Edoxaban May Be Alternative to Warfarin After Surgical Valve Implantation or Repair

For patients requiring short-term anticoagulation after surgical aortic or mitral valve procedures, edoxaban (Savaysa; Daiichi Sankyo) is noninferior to warfarin for preventing thromboembolism and major bleeding, the Korean ENAVLE study suggests.

Although anticoagulation after surgical or transcatheter valve therapy is not routinely used in the United States and many other parts of the world unless patients are at high bleeding risk or have an indication for some other reason, it is common in Korea, noted Geu Ru Hong, MD, PhD (Yonsei University College of Medicine, Seoul, South Korea), in a featured clinical research presentation at the recent virtual American College of Cardiology 2020 Scientific Session.

“Usually in Korea even in low-risk [patients] we give at least 3months anticoagulation with warfarin,” Hong said. “If patients are at low risk of bleeding, I think it’s very important to preserve anticoagulation in the early postoperative period.”

He also pointed out that Koreans have some unique dietary issues such as regular consumption of soybeans and other foods high in vitamin K that can interfere with the efficacy of warfarin, making it of interest to look for alternatives among the direct oral anticoagulants (DOACs) that may provide adequate protection.

Commenting on the study, Daniel H. Steinberg, MD (Medical University of South Carolina, Charleston), said it fills an important gap.

“This is a very interesting topic. Something that we all struggle with in surgical and transcatheter valve therapies is what to do with these patients,” he observed. “Our guidelines are not crystal clear on that, and there is question at some centers whether to anticoagulate in the first place.”

Hint of Benefit, But Numbers are Small

From late 2017 to September 2019, Hong and the ENAVLE researchers randomly assigned 220 patients to receive edoxaban (60 mg or 30 mg once daily) or warfarin for 3 months after surgical aortic or mitral valve replacement or repair. Baseline characteristics were well balanced between the two groups, with a high proportion in each having hypertension, hyperlipidemia, and A-fib.

Randomization occurred an average of 5 to 9 days after surgery. In the warfarin group, dose adjustments were made to maintain an INR between 2.0 and 3.0. Outpatient clinic visits occurred at 2, 4, and 12 weeks. Echocardiography and CT scans were performed at the final follow-up visit.

The primary composite endpoint of death from any cause, clinical thromboembolic events, or asymptomatic intracardiac thrombosis occurred in four patients in the warfarin group and no patients in the edoxaban group (noninferiority P < 0.001). Of these primary endpoints, one was a left atrial thrombus after mitral valve replacement, one was an MI after aortic valve replacement, one was an aortic valve thrombus after aortic valve replacement, and one was a left atrial appendage thrombus after mitral valve repair.

We need larger numbers of patients to be assured of safety, and to be at least somewhat confident that there really is an efficacy benefit relevant to no therapy, which is generally the standard in the United States. Martin Leon

Additionally, edoxaban was found to be noninferior to warfarin for the primary safety endpoint of the occurrence of major bleeding according to ISTH criteria. This endpoint occurred in three patients in the edoxaban group and one in the warfarin group (noninferiority P = 0.013). Clinically relevant nonmajor bleeding, a secondary endpoint, occurred in one patient in each group, while major bleeding plus clinically relevant nonmajor bleeding occurred in four edoxaban patients and two warfarin patients; noninferiority was established for both endpoints.

Among the major bleeds reported were intracranial hemorrhage in a warfarin patient with aortic valve replacement, GI bleeding in two edoxaban patients with aortic valve replacement, and pericardial bleeding in one edoxaban patient with mitral valve repair. Clinically relevant nonmajor bleeding consisted of an organized hematoma in the anterior mediastinum in a warfarin patient after mitral valve replacement and hematuria in an edoxaban patient after mitral valve repair.

GI Protection Considerations

Given the higher risk of GI bleeding in the edoxaban group, Hong said GI protection is probably an important consideration in this population.

During the discussion, Alan Zajarias, MD (Washington University School of Medicine, St. Louis, MO), asked Hong if the event rate of the primary outcome with warfarin was as high as what the ENAVLE investigators expected and if a larger clinical trial might show a different result.

Hong said while it was a little higher than expected, it is consistent with other studies of warfarin in this setting. As for a larger trial, he said it would be good to have, but he expects the findings would likely be similar.

Nevertheless, session co-moderator Martin B. Leon, MD (NewYork-Presbyterian Hospital/Columbia University Irving Medical Center, New York, NY), said, “If we’re going to change the guideline recommendation and begin using a DOAC post-valve therapy, we need more data. We need larger numbers of patients to be assured of safety, and to be at least somewhat confident that there really is an efficacy benefit relevant to no therapy, which is generally the standard in the United States.”