Direct Oral Anticoagulants versus Warfarin in Patients with Atrial Fibrillation: Patient-Level Network Meta-Analyses of Randomized Clinical Trials with Interaction Testing by Age and Sex

Background: Direct oral anticoagulants (DOACs) are preferred over warfarin for stroke prevention in atrial fibrillation (AF). Meta-analyses using individual patient data offer significant advantages over study-level data. Methods: We used individual patient data from the COMBINE AF database, which includes all patients randomized in the 4 pivotal trials of DOACs vs warfarin in AF (RE-LY, ROCKET AF, ARISTOTLE, ENGAGE AF-TIMI 48), to perform network meta-analyses using a stratified Cox model with random effects comparing standard-dose DOAC, lower-dose DOAC, and warfarin. Hazard ratios (95% CIs) were calculated for efficacy and safety outcomes. Covariate-by-treatment interaction was estimated for categorical covariates and for age as a continuous covariate, stratified by sex. Results: A total of 71,683 patients were included (29,362 on standard-dose DOAC, 13,049 on lower-dose DOAC, 29,272 on warfarin). Compared with warfarin, standard-dose DOACs were associated with a significantly lower hazard of stroke/systemic embolism (883/29312 [3.01%] vs 1080/29229 [3.69%]; HR 0.81, 95% CI 0.74-0.89), death (2276/29312 [7.76%] vs 2460/29229 [8.42%]; HR 0.92, 95% CI 0.87-0.97) and intracranial bleeding (184/29270 [0.63%] vs 409/29187 [1.40%]; HR 0.45, 95% CI 0.37-0.56), but no statistically different hazard of major bleeding (1479/29270 [5.05%] vs 1733/29187 [5.94%]; HR 0.86, 95% CI 0.74-1.01), whereas lower-dose DOACs were associated with no statistically different hazard of stroke/systemic embolism (531/13049 [3.96%] vs 1080/29229 [3.69%]; HR 1.06, 95% CI 0.95-1.19) but a lower hazard of intracranial bleeding (55/12985 [0.42%] vs 409/29187 [1.40%]; HR 0.28, 95% CI 0.21-0.37), death (1082/13049 [8.29%] vs 2460/29229 [8.42%]; HR 0.90, 95% CI 0.83-0.97), and major bleeding (564/12985 [4.34%] vs 1733/29187 [5.94%]; HR 0.63, 95% CI 0.45-0.88). Treatment effects for standard- and lower-dose DOACs versus warfarin were consistent across age and sex for stroke/systemic embolism and death, whereas standard-dose DOACs were favored in patients with no history of vitamin K antagonist use (p=0.01) and lower creatinine clearance (p=0.09). For major bleeding, standard-dose DOACs were favored in patients with lower body weight (p=0.02). In the continuous covariate analysis, younger patients derived greater benefits from standard-dose (interaction p=0.02) and lower-dose DOACs (interaction p=0.01) versus warfarin. Conclusions: Compared with warfarin, DOACs have more favorable efficacy and safety profiles among patients with AF.