Mid-Term Outcomes of Screw Loosening in Lumbar Dynamic Stabilization with Polyetheretherketone Rods versus Titanium Rods: A Minimum 4-Year Follow-Up.

Abstract

To retrospectively analyze screw loosening rates following dynamic fixation with polyetheretherketone (PEEK) rods and rigid fixation with titanium rods and assess mid-term outcomes of loosened screws. This retrospective analysis included 203 patients who underwent lumbar pedicle screw fixation between March 2017 and June 2020 (57 with PEEK rods, 146 with titanium rods). Patients were followed for at least 48 months to evaluate screw loosening postoperatively and investigate outcomes of loosened screws. Multivariate logistic regression identified factors influencing screw restabilization. Both PEEK rod and titanium rod groups exhibited peak screw loosening rates approximately 1 year after surgery, decreasing with longer follow-up. At 48 months, screw loosening rates were 5.3% and 15.8% for PEEK and titanium rods, respectively (P < 0.05). Among patients experiencing early loosening (within 12 months), the proportion of stabilized screws was significantly higher with PEEK rods (84% vs. 34%, P < 0.001). Regression analysis revealed dynamic fixation (odds ratio 4.579; 95% confidence interval 1.611-12.519), lowest fixed vertebra S1 (odds ratio 3.151; 95% confidence interval 1.352-9.233), and L1-L4 average computed tomography value (odds ratio 1.132; 95% confidence interval 1.015-1.263) as independent risk factors for screw restabilization. The area under the receiver operating characteristic curve for L1-L4 average computed tomography value predicting restabilization was 0.713 (P < 0.05), with an optimal threshold of 106 Hounsfield units (sensitivity 0.771, specificity 0.803). Following PEEK rod dynamic fixation surgery, a certain proportion of screw loosening may occur in the short term. With prolonged follow-up, screws gradually restabilize at the bone interface, with most loosened screws returning to a stable state.