Do ICDs Have a Role in Nonischemic Cardiomyopathy?

Naveen Pereira, MD; Paul Friedman, MD; Abhishek Deshmukh, MBBS


June 19, 2017

Editorial Collaboration

Medscape &

Naveen Pereira, MD: Greetings. My name is Naveen Pereira, and I am a professor of medicine at Mayo Clinic in Rochester. During today's recording, we will be discussing implantable cardioverter defibrillators (ICDs) and the important role that they play in congestive heart failure, specifically nonischemic cardiomyopathy.

It is a privilege and pleasure to have my colleagues, Dr Paul Friedman and Dr Abhishek Deshmukh, who are electrophysiologists and are well reputed in this area, join me. Welcome, Paul, Abhishek.

Paul Friedman, MD: Pleasure to be here.

ICDs in Nonischemic Cardiomyopathy: A Brief History

Dr Pereira: We know that ICDs play an important role in reducing mortality in congestive heart failure. Patients with congestive heart failure die from progressive heart failure or sudden cardiac death. ICDs kind of preempted sudden cardiac death. Initially, ICDs were shown to improve survival in patients who survived their [first] episode of sudden cardiac death. I think those were the earliest studies historically.[1] Then subsequently, [researchers] went a step further to ask: Can we put ICDs in patients to prevent that first episode of sudden cardiac death? Is that how the history unfolds?

Dr Friedman: That is exactly right. What is striking is if you [compare] the mortality reduction between the two groups, it was higher in patients who had not yet had an event because in order to get into those early trials—the first MADIT study, the MUSTT[2] study—patients had [to have] more ventricular dysfunction. The key point to come from those early trials was that defibrillators can extend life in people who are at high risk for events, and it is the degree of structural abnormality that leads to the arrhythmic events and premature death.

Dr Pereira: We have this really interesting study called the [Danish ICD Study in Patients With Dilated Cardiomyopathy] DANISH trial.[3]Can you put it in the context of available literature? Why now? Why another ICD trial now when it has been pretty well established in the community that ICDs reduce mortality? It is obvious that it can help in treating sudden cardiac death in patients with congestive heart failure, so why do we need another trial?

DANISH: A Contemporary Trial With Optimal Background Therapy

Abhishek Deshmukh, MBBS: That is a great question. If you look at the history of the trials performed for ICDs, particularly in nonischemic cardiomyopathy, it started off with the [Cardiomyopathy Trial] CAT study and [Amiodarone Versus Implantable Cardioverter-Defibrillator Randomized Trial] AMIOVIRT in the late 1990s.[4,5] Then, we had the [Sudden Cardiac Death in Heart Failure Trial], SCD-HeFT[6] analyzing a subgroup of patients with nonischemic cardiomyopathy, followed by the [Prophylactic defibrillator implantation in patients with nonischemic dilated cardiomyopathy] DEFINITE trial.[7]

Dr Pereira: The SCD-HeFT subgroup analysis showed that patients with nonischemic heart failure and an ICD survived at greater rates than patients without ICDs. Is that right?

Dr Deshmukh: SCD-HeFT did show that ICDs reduce the risk of sudden cardiac death in patients with nonischemic cardiomyopathy. That led to guideline recommendations from the scientific societies. However, medical management has dramatically improved. There has been more [emphasis on] cardiac resynchronization therapy (CRT).

This was probably an ideal time to reassess whether ICDs on top of the medical management and CRTs confer an incremental benefit compared with, say, no ICDs. That was the background under which [the DANISH] trial was conducted. The trial started in 2007, and it took a while to enroll and present the data. Although the data are contemporary, it is still very important for us to know whether ICDs [are effective] in patients with nonischemic cardiomyopathy and their [longer-term] implications as far as costs and quality of life are concerned, particularly in societies where ICDs are not readily available.

Dr Pereira: Regarding improved medical therapy, we have the PARADIGM-HF trial of Entresto (sacubitril/valsartan, Novartis)[8] in patients with heart failure with reduced ejection fraction that showed a rather dramatic relative risk reduction in mortality. I believe fewer than 20% of those patients had ICDs. This does reflect the fact that medical therapy has advanced.

In the context of modern medical therapy, we needed to do another ICD trial. Can you comment, Paul, about cardiac resynchronization therapy (CRT)? How does that affect patients with nonischemic cardiomyopathy?

Dr Friedman: The real difference in the DANISH study is that roughly 60% of patients in both arms had CRT devices. So this is not a study of medical therapy vs medical therapy and device as most of the previous studies have been. This one is [about the] best heart-failure therapy, best drugs; add CRT to that, and then does a defibrillator help?

Some of the key take-home messages are: First, medical therapy is powerful and important, and our current guidelines indicate patients need to be on good medical therapy before consideration for an ICD, and this reinforces that finding.

Second, it highlights the role of resynchronization, because one of the reasons the ICD benefits may be blunted was the fact that patients did receive CRT. Importantly in this trial, patients were getting good medical therapy—100% essentially on beta-blockers and ACE inhibitors or [angiotensin-receptor blockers] ARBs, over half on mineralocorticoid therapy—so medical therapy was robust. That is an important factor and reminds us that it is critical that we have patients on good medical therapy before we offer device therapy.

The other [point] it highlights is that one-third of patients with heart failure die from noncardiovascular causes, meaning no device can protect against those [events], unfortunately. That may factor into which device for which patient given their age.

Dr Pereira: Abhishek, can you briefly summarize the design of the trial?

Dr Deshmukh: The DANISH study was a randomized controlled trial that enrolled approximately 1150 patients and divided them into two groups, one receiving an ICD in addition to optimal medical management and CRT whenever indicated. A second group did not receive an ICD but did get a CRT pacemaker, if indicated. The mean follow-up was 67 months. The primary end point was all-cause mortality, and there was essentially no difference between the two groups.

Interestingly, the researchers did note that in patients who had an ICD, the chance of sudden cardiac death was about 50% less than in patients who did not have an ICD. What ICDs typically do is prevent sudden death, and that was an important secondary end point that was met.

In a prespecified subgroup analysis, [the study] showed that patients younger than 68 years of age tend to have mortality benefit from an ICD compared with certain older patients. The study had a good follow-up, and we look forward to more extended follow-up.

Dr Friedman: It is interesting when you look at the survival curve of the whole population. At 2 years, you start to see a separation: ICD is beneficial. But as you keep following patients out to 6 years, the survival curves start to come back together.

It may be that the patients who were older were starting to die from other causes. In my mind, one take-home message is that a CRT device may make sense in the older, frailer patient if [a potential benefit is] indicated. [Conversely, it] may be appropriate to use the more aggressive ICD therapy in younger patients with nonischemic cardiomyopathy.

Dr Pereira: What is the average age of the patient in the trial?

Dr Deshmukh: The median age was 64 years.

Dr Pereira: That is interesting. In a modern-era trial of heart-failure therapy—great medical therapy, great mechanical therapy, almost 60% on CRT—[there is] no difference in terms of total mortality or a reduction in sudden cardiac death. How do you define "sudden cardiac death" in patients with or without an ICD? I thought that was a little confusing.

Defining "Sudden Death" in ICD Trials

Dr Deshmukh: The way the [researchers] describe "sudden death" is [simply] dying suddenly. They do make a disclaimer that [this includes] sudden vascular death, but it is hard to know how [patients] died suddenly.

Dr Pereira: Right, and there was no obvious secondary cause.

Dr Deshmukh: Right.

Dr Pereira: So those with an ICD had sudden death too?

Dr Deshmukh: It is interesting that ICDs do not prevent all sudden deaths. Because as people get sicker and if they have other metabolic issues—refractory acidosis, hyperkalemia—ICDs' function may become limited over a period of time. Also, it depends on programming. If the device has not been programmed optimally, there is obviously a chance of proarrhythmia, which can be potentially detrimental.

Dr Friedman: It can be pulseless electrical activity, it can be noncardiac pulmonary embolism, which may look the same. Those are the sorts of things that lead to someone dying suddenly or someone who was well the last time they were seen and they were found dead and they have an ICD, and the ICD may not have stored an episode if there was not tachycardia that would trigger storage of the event.

Dr Pereira: That is interesting. In this trial or previous trials, have [researchers] interrogated the devices when people die suddenly?

Dr Friedman: That has been done. The cause of death in patients with heart failure, as you noted, depends in part on the stage of congestive heart failure. In patients where the heart failure is less severe (New York Heart Association class 2 or class 3), they are more likely to have tachyarrhythmias that lead to recurrent shock, [ventricular tachycardia] VT storm, those kinds of things. Patients with more severe heart failure are more likely to have [pulseless electrical activity] PEA, bradyarrhythmic, or systolic failure–type heart deaths.

What to Do Now?

Dr Pereira: If we have to advise a patient today—say a 70-year-old patient walks into our office and he has nonischemic cardiomyopathy, how would you initiate the discussion?

Dr Friedman: A couple of points are worth making. First, the trial highlights the importance of good medical therapy. Second, we rarely change practice strictly on the basis of one randomized controlled trial. Third, since the trial was completed, meta-analyses[9] have included greater numbers of patients and have demonstrated a benefit in nonischemic cardiomyopathy of ICD in addition to other therapies.

I think we have to be careful before we take the results of this single trial and tell a patient, "You do not need an ICD." On the flip side, [the study offers] useful information. Remember, 5% of patients had an inappropriate shock, and so there is something to be said for a device that cannot give you a painful shock but can extend your life. The more aggressive ICD CRT pacemaker (CRT-P) would be appropriate for someone who has a number of other comorbidities. I think it highlights the thoughtful consideration of that on an individual basis, but I do not believe the evidence is strong enough to suggest that we should say an ICD is no longer warranted in nonischemic cardiomyopathy.

Dr Pereira: Should we be considering CRT-P first because we see that, especially in patients with nonischemic cardiomyopathy, ejection fractions, for example, improve? Say a patient comes in on maximum medical therapy and is a prime candidate for resynchronization therapy. The ejection fraction was 25%; now it becomes 50%, so it is an interesting dilemma. Should we use CRT-P, because now he is not eligible for an ICD because his ejection fraction has improved?

Dr Deshmukh: There have been studies—particularly the [Comparison of Medical Therapy, Pacing, and Defibrillation in Heart Failure] COMPANION trial[10]—which looked at medical management vs CRT pacemaker vs CRT defibrillator(CRT-D). Interestingly, comparing CRT-P and CRT-D found no difference in outcomes in cardiomyopathy. So, it is helpful that CRT pacemaker could be a potential [option] in patients who may not want an ICD, especially where cost is a consideration.

Dr Friedman: In the younger patient, I would still offer a CRT defibrillator device. In the patient who is older, who has comorbidities, who does not want the risk of an inappropriate shock (which is not huge, but it was 5% to 6% over the course of the study), the CRT-P is an attractive option.

Certainly in nonischemic cardiomyopathy, the incremental benefit of defibrillation above what the CRT offers is more modest. It is important to emphasize we are referring to nonischemic cardiomyopathy. In ischemic heart disease where you have a fixed scar, modus for reentry, the benefit of defibrillation is far greater.

Dr Pereira: What is our take-home? Do we say medical therapy is great? CRT can help these patients? Are we going to change our practice? Are the guidelines going to change?

Dr Deshmukh: I am not sure if the guidelines will change right away, but it emphasizes, as Dr Friedman mentioned, about CRT, about medical management, and an ICD. It's important to note that not everybody is going to have a cardiovascular death. That should be discussed with the patient. Some people may prefer dying suddenly without having an ICD. Those realistic expectations should be discussed with the patients.

As far as future investigations are concerned, we need to come up with a better marker to assess who will need an ICD beyond the ejection fraction, beyond ECG—maybe MRI and delayed enhancement. Some of this work is currently in progress.

Dr Pereira: So we need to figure out how to risk-stratify these patients better.

Dr Friedman: This has always been an issue. Predictions about the future are challenging. And yet we have to make decisions based on our best available evidence, and so to get to your question of—what do we do right now? In the nonischemic cardiomyopathy patient:

  1. Optimize medications; there are powerful data[11] that show they are very impactful, so they need to be on best medical therapy. We cannot shortcut that.

  2. For the younger patient, in particular (you know, under 68, under 70), I would still offer CRT-D based on meta-analysis and the findings from DANISH itself.

  3. For patients who are older, have more comorbidities, I think an individualized informed discussion saying the additional benefit of the ICD component is less robust, and then proceed on that basis.

Dr Pereira: That is a great summary and conclusion. Paul, Abhishek, thank you so much. This has been, certainly very informative, and I appreciate you all taking the time to do this. Thank you for joining us on on Medscape.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.