Loss of coronal correction following instrumentation removal in adolescent idiopathic scoliosis

Study Design. Retrospective radiographic analysis of patients with adolescent idiopathic scoliosis (AIS) status after instrumentation removal. Objective. To evaluate the effect of instrumentation removal in surgically corrected AIS. Summary of Background Data. Spinal instrumentation is occasionally removed for various reasons, most commonly for postoperative pain or infection, in surgically corrected AIS. The fate of instrumentation removal in adults has been previously reported with documented loss of sagittal alignment. However, to our knowledge, the long-term follow-up after instrumentation removal in AIS has not been reported. Methods. We retrospectively reviewed the preoperative, pre-instrumentation removal, postoperative following instrumentation removal, and latest follow-up radiographs of all patients with a primary diagnosis of AIS who underwent instrumentation removal after posterior spinal fusion. There were 21 patients (15 females, 6 males) from 2 institutions, at an average age of 14.8 years (range 9-19), who were originally treated between 1988 and 2002. Instrumentation removal occurred at an average of 2.4 years after surgery (range 8 months to 4 years, 2 months), with an average follow-up of 5.2 years (range 2-11). Fifteen patients underwent removal secondary to pain (2 of these with undetected infection) and 6 for known infection. Evaluation included coronal proximal thoracic, main thoracic, thoracolumbar/lumbar (TL/L), and sagittal T2-T5, T5-T12, T2-T12, T10-L2, T12-S1, and sagittal balance before surgery, before instrumentation removal, at immediate post-removal evaluation, and at latest follow-up. Results. There were 12 main thoracic curves (Lenke type 1), 6 double thoracic curves (Lenke type 2), and 3 double major curves (Lenke type 3) in the series. Average proximal thoracic curve was 19.7° (range 5°-35°), the main thoracic curve 63.3 (range 42°-112°), and the TL-L curve averaged 31.4° (range 17°-53°). There was an immediate loss of approximately 4° (range 0°-8°) in the main thoracic curve and 6° (range 1°-15°) in the TL/L curve after removal, with continued settling of an additional 6° (10° total, P = 0.002) in the main thoracic curve, and 3° in the TL/L curve (9° total, P = 0.01). There was also a significant difference in the group that underwent instrumentation removal <2 years after surgery compared to >2 years (main thoracic curve 13° vs. 7°, P= 0.017; TL/L 11° vs. 7°, P = 0.036). There were no significant changes in sagittal curvature or sagittal balance in either group (P > 0.39). Conclusions. Instrumentation removal in AIS is not always a benign process because the long-term follow-up of this cohort of patients shows a "settling" effect in the coronal plane of the main thoracic and TL/L curves after instrumentation removal. Interestingly, there was no change in the sagittal plane with time. Parents and patients should be counseled for this result when instrumentation removal is contemplated, and limited removal of focally symptomatic implants should be considered.