Role of Warfarin in Preventing Stroke in AFib Patients

Manesh Patel, MD: Rob, maybe this is a good time to switch back to where we were in 2010 or 2006, even when we were planning these trials and thinking about what Chris started us with, which was there’s an unmet need. Even though warfarin works well, people weren’t taking it. How did we decide how close we had to get to warfarin? Tell us a little about the thinking there. One of the things I think you probably have the unique perspective on is that when we started it, nothing had beat warfarin for 50 years. I used to say it was the Mohammad Ali. Even though we tried a bunch of stuff to beat it, we couldn’t. These new drugs were coming. What were we going to have to show to get people to use them?

Robert M. Califf, MD, MACC: It was really a problem and not unique in this area, but this was really a hot spot because we knew there were so many problems with the impending epidemic and people being unhappy with warfarin even though it works, as you said. Chris and I both had parents who were dealing with this. People don’t like to have to change their diets and all the monitoring that has to be done. It’s painful. Yet on the other hand, if you have a treatment that’s highly effective and reduces mortality, you shouldn’t just replace it willy-nilly without proof that you’re at least as good.

As almost everyone now knows, when you look at a noninferiority-type trial to prove that you’re at least as good, it takes a big trial. Of course, if you think your treatment is a whole lot better for the primary outcome, it doesn’t take such a big trial. But the real hope with these drugs is that they would provide the benefit of warfarin without all the complexity and some of the risk. So we actually had to develop a plan with the FDA in a whole series of meetings.

This was very entertaining for me because little did I know I’d end up being FDA commissioner. But we ended up with a manuscript really that had FDA coauthors that laid out what you had to do to meet the bar. Then the race was on to see who could meet the bar. It was also interesting that there were several options, 1 of which was an unblinded trial and that was chosen by 1 of the new oral anticoagulants. The other would be a requirement for a double-blinded drug, a much tougher hurdle logistically at a much higher cost. And 2 or 3 of the agents took that approach.

Manesh Patel, MD: Yeah.

Christopher Granger, MD: And the third, Rob, was using a non-anticoagulative comparator, which was done in AVERROES, where apixaban was compared with aspirin. And the FDA wasn’t particularly enthusiastic about that. The results of that trial are important because they showed NOACs [novel oral anticoagulants] are clearly better than aspirin and not that much higher risk, for example, for intracranial hemorrhage.

Manesh Patel, MD: I think that’s a key point, Chris. And going back to 1 of the things Rob just said. When we were planning these trials, I remember you asked me, “Are you sure you’re going to do it double-blinded, Manesh?” And I remember saying, “Well, if you’re a patient, and you got randomized to open-label warfarin, would you stay in the study? One of the fears we had was about keeping patients in the study and some biases that happened with the open-label versus closed or blinded studies. There are pluses and minuses on both ends, but I think this was something we struggled with a lot.

Robert M. Califf, MD, MACC: It takes a lot of discipline, and the patient part can go both ways, right?

Manesh Patel, MD: Yeah.

Robert M. Califf, MD, MACC: Because a new drug has risk, some people gravitate to it, hoping that it’s going to be better. But the fact is that 90% of drugs that get into clinical trials never make it to market because they’re not better. Doing double-blinded warfarin is not a trivial thing because you have to have sham up titrations and down titrations. It’s very complicated. Then we had this very difficult problem of what is the community standard that should be used to judge whether the warfarin arm is adequate to be a comparator. That was a hotly contested issue because some places seem to be able to do it almost perfectly, but 95% of places were certainly less than perfect. The average, you’d say, is pretty mediocre.

Manesh Patel, MD: Yeah.

Bernard J. Gersh, MBChb, DPhil, MACC: One of the things that came out of the AVERROES trial that is really important, and that we’ve learned particularly in the last 5 years, is the myth that aspirin works for atrial fibrillation [AFib], for thromboembolism prophylaxis. It does not. If you look at the 5 trials of aspirin versus warfarin, or placebo at least, 1 is a very flawed trial and 25 years ago showed a benefit for aspirin. If you look at the other 4 trials, there was no benefit from aspirin. What AVERROES showed was that the bleeding rate was slightly less, I think. Chris, right?

Christopher Granger, MD: Slightly.

Bernard J. Gersh, MBChb, DPhil, MACC: Slightly less than warfarin. I was involved in an editorial several years ago in the European Heart Journal, in which we said the misperception that aspirin alone is effective for thromboembolism prophylaxis in atrial fibrillation is 1 of the reasons we so underutilize warfarin. Because patients don’t like warfarin and the physician could say, “Well, this is not quite as good as warfarin, but we can give that to you.” It’s ineffective.

Manesh Patel, MD: That’s a great point.

Bernard J. Gersh, MBChb, DPhil, MACC: It may be effective for other reasons, namely people with atrial fibrillation who have coronary disease and need it for secondary prevention and so on. But for thrombo prophylaxis, it’s completely ineffective.

Christopher Granger, MD: And it’s not so safe.

Bernard J. Gersh, MBChb, DPhil, MACC: And with a risk profile almost as high as warfarin. That’s 1 thing we really need to propagate: that it’s no good for this.

Manesh Patel, MD: We’re certainly going to spend some time talking about what aspirin does. And in the British Heart Foundation or other trials, as you said, warfarin use in patients over 80 had a risk almost or even greater than warfarin in patients with AFib with no stroke benefit. Obviously, before we started the non—vitamin K drugs, there was the ACTIVE trial.

Bernard J. Gersh, MBChb, DPhil, MACC: I think you all have seen the cartoon of the 2 people in heaven, and there’s a conversation between 2 atrial fibrillators. And 1 says, “Well, just to be on the safe side, I still take 1 aspirin a day.”

Manesh Patel, MD: Yeah, that’s good.

Transcript edited for clarity.