Rahul N. Doshi, MD: Hello. Welcome to this program titled "Early Risk of Mortality After Revascularization: What Do We Need to Do?" I am Rahul Doshi from the University of California at Irvine. I am serving as moderator of the program. I am an electrophysiologist. We will introduce our distinguished panel, but a few introductory comments first.

Together with bypass surgery, the widespread adoption of percutaneous coronary intervention (PCI) has allowed for tremendous improvement in the management of acute myocardial infarction (MI). However, despite this there is the risk of arrhythmic death (something near and dear to my heart, of course), and the risk is highest shortly after MI. Identification of which patients are at highest risk and the use of appropriate therapy can save lives, but considerable controversy exists in this setting, specifically what should we do in this period post MI, postrevascularization.

Today, we will examine those risks; we will evaluate measures to prevent sudden cardiac death (SCD) in patients with compromised left ventricular ejection fractions (LVEFs) after surgical and/or percutaneous revascularization.

I am pleased to introduce Dr Mina Chung from the Cleveland Clinic, who is also an electrophysiologist; Dr David Kandzari from the Piedmont Heart Institute in Atlanta, Georgia, who is an interventionalist; and Dr Eric Roselli from the Cleveland Clinic, who is a cardiothoracic surgeon. They will be able to offer their unique perspectives. Welcome everyone. Thanks for being here.

Let us start off specifically addressing the risk of SCD after MI. There are numerous data sets and evaluations looking at this issue. Dave, could you address some of the data from VALIANT and other data sets post MI?

David E. Kandzari, MD: Sure, Rahul. To begin, this is a challenge that affects all of us as cardiovascular practitioners independent of our disciplines. In particular, there is a remarkable consistency across the data that highlights the fact that post-MI patients, and especially those with LV dysfunction, are at especially high risk for SCD. That risk, that sudden cardiac arrest risk, is one that really never mitigates over time. If we follow these patients long term, the event rates never really converge to those [of] individuals who have preserved ventricular function. This risk is especially high within the first 30 days, and it is high even extending on a little bit further into the first 90 days.

We know this from a succession of trials from more dated studies like the CADILLAC trial^[1] to even more recent observations like the American College of Cardiology (ACC)- National Cardiovascular Data Registry (NCDR) following MI that the presence of LV dysfunction post MI portends a risk of mortality even at 90 days that may exceed 11%, maybe as high as 30% in selected patient cohorts.^[2]

VALIANT, as you have referenced, highlighted the fact that the risk of SCD in that early 90-day period was at least 2- to 2½-fold higher compared with other individuals with preserved ventricular function.^[3]

What is also striking to me with regard to the uncertainty of the timing of these events, the unpredictability of them, is that over 80% of these events occur after hospital discharge. We can develop risk models that we can discuss, and we can quantify that risk numerically for mortality in that early phase, although it probably raises more questions than answers.

Dr Doshi: You mentioned these very high event rates after ST-segment elevation myocardial infarction (STEMI). Certainly, these patients with low EFs in this post-MI, post-PCI setting are at a high risk. Are there any unique challenges that you think we need to overcome clinically? For example, should these patients be encouraged to have an automated external defibrillator at home? That has been addressed in a clinical trial, the HAT trial.^[4] Are there any other recommendations that you want to make?

Dr Kandzari: One of the challenges or dilemmas of being a critical care physician is that while we may feel reassured that we do primary PCI for these patients and we revascularize them, I am humbled by the fact that revascularization does not necessarily protect against or mitigate against that risk. Many of these people have other residual disease after primary PCI for STEMI. We have this kind of staging period where there is still residual disease that portends even further risk for these patients. Are these people at greater risk than someone who had single-vessel disease who was revascularized? For me, the biggest challenge is the unpredictability of recovery of LV function. We have individuals post MI who may be awaiting bypass surgery, who may be awaiting stage revascularization, who may have been revascularized. The big challenge is, can we identify those individuals whose LV dysfunction is going to improve, remodel, supersede an EF of 35% to 40%? We just do not know.

Dr Doshi: Dr Roselli, could give us your thoughts from the surgical standpoint, following these patients clinically after bypass?

Eric E. Roselli, MD: Sure. First of all, it is great to be on this multidisciplinary panel because you are right; it is a common problem for all of us. The surgery patients who we see often have multivessel disease, and we may get to them in a delayed fashion after their acute MI. The fundamental problem with the patients we see, and it is true preoperatively and postoperatively, is we do not know how any one specific patient is going to respond to revascularization, or how any territory of myocardium is going to respond to revascularization.

One of the things we do before we operate is a lot of viability studies in these folks with low EF.

The problem, and we have seen this from the Surgical Treatment for Ischemic Heart Failure (STICH) trial analysis, is that it is not that predictive of exactly how these patients are going to respond. It does not give us the data that we really want to know. What viability studies give us is a sense of the safety of revascularizing patients. In other words, we know that if a patient has a certain amount of scar burden in an area that we plan to revascularize, it is probably not safe to go ahead with that revascularization. We should think about another line of therapy.^[5]

We can get them through bypass surgery for multivessel disease if they have some viable myocardium in the areas that we want to revascularize, because we know that doing so prolongs survival and takes care of angina, but we do not know that we are going to make their myocardium work better; we do not know if we are going to improve their EFs. We do not know necessarily whether we are going to eliminate the SCD risk. Even in the postoperative period when we have patients who we are more worried about, we cannot predict exactly which one is the patient who [we will] send home and is going to have SCD.

We need to understand these patients better and understand exactly what is happening to them, to protect them better. That is where this discussion of the wearable cardioverter-defibrillator (WCD) comes into play, or when is the right time to place an implantable defibrillator in these patients. The 90 day waiting period to some patients is a really scary period of time.

Dr Doshi: As electrophysiologists, we have conflicting viewpoints, and we have conflicting data sets. We know there is a very high-risk population that exists post MI, postrevascularization, whether it is with PCI or bypass surgery.

We have these imposed waiting periods, 40 days post MI, 90 days postrevascularization, and this is based on clinical trials.

DINAMIT was a negative clinical trial looking at the early adoption of an implantable cardioverter-defibrillator (ICD), although it did dramatically reduce arrhythmic death.^[6] Is this an issue with competing mortalities with nonarrhythmic death rates, or higher in the ICD arm in DINAMIT, as well as in IRIS^[7]?

There are other unique challenges that bring up the issue of the WCD. I mentioned the HAT earlier, which was a negative trial using automated external defibrillators. These would be fine if you think that all these events are happening at home after discharge. It is not so nice if they are not witnessed or they happen during sleep, where you are not necessarily going to have that type of resuscitation.

Dr Kandzari: Rahul, you bring up a very good point, though, that maybe not all of these individuals who we are describing are going to merit an early ICD implantation if it were available to them, or even a WCD, but they all should be screened for such. When we talk about trials like DINAMIT being a negative study overall, it may be negative in mortality because there are other, competing factors of renal failure, infection, and other issues in a very sick population post-MI that contribute to mortality. Still, at least more than half, maybe perhaps 60%, of the mortality events are SCD. The trial did show a reduction in arrhythma-mediated death. Perhaps that is the element for the art of medicine rather than the science. We need to have better predictive models of identifying who is going to be at risk for SCD and who would benefit from early intervention versus just the broader population itself of people post MI.

Dr Roselli: We saw the same thing in folks who had defibrillators after coronary artery bypass graft surgery (CABG), too. If we go back to the 1990s in the Coronary Artery Bypass Graft (CABG)-Patch trial, a negative study, but a 45% reduction in arrhythmia deaths when you do a focused cause of death analysis.^[8] It is all supportive...

Dr Kandzari: With a very difficult surgery.

Dr Roselli: Exactly.

Mina K. Chung, MD: That was an epicardial defibrillator that was put in very early generations, so theoretically our morbidity and mortality rates after defibrillator implantation transvenously should be much lower today than they were back then. The difference, I think, between a post-MI population and a revascularized population is that we actually have IRIS and DINAMIT data looking at defibrillator versus no defibrillator implantation within the first 30 to 40 days, whereas we do not have that data with modern-day defibrillators for a postrevascularization population. People have tried to look at the randomized clinical trials, primary prevention trials, of defibrillators. Most of those primary prevention trials excluded people within the first 2 to 3 months after revascularization. The substudies that have tried to look at early benefit of mortality, even those, they start at 6 months, so we really have not looked at the very early mortality within the first 3 months.

Dr Doshi: Obviously this is a very complex issue. There are many controversies here. Maybe that is the perfect segue to speak about identifying risk in these patients. Maybe we should not be committing them to an ICD. Mina, could you show us your data related to the WCD?

Dr Chung: As you know, ICDs for primary prevention of SCD have been shown to improve survival in patients with EFs of 35% or less. They are only reimbursed when they are implanted 3 months after revascularization, or 40 days after MI, or 3 months after newly diagnosed cardiomyopathy. If you look at event curves in many of these conditions, mortality risk is usually much higher early after cardiac events, including after MI or cardiac revascularization.

If you look at the large registry from the Society of Thoracic Surgeons Adult Cardiac Surgery Database (ASCERT) study, that looked at 300,000 or so primary CABG patients over age 65 that went to the Centers for Medicare and Medicaid Services (CMS) databases, there is an early mortality risk if you look at the survival curves. It is kind of a dual-phase mortality curve.^[9]

If you look at patients who have an EF less than 30% in that cohort, it is even more marked. This early phase is actually within the first 3 to 6 months.

That also is seen in the NCDR CathPCI registry, perhaps even more markedly. They divided patients into those who had STEMI and those who did not have STEMI. Again, this registry had over 300,000 patients in it. With or without STEMI, you see this marked high mortality that is very early on, again within the first 3 to 6 months. That is particularly marked if you look at the population with low EFs, and they used a cut point of 30%.^[2]

We wanted to look at this. For years I have been intrigued by the mortality curves we see after CABG. In MUSTT and MADIT, they excluded the first 3 months. I looked at our mortality curves and there was a little bit of dip early on. We decided that we would look at our revascularization population at the Cleveland Clinic who were not discharged with a defibrillator or a WCD and compare those patients with EF less than or equal to 35% to patients who were in a national registry of WCD users who had their WCD put on because they were being discharged from a PCI or a CABG and had an EF less than or equal to 35%.

If you look at those data, there were about 4900 patients. If you look at PCI plus CABG patients shown in the red, you see this early mortality that I was mentioning. That is in the inset of the figure and you see an instantaneous hazard curve. It shows that there is very early instantaneous hazard risk, or early hazard risk. If you look at the blue curve for the national registry of WCD users, you do not see that early hazard, or it is less marked.

If you break this down into looking at the survival early in the first 90 days, you see that it is significantly different. Then if you look at after 90 days, the curve is still split a little bit. There is some potential early benefit. Now this was not randomized, but it is a comparison. Eric, you were also on the report. We also broke it down by CABG and PCI.

Dr Roselli: Yes. It is interesting that the type of revascularization did not seem to matter much. When we look at the curves, we see an early hazard risk shown both in the inset, and you can see by the big dip in the red curve on this figure. The late hazard phase does not seem to be that impressive; the curves flatten out later on. There is an early separation. When we look at these folks, there is now a more focused view on the curve on the left where the x-axis represents months.

When we look at the first 90 days, we see that there is a separation between those with and without the WCD, such that the mortality was only 2% in that group as compared to 7%, and that was significantly different. Beyond 90 days, you take out the folks who did not survive the first 90 days. Again, we still see this separation in the curves.

I think that this is definitely telling us something about these patients. There certainly are limitations to where the data came from, but it is a big study and it is consistent across all these analyses.

Dr Kandzari: As we introduced this topic, we felt like revascularization, interestingly, did not protect against an arrhythmic event. If anything, we see that postrevascularization, there is this early phase where patients are at risk for greater events. Perhaps this is an opportunity for early intervention or early screening for these types of people. As someone who revascularizes, what is it, Eric? Is it the inflammation postoperative that is contributing? It is not necessarily that because it happens with PCI as well.

Dr Roselli: True. What we are showing is death. These are sick people. Certainly there is more of an inflammatory burden on the patient who undergoes surgery, but when we look at these next set of slides, it shows PCI. We see that the early hazard risk is even higher. It is not just the type of procedure; it is not just what we are doing to them. We see this greater separation in the early hazard risk in the PCI patients because we have more patients having acute MIs. Again, there is your inflammation hypothesis.

Dr Kandzari: Before we go to the PCI data, I’ll ask you about this: Are you suggesting an enduring benefit even after the 90-day period or the period of a WCD? Is that just a function of us following these patients more carefully, or is it that they all then transition over to an ICD? What is the mechanism there?

Dr Chung: It could be because the split is more marked early on. That was the period that they wore a WCD. The splitting of the curve after 90 days is hard to explain, but it is more marked in the PCI cohort. It is hard to know if maybe that cohort is not getting as much follow-up to reassess their LV function. I think after CABG, they come back for their follow-up. After PCI, we think, “Oh, we have opened the artery and you are done, you are good,” but they may not be good if their EF is still low.

Dr Doshi: That is the plug for seeing an electrophysiologist after their revascularization.

Dr Chung: More job security.

Dr Doshi: More job security. It is interesting. You would hope, and I think this is borne out in the data sets as well, that these patients were more likely to receive an ICD than they would having the vest. Could you specifically address the event rates with the WCD? Could this be the explanation for these dramatic differences in that 90-day period?

Dr Chung: It does not explain everything. There is a several percent difference in their mortality rates in the first 90 days. The mean or median duration of use of the WCD was 3 months or less. There were events that happened in about 1.3% of the patients. It does not explain the entire benefit. It is possible that they could have gotten attention earlier, maybe moved on to a defibrillator. We know that almost a third of the WCD population ended up getting defibrillators. We do not have that number for the no-WCD group. It is possible that they got better follow-up, and they potentially could have had arrhythmic symptoms, or atrial fibrillation, or warnings that the defibrillator was about to go off, and that is not accounted for by the actual shocks that were delivered by the WCD.

Dr Kandzari: Mina, I am always impressed about the lack of compliance and adherence to medical therapy after revascularization and after acute MI. We have studies suggesting that discontinuation of antiplatelet therapy may exceed 10%; β-blockers, other medications similarly. What struck me about WCD data is that, maybe it is a part of natural selection, but for patients who wear it for a short period of time, they do not typically wear it throughout much of the day; but for people who adhere to it longer term, the compliance or adherence tracks more throughout the 24-hour period. Did you see this? Have you tracked compliance in your study?

Dr Chung: Not in this study, but in an earlier study we did. There was actually pretty good compliance, something like 85% to 90% of patients wore the WCD for more than half the time.^[11] The median use was actually not as low as what you would think. They are using it. It also could potentially make them seek follow-up, asking should I continue to wear this, or should I not continue to wear this? Patients want to know that.

Dr Roselli: You could probably guess that someone who has to wear the WCD is being constantly reminded to be compliant with all of their care.

Dr Chung: That is a good point.

Dr Roselli: It sort of scares them into being compliant with everything.

Dr Kandzari: Do we need randomized trials in this situation, or is this bordering on unethical that we should do this? What are the situations where we might find an opportunity for randomized trials based on DINAMIT and studies that we have referenced already?

Dr Doshi: It is a great question. There is an ongoing prospective clinical trial that is larger but similar to DINAMITcalled the DAPA trial. It is a larger cohort. It is a little bit different. It is 30 to 60 days post MI. I think the EF cutoff is 40%. This will be interesting data if a mortality benefit can be demonstrated. Then we could make further distinctions based on what we think are comorbidities, whether the patient should have an implantable device or be bridged to see if they survive, or they are capable of survival for non-SCD reasons. I think these data are still hypothesis generating. It is interesting that the whole story is not just arrhythmic survival. Is this just a function of good follow-up plus the added benefit of clear event rates that are real in this high-risk population?

Based upon the data we have, and I think we have fairly good data suggesting who is at high risk, are all of you taking data from CADILLAC or from VALIANT or whatever else, are you risk stratifying your patients, taking into account low-EF patients? Age, presence of diabetes, renal insufficiency, what if they were not completely revascularized, you have a low EF and making a decision to get a WCD? Or, you are not a high-risk patient so we will stratify you differently.

Dr Roselli: Certainly there are patients we are more worried about based on their anatomy, their burden of disease, any other comorbidities that we know put them at high risk, but we do not have a dedicated way to assess that. It is still the art, as you described earlier.

Dr Kandzari: The first step is in screening these patients. As an interventional cardiologist, we need to take some responsibility as the first point of contact for many of these people coming into the hospital for their primary PCI event at initiating the screening process, and then working with our colleagues, our consultants, electrophysiologists, as well. There is this window where the patient gets discharged, goes through a fairly uneventful course, and then, while they are waiting to go see Mina or Rahul, some adverse event occurs. There are predictive models, like from CADILLAC, that we can apply to identify who is going to be at greater risk, people who are elderly people who have renal impairment, but by far the greatest predictor that is going to drive those risk models is LV dysfunction.

I do not know exactly who is going to have an event with LV dysfunction. We know the risk is considerably higher for them. Maybe not everybody needs to have a WCD or go on to an ICD implantation, but screening all of them should be a quality core measure in our hospitals.

Dr Chung: Often we do not see them after PCI. They get discharged very soon after that. We see the post-CABG patients if they are having nonsustained ventricular tachycardia or some worrisome arrhythmic event. As an electrophysiologist, we do not have a chance to intervene prior to discharge, except in those very high-risk patients. It is actually the surgeons and the interventionalists who are the first line.

Dr Roselli: We are a little bit limited though sometimes. We call you to see a patient who maybe has not had an arrhythmic event, but they have an especially low EF and we get them plugged into the system, but then we have to wait.

Dr Chung: That we have to think about, whether or not we should be protecting them in those first 3 months. If you are very worried about them, a bridge is reasonable. I am struck by how many of these patients with an LVEF less than 35% actually make it to over 35% by 3 months. It is between a quarter to three quarters in a couple of reports, so you cannot really predict. With that, then you do not want to go and slap a defibrillator into everybody. Perhaps a WCD is a good bridge. I would love to see a randomized trial.

Dr Roselli: I would agree.

Dr Chung: There is a randomized trial, VEST, going on post-MI. I am really interested in the outcome of that.

Dr Doshi: We have not addressed some of the other issues, such as whether we are fixated seemingly after their intervention, their bypass, or their percutaneous intervention. What about the patients who are waiting for revascularization who are at high risk? Most CABG is still elective. A staged PCI would be another good example of a high, high-risk population.

This has been a great discussion. It illustrates a lot of the controversies, but I think we are speaking about a very high-risk population that we need to do something about.

Any final thoughts as far as any challenges with this therapy?

Dr Kandzari: I would just highlight that there is this large unpredictability that you characterized of who is going to improve and who does not. This is such a common issue for all of us across our disciplines that screening really is probably the first step for us to begin.

Dr Doshi: With that, Eric, David, and Mina, thank you very much for participating on the panel. You have shared a great amount of data with everybody. Thank you all for participating in this activity.

To proceed to the CME posttest, you have to click the Earn CME Credit link on this page. Again, thanks to everyone.