Risks and Benefits of Direct Oral Anticoagulants across the Spectrum of GFR among Incident and Prevalent Patients with Atrial Fibrillation

Abstract

Background and objectives: All randomized trials of direct oral anticoagulants in atrial fibrillation excluded patients with severe kidney disease. The safety and effectiveness of direct oral anticoagulants across the range of eGFR in real-world settings is unknown. Our objective is to quantify the risk of bleeding and benefit of ischemic stroke prevention for direct oral anticoagulants compared with warfarin in patients with atrial fibrillation with and without CKD.

Design, setting, participants, & measurements: We created a propensity score-matched cohort of 3206 patients with atrial fibrillation and direct oral anticoagulant use and 3206 patients with atrial fibrillation using warfarin from October of 2010 to February of 2017 in an electronic health record (Geisinger Health System). The risks of bleeding and ischemic stroke were compared between direct oral anticoagulant and warfarin users using Cox proportional hazards regression, stratified by eGFR (≥60 and <60 ml/min per 1.73 m²).

Results: The mean (SD) age of the 6412 participants was 72 (12) years, 47% were women, and average eGFR was 69 (21) ml/min per 1.73 m². There were 1181 bleeding events and 466 ischemic strokes over 7391 person-years of follow-up. Compared with warfarin use, the hazard ratios (HRs) (95% confidence interval [95% CI]) of bleeding associated with direct oral anticoagulant use were 1.01 (0.88 to 1.17) and 1.23 (1.02 to 1.48) for those with eGFR≥60 and eGFR<60 ml/min per 1.73 m², respectively (P-interaction=0.10). There was no difference between direct oral anticoagulant and warfarin users in the risk of ischemic stroke: HRs (95% CI) of 0.94 (0.74 to 1.18) and 1.02 (0.76 to 1.37) for those with eGFR≥60 and eGFR<60 ml/min per 1.73 m², respectively (P-interaction=0.70). Similar findings were observed with individual drugs.

Conclusions: In a large health care system, patients with eGFR<60 ml/min per 1.73 m² who took direct oral anticoagulants for atrial fibrillation had slightly higher risk of bleeding compared with those on warfarin, but similar benefits from prevention of ischemic stroke.

Keywords: Anticoagulants; Atrial Fibrillation; Brain Ischemia; Confidence Intervals; Direct Oral Anticoagulants; Electronic Health Records; Female; Follow-up Studies; Hemorrhage; Humans; Propensity Score; Renal Insufficiency, Chronic; Risk Assessment; Stroke; Warfarin; chronic kidney disease; glomerular filtration rate; kidney.

Graphical abstract

Figure 1.

Direct oral anticoagulant prescriptions increased over time among patients with atrial fibrillation in all stages of CKD in the Geisinger Health System (n=20,727). The total number of patients in each year from 2010 to 2016 was 6622, 6826, 7282, 7964, 8445, 8781, and 9001, respectively. CKD stage G1–2: eGFR≥60, G3: eGFR 30–59, and G4–5: eGFR<30 ml/min per 1.73 m².

Figure 2.

Patients with low eGFR on direct oral anticoagulants for treatment of atrial fibrillation experienced bleeding events more frequently than those on warfarin, but had similar rates of ischemic stroke. Incidence rates of bleeding (A) and ischemic stroke (B) by eGFR category among propensity-score matched patients with atrial fibrillation using direct oral anticoagulants or warfarin. IR, incidence rate; PYs, person-years.

Figure 3.

Pattern of prescribed direct oral anticoagulant dose by eGFR and the FDA recommended dose level. Adding the sample size of cohort of each drug in (A) does not give 3206, the total number of matched direct oral anticoagulant users, because individuals changing from one type to the others were included in two or three different cohorts of each drug. The sample sizes in (B) are slightly smaller than those in (A) due to missing body weight. The current FDA dosing guidelines were applied to all past dosing of each direct oral anticoagulant: Dabigatran 150 mg twice a day and 75 mg twice a day for those with creatinine clearance >30 and creatinine clearance 15–30 ml/min, respectively. Rivaroxaban 20 mg every day and 15 mg every day for those with creatinine clearance >50 and creatinine clearance 15–50 ml/min, respectively. Apixaban 5 mg twice a day unless dose reduction is recommended. Apixaban 2.5 mg twice a day for those with at least two of the following conditions: age≥80 years, body wt ≤60 kg, or serum creatinine ≥1.5 mg/dl. FDA, Food and Drug Administration.

Figure 4.

Cumulative incidence of treatment change during follow-up among matched direct oral anticoagulant users and warfarin users with atrial fibrillation. *Baseline time for warfarin users (A) is the warfarin prescribed date matched with duration of previous warfarin use, in comparison with direct oral anticoagulant users (B). Thirty-nine percent of discontinuation of anticoagulant occurred after bleeding events.