592 - Diabetes and Long-Term Complications After Vertebroplasty and Kyphoplasty: A Propensity Score–Matched Analysis

Purpose: Percutaneous vertebroplasty and kyphoplasty are recognized treatment options for osteoporotic vertebral compression fractures. Current Society of Interventional Radiology (SIR) guidelines outline procedural risks but do not address diabetes as a potential modifier. Considering the strong association between diabetes, increased risk of infection and impaired bone healing, this study assessed whether diabetic status contributes to higher complication rates following vertebral augmentation.

Materials and Methods: A retrospective cohort study was performed using the TriNetX research network, covering the period from January 1, 2015, to September 1, 2025. Adults aged ≥18 years who underwent vertebroplasty or kyphoplasty were identified using relevant CPT and ICD-10 procedure codes, with exclusions for malignancy, traumatic fractures, and prior augmentation within 12 months. Diabetic and non-diabetic cohorts were propensity score–matched (n=6,565 each) to balance baseline demographics and comorbidities. Outcomes at 90 days, 6 months, 1 year, 3 years, and 5 years included infectious, spine-related structural, and healthcare utilization complications. Risk ratios (RR), 95% confidence intervals (CI), and p-values were calculated to determine significance.

Results: In matched cohorts, diabetes was associated with higher spinal stenosis rates at 90 days (8.6% vs 7.4%, RR 1.16, p=0.013) and persisting through 5 years (24.5% vs 22.4%, RR 1.09, p=0.005). Spinal fusion was also increased at 3 years (1.6% vs 1.1%, RR 1.39, p=0.029) and 5 years (1.7% vs 1.2%, RR 1.37, p=0.030). Sepsis was consistently elevated, including 1 year (7.7% vs 5.9%, RR 1.30, p< 0.001) and 5 years (15.5% vs 12.6%, RR 1.23, p< 0.001). Emergency department utilization was higher across all timepoints, reaching 44.8% vs 41.6% at 5 years (RR 1.08, p< 0.001). Thromboembolic complications were not different early but became modestly elevated by 5 years (10.2% vs 8.9%, RR 1.14, p=0.013). Chronic pain, repeat procedures, and mortality were not significantly different.

Conclusion: Diabetic status was associated with higher rates of spinal stenosis, sepsis, emergency department utilization, and spinal fusion over mid- to long-term follow-up, while repeat procedures, chronic pain, and mortality did not differ. Thromboembolic complications were not significantly different early and showed only modest increases at longer follow-up. These findings highlight the need for risk stratification, closer monitoring, and recognition of diabetes as a potential risk-modifying factor in vertebral augmentation.