Applying Real-World Evidence in the Management of CAD/PAD - Episode 2

COMPASS Trial: Subgroup Analysis and Implications

Published on: 


Deepak Bhatt, MD, MPH: Kelley, I’ll turn to you. You helped lead the COMPASS trial efforts in the United States. Fortunately, a significant number of patients from the United States got into the trial. It’s great to see the United States represented in these international trials, but it’s not always easy to do these days. There was a lot of great work by you and your colleagues at the University of Washington. What would you say in terms of some of the other analyses that have come out looking at patients with CKD [chronic kidney disease] and of course patients with a history of heart failure, an analysis you utilized and published in Circulation. Did these subgroups that are known to be high risk provide any additional insight into the potential utilization and value of this vascular dose of rivaroxaban?

Kelley Branch, MD, MS: Yes, Deepak. Thanks for that question. It’s an interesting question because whenever we’re looking at specific subgroups within these analyses, we cardiologists always have that penchant to want to dig a little deeper once we see the primary end point, and we’ll say, “Whom are we going to treat? Who are the patients with the highest risk? Whom do I start with or whom do I end with?” A lot of the analyses that were done subsequently lent more information on who the higher-risk patients are.

Sonia Anand completed a wonderful analysis looking at multiple risk factors to see which 1 of those was driving the end points for the patients within the COMPASS trial. Of those, it came down to 4, and I don’t think these are too surprising, but it’s good to see the consistency of effects along different risk factors.

Coming down to the 4, the first is patients with polyvascular disease. Deepak, you’ve done some wonderful work showing that that was the highest-risk patient population. This was borne out in the COMPASS trial, and the highest-risk patients actually had the most benefit.

Next was patients with a history of heart failure. Generally, these were patients with preserved ejection fraction. We did that analysis that, as you noted, was published in Circulation. Most of the patients had preserved ejection fraction. We had a proportion of patients, about 60%, who did not have ejection fractions but had a predominance of patients had HFpEF [heart failure with preserved ejection fraction]. Remember that the COMPASS trial did not allow patients who had severe heart failure, New York Heart Association class 3 or 4, although a couple of class 3 patients snuck in there. Then, if you had an ejection fraction of less than 30%, you were not allowed. There was obviously the COMMANDER HF trial that looked at that. We saw a real consistency within the heart failure population that these were high-risk patients. They had similar benefits or a nice consistency.

Then you had patients with EGFRs [estimated glomerular filtration rates] less than 60 mL/min, so renal insufficiency, and then patients with diabetes. Deepak, you published that nice paper in JACC [Journals of the American College of Cardiology]. These are definitely risk factors. There’s consistency within the forest plots for all these risk factors, but if we concentrate on the big 4, polyvascular disease, heart failure, renal insufficiency, and diabetes, those are the highest-yield populations. A consistency through all risk factors suggested that it may be more applicable to an even broader audience than those 4.

Deepak Bhatt, MD, MPH: That’sgreat information. I’ll turn to you, Marc Cohen. You’ve been leading the way in anticoagulation going back even before the NOAC era to enoxaparin and helping us move away from unfractionated heparin as the go-to anticoagulant. What have you thought about the COMPASS trial data in terms of your own practice [at Newark Beth Israel Medical Center]? Has it been data that have been useful and actionable in your own patients?

Marc Cohen, MD: The short answer is that it is very compelling in the sense that it’s the concept of combining anticoagulation with the right dose and the right drug with antiplatelet therapy with the right doses and the right drugs. The notion of putting those together in an optimal manner dates to my fellowship. I remember reading Alexander Turpie’s paper combining Coumadin [warfarin] and aspirin in women with rheumatic mitral valve prostheses. That had a dramatic benefit, but it suffered terribly from a lot of the bleeding. Bleeding in the United States always sets us back, and it had suppressed the notion or the hypothesis of combination therapy until a lot of work went into identifying better agents and more optimal doses.

Manesh, in a burst of transparency, suggested that he was a bit surprised that 2.5 mg bid [twice daily] would even have an effect. Frankly, if you look at the pharmacodynamic data in terms of what kind of anti-Xa assay activity you get, it’s a little shocking that we get these overwhelmingly large trials confidently supporting a benefit from such a modest—I’ll use the word modest—intervention. This is incredibly compelling. The concept of combination therapy is here to stay, and we have enough data from the people who, like myself, were very much on the dual antiplatelet train and on the “platelet is the center of the universe” train. We ourselves were on a path in 1 direction, and little by little, conceptually, we’re veering in a different direction.

Transcript Edited for Clarity