Bleeding and Ischemic Events Fluctuate Out to 1 Year Post-STEMI: HORIZONS-AMI

Following a STEMI, the daily risks of both ischemic and bleeding events fluctuate but eventually decrease over time, with bleeding more likely to occur early and ischemic events later, according to a subanalysis of HORIZONS-AMI. The data support the use of potent platelet inhibition throughout the first year following primary PCI, the authors say.

The issue of which antiplatelet regimen ACS patients overall should be placed on following their event has been up for discussion in 2017. The TOPIC trial, presented in May at EuroPCR, suggested that ACS patients can be switched from more potent P2Y12 inhibitors like prasugrel (Effient; Eli Lilly) or ticagrelor (Brilinta; AstraZeneca) to clopidogrel after 30 days, to minimize bleeding risk. Additionally, two other ACS trials, CHANGE DAPT and TROPICAL-ACS, both presented at the European Society of Cardiology Congress in August, questioned the superiority of the more potent antiplatelets over clopidogrel in the long term.

“Every antithrombotic and antiplatelet agent that we use has benefits of reducing ischemia, but also [has] potential risk of increasing bleeding, so there's no such thing as a free lunch,” senior author on the current study, Gregg Stone, MD (NewYork-Presbyterian Hospital/Columbia University Medical Center, New York, NY), told TCTMD. “There’s always a trade-off. So the way we decide when to use more potent agents—which may have greater efficacy in reducing ischemia but may also have greater hemorrhagic complications—is to consider different time periods.”

For the study, Stone along with lead author Gennaro Giustino, MD (Icahn School of Medicine at Mount Sinai, New York, NY), and colleagues looked at 3,602 STEMI patients from the HORIZONS-AMI trial who received primary PCI and were treated with aspirin and clopidogrel through 1 year. The researchers noted 279 discrete ischemic events (cardiac death, MI, and definite stent thrombosis) and 271 non-CABG-related TIMI major and minor bleeding events, including recurrent events.

Both average daily bleeding and ischemic event rates decreased from the acute period (≤ 24 hours after PCI) through the late period (30 days to 1 year; P < 0.0001 for both), with the former peaking between day 2 and 3 and the latter within the first 24 hours. There were no differences between bleeding and ischemic risk in the acute phase, but the average daily bleeding risk was higher than that of ischemic events between days 1 and 30 (subacute phase; P < 0.0001). After 30 days, the absolute rate of ischemic events was 1.5% greater than the absolute risk of bleeding (P < 0.0001).

Of note, patients treated with bivalirudin as compared with unfractionated heparin plus glycoprotein IIb/IIIa inhibitors (GPIs) saw lower rates of bleeding and mortality at 30 days, but there were no differences seen in average daily ischemic risk. There were no additional risks seen in either bleeding or ischemic events past 30 days.

The findings were published in the October 10, 2017, issue of the Journal of the American College of Cardiology.

‘A Good Reminder’

Over the entire 1-year study period, the ongoing risk of ischemia and the fact that it tended to be greater than that of bleeding after the first month are important to note, Stone said. This supports “recommendations to continue potent antiplatelet agents for at least 1 year after STEMI, and also I think provides the mechanistic underpinnings for the advantages of those potent agents that were seen in studies such as PLATO with ticagrelor and TRITON with prasugrel.”

The analysis is “very unique” as it represents the first time that bleeding and ischemic risks have been analyzed as a daily average after STEMI, he observed. The extent to which the rates of both types of risk were heightened early after STEMI was “striking,” he commented, adding that he was also surprised at “the fact that bleeding was even more common than ischemia early after myocardial infarction.”

However, “that really depended on which antithrombotic regimen we used,” Stone noted. “In the bivalirudin arm, . . . there was not an increase in bleeding compared to ischemia in the early period, which is one of the real advantages of bivalirudin.”

In an editorial accompanying the study, Derek Chew, MBBS, MPH (Flinders University of South Australia, Adelaide), and Deepak Bhatt, MD, MPH (Brigham and Women’s Hospital, Boston, MA), write that this analysis “demonstrates the well-recognized higher recurrent risk of ischemic and bleeding events within the first 30 days of MI. Less commonly appreciated is that in this early phase, the daily risk of bleeding events exceeds that of ischemic events.”

Likewise, commenting on the study, Robert Yeh, MD, MSc (Beth Israel Deaconess Medical Center, Boston, MA), told TCTMD that this analysis is “a good reminder that really that first 30-day period is a period of heightened concern for recurrent events, stent thrombosis, myocardial infarction, bleeding, and then outside of that window these risks really decline substantially.”

Yeh added that the long-term results show the “late risks of ischemia appear to be greater at least on absolute terms than the late risks of bleeding,” which reinforces current guideline indications that recommend maintaining DAPT for a full year after STEMI. “It might also support this idea that maybe more potent antiplatelet inhibition is preferred in this situation over the long term, . . . which our current guideline recommendations have a preference for [in] patients with acute coronary syndrome and STEMI in particular,” he said.

Stone also emphasized that “there is an ongoing risk of ischemia, which is even greater than bleeding on clopidogrel. This warrants the use of more potent agents such as ticagrelor, especially in patients who are not at high risk for bleeding complications.”

Generalizable Findings?

Those newer drugs, however, were not tested in HORIZONS-AMI and indeed, a number of practices common when HORIZONS-AMI was conducted are no longer standard, the editorialists note—specifically, the use of routine rather than bail-out GPIs and preferential use of transfemoral rather than transradial access.

“These data may not fully reflect current practice and the associated contemporary ischemic and bleeding event rates,” Chew and Bhatt write. “Nevertheless, the demonstration of a clear persistent risk of recurrent ischemic events among patient with [STEMI] receiving culprit-lesion [PCI] provides a strong rationale for the effective use of secondary prevention strategies in this patient population.”

Stone admitted that the older data “explain some of the relative rates, but I think the concepts still hold. For example, many people do still use bivalirudin in STEMI, and you see that did change the relative rates of bleeding versus ischemia in that early period. If by extension, if we were to apply even greater bleeding avoidance strategies, such as radial access, then the early relative rates of ischemia compared to bleeding would be even greater.”

He also wanted to highlight the fact that the study counted each and every event, something not often done in the literature, “so this really provided an important estimation of the burden of both cardiovascular ischemic events and hemorrhagic complications over time.”

Chew and Bhatt praised the study for including these data. “Patients are rightly concerned about the risk of the ‘next event’ and not only the event they have just sustained,” they write. This provides “a more patient-relevant assessment of risk.”

Yeh agreed. The findings might “help us determine how much concern, how much vigilance we need to have with regard to certain patients at different times in their course after STEMI. They might help us to think about what is the true magnitude of risk that a patient has,” he said. “Patients often ask you, ‘When am I going to get back to normal? Can I travel?’ and things of that nature. And really when you see these curves, they help you . . . get a general feel for when the patient has sort of plateaued in terms of their acute risk of having a recurrent event.”

Yeh observed that the results lead him to question whether the ischemic risks could be reduced even further potentially with non-vitamin K antagonist oral anticoagulants, particularly in the wake of trials like COMPASS and ATLAS. “It does sort of reintroduce that possibility again that maybe we haven’t mitigated the ischemic risk in the first year enough in a high-risk population like STEMI,” he said.

Then there are the difficulties of applying data from clinical trial populations to real-world practice, Yeh commented. He suggested undertaking a future all-comers trial of STEMI patients with adjudicated endpoints “to see if the transitions in bleeding and ischemic risk were similar to what they observed here.” In particular, Yeh added, the population should be representative of current transradial practice and use of anticoagulants.

Note: Stone and several co-authors are faculty members of the Cardiovascular Research Foundation, the publisher of TCTMD.