Preoperative Lp(a) of 30 mg/dL or more independently predicts peri-procedural stroke after intracranial stenting (Neurol Res 2026)

Background And Objectives: Peri-procedural ischemic stroke (PPIS) is the primary complication limiting the benefit of intracranial stenting for symptomatic intracranial atherosclerotic stenosis (ICAS). Lipoprotein(a) [Lp(a)] is a known risk factor for long-term vascular events, but its role in predicting acute procedure-specific risk is unknown. This study aimed to determine whether preoperative Lp(a) independently predicts PPIS, characterize its dose-response relationship, and evaluate its incremental value for risk stratification.

Methods: We retrospectively analyzed patients with severe symptomatic ICAS who underwent stenting (2020-2024). Patients were stratified by preoperative Lp(a) (cutoff: 30 mg/dL). Primary outcome was PPIS within 30 days. Multivariate logistic regression assessed the independent association. Model performance was evaluated using ROC and continuous net reclassification improvement (NRI). A restricted cubic spline model visualized the dose-response relationship.

Results: PPIS occurred in 33 patients (7.7%). Incidence was higher in the high Lp(a) group (14.6% vs. 4.4%, p=0.001). After adjusting for stenosis rate and preoperative hyperglycemia, high Lp(a) remained an independent predictor (aOR=2.82, 95%CI 1.35-5.89). Adding Lp(a) improved AUC from 0.727 to 0.766 (NRI=0.421, p=0.019). Restricted cubic spline revealed a monotonically increasing, non-linear relationship between Lp(a) and PPIS risk.

Conclusion: Preoperative Lp(a) independently predicts PPIS with a dose-response relationship, offering ancillary information for risk stratification. Given the single-center design and limited events, these findings require multicenter validation.