Identifying the Cause of Cardiomyopathy Using CMR (Chetan Shenoy, MBBS, MS)

0:00 Before I introduce, today's speaker, just a few mandatory housekeeping items. for the audience here, we are broadcasting live via, YouTube and livestream. So, please, if you have questions, use a microphone. the recording will be available for viewing afterward, as well, as viewing in real time. and for those of you out there, please consider, subscribing to our YouTube channel to get notification when new videos are added. for viewers that are out there, if you have any questions that you want to submit, please text the the keyword DeBakey to 37607.

0:39 and you can submit questions via livestream at the, link livestream.com/hmh-edu. some upcoming conferences, March 25th and 26th will be the Robotic Vascular Surgery Summit. April 27th will be the Heart of a Woman Summit. And May the 4th will be the CV Prevention Conference. So, please, look forward to any of those conferences.

1:09 So, really, it gives me a great pleasure today to introduce, the speaker. So, Dr. Chetan Shenoy is associate professor of medicine in the Cardiovascular Division at the University of Minnesota. He is a NIH-funded physician-scientist. His, expertise both for clinical standpoint and research really evolves around cardiac MRI. So, Chetan, went to med school in India. He completed his residency at the Guthrie Clinic in Pennsylvania.

1:43 Did cardiology fellowship at Tufts University. and then did and also did CMR fellowship at Duke. And that's how I know know Chetan from his, time that he spent at Duke. So, couple of things about Chetan. So, I've known him for about 15 years. even when he was a fellow, when he was getting started, I knew this is somebody who's going to go far. And I think, it speaks to the not just the intellect, but the intellectual curiosity that he has. And also the rigor that he has for doing scientific research projects. And so, I think you'll if you look through the literature, in the cardiac MRI space, whether it comes to cardiomyopathies, to cardiac sarcoid, to transplantation, you see Chetan's name all over the literature. And I think the the really the unique thing about it is he comes up with unique ways to address questions.

2:34 So, I think, you know, what I really enjoy is the way that he puts studies together to really answer clinical questions, but he puts it together in a very unique and thoughtful manner. So, Chetan, I'm really, excited to have you here to talk to us about identifying causes of cardiomyopathy using cardiac MRI. And I'm sure we'll have lots of questions for you. So, welcome, Chetan. Well, thank you, Deepan. That was very kind of you. good morning, everyone.

3:03 I'm delighted to be here. So, these are my disclosures. So, I'll start with the current classification of heart failure. And so, this is something we all know. And this is a picture from the universal definition and classification of heart failure. So, we classify heart failure, based on ejection fraction. And as you can see, preserved ejection fraction, improved ejection fraction, mildly reduced, or reduced. So, how well does this work? So, I'll start with this, opinion piece from two leaders in the heart failure field, Mark Constan and Frank Aboud, where they talked about ejection fraction and they titled their opinion as misunderstood and overrated. And I'll actually read the text because it summarizes their, entire, viewpoint very nicely. This abstract, "The long-standing emphasis on ejection fraction is misguided. EF is erroneously assumed to be a measure of myocardial contractility.

4:08 Of greater concern is the widespread classification of patients with heart failure based on whether EF is preserved or reduced. In fact, EF does not provide any specific information on causation or underlying mechanisms. We believe that a revision or abandonment of the nomenclature is warranted. And categorization of patients with heart failure should strongly emphasize the underlying pathophysiology." There's more papers. So, this one is also an opinion piece with 42 heart failure experts. Again, bringing up somewhat similar points. LVEF has imprecise physiologic implications, substantial variability between measurement modalities.

4:52 The cutoffs used are arbitrary. And EF, as we know, changes during the heart failure course. And RV dysfunction can affect the LVEF. Now, let's look at how we treat, heart failure and how the categories help with the treatment. So, if you look at heart failure with reduced ejection fraction, or mildly reduced, or improved ejection fraction, they all benefit from the same, GDMT, or guideline-directed medical therapy.

5:25 But when you look at the pathophysiology, or the cause, the treatment is very different. So, ischemic cardiomyopathy, we think about revascularizing them. Even in the group of inflammatory cardiomyopathy, the specific cause of the inflammation is important and it guides management. So, viral, eosinophilic, infectious, sarcoid, giant cell, these are some of the inflammatory cardiomyopathies. And as you can see, the treatment is very different for each of these. For viral, typically, we don't really need to address the cause. Eosinophilic, it depends on what's driving it.

6:02 Infectious, so, things like Chagas or Lyme's, we have to treat with antibiotics. Cardiac sarcoid, immunosuppression. Giant cell is mega doses of immunosuppression. Similarly, these other non-ischemic cardiomyopathies I've listed here, stress cardiomyopathy, genetic, cardiac amyloid, toxic, iron overload, tachycardiomyopathy, each has a very different and cause-specific treatment. And in fact, there is a whole, scientific statement from the AHA on the diagnostic and treatment strategies for these specific non-ischemic cardiomyopathies.

6:41 How about outcomes? So, if you look at the outcomes of heart failure by EF categories, and this is a paper from the Get With The Guidelines registry published in 2017 with the largest numbers comparing HFpEF to HFrEF to heart failure with borderline EF. And as you can see in this Kaplan-Meier curves, the curves literally superimpose. And there is no real difference in the mortality based on these EF categories.

7:12 But when you look at the, outcomes by cause, and, most of us in the room will be familiar with this paper by Felker, from, NEJM now 24 years ago, the outcomes are very different based on, the type or the cause of cardiomyopathy. Now, you could be wondering, is this different simply because some of these cardiomyopathies have low EF, some of them have high EF? So, could the difference be still driven by the ejection fraction?

7:44 And so, this question was, answered by, this paper, which was just published, few months ago, in the journal Heart, where they did a sophisticated mediation analysis. And so, on the upper right is this, DAG, or directed acyclic graph, which shows the association, where the etiology has a direct effect on the mortality. And there is an indirect effect through LVEF. So, we all know that, you know, things like hypertrophic cardiomyopathy have a normal or close to normal EF and they do well, as opposed to some of the other, causes of HFrEF.

8:27 So, this is trying to tease out the contribution of the EF to the mortality, as opposed to the direct effect of the mortality. And, in the table on the, bottom on the right, you can see the contribution of the EF. And for all the causes, except for cardiac amyloid, the contribution of the EF was less than 20%. So, it was anywhere from 2 to 19%. And the conclusion is the effect of the cause on the mortality is largely driven by the cause itself, rather than by the EF.

9:06 So, how often do we look for a cause in new-onset heart failure? So, these are a bunch of papers which have looked into this question over the last decade, looking at how often do we look for a cause, in patients who are presenting or admitted with new-onset heart failure. And when we, look at the numbers, it's a shocking 20 to 40%. So, in more than half of these patients across the country and across the world, we don't really investigate for why they have heart failure. They are treated, and that's a whole, different topic of how well we treat them. But, there is no investigation. And all these papers looked at a period of time before their hospitalization and a period of time after their hospitalization, typically 3 to 6 months to see what testing was done to investigate why they had heart failure.

10:05 And so, how do we look for a cause? So, among the patients, the 20 to 40% of patients where they looked for cause, how do we look for it? And so, this is the largest study and the most contemporary study from the Optum database, which is a large American database of patients with both commercial insurance and Medicare. And this spanned from 2003 to 2020. And this was over half a million patients with new onset heart failure who was who were hospitalized. And in this study, 35% had any testing.

10:43 But this is a breakdown of what testing they had, and you can see the most common testing was invasive coronary angiogram, followed by stress ECG, nuclear imaging, stress echocardiogram. And I have highlighted MRI, which was done in 2.5% of these patients. And then CTA was done even lower in less than 2% of patients. So, we don't really use a whole lot of cardiac MRI and I'll explain why this should change.

11:19 So, before I explain that, does it really matter? Does it matter that we are not looking for the cause or how we are looking for the cause? And the way to understand this is by looking at cases of patients who had heart transplantation and had explant pathology evaluation and comparing the diagnosis on the pathology, because that is the reference standard or the gold standard, and comparing those diagnoses to the clinical diagnosis prior to the transplant. And again, there's a whole bunch of studies, at least 10 of these studies that I have listed here from transplant centers across the world. Two of them here in Texas, UT Southwestern and Baylor. And among 2,500 patients, approximately 10% of the patients had the entirely wrong cause of cardiomyopathy identified.

12:15 And this is the largest paper, which is from Baylor. And they looked at all their transplants from 1993 to 2020, And in this group, 9% of the diagnosis, the clinical diagnosis for the cause of the cardiomyopathy, were wrong. And the most common diagnoses that were missed were cardiac sarcoidosis, where 75% Think about that. 75% of the diagnoses were not recognized before the transplantation. ARVC and HCM. And these are, you will see, some of the diagnoses that we can reliably make using cardiac MRI.

13:00 Now, I would argue that this is simply the tip of the iceberg, because these are transplant patients who are very deeply phenotyped. So, virtually every transplant patient has a coronary angiogram, and we do a lot of testing in these patients. And in the real world of people with heart failure who are not so closely tested or phenotyped, the problem, I would argue, is much worse.

13:30 So, the question is, why are we wrong sometimes? So, the first reason, as I showed here, the testing is directed at coronary artery disease. And this approach of looking for coronary artery disease to explain the cause of the cardiomyopathy is flawed. Why is it flawed? It's because the presence of coronary artery disease does not always indicate that this is an ischemic cardiomyopathy. And conversely, the absence of coronary artery disease does not always indicate that the patient has a non-ischemic cardiomyopathy.

14:09 This has been shown for at least 15 years using cardiac MRI data. This is paper dating back to 2011 from the London Royal Brompton Group, where they looked at 100 patients and described both these instances. On the top is a patient with dilated cardiomyopathy or a non-ischemic cardiomyopathy who has bystander coronary artery disease. And on the bottom is a patient who has no coronary artery disease or unobstructed coronary arteries. And on cardiac MRI, they have ischemic cardiomyopathy or ischemic heart failure.

14:51 So, we looked at this in larger detail in our database, and we looked at over 3,000 patients who had documented coronary artery disease and also had CMR for clinical indications. And we found that 9% of these had non-ischemic cardiomyopathy and 7.7% had evidence of both ischemic and non-ischemic or dual cardiomyopathy.

15:22 And we also looked at their long-term outcomes. And these patients who had either non-ischemic or dual cardiomyopathy had a worse outcome in terms of death or hospitalization for heart failure compared to the patients who truly had ischemic cardiomyopathy. Now, the converse situation of ischemic cardiomyopathy in patients without CAD. So, there's no large-scale data on this, and we hope to investigate this. But we know that we know or myocardial infarction with normal coronary arteries. So, this is the acute setting.

16:01 It occurs in around 6% of all acute MIs who undergo coronary angiography. And this these patients will also be seen in the chronic setting as patients with ischemic cardiomyopathy who do not really have coronary artery disease. So, what about the diagnostic accuracy of CMR? So, this is a paper that did a meta-analysis of 12 studies, 1,217 patients.

16:32 And they defined the sensitivity of CMR with LGE as 91% and the specificity was 90% and this was to detect ischemic cardiomyopathy. And they defined it as LVEF less than 50% and greater than 75% stenosis of epicardial coronary arteries. So, if you think about what I just said, does this really make sense? So, are they really looking for ischemic cardiomyopathy or are they looking at CAD?

17:06 So, this is actually looking at CAD. And so, the sensitivity and the specificity is for CMR to detect more than 75% stenosis rather than truly ischemic cardiomyopathy. So, this whole meta-analysis and many of the studies included in here are flawed, and this is not really looking at ischemic cardiomyopathy. In fact, the reference standard should be CMR rather than coronary angiography.

17:42 So, what's the second reason? So, the second reason is not only is the coronary approach flawed, but the coronary approach is also incomplete. And what do I mean by that? So, in the absence of coronary artery disease, often patients are labeled with non-ischemic cardiomyopathy. But it's not really a diagnosis. It simply tells us that the patient does not have ischemic cardiomyopathy. It doesn't really tell us what the patient has. And so, I would argue for approach where we are looking at the myocardium. So, this is William Sutton. And you may have heard about this story. This is a notorious bank robber in the 1950s who would rob a lot of banks. And when asked why he robbed banks, his answer was because that's where the money is.

18:34 So, if we are looking to understand the cause of a myocardial problem or cardiomyopathy, we should be looking at the myocardium, not the coronary arteries. And so, CMR to look for the myocardium or look at the myocardium is the reference standard for function and morphology. And this cartoon you have you may have seen before in Deepen stocks and many other talks. It illustrates the different pathologies and the different patterns of LGE, which we can use to identify the cause of the cardiomyopathy.

19:12 And CMR offers a virtual gross pathology view of the entire heart. And many, if not most, cardiomyopathies have unique features, and these can be used to identify the cause and to distinguish between the different causes. And I'll show a few examples, and these are pathology images and CMR images. These are not the same patients. All the pathology images are from Bill Roberts from Baylor, his papers. And this is an example of ischemic cardiomyopathy top, and you can see how you can see the same scarring or injury or damage on CMR using the late gadolinium enhancement pattern. In both these examples, there is a inferior myocardial infarction, which is a chronic MI.

20:13 This is examples of hypertrophic cardiomyopathy. Again, not the same patients. The top ones are from Bill Roberts' papers. The bottom ones are CMRs from our institution. But, you can see the appearance, and in some ways, it's even better on CMR because the LGE pattern is more of a black and white. But, you can see the mid-myocardial septal fibrosis in the pathology images, which is also seen on the LGE images.

20:46 Similarly, cardiac amyloid. the appearance of the gross pathology image is sort of a pale appearance, and there is significant thickening. And on CMR, the LGE appearance is very characteristic and pathognomonic. This is genetic cardiomyopathy or a dilated cardiomyopathy with a mid-wall scar. on the top again, you can see a faint septal mid-wall scar, which is seen also on the CMR images at the bottom.

21:24 So, our contributions to some of these. So, this is we looked at anthracycline cardiomyopathy. so, the prior literature talked about different patterns of LGE. Some papers mentioned no LGE. Others mentioned an inferolateral LGE. so, we looked at a large cohort of 300 patients who were treated with anthracyclines or trastuzumab. And we actually found 10% of them had LGE, but they were in different patterns, different locations. And when we looked at their clinical history, they were all explained by other cardiomyopathies and mainly coronary artery disease or ischemic cardiomyopathy. And so, our conclusion was that anthracycline cardiomyopathy does not have any LGE, and this feature can really help distinguish it from other cardiomyopathies when you have patients with suspected cardiotoxicity.

22:21 And here's an example. again, the pathology images from Bill Roberts' paper, and there is no scar, no grossly visible scar on the pathology image. And this is a patient with anthracycline cardiotoxicity from our institution who was transplanted, and there was no gross scar on the pathology specimen or the CMR. So, here's two examples, and this is again to illustrate the point that we can really use the CMR features to get at the true cause of cardiomyopathies.

23:03 So, the first one is a 54-year-old man with a family history of sudden cardiac death in his father, who presented with sustained VT and was found to have a EF of 30%. And the second one is a 51-year-old with biopsy-proven pulmonary sarcoidosis who presented with sustained VT and had a EF of 20%. And so, these are the CMR images. So, cine images four-chamber cine view and single-shot axis LGE image for both these patients.

23:38 And so, what do we see here? So, the first patient has somewhat proportionate LV dysfunction. So, the EF is 30%. He has pretty extensive LGE. the lateral wall is moving better than the septum. The LGE is septal predominant. so, more involvement of the septum rather than the lateral wall. The RV aspect of the septum is involved, and there is also involvement of the right ventricular free wall here with some sparing of the inferior portion of the free wall. So, it's not involvement of the entire RV free wall, but part of the RV free wall.

24:22 On the right, we see somewhat disproportionate LV dysfunction. This is a dilated LV with a EF of 20% way beyond the extent of LGE. The LGE is more on the lateral wall, so lateral predominant, and there is no RV aspect of the septum LGE. So, based on the CMR findings, we diagnosed these patients. So, the first one we diagnosed as cardiac sarcoidosis, and the second one we diagnosed as genetic dilated cardiomyopathy. So, somewhat goes against the clinical history, but both these cases were confirmed.

25:04 our diagnosis were confirmed. The first one was confirmed on explant heart pathology. the patient underwent a transplant within a year, and the second one had genetic testing, and it showed a pathogenic filamin C variant. So, there are many features to distinguish these two commonly confused pathologies. And I won't go into detail on into each one of these, but simply looking at what we should be looking at, looking at proportionality of LV dysfunction, presence or absence of fat, how the hearts look like when there is a edema or acute inflammation, the presence of microvascular obstruction, the uniformity of the LGE, the location within the LV, is it basal, mid, or apical, the location between the segments and within the segments, and whether it's more on the septum or the lateral wall, and the presence of RV LGE, and whether it's focal or it involves the entire RV free wall.

26:18 Now, for the CMR physicians in the audience, I don't recommend using this as a checklist when you're reading CMRs, but what I would suggest is actually looking at some of these patients with a definite diagnosis. So, definite cardiac sarcoidosis or genetically proven genetic cardiomyopathy, and paying attention to some of these details, and you will learn and you will notice these differences, and you will recognize that these are actually quite distinct cardiomyopathies.

26:52 And one other way to look at this, especially for cardiac sarcoidosis, since we don't really have a lot of definite cases where we have proven it on endomyocardial biopsy, is by looking at pathology images. And so, that's what we did in this paper. We examined patients with explanted hearts after they had a heart transplant or after they died suddenly, and we looked at the patterns of cardiac damage and involvement, and we did a sort of a meta-analysis, and we identified features which are very common and seen in more than 90% of these patients. And then we also saw features which were previously described as possible features of cardiac sarcoidosis, which were almost never seen in these patients.

27:46 And so, this is what we described as the pattern of LGE in patients with cardiac sarcoidosis, although this was not really looking at CMRs. And then we proceeded to sort of validate this in patients with CMR, but this is not really a true validation because we don't really have biopsy proof in all these cases. So, instead of looking at the diagnostic accuracy, we looked at outcomes. So, which of these patients behaved like they had cardiac sarcoidosis? And the patients who had these common features, which we call pathology frequent LGE features, behaved like cardiac sarcoidosis and had a really high rate of arrhythmias, as opposed to the other patients, even despite an abnormal EF or despite LGE which did not really this pathology frequent pattern.

28:53 They really did well and did not behave as if they had cardiac sarcoidosis. So, this is indirectly showing that these are the features or LGE features of cardiac sarcoidosis, and they can be used to diagnose and treat these patients, and also make decisions regarding ICDs. So, this is one of the disease entity that we looked at and described a unique LGE pattern. So, systemic sclerosis, and there is some overlap with systemic lupus erythematosus.

29:31 we noted that some of these patients had a unique subendocardial involvement, but it was virtually always lateral wall predominant, and sometimes it extended to the septum, but not always. And interestingly, all the papillary muscles were also involved and enhanced in these cases, which was kind of unusual. And this is sometimes labeled as myocarditis or inflammatory cardiomyopathy, but this is distinct from coronary artery disease or ischemic cardiomyopathy, which is also very common in patients with these connective tissue diseases.

30:12 On the right is case report that I found which also described the same lateral wall predominant subendocardial fibrosis in these patients. So, the next question is how often can we identify a cause with CMR? So, this is something that comes up when I give this talk. I would say most of the time, but there are a few caveats and requirements. So, we need good quality imaging. So, not always possible or not always available.

30:50 There is also an element of clinical correlation. It's a It's a joke that we talk about clinical correlation needed, but truly a lot of these diagnosis depend on what the other non-CMR clinical data are. And so, we have to interpret this in the context of the clinical data and non-CMR imaging data. And that's true of many of these non-ischemic cardiomyopathies such as tachyarrhythmia for instance, tacky cardiomyopathy, or stress cardiomyopathy.

31:24 So, the clinical history really helps. And then, this is also important when we suspect that the patient may have say an inflammatory or genetic cardiomyopathy, there should be directed workup to clarify this and to confirm this. And sometimes this may be endomyocardial biopsy or genetic testing or some other testing that will get us the specific diagnosis. So, in our circulation paper, there were about 100 cases of the 512 cases with non-ischemic or dual cardiomyopathy. And this was a retrospective study where we looked at all the clinical data and the CMR data, and we really couldn't pinpoint a cause.

32:13 So, this would be one in five. But remember that this is among 3,000 patients. So, this does not include the patients where we were able to diagnose ischemic cardiomyopathy or no cardiomyopathy. So, 100 in 3,000 would be less than 5% where we couldn't really pinpoint the cause of the cardiomyopathy. So, a few nuances. So, how often does ischemic cardiomyopathy occur with no LGE? So, this is something that often comes up, especially when CMR physicians talk about using CMR.

32:52 On pathology studies, ischemic cardiomyopathy almost always has gross cardiac damage in the form of scar or necrosis. And this is true regardless of whether they have a clinical history of MI or not. And so, hibernating myocardium without any LGE is extremely rare. And we had the opportunity to look at this in our study. And so, among patients with ischemic cardiomyopathy and EF of less than 50%, the prevalence of no LGE was anywhere from 1.1 to 2.3% and this depends on the definition of severe CAD, which is severe enough to cause hibernating myocardium. Now, we don't really know how severe it should be. And I think it would depend on the patient, and there is no single threshold. It would also depend on microvascular dysfunction and many other aspects. And so, it's somewhere between 1.1 to 2.3% the occurrence of ischemic cardiomyopathy without any LGE. So, this means that if you see a patient with LV dysfunction and they have no LGE, then there is a 98 to 99% negative predictive value for diagnosing ischemic cardiomyopathy.

34:23 The next nuances do we need to do stress perfusion? And this is also question that often comes up. Can we just do CMR with contrast or do we need to do a stress perfusion to distinguish between ischemic and non-ischemic cardiomyopathies? And this is a paper that looked at the specific question. They looked at 100 patients, and they found that stress perfusion CMR changed the diagnosis only two two of 100 patients. So, in some ways this matches our finding that there's only 1 to 2% rate of LGE no LGE and ischemic cardiomyopathy.

35:08 And so, their conclusion was that stress perfusion provides minimal incremental diagnostic value over LGE and cine for identifying the cause of the heart failure in patients with severe LV systolic dysfunction. But in these 1 to 2% of cases, these are people with hibernating myocardium who would really benefit from coronary revascularization such as CABG because these are patients who either have multivessel CAD or left main disease. And so, stress perfusion may really help, but in a small number of people. So, should we do it in everyone?

35:51 I don't have an answer. It depends. And sometimes we also look at the pretest probability of whether a patient may have multivessel or left main disease. For instance, if you have chest CT showing no coronary artery calcium, then we can use the power of zero, and we can say the likelihood of this patient having multivessel CAD or left main CAD is virtually zero. So, we can skip the stress perfusion part, and this patient may still have ischemic cardiomyopathy.

36:26 So, we do need a CMR, and then if it is non-ischemic to identify the type of the non-ischemic cardiomyopathy. More nuances. So, this is a nuance in the interpretation. So, the interpretation should be holistic. So, we should look at the whole heart and look at all the images. And so, one part of ischemic LGE may appear to be non-ischemic. And so, in this case, you can see the first image showing what appears to be subepicardial or on the RV subendocardium.

37:04 LGE. But if you look at the entire scan, this appears to This is very clearly large LAD MI. So, the basal most portion of the ischemic LGE may not be subendocardial, but this is still only ischemic cardiomyopathy. Similarly, midwall LGE. This is often associated with non-ischemic cardiomyopathy, but it is actually relatively non-specific, and it may occur from long-standing remodeling in ischemic cardiomyopathy.

37:41 And so, this is a case where there is some septal midwall LGE and also some lateral midwall LGE in the base. And when you look at the entire scan, this is a large LAD MI. And there have been papers that have looked at this and described this. And the most recent paper here in European Journal of Heart Failure looked at bunch of these patients and did genetic testing to see whether this could be dual cardiomyopathy.

38:12 And it was not it did not have a genetic basis. And so, they concluded that this is simply a part of the remodeling with ischemic cardiomyopathy. So, this is not a dual cardiomyopathy. This is simply a ischemic cardiomyopathy. And the same with insertion site LGE. This is non-specific, but it is also non-specific between ischemic and non-ischemic cardiomyopathies and simply a feature of badly remodeled hearts and this does does not really indicate the cause of any specific cardiomyopathy.

38:55 The next nuances do we really need T1 T2 mapping to identify the cause of the cardiomyopathy? And so, just as a quick primer, abnormal T1 indicates diffuse fibrosis, edema, or infiltration. And infiltration examples include sphingolipids as in Fabry cardiomyopathy, iron, cardiac amyloid. And low T1 can help identify Fabry cardiomyopathy or iron overload. And this can happen in the absence of LGE.

39:28 So, there is a small proportion of cases where this may help. High TI or T1 can help identify early cardiac amyloidosis in the absence of LGE. but when there is no LGE, the elevations in T1 are really mild and often non-specific. These are older patients who have hypertension, diabetes, and many other abnormalities.

39:59 And when you look at the ECV, which is a correlate of the T1, it's only about 32 or 33 or 34. And so it's not really at the amyloid levels of 50 or so. So even though this is abnormal, you're not really confident about calling it as cardiac amyloidosis. And similarly, high T2 indicates edema, but it does not really tell us why the patient has edema or inflammation, so it is non-specific.

40:34 So overall, we do this routinely as part of our protocol, but it has relatively low incremental value for identifying the cause of cardiomyopathy above and beyond LGE. So which patients should get a CMR? So obviously, ideally everyone. That would be the clear answer, and it's not really such a terrible answer because if you look at the 2023 ESC cardiomyopathy guidelines, CMR is recommended in patients with cardiomyopathy at initial evaluation.

41:11 So they recommend, and this is a class 1B recommendation. But practically speaking, when you have limited resources in most parts of this country, I would argue that we should use this in every patient other than those who have clear-cut ischemic cardiomyopathy. And what do I mean by that? So the patients with ischemic cardiomyopathy who may have echocardiographic features suggestive of non-ischemic cardiomyopathy, perhaps they have asymmetric left ventricular hypertrophy or symmetric left ventricular hypertrophy, or maybe they have disproportionate LV dysfunction. So these are the cases where even with ischemic cardiomyopathy, they would benefit from a CMR. And then all non-ischemic cardiomyopathies, and again, this is because non-ischemic cardiomyopathy is not a diagnosis, so we need to know the true diagnosis and the cause of the cardiomyopathy. And so they would all benefit from a CMR at least at the initial evaluation.

42:19 And there is a randomized clinical trial that is just starting up, Cross-HF, where they are randomizing three arm patients to three arms. This is in the UK, the University of Leeds. Peter Swoboda is the PI. This was just funded, and they're setting it up. Thousand patients each will be randomized to cardiac CT, stress CMR, and invasive coronary angiogram to see how these at the time of initial evaluation of new-onset heart failure will help. And they're looking at outcomes with a minimum 12-month follow-up for death or hospitalization for either MI or heart failure.

43:07 So extending this, I would say that everything I've spoken about so far is about LV dysfunction and heart failure, but cardiomyopathy is not really limited to LV dysfunction. So cardiomyopathy is heart muscle abnormality, and we can see this in settings other than heart failure. And so examples here are MINOCA, which is MI with non-obstructive coronary artery disease, where it is acute cardiomyopathy. And only a quarter of them truly have an MI when you look at look at them with the CMR, a quarter of them have stress cardiomyopathy, another quarter have other cardiomyopathies, and a quarter don't really have anything on their CMR, so you can clarify that. Young or middle-aged patients with AV block or heart block. So these are again patients who present with some cardiomyopathy that's causing the AV block. Young patients with atrial fibrillation. So this may be a harbinger of genetic cardiomyopathy, and this has been described in epidemiologic studies.

44:18 Patients with VT or VF without previously established substrate. So you're not sure why they're having a VT or VF. So CMR can really help identify the cause of their arrhythmia. And so in EP, we often talk about the absence of structural heart disease when we're talking about therapies. And so I would argue that this phrase, which typically means a normal EF or a normal echo, should be defined by CMR because many pathologies can have normal EF and still a substrate that can be identified on CMR. And for each of these scenarios, the story is very similar to the story I presented with heart failure patients. And I will give an example just to illustrate this point. This is the example of young or middle-aged patients with AV block. So a Danish study of a thousand patients looked at AV block in patients less than 50 years who got their first pacemaker between '96 and 2015.

45:29 The cause of the AV block was known in only half the cases, 50% of patients. There was a similar study looking at US patients using the national inpatient sample. 2,900 patients aged less than 60 who were hospitalized, and the cause was only known in a third of patients. And then the Danish folks did a long-term follow-up, and they showed that the patients who had AV block of unknown cause had a 3.8 times higher hazard. So the hazard ratio was 3.8 for death, heart failure hospitalization, VT, or VF. So that's really bad.

46:11 And the US study looked at how often CMR was done, and this was done in only 0.2% of patients. So this is two in a thousand patients. And so how can CMR clarify? So we wrote this editorial, and this is the image that shows how the diagnosis matters and the CMR can help with this. So these are four examples from our institution of patients who came in with AV block and had CMR. First is cardiac sarcoidosis, second is LMNA or laminin cardiomyopathy, a genetic form of cardiomyopathy where they can present with AV block, ischemic cardiomyopathy, and Lyme carditis. So if you look at the bottom two rows, the medical treatment is somewhat different between these. So immunosuppression for sarcoid, genetic testing and family testing for LMNA, revascularization possibly, and secondary prevention of CAD for ischemic cardiomyopathy, and antibiotics for Lyme carditis. If you look at the device treatment, it's also very different. So the first two diagnoses, sarcoid and LMNA, both of these would benefit from an ICD even when they present with AV block because these are the two conditions where they are at risk for sudden death and ventricular tachycardia. So when they present with AV block, the guideline recommendation is to put an ICD. With ischemic cardiomyopathy, generally speaking, it's a permanent pacemaker. In rare cases, in RCA disease, maybe they can recover and don't need a pacemaker, but for the most part, they would benefit from a permanent pacemaker. And with Lyme carditis, they almost never really need a permanent device. They may need temporary pacing, but with antibiotics and time, they recover completely their conduction and don't really need a device. So you can see how the treatments are very different based on the cause, and CMR can help clarify. Now if you're thinking about how can we do a CMR in patients with heart block, especially complete heart block, who may need temporary pacing, that is true. A temporary pacing wire is not MRI compatible, but the paper that we wrote this editorial for actually described this way of doing CMRs using TPPM, which is temporary permanent pacemaker. This is something that the EP folks have been using for a long time in patients who have pocket infections and other reasons why they cannot have permanent device placed. So the lead is a permanent fixation lead, which is MRI compatible, and the device is put outside the body. It's just taped to the body, and you can place it far away from the heart to minimize the artifacts, and it can be reused in patients. And in this Finnish study, they did this in 18 patients and found cardiac sarcoidosis in about half of these patients, and it really changed the management because these patients needed immunosuppression and ICDs.

49:40 So, in conclusion, I would say that CMR can really help identify the cause of cardiomyopathy in most patients and in many different clinical scenarios. And knowing the cause really makes a difference in how we treat these patients and in the patients' outcomes. And in addition to what is the EF, I think we should also ask why more often and we owe it to our patients. Thank you.

50:17 All right, Chetan. That was again, as always, a really phenomenal and very insightful talk. So, I know there's going to be lots of questions, so I'm going to just head right to the questions, Arvin. Thanks. Thank you, Dipen. Thank you, Chetan. It's an amazing talk and for those of you might not know, Chetan and I went to med school, so he was always a professor from day zero, so his title doesn't matter. very very elegantly put as a cardiomyopathy cardiologist, I think the pearls that you gave are very important for clinicians and fellows to understand that we are thinking about heart failure probably 100 years older than what we need to think about.

50:56 my question I have a couple questions if I have the liberty to take. The The first one is as a you know, person who's drank the Kool-Aid of CMR, I think it's not an overstatement to say that somehow we have to make the guidelines almost mandate a CMR in the early stage of diagnosis. Access will always be an issue, but people drive 10 hours to get a VAT or a transplant. It's almost criminal that early on we shouldn't promote them to drive 10 hours to get a CMR so they can avoid a VAT or a transplant.

51:33 And of course, I'm sure we never did an RCT to do an echo in a heart failure patient. So, it just comes down to this whole evidence-based practice. So, I don't know how if you have any thoughts on how the imaging societies and and the heart failure societies together are working towards the implementation aspect of changing the paradigm that CMR is mandated and not like an option. Yeah, well, that's very great points and you know, I don't really have an clear answer. It's something that we struggle with every day. And I would say that it's not just the access issue. Access is a big issue, especially in the US.

52:18 CMR is not really done a whole lot and they say that in London CMR, they do more CMRs than all of the US. So, there is a lot of resources and access that we need to talk about, but the other aspect is also expertise. And so, you know, things that I presented require some expertise and you know, we can do that at expert centers, but not at every center. And so, we can talk about increasing CMR centers and doing more CMRs, but we also have to talk about increasing the expertise and educating the people so that we don't get reports with quote unquote non-ischemic cardiomyopathy or a long differential of multiple diagnosis, which doesn't really serve the patient and defeats the whole point of doing advanced imaging tests like CMR.

53:18 Yeah, quick I think you answered my second question actually on the pitfalls and the expertise. Yeah. But a quick follow-up, is the technology of the scanner going to ever be something that you can do a mobile CMR scanner? Or I know it always goes to the basement with the weight of magnets. Is that something that the field is looking into for a scanner that can go into a truck? Yeah, so there are efforts to do cheaper CMRs, low field CMRs, protocols where it's done faster in like 30 minutes or even less. And I'm working with collaborators in engineering and physics to look at how we can do this without any breath holds, accelerated CMR. So, there are ways to our efforts to speed up the process, which indirectly will allow more patients to have this done. But to answer your question specifically, cardiac MRI can be done on any CMR magnet. It needs some additional cardiac software package, but every MRI scanner out there that can do brain imaging can also do cardiac imaging if if they get the software. So, it's not a huge investment needed to do more CMR.

54:42 Thank you very much. That was That was very nice, actually. And compelling argument to use it. I could see multiple areas where you have issues of implementation. First, I'm interested in your data. First, that was a 10-year span. I'm just wondering about the last 2 years of the span of utilization of MRI, about 2% or so. Has that increased over time?

55:13 Or is it still incredibly low? I don't think there has been a lot of change. There's change is slow and the numbers. So, there are some data on the number of scanners, number of CMR readers, number of CMRs done, especially from Medicare over the last 10 to 15 years. There's been an uptick, but it's gradual and slow, not a dramatic rise in the number of CMRs.

55:44 along these lines, I'm just wondering, obviously we know that MRI units, although they're available for radiologists, they have not been as much used for cardiac. I mean, if I look at the Houston greater Houston area, obviously there are pockets. So, the question to you is besides availability, is I'm wondering of how much the physicians treating these patients feel compelled in a way or is it even on their radar screen?

56:16 So, I would you you are the leaders in the field. I would ask Heart Failure Society of America in addition to SCMR to do such, meaning send a survey to the physicians that do you think MRI is important? And if they don't feel it's important at this stage in in they're not going to push for it. And I really feel at this stage, you know, we're talking among people who know the field, is I think physicians out there are are not really pushing for it.

56:53 And I agree with Arvin that, you know, people can travel an hour to get a diagnosis that may change their therapy in so many ways. And and you know, these are usually not urgent situations. I mean, we're not necessarily all of them are in shock. So, I'm just wondering how much of the problem is each one of them. So, you can tackle it. Otherwise, we can keep talking about it and nothing happens. Yeah, I agree. So, the education has to be beyond, you know, us talking at the CMR meetings. So, this should be more looking at heart failure folks and EP folks who deal with most of these patients. And there are some efforts from the societies to do that.

57:40 There are joint sessions by SCMR at HFSA and, you know, many of the other meetings. And the other thing we as researchers, people like Dipen and me, can do is to put more data out there showing the benefits, large compelling arguments so some of this eventually gets into the guidelines like in the ESC guidelines. So, our recent paper looking at this non-ischemic cardiomyopathy, it's not entirely novel. It was described in 2011 by, you know, the Royal Brompton Group. We only did a large data set and looked at outcomes to make that argument compelling because, you know, despite the 2011 paper, you know, there's still a little understanding of this issue and as illustrated by the Jack Imaging meta-analysis.

58:33 And so, what we as researchers trying and we should do more of is to do such studies where we are demonstrating the power of CMR and try to get that into the guidelines, which the other clinicians will read and follow. Question in the back. Thank you very much. This one was very educational, especially as a heart failure surgeon. my question is related to recovery. And that's point, we know that level of fibrosis is is predictor of the recovery. And is there any focus on that point which hearts will recover, which hearts after recovery even will be sustainable as and also including that ischemic cardiomyopathy, the we got viable studies and they are always viable. But you know, how much is viable? It's like, you know, like you said, election, you know, like 30% Democrat, 70% it will win, but it's not doesn't make sense to now with the modalities just taking someone where a high risk coronary artery surgery and putting an LVAD in 6 months, in 1 year.

59:49 So, is there any focus to look at the myocardium of the that level of fibrosis that predict going to predict us to recover process? Yeah, so there's decent amount of data looking at simply the extent of LGE to predict recovery of LV function. So, LGE is you know, damaged myocardium or dead myocardium. And so, it cannot really recover. So, it makes sense that more the LGE, less the likelihood of recovery. And this has been demonstrated in various cardiomyopathies.

60:27 But, there is also the influence of the cause of the cardiomyopathy. So, not all cardiomyopathies recover the same way. and then even among non-ischemic cardiomyopathies, there's data that some recover better than others. So, in genetic dilated cardiomyopathies, folks with titin mutations, they tend to recover with GDMT at least for a good period of time as opposed to some other mutations.

60:58 So, it is a combination of the amount of LGE, amount of cardiac damage, and the cause of the cardiomyopathy. And with respect to your question about viability and recovery, so again the extent of LGE and the extent of viable myocardium predicts the amount of recovery. but, it's not just that. It also depends on the completeness of revascularization, the control of optimal medical therapy, how how well their diabetes is controlled, whether they're doing cardiac rehab. So, the milieu to help the LV recover, how well is it optimized? So, that also has a contribution to the extent of recovery, which is why there is a disparity in the recovery that we see even when we say it's all viable, some recover, some don't. And it's because it's not just the amount of myocardium that's viable, but also in other aspects such as the completeness of the revascularization.

62:09 So, Jason, thanks for a great talk. you know, you you never realize sometimes how good you have things till you miss till they're not there. and I can say, you know, being able to work with you for many years, you spoiled me with your expertise that you have. And it's really fun to start establishing some of those relationships here, you know, in Houston Methodist. but, so two things. One, how I I think Arvin, you know, talked on this a little bit. I think one of the things that I would ask as we are continuing to look at promoting the use of CMNR is also continuing to promote the importance of CMNR expert. And I think in that expertise because I think we've, you know, come across several of those instances where MRIs were done at centers and maybe smaller centers, and the interpretation was completely wrong and erroneous. And that led people down a very different pathway. And so, I think, you know, getting your thoughts on how we continue to make sure the importance of the subspecialized training in imaging is supported. And also, especially in the era of RV reimbursement and multimodality everything of making sure that depth is there. But, my second quick question and or just a thought about, I know this talk was a lot about initial CMR readings and stuff. Maybe give us a little bit of words on where you think like serial CMRs are helpful in which particular cardiomyopathies.

63:29 So, I'll answer the second question first. So, the serial CMRs, it depends. And I don't really feel as strongly that we need CMR to follow these patients once we know what we are dealing with and what we are treating. and we are looking essentially at recovery of function or you know, how their heart is remodeling. And for that, echocardiogram is just as good. And and the LGE that we see, typically, you know, it doesn't get better. It never gets better. And typically, it doesn't get worse or a whole lot worse unless it's a progressive cardiomyopathy.

64:07 So, I don't really recommend serial follow-ups or talk about in my reports when they should come back. some of them do for clinical reasons, but I don't think there's mandated follow-up for any type of cardiomyopathy. the first question of expertise, yes. So, that is a challenge. and there's always this tension between increasing CMR access and increasing, you know, getting CMR to more patients versus requiring the expertise, which requires training and time. And you know, it's not always easy. And so, on one hand, the society and us CMR folks want more people to do it and more patients to get it. But, on the other hand, as you mentioned, you know, improving it or increasing it rapidly means that we have to sometimes compromise on the quality. And that is not optimal. And so, again, there's no easy answers. But, we try with our training programs.

65:21 And here, Deepan has a large training program training fellows to go out there and do more. I think you know, education through conferences and societies and fellowships, that's the way to go. Awash. So, thank you, Jason. That was an excellent talk. I I mean, I learned a lot from your talk today. You showed an excellent correlation of how path correlates with CMR. My problem with CMR or the real issue that I have is that when I see patients who have sarcoidosis, their LGE is limited.

65:52 Yet, the extent of inflammation that I see on PET scan is is very disproportionately high. So, it's clearly there are things that the CMR cannot see with the naked eye. So, what is your perspective on that? And second, where do you think is the role of radiomics for for CMR? Because clearly, there are things that we cannot see with our naked eye. Yeah, so LGE versus FDG PET. So, in cardiac sarcoidosis, LGE shows cardiac damage. And so, that may be permanent damage and scar, but it may also include some edema intermixed with it. So, it's not just fibrosis as some people believe. So, it's a combination of necrotic myocardium and which is active inflammation and fibrosis.

66:42 And there may be other areas of the myocardium where there isn't damage yet, but there may be inflammation. And that's typically the area surrounding the areas of LGE. And so, that may be one reason why you see the extent of FDG uptake to be greater than the extent of LGE. And we also see FDG uptake in many other inflammatory conditions. So, it's not just cardiac sarcoidosis. And I believe we also see some degree of FDG uptake simply due to passive inflammation in inflammatory systemic conditions such as sarcoidosis. So, there are descriptions of FDG uptake in sarcoidosis, systemic sclerosis, COVID. And there's a parallel literature on T2 abnormalities in these cases without any LGE. And outcome data show that these patients do okay. They don't really have any cardiac abnormalities. And they don't progress to developing cardiac involvement in the form of LGE. So, if you see patients with isolated FDG uptake without any LGE, then it's highly unlikely that the patient has active cardiac sarcoidosis that needs to be treated even though they may have some inflammation in the heart.

68:22 So, Jason, again, I mean, just phenomenal talk. And And maybe I'm just going to kind of wrap it up with the fellows. I think, you know, correct if I'm wrong, I think there's two kind of interesting concepts which are a little bit foundation shifting that you put forward today. One is maybe we should move beyond ejection fraction as a way that we think about patients with heart failure and cardiomyopathy. And the second is that maybe we should think about ischemic cardiomyopathy less about the status of the coronaries and more about the status of the myocardium.

68:54 And that's a bit foundation shifting as well. And I think that, you know, what this speaks to and I've learned as as I've been doing this a little bit longer than you, is that, you know, things that you learn when you're in med school that are kind of foundational concepts or dogmas, you'll always have to be open to the possibility that things can shift over time. And I think that that's really an important lesson that you've taught us here today is that we have certain basic tenants that we started off with, but it's going to continue to evolve over time. And I want to really congratulate you for the great work that you've been doing. And thank you for joining us.

69:28 Thank you, Deep Fund.