Outcomes in Older Patients After Switching to a Newer Anticoagulant or Remaining on Warfarin

BACKGROUND Whether frail, elderly patients with atrial fibrillation (AF) on a vitamin K antagonist (VKA) should switch to a direct-acting oral anticoagulant (DOAC) was studied in the FRAIL-AF trial and remains controversial. OBJECTIVES The purpose of this study was to evaluate, in the COMBINE-AF data set, the impact on clinical outcomes of switching frail, elderly AF patients from VKA to DOAC. METHODS COMBINE-AF consists of individual patient-level data from 71,683 patients with AF in 4 randomized clinical trials comparing DOAC vs warfarin. Frailty was evaluated using a frailty index derived from a modified Rockwood's Accumulation Model including 18 age-related conditions. Patients with a frailty index score above the median were considered frail. Prespecified outcomes were stroke or systemic embolic events, bleeding events, death, and a net clinical outcome combining these events. RESULTS We identified 5,913 patients who were frail, elderly (age ≥75 years), and VKA-experienced and 52,721 patients who did not meet all 3 of these criteria. Patients were randomized to a standard-dose (SD) DOAC or warfarin. After 27 months median follow-up, there was no heterogeneity in treatment effect with SD-DOAC vs warfarin among those who met all 3 criteria vs those who did not for the endpoints of stroke or systemic embolic events (HR: 0.83 vs 0.81; Pint = 0.75) or for death (HR: 0.95 vs 0.91; Pint = 0.54). Major bleeding was similar with SD-DOAC vs warfarin in frail, elderly, VKA-experienced patients (HR: 1.06 95% CI: 0.90-1.25), while it was significantly reduced with SD-DOAC in patients without all 3 criteria (HR: 0.82 95% CI: 0.76-0.89; Pint = 0.007). Likewise, the net clinical outcome was similar in the frail, elderly, VKA-experienced patients with SD-DOAC vs warfarin (HR: 1.01 95% CI: 0.91-1.13), while significantly reduced with SD-DOAC patients without all 3 criteria (HR: 0.89 95% CI: 0.85-0.93; Pint = 0.028). Fatal and intracranial bleeding were significantly reduced with SD-DOAC in both subgroups to a similar degree (both Pint > 0.05), while gastrointestinal bleeding with SD-DOAC was increased to a greater degree in frail, elderly, VKA-experienced patients (HR: 1.83 95% CI: 1.42-2.36) compared with those without all 3 criteria (HR: 1.23 95% CI: 1.09-1.39; Pint = 0.006). CONCLUSIONS Frail, elderly, VKA-experienced patients with AF switched to SD-DOAC experienced significant reductions in stroke or systemic embolism, fatal and intracranial bleeding, and death. Gastrointestinal bleeding was increased with SD-DOAC, while major bleeding and the primary net clinical outcome were similar. Based on these findings, SD-DOAC is a reasonable choice for frail, elderly, VKA-experienced patients to reduce stroke and systemic embolism, death, and the most serious types of bleeding.