LAS VEGAS -- The ISCHEMIA trial pitting early invasive management of stable angina against medical therapy alone is sure to be a landmark in the field when the results are reported this fall.
But concerns about the trial go well beyond a controversial change to its primary endpoint, as a debate made clear here at the Society for Cardiovascular Angiography and Interventions meeting.
ISCHEMIA "will change the course of interventional cardiology," argued Nathaniel Smilowitz, MD, of the NYU Langone Medical Center in New York City, for the .
Enrollment has already finished for the trial and follow-up is almost complete, noted Smilowitz, who added the disclaimer that he is not part of the trial's clinical coordinating center his institution hosts.
Endpoints and Implications
However, controversy kicked up last year when the trial's prespecified contingency plan kicked in to boost statistical power by changing the primary endpoint from just hard events (cardiovascular death and myocardial infarction) to include the softer outcomes of hospitalization for unstable angina or heart failure and resuscitated cardiac arrest in a five-point composite.
Criticism of the trial organizers over having "" was not really warranted, as it was a blinded, planned change, acknowledged Usman Baber, MD, of Mount Sinai Hospital in New York City, who argued that .
But the shift is still problematic, he pointed out.
Including hospitalization for unstable angina could yield ascertainment and reporting bias in favor of the intervention arm in an unblinded treatment setting and because hospitalization for heart failure lacks specificity and a standardized definition, Baber added. The OAT and trials showed no hint of benefit from percutaneous coronary intervention (PCI) for heart failure hospitalization, which could mean dilution of a potential treatment effect from intervention in ISCHEMIA, he said.
"Naturally, these are very difficult to interpret," Baber said. "What do you do if the trial results are positive but entirely driven by soft endpoints? Should this impact practice if we see no movement on death or MI, or guidelines?"
"I think ultimately it will be important for us to focus on the key secondary endpoint of CV death or MI, number one. And number two, to look for the consistency of that endpoint with respect to the other endpoints that are presented," he concluded.
Prior landmark trials in this arena -- , BARI 2D, and -- have consistently shown that coronary revascularization for stable ischemic heart disease doesn't reduce death or MI, noted Smilowitz.
It's possible that those trials failed to show an effect because they were underpowered or didn't have a sufficiently high risk population, he suggested.
Smilowitz proposed that if ISCHEMIA ends up neutral, then practice should be a cardiac CT angiography to exclude left main disease followed by optimizing medication and lifestyle then considering invasive management based on patient preference but not in the absence of symptoms.
However it turns out, "I think this is going to help us with shared decision making," he told attendees.
Limitations
Despite being more than twice as large as the previous randomized trials of stable ischemic heart disease, ISCHEMIA did have some important exclusion criteria to note.
Patients failing medical therapy were not included, nor were high risk patients with low ejection fraction or left main coronary artery disease on CT angiography, Smilowitz noted.
Many women were excluded due to insufficient ischemia or nonobstructive coronary artery disease, so they represented only 22.6% of the trial population, he said. By comparison, women account for 43% of PCI done for stable disease in the U.S., Baber pointed out.
Also, treatment allocation was unblinded to both patients and their clinicians.
ISCHEMIA should get some higher risk patients because coronary anatomy was not known at time of randomization, unlike in prior trials, Baber noted.
But it had a much lower proportion of patients with prior MI than prior trials and fewer with prior PCI than seen in routine practice, based on the National Cardiovascular Data Registry in the U.S. for nonacute indications.
Severe and unacceptable angina were excluded from the trial, but notably about half of the cohort had none or class I angina -- much higher than prior trials or the 25% in typical practice getting PCI, Baber noted.
Overall, ISCHEMIA has a lower clinical risk profile and symptom burden than other trials and usual practice, although greater anatomic extent of disease, he added. "I think those will certainly be provocative when the trial results come out irrespective of which direction with respect to these differences," he said.
Crossover is unavoidable in a strategy trial like this, diluting effects and complicating communication of the trial results, he added.
However it turns out, ISCHEMIA "is going to be part of our conversation for many years to come in the management of stable ischemic heart disease," Baber said.
Disclosures
Baber disclosed relationships with Boston Scientific and AstraZeneca.