Preoperative Lipoprotein(a) and Dose-Dependent Risk of Stroke After Intracranial Stenting for Symptomatic Atherosclerosis

Objectives: Lipoprotein(a) Lp(a) is an established risk factor for atherosclerotic disease and predicts events after coronary or carotid revascularization. Its role after intracranial artery stenting remains undefined. We investigated the association between preoperative Lp(a) and 1-year stroke risk post-stenting for symptomatic intracranial atherosclerotic stenosis (ICAS), as well as to characterize the dose-response relationship. Methods: In this retrospective analysis of a prospectively maintained registry, we enrolled consecutive patients undergoing stenting for symptomatic ICAS (2022 to 2024). Patients were categorized by preoperative Lp(a) (<30 vs. ≥30 mg/dL). The primary endpoint was ischemic stroke (30 d to 1 y). Associations were assessed through Kaplan-Meier, Cox regression, and restricted cubic splines (RCS) for dose-response analysis. Sensitivity and competing-risk analyses were performed. Results: Among 405 patients, the high Lp(a) group (n=132) had a significantly higher stroke incidence (12.87% vs. 6.27%, P =0.024). After adjusting for diabetes and follow-up LDL-C, Lp(a) ≥30 mg/dL was independently associated with stroke risk (adjusted HR=2.05, 95% CI: 1.04-4.04; P =0.038). Results were robust in sensitivity analyses. RCS analysis confirmed a significant nonlinear dose-response relationship ( P for nonlinearity=0.010). The risk increased progressively across the Lp(a) distribution, with a markedly attenuated slope of increase at concentrations above ~50 mg/dL. A post hoc cutoff of ≥50 mg/dL yielded a higher HR (2.73) but with wider confidence intervals. Conclusions: In a contemporary ICAS stenting cohort, elevated preoperative Lp(a) is an independent predictor of 1-year stroke recurrence, demonstrating a nonlinear dose-response pattern. The 30 mg/dL threshold is pragmatic for clinical risk screening, whereas the 50 mg/dL level may serve to enrich populations for trials of novel Lp(a)-lowering therapies. These findings require prospective, multicenter validation.