Oral Anticoagulation Post-AVR Safe but Not Routinely Needed: PARTNER II

Oral anticoagulation after aortic valve replacement—whether TAVR or SAVR—is safe when needed and is not associated with increases in adverse clinical events or changes in valve hemodynamics, a new analysis of PARTNER II data shows. For SAVR, the data also hint at a potential stroke reduction, though the researchers say the mechanism behind this finding isn’t clear.

Importantly, though, the choice to administer anticoagulants at discharge “was based on the clinical indications, as determined by the treating physicians,” senior investigator Raj Makkar, MD (Cedars-Sinai Medical Center, Los Angeles, CA), and colleagues caution.

“It’s been a hot issue whether or not we should anticoagulate patients after a TAVR procedure,” Makkar observed to TCTMD. “Some of the studies we did earlier showed that there was this finding of subclinical leaflet thrombosis on almost 15% of patients after TAVR, and there was some association of this finding with TIAs, [though] not necessarily with stroke.”

The question, then, is whether short-term anticoagulation might provide long-term protection against valve degeneration.

Based on these latest results, “I think we can step back and we can take the approach of not rushing to anticoagulate after TAVR” unless there’s another indication, Makkar said. “We would not withhold anticoagulation for other clinical reasons and we would not suggest routine use . . . just for the purpose of the valve.”

I think we can step back and we can take the approach of not rushing to anticoagulate after TAVR. Raj Makkar

Harold L. Dauerman, MD (University of Vermont Medical Center, Burlington), told TCTMD that despite being nonrandomized, the study is interesting and provides confirmation of prior research. Both this evidence and the prematurely halted GALILEO trial should make clinicians wary of routinely prescribing anticoagulants post-TAVR, he urged. “Our own data suggests that there is biologically a reason to avoid that strategy, as these patients may be at a higher intrinsic risk of bleeding, based on platelet- and nonplatelet-mediated factors.”

American Heart Association/American College of Cardiology TAVR guidelines currently say 3 months of anticoagulation is reasonable when patients don’t have an underlying need for the medication (such as A-fib or prior stroke) and when bleeding risk isn’t a concern, he said. “I think that strategy still can be utilized, but this is really an area where we lack clear data to even give a guideline statement.”

The study, led by Tarun Chakravarty, MD (Cedars-Sinai Medical Center), was published online ahead of print in the September 3, 2019, issue of the Journal of the American College of Cardiology.

Very Few Differences Seen

Chakravarty and colleagues looked at 4,832 patients in the PARTNER II randomized trials and registries, of whom 3,889 underwent TAVR and 943 had SAVR. They adjusted for valve size, annular diameter, comorbid A-fib, and ejection fraction at the time of hemodynamic assessment; after adjustment, there were no differences in aortic valve mean gradients or areas based on anticoagulation at discharge after either procedure. However, patients not given anticoagulants were more likely to see an increase in mean gradient > 10 mm Hg from 30 days to 1 year (2.3% vs 1.1%; P = 0.03).

Makkar pointed out that, though statistically significant, the absolute difference in mean gradient is quite small.

That mean gradients held steady for such a large segment of patients is “remarkable,” Dauerman agreed. “Overall, 98% of people had a superb hemodynamic outcome over the first year. . . . And no patient had significant structural valve deterioration in either arm.” As such, improving hemodynamic performance should not be the goal of anticoagulation, he added, explaining that this can inform the risk-benefit ratio: “If your mean gradient goes up by 12 mm Hg, I’m not sure that’s worse than a minor bleed that requires hospitalization.”

For TAVR, anticoagulation use was not independently associated with adverse outcomes. For SAVR, only stroke showed a link, with discharge on an anticoagulant reducing risk by 83% (HR 0.17; 95% CI 0.05-0.60).

Makkar stressed that the risk-benefit tradeoff demonstrated here only applies to the older patients studied in PARTNER II, who had a mean age of 82 years. “[It] may be completely different in a 65-year-old patient who is undergoing a TAVR procedure, where the risk of bleeding may be very low and there may be benefit. We cannot make any assumptions. . . . We need a clinical trial,” he said.

Dauerman, too, called for a cautious approach. He said that, going forward, the ATLANTIS trial will be “very important—we really do need randomized trial information. And if nothing more, the halting of GALILEO should raise clinicians’ ears to the fact that we don’t know the answer.”

Writing in an editorial, Philippe Pibarot, DVM, PhD (Université Laval, Québec City, Canada), along with C. David Mazer, MD, and Subodh Verma, MD, PhD (both from St. Michael’s Hospital/University of Toronto, Canada), advise clinicians to make decisions on a case-by-case basis.

“The data from the present study and recent previous studies do not support systematic anticoagulation therapy for 3 to 6 months after AVR,” they write. “It may be preferable to adopt an individualized approach for early and long-term anticoagulation based on the presence or absence of procedural and patient factors that increase the risk of thromboembolism, valve thrombosis, or bleeding for thromboembolisms or valve thrombosis.”