ISCHEMIA Questions? Some Answers From David Maron
In an in-depth interview, the principal investigator for ISCHEMIA tells TCTMD what he thinks the key trial take-homes should be.
I n the month since its presentation at the American Heart Association 2019 Scientific Sessions, the ISCHEMIA trial has continued to generate discussion and debate. The trial was presented to a packed hall in Philadelphia by study chair Judith Hochman, MD (NYU Langone Medical Center, New York, NY). TCTMD caught up with ISCHEMIA co-chair and principal investigator David J. Maron, MD (Stanford University School of Medicine, CA), in the weeks after the meeting to get his take on the trial and what he believes are the key outstanding questions raised in its wake.
Now that you have had some time to digest the results and have seen people’s reactions, what's your main takeaway from this whole experience after more than a decade of working on the trial?
I think we answered the question. We showed that there's no net benefit from an invasive strategy during this period of follow-up. It was really very interesting and in retrospect not that surprising that we saw what we did. We expected more procedural MIs [in the intervention group], but it was interesting to see the late benefit that was not clear in COURAGE and started to emerge in FAME 2, which was the crossing of the curves [for MI]. I think that we all expected the quality-of-life benefit, and it was reassuring to see so conclusively that for people with angina, an invasive strategy with revascularization improves symptoms.
Given the crossing of the lines, as you say, for the MI outcomes: two questions. The procedural MI outcomes were quite low—does this reflect real-world practice? On the other hand, the late, spontaneous MI rate in the medical therapy arm has produced a lot of discussion—what’s your read on that?
Yes, there is a question about how common these [procedural MIs] would be in centers that have smaller volumes or outcomes that are not as good as the sites that we selected, so that's something to consider. The findings may not be generalizable to centers that don't have as high-quality PCI. But really the big question is what is the prognostic significance of the procedural MIs compared with the procedural significance of spontaneous MIs? And during the period of follow-up in this trial, we found no net difference, but it's really one of the important reasons and it’s why we should do the long-term follow-up to see if there's any long-term mortality benefit from these procedures.
Mortality in this group of stable patients would presumably be pretty low. Do you think you have the numbers to detect a difference, even if you follow patients for longer?
Well, that depends on how long we follow up. But yes, if there is a mortality benefit, I think that long-term follow-up should be able to detect that. An additional 5 years or so should be adequate.
I know that at the time these results were presented, investigators were seeking funding for that additional follow-up. Any word on whether that funding has been realized?
No, we haven't yet resubmitted the application. I think that we are submitting that in February.
What kind of impact should the ISCHEMIA results have? I've heard from some people, and seen some polls on Twitter, suggesting that many cardiologists believe these results won't have much impact on their practice.
I think there are a number of implications for how practice should change. This is my opinion, and goes beyond the scope of what we studied. I think that we provide really powerful evidence for patients who are hopefully engaging in a discussion with their physicians about whether or not they should undergo cardiac catheterization and revascularization—that is, a patient who has had an abnormal stress test already. I think that this says that there's not an urgency to go to the cath lab, because there's not evidence from this trial that they are exposing themselves to excessive risk by just starting medical therapy and considering their options. It takes away urgency [and] it provides patient and physician with some data that they can use to make a good decision based on the patient’s preferences and values. . . . And then ultimately for the patient, is it worth it for them? I'm not sure that patients really understand before they have a stent or bypass operation what they are going to get out of it. I think a lot of patients still believe that it's going to save their lives, and it's a nuanced discussion about procedural MI versus later spontaneous MI, the early hazard and the late benefit.
I think that this says that there's not an urgency to go to the cath lab. David J. Maron
One thing we’re doing is, we are going to develop an app that will go on a smartphone with questions that patients can complete regarding their symptoms. That will be based on the Seattle Angina Questionnaire, just like the patients completed for the trial.
Another other implications?
I think ISCHEMIA really sheds some light on the need to identify whether or not left main disease is present. As you know, we built in a coronary CT angiogram so we could exclude people with left main disease, for patient safety, and also for the pragmatic issue regarding recruitment. A lot of physicians were reluctant or would have been reluctant to allow their patients to participate if we couldn’t exclude left main. But a nice benefit from doing that procedure is that we identified patients who did not have obstructive disease and therefore they were excluded from the trial—these are patients that we would not want to send to the cath lab. So not only would CT angiography, in practice, help us identify the patients who have left main disease, which among patients with at least moderate ischemia turned out to be about 9% of the patients that we enrolled, . . . but we identified 20% of patients who had the CT angiogram who had no obstructive disease. That's very useful information and I think that would work its way into our practice.
People have said to me that even though CT wasn't part of the experiment, as it were, it has become one of the many talking points of the trial. Do you think this should change the way patients are triaged when they have chest pain? What is the value of especially an imaging stress test in this setting if you have to do a backup CT anyhow?
I'm grappling with these questions myself. Again, I'm only telling you my own personal impressions, but I think that if I had to choose one or the other, stress test versus a CT angiogram, I would choose the CT angiogram. And the reason is that the severity of ischemia did not predict risk for having an event and it didn't identify patients who would benefit from a revascularization procedure, so knowing the severity of ischemia is not as valuable as knowing the severity of the coronary disease, which is: the more severe the anatomy, the higher the risk, even though revascularization did not reduce that risk. I think that the imaging test, which tells us about the severity of ischemia, is not as valuable as the physiologic information that we get from a non-imaging exercise stress test where we can learn about exercise endurance, response in terms of blood pressure, arrhythmias, and symptoms. So if we're going to be continuing to rely on stress testing, I think there should be a greater value placed on exercise testing as opposed to pharmacologic stress testing, and I'm not sure about the value of stress imaging.
The depth of these conversations with patients, and the choice of tests, seems to depend on who is having those conversations with patients and who makes the decision as to when to send someone for further tests. Is that mostly happening with a general practitioner, is it happening with a general cardiologist? Where are those discussions happening and where should they happen?
I actually don't know what proportion of stress tests are ordered by a primary care provider as opposed to a cardiologist. I wouldn't be surprised if it was roughly 50/50. An important question is what is the purpose of the stress test? And if it's to make a diagnosis, then I find myself both before and after knowing the results of ISCHEMIA still ordering stress tests because people describe certain symptoms associated with exercise and then I want to get some information about what happens to the patient with exertion. But I'm biased because as a cardiologist I see patients that underwent stress testing by a primary care physician and now they're being referred for an abnormal stress test, so the conversation that I'm having with the patient as a cardiologist is complex. Some patients come in with the expectation that they should have an angiogram, and if patients are referred directly to an interventional cardiologist it gets complicated if there's an expectation on the part of the patient or the referring physician. I think that these conversations do occur, but I don't know at what frequency and I don't know how well informed they are.
So in your mind, the trial met its goals?
The goal of the trial was to try to fill the gaps, to answer the question of what is the value of adding revascularization to good medical therapy. I think we have an answer, and quality of life is really important—it's one of the two reasons why we do anything in medicine. And if a patient needs relief of symptoms, then the revascularization should take place. I just want to be sure that patients understand and physicians understand what the proposition is, and the value proposition is: better symptom relief. And that's just a really clear message from the trial. I really hope that we can verify that those mortality curves that are virtually identical between the two strategies now, whether they separate or stay completely overlapped with longer-term follow-up.
I’m sure cost-efficacy analyses are planned or underway, but one of the points made by Dr. Hochman during her presentation was that despite the cost of this trial, it should lead to cost savings.
Yes, the low-hanging fruit here is that we shouldn't be doing revascularization procedures in stable patients who have no symptoms, and who have no left main disease. This trial cost taxpayers roughly $100 million spread out over 8 years, and people have questioned whether that was a good use of taxpayer dollars. And one of the take-home points from the trial is that there is no benefit to patients who have no symptoms, who are stable, and so if we just translated that finding into practice, with very conservative estimates, we would save the healthcare system $500 million per year. That’s a back of the envelop calculation—it's probably significantly more—but if we just stopped doing PCI on asymptomatic patients, it would be a huge saving.
So this $500 million dollars in costs saved, that covers only the asymptomatic patients? Those aren't also the patients who are told, ‘You're not going to live longer or avoid an MI; therefore you can choose?’
Oh no. That is strictly for asymptomatic, elective PCI.
It seems extraordinary to me that so many patients with zero symptoms were even ending up in cardiologists’ offices and getting these tests—it's a huge dollar number. What on earth drove them in there if they weren't even feeling anything?
Well, it's not that uncommon for screening stress tests to be done just looking to see if somebody has a problem.
ISCHEMIA grew out of the unanswered questions in COURAGE. What is the next trial to do in this space?
That's a good question. One area that I do not personally intend to explore is moderate left main disease, moderate meaning in the 50% to 70% range. When left main is greater than 50%, it's just a rule: patients need to be revascularized. I feel that way when I see a left main stenosis greater than 70%, but I don't know where the truth is. We're now talking about a smaller population, and it's just not something that I personally want to devote a number of years to discovering, but that's one unanswered question. Unanswered from our trial, but we hope to answer it, is completeness of revascularization and the impact of complete revascularization on prognosis and symptoms. I hope that we can answer that with our data.
What about “completeness” of adherence to guideline-directed medical therapy?
That certainly is important—not necessarily comparing revascularization with medical therapy. I think we have enough information about the incremental value of revascularization. Where we really need to do some work is in implementation science. We need to improve the adherence to medical therapy.
Can we take a break from ISCHEMIA for a bit? When will more data start to come in?
You can take a bit of a break. Then, starting in the spring, we plan to start publishing a lot of manuscripts.
This interview has been edited for brevity and clarity.