Hi, this is Eric Prostowski, and welcome to another episode of EP on EP. It's an absolute delight to have a friend of mine for many years, Dr. Ed Gerstenfeld, who is professor of medicine and director of clinical electrophysiology service at UCSF, to discuss a topic that I think is very important, and that's whether scar in the heart makes a difference in how you select people for defibrillators who are at risk for sudden death. So Ed, welcome to the show. Thanks, Eric. It's a pleasure to be here. Now let's get right on with it. These are two patients from my own collection, Ed. The first is a 42-year-old man who underwent optimal medical therapy and had a non-ischemic dilated cardiomyopathy with class two heart failure symptoms. Sort of a routine case that one would see for possible ICD involvement. I went ahead and did a CMR before considering an ICD, and in his case there was negative delayed enhancement. So how does that affect your judgment, Ed? How would you approach a patient like that? He has no symptoms, and you said you've done some monitoring and any non-sustained ventricular tachycardia? Good point. We didn't do extended monitoring. I think it was maybe a Holter, and PBCs, no non-sustained VT, and symptoms of heart failure, but nothing else. Gotcha. And his EF, you said, is around 30%. 30% after several months of good medical therapy. This is obviously a common problem that has been somewhat controversial over the years, which is obviously why you're bringing it up. We know that the risk in general in patients with non-ischemic cardiomyopathies is lower than patients with post-myocardial infarction cardiomyopathies. There's a fair bit of controversy in the literature. There was this definite study, which was a large study run by Alan Kadish in patients with non-ischemic cardiomyopathy, randomized the defibrillators or not. And although there was a trend, that did not hit significance in terms of benefit. And the real study that gained approval, I would say, for implantation of a defibrillator and showed a benefit was a ScudHeF study, which did show a mortality benefit to implantation of an implantable defibrillator. But I use a ScudHeF study, and I keep those numbers in mind. If you look at the mortality benefit in general, it was around 7% over five years. I keep that in mind when I'm talking to patients, that over five years is about a one and a half percent per year improvement in your mortality. So it's not 50%, right? It's not 30%, but one and a half percent. Is that important? While it may be important to some patients, it may be less important to others. There's also obviously downsides, which is why you're bringing that up, that go along with a defibrillator. There's risk of shocks when you don't need them from a normal fast heart rhythm. There are infections, and those on average, most studies may be 5% or so over those five years or higher. So you're looking at about a 1% risk of not mortality, but some potential downside, and maybe one and a half percent of a mortality upside. This whole area has become more controversial since the Danish study, which randomized about 500 patients, again, with modern therapy. So the ScudHealth study, you could say, was done probably 10 years ago. We didn't have as good medical therapy. We didn't have biventricular devices. The Danish study randomized patients with modern medical therapy, and about half the patients in those studies, and that study got biventricular pacing. I didn't ask you about the EKG in this patient, so I'd be interested to know if it's a narrow QRS or wide QRS. Yeah. No, it's not a CRT candidate. Good point. Okay. Narrow QRS. So in that study, there was no overall mortality benefits to non ischemic cardiomyopathies. Again, ejection fraction under 35% who receives an implantable defibrillator. If you look at the subgroups, though, one important subgroup is younger patients, right? Patients under 70, in general, did have a mortality benefit. And then you bring up an important point, which I think when you're sort of on the fence is imaging. And obviously, intuitively, I mean, I tend to think, I'm sure you do, that you need some scar potentially to facilitate reentry or VF. And studies have shown that, although, as we've mentioned, there's some controversy. So this Danish study did have an MRI sub-study where they looked at it, and about half the patients had good quality MRIs. And it turned out if you had scar that was associated with higher mortality, but the defibrillator, at least in that study, did not change that mortality. It reduced sudden death by about half, but didn't change overall mortality. The other big recent study is this Gutman study that was done in Australia. Again, about 500 patients, retrospective. But in that group, people who had a scar on MRI were the ones who benefited from the defibrillator. The ones without a scar did not benefit. So again, if you're on the fence, I think the presence of scar does certainly make me worry more. And I would say, you know, increase the risk. You know, in this patient, EF's 30, so lower EF makes me more worried. So EF in the, you know, 35% makes me less worried than an EF of 25%, so 30's somewhere in the middle. But 30% is getting me a little worried. They have PVCs but not non-stained VT, and they're in their 40s. They have some, sounds like class II heart failure. So again, what I do, and I think, you know, some of us think, you know, we have to have the answer, right? When we go talk to a patient, this is what you need. But as you know, it's a discussion, and you know, the hashtag is the sort of shared decision making, but it really is. You sit with your patient, and you don't have to have the answer, you say, listen, there's a chance that you may die suddenly, right? That it could happen. That chance is not huge. I give them the SCUDF data, I say, you know, probably 1.5% per year over the next five years, and this device could save your life. On the other hand, there are risks. You could get a shock when you don't need it. You could end up having a problem with a device or an infection. And then you have that, you know, discussion with the patient. I think in someone like this, where they're in their 40s, they have a lot of potential life ahead of them. They don't have severe class IV heart failure, where they have competing causes of death. They don't have renal failure on dialysis. I think they have a lot of potential life to be gained. So even though, you know, the MR showed a scar, it might be a little more forceful. I would say that I think an ICD is certainly reasonable in this patient. Well, that's a great discussion. So I'm a little, I've changed over the years. I was, as you know, most people know, I've been fairly bullish on ICDs over the years. And there are multiple papers now, you alluded to a couple, that have shown that in the absence of scar, and you and I are EP, so we understand the concept, scar does tend to promote reentry and it makes sense to me pathophysiologically, Ed, the data that's been coming out with scar burden, at least in the non ischemics. So it's interesting. I was leaning with this patient actually to not put an ICD, but as you said, it was a shared decision making discussion. And the other thing is, there are some of these patients who developed, who have, you know, late recovery of ejection fractures, you know, we've all seen that. So I did just what you suggested, I went and spoke to him and I did tell him that here's the risk pool, you're in it, but you're at, but the scar not being there. I told him, honestly, you're at the lower end of the risk pool. Yeah. I would agree. Yeah. And he actually opted for the moment not to have one, but he wants to think about it some more, which I think is very reasonable. And I gave him the same discussion you did. I think one of the reasons I wanted to discuss this with you is, I think a lot of our colleagues forget to get the CMR before they go right for a device. And I'd like to get your opinion. Do you think this should be a routine test before you consider device? I think so. It's certainly come to be our routine. In addition, I don't know if it'll add much all the time, but I usually will get genetics as well. Cause sometimes you might, although it's rare, see I've picked up some high risk mutations of filament C or Titan splicing mutations that may put them at higher risk. So I think these days in 2020, I would get genetics and I would absolutely get a CMR. I mean, there are, you know, one of the challenges is once the device is in, you know, there are techniques to minimize noise, but you'd never get as good a quality CMR. And you see patients all the time who've had this device implanted and the images aren't as good. So I think in terms of looking at the risk stratification, looking at baseline, it is useful. And I want to emphasize one other point that you made, because I've had lots of these patients I follow as well, who looking at the risks and benefits decide or, you know, not to get an ICD. But the important thing is, you know, you don't have to make the absolute yes or no decision today with me in the office, right? You can think about it. And if you don't want the ICD, we'll continue to follow you. I think that's important in terms of documentation, you know, I think one of the reasons people just put an ICD in everyone is they're worried about, you know, medical legal consequences. What if two years later, this patient drops dead, right? Are you going to be liable? Well, I think you document that you had the discussion and you tell the patient, I'm going to continue to follow you. I'm going to get, you know, another monitor in a year, maybe another MR in two years. And you know, we can see if that fibrosis has changed. And if things change, you start having non-stained BT, we can decide to put the defibrillator in later. So I think documenting that you're going to continue to follow the patient, that you may decide to put it in if things change that you had that discussion are all, you know, important. So just, we only have a couple of minutes, but let me ask you something that's bothered me. It bothered me from the view that I don't quite understand it. Most of the literature would suggest when it's looked into subgrouping, that just the presence of fibrosis in a non ischemic, not the percentage of fibrosis, puts you at risk. Not just that sort of mid-wall fibrosis you typically see on these CMRs, but for the hypertrophs, it seems to be different. I mean, there are several papers suggesting that it's a percent of fibrosis. For example, 5% doesn't put you at anywhere near the risk if you had 15 or 20%. Why do you think, and I don't have an answer to this, but why do you think if you can have 5% fibrosis in a non ischemic dilated myopathy and put you at a higher risk, but in a hypertroph, it seems to be not as much? I don't understand that. Yeah. I mean, my reading of the literature, Eric, is that it is a gradation, you know, for when you're writing your paper, you like a cutoff and I view it, you know, again, the data clearly the 15% for hypertrophic cardiomyopathy was the best cutoff in terms of risk. And, you know, no one wants to hang their hat and put an ICD on a single risk factor of fibrosis alone. But I think even from that data, there is a gradation that, you know, 10 to 15 is higher than five to 10, which is higher than less than five. I think that's true in non ischemics as well. I think it's similar to hypertrophic cardiomyopathy, the wall thickness, right? So everyone knows greater than 30 is high risk, you absolutely would get an ICD. But what people don't realize is there's a gradation. So 20 to 25 is higher than less than 20 and 25 to 30 is high. So if I have a hypertrophic with a wall thickness of 28 and his fibrosis burden is five, that makes me more worried than if their thickness is 18. So I think these are statistical cutoffs and, you know, so I don't think that 14 is necessarily that much lower than 16, but five probably overall is lower risk. So it makes sense in general that more scar is higher risk, and that there are these cutoffs that help differentiate them, I wouldn't hang my hat on a particular number. But I will tell you this, it's funny, you said 28%, if you will believe this, I had a patient with 28, I mean, 28, not percent, 28 millimeters, and the insurance company turned down this person who didn't have non-sustain at that point, but had fibrosis that had 28 millimeters, typical Anthem, you know, kind of request, and they said, no, they actually turned it down. I had to monitor this patient for like three weeks, and I got, I think, one four-beat run of VT, you know, and then they said, okay, now, I will tell you personally, I think that fibrosis, significant fibrosis is a higher risk factor than non-sustained VT in at least over than 30 hypertrophy. I have found non-sustained VT to almost be nonspecific in my experience over the years. Yeah, well, I think it depends, and as you know, nowadays with these 14, you know, 21-day continuous monitors, it's different than the 24-hour holter we used to do. It's pretty unusual for a 14-day monitor not to see one run of non-sustained VT, and again, non-sustained VT, it's not a dichotomous variable, right? You see runs of 20 beats, you know, of pleomorphic, that's going to make you more worried than one or two four-beat runs. The other thing I was going to mention, just to wrap up, when it comes to quantifying the delayed enhancement, that's also not necessarily a black or white process, right? So I've shown two different radiologists and gotten two different quantifications, and there's different methodology for that. There's the, you know, based on standard deviation of normal, there's based on this half, you know, half at pulse width. So you shouldn't talk to your radiologist, as you probably do. I sit down with the radiologist and go over these scans myself and take a look, because I've seen a report, but then I sit with Charlie Higgins, you know, who developed delayed enhancement, and he says, oh, no, it's much more than that. So I think looking at the primary data yourself, looking at the, you know, the strips from the monitor, looking at the MR all the way in. Well, listen, it's been a great time with you, Ed, as always. And I think the message I hope it gets out to our colleagues is that doing a CMR gives you important information. It's part of, but not the final, right, piece of information to help you with a decision. But if you don't get one and then put it in ICD, like you said, you really can't get the pictures that you want. It's not like a pacemaker where you can still get good pictures. At least that's been in our own experience. So I think your message is it's useful. If I'm summarizing, it's part of how you stratify risk, and it adds to making a clinical decision. Is that fair? I think so. That's very fair. And again, some people may say, well, the cardiac MR at my hospital, you know, they're not as good, but there's always going to be a center probably that has good quality MRs. And obviously, you need a good quality MR and good quality readers. But I agree, especially, I don't know if you've seen it, but with the subcutaneous ICDs, there's just so much artifact that it's hard to get a good quality MR once the ICD is in. We're dealing with a patient right now, and I'll end with this, who I got involved in peripherally, a young gal who's seven months pregnant who came in with syncope and ugly looking polymorphic PT that we've quieted down in the hospital, but you can't give GAD during pregnancy. And so we've made a decision to use a wearable cardioverter defibrillator for the last, hopefully, month, then we'll hopefully get the baby out safely and cover it with some beta blockers and so forth. But just what you said, because we're headed in that direction, and if we go do that now, we will never get a good look at what's going on. And I think that's something she's going to probably need. So very good point you raised with the SICD. Ed, as always, wonderful to spend some time with you, and I'm sure our listeners will have benefited from it. Stay safe and stay healthy. You too. Thanks, Eric. It's great being here, and I hope this is helpful for the Heart Rhythm audience.

EP on EP

Eric Prostowski

Dr. Ed Gerstenfeld

scar in the heart

defibrillators