Abstract
OBJECTIVES: It is uncertain whether the protein targets of direct oral anticoagulants (DOACs) causally influence cerebral venous thrombosis (CVT) risk. We used Mendelian randomization (MR) to examine the association between genetically proxied levels of DOAC protein targets and CVT risk.
METHODS: We identified genetic proxies for the protein targets of established and emerging DOACs: circulating thrombin, activated factor X (FXa), and factor XI (FXI). CVT cases were ascertained from the UK Biobank (n = 337,134) and FinnGen cohorts (n = 454,144). For each drug target, we estimated the association between a genetically proxied one-standard deviation reduction in circulating protein levels and CVT risk using the Wald ratio. Cohort-specific estimates were meta-analyzed using the fixed-effect inverse variance-weighted method.
RESULTS: We identified 422 total CVT cases across both cohorts. Genetically proxied lower levels of all 3 DOAC protein targets associated with reduced CVT risk: FXI (odds ratio [OR] 0.38, 95% CI 0.26-0.55), FXa (OR 0.28, 95% CI 0.10-0.78), and thrombin (OR 0.22, 95% CI 0.07-0.74).
DISCUSSION: We found strong genetic evidence linking FXI to CVT pathogenesis and suggestive genetic evidence for FXa and thrombin. Inhibition of these proteins with DOACs may represent an effective therapeutic strategy for CVT.