Risk factors of distal adding-on phenomenon after posterior correction of congenital thoracic deformity with long fusion

English Keywords:Congenital Scoliosis  Distal coronal decompensation  Adding-on phenomenon  Long segment  Risk factors

English Abstract:

  【Abstract】 Objectives: To investigate the incidence and risk factors of postoperative distal adding-on phenomenon after posterior correction of congenital thoracic deformity with long fusion. Methods: 78 patients were analyzed who underwent posterior pedicle screw correction with longfusion for congenital thoracic deformity from April 2006 to April 2016 in our hospital. There were 35 males and 43 females with a mean age of 12.6±2.9 years. The average follow-up was 42.4±27.3 months. All patients were categarized into adding-on group and non-adding-on group according to definiton of postoperative adding-on phenomenon. Factors which might cause adding-on phenomenon occurrence were collected: age, gender, Risser sign, body mass index, bone graft, segments, pre- and post-main thoracic cobb, cranial curve, caudal curve, coronal balance distance, pre-post coronal balance distance, radiographic shoulder height(RSH), apical vertebra translation(AVT). Chi-square test, Fisher exact test, t-test, Mann Whitney U test and multi-factor logistic regression analysis were used to investigate the correlation of possible risk factors and adding-on phenomenon; ROC curve analysis was used to determine the cut-off score between factors and adding-on. Results: Among the 78 patients, 7 cases (8.97%) presented with distal adding-on phenomenon. There were significant differences of pre-operative coronal balance distance(-0.3±1.6mm vs 2.3±2.7mm), LIV location(59/12 vs 0/7), post-operative main thoracic Cobb(19.0°±7.8° vs 28.7°±9.5°), post-operative cobb angle of caudal curve(5.9°±3.6° vs 9.1°±7.4°), pre-post coronal balance distance(-0.9±1.6mm vs 3.4±1.7mm), LIV disc angle(1.4°±1.7° vs 3.5°±3.1°) between the two groups in univariable test(P＜0.05). There were no signficant differences between two groups in gender(32/29 vs 3/4), age(12.5±2.8 years vs 13.2±5.2 years), Risser sign(18/53 vs 2/5), BMI(17.8±3.8kg/m2 vs 18.5±5.1kg/m2), segments(7.9±3.4 vs 8.3±3.0), bone graft(40/31 vs 4/3), pre-operative main thoracic Cobb(47.4°±14.0° vs 55.1°±12.0°), pre-operative cranial curve Cobb(16.5°±12.6° vs 24.5°±13.7°), pre-operative caudal curve cobb(19.1°±12.3° vs 26.1°±14.3°), pre-operative RSH(0.4°±1.4° vs 0.0°±1.4°), pre-operative AVT(1.7°±2.7° vs 2.0°±2.7°), Pfirrmann classification(64/7 vs 4/3), post-operative Cobb angle of caudal curve(5.9°±5.0° vs 8.4°±6.3°), post-operative coronal balance distance(0.7±0.9mm vs -1.1±1.2mm), post-operative RSH(0.2°±2.3° vs 0.7°±0.9°), post-operative AVT(1.0°±1.2° vs 2.2°±1.8°), post-operative ratio of major curve/caudal curve(5.8°±6.7° vs 6.8°±12.2°) and major curve correction(28.4°±17.0° vs 34.5°±15.1°, P>0.05). In multivariate logistic regression analysis between each two groups, postoperative LIV disc angle(OR=1.72, P=0.003), LIV cephalad to the SV(OR=2.31, P=0.007), pre-post CBD (OR=1.21, P=0.014) were identified as independent factors positively correlated with distal Adding-on phenomenon. The ROC curve revealed that the optimal cut-off point of LIV disc angle was 8.05°, the area under the ROC curve was 0.821, and the optimal cut-off point of pre-post CBD was 3.45cm, the area under the ROC curve was 0.915. Conclusions: Larger LIV disc angle， larger pre-post CBD, LIV cephalad to the SV are independent factors for postoperative adding-on in patients with congenital thoracic scoliosis undergoing posterior correction with long fusion.