This transcript has been edited for clarity.
Michelle L. O'Donoghue, MD, MPH: Hi. I'm Dr Michelle O'Donoghue, reporting from the American Heart Association Scientific Sessions in Philadelphia. Joining me today is Dr Jacqueline Tamis-Holland from Mt Sinai in New York, as well as Dr Rasha Al-Lamee from Imperial College in London. We're going to be talking about the ISCHEMIA trial. Obviously there has been a lot of talk about the results of this study. Perhaps you'd like to lead off and walk us through the study design and the top-line findings.
ISCHEMIA Trial Design
Rasha K. Al-Lamee, MBBS, MA, PhD: This was a landmark international study with over 38 countries taking part. For study design, they looked at patients with stable coronary artery disease who had moderate to severe ischemia as assessed by exercise testing, nuclear scanning, stress echocardiography, and MRI. Patients went on to have a CT to rule out left main stem in particular and were randomized to either an invasive or conservative arm, with optimal medical therapy given to both groups. This was an unblinded trial, so patients were aware of their treatment allocation as were the physicians. They evaluated hard endpoints between the two groups and then also symptomatic secondary endpoints. The primary outcome included death, myocardial infarction (MI), hospitalization for unstable angina, resuscitated cardiac arrest, and hospitalization for heart failure.
O'Donoghue: I think many people thought about ISCHEMIA as a "sequel" to the COURAGE trial. You hit on a very important point that within the ISCHEMIA cohort, everybody did undergo a coronary CT angiography (CTA), so you did have a sense of their coronary anatomy prior to their entering the study. Similarly with COURAGE, the patient population had all undergone coronary angiography before being randomized.
Al-Lamee: What really stood out for this trial was the fact that they randomized them ahead of a coronary angiogram. Hopefully we got away from one of the criticisms that COURAGE received, which was that patients with proximal left anterior descending (LAD) disease would never make it to trial; they would never be randomized, so there was some selection bias. I think some of that was minimized by this very groundbreaking trial design.
O'Donoghue: What do you view as the top-line results that you want viewers to walk away with?
Jacqueline E. Tamis-Holland, MD: The most important thing to me is the fact that their mortality was incredibly low and that the event rate was relatively low. These patients, with the exception of those who might have left main disease, have a relatively good outcome regardless of what we do in the initial setting. I'm comfortable saying that we can take our time deciding what we want to do. We don't have to rush to the cath lab that evening after their stress test, and we can have a conversation with them.
O'Donoghue: I was really quite struck by the fact that this was a patient population where a lot of them had proximal LAD disease—things that would give us pause in a clinical setting. Yet the event rate was quite low. Do we think that optimal medical therapy is what is making a big difference in the backdrop for these patients?
Tamis-Holland: I definitely think it's helping—no doubt about it. It makes a difference in our outcome compared to what we used to do many years ago, where you saw different degrees of aggressiveness. It does help contribute to that.
Al-Lamee: The investigators published a paper in Circulation: Cardiovascular Quality and Outcomes a week ago looking at the level to which they delivered that medical therapy. They tried very hard to deliver optimal medical therapy. Patients were judged in terms of their adherence; nonadherent patients were not even included in the trial. They tried very hard to get to target on lipids and to target on blood pressure, and also give them some pretty decent antianginal therapy, which we have not necessarily seen in some of the trials that preceded it and we're not so good at doing in clinical practice.
Tamis-Holland: Although the percentage of patients who actually were optimal was really not that high.
Al-Lamee: And they didn't change so much within the trial. But it's hard to do that with such an international trial. Even at the point when they realized that some sites were not going to target, they went back out to those sites and tried to reeducate them to do some extra strategies to make it better.
O'Donoghue: Perhaps even more remarkable is that even though medical therapy was not fully optimized, there still was not any clear benefit of stenting these patients in terms of mortality reduction or net MI reduction.
Al-Lamee: These stable patients probably have a very different disease than the acute patients that come to our cath labs. We very rarely have patients for whom we perform primary percutaneous coronary intervention (PCI) who tell us they have had angina over the past few weeks. Often this is the first time they presented with it. There must be a very different disease process going on between these two groups.
O'Donoghue: It's always important to underscore that this was a stable patient population. There was discussion that when you stent a patient, there seems to be a little bit of early hazard in terms of earlier risk for MI, many of those probably periprocedural. And a little bit of a benefit was emerging late in terms of MI reduction on the late side. But this was overall sort of offset by that early hazard.
Tamis-Holland: We still have to wait to see the manuscript regarding the clinical relevance, although they met the criteria and they had very strict criteria for postprocedure MI. I would be curious to know how clinically relevant those postprocedure MIs were. More importantly, I would like to know more about the spontaneous MIs, which were clearly higher in the group of patients who got conservative care.
Al-Lamee: Once we saw those curves cross, seeing the 5-year data is going to be key to see whether things change and we suddenly have a reversal in the primary endpoints essentially over time.
Effect on Symptomatology
O'Donoghue: Perhaps you'd like to talk about the symptomatology piece and whether you feel that ISCHEMIA demonstrated a reduction in overall anginal symptoms. How do we put that in the context of ORBITA?
Al-Lamee: I guess the good-news story from ISCHEMIA is that there does seem to be a significant improvement in symptoms for those patients who have the invasive strategy. I do caveat that with the fact that this was an unblinded trial—but the follow-up goes out to 36 months. You would expect placebo to attenuate over time, although in the later myocardial revascularization trials, we saw that the placebo effect can last out to 30 months. That is hard to tell. Having said that, symptom improvement in terms of freedom from angina was actually quite similar to ORBITA. In our secondary analysis, patients were 20% more likely to have freedom from angina if they got angioplasty. And there, too, the rates were quite similar. They do kind of tie up. We recently published our stress echo–stratified analysis looking at all patients stratified on the basis of their burden of stress echo ischemia. Patients with the highest stress echo ischemia at pre-randomization benefited the most in terms of symptom reduction on angina frequency with PCI versus placebo. I would say that the two studies are quite in parallel.
O'Donoghue: As I think back to COURAGE, it seemed like the improvement in symptoms started to wane over time. It started to close between the treatment arms.
Tamis-Holland: BARI 2D showed the same thing—that initial improvement was seen and then over time they converged. I feel very strongly that it is a definite improvement in symptoms as opposed to a placebo because of the fact that the two trials similar to this did show an attenuation in the difference in ischemia and it did stay out to 3 years.
Al-Lamee: Symptoms are important to our patients, right? They want quality-of-life differences. Yes, some of that may have been placebo, but there is probably some definitive improvement.
O'Donoghue: The point made yesterday during the discussion was that ISCHEMIA reassures us that we can go that route for a patient who is appropriate for initial medical management. If they continue to have intractable symptoms or it's really interfering with their quality of life, ISCHEMIA also gives us some reassurance that there is no clear net harm from going ahead and stenting patients in that situation. There might be the early hazard, and that is a conversation we need to have with our patients beforehand. But nonetheless, it ends up being net-neutral in terms of the hard outcomes of death or risk for MI.
Tamis-Holland: It reflects back to where we started, which is that regardless of the degree of ischemia, you can comfortably say that if you are taking care of a patient, you are doing it to try to improve their quality of life or symptoms. So if you need to do a revascularization procedure to help make them feel better, it's still fine. It was fine before [ISCHEMIA] and it's still fine.
Al-Lamee: Sometimes I use an analogy with orthopedic surgery: You don't do a total hip replacement or knee replacement to save someone's life; you do it to improve their quality of life. And those patients still want that procedure. Maybe that is the place for angioplasty in stable coronary artery disease.
Tamis-Holland: But it's surgery too.
Al-Lamee: Absolutely. That is key. You make a great point. A quarter of these patients had coronary artery bypass grafting (CABG). It's revascularization as a whole, I suppose.
O'Donoghue: It's not just a stenting trial. It's a little bit different from COURAGE that way because coronary anatomy was not known upfront.
Role of Stress Test
O'Donoghue: Let's say that people watching this are thinking, How am I going to evaluate the next patient who comes in to my clinic who has been having anginal symptoms? Maybe you do or do not yet have a stress test. Even though it was part of ISCHEMIA, is the stress test going to modify your thinking, necessarily? And do we need to have knowledge of their coronary anatomy once you're aware of the fact that they have moderate to severe ischemia in any given distribution? When do we say, "I'll give medical therapy; I don't need to know any more." Or do we need to investigate a little bit further?
Tamis-Holland: Before we can extrapolate the results of the trial to our patient in the office who comes to us with an abnormal stress test, I think we do need to define the anatomy, whether by protocol with CT scan or whether they are taken to the lab to understand whether this patient with this degree of ischemia would have met the criteria to be enrolled. Otherwise, if they had severe left main, they would not have been included.
I think that is important, but I'm also not really sure. I'm curious to know what your feelings are, of where you think the role of stress testing versus just defining the anatomy comes in. I still feel that functional testing is sometimes helpful in those 75-year-olds who are going to have incidental coronary disease anyway. What do you do with that?
Al-Lamee: The pathway in the United Kingdom is really quite different to the United States. Our gatekeeper has become the CT. This was National Institute for Health and Care Excellence (NICE) guidance a while ago. In fact, when I do the rapid-access chest pain clinic, the majority of patients have a CT.
Tamis-Holland: If they are abnormal, do you then follow it with a functional test?
Al-Lamee: I check their CTs, and if they have very significant symptoms which I think are cardiac, the predominant strategy has been to get these guys to the lab. Then I go for coronary angiogram, and at that point we do invasive physiology on the majority.
Tamis-Holland: What if they are 75 years old and they get short of breath or tired after two blocks? Not really angina symptoms.
Al-Lamee: Or you see kind of moderate disease on CT. For those patients I do functional testing. But the predominant strategy has been CT for us. And then when we see very significant disease, it's very heavily calcified and you can't define it any further, or it may be triple vessel or left main, we send those patients straight to the lab. I don't know if that is the right strategy, but that has been our strategy.
O'Donoghue: In some ways, you're now exposing them to double the contrast load and extra radiation. I suppose you could argue certainly for the three-vessel disease or left main, where it's going to definitively change your management. But what if you saw a 90% proximal LAD lesion on the coronary CTA? Would you feel the need to bring them back to the cath lab to confirm that?
Al-Lamee: I find it really tricky. We had a bunch of patients with proximal LADs in ORBITA and they clearly had a load in ISCHEMIA too. I still feel like that 50-year-old guy with that proximal LAD wants that fixed, and I want to fix it. The reality is that we don't necessarily have the evidence that you change that person's hard outcomes. I still find it uncomfortable to leave those patients alone. I don't know what to say.
O'Donoghue: A lot of interventional cardiologists would say the same thing.
Al-Lamee: I ran ORBITA and I'm saying that. To be honest, I have become much more conservative in patients with circumflex lesions and right coronary lesions, and I often start with medicine and see where it goes.
Tamis-Holland: I would agree with that. I'm still perplexed about the severe triple or the very proximal LAD. I would like to really see a drill-down of the data. I know that they said there was no difference. I believe that one of the interactions was proximal LAD versus not, and there was no difference.
I'd be especially curious about the breakdown of the CABG patients to see whether there was a difference in outcome between patients who had enough disease to undergo CABG compared with those whose disease just required PCI.
Al-Lamee: Definitely. And also even for the quality-of-life data. My sense is that the quality of life is quite different for those having had a CABG versus PCI. I would be interested to see if CABG patients have exactly the same quality-of-life improvement or whether having had that big procedure makes a difference going forward.
O'Donoghue: You raise concerns about the "oculostenotic reflex" when they are in the cath lab. On some level, the beauty of doing a coronary CTA is that it might make it easier to make a decision or have a conversation with the patient first before thinking about stenting them.
Al-Lamee: It is important when we discuss it with them now that we do not scare the patient. In the United Kingdom there is a waiting list for coronary angiography, and sometimes you have these patients asking whether they should be paying for this to happen privately so they can have it quickly. I don't think there's any need for that now. I think we can calm them down. We can say it's okay to get them on the right meds and then wait until the cath to see.
Role of FFR and iFR
O'Donoghue: One of the comments you are making about bringing them to the cath lab is about getting a better sense of the severity of the stenosis. One topic that has been much discussed after ORBITA was the concept of fractional flow reserve (FFR). How much should FFR and instantaneous wave-free ratio (iFR) be guiding our decision-making in this situation? What are your thoughts?
Al-Lamee: It's tricky because the problem with the invasive physiology data and FFR data is that the ischemic threshold for FFR was much lower than the critical threshold we are now using; 0.75 was the initial threshold. And when we looked at the ORBITA dataset in terms of FFR and iFR stratification, we found a really distinct relationship between FFR and iFR burden at pre-randomization. The lower the FFR, the lower the iFR, the more likely it was that angioplasty versus placebo would have an impact on stress echo ischemia or reduction ischemia. But we didn't find any impact on symptoms or exercise time. That may be because we were underpowered. It may also be over time because from our stress echo dataset, we've seen something different. We've seen a relationship between stress echo and symptoms. I'm starting to wonder if the FFR and iFR really tell us about the burden of disease and myocardial mass that is actually affected by that stenosis. Maybe an iFR of 0.4 in the circumflex is quite a different thing to an iFR of 0.4 in an LAD. So, I don't know. I'm still using physiology a lot and I will continue to use physiology a lot because I believe the data. I believe that it's important. But I am slightly starting to change my threshold of treatment. And I don't suddenly think that with a 0.79 on FFR I need to get on and treat. I'm trying to think about the patient more.
Tamis-Holland: I agree. To be honest with you, most of the data on FFR is really for the intermediate stenosis. It's definitely helpful in the symptomatic patient with intermediate disease. For the asymptomatic patient with severe disease, I'm not sure I would necessarily use that entirely as a marker of whether I intervene or not, especially now. So I take them to the lab or it's okay to intervene because the FAME study would show a benefit with an FFR that is abnormal. I don't know if I would have that same approach. In fact, the FAME study only included symptomatic patients. I'd be careful about asymptomatic patients.
Al-Lamee: The other thing with FAME 2 for me is that those urgent revascularization rates in FAME 2 have never been replicated in another trial again. They had urgent revascularization rates of over 20%. That is a little bit crazy to me because when we see the hospitalization for unstable angina in ISCHEMIA, it's much lower, and in ORBITA we didn't have these patients suddenly presenting with acute coronary syndrome or needing to be taken to the cath lab. It didn't happen. So, it speaks likely to the design of the trial, I think. And since it's not been reproducible, I'm a bit wary about FAME 2.
O'Donoghue: There was no clear indication that death or MI is modified.
Tamis-Holland: It's entirely driven by revascularization.
O'Donoghue: Which many would consider to be a softer outcome.
Tamis-Holland: Particularly when a physician and patient know of a lesion that has never been taken care of.
Al-Lamee: One thing that I was really pleased with was how [the ISCHEMIA investigators] defined hospitalization for unstable angina and they adjudicated it. You had to have more than just chest pain. You had to have ECG changes, you had to have troponin checked. They didn't do this in FAME 2, which meant that the vast majority of the revascularizations just happened for a chest pain presentation, troponin negative, ECG negative. And that could have been affected by unblinding, potentially.
O'Donoghue: One observation I heard somebody make yesterday was that, not surprisingly, for those patients who ended up having revascularization, at least PCI, there was a higher usage of dual antiplatelet therapy (DAPT) in that setting. I've heard some people postulate that some of that later MI benefit may in fact be some of those differences in therapy rather than the stent itself. It may be some additional benefit of having a P2Y12 inhibitor on board. Any thoughts on that?
Al-Lamee: We need to know more about that. It's possible that there is a definite contribution. It would make sense that the DAPT might make a contribution. Perhaps this is a dual effect—some of it from the revascularization, some of it the medications.
Tamis-Holland: I agree completely. This was not a trial on bleeding versus ischemic events in patients on DAPT. So that whole issue of bleeding events, which would be adjudicated in a trial that was looking specifically at that, is not. So you are getting, in a sense, the "DAPT trial" benefits of long-term DAPT without looking at the bleeding risk. We know from the CAPRIE trial that a single antiplatelet with a P2Y12 inhibitor is superior to aspirin alone in the vascular disease patients. One would think that it's a similar kind of situation.
O'Donoghue: It will be interesting to see whether this is something we should be considering for our patients with stable coronary disease who are being medically managed without a stent. Should we be thinking about a P2Y12 inhibitor for those patients? Obviously, it has been a little harder to demonstrate net clinical benefit in the past. Nonetheless, thinking about how best to optimize medical therapies is most important.
Al-Lamee: That is a trial we need.
Tamis-Holland: I was just going to say that. I was thinking we need to plan a trial.
O'Donoghue: That is the perfect place to wrap up. We'll start working on our trial designs. Thank you both for joining me today to discuss this very exciting topic.
Tamis-Holland: Thank you for having me.
Al-Lamee: Thank you.
Follow Medscape on Facebook, Twitter, Instagram, and YouTube
Cite this: Michelle L. O'Donoghue, Rasha K. Al-Lamee, Jacqueline E. Tamis-Holland. Managing the SIHD Patient in a Post-ISCHEMIA World - Medscape - Nov 20, 2019.