Direct Oral Anticoagulants and Risk of Acute Kidney Injury in Patients with Atrial Fibrillation

Patients receiving warfarin are at higher risk for acute kidney injury (AKI), presumably due to a tendency towards glomerular hemorrhage, a condition known as anticoagulation-related nephropathy (1). Recently, Chan et al. found that dabigatran, a direct oral anticoagulant (DOAC), was associated with lower risk of AKI compared to warfarin among Asian patients with atrial fibrillation (AF) (2). This study, however, did not account for the severity of pre-existing chronic kidney disease (CKD), which is strongly associated with AKI risk, and likely influences the choice of anticoagulants. Whether the risk of AKI remains lower across different levels of glomerular filtration rate (GFR), and in non-Asian populations, requires further investigation. Using a large, tertiary health system and propensity-score matching techniques, we compared AKI risk associated with the 3 available DOACs (dabigatran, rivaroxaban, or apixaban) versus warfarin overall and stratified by estimated GFR (eGFR) (≥60, 30-59, and <30 mL/min/1.73 m²) (Chronic Kidney Disease Epidemiology Collaboration equation) (3).

We identified 20,727 patients with AF (ICD9-CM code 427.x) who filled >1 prescription of DOAC (dabigatran, rivaroxaban, apixaban) or warfarin between October 19, 2010 (the date of the US FDA approval of dabigatran) and February 2, 2017 in a fully integrated rural health care system in Pennsylvania. Among 3,647 patients with AF who initiated DOACs during the study period, we used 1:1 propensity-score matching to select 3,206 patients with AF on DOACs and 3,206 similar patients with AF on warfarin. The study outcome, AKI, was ascertained using ICD-9-CM codes 584.x. Patients were followed from baseline (date of first DOAC prescription or matched date of warfarin prescription) until the first development of AKI, death, discontinuation of anticoagulants, prescription of a different anticoagulant, or the end of study period. The risk of AKI was compared between DOAC and warfarin users using Cox proportional hazards regression. The study protocol was reviewed and exempted by the institutional review boards of Geisinger Medical Center and Johns Hopkins University.

Mean baseline age of the 6,412 participants was 72 years, 47% were women, and 98% were non-Hispanic whites. Mean baseline eGFR was 69 mL/min/1.73 m². Baseline age, sex, race, eGFR, prescription year, duration of prior anticoagulation, CHA₂DS₂-VASc score, HAS-BLED score, comorbidities, and medication use were well balanced between DOAC and warfarin users (all standardized mean difference <0.1). Among warfarin users, 56.7% had poor anticoagulation control (time in therapeutic range <60%) during follow-up. A total of 659 AKI events (72.2% were inpatient) occurred among 6,412 patients with 6,945 person-years of follow-up (incidence rate = 9.49 per 100 person-years). Overall, DOAC users had a 21% lower risk of AKI compared to warfarin users (hazard ratio [HR] 0.79, 95% confidence interval [CI] 0.68-0.92, p = 0.003, Table 1). When stratified by baseline eGFR, DOACs were associated with a lower risk of AKI compared to warfarin among patients with eGFR ≥60 mL/min/1.73 m² (HR 0.72 95% CI 0.57-0.91, p =0.007). Among patients with eGFR 30-59 mL/min/1.73 m², dabigatran was associated with a lower risk of AKI compared to warfarin (HR 0.60, 95% CI 0.37-0.98, p =0.040), but there was no statistically significant association between DOACS overall and AKI (HR 0.80 95% CI 0.64-1.01, p = 0.067). DOAC users tended to have a slightly higher risk of AKI compared to warfarin users among those with eGFR <30 mL/min/1.73 m² (HR 1.31, 95% CI 0.81-2.13, p = 0.271).

Table 1. Risk of acute kidney injury among patients receiving DOACs vs. warfarin overall and stratified by eGFR.

	DOAC* (N = 3,206) vs. Warfarin (N = 3,206)		Dabigatran (N = 852)† vs. Warfarin (N = 852)		Rivaroxaban (N = 1,325)† vs. Warfarin (N = 1,325)		Apixaban (N = 1,029)† vs. Warfarin (N = 1,029)
	HR (95% CI)	p-value	HR (95% CI)	p-value	HR (95% CI)	p-value	HR (95% CI)	p-value
Overall (N = 6,412)	0.79 (0.68-0.92)	0.003	0.70 (0.52-0.96)	0.025	0.83 (0.66-1.05)	0.114	0.86 (0.68-1.10)	0.233
eGFR ≥60 mL/min/1.73 m2 (N = 4,169)	0.72 (0.57-0.91)	0.007	0.66 (0.43-1.02)	0.063	0.79 (0.50-0.98)	0.037	0.67 (0.45-1.00)	0.052
eGFR 30-59 mL/min/1.73 m2 (N = 1,990)	0.80 (0.64-1.01)	0.067	0.60 (0.37-0.98)	0.040	0.95 (0.68-1.33)	0.764	1.01 (0.72-1.41)	0.956
eGFR <30 mL/min/1.73 m2 (N = 253)	1.31 (0.81-2.13)	0.271	1.52 (0.53-4.38)	0.440	1.48 (0.63-3.49)	0.368	1.23 (0.58-2.61)	0.593

^*DOACs, direct oral anticoagulants, including dabigatran, rivaroxaban, and apixaban.

^†Dabigatran (mean age 68, 76, and 79 years; 36%, 53%, and 70% were female in patients with eGFR ≥60, 30-59, and <30 mL/min/1.73 m², respectively); rivaroxaban (mean age 67, 76, and 80 years; 39%, 53%, and 64% were female in patients with eGFR ≥60, 30-59, and <30 mL/min/1.73 m², respectively); apixaban (mean age 70, 78, and 78 years; 44%, 56%, and 57% were female in patients with eGFR ≥60, 30-59, and <30 mL/min/1.73 m², respectively).

CI, confidence interval; eGFR, estimated glomerular filtration rate; HR, hazard ratio.

Data comparing the renal adverse effect of DOACs to that of warfarin are limited, particularly among patients with preexisting moderate to severe kidney dysfunction. This study, conducted in a large US-community-based cohort across the spectrum of eGFR, showed that DOACs were associated with a lower risk of AKI, but mainly among patients with relatively preserved kidney function (eGFR >60 mL/min/1.73 m²). Among patients with eGFR 30-59 mL/min/1.73 m², only dabigatran was associated with a lower risk of AKI compared to warfarin, and there was a trend towards higher AKI risk among all DOAC users compared to warfarin users with eGFR <30 mL/min/1.73 m². Coupled with case reports of fatal bleeding events in patients with severe renal impairment on dabigatran (4,5), caution may be needed when DOACs are prescribed in patients with advanced CKD.