Distraction-Based Treatment Maintains Predicted Thoracic Dimensions in Early-Onset Scoliosis

Abstract

Study DesignRetrospective cohort study. ObjectiveExamination of distraction-based treatment effect on thoracic dimensions in patients compared to predicted individual normal values, at initial treatment and subsequent follow-up after lengthenings. Summary of Background DataChange in thoracic dimensions and spine length is an important outcome measure in treatment of children with early-onset scoliosis; however, it is difficult to use to make comparisons between patients and the normal population because of the heterogeneous nature of early-onset scoliosis. MethodsEarly-onset scoliosis patients treated with distraction-based therapy who had radiographic parameters (pelvic inlet width, chest width, and thoracic height) preoperatively, immediately postoperatively, and at a minimum 5-year follow-up were included. Individual thoracic measurements were compared with predicted normal measures based on pelvic inlet width, and expressed as a percentile of predicted measure. ResultsComparisons were made in 41 patients; mean age at time of primary surgery was 4.5 years, and median follow-up was 6.5 years. Thoracic height percentile increased from a mean preoperative value of .78 to a postoperative percentile of .88 (p < .001); at long-term follow-up, it was .85. Absolute thoracic height increased at all 3 time points: 141.6, 159.79, and 203.45 mm, respectively Mean chest width was similar preoperatively (170 mm) and immediately postoperatively (166.5 mm) but increased at latest follow-up (206.9 mm). Chest width percentile was similar at all 3 times (.93, .90, and .91). ConclusionsDistraction-based treatment increases absolute thoracic height over time. There is significant improvement in the thoracic height percentile normalized after initial surgery, which was maintained over time. Measuring expected gains as a percentile normalized for pelvic width may be a more relevant outcome measure compared with measuring only absolute values.