Nonculprit PCI Now or Never?
Gilles Montalescot, MD, PhD: Hello. I am Gilles Montalescot, a cardiologist from Paris, and I’m here in Barcelona at the European Society of Cardiology Congress with Tony Gershlick. Welcome, Tony.
Anthony H. Gershlick, MBBS: Thank you.
Dr. Montalescot: You are presenting a very important studyhere in the Hot Line Session. It’s the same question that we have been asking for years. What should we do with the coronary arteries other than the culprit when we are in the middle of a percutaneous coronary intervention (PCI) in a patient with ST-elevation myocardial infarction (STEMI) in the cath lab? Of course we treat the culprit artery, but what should we do with the others? You did a fantastic randomized study. Can you tell us about it?
Dr. Gershlick: You and many other clinicians who are watching this know that when you do a primary PCI in a left anterior descending artery, and there is a very tight stenosis in the right coronary artery, we have never known what to do. Some people leave it. Some people treat it. Some people defer treatment. Some people wait for a myocardial perfusion scan (MPS) or a nuclear scan. Some people don’t do anything. The original question (and there are many different questions, including how to judge severity) was: What should we do? Should we treat it during the admission, or should we not treat and wait and see what happens? That was the basic question that struck me as a clinician.
We had STEMI patients who had lesions in their non-infarct-related artery. The previous guidelines and the current American guidelines suggested that these non-infarct-related lesions shouldn’t be treated. We can — and should — discuss why what we did was different from those guidelines. Patients were randomly assigned after the PCI had been started on their infarct-related artery. If there was a lesion in the non-infarct-related artery that was more than 70% in a single plane or more than 50% in orthogonal planes, patients were randomly assigned to either treating the infarct-related artery alone and leaving any other significant stenosis or to treating both, preferably at the same session so the patient didn’t have to come back and have a second puncture. The interventionist could defer it if clinically he or she felt it was right to do so. I deferred one case when it was 2:00 AM and the team was tired. It was a right coronary artery occlusion, a significant stenosis in the circumflex artery.
Dr. Montalescot: So, the other lesion would be treated but not in the same sitting?
Dr. Gershlick: It had to be done during the same hospital admission, and in 59% of patients, it was done at the same sitting. We preferred to do it in the same sitting because we might have lost patients who did not want to come back and also because we wanted to shorten the length of stay.
We didn’t want patients to have more than one puncture. There were all sorts of reasons why we wanted to do it during the same admission.
These were significant stenoses. The randomization was different from some of the other trials in that it was done while the infarct-related artery was being treated. In PRAMI, the infarct-related artery treatment was finished, and then randomization was done, so there could have been a degree of selection.
Total Revascularization Cuts MACEs in Half
Dr. Montalescot: How many patients were there? What did you find?
Dr. Gershlick: We planned for 144 patients per group, based on the Politi data. There were some crossovers. It was an intent-to-treat protocol, and I have some per-protocol/per-treatment analyses as well. We found a 55% reduction in composite major adverse cardiac events (MACEs) at 12 months in patients who had complete revascularization at the time of their index admission. This was across the board. There was as much difference in hard endpoints — death and myocardial infarction — as there was in revascularization. It’s always different with open trials involving revascularization when the patient [and physician] knows they still have a lesion. We did a subanalysis looking at only the hard endpoints, and they were still significantly different, regardless of concerns about revascularization and objectivity.
The hazard ratio was 0.45 with a P value of .009 across all endpoints.
Dr. Montalescot: That is good.
Dr. Gershlick: It is interesting. It’s difficult because it prevents us from understanding the mechanism when you see differences across all composite endpoints. But we do have some cardiac magnetic resonance substudy data. We have a nuclear substudy that we put in at six weeks for safety, so that patients with a high ischemic burden weren’t left, but we nested it. We didn’t look at it unless we were notified by the core lab that there was more than 20% ischemic burden.
Dr. Montalescot: The curves diverged quite rapidly, didn’t they? The advance in this control arm occurred when — within the first two weeks, three weeks, four weeks?
Dr. Gershlick: We did the analysis with the Kaplan-Meier curves up to one month. Although the P value wasn’t significant at .055, it told me that if you look at the curves, there is early divergence across all the endpoints. No one thing in particular caused this early divergence. But it tells you, although this is all speculative because it’s a small study, that if you decide to bring the patient back in two months’ time to treat the non-infarct-related artery stenosis, you may have missed your chance because the divergence with these curves was very early.
No Safety Issues
Dr. Montalescot: Many times, we would take the patient back to the cath lab two or days later. This is different from what you did in the control arm, right?
Dr. Gershlick: You could do that. Is there another study where you randomize to treating at the time or doing it within two or three days? I don’t think so. Clinicians know when not to do something at the time of the primary PCI. If you had a patient, you would know immediately. The blood pressure is a little low. The patient is a bit peaky. It’s complex. You have spent time on the primary, but there is a tight stenosis somewhere else. You wouldn’t do it at the time. We don’t need to do that trial. There are other trials that need to be done, such as fractional flow reserve for determining stenosis, doing nothing vs doing it at admission vs stage. Other trials have asked slightly different questions, such as doing nothing vs a staged procedure, but they may have missed the in-hospital patient.
Dr. Montalescot: Did patients in your control group receive medications including antiangina agents?
Dr. Gershlick: Yes. We had a very high incidence of secondary prevention medications — 98% were on antiplatelet therapy or beta blockers, and 100% were on statins. It was very good. One of the pleasing things about the study is the optimal medication arm — we can’t be criticized there. Then we waited. Obviously we had a data and safety monitoring board (DSMB). Unlike PRAMI, the trial was not interrupted. The DSMB let us continue to the end of the study. That was very useful — we were not exaggerating the effect by stopping the trial early.
Dr. Montalescot: Did the physicians feel confident in leaving 98% stenosis in the contralateral artery?
Dr. Gershlick: We reassured them by having the MPS. We don’t know the results of the MPS yet because it’s still being analyzed. We know that no patients had more than 20% ischemic burden because no one was informed by our independent callout that such a situation needed to be addressed. These are hard studies. I take my hat off to all of the physicians who were prepared to leave the 98% right coronary artery stenosis because no trial existed previously to tell us that we shouldn’t do that. PRAMI had been completed. There were no safety issues in doing both lesions at the same time, even though it was a longer procedure and it was more dire, but it was not associated with any excess bleeding or contrast-induced nephropathy — no safety issues at all.
A Change in Thinking
Dr. Gershlick: This was in contrast with data from many of the previous registries[5,6] that said that this approach was associated with harm. I suspect that the reason for those findings (which led to the American guidelines giving this a class III indication) was selection. We are taught that patients with cardiogenic shock(although we had very few cases because they shouldn’t be included) need total revascularization. If a sick patient comes in, we do everything. If you look at the registry data, many of the patients who had complete revascularization were sicker. Only approximately 20% of patients in the registries had complete revascularization initially. That suggests that they may not have died because they had complete revascularization but because they were sicker, which led them to do the revascularization.
It has changed my thinking. People who are better at this than me will decide about guidelines, but for now, we have two trials with very strong signals.
Dr. Montalescot: Two trials from your country, PRAMI and CvLPRIT, say that we should look very carefully at these other coronary arteries and probably treat them rapidly.
Dr. Gershlick: One would believe that you would now question leaving a tight stenosis in a non-infarct-related artery. They are not the same as stable patients.
Dr. Montalescot: While awaiting the change in the guidelines, what are you going to do tomorrow? Are you going to treat all of the arteries in the same sitting? Or are you going to take the patient back to the cath lab the next day?
Dr. Gershlick: It depends. We can only be guided by the evidence base. If people believe that this is part of the evidence base, then we have to be guided by it. It’s not the whole evidence base — it’s a small part of it. We should dampen down any extreme views about this.
What would I do tomorrow? If my patient had a blocked left anterior descending artery, with a very tight right coronary artery, I would treat it at the time unless the patient wasn’t doing well. It’s always easier to do things in trials than it is in real life. We have to believe the evidence. This was a very robust trial, independently funded by the British Heart Foundation and the National Institute for Health Research. We had an independent clinical trials unit and an independent Chair of the Steering Committee, with lay membership. I was very “hands-off” during this trial.
Dr. Montalescot: It is beautiful work. Congratulations to you and the others. I’m sure that it is going to have an impact on the guidelines. Thank you very much.
Dr. Gershlick: Thank you.