DOACs May Up Post-TAVR Survival in A-fib: OCEAN-TAVI

The registry data suggest an advantage for the newer drugs versus a VKA, but RCT results would be more convincing, experts say.

DOACs May Up Post-TAVR Survival in A-fib: OCEAN-TAVI


(UPDATED) For patients with atrial fibrillation who undergo TAVR, prescription of a direct oral anticoagulant (DOAC) at discharge may translate into lower long-term mortality compared with a vitamin K antagonist (VKA), Japanese registry data suggest.

Both European and US guidelines currently lean toward VKAs for this population, Hideyuki Kawashima, MD (Teikyo University School of Medicine, Tokyo, Japan), and colleagues point out in their paper. “However, the requirements for high patient compliance, narrow therapeutic window, and multiple food and drug interactions hamper the use of VKAs, particularly among multimorbid and elderly patients,” they note, so DOACs are considered an alternative.

The guidelines currently exist as they do “because there haven’t been prospective randomized trials of DOACs after TAVR, and [warfarin] has the long track record of treatment for A-fib,” said Howard Herrmann, MD (Hospital of the University of Pennsylvania/Penn Medicine, Philadelphia).

Most trials investigating new versus old anticoagulants in A-fib patients have shown that DOACs are safer and more effective than VKAs, Herrmann observed, so it’s not unusual to see this pattern in TAVR patients with the arrhythmia. “The surprising thing in this study is the magnitude of the benefit—it’s much larger than any of us would expect and it doesn’t appear to be due primarily to differences in bleeding,” he commented to TCTMD, adding that these latest results, though “provocative, . . . should not change practice in the absence of confirmation by larger studies and/or randomized trials.”

He also cautioned that it’s important to wait for more data and to not assume where the results will land. “It’s not a given that one would expect the DOACs to be better or not better given the results of GALILEO, which suggested an increase in mortality with even low-dose rivaroxaban [compared to antiplatelets alone],” said Herrmann, an investigator for that prematurely halted trial. “So not all DOACS are likely to be the same, and we don’t really understand the mechanism for the increase in mortality in GALILEO.”

Toby Rogers, MD, PhD (MedStar Heart & Vascular Institute, Washington, DC), also commenting for TCTMD, said that it’s hard to interpret observational data like these “where there is clearly such a big influence of operator discretion with regards to therapy,” even with strong attempts at statistical adjustment.

Yusuke Watanabe, MD, PhD (Teikyo University School of Medicine), corresponding author of the new paper, said that the researchers also had not expected to see the observed survival benefit here. In his opinion, “DOACs should be the first choice for an antithrombotic regimen” in TAVR patients at high risk for bleeding, Watanabe told TCTMD via email. However, a “contraindication is severe renal dysfunction, [and] concomitant use of antiplatelet therapy should be minimum.”

The results, from the prospective OCEAN-TAVI registry, were published online yesterday in JACC: Cardiovascular Interventions.

Differences Across Antiplatelets, Too

Using data from OCEAN-TAVI, the researchers analyzed results for 2,588 patients who were successfully discharged after TAVR between October 2013 and May 2017. Among them, 403 individuals (mean age 84.4 years; 33.3% men) had A-fib and were prescribed an anticoagulant: 56.3% a DOAC and 43.7% a VKA. The mean CHA2DS2-VASc score was 5.1.

Patients taking DOACs tended to have lower HAS-BLED scores (mean 2.6 vs 2.9) and STS scores (mean 7.7 vs 9.5) than those on a VKA. They also were less apt to have NYHA class III or IV symptoms (51.5% vs 63.6%), history of stroke (10.6% vs 19.3%), and CAD (26.0% vs 35.2%). TAVR was more likely to be transfemoral for the DOAC group (90.7%) than the VKA group (76.7%). At discharge, DOAC patients were more likely to be on no antiplatelet therapy, whereas VKA patients were more likely to be on single or dual antiplatelet therapy (DAPT).

Median follow-up duration was 568 days. Bleeding and stroke rates were similar in the two groups.

All-cause mortality, however, was lower in the DOAC group than in the VKA group in two analyses: a multivariate Cox regression model (10.3% vs 23.3%; HR 0.39; 95% CI 0.20-0.75) as well as one using inverse probability of treatment weighting based on the propensity score (10.2% vs 20.6%; HR 0.53; 95% CI 0.25-0.96). Landmark analysis between 31 days and 2 years after discharge also showed lower all-cause mortality with DOACs.

Kaplan-Meier estimates showed significant differences in mortality by anticoagulant/antiplatelet strategy at 720 days after discharge. All-cause death was least common in the single antiplatelet-plus-DOAC arm and highest in the VKA-alone arm (log-rank P = 0.004).

All-Cause Mortality Among TAVR/A-fib Patients at 720 Days




Single Antiplatelet



Dual Antiplatelet



Anticoagulant Alone



The signal of lower mortality “might support the uniform use of DOACs in patients with atrial fibrillation undergoing TAVR who are indicated for anticoagulation therapy,” Kawashima et al say, though it still “warrants investigation in ongoing prospective randomized trials.”

Association or Causation?

Bleeding might be one explanation for their findings, Watanabe said, as their study found that event rates trended lower with DOACs. 

Chronic kidney disease (CKD) is also a possibility. VKAs are preferred in patients with advanced CKD, the researchers observe, whereas DOACs are contraindicated. In this study, though, the CKD prevalence and estimated glomerular filtration rates didn’t differ by anticoagulant type. But since renal dysfunction factors into HAS-BLED and STS scores, it may still be a confounder, they note.

“Another potential explanation is that the combination of anticoagulant and antiplatelet agents influenced long-term all-cause mortality,” Kawashima and colleagues continue. While the subgroup analyses do support this possibility, they write, the small number of events and observational study design mean that the findings must be replicated.

Herrmann, too, cited limitations: the study’s observational design, small numbers, lack of data on international normalized ratio (INR) for VKA-treated patients, differences in antiplatelet regimens, and no details on why patients were prescribed their particular anticoagulant. “There are likely a number of unmeasured confounders in this study,” he said.

Interestingly, “most of the divergence in the adjusted analysis is between 1 and 2 years,” Herrmann pointed out, adding that it’s unknown whether patients stayed on their initial anticoagulant during this time.

For Rogers, another issue is that without a control group of patients on antiplatelet therapy but not anticoagulation—the group that came out ahead in GALILEO—the data set doesn’t capture one option being pursued in real-world practice. “So it’s okay to say there’s no bleeding difference between these two groups, but we all know that if you put a lot of 80-year-olds on anticoagulation of any form you’re going to get more bleeding,” he observed.

Weighing the Options

So what to do while awaiting RCTs specific to both atrial fibrillation and TAVR?

Rogers said that for TAVR patients as a whole, he goes with aspirin for everyone and only adds on a second antiplatelet for people who have had a recent PCI or have some other indication; published guidance continues to call for DAPT. The guidelines also emphasize anticoagulation in A-fib, he noted. “That’s not wrong—it’s just I think these patients are elderly, they have many medical comorbidities, and it’s always about individualizing the decision-making. . . . You can’t paint these patients with a broad brush, because they are so complex, particularly [when it’s] this older, high-risk TAVR population.”

When balancing the risks of bleeding and thrombotic events, aspirin alone may suffice, Rogers suggested. “You just have to tailor that to the individual patients in front of you, even if they have A-fib.”

He pointed to LRT 2.0 trial, which he presented at CRT 2020, for context. That randomized study found that warfarin on top of aspirin reduced leaflet thrombosis versus aspirin alone, without excess bleeding—“but the really key thing there is that was in a younger, low-risk patient population,” Rogers stressed. “So targeted anticoagulation may be appropriate in low-bleeding-risk patients and most of those low-bleeding-risk patients are young.”

For patients who do need an anticoagulant, Herrmann said, most clinicians choose whichever one makes sense in particular circumstances but, based on results from POPular TAVI, do not give an antiplatelet agent unless there’s a reason to.

Elderly patients, for instance, are more apt than younger individuals to have undesirably low or high INR values on a VKA, Herrmann noted. This is worth recognizing here because patients undergoing TAVR tend to be older and have more comorbidities than the general A-fib population, Herrmann explained. Patients who have difficulty keeping track of their INR or who have wide variation in INR values may benefit from a DOAC. On the other hand, depending on insurance coverage, cost can also factor into the choice between DOAC and VKA therapy.

Kawashima and colleagues, as well as Herrmann and Rogers, cited several studies that may provide guidance when the data are in. ATLANTIS, comparing apixaban (Eliquis; Bristol-Myers Squibb) to standard of care (VKA or antiplatelet therapy) after TAVR, will include some A-fib patients. Then there’s ENVISAGE-TAVI AF, taking a look at edoxaban versus a VKA specifically in A-fib patients who’ve undergone TAVR. Another trial, highlighted by Herrmann, is taking an entirely different approach: WATCH-TAVR is looking at how percutaneous left atrial appendage closure with the Watchman device (Boston Scientific) stands up against medical therapy in the A-fib/TAVR setting.

  • The OCEAN-TAVI registry is supported by Edwards Lifesciences, Medtronic, Boston Scientific, and Daiichi-Sankyo Company.
  • Kawashima reports no relevant conflicts of interest.
  • Herrmann reports receiving institutional funding from Bayer as an investigator for GALILEO; research funding from Abbott, Boston Scientific, Edwards Lifesciences, and Medtronic; and consulting fees from Edwards Lifesciences and Medtronic.
  • Rogers reports being a proctor for and consultant to Edwards Lifesciences and Medtronic.
  • Watanabe reports being a clinical proctor for Edwards Lifesciences and Medtronic.