Initiating the oral anticoagulant apixaban in patients undergoing kidney dialysis who also had newly diagnosed atrial fibrillation (AF) was associated with fewer strokes and deaths but higher rates of bleeding events, a new study found.

Wolfgang C. Winkelmayer, MD, ScD, of Baylor College of Medicine in Houston, and colleagues conducted a retrospective cohort study to weigh the benefits and risks of oral anticoagulants in this population.

To improve comparability, the researchers used propensity score matching to balance baseline characteristics between patients who received apixaban and those who did not. This approach addressed potential bias toward prescribing apixaban to healthier patients. For example, after matching, prior bleeding hospitalization rates were similar between groups (15.7% in the control group vs 15.1% in the apixaban group) compared with larger imbalances before matching.

In intention-to-treat analyses, apixaban use was associated with a 25% lower rate of ischemic stroke and a 24% lower rate of a composite outcome of thromboembolic events and cardiovascular deaths.

Conversely, patients who received apixaban had a 55% higher rate of hemorrhagic stroke and a 29% higher rate of clinically important bleeding. The hazard ratio for all-cause mortality was 0.61 (95% CI, 0.56-0.67).

The study was published in the American Journal of Kidney Diseases.

Adding Clarity Around Apixaban Use

These findings add to the sparse observational evidence guiding apixaban use in patients undergoing kidney dialysis. Major clinical trials of apixaban did not include patients with end-stage renal disease, as noted in the drug’s US labeling.

As a result, clinicians have relied heavily on observational data when managing AF in people undergoing dialysis. The current study includes both intention-to-treat and as-treated analyses to better inform risk-benefit discussions.

Winkelmayer told Medscape Medical News that these data may help guide difficult conversations with patients about whether to proceed with oral anticoagulation. “Physicians can have meaningful discussions with their patients and consider whether the patient is more scared of a stroke or of a bleeding event and incorporate this into their decision-making,” he said.

Winkelmayer, section chief of nephrology at Baylor College of Medicine, has studied AF in patients undergoing dialysis for several years. His earlier research estimated that AF prevalence among US hemodialysis patients exceeded 10% in 2006, whereas more his recent rhythm monitoring studies suggested that up to one third of patients may experience clinically meaningful AF episodes over 6 months.

Despite this, clinicians often hesitate to prescribe anticoagulants because of bleeding concerns. 

Use of oral anticoagulants in patients undergoing dialysis has increased but remains relatively low. Winkelmayer and colleagues previously reported that among patients on hemodialysis with newly diagnosed AF, oral anticoagulant initiation within 90 days rose from 20.5% in 2007 to 34.1% in 2020. In contrast, anticoagulation rates in the general population increased from 56.3% to 64.7% between 2011 and 2020, driven by the growing use of direct oral anticoagulants (DOACs).

Study Design

Winkelmayer and colleagues used data from the US Renal Data System, including Medicare Part D claims on filled prescriptions, to identify patients with a first AF diagnosis between January 1, 2014, and December 31, 2019. 

Effectiveness outcomes included ischemic stroke (including transient ischemic attack), myocardial infarction, systemic embolism, cardiovascular death, and all-cause mortality. A composite cardiovascular outcome of time to the earliest among ischemic stroke, myocardial infarction, systemic embolism, and cardiovascular death was also assessed.

Safety outcomes included intracranial hemorrhage, major bleeding, and clinically important bleeding (composite of major bleeding and other bleeding requiring hospitalization).

Treatment discontinuation was common: 33% of patients did not refill their initial 30-day prescription, and the median duration of apixaban treatment was 59 days.

In the as-treated analysis, the rate of clinically important bleeding was 43% compared with 29% in the intent-to-treat analysis.

All-cause mortality was reduced by 39% in intent-to-treat analyses but by 56% in as-treated analyses.

The magnitude of the all-cause mortality reduction raises the possibility of residual confounding, the authors noted, although similar trends have been observed in other anticoagulation studies, such as the COMPASS trial of rivaroxaban.

Trials Underway

There is strong interest in obtaining clinical trial data to clarify the role of anticoagulation in patients undergoing dialysis. 

The RENAL-AF trial comparing apixaban with warfarin was stopped early due to slow enrollment, with only 154 participants out of a target of 762.

However, a planned randomized controlled trial in Canada may provide more definitive evidence, Winkelmayer said. Funded by the Canadian Institutes of Health Research, the trial will compare apixaban with no anticoagulation in patients undergoing dialysis.

Primary outcomes will focus on stroke, systemic embolism, and bleeding events, said Ziv Harel, MD, of St. Michael’s Hospital in Toronto, Ontario, Canada, in an email exchange with Medscape Medical News. The trial is expected to begin later this year.

This trial builds on a successful SAFE-D pilot study comparing warfarin, apixaban, and placebo. Despite being conducted during the COVID pandemic, the trial enrolled 151 patients, with 83% completing follow-up.

Winkelmayer said this new trial could provide more authoritative data within a few years.

“But for now, we are really dependent on high-quality observational studies,” he said.

Expert Perspective

Paul M. Palevsky, MD, who was not involved in the study, said the findings reinforce the importance of ensuring that patients undergoing dialysis who have newly diagnosed AF understand the risk-benefit profile.

“The big takeaway from this is we need to talk to our patients about the risk of stroke, evaluate them on an individual basis for their various risks of complication, and have a very frank discussion about whether they want to be on anticoagulation,” he told Medscape Medical News.

Palevsky described the study as well conducted but emphasized the persistent challenge posed by residual confounding in observation research, “No matter how sophisticated the statistical hoops that the authors manage to jump through,” he said.

In an accompanying editorial, An S. De Vriese, MD, PhD, Ghent University, Ghent, Belgium, called for coordinated international efforts to conduct adequately powered trials in this complex population.

“Whether we opt for (kidney failure requiring hemodialysis) tailored risk stratification, DOAC dose reduction, or a new class of anticoagulants, we cannot resign ourselves to seeing our (kidney failure requiring hemodialysis) patients caught between the ‘rock’ of thrombosis and the ‘hard place’ of bleeding,” De Vriese wrote.

This study was funded by a National Institutes of Health grant. Winkelmayer also reported receiving salary and research support through the endowed Gordon A. Cain Chair in Nephrology at Baylor College of Medicine. Additional author support came from a philanthropic gift from Dr Harold and Ms Martha Ann Selzman to Baylor College of Medicine.

Winkelmayer reported having served as a scientific advisor or consultant to Akebia, Anthos, AstraZeneca, Bayer, Boehringer Ingelheim, Cadrenal, City, GlaxoSmithKline, Idorsia, Merck, Natera, Novartis, and Vera; serving on clinical trial committees for Akebia, Bayer, Merck, and NephroNet; and receiving speaking honoraria from Pharmacosmos. Tara I. Chang reported having served as a consultant for Bayer, Janssen Pharmaceuticals, Novo Nordisk, Fresenius Medical Care, Tricida, Gilead, and AstraZeneca and having received salary support from CSL Behring through funds paid directly to Stanford University. The other authors declared having no relevant financial interests. 

Palevsky and De Vries reported having no relevant financial disclosures. 

Kerry Dooley Young is a freelance journalist based in Washington, DC. She has covered medical research and healthcare policy for more than 20 years.