Efficacy and safety of factor XIa inhibitors for stroke prevention: a reconstructed individual patient data meta-analysis

Background Factor XIa inhibitors represent a novel anticoagulant class targeting the intrinsic coagulation pathway, potentially offering stroke prevention with reduced bleeding risk compared to direct oral anticoagulants. Despite promising phase 2 data, comprehensive evaluation across diverse patient populations remains limited. Traditional aggregate meta-analyses cannot adequately address heterogeneity in patient characteristics and comparator regimens. Purpose We performed an individual patient data meta-analysis to evaluate efficacy and safety of Factor XIa inhibitors (milvexian, asundexian, abelacimab) for stroke prevention across multiple clinical settings, using reconstructed pseudo-IPD from published Kaplan-Meier curves. Our objective was to determine pooled treatment effects on ischemic stroke/systemic embolism and major bleeding while accounting for between-trial heterogeneity. Methods We reconstructed pseudo-IPD for 31,681 patients from four phase 2/3 trials (AXIOMATIC-SSP, PACIFIC-STROKE, AZALEA-TIMI 71, OCEANIC-AF) using the Guyot algorithm. Validation against published hazard ratios demonstrated 65% achieved good-to-excellent concordance (mean relative difference 18.4%). We conducted both one-stage (stratified Cox regression) and two-stage (random-effects pooling) IPD meta-analyses. Pre-specified sensitivity analyses excluded trials with poor validation quality. Comprehensive audit of 51 verification checks confirmed 94% pass rate, validating analysis integrity from raw data to final estimates. Results Among high-quality trials (n=8,653 patients, 613 events), Factor XIa inhibitors demonstrated significant 29% relative risk reduction in ischemic stroke/systemic embolism (pooled HR 0.71, 95%CI 0.56-0.90, p=0.005; I²=22.1%) with low heterogeneity. Major bleeding showed neutral effect (HR 1.03, 95%CI 0.67-1.57, p=0.902), though substantial heterogeneity reflected diverse comparators (placebo vs. rivaroxaban). One-stage and two-stage approaches demonstrated excellent agreement in validated trials. The stroke risk reduction translates to approximately 1.5% absolute risk reduction annually in high-risk populations (NNT=68). Conclusion Factor XIa inhibitors demonstrate clinically meaningful stroke prevention efficacy with potentially favorable bleeding profiles. The 29% relative risk reduction represents a significant advancement over standard anticoagulation. However, inter-trial heterogeneity driven by methodological quality differences underscores the critical need for rigorous data validation in reconstructed IPD meta-analyses. Future head-to-head trials comparing Factor XIa inhibitors directly to DOACs in adequately powered atrial fibrillation cohorts are warranted to definitively establish their clinical role.