March 23, 2016 – Patients who are older and have diabetes are among those at greater risk of spontaneous heart attack.
Spontaneous heart attacks are common in patients with acute coronary syndrome (ACS) who do not undergo revascularization. However, a new tool could allow clinicians to predict those patients’ risk for subsequent spontaneous heart attack.
ACS patients who do not undergo revascularization are already known to be at greater risk of spontaneous heart attack, but the frequency and predictors of such events have not been studied until now. In an article published this week in the Journal of the American College of Cardiology (JACC), the DCRI’s Renato Lopes, MD, PhD (pictured), and his colleagues examined data from the TRILOGY ACS (TaRgeted platelet Inhibition to cLarify the Optimal strateGy to medically manage Acute Coronary Syndromes) trial to answer those questions.
TRILOGY ACS studied the effect of platelet inhibition in patients with ACS managed medically without revascularization. In this substudy, the DCRI researchers sought to characterize patients with spontaneous heart attack during long-term follow-up through 30 months to develop a prediction model for time to first spontaneous heart attack.
To do so, they examined data from 9,294 patients with non–ST-segment elevation myocardial infarction (NSTEMI)/unstable angina (UA) who were managed medically without planned revascularization. Of these patients, 983 experienced heart attacks over 30 months. Of these, 737 (75 percent) were first heart attacks and 246 (25 percent) were second or recurrent heart attacks. Of the 737 first heart attacks, 695 (94 percent) were classified as spontaneous heart attacks.
Using these data, Lopes and his colleagues were also able to identify what patient characteristics were most strongly associated with increased risk of spontaneous heart attack. Patients who were older, had diabetes, were initially diagnosed with heart attack (as opposed to unstable angina), had no pre-randomization angiography, and had higher baseline creatinine values were found to be most at risk.
With these findings, the researchers were able to construct a new predictive model for physicians that provides real-time and individualized time-varying risk estimates on the basis of 17 variables. This model could be highly useful in assessing the likelihood of spontaneous heart attack in multiple populations, Lopes said.
“This is a well-developed, very good predictive model, not just for patients who are at high risk of a heart attack, but particularly for low- and intermediate-risk patients,” said Lopes. “It’s user-friendly, and we hope to make it widely available. It’s just a question of familiarizing physicians with the tool.”
“We used predictive model markup language to deliver our model, making interacting with our model easier and more comprehensive,” said Benjamin Neely, MS, the main DCRI statistician for the project. “We encourage those interested in learning more about our model to explore the utility of this delivery mechanism by downloading the supplemental materials from JACC or visiting our GitHub repository.”
In addition to Lopes and Neely, DCRI authors included Megan L. Neely, PHD; E. Magnus Ohman, MD; and Matthew T. Roe, MD, MHS.