Edoxaban: A Mixed Bag for TAVR Patients Needing Anticoagulation

A direct oral anticoagulant (DOAC) passed a test of efficacy but tripped up over bleeding rates when used for patients who came out of successful transcatheter aortic valve replacement (TAVR) and needed oral anticoagulation for their atrial fibrillation (Afib), the randomized ENVISAGE-TAVI AF trial showed.

The good news was that edoxaban (Savaysa) passed the hurdle of noninferiority against standard vitamin K antagonists (VKAs) like warfarin in terms of efficacy in intention-to-treat analysis. Net adverse clinical events (NACE) were 17.3 per 100 person-years with DOAC use compared with 16.5 per 100 person-years with VKAs (HR 1.05, 95% CI 0.85-1.31), allowing edoxaban to meet non-inferiority under a HR margin of 1.38.

Notably, ischemic stroke and all-cause mortality rates were numerically lower for the edoxaban group, reported George Dangas, MD, of Icahn School of Medicine at Mount Sinai, New York City, during a Hot Line session at the European Society of Cardiology (ESC) virtual meeting. The findings were simultaneously published in the New England Journal of Medicine.

The ENVISAGE group was congratulated by ESC session discussant Jean-Philippe Collet, MD, PhD, of Sorbonne University and Hôpital de la Pitié-Salpêtrière in Paris, for conducting the “only positive” DOAC trial in TAVR so far.

Collet’s own ATLANTIS trial had shown another DOAC, apixaban (Eliquis), to be no better than antiplatelet therapy or warfarin after TAVR — a finding that was consistent whether patients had an indication for oral anticoagulation or not — and worse yet, that ischemic event rates were even numerically higher comparing apixaban versus standard care.

The point of concern in ENVISAGE was major bleeding, which reached 9.7 vs 7.0 per 100 person-years (HR 1.40, 95% CI 1.03-1.91). Thus, edoxaban failed to meet noninferiority in safety, its disadvantage driven by more gastrointestinal bleeds despite similar incidences of administration of proton-pump inhibitors between groups.

The bottom line is that more bleeds puts a limit on potential patient eligibility for edoxaban, said Collet.

Several factors may account for the excess bleeds in the study, Dangas suggested, citing the warfarin arm’s suboptimal treatment — having spent less than 70% of the time within the therapeutic international normalized ratio (INR) range — and higher incidence of drug discontinuation (40.5% vs 30.2% for edoxaban) over the course of the study.

Furthermore, the bleeding curves started diverging only after the first few months, and the subgroup that had edoxaban dosing adjusted to 30 mg shared similar NACE and major bleeding rates as the VKA group.

“I would say that the hazard [with edoxaban] seems to be with long-term use and higher dose,” Dangas said during an ESC press conference.

Collet also warned against concomitant use of antiplatelets, as the nearly 60% of ENVISAGE participants taking concomitant antiplatelet therapy at any time after randomization (allowed at their physicians’ discretion) suffered more major bleeds on edoxaban vs warfarin.

The trial ultimately adds high-quality evidence on what the “sweet spot” is regarding the dosing and duration of anticoagulation for each patient such that they achieve the greatest reduction of ischemic events while minimizing bleeding, “which we know is never going to be zero,” commented Renato Lopes, MD, PhD, of Duke University Medical Center in Durham, North Carolina, during the ESC panel discussion.

Just hours before ENVISAGE was presented, ESC released the new European guidelines on valvular heart disease .

Collet suggested that more answers are needed before another guideline update is warranted to incorporate the trial’s findings. It is currently not known if all DOACs perform similarly after TAVR, for instance.

“If you pool your data you’ll have a better answer,” offered panelist Alec Vahanian, MD, of University of Paris Descartes, the guideline writing committee chair. “Let’s see within a couple of years.”

ENVISAGE-TAVI AF was conducted at 173 centers in 14 countries. Participants were people who had prevalent or incident Afib soon after receiving successful TAVR for aortic stenosis.

They were randomized to edoxaban 60 mg once a day (though 46.4% had their dose adjusted to 30 mg once daily on the basis of creatinine clearance, weight, or P-glycoprotein inhibitor use) or usual VKA in local practice (targeting an INR of 2.0-3.0). Randomization largely occurred within 3 days after TAVR.

The 1,426 study participants had a mean age of 82.1 and 47.5% were women. This was an intermediate-risk cohort given a mean STS score of 4.9%. The mean CHA₂DS₂-VASC score was 4.5, and 99% of people had Afib prior to TAVR.

NACE was a composite of all-cause death, myocardial infarction, ischemic stroke, systemic thromboembolism, valve thrombosis, and major bleeding. Follow-up lasted approximately 18 months.

ENVISAGE-TAVI AF’s open-label design left room for reporting bias, Dangas and colleagues acknowledged.

Additionally, as enrollment only finished in January 2020, the COVID-19 pandemic “affected the outpatient clinic follow-up routine and may have resulted in underassessment of laboratory data and mild-to-moderate clinical events,” the investigators wrote.

Dangas reported that a separate bleeding analysis is underway tying outcomes to a patient’s exact antithrombotic regimen after TAVR.