Clinical trials supporting FDA 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 CVD.
Nearly 40% of trials did not report any CVD events in follow-up, the authors report online February 10 in the Journal of the American College of Cardiology.
“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, told theheart.org | Medscape Cardiology.
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,” 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 US Food and Drug Administration (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 1148 incidents of MACE (375 heart failure, 253 myocardial infarctions, 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 < .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 6000 similarly aged participants in the Multi-Ethnic Study of Atherosclerosis (MESA).
The overall weighted-average incidence rate was 1408 per 100,000 person-years among MESA participants, compared with 542 events per 100,000 person-years among oncology trial participants (716/100,000 in the intervention arm). This represents a reported-to-expected ratio of 0.38 — a 2.6-fold lower rate of reported events (P < .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 < .001).
There was no significant difference in MACE reporting between independent or industry-sponsored trials, the authors report.
No Malicious Intent
“There are likely a few or some that might lean toward not wanting to attribute blame to a new drug when the drug is in a study, but I really think that the leading factor is lack of awareness,” Addison said.
“I’ve talked with several cancer collaborators around the country who run large clinical trials, and I think often when an event may be brought to someone’s attention, there is a tendency to just write it off as kind of a generic expected event due to age, or just something that’s not really pertinent to the study. So they don’t really focus on it as much,” he said.
“Closer collaboration between cardiologists and cancer physicians is needed to better determine true cardiac risks among patients treated with these drugs.”
Breast cancer oncologist Marc E. Lippman, MD, Georgetown University Medical Center, Georgetown Lombardi Comprehensive Cancer Center, Washington, DC, isn’t convinced a lack of awareness is the culprit.
“I don’t agree with that at all,” he told theheart.org | Medscape Cardiology. “I think there are very, very clear rules and guidelines these days for adverse-event reporting. I think that’s not a very likely explanation — that it’s not on the radar.”
Part of the problem may be that some of the toxicities, particularly cardiovascular, may not emerge for years, he said. Participant screening for the trials also likely removed patients with high cardiovascular risk.
“It’s very understandable to me — I’m not saying it’s good particularly — but I think it’s very understandable that if you’re trying to develop a drug, the last thing you’d want to have is to have a lot of toxicity that you might have avoided by just being restrictive in who you let into the study,” Lippman said.
The underreported CVD events may also reflect the rapidly changing profile of cardiovascular toxicities associated with novel anticancer therapies.
“Providers, both cancer and noncancer, generally put cardiotoxicity in the box of anthracyclines and radiation, but particularly over the last decade, we’ve begun to understand it’s well beyond any one class of drugs,” Addison said.
“I agree completely,” Lippman said. For example, “the checkpoint inhibitors are so unbelievably different in terms of their toxicities that many people simply didn’t even know what they were getting into at first.”
One Size Doesn’t Fit All
Javid Moslehi, MD, director of the cardio-oncology program at Vanderbilt University, Nashville, Tennessee, said echocardiography — recommended to detect changes in left ventricular function in patients exposed to anthracyclines or targeted agents like trastuzumab (Herceptin) — isn’t enough to address today’s cancer therapy-related CVD events.
“Initial drugs like anthracyclines or Herceptin in cardio-oncology were associated with systolic cardiac dysfunction, whereas the majority of issues we see in the cardio-oncology clinics today are vascular, metabolic, arrhythmogenic, and inflammatory,” he told theheart.org | Medscape Cardiology. “Echocardiography misses the big and increasingly complex picture.”
His group, for example, has been studying myocarditis associated with immunotherapies, but none of the clinical trials require screening or surveillance for myocarditis with a cardiac biomarker like troponin.
The group also recently identified 303 deaths in patients exposed to ibrutinib, a drug that revolutionized the treatment of several B-cell malignancies but is associated with higher rates of atrial fibrillation, which is also associated with increased bleeding risk. “So there’s a little bit of a double whammy there, given that we often treat atrial fibrillation with anticoagulation and where we can cause complications in patients,” Moslehi noted.
Although there needs to be closer collaboration between cardiologists and oncologists on individual trials, cardiologists also have to realize that oncology care has become very personalized, he suggested.
“What’s probably relevant for the breast cancer patient may not be relevant for the prostate cancer patient and their respective treatments,” Moslehi said. “So if we were to say, ‘every person should get an echo,’ that may be less relevant to the prostate cancer patient where treatments can cause vascular and metabolic perturbations or to the patient treated with immunotherapy who may have myocarditis, where many of the echos can be normal. There’s no one-size-fits-all for these things.”
Wearable technologies like smartwatches could play a role in improving the reporting of CVD events with novel therapies but a lot more research needs to be done to validate these tools, Addison said. “But as we continue on into the 21st century, this is going to expand and may potentially help us,” he added.
In the interim, better standardization is needed of the cardiovascular events reported in oncology trials, particularly the Common Terminology Criteria for Adverse Events (CTCAE), said Moslehi, who also serves as chair of the American Heart Association’s subcommittee on cardio-oncology.
“Cardiovascular definitions are not exactly uniform and are not consistent with what we in cardiology consider to be important or relevant,” he said. “So I think there needs to be better standardization of these definitions, specifically within the CTCAE, which is what the oncologists use to identify adverse events.”
In a linked editorial, Lippman and cardiologist Nanette Bishopric, MD, Medstar Heart and Vascular Institute in Washington DC, suggest it may also be time to organize a consortium that can carry out “rigorous multicenter clinical investigations to evaluate the cardiotoxicity of emerging cancer treatments,” similar to the Thrombosis in Myocardial Infarction (TIMI) Study Group.
“The success of this consortium in pioneering and targeting multiple generations of drugs for the treatment of MI, involving tens of thousands of patients and thousands of collaborations across multiple national borders, is a model for how to move forward in providing the new hope of cancer cure without the tradeoff of years lost to heart disease,” the editorialists conclude.
The study was supported in part by National Institutes of Health grants, including a K12-CA133250 grant to Addison . Bishopric reports being on the scientific board of C&C Biopharma. Lippman reports being on the board of directors of and holding stock in Seattle Genetics. Moslehi reports having served on advisory boards for Pfizer, Novartis, Bristol-Myers Squibb, Deciphera, Audentes Pharmaceuticals, Nektar, Takeda, Ipsen, Myokardia, AstraZeneca, GlaxoSmithKline, Intrexon, and Regeneron.