Clinical trials supporting Food and Drug Adminstration approval of contemporary cancer therapies frequently failed to capture major adverse cardiovascular events (MACE) and, when they did, reported rates 2.6-fold lower than noncancer trials, new research shows.
Overall, 51.3% of trials did not report MACE, with that number reaching 57.6% in trials enrolling patients with baseline cardiovascular disease (CVD).
Nearly 40% of trials did not report any CVD events in follow-up, the authors reported online Feb. 10, 2020, in the Journal of the American College of Cardiology (2020;75:620-8).
“Even in drug classes where there were established or emerging associations with cardiotoxic events, often there were no reported heart events or cardiovascular events across years of follow-up in trials that examined hundreds or even thousands of patients. That was actually pretty surprising,” senior author Daniel Addison, MD, codirector of the cardio-oncology program at the Ohio State University Medical Center, Columbus, said in an interview.
The study was prompted by a series of events that crescendoed when his team was called to the ICU to determine whether a novel targeted agent played a role in the heart decline of a patient with acute myeloid leukemia. “I had a resident ask me a very important question: ‘How do we really know for sure that the trial actually reflects the true risk of heart events?’ to which I told him, ‘it’s difficult to know,’ ” he said.
“I think many of us rely heavily on what we see in the trials, particularly when they make it to the top journals, and quite frankly, we generally take it at face value,” Dr. Addison observed.
Lower Rate of Reported Events
The investigators reviewed CV events reported in 97,365 patients (median age, 61 years; 46% female) enrolled in 189 phase 2 and 3 trials supporting FDA approval of 123 anticancer drugs from 1998 to 2018. Biologic, targeted, or immune-based therapies accounted for 72.5% of drug approvals.
Over 148,138 person-years of follow-up (median trial duration, 30 months), there were 1,148 incidents of MACE (375 heart failure, 253 MIs, 180 strokes, 65 atrial fibrillation, 29 coronary revascularizations, and 246 CVD deaths). MACE rates were higher in the intervention group than in the control group (792 vs. 356; P less than .01). Among the 64 trials that excluded patients with baseline CVD, there were 269 incidents of MACE.
To put this finding in context, the researchers examined the reported incidence of MACE among some 6,000 similarly aged participants in the Multi-Ethnic Study of Atherosclerosis (MESA). The overall weighted-average incidence rate was 1,408 per 100,000 person-years among MESA participants, compared with 542 events per 100,000 person-years among oncology trial participants (716 per 100,000 in the intervention arm). This represents a reported-to-expected ratio of 0.38 – a 2.6-fold lower rate of reported events (P less than .001) – and a risk difference of 866.
Further, MACE reporting was lower by a factor of 1.7 among all cancer trial participants irrespective of baseline CVD status (reported-to-expected ratio, 0.56; risk difference, 613; P less than .001).
There was no significant difference in MACE reporting between independent or industry-sponsored trials, the authors report.