.69) and RVAD of 37° (p =.94). In patients currently being casted, 19 IS patients had a decreased CA, from 47° to 27°. The 9 NIS patients had a decreased CA, from 62° to 57° (p =.0002). Cobb angle improvement was significantly better in IS (p =.0005). In the TLSO group the 17 IS patients had a decreased average CA, from 46° to 18°, after serial casting and the 13 NIS patients decreased CA from 42° to 32°. Patients with IS had better improvement in CA than the NIS group (p <.001). At last follow-up, this was reduced to 11° in the IS group and maintained at 32° in the NIS. In the IS group, 5 of 41 patients were converted to growth constructs, and 4 of 26 in the NIS group. Casting initiated before age 2 years yielded better curve correction for IS (p <.01) compared with NIS. Conclusions Progressive idiopathic scoliosis patients had better curve correction with casting than NIS patients. Casting in IS patients before age 24 months yielded better curve correction. Patients who required surgery had a higher age and Cobb angle at presentation than those who transitioned to a TLSO. The surgical group was observed for a similar duration of time and there was no significant statistical difference. Although RVAD is a predictor of progression in infantile IS, it did not show a predictive value in the response to casting of either the IS or NIS groups.