Abstract
Aim:
To determine radiographic variables that predict the need for distal extension of the fusion beyond Cobb-to-Cobb levels in treating thoracolumbar/lumbar (TL/L) scoliosis (Lenke 5) in adolescent patients.
Method:
We reviewed the medical notes and radiographs of the senior author's consecutive series of 53 adolescent patients with TL/L scoliosis treated by posterior instrumented spinal arthrodesis using an all-pedicle screw construct. Our patients were categorised into 2 groups: patients with instrumented fusion between Cobb-to-Cobb levels of the TL/L curve (Group 1), and patients that required distal extension beyond the caudal Cobb level (Group 2). Pearson correlation and binary logistic regression analyses (significance p<0.05) were performed to identify variables that predict the need for distal extension.
Results:
Groups 1 and 2 comprised 36 and 17 patients, respectively. The following preoperative parameters significantly correlated with distal extension of the fusion: TL/L scoliosis angle (TL/L), TL/L supine maximum lateral bending angle, TL/L apical vertebral translation (AVT), TL/L flexibility index (FI), lowest instrumented vertebra angle (LIVA), and compensatory thoracic scoliosis angle (TH). Binary logistic regression analysis optimised a predictive equation incorporating TL/L, AVT, FI, LIVA, and TH parameters that provides an 81% accuracy in predicting the need for Cobb-to-Cobb fusion or distal extension. There was no difference in demographic data or SRS-22 scores between the 2 groups.
Discussion:
Regression analysis of preoperative radiographic variables can accurately predict the need for distal extension of the fusion beyond the preoperative Cobb-to-Cobb levels during posterior spinal arthrodesis in patients with adolescent idiopathic thoracolumbar/lumbar scoliosis.
Conflict Of Interest Statement: No conflict of interest.
