Kuno T, Takagi H, Ando T, et al. Oral Anticoagulation for Patients With Atrial Fibrillation on Long-Term Hemodialysis. J Am Coll Cardiol. 2020 Jan 28;75(3):273-285. doi: 10.1016/j.jacc.2019.10.059. (Systematic review)

BACKGROUND: Patients on long-term dialysis are at increased risk of bleeding. Although oral anticoagulants (OACs) are recommended for atrial fibrillation (AF) to reduce the risk of stroke, randomized trials have excluded these populations. As such, the net clinical benefit of OACs among patients on dialysis is unknown.

OBJECTIVES: This study aimed to investigate the efficacy and safety of OACs in patients with AF on long-term dialysis.

METHODS: MEDLINE and EMBASE were searched through June 10, 2019, for studies that investigated the efficacy and safety of different OAC strategies in patients with AF on long-term dialysis. The efficacy outcomes were ischemic stroke and/or systemic thromboembolism, all-cause mortality, and the safety outcome was major bleeding.

RESULTS: This study identified 16 eligible observational studies (N = 71,877) regarding patients on long-term dialysis who had AF. Only 2 of 16 studies investigated direct OACs. Outcomes for dabigatran and rivaroxaban were limited to major bleeding events. Compared with no anticoagulants, apixaban and warfarin were not associated with a significant decrease in stroke and/or systemic thromboembolism (apixaban 5 mg, hazard ratio [HR]: 0.59; 95% confidence interval [CI]: 0.30 to 1.17; apixaban 2.5 mg, HR: 1.00; 95% CI: 0.52 to 1.93; warfarin, HR: 0.91; 95% CI: 0.72 to 1.16). Apixaban 5 mg was associated with a significantly lower risk of mortality (vs. warfarin, HR: 0.65; 95% CI: 0.45 to 0.93; vs. apixaban 2.5 mg, HR: 0.62; 95% CI: 0.42 to 0.90; vs. no anticoagulant, HR: 0.61; 95% CI: 0.41 to 0.90). Warfarin was associated with a significantly higher risk of major bleeding than apixaban 5 min/2.5 mg and no anticoagulant (vs. apixaban 5 mg, HR: 1.41; 95% CI: 1.07 to 1.88; vs. apixaban 2.5 mg, HR: 1.40; 95% CI: 1.07 to 1.82; vs. no anticoagulant, HR: 1.31; 95% CI: 1.15 to 1.50). Dabigatran and rivaroxaban were also associated with significantly higher risk of major bleeding than apixaban and no anticoagulant.

CONCLUSIONS: This meta-analysis showed that OACs were not associated with a reduced risk of thromboembolism in patients with AF on long-term dialysis. Warfarin, dabigatran, and rivaroxaban were associated with significantly higher bleeding risk compared with apixaban and no anticoagulant. The benefit-to-risk ratio of OACs in patients with AF on long-term dialysis warrants validation in randomized clinical trials.

Discipline Area Score
Nephrology 7 / 7
Family Medicine (FM)/General Practice (GP) 6 / 7
General Internal Medicine-Primary Care(US) 6 / 7
Internal Medicine 6 / 7
Cardiology 6 / 7
Hemostasis and Thrombosis 6 / 7
Comments from MORE raters

Cardiology rater

A very thought-provoking study. This strongly suggests that a randomized placebo-controlled trial using warfarin, placebo, and, possibly, a non-vitamin K agonist should be done.

Cardiology rater

This study provides an important message about the limitations of warfarin and direct-acting oral anticoagulants to prevent stroke and thromboembolic events in patients with atrial fibrillation on dialysis for end-stage kidney disease.

Hemostasis and Thrombosis rater

Findings have very limited validity because of significant potential for selection bias and confounding (observational study).

Nephrology rater

Important summary of observational data raising the issue of OAC in the dialysis population.

Nephrology rater

Very important information for a vexing clinical problem. Unfortunately, data are still insufficient for guideline recommendations, but this meta-analysis is an important first step.

Nephrology rater

Very nice summary of the individual studies that have accrued, confirming that in aggregate there is no suggestion of reduction in risk for stroke in ESRD. It's ethically acceptable not to offer this prophylactic strategy to patients treated with maintenance dialysis, given the absence of benefit and the likelihood of harm suggested by these observational studies and the absence of randomized evidence. It's also ethically acceptable to conduct placebo-controlled trials in this area.

Nephrology rater

It seems from this that we lack clear, substantive, high-quality evidence on whether or not to use anti-coagulation for AF in ESRD. But from what we have, it does not seem wise to advise anti-coagulation for these patients. If I were on dialysis and had AF, I would not choose to be anti-coagulated.
Comments from EvidenceAlerts subscribers

Dr. Rajeev Gupta (3/1/2020 2:47 PM)

To me, the message is very clear: No OACs in victims with ESRD/dialysis. An RCT is sorely needed as apixiban 5mg shows a mortality reduction.