Cerebral Microbleeds, Anemia, and Estimated Net Benefit of DOACs versus Warfarin in Older Atrial Fibrillation: A Single-Center MRI-Selected Cohort

Background: Randomized trials favor direct oral anticoagulants (DOACs) over warfarin for eligible atrial fibrillation (AF), but whether cerebral microbleeds (CMBs) and anemia identify older patients with narrower benefit–harm margins remain uncertain. We estimated 2-year net clinical benefit (NCB) overall and by CMB/anemia strata. Methods: This single-center, MRI-selected prospective observational cohort included patients aged ≥65 years with non-valvular AF receiving a DOAC or warfarin. Follow-up began at susceptibility MRI; CMB presence (0 vs ≥1) was rated by two blinded raters, and anemia used World Health Organization hemoglobin thresholds. DOAC–warfarin associations were estimated using propensity-score overlap weighting and competing-risk models for ischemic stroke/systemic embolism (IS/SE) and intracranial hemorrhage (ICH). Two-year NCB was expressed as net events prevented per 1000 with bootstrap 95% confidence intervals (CIs). Results: Among 797 patients (DOAC n=517; warfarin n=280), 31.0% had ≥1 CMB and 33.8% had anemia. During 2463 person-years, 40 first IS/SE events, 22 first ICH events, and 89 deaths occurred as first components of net adverse clinical events (NACE). After weighting, DOAC use was associated with lower NACE (2-year risk 14.9% vs 21.8%; hazard ratio 0.69, 95% CI 0.52–0.91), ICH (1.4% vs 4.3%; subdistribution hazard ratio 0.39, 95% CI 0.21–0.72), and IS/SE (3.9% vs 6.4%; subdistribution hazard ratio 0.63, 95% CI 0.41–0.96). Estimated NCB favored DOACs overall (67.0/1,000; 95% CI 28.4–102.7) and was attenuated in patients with both CMBs and anemia (38.0/1,000; 95% CI 4.2–71.5). Conclusions: DOAC use was associated with a favorable estimated 2-year NCB vs warfarin in this MRI-selected cohort. Combined CMB/anemia findings are exploratory and hypothesis-generating and should not guide prescribing without multicenter validation.