Good afternoon, ladies and gentlemen. I will open the session on phenotypic differences in genotypic arrhythmia cardiomyopathy subcategories. This is a joint session between HRS and ECAS, the European Cardiac Arrhythmia Society. My name is Richard Hauer. I am from the University of Utrecht, and my co-chair is Hugh Colkins from Johns Hopkins in Baltimore. The first speaker will be Dr. Platonov. He will give a presentation on electrocardiographic differences in genotypic ACM, arrhythmogenic cardiomyopathy subcategories. Dr. Platonov. Thank you very much for this introduction, Richard. It is really my pleasure to stand here and to talk about the ECG features of genotype-specific phenotypes of arrhythmogenic cardiomyopathy. And I'm going to start, sorry, let's see how it works here, there we go. These are the features that are most commonly described, and among the classic diagnostic criteria included in the RBC diagnosis is the T-wave inversion in right tricorder leads that varies regarding the number of leads involved, and depending on their number, it could constitute either minor or major criterion, according to Task Force 2010. There are depolarization abnormalities. One of them is the terminal activation duration, which if it exceeds 55 milliseconds, that constitutes minor criterion, and that's measured in right precordial leads from S-wave to the end of QRS complex. And finally, the feature that we've been struggling to define exactly what it is, epsilon wave, it is there, but a number of studies show that we are not always in agreement when we try to identify what it is, and when it's present, most of other criteria are also present, but it's still there. And these are the classical features of ARVC, and that is matching the most often the phenotype associated with the plocafilin-2 mutations in these patients. When it comes to non-PKP2 ACM genotypes, then the most common are DSC2, DSG2, and desmoplacin variants, and the DSC2 and desmoglien, they exhibit a lot of similarities with regard to ECG manifestations. One of them is the T-wave inversion in right precordial leads, which is rather consistently showing up in the same leads that are observed in the PKP2-related ACM genotypes, and also the depolarization abnormalities also confined to the right precordial leads. They may differ to some extent in the appearance, but still that's measured from the S-wave to the end of QRS complex. It's fairly well matching the definitions used in the PKP2-related cardiomyopathy. When it comes to desmoplacin-related ACM, then the phenotype is different. As you see in this example, the T-waves are not necessarily inverted, but what is striking is that the amplitude of the QRS complex measured in the extremity leads is low, and that's the feature of desmoplacin-associated mutations. This ECG series illustrates the progression of electrical substrate over 10 years of time in a patient with a DSC2 mutation. As you see here, it starts very early with a T-wave inversion observed in right precordial leads, and at that time, there is not much seen as depolarization disturbance in right precordial leads that we would expect to see. We have it here, but not in the V1, for example. A few years later, it develops and persists until, in this case, over 10 years of follow-up time, the patients advance to the transplantation stage, and it was the last ECG before cardiac transplant was recorded at 49 years of age. And at that time, also, the more dramatic, more visible T-wave inversions were observed. But the changes in ECG not only confine to the right precordial leads, the QRS amplitude also diminishes over time, not as dramatically in desmoplacin-related SEM phenotypes, but it's also consistently observed in other phenotypes as well, but rather as a more late sign of the disease. When it comes to DSG2, which is more common than DSC2 variants, these are the three patients that we had in our database that I wanted to show with you what the ECG may look like. These three patients carrying different DSG2 variants exhibit ECG feature matching classic description of the ARVC phenotype, as in placophilin 2. You see here the T-wave inversions observed in the right precordial leads. You see the depolarization disturbance that could be very dramatic in these patients with very, very delayed epsilon wave, I would say, here. But in other cases, it's just prolonged terminal activation depolarization. Very consistent finding observed in these patients. This is one patient for whom we had 24 years of ECG follow-up with a DSG2 variant, and this patient was diagnosed after cardiac transplant when the heart was examined and genotyping was performed after that. And as you see here, this patient started with showing the T-wave inversion in right precordial leads, not necessarily affecting the depolarization phase, QRS complex, which came later, 10 years later here at 50 years of age when this started to get prolonged. And these changes were not confined to precordial leads only. It was also affecting, if you look at the amplitude of QRS complexes, over 24 years of time, the amplitude got, it was small in the beginning, got even smaller at the end of the... of the ECG taken before a cardiac transplantation was performed. So if we talk about DSC2, DSG2 cardiomyopathy, then the ECG phenotype is very similar to the classic glycophylline-related phenotype. Diagnostic test for criteria are useful. T-wave inversion in right precordial leads may appear early in the course of the disease, and low-voltage QRS may develop, but not necessarily in the early phase. We looked at this in a large study performed based on the Nordic ERVC registry, where we followed up more than 350 patients with ERVC, and analyzing nearly 7,000 ECGs. And in this data set, over 80% of patients had either plocophylline 2, desmogline, or desmocolline mutation carrier. So this is really, these findings from this study really stand for the typical arrhythmogenic cardiomyopathy phenotype. What we observed is that the... all ECG parameters we looked at appeared trend that was measurable before patients fulfilled diagnostic criteria, which is time zero here. And also in comparison with the control population, that was a background group of patients, 18,000, that were admitted to hospital for various reasons, but not having arrhythmogenic cardiomyopathy as a diagnosis. And you see that terminal activation depolarizations was growing up, age-related, and beginning to be prolonged before diagnosis, and continues going up afterwards. When we looked at the QRS voltage, it appeared the same dynamics, was starting to go down years before the diagnosis, and continues afterwards. And this is a prevalence of the... this is the ERV amplitude at the first ECG recorded, diagnostic ECG, and the latest ECG recorded. And you see that the trend is very, very clear in all leads that we looked at. Same was observed for T wave amplitude and the percentage of patients experiencing T wave inversion in different leads. Again, that prevalence growed from the first ever ECG recorded to the diagnostic and the latest ECG recorded. When it comes to desmoplacan-related cardiomyopathy, then the situation is different. This is a patient that was diagnosed at 19 years of age, and over two years follow-up, we've observed a reduction of the QRS voltage, quite dramatic. And even at... already at 19 years of age, this patient had a number of PVCs, fibrosis by MRI, and we chose to implant ICD because her mother was resuscitated from cardiac arrest a few years before that. We observed that the T wave inversion is not present, even though the QRS voltage is very low already in the beginning. So when it comes to desmoplacan-related cardiomyopathy, then task force criteria often not fulfilled. Repolarization is frequently normal or show unspecific changes. Low-voltage ECG is very common, and occasionally we may see left-recorded T wave inversion V5, V6. We've even looked at signal-average ECG in the large group of patients in the Nordic RBC registry, and compared the value of that sign for the diagnosis, and did not observe any differences between different genotypes in terms of diagnostic validity, probably because the signal-average ECG is not lead-specific, and so that it sort of performs exactly the same way in different. For LAMB-BAN patients, I don't have any, so I had to borrow this ECG from Richard, and this is a specific type of arrhythmogenic cardiomyopathy that is characterized by low-voltage ECG in inferior or left precordial leads with common left QRS axis deviation and frequent T wave inversion in right precordial leads and prolonged TAD. Also frequent T wave inversion in the left precordial leads. So to conclude, PKP2, DEC2, and desmoglin2, they share classic ECG phenotype where task force criteria are useful. Even RBC risk calculator is useful because the ECG signs of the phenotype are very, they share the same pattern. Desmoplakian is characterized by low-voltage QRS that may appear early in the course of the disease, and T wave inversion in left precordial leads that also may appear, but not necessarily in right as for the classic RBC phenotype. For that reason, task force criteria are less useful, and RBC risk calculator is not useful either. So alternative risk calculators are developed. Phospho-LAMB-BAN, limited number of genetic variants with described ECG phenotype, and for that reason, again, task force criteria are less useful. And we don't have that many patients with our genotypes to speak about any systematic ECG patterns, and data on that sort of, that type of ACM is limited. And I thank my colleagues from the Nordic RBC Registry who stand for a large work that our registry is performing, and thank you very much. Thank you very much. It was a nice presentation, nice schematic presentation. I have just one question. Your example doesn't show any negative T waves in desmoplakian mutations in the left precordial leads. Is that your usual pattern? You mentioned already that it was rare. We have the same experience, but I have no explanation. I'd say that we have few patients with desmoplakian, and that particular patient that I showed, she's very young, and she did not develop, she did not appear to have the desmoplakian to T wave inversion in left precordial leads, but older patients that we have, they develop that. But QRS, reduction of QRS voltage is apparently the most clear, most obvious sign of the disease of those patients. And can you make another comment about epsilon waves? I know you wrote a beautiful paper about epsilon waves in terms of how they should be used in clinical practice. Do you think they're part of the 2010 diagnostic criteria? Do you think they belong there, or do you think they need to be eliminated? We looked at this question a few years ago by assessing how different experts from different registries would classify ECG patterns in terms of epsilon waves, and we found ourselves in a situation that we do not agree on the basic ECG characteristic of the disease. And in addition to that, we looked at whether it actually contributes to diagnosis, and it appeared that in those cases where some of us thought that this is epsilon wave, there were all other criterias necessary for diagnosis, where epsilon wave was not necessary. So in my opinion, and I think it's shared by many of our colleagues, it doesn't have to be included in the diagnostic criteria. Okay, well, thank you very much, it's a beautiful talk, I appreciate it. So our next speaker is Dr. Agarwal, who will speak on imaging differences in genotype ACM subcategories. Thank you very much, it's a pleasure to be here. Can we go back? I think we have the prior speaker here. So, I'm an imager, and I'm going to talk about the imaging diagnosis today, and we're going to look at the diagnosis of arrhythmogenic cardiomyopathy from an imaging lens. So in the next few minutes, we'll discuss the imaging phenotypes, discuss about the differential diagnosis, and from the imaging point of view, discuss the challenges as well as the opportunities. And our understanding of genetic arrhythmogenic cardiomyopathy has really grown from having arrhythmogenic cardiomyopathy with an RV predominant phenotype to having an LV predominant phenotype and everything in between, a biventricular type. We've had several fantastic talks about genotype to phenotype. We've discussed desmosomal genes, placofilin 2 being the most common one. Among the LV predominant types, there are desmosomal as well as particularly the desmoplakin, but also nondesmosomal genes like FLNC and phospholambin. And we'll discuss a little bit about how this translates to the different phenotypic presentations. And this has really transcribed into our evolving understanding of how these translate to the criteria for imaging and diagnosis and imaging REM. If you look at the very, very original criteria, this is the 1994 criteria, and what really strikes you is that these are really primitive. You can see that they are very broad strokes. So we're talking about severe dilatation and reduction of RVEF. So it's very subjective. We're talking predominantly or rather just exclusively about the right ventricle, and they're completely qualitative criteria. It took almost 16 years, so by 2010, we had these criteria, the task force criteria, where we had the more quantitative parameters at this point. But even at this time, there was no framework for LV involvement or diagnosing LV predominant diseases. Also, one major lacking at this point was we did not have or we did not utilize one very important feature of cardiac MR, which is tissue characterization. So there was no way of utilizing delayed enhancement or utilizing tissue characterization of cardiac MR. So we are, and we look at in subsequent slides, how we can really utilize some of the delayed enhancement in diagnosing LV predominant cardiomyopathies. So this gave way to these Padua criteria. We'll discuss this. There were some revisions to the RV criteria, but one of the biggest advances was that for the first time, there was a discussion about LV predominant criteria. There was incorporation of delayed enhancement in these criteria. And with some minor tweaks emerged these European task force criteria for diagnosis of arrhythmogenic cardiomyopathy. And one of the biggest takeaway messages here is the diagnosis or the element of ring-like left ventricular LGE. So there's one thing you take away as clinicians, and a lot of you might not be interpreting cardiac MR, but if there's one thing you take away from this talk is that if you see this kind of ring-like left ventricular delayed enhancement, then it should evoke or it should at least make you think about the possibility of arrhythmogenic LV predominant cardiomyopathies. And when we say ring-like LGE, we're talking about involvement of at least three or more segments of delayed enhancement. And these are typically sub-epicardial, linear, or mid-myocardial delayed enhancement. So think about that possibility in your mind. So with this, let's talk about what are the various imaging phenotypes, and we'll start off with the RV predominant phenotype. This is a very, very typical example of that. If you look at the right ventricle, you can see this patient had a dilated RV. The MR is really a great tool for this. The end-diastolic volume was 120 milliliters per square meter. There was a depression of systolic function. The EF was estimated to be 38%. But very nicely, you can see these kind of focal dyskinetic areas within the RV free wall, and that really clinches one of the major criteria for RV involvement in this particular case. Now what are some of the challenges? It's always not that easy on MRI. This was one of the very striking cases that I shared with you, but unfortunately, we grapple all the time. And this is because the clinical picture is unfortunately not clear all the time. The imaging findings are often subtle, and this is a very high-stakes diagnosis. So it's often you are dealing with the situation of, you know, the findings are subtle, should we raise that possibility or not? And I'm just showing you a few situations. We often have these kind of areas near the moderator band where you see these areas of hypokinesis. Fatty infiltration, although it's not in the criteria, it's a very, very frequent finding on imaging. If you have pectus excavatum or you might have RV tethering, then kind of finding these focal areas of dyskinesis can be very challenging. And if you have been referred a patient with suspected ARVC, then there is often incumbent on that imager to find those areas, and it can become challenging. So just to share a clinical case with you, this was a patient who had frequent PVCs, was referred for the possibility of finding ARVC. This patient had a dilated right ventricle. The end-diastolic volume was elevated at 212 milliliters per square meter, had a depressed, mildly depressed EAF. And now, of course, the most important thing is to look for any regional wall motion abnormality. And you can see that in this situation, the readers start finding subtle wall motion abnormalities. So this was suggested that this was fulfilling a possible major criterion. But it is in situations like this, it's very important to look for any alternative causes of RV dilatation. Now a lot of times when you have a suspected ARVC, we will do black blood imaging to look for fat. At least earlier, this is what we used to do. And when you look at this, if you pay attention to the innominate vein, you can find the cause for RV dilatation. This patient actually has this anomalous pulmonary vein draining into the innominate vein, which explains why this patient's right ventricle is dilated. So you can see how in this particular case, we have a clear explanation for right ventricular dilatation. This was an anomalous pulmonary vein leading to right ventricular dilatation. Definitely not a case of ARVC, and there was no regional wall motion abnormality. I'll just show you, this is basically, so what are some internal checks you have to make sure. One thing very easy to do is compare internal stroke volumes, ventricular stroke volumes. If you have a dilated RV, you normally, if you don't have any significant regurgitation, the stroke volume should be roughly similar. So if you have a discrepancy, think about why. Is there a shunt? Is there something else that explains that enlarged RV? When in doubt, don't be afraid to check out face contrast imaging, look for that septum, look for the pulmonary veins. Just to show you another companion case, and this is something I just saw recently. This was a patient, a 15-year-old kid who had a pulmonary embolism CT, was tachycardic, had dyspnea. And this time, this patient had very nonspecific symptoms, was clear on pulmonary embolism CT, had no PE. But one thing that was overlooked was a dilated right ventricle. In this particular patient, there was no anomalous vein, there was no evidence of any septal defect. Now, this patient was, however, lucky, had an uncle who was a cardiologist who insisted that the patient get an echocardiogram, which he did. And when the dilated right ventricle was discovered, the patient underwent a cardiac MR. And at this point, you can see that this patient has a very, very abnormal right ventricle with multiple areas of dyskinesia in the right ventricle. In fact, you can even see these areas of delayed enhancement. You can see it in the inferior RV wall, along the septum, along the subepicardial LV, and later was found to be plaquephyllin 2 positive and was diagnosed with ARVC with a biventricular subtype. So again, this was, you can see how the clinical picture can be extremely confusing. You may not have a suspicion of ARVC or you may come up with a clinical suspicion and, you know, not have that disease. Moving on to the other end of the spectrum of a left-dominant arrhythmogenic cardiomyopathy, like I said, look for that ring-like pattern of LGE. And this is what I mean. You have involvement of multiple contiguous segments in this particular case, very characteristic subepicardial LGE. This is a ring-like LGE. In this particular case, it's involving at least two segments. These are both patients with known LV-predominant arrhythmogenic cardiomyopathy. Now what should not be confused is this kind of patchy enhancement at the insertion points. This is very common. It's almost within the realms of normal. Now again, just to show you another differential on the LV side, this is a patient, a younger kid, 18-year-old. This patient had a PET scan and was found to have some kind of inflammatory changes on the PET scan. And obviously, one of the first things we think about is cardiac sarcoid. So it was being considered and was being investigated for cardiac sarcoid, but the biopsy came repeatedly negative, did not have any lung sarcoid. And this is what the cardiac MR looked like. So I hope all of the people in the audience can appreciate this kind of ring-like enhancement in sub-epicardial LGE. And at least this should make us now, all of us think that could this be a genetic form of LV-predominant cardiomyopathy. So this is what it was proven to be. Now I just want to, as a parting note, say that apart from MR, here is an example of arrhythmogenic LV-predominant cardiomyopathy on CT, on a photon-counting CT, where you can see not only this kind of fat, but you can also see on the late iodine enhancement images, very similar to the MRI. If you have a patient with an ICD, you can now do photon-counting CTs with late iodine enhancement, and you can see this kind of very beautiful late iodine enhancement along the sub-epicardial part of the left ventricle and see that enhancement in the LV. So just very quickly, I want to show you some of the differential diagnosis. We can have dilated cardiomyopathy, sarcoid, myocarditis. I think dilated cardiomyopathy, to my eye, is probably one of the easiest ones to differentiate, and the reason being that the enhancement in dilated cardiomyopathy is more of an epiphenomenon. So it's typically a lot less distinct. It's mid-vowel, it's usually stria-like, and usually the LV is a lot more dilated and more dysfunctional than it is in cases of left-predominant arrhythmogenic cardiomyopathy. Sarcoid is, again, a lot more patchy in appearance. You can sometimes have that hook sign in sarcoid. The one that is the most difficult to diagnose is usually, and there's often a lot of overlap also, is myocarditis, also because we know that there is a significant amount of overlap between the presentation of arrhythmogenic cardiomyopathy and myocarditis. So here's a patient who came in with chest pain, had cardiac MR, and just to show you what this looked like, has a very abnormal RV as well as LV, look at that septum there, and has high T2 signal, extensive LGE, both along the LV as well as along the RV. And in this particular case, even though the clinical presentation was, the clinicians were very concerned about myocarditis, we raised the possibility of a biventricular form of arrhythmogenic cardiomyopathy. Genetic testing was done, and this was proven to be arrhythmogenic form of cardiomyopathy. So it is well-known that these patients often have these hot-face clinical presentations, and we've seen several population-based studies where there is a significant overlap in these patients with myocarditis and arrhythmogenic cardiomyopathies. And at the very end, I just want to show you that how difficult on the imaging side these diagnoses can sometimes be. So this is a patient with PVCs, and you know, I'm very attuned to looking at it and raising that possibility, but you don't want to overdo that. So this was a patient, came in with PVCs, and you saw this kind of what I thought was a very ring-like enhancement pattern, really thought about the possibility. And then I looked up the patient's prior PET examination. So here was the LGE, really made me think, oh wow, this is such a perfect ring. And then I just happened to look at the patient's non-contrast CT, which was done just two weeks ago, and then I found, oh my gosh, there is so much calcification in the patient's myocardium. And all of this calcification was very new, had another CT done just a month ago, and had developed this dystrophic calcification like within a few weeks. So this was all acute myocardial calcification that the patient had developed, and this was likely related to an acute period of sepsis and had developed acute myocardial calcification. And this was not arrhythmogenic cardiomyopathy, but was all related to this acute myocardial calcification. And we typically think about calcium salts as being dark, but typically when they are in this acute myocardial calcification with this necrosis and edema, they can be bright as in this case. So just wanted to share this. This was a completely very strange case, so just wanted to share this with you. Thank you very much. Okay, we are open for questions. Can you comment on doing MRIs in patients with defibrillators, how much does that interfere or not? Yes. So I think that's an excellent question, and in our practice we do a lot of these patients with, we do a lot of patients, MRIs in patients with defibrillators. There are several things that go into it. I would say the position of the defibrillator, the type of defibrillators. So one is the logistics, the second is the image quality. These are the two things that come into play. The logistics is that we do almost all of these patients. In terms of image quality, there are several things we try to take into account, so we use wideband imaging in these patients. We always have the patients raise their arms whenever possible, and that a lot of times alleviates the artifact. But having said that, depending on the position of the generator pack, sometimes no matter what you do, or even with wideband imaging, there is a significant amount of artifact despite the use of wideband imaging. And in those situations, it may not help despite whatever kind of sequencing you do. Question. Question, please. Excellent talk. Thank you very much. My question is, do you think there's, do you think in the community setting, the diagnostic accuracy of cardiac MRI in the setting of non-fellowship trained imagers, this is as accurate? What has been your personal experience, especially since cardiac MRI plays such a large role in diagnosis? Non-fellowship trained, did you say? Yes. Yes. No, I don't think that that's a good way to go. I think you need very well-trained MR technologists. The image quality plays a really big role, and then you definitely need very well-trained interpreters to make a good diagnosis. Can I just ask one follow-up? Yes. So in your experience, when you have patients coming in from outside, from the community with other cardiac MRIs, what's your experience been in the changes in diagnoses when it's been re-imaged? Yes. I think the changes are substantive. We've seen a lot of patients who've come in from outsides with interpretations where technical challenges with delayed enhancement were, where there was inadequate nulling was called extensive delayed enhancement when it was all technical factors. So yes, there are significant challenges to that. So I do think training plays a very important role. Thank you. We put together a paper a while ago on misdiagnosis of AR. Fact is, other things you pointed out, it's really very... Last question, short question, short answer. Please. Sorry, it's short. Hi. As an imager, do you see any value in measuring ECV? And if so, the sensitivity and specificity for ACM? For ACM, I don't, we don't routinely do it because a lot of our ACM patients, we've not looked into how it's helped in making a diagnosis of ACM. Let me put it that way. I think it's the pattern of delayed enhancement that has really helped us in making a diagnosis of ACM versus a more infiltrative cardiomyopathy like amyloid. So it's a more pattern that really helps us suggest that possibility. Well, thank you very much. Thank you. We will go to the next presentation by Dr. Cindy James from Johns Hopkins University Hospital, and she will give a talk on exercise tolerance in ACM subcategories. All right, well, thank you, Richard, and thank you, everyone, to the conference organizers for this opportunity to speak with you today about exercise tolerance in the ACM subcategories, and in particular, why genotype matters in exercise guidance. For those of you who I don't know, I'm a genetic counselor and geneticist. I've spent my career in Baltimore with an interest in arrhythmogenic cardiomyopathy, and these days, I'm the research director of our Center for Inherited Heart Diseases. What I'll do in the next dozen minutes or so is provide a little bit of background on what we know about the association of exercise with penetrance and clinical outcomes in arrhythmogenic cardiomyopathy and a little bit on the methods of how we got there. I'll spend most of the time talking about and walking you through the emerging evidence based for genotype-specific associations with exercise in the ACM genetic subtypes, and then talk about how we can put all this information together to provide exercise guidance in a shared decision-making context for the patients and families we work with. All right, so we've been working on this question of the influence of exercise in ARVC for a long time. Over the last 15 years, the community really has put together evidence that when we consider ARVC classic right-dominant ARVC as defined by 2010 task force criteria, we know that frequent vigorous endurance exercise is associated with earlier onset, increased penetrance, worse survival free from sustained ventricular arrhythmias and more frequent ventricular arrhythmias, worse right and in some studies left ventricular structure and function, as well as greater progression of structural deterioration, including an increased likelihood of heart failure and eventually transplant. And really in a gratifying way, those patterns are seen across cohorts, across nations, across studies and indeed even across different methodologies of collecting exercise history. In our center, we do these lifetime exercise interviews from age 10, where we ask people about exercise they've participated in regularly and then score it by duration and intensity. Other approaches include focusing on exercise really in the years preceding diagnosis, but regardless of exactly how you do it, regardless of how you cut athletes and not athletes, these patterns hold. But most of this evidence-based is made up of PKP2 cardiomyopathy patients and gene elusive ARVC. So we started on this journey of looking into gene-specific ACM subtypes in considering the influence of exercise on outcomes. And the first thing we did, now more than a decade ago, was look at this group of gene elusive ARVC patients. So we stratified by genotype and family history. Gene elusive patients were patients who met 2010 task force criteria, and they'd had the best genetic testing we had available at the time, and there was no pathogenic or likely pathogenic variant found. As you'll see here, what we found, which had been seen by Andre Laguerche and his colleagues previously, was people with gene elusive ARVC and no family history were overwhelmingly high-level athletes. They had done twice as much, twice the met hours per year of exercise, in comparison to individuals with desmosomal ARVC who were in the ACWA, who were themselves disproportionately athletes. So that was already giving us a hint that genotype mattered. Moving on, we specifically looked at exercise in PKP2 only, so this is a family study looking at 28 relatives from 10 families segregating pathogenic PKP2 variants. And as you can see here, this study confirmed that probands in the red and family members in the black who developed ARVC, as defined by 2010 task force criteria, or who had had a VT, had done high levels of exercise, and those levels of exercise were indistinguishable and much greater than the relatives in the gray who had not developed ARVC or ventricular arrhythmia at last follow-up. So that's telling us that in the case of PKP2 cardiomyopathy, exercise is associated with penetrance. The story is very different for phospholambin cardiomyopathy. This is the one Dr. Platnoff told us that he had to borrow from our Dutch colleagues, so I was lucky to be able to work with the Dutch team and looked at the association of lifetime exercise history with likelihood of developing a ventricular arrhythmia, of developing heart failure, of being diagnosed with either DCM or ARVC. And as you can see, the more and less active half of the cohorts, the red and blue lines are on top of each other in these survival curves, there was no association of lifetime exercise history with penetrance of phospholambin cardiomyopathy. I will say that in other studies, after diagnosis of PLN cardiomyopathy, exercise may still be a trigger of ventricular arrhythmia events. A lot of those events are associated with exercise. All right, the previous speaker raised the question of desmoplacan like phospholambin cardiomyopathy, a biventricular or disproportionately left dominant disease. The question of how and whether exercise is associated with outcomes was really raised by Adam Helms and Eric Smith in their very nice paper in circulation a couple of years ago, where they observed that among the 80 patients in whom they had at least some exercise history, that exercise was less common in their probands, completely the opposite of what we saw for PKP2 cardiomyopathy, and there was no association with the likelihood of having a life-threatening ventricular arrhythmia, and it was unassociated with right or left ventricular dysfunction. So what are we going to do about that? We had the opportunity last year to look at 100 patients from Johns Hopkins in our ARVC ACM registry who had a pathogenic or likely pathogenic desmoplacan variant, and what we found was our athletes in the blue were indistinguishable from our non-athletes in lifetime survival, free from sustained ventricular arrhythmias or heart failure, which is what the group of the University of Michigan saw as well, but, and this is important, our athletes in the blue were significantly more likely to want to have one of those characteristic myocardial injury episodes that the previous speaker was describing. These are these myocarditis-like episodes characterized by chest pain and elevated troponin that bring patients into care. This is important because in another series of studies that I'll talk about in great detail in two days, these myocardial injury episodes were associated with significantly worse disease trajectory for desmoplacan cardiomyopathy thereafter. These Kaplan-Meier curves are a little bit hard to interpret because they're longitudinal analyses, but the take-home message is, after you have one of these myocardial injury episodes, patients, the red lines, have a much higher likelihood of developing sustained ventricular arrhythmia and developing a heart failure hospitalization. Team M43 is most commonly observed in founder populations in the Netherlands. This is a condition with a very, very high sudden death rate and a very high risk of arrhythmias, particularly in males. So our colleagues, led by Sean Connors, took a look at 80 patients who had recently had a primary prevention ICD placed, and this disease, primary prevention ICDs, are recommended for young men nearly universally, given this very, very high risk of sudden death. And here you can see, like PKP2 cardiomyopathy, like gene-elusive ARVC, the athletes in blue had a much worse arrhythmia course than their non-athletes in the red. And finally, we're beginning to get some data on desmoglien-2, desmocolin-2, associated with ARVC. This is a paper that was published a couple of weeks ago, led by the group in Fouay, as well as some colleagues in Switzerland and France that had the opportunity to be a very small part of. And what we see here is an interesting interaction between genotype and exercise exposure. So desmoglien-2 and desmocolin-2, associated with ARVC, is disproportionately likely to occur with multiple variants. If you have multiple variants, those are the red and orange curves, your outcomes appear to be more severe and bad, regardless of the amount of exercise you've done. But at least in this more exploratory study, it looks like for those with a single variant, those participating in intense exercise had earlier-onset disease, although there wasn't a statistically significant difference in course of disease. So more to come on that. Putting this all together reminds us that ACM really is an oligogenic disease. Inheritance is not purely Mendelian. There is certainly a mix of major pathogenic or likely pathogenic variants. Some amount of lifestyle exposure, which we're discussing here, is exercise. If you already have two pathogenic variants, exercise exposure doesn't matter. If you're gene elusive, exercise exposure matters a whole lot, as well as a number of modifiers that we are still learning about. Putting it all together, this has taken us as a community from in the 2019 HRS-ACM guidelines a phenotype-based set of exercise recommendations to limit exercise, if you have ARVC, to really tailor genotype-informed exercise guidance, both for individuals diagnosed with ARVC or arrhythmogenic cardiomyopathy, and genetically at-risk family members. Within this document, there is a table we put together that summarizes the quantity and quality of evidence for or against the association of endurance exercise and arrhythmia outcomes progression and penetrance. I've highlighted a couple boxes here because I think new data has probably changed the weighting we would give that today. And then finally, how do we bring this all to the families we serve? I think humility is an important first step. There's a lot we still don't know. Those guidelines I just shared made a very large point of the importance of a shared decision process in making decisions about exercise. We had the opportunity to study the extent and implications of shared decision-making a year and a half ago for ARVC, and I don't have time to go into the details of this, but really the take-home message was there was a wide variety in the extent of shared decision-making employed by the physicians and the other providers that these patients saw. But shared decision-making was strongly associated with better psychosocial outcomes and led to absolutely no difference in the exercise decision made. I wish all my data had that sort of correlation coefficient that we're seeing here. So to end with this, genotype matters a lot to exercise guidance in ACM. In my opinion, higher-risk genotypes include PKP2-associated ARVC, gene-elusive ARVC, TMEM43 cardiomyopathy, and although I didn't talk about it, Lamin-associated cardiomyopathy. There may be a weaker impact. There's a little bit of less evidence for desmoplacan cardiomyopathy, desmoglianin-2, and desmocolin-2, but there's certainly some evidence that there is a risk, lower risk at least in terms of penetrance for phospholamban, shared decision-making for all, better psychosocial outcomes and no harms. And in these discussions, it's important to remember, even though I know this does not address the psychosocial needs of patients, that many benefits come from modest exercise. So thank you for your attention. I'm always delighted to talk about exercise in ARVC. And really thank you to all my friends and colleagues who have made this work possible. Thank you. Great talk. Marina? Yes. So, well, great talk, Cindy, of course. Just to play a little bit devil advocate. I didn't see a mention on two genes that are like filamine C and RBM20, and I know that there is probably not much up there, but maybe for me it's more like, what's your or your clinical colleague recommendations on those patients? Do you tell them to restrict exercise while waiting for data or the opposite? You know, that's such a good question. And you know how much I love evidence. And so I'm like, I'm going to walk through the evidence. Filamine C cardiomyopathy. I think it's probably, my guess is there's a risk, at least, it's kind of like PLN cardiomyopathy. My guess would be once the diagnosis is there, it probably serves as an arrhythmia trigger. RBM20, I just don't know. I'll turn it back to you. You don't understand mechanisms better than I do. What do you think? I feel that there is a lot of arrhythmias in RBM20, but again, we don't have the numbers and we're not like a big center for that, so I don't have an advice. And so I was wondering what your center is advising patients because we keep them in a gray zone because really there is no evidence-based data. And we only have like five or six in our registry, so yeah. Go ahead. Piotr. Thanks. So thanks, Cindy. How much evidence do we have to link the myocardial injury in desmoplacian to the exercise? Because from single signal perspective, those few patients that have atropine spikes, I have no connection with the exercise they had. Yeah. So one thing I didn't point out that I think is important is that there is a time lag here. It's not like when we talked about the myocardial injury setting people on a worse trajectory later, it's not an immediate thing. It's not an acute arrhythmia associated with a myocardial injury, which didn't quite answer your question, but it's important to where I'm going. In terms, likewise, I don't think this is an acutely triggered, in most cases, myocardial injury event. It's not usually like someone's gone out and, I don't know, run a 10K and then they wind up in the hospital. Although, I'll ask Hugh to phone a friend here to see if he's heard that. We have heard, I have heard other people claim that they've seen it, so it's interesting to hear that you haven't. Hugh, what do you think? No, I can't recall that immediate link. Last question by Mario Delbar. Just a comment. I think that we started this journey of trying to understand the ERVC on the idea that these molecules are in the desmosome and therefore the desmosome loses integrity and disrupts and bad things happen. Now with all of this evidence, it's very clear that that is not the whole story because both placofilin and desmoplakin are components of the desmosome and yet you are getting some very different outcomes. Something about placofilin that has probably little to do with being a component of the desmosome that triggers a pathophysiology that we see versus desmoplakin in spite of the fact that they both belong to the same components. Okay, thank you Cindy. And we will move on to our final presentation, which is by Julia Cadran-Teregni from Montreal and the title of her lecture is Risk Dependence in ACM Subcategories. Perfect. Hello everyone. Good afternoon. It's an honor to talk about risk in ACM, so the last presentation of this session. So risk-wise in ACM, we can see things in two different ways, phenotype as we're used to seeing patient, phenotype of ARVC, phenotype of non-ischemic or dilated cardiomyopathy. But more recently, we have discovered that some genes are diseases by themselves and they warrant their own risk stratification. And we now even have enough data to have some risk calculators in some of these specific genes. So let's start with the one we know the most, ARVC. Over the past 25 years probably, there has been a lot of effort trying to find the predictors of ventricular arrhythmia in ARVC and this work has progressively been integrated into different guidelines and consensus. Well, pretty much everyone recommends ICDs in secondary prevention. In primary prevention, it's always more difficult. So we started with, I'm just going to take the pointer. So initially we started with standalone criteria warranting an ICD and then the idea of combining multiple criteria came together, culminating in the idea that a risk calculator could be used as suggested by the ESC cardiomyopathy guidelines. So why do we care about doing risk calculator? The idea is always to try to improve risk prediction by trying to put the whole literature together in one single number, taking multiple predictors that may have different ways, interaction, additive value, continuous value, and then come up with a number that we can use in discussions with patients, but also in discussion amongst us when we're hesitating about the defibrillator. And then how they are made, and I think this is important to understand how we should use them. So first, they are developed on a certain population and knowing the characteristics of that population helps us know in whom we should use it. Then we determine the outcome and you'll see in all of these different risk calculators, we always used any sustained ventricular arrhythmia, which includes the sudden cardiac arrest, but also sustained and ICD treated events. The performance has to be assessed and there needs to be external validation that is of adequate size. And this is what we have for the RVC risk calculator, an initial development study that published in 2019 and two validation study of large size published in 2022. The risk calculator has seven predictors that stayed in the model. There was an eighth one that didn't stay in, which was left ventricular ejection fraction. So young age, male sex, syncope, number of PVCs on the 24-hour holter, number of leads with T-wave inversion, non-stitch sustainability, and RVEF were the different predictors in the model. The outcome, sustained ventricular arrhythmia for sure, and the population, only definite ARVC patients. So this is not the gene-first approach. Patient needed to fulfill the definite task force criteria of 2010. Patients didn't have sustained ventricular arrhythmia at the time of their inclusion, at the time of their diagnosis, however, 40 to 50% of them already had an ICD. And if you look at the derivation study and the first validation study, you see that the proportion of PKP2, the most common gene associated with ARVC that you just heard a lot about, this was the predominant genotype, expectedly because it's the most common, but it's approximately half of the cohort. In the second validation study, you see that the proportion is much lower, 21% of PKP2 patient. You also see that decimal plaquins are well represented here in the second validation study. And the event rate is pretty different. The first validation study, about 5% per year, while the second validation study had an event rate half of that, 2.6% per year. Then performance. So we usually measure performance of these models in all these publications into discrimination and calibration. Discrimination is considered to be good if at 0.70 or above. But we always need to remember what that means. So basically it means that 0.70 to 78% of the time, the model will give a higher risk to someone with an event versus someone without an event. So always have to keep that in mind and realize that these models are useful, but not perfect. And then calibration, or the correlation between prediction and observation, for which the perfect results, the dotted line. So we want to be on that dotted line. So derivation, first validation study, things went well. Second validation study, you see the overall cohort, there seem to be an overestimation of the risk with the model. And when looking at gene-positive patient, well, you see it looks much better. So going deeper into that publication, gene-positive patient, the model works well. PKP2 patient, the model works super well. Desmosomal patient, not so well. And so we remember there were some desmoplakins in there. And gene-negative patient in that specific cohort, the model didn't work so well. So going more specifically towards desmoplakins here, there were 79 patients with desmoplakins versus 118 for PKP2. And you see here the predictors that stay for desmoplakins and PKP2 are super different. LVEF seems to play a role, while all the other factors that are supposed to be in red here are not. So there's something different about desmoplakin and risk. And this has been tested. So does the ARVC risk calculator works in desmoplakin patient? And the answer is unfortunately no. With poor C-statistic, as you can see here, but it gets better if you exclude all those who have live ventricular dysfunction. And it's a shame because a lot of desmoplakin patient do have that problem. So if we want to globally simplify things about genes and ARVC, we can say that there are some classical genes. The most classical of the classical is PKP2. That is the most common and kind of influences how we see ARVC. Desmoglienin and desmocolidin, I would say risk-wise, are probably relatively similar. And plaquoglobin also kind of seem to be similar, but I would be more cautious about this one because we have way less data. On the other hand, those who are clearly non-classical and on whom we should not use the ARVC calculator include desmoplakin, TM43, phospholamban, desmin, and filamency. So moving on to these genes that also give left-sided involvement. So arrhythmogenic left-sided cardiomyopathy or non-ischemic cardiomyopathy, dilated cardiomyopathy, or whatever you want to call it. So for a long time, we use a one-size-fits-all approach. So if someone with non-ischemic cardiomyopathy had an ejection below 35%, then magic, we would implant an ICD and hope for the best. However, in 2016, a study brought a lot of humility to that one-size-fits-all approach, showing that it did not improve survival. And one way to think about that is that there may be a better way to risk stratify patient than using this one-size-fits-all approach. And genes seem to be very important, especially genes that confer a higher risk of ventricular arrhythmia, such as the desmosomal genes, such as desmoplakin, or here, lamin, and there are others. Interestingly, left ventricular ejection fraction doesn't have the same effect depending on which genes we're talking about. In desmoplakin, interestingly, the risk increases when you get below 50%, and then it stays stable. It doesn't continue to increase when your ejection fraction decreases. And in one study in filaments C truncating variants, there was absolutely no impact of left ventricular dysfunction on the risk. So now, knowing that the risk calculator doesn't work for desmoplakin, the idea of having a specific calculator came. And this risk calculator is a risk calculator, specifically a gene-first approach, meaning that anyone with a desmoplakin pathogenic or likely pathogenic variant is it. The first thing that we see is that while the sex has an opposite role, so this time it's bad to be a female, it's also important to keep in the ejection fraction when compared to ARVC, while non-sustained VT, PVC count, and RV dysfunction also stay in the model. Another very specific gene associated with high risk of ventricular arrhythmia is lamin. And lamin is a very different disease, so we can't use the same predictors because the specific thing with lamin is the risk of atrioventricular block, which is a risk of ventricular arrhythmia but should also be kept in mind when treating these patients. If a pacemaker is needed, we should put an ICD. If an ICD is needed, we should put something with a pacing capability. So we started doing, at my institution, rapid testing for LMNA when we have to take a decision quickly about a device. So we received the results in less than 72 hours and then can decide if we put an ICD or a pacemaker or a transvenous system. So there's a nice poster about that on Sunday. Then interestingly, the risk of desmoplacan, filamency, and LMNA are relatively the same. But the other things are not the same, so I don't want to give you a seizure looking at this very busy slide. But the message here in the different types of arrhythmogenic cardiomyopathy, the RVC and the different genes, and all the predictors is that you can see there's no one-size-fits-all approach working because the predictors are different. Some predictors are stronger for some genes. Some predictors are very popular, like PVC count or LVEF. But nothing is standard for these different genes. And we have the most data for the first four ones, ARVC, desmoplacan, LMNA, and phospholambin, with a validation only for ARVC. And one last thing, when comparing these different risk calculators, Cindy is laughing because she knows this slide very well, the outcome. So it's always predicting sustained ventricular arrhythmia, which includes sudden cardiac arrest, but also sustained VT and ICD-treated VT. And when thinking about the threshold we should use, it's the same for all these ACM risk calculators. We're tempted sometimes to use the same threshold as HCM because we know the HCM calculator so well, and it's 6% at five years. But this threshold should probably be different because we're predicting way more sustained and ICD-treated VT than sudden cardiac death in red on this slide. The other thing is the C-statistic, which, as you can see, is all in the same realm, telling you to stay humble when using these models. So in conclusion, our understanding of arrhythmogenic cardiomyopathy is definitely improving. A personalized approach is better than a one-size-fits-all approach. Risk prediction models are useful, but we have to stay humble when using them. Validation is important, so it's yet to come for many of these risk calculators. The calculator for ARVC, it works, it's been validated, but it works especially well for PKP-2 patients. Some genes warrant really their own risk stratification. We should know about these genes and know what are these factors. There is always room for improvement. All these models will improve with time. And I think a challenge remains a gene-negative patient, which may be a more heterogeneous group. We love to have positive genetic testing. So thank you very much for your attention. Thank you. A really strong support of the genotype analysis and the clinical perspective. Thank you. Please, for the audience. Thanks, Julie. Is there any room for risk calculator for gene-elusive ACM? So it's a good question. So this validation study showed, well, one of the validation studies showed that it didn't work very well. And I think my take for now, but it didn't work very well in the gene-elusive patient. However, it worked. The calculator worked in the first derivation study and the first validation study. My take is I think that the gene-negative patient might be different in different cohorts, depending on where you kind of, so what type of centers is enrolling the patient. So I want to have a look at our data and see if we can find something there. But I think that the problem is heterogeneity between the different cohort to make that work in that group. Another variable is obviously if someone gets a diagnosis of ARVD and they dramatically reduce their exercise by 85%, their rhythmic risk goes down, whereas if people continue to exercise, it stays high. Have you been able to factor that into your analysis? So that's a super good question. So in fact, this has been looked at mainly by Cindy James with Lawrence Bozeman. And so what we see is exercise, yes, is a risk predictor of ventricular arrhythmia in ARVC, but the different risk factors that are in the prediction model include this effect of exercise. So if you exercise, you're going to be worse. You're going to have more PVCs, non-sustained VT, lower right ventricular ejection fraction. And so it's not an independent variable. It doesn't add up to the model when you add up exercise. And the other thing about what you mentioned, and I think it's important, if you stop doing exercise, so there is a change in what you're doing. So another nice study also from the Hopkins group with Rick Carrick showed the longitudinal effect of, well, the effect of time on the risk calculator. And the best way to use it is to update it at each visit with the new predictors. So if you do less exercise, you probably have less PVCs. And we see the global number of these PVCs decreasing with time, and the reduction of exercise, because we tell our patient to reduce exercise, probably comes into play in that effect. So by updating the model, we probably capture this effect of reduction of exercise. And what is the risk where a defibrillator is indicated? Oh, that's the tough question. So I never write this down. I always tell it verbally. So I think it's higher than for HCM, as I mentioned, because I'm not a guideline, or I don't want to put money on this. But in real life, I think the sweet threshold is probably between 10%, 15% of risk where we think about defibrillator and we discuss. No more questions? Then we have to close the session. Thank you very much, and have a good evening. Thank you.

phenotypic differences

genotypic arrhythmia

Arrhythmogenic Cardiomyopathy

ECG manifestations

genetic subtypes

cardiac MRI

exercise guidelines

risk stratification

genotype-specific features

personalized approach