Computed Tomography Lung Volume Changes After Surgical Treatment for Early-Onset Scoliosis

Study DesignSingle-center review of prospectively collected data.ObjectivesTo document anatomic lung volume and thoracic parameter changes in early-onset scoliosis patients undergoing rib-based (RB), or spine-based (SB) distraction surgical treatment who were too young to perform pulmonary function tests.MethodsTwenty patients undergoing growth-sparing treatment had computed tomography lung volumes (CTvol) determined by controlled-ventilation CT scanning preoperatively and at a mean of 2.7 years later under an institutional review board (IRB)-approved protocol. Twelve patients had non-congenital curves and 8 had congenital curves. Eleven patients had SB constructs and 9 had RB ones. Computed tomography lung volumes were correlated to T1–12 length, T6 coronal width, pelvic width, and curve magnitude, and were normalized by comparison with age standards and pelvic width.ResultsAll patients had increased CTvol at follow-up (RB 51%, SB 46%; p < .001). All increased T1–12 length from 128 mm (range, 39–160 mm) preoperatively to 154 mm (range, 61–216 mm) at follow-up. Both RB and SB gained 2.6 cm; this measurement was significant in RB (p < .001) owing to the shorter preoperative length. The T1–12 length correlated well with CTvol preoperatively (p = .002) and at follow-up (p = .007). The T6 width correlated best with CTvol (r = 0.76; p < .001 preoperatively and at follow-up). Main thoracic curves improved 21° in SB (preoperatively, 78°) versus 1.5° correction in RB (preoperatively, 60.2°). There was no correlation between curve magnitude and CTvol preoperatively or at follow-up. Follow-up CTvol percentile decreased in 10 patients, increased in 6, and was unchanged in 4. The T1–12 length was less than the fifth percentile in all patients preoperatively and increased in 9 patients at follow-up, whereas 11 remained at less than the fifth percentile.ConclusionsThe CTvol quantitates anatomic results of early-onset scoliosis growth-sparing surgery in patients too young for standard pulmonary function tests. Thoracic length and width correlate well with absolute CTvol and are possible surrogate measures. Curve magnitude and correction correlate poorly and assume less importance in outcome evaluation. Thoracic volume and length gains exceeded normal growth in about half of the patients.