Risk of Cardiovascular Events and Mortality Among Elderly Patients With Reduced GFR Receiving Direct Oral Anticoagulants

RATIONALE & OBJECTIVE: Evidence for the efficacy of direct oral anticoagulants (DOACs) to prevent cardiovascular (CV) events and mortality in older individuals with a low estimated glomerular filtration rate (eGFR) is lacking. We sought to characterize the association of oral anticoagulant use with CV morbidity in elderly patients with or without reductions in eGFRs, comparing DOACs with vitamin K antagonists (VKAs). STUDY DESIGN: Population-based retrospective cohort study. SETTINGS & PARTICIPANTS: All individuals 66 years or older with an initial prescription for oral anticoagulants dispensed in Ontario, Canada, from 2009 to 2016. EXPOSURE: DOACs (apixaban, dabigatran, and rivaroxaban) compared with VKAs by eGFR group (≥60, 30-59, and<30mL/min/1.73m²). OUTCOMES: The primary outcome was a composite of a CV event (myocardial infarction, revascularization, or ischemic stroke) or mortality. Secondary outcomes were CV events alone, mortality, and hemorrhage requiring hospitalization. ANALYTICAL APPROACH: High-dimensional propensity score matching of DOAC to VKA users and Cox proportional hazards regression. RESULTS: 27,552 new DOAC users were matched to 27,552 new VKA users (median age, 78 years; 49% women). There was significantly lower risk for CV events or mortality among DOAC users compared with VKA users (event rates of 79.78 vs 99.77 per 1,000 person-years, respectively; HR, 0.82 [95% CI, 0.75-0.90]) and lower risk for hemorrhage (event rates of 10.35 vs 16.77 per 1,000 person-years, respectively; HR, 0.73 [95% CI, 0.58-0.91]). There was an interaction between eGFR and the association of anticoagulant class with the primary composite outcome (P<0.02): HRs of 1.01 [95% CI, 0.92-1.12], 0.83 [95% CI, 0.75-0.93], and 0.75 [95% CI, 0.51-1.10] for eGFRs of≥60, 30 to 59, and<30mL/min/1.73m². No interaction was detected for the outcome of hemorrhage. LIMITATIONS: Retrospective observational study design limits causal inference; dosages of DOACs and international normalized ratio values were not available; low event rates in some subgroups limited statistical power. CONCLUSIONS: DOACs compared with VKAs were associated with lower risk for the composite of CV events or mortality, an association for which the strength was most apparent among those with reduced eGFRs. The therapeutic implications of these findings await further study.