87. Rates of loosening, failure, and revision of iliac fixation in adult deformity surgery

Abstract

BACKGROUND CONTEXT Traditional iliac (IS) and S2-alar-iliac (S2AI) pelvic fixation methods have unique technical characteristics in their application, and result in varied biomechanical and anatomic impact. These differences may lead to variance in lumbopelvic fixation failure types and rates. PURPOSE We intended to evaluate the influence of type of iliac fixation utilized for adult spinal deformity (ASD) correction on the frequency and manner of lumbopelvic fixation failure. STUDY DESIGN/SETTING Retrospective review of prospective, multicenter ASD database. PATIENT SAMPLE This study included 410 patients. OUTCOME MEASURES HRQL (ODI, SF-36, NRS), rod fracture, screw fracture, screw loosening, revision surgery, pseudarthrosis METHODS Inclusion criteria included ASD (coronal Cobb≥20°, sagittal vertical axis (SVA) ≥5cm, pelvic tilt ≥25° and/or thoracic kyphosis >60°) ≥ 18 years old, 2yr follow-up, and >5 level fusion with pelvic fixation. Cohort subdivided by type of pelvic fixation (IS vs S2AI). Loosening was defined by lucency around the screw shank on radiographs. Univariate testing was performed using t-tests and chi-squared tests. Multivariate logistic regression, accounting for significant confounders, was used to examine differences between the two groups for screw loosening/fracture, rod fracture, and revision surgery. Level of significance = p RESULTS Of the total of 1422 patients, 410 met inclusion criteria (IS=287, S2AI=131). The two groups had similar age, BMI, baseline co-morbidities, and number of levels fused (p>0.05), as well as similar baseline HRQLs (SF-36, ODI, SRS-22, NRS leg and back, p>0.05) and deformity (PT, PI-LL, SVA, p>0.05). Patients in the IS group had a higher proportion of unilateral fixation compared to S2AI group (12.9% vs 5.6%; p=0.001). The overall pelvic fixation failure rate (screw loosening/screw fracture/rod fracture) was 29.4%. Loosening of pelvic fixation occurred in 13.4% of patients, and was more prevalent in the S2AI fixation group (OR 2.74, p=0.001). The rate of S1 screw loosening was 2.9%, and more likely to occur in the S2AI group (OR 4.17, p=0.045). The rate of pelvic fixation fracture in the overall cohort was 2.3%, with no difference between groups (p=0.37). Rod fracture occurred in 14.1% below L4, with a trend toward less occurrence in the S2AI group (OR 0.47, p=0.06). Revision surgery was required in 22.7% of our cohort, with no difference between groups (p=0.449). Patients with failure of the pelvic fixation had less improvement in their HRQL at 2years (PCS 7.69 vs 10.46 p=0.028; SRS 0.83 vs 1.03 p=0.019; ODI 12.91 vs 19.77 p=0.0016). CONCLUSIONS Pelvic fixation is commonly used in long segment adult spinal deformity surgery to improve the rate of L5-S1 fusion and protect from construct failure at the lumbosacral junction. Our results demonstrate a substantial rate of pelvic fixation hardware issues following ASD correction. Lumbopelvic fixation failure occurred at a rate of 29.4% following surgical correction of ASD that involved iliac fixation and was associated with diminished clinical outcomes. Additionally, S2AI screws were more likely to demonstrate loosening, but less commonly associated with rod fractures. Patients who experienced failure of their pelvic fixation had less improvement in their HRQL two years postoperatively. FDA DEVICE/DRUG STATUS This abstract does not discuss or include any applicable devices or drugs.