The Association of Scoliosis Properties with Spinal Cord Tethering: A Statistical Model for Prognostication

ObjectiveTo evaluate the relationship between the structural measures of scoliosis and underlying spinal cord tethering (SCT) and proposing a statistical prognostication model. Study designCross-sectional. SettingAcademic healthcare center Methods128 definite scoliosis cases that were candidates for corrective surgery were enrolled. Anterior-posterior whole column digital radiographs and whole-spine MRI (supine for all samples and adjuvant prone MRI for suspected cases with tight filum terminal) were performed. Univariate and multiple logistic regression were used for the analysis of association and interaction. Association of SCT with structural features of scoliosis -Cobb angle, convexity, and type (idiopathic and congenital)- age, and sex were assessed. ResultsNone of the study variables showed a statistical association with SCT in univariable and multiple logistic regressions. After inclusion of Cobb angle-convexity-type interaction, higher Cobb angle, idiopathic scoliosis, dextrosoliosis, and male gender had a significant effect. Stratification for convexity discovered a positive association of Cobb angle and SCT in idiopathic patients with dextroscoliosis (1.02 [1.01-1.03], 0.049). In contrast, in congenital cases, the rate of SCT decreased by higher left-sided Cobb angles but it was not statistically significant (0.94 [0.88-1.01], 0.104). ConclusionThe risk of spinal cord tethering was not zero in any of the subgroups and no SCT-free group could be detected. Conventional MRI should be preoperatively performed for every case of scoliosis and thoroughly examined for signs of tethering. Clear imaging of patients at higher risk of SCT should not be decisive and further workup should be utilized before proceeding with reconstructive surgery.