Jody Hurwitz from Dallas, Texas, welcome. Is this on? Great. Thank you so much, Madam Chairwoman. It's a pleasure to be here, and I appreciate the invitation to speak. I interpreted this topic as primarily geared at defibrillator therapy for treatment of sudden death and ventricular arrhythmias in older patients with heart failure, so that's primarily what I'll focus on, it looks like. Hopefully you can see that pointer, at least for now. My disclosures were shown previously. And just by way of outline, I'll address kind of does age matter, are ICDs effective in older patients, and then just because we can put them in, it should be. So I'll start with this really nice figure that was in a review by Dr. Curtis and her group looking at arrhythmogenesis in older adults. It essentially highlights the many translational and basic science factors that go into different arrhythmogenesis contributors in older adults across the spectrum of cardiovascular disease, including those with normal hearts and heart failure. And I think it nicely highlights many of the issues. This is also brought up in spades in this older kind of meta-analysis of the amiodarone trialists group of about 6,000 patients, which demonstrated, as you can tell in the red, increased incidence of mortality overall with older age groups, but also higher rates of sudden death. So I think there's two take-home points here. One is that as the age of the patients gets older, the rate of sudden death increases. However, the proportion of overall deaths at sudden death is relatively lower. So the incidence rate goes up, but the proportion of overall deaths is lower. This is also highlighted to some extent in this figure that we borrowed from Dr. Barty and Dr. Mitchell in a review looking at causes of death across heart failure class, but I think could be extended to across the spectrum of age, where you have a large proportion of patients with ventricular arrhythmias contribute to death. But a not insignificant group of patients that have Brady-related arrhythmias and maybe important implications for older patients that receive devices, we may be preventing deaths due to non-tachyarrhythmias, but more potentially Brady events as well. So I think that's an important implication in this patient group. Nevertheless, patients that are older are less likely to get defibrillators. These are data from Dr. Hess at the University of Colorado looking at Get With The Guidelines and Medicare-linked analysis. From 2005 to 2011, there was an increase in rates of patients getting ICDs, but there still remained a gap, particularly among the oldest patients. And over that period of time, the patients that were admitted with ICDs, the rates primarily were derived by previously implanted rates. In other words, the rates of actual implantation during the Get With The Guidelines admission didn't actually go up. So it appears that these patients are, I don't know if you want to use the term undertreated, but are certainly receiving ICDs at lower rates, whether or not that's appropriate. This is probably partly derived from guidelines which historically have cautioned against the use of primary prevention defibrillators in patients who do not expect to survive with good functional capacity for at least one year. Certainly that's, for some patients, easier to judge than others, but obviously as you get older, the likelihood of survival for one year goes down. So what evidence do we have for the use of defibrillators in older patients? Well, this is a sampling of some of the major defibrillator trials, both primary and secondary prevention, and the mean or median ages for these groups. As you'll note, all were 65 or under for mean or median age, so the vast majority of the patients younger than 70 and very low proportions of patients over the age of 75 to 80. So our randomized trial comparisons are relatively limited in this group. So how do we know if ICDs are effective for older patients? This is another analysis from Get With The Guidelines Medicare, and obviously there are limitations to the Medicare analyses in that over 65 has become a much less stringent definition of older, and we tend to look more at the 75 to 80 group. But nevertheless, in Medicare patients that are hospitalized for heart failure in a non-randomized but adjusted comparison, patients with ICDs tend to appear to have better survival. My colleagues at Duke looked at this in a meta-analysis and the influence of age and the hazard ratio of death receiving an ICD versus not. So this is a hazard ratio of equivalence and a probability with 95 percent confidence intervals. As age goes up, we cross the unity line, and the benefit of defibrillator becomes less clear. It appears that kind of at the 65 to 70 range, at least by this analysis, there may be a questionable benefit or benefit that comes into question as people age. I think this is also highlighted in spades in the recent or less recent now, but in the Danish trial, which as many of you know, looked at ICDs in the non-ischemic primary prevention population, and among the interactions was that with age, with a significant interaction value as patients got older. So that's primary prevention. This is data in secondary prevention. This is a nice NCDR analysis of patients receiving defibrillators for secondary prevention. I think there's two important messages here. As people get older, from the blue, red, purple to green, obviously the risk of adverse events goes up, whether it's cumulative death, hospitalization, heart failure, or SNF admission. But even in these older patients that have had an event and they're getting a device for secondary prevention, mortality is high. But on the flip side, 70% of them survived at two years. And so this question of what's the likelihood of the patient surviving for one year with good quality of life, it can be a challenging one to answer when you have a greater than 80-year-old that's getting a secondary prevention ICD that has more than 50% survival at two years. And I think one other consideration to think about is the utility of CRT in older patients. Again, non-randomized trials, Medicare patients, so many closer to 60 than 80, but demonstrating consistent effects of CRT for all-cause mortality, readmission, cardiovascular, and heart failure, indicating probably similar mechanisms of benefit. Again, non-randomized data, but left bundles doing better than non-left bundles, again, suggesting similar types of benefits in the older patient population. But clearly not one size fits all, particularly as patients age. So I think there are a few considerations, and I'll spend a little bit of time on these. But in terms of ICDs in older patients, certainly we need to weigh the benefits, the primary versus secondary prevention, whether they're non-ischemic versus ischemic, whether they're candidates for cardiac resynchronization therapy, and what their comorbid disease is and whether that contributes to their benefits. Obviously we'll also need to balance the risks of placing an ICD in terms of whether they have another indication for a device anyway and will it be receiving a procedure nonetheless. What their comorbid disease is will also influence their risk of the procedure, obviously. And then certainly I think among the most important are the patient's quality of life as they age and personal considerations. So some of the data I showed earlier demonstrated a questionable interaction of age and treatment benefit, and I think one of the things that does not get included in many of those analyses is the comorbidities of those patients. This is an older analysis from the MADE IT 2 study looking at likelihood of benefit and a risk score, and so they took these several parameters, renal dysfunction, neurocardioassociation class, presence of atrial fibrillation, older age, and QRS duration, and found that as you accumulated more and more of these, your benefit from ICD, at least in that trial, went down. So for the lower-risk patients, those with zero to about two of these factors were more likely to benefit from ICD than those with more. So even if you didn't have age, if you had three or four other factors, you were less likely to benefit. On the flip side, you could have just had age and potentially still benefited in that trial. More recently, we looked at this again in another meta-analysis of primarily non-ischemic ICD trials stratifying patients with less than two comorbidities or two or more. There appeared to be a benefit of ICD, but that benefit was attenuated for patients with more than two comorbidities here on the right. So in the blue is those with ICD, in the red those without. We did a similar figure to the one you saw earlier. So on the X, the number of comorbidities, and on the Y, the expected hazard ratio of ICD for all-cause mortality versus none. And again, 95% confidence intervals with a hazard in the line of unity. So at about two-ish comorbidities, it appears a challenge in terms of benefit for defibrillator therapy in patients. These comorbidities didn't include age, so take that with a grain of salt. But nevertheless, is age a surrogate for comorbidities in the older patients that might not have many of these comorbidities? Certainly much more likely to benefit. I think lastly, the question of quality of life is a challenging one, but one that's extremely important, particularly in our older patients. And so this is a recent meta-analysis of a number of ICD trials that looked at quality of life among patients that received ICD versus those that did not. By and large, they found the vast majority had no significant difference. There's some differences, but overall they concluded not a dramatic difference in quality of life among patients that received an ICD versus those that did not. This included all comers, so as you noted earlier, the ages of patients included in those trials was relatively younger, and so this did not focus on the older patient group. And in fact, this is largely the amount of data on older patients in quality of life for ICDs. At the University of Utah, however, we're relatively fortunate to have a veritable wealth of data on quality of life. And so I've pulled together some of the preliminary data on patients in our heart failure clinic that have ICDs. And so shown here is a variety of quality of life metrics, so a visual analog scale, a similar five-point general health scale, a measure of depression, and a measure of satisfaction with their life roles and activities. I'll call your attention to the depression or the scale, which is higher is generally better quality of life with the exception of depression. The higher your score, the more depressed the patients are. And when you look at patients with heart failure that have defibrillators in increasing age groups, they note lower health status. Their depression scores, however, are not higher, and their satisfaction with their roles and activities is not necessarily lower. And so these are preliminary unpublished data in a relatively smaller sample size that don't necessarily suggest a dramatic impact of quality of life for ICD patients as they get older. Lastly, there are some data on cost effectiveness in older patients with ICDs. Again, limited to Medicare age patients, so 65 and older. This is Jillian Sanders from Duke and her group looked at quality of life across the different trials of defibrillators. So this first figure is looking at age at implantation, and cost effectiveness of defibrillator therapy per quality adjusted life here gained. The scud heft is the outlier here, but as you can see, as the patient population ages, the cost effectiveness, the cost goes up, not necessarily dramatically, and there's not really an inflection until you get at about 80, and even then it's still kind of in the $50,000 to $60,000 range for what that's worth. They looked at this also by frequency of generator change and duration of battery. So at the time this was done, they used a base case of about five years. I think that's relatively modest in terms of battery. Longevity for these primary prevention devices, and as you get out longer in terms of longevity, obviously the cost effectiveness becomes more favorable. Lastly, they did also look at quality of life reported after the placement of a defibrillator, and not surprisingly, as the quality of life decreased, the cost effectiveness became less favorable, but at a base case of about .5 or .9 remained in this $50,000 to $75,000 per quality adjusted life here gained. The authors of this concluded that there still remains unclear in terms of cost effectiveness, depending on the threshold that you might set for the use of ICDs in primary prevention for older patients. So in conclusion, ICDs can be effective in older patients. Certainly we need to weigh the benefits versus risks. I tend not to look only at chronologic age, but much so on biologic age, and a lot of that is influenced by comorbidities. I think cost effectiveness remains unclear, but what remains extremely important is that, particularly for older patients, this remains a very personalized decision. And age sometimes is one, maybe a major factor, but is often only just one in many factors that might influence our use of defibrillators for the prevention of sudden cardiac death. So this is a nice figure from a recent review on the topic that contributes nicely. So with that, I will conclude, and I'm happy to take any questions. Thank you. Let's open for questions. We have one here. Thank you for your wonderful presentation. In your figure where you showed that by ICN analysis the benefit of ICD goes down with more comorbidities or age, did you take into consideration ischemic versus non-ischemic? So in the analysis we did with the five trials and the comorbidities, ischemic versus non-ischemic actually was one of the comorbidities. So it was treated as did you have coronary disease as an additional comorbidity. The others included renal dysfunction, lung disease, peripheral disease, a variety. It's complicated, right, because not every trial reported them exactly the same. But in that particular analysis we used coronary disease as a comorbidity, yes. But there are other analyses out there that have used a variety of different comorbidities that have consistently shown, not surprisingly, that the higher the number of comorbidities, the lower the likelihood of benefit, very likely due to the competing risk of non-arrhythmic death. I would expect the non-ischemics to benefit even less. I agree, I agree. Because the risk is lower. Yeah, and that wasn't necessarily the target of that analysis, but I agree, and that's why I put up kind of ischemic versus non-ischemic as a consideration, not just in light of the Danish trial, but in the setting of the ischemic that may be much more responsive to ATP therapy for their arrhythmic event and less likely to have a decreased quality of life from shocks in that setting, might more likely have stable VT that can be, again, treated and maintain a quality of life. So, yeah, I agree. Thank you. This was a really very nice review of an important issue with, of course, data that isn't perfect. Since all we have in the older population is observational data, we have to realize that the people who didn't get ICDs and did get ICDs, that includes a clinical decision made by somebody, which probably already integrated many of these factors. We presume that people make decisions rationally. So just as a thought experiment, if you actually were going to randomize a group of people between 70 and 90 to an ICD versus no ICD for usual primary prevention, what would you expect to find? I think you would defend heavily on the composition of that cohort, right? So if it's a primary prevention, ischemic, otherwise relatively low comorbidity cohort, I would expect to find a benefit for those patients with our relatively advanced therapies for coronary disease and heart failure at this point that have reduced mortality. Alternatively, if the enrollment includes a predominantly non-ischemic cohort with multiple comorbidities, I think you would be challenged to find a treatment benefit. And that also is due to the, and I didn't show this, but as you know, there's a number of figures out there demonstrating the incremental benefit of ICDs having decreased as our medical therapy has improved and our incidence of arrhythmic death has decreased. It's been increasingly challenging to improve mortality in that cohort. So forgive me, I will push a little bit and say it depends on what the population is that gets enrolled. And I would be amiss if I didn't also add the fact that a lot of these studies include many more men than women. And, of course, as people get older, that tends to change. There are more women. So that could answer your question as well, couldn't it? Yes, and I'm embarrassed to say that there are data out there, as you know, in GEM and other analyses looking at the underutilization in women, but also the maintained benefit in women. So I apologize, that was omitted. No, that was great. Very nice job. Thank you.

defibrillator therapy

sudden death

ventricular arrhythmias

older patients

heart failure