No Increased Risk of Serious Liver Injury Seen With NOACs

Liver status should not be a major part of the decision when choosing an oral anticoagulant for stroke prevention in A-fib, researchers say.

No Increased Risk of Serious Liver Injury Seen With NOACs

Patients newly diagnosed with nonvalvular A-fib do not have a greater risk of serious liver injury if they start taking a non-vitamin K antagonist oral anticoagulant (NOAC) instead of a vitamin K antagonist (VKA), a new study shows.

That finding was consistent regardless of whether patients had a history of liver disease at baseline, which is reassuring, lead author Antonios Douros, MD, PhD (Jewish General Hospital, Montreal, Canada), and colleagues report in a study published in the March 13, 2018, issue of the Journal of the American College of Cardiology.

“VKAs are known to be safe with respect to this outcome,” Douros told TCTMD, “so NOACs should also be considered safe with respect to this outcome.”

Senior author Christel Renoux, MD, PhD (Jewish General Hospital), added: “Risk of liver injury should not be a major part of the decision to initiate a NOAC or VKA.”

Concerns about liver risks with the NOACs linger and, in fact, liver injury doomed development of the first NOAC, ximelagatran. There are conflicting results in the literature, however. Some case reports and pharmacovigilance studies have suggested a potential risk of hepatotoxicity with the NOACs, but a prior observational study demonstrated just the opposite: a lower risk of liver injury with NOACs versus VKAs. But the latter analysis had some limitations and did not stratify patients by baseline liver disease status.

‘Clinicians Should be Reassured’

Seeking to address some of those issues, Douros et al examined data from administrative health insurance databases in the province of Quebec on 51,887 patients newly diagnosed with nonvalvular A-fib between 2011 and 2014; 7.3% had prior liver disease.

During follow-up, the rate of serious liver injury—requiring hospitalization or causing death—was numerically lower in patients taking NOACs versus those taking VKAs, both for those without a history of liver disease (3.9 vs 4.5 per 1,000 person-years) and for those with prior liver disease (24.5 vs 44.8 per 1,000 person-years).

After adjustment, however, the difference was not statistically significant in the former group (adjusted HR 0.99; 95% CI 0.68-1.45) or the latter group (HR 0.68; 95% CI 0.33-1.37). The results were consistent regardless of duration of NOAC use and type of NOAC.

Although there is a signal toward a decreased risk of liver injury with NOACs in the subset of patients with prior liver disease, Renoux said firm conclusions cannot be made because the finding was based on a low number of events and the estimates are not precise.

She noted, too, that multiple observational studies, including this one, have shown that clinicians tend to prescribe VKAs to sicker patients because of the long history of use and well-established safety profiles of the agents. That bias could artificially make NOACs look better in terms of adverse events.

That said, Renoux stated, “We are confident that there is no greater risk with NOACs versus VKAs in patients with or without liver disease.”

Douros indicated that the findings of this and other studies might mitigate clinicians’ tendency to prescribe VKAs over NOACs in patients with a perceived risk of liver injury. “We think that with accumulating data on [NOAC] safety there’s going to be a shift there,” he said.

In an accompanying editorial, Mark Crowther, MD, and Wendy Lim, MD (McMaster University, Hamilton, Canada), point out some limitations of the study—it was restricted to patients with nonvalvular A-fib, it was not randomized, and the outcome was assessed using coding data—but say that the statistical approach was rigorous.

Thus, “clinicians should be reassured by the results of this analysis,” they say. “Outside of bleeding, which is a predictable complication of any anticoagulant, [NOACs] appear highly safe and have become the preferred agents for the prevention of thrombotic complications in many patients who have, or are at risk of, this complication.”


Renoux said further studies are needed to evaluate the risk of liver injury in individual NOACs because the current analysis was underpowered to do so.

  • The study was supported by an infrastructure grant from the Canadian Foundation for Innovation, and the database was acquired using unrestricted funding from Bayer Pharma AG.
  • Douros reports having received a research fellowship from the German Research Foundation (Deutsche Forschungsgemeinschaft).
  • Renoux reports no relevant conflicts of interest.
  • Crowther reports having participated in data safety monitoring boards for Bayer and Daiichi; having received personal fees and/or having participated in advisory boards for Shionogi, Octapharma, and BMS Canada; having received institutional funding from Bayer and Leo Pharma; having received fees for educational material preparation and/or presentations for Bayer, Pfizer, and Alexion; and having individual stock ownership in Alynlam.
  • Lim reports having served as a consultant and/or on advisory boards and receiving honoraria (for CME) from Leo Pharma, Pfizer, and Portola.

We Recommend