Abstract / Summary
Anticoagulation is recommended for atrial fibrillation (AF) in cardiac amyloidosis regardless of CHA₂DS₂-VASc score, but no guidance exists on agent selection. Using the TriNetX Research Network, we identified adults with amyloidosis (ICD-10 E85) and non-valvular AF/flutter (I48). Prosthetic valves and rheumatic mitral stenosis were excluded. In an active-comparator new-user design, patients initiating a direct oral anticoagulant (DOAC) or warfarin were propensity matched 1:1 across 30 covariates. The primary outcome was thromboembolic events (stroke or systemic embolism). Secondary outcomes included major bleeding, heart failure hospitalization, myocardial infarction, and cardiac arrest. Exploratory subgroup analyses assessed light chain (AL) and wild-type transthyretin (ATTR) amyloidosis. Among 1,110 matched pairs, no significant difference in thromboembolic events was observed (HR 0.798, 95% CI 0.525 to 1.213, p = 0.289). Major bleeding (HR 1.164), heart failure hospitalization (HR 0.848), myocardial infarction (HR 1.162), and cardiac arrest (HR 1.190) did not differ. Results were consistent after excluding events in the first 30 days (HR 0.971) and requiring a coded cardiac diagnosis (HR 0.702). In the 2019 to 2025 era, DOACs were associated with fewer thromboembolic events (HR 0.554, 95% CI 0.326 to 0.941, p = 0.027). In exploratory analyses, DOACs were associated with more thromboembolic events in AL amyloidosis (549 pairs, HR 2.206, 95% CI 1.186 to 4.103) but not ATTR (756 pairs, HR 0.741). No significant difference in thromboembolic or bleeding outcomes was observed between DOACs and warfarin in amyloidosis with non-valvular AF. The AL subgroup signal is hypothesis-generating. Prospective trials stratified by amyloid subtype are needed.