ENVISAGE-TAVI AF: Edoxaban Disappoints in TAVI Setting

“The enthusiasm for NOACs in this particular population is maybe a little bit less than it was before,” Jurriën ten Berg says.

ENVISAGE-TAVI AF: Edoxaban Disappoints in TAVI Setting


(UPDATED) Hopes are waning for the use of direct oral anticoagulants (DOACs) in the post-TAVI setting, after the ENVISAGE-TAVI AF trial demonstrated an excess risk of bleeding with edoxaban (Savaysa; Daiichi Sankyo) versus vitamin K antagonists (VKAs) in patients with atrial fibrillation (AF) treated with transcatheter valves.

Edoxaban was noninferior to VKAs in terms of net adverse clinical events (NACE), but the rate of ISTH major bleeding was higher (9.7 vs 7.0 per 100 person-years), failing to meet the bar for noninferiority (P = 0.93). The difference was mostly attributed to more GI bleeds with edoxaban, including one that was fatal. Rates of intracranial hemorrhage and fatal bleeding were similarly low in both groups.

Exploratory analyses suggested that the risks and benefits with edoxaban were better balanced in patients who met criteria for using a half dose of the medication and those who were not also taking antiplatelet therapy, two groups in which use of the DOAC might be reasonable or appropriate, George Dangas, MD (Icahn School of Medicine at Mount Sinai, New York, NY), reported during a Hot Line session at the virtual European Society of Cardiology Congress 2021.

“At the same time, more research is required, including our own bleeding-oriented analysis that’s going to look more into the details of how the major bleeding occurred: not just the types, etc, but also [whether it was] related to any ingestion of a study drug,” he told TCTMD.

For Jurriën ten Berg, MD, PhD (St. Antonius Hospital, Nieuwegein, the Netherlands), the results fell short of expectations. “To be honest, I was a little bit disappointed by the results,” he commented to TCTMD, particularly in light of the ATLANTIS trial, which showed no benefit for another DOAC—apixaban (Eliquis; Bristol-Myers Squibb)—over standard care but also demonstrated no increase in bleeding.

“It’s not the end of NOACs, but the enthusiasm for NOACs in this particular population is maybe a little bit less than it was before reading the two trials,” he said, referring to ATLANTIS and ENVISAGE-TAVI AF.

The new trial findings were published simultaneously online in the New England Journal of Medicine.


Roughly one-third of patients undergoing TAVI have either preexisting AF or develop it after the procedure, and thus have an indication for oral anticoagulation. It remains unclear whether DOACs or VKAs are the preferred option, although Cohort B of the POPular TAVI trial showed that when oral anticoagulation is needed, it should be used alone and not in combination with antiplatelet therapy. A recommendation for this leaner antithrombotic regimen made its way into a European consensus statement, written by a group led by ten Berg, earlier this year.

In the pivotal trials for stroke prevention in patients with AF, the DOACs were shown to be at least as effective as warfarin, with lower risks of major bleeding. In ENGAGE AF-TIMI 48, edoxaban was noninferior to warfarin for the prevention of stroke and thromboembolic events, with lower rates of bleeding and CV death; that trial, though, didn’t include patients undergoing TAVI.

To be honest, I was a little bit disappointed by the results. Jurriën ten Berg

ENVISAGE-TAVI AF, conducted at 173 centers across 14 countries in North America, Europe, and Asia, randomized 1,426 patients (mean age 82.1 years; 47.5% women) who underwent successful TAVI and who required chronic oral anticoagulation for AF to edoxaban 60 mg/day or VKA therapy. Patients received an edoxaban dose of 30 mg/day if they had a creatinine clearance of 15 to 50 mL/min, had a body weight 60 kg or less, or used certain P-glycoprotein inhibitors (depending on geographic location). Nearly half of patients received that lower dose (46%) and received concomitant antiplatelet therapy (48%). Almost all (99%) had AF before undergoing the procedure.

The primary efficacy outcome was a NACE endpoint encompassing all-cause death, MI, ischemic stroke, systemic thromboembolism, valve thrombosis, and ISTH-defined major bleeding. Through a median follow-up of about 1-and-a-half years, the rate per 100 person-years was 17.3 with edoxaban and 16.5 with VKA, a difference that met criteria for noninferiority (P = 0.01).

The primary safety outcome was ISTH-defined major bleeding (clinically overt bleeding associated with reduced hemoglobin, blood transfusion, symptomatic bleeding at a critical site, or a fatal outcome). Risk was a relative 40% greater with edoxaban (HR 1.40; 95% CI 1.03-1.91), driven mostly by a higher number of GI bleeds (5.4 vs 2.7 per 100 person-years; HR 2.03; 95% CI 1.28-3.22).

There were no differences between trial arms in ischemic stroke, all-cause death, MI, or other clinical outcomes; no cases of valve thrombosis were detected in either group.

What Explains the Bleeding?

Asked why edoxaban would be associated with more major bleeding versus VKA therapy in a TAVI population but not the broader AF population, Dangas pointed to the older age of the patients being treated for aortic stenosis—who were about a decade older, on average, than patients in the AF trials—and to a heavier comorbidity burden.

“Acquired von Willebrand’s disease and arteriovenous malformations may also contribute to gastrointestinal bleeding in patients with severe aortic stenosis,” the paper notes.

Ten Berg agreed that the older age of the patients undergoing TAVI likely explains the bleeding findings. “It seems to be that especially elderly patients on NOACs bleed from their intestines,” he said, noting that as a whole, the class of drugs has been shown to carry a risk of GI bleeding.

Renato Lopes, MD, PhD (Duke Clinical Research Institute, Durham, NC), underscored that point during a panel discussion following Dangas’ presentation, saying the greater risk of GI bleeds with DOACs versus VKAs probably has to do with how they’re metabolized—through the GI tract for DOACs and the liver for VKAs. “So finding a 40% increase in GI bleeding, [in] my view was somehow expected.”

Patients undergoing TAVI could represent a group particularly at risk for bleeding with any treatment, Dangas told TCTMD, suggesting that the findings of ENVISAGE-TAVI AF highlight the need to revisit the pharmacodynamics and dosing of DOACs in older age groups that were not the focus of earlier trials.

How to Choose an Anticoagulant

While awaiting a more-detailed look at bleeding within the trial, Dangas noted that the default strategy for post-TAVI anticoagulation remains VKA therapy. If one were to use edoxaban, he added, “you have to consider the lower dose of edoxaban more so in order to be protected from bleeding.”

Still, he stressed, there is a need for continued research into the efficacy and safety of each of the DOACs in elderly patients.

The main message for ten Berg is “that the risk for bleeds is much larger than the risk for thrombotic events” in this setting. “So we should reduce the antithrombotic regimen as much as possible.” What that means is that if antiplatelet therapy is indicated, it should be a single agent, and if oral anticoagulation is indicated, it should be the anticoagulant alone, without an antiplatelet agent like aspirin or clopidogrel. “We should take care more of the bleeding risk than of the thrombotic risk,” he said.

When it comes to choice of anticoagulant, ten Berg said that if patients are tolerating either a VKA or DOAC prior to TAVI, they should usually remain on that particular agent after the procedure. For patients with new AF after TAVI, the choice is less clear, particularly because ATLANTIS showed that use of a DOAC had similar efficacy without an increase in bleeding compared with VKA therapy. However, those with a high risk for GI bleeds—because of a prior history of ulcer, for instance—should probably be given a VKA, ten Berg advised.

Regardless of the agent chosen, Lopes said, “it’s very important to find what I’ve been calling the antithrombotic sweet spot.” For each individual patient’s clinical situation, he explained, “it’s important to find the right dose [and] the right combination of agents for the right duration that can give us the greatest reduction in ischemic events at a minimal cost of bleeding, which we know is never going to be zero. But if we can minimize bleeding risk and preserve the efficacy by finding the sweet spot, I think that’s what we should do [for] our patients. And this trial adds to the body of evidence with high-quality evidence to better inform what this sweet spot might be.”

Commentary delivered by Jean-Philippe Collet, MD (Pitié-Salpêtrière Hospital, Paris, France), principle investigator of ATLANTIS, delved into the many differences between that trial and ENVISAGE-TAVI AF, particularly the higher rates of concomitant antiplatelet therapy, dose adjustment of the DOAC, and discontinuation of the study drug in the latter study, as well as the discrepant findings regarding bleeding risk.

Collet said applying the results of ENVISAGE-TAVI AF to practice is “tricky” because of the excess bleeding risk with edoxaban, and raised a question about whether all DOACs will provide similar results after TAVI. “Clearly, NOAC safety depends on the concomitant use of antiplatelet therapy and dosing,” he concluded, adding that he’s not sure whether the time has come for guideline updates.

To that point, Alec Vahanian, MD (Universite ́ de Paris, France), one of the chairs of the newly updated European valvular heart disease guideline, stressed the importance of data to support recommendations, adding that the questions raised by Collet and others “have to be answered before moving to a guideline recommendation.” He suggested that pooling the data from the two trials might help provide clarity.

Todd Neale is the Associate News Editor for TCTMD and a Senior Medical Journalist. He got his start in journalism at …

Read Full Bio
  • The trial was funded by Daiichi Sankyo.
  • Dangas reports institutional research grants/contracts from Bayer, Boston Scientific, Daiichi Sankyo, and Medtronic.