Christopher P. Cannon, MD
See a slide and audio version of this commentary at our partner site, QuantiaMD (free account required)
One of the biggest problems with antithrombotic therapy is the ever-present balance of reductions in ischemic events with increases in bleeding events. The analogy of “there is no free lunch” seems to apply: When more antithrombotic agents are added, or more potent ones are used, there is a benefit with a reduction in recurrent myocardial infarction and stent thrombosis but unfortunately an increase in bleeding. This issue is present for patients undergoing coronary stenting, especially drug-eluting stents, for whom dual antiplatelet therapy (DAPT) for 1 y after coronary stent implantation has been recommended in most guidelines for the past decade. However, because of increased risk for bleeding, many trials sought to reduce the duration, and indeed the most recent guidelines from the American College of Cardiology and American Heart Association and from the European Society of Cardiology recommend 6-12 mo.
Others have noted that extended duration beyond a year could be beneficial, and this was seen in the large DAPT trial involving nearly 10,000 patients. It found a 71% reduction in the risk for stent thrombosis and 29% reduction in the risk for major adverse cardiac events in patients with drug-eluting stents who received 30 vs 12 mo of DAPT. However, there was also a significant 36% increase in moderate to severe bleeding events with extended DAPT duration. Thus, we had the same conundrum—should we treat and prevent more events? Or stop and not induce more bleeding?
The DAPT Score
Yeh and the DAPT investigators (in a study summarized here) worked on a new score to try to predict which patients would have benefit of lower risk for thrombotic events without increased bleeding. They did this in a unique way that I have never seen done before. Most risk scores predict the risk for an event, but this score predicts the predicted benefit (and in this case the predicted benefit without risk):
- First, they developed standard risk scores using multivariable models to predict the incidence of ischemic events (stent thrombosis or myocardial infarction) and bleeding among the 11,648 patients. Then—and this is the unique “secret sauce” of this score—they took each patient in the trial and calculated the predicted thrombotic risk in each treatment group (shorter vs longer duration of DAPT) and could see the difference between the two (ie, the predicted benefit for that individual patient). Then, they did the same thing for bleeding: They predicted each patient’s risk for bleeding with shorter vs longer duration of DAPT and could determine what that individual’s predicted harm would be with extended therapy. Then, because the question is who will benefit without harm, they subtracted the harm from the benefit. This, then, is the predicted benefit without harm.
- The investigators then developed a risk score to predict this predicted benefit (without harm). They put a similar group of individual prediction factors in a multivariate model and developed a simplified integer score. The higher the score, the higher the predicted benefit without bleeding.
- Finally, they applied the score to the tot...