[MUSIC] Today, I want to talk about the problem of idiopathic heart failure, and I will telegraph one of my main messages, is that I absolutely despise the word idiopathic. The reason I don't like it is that I think idiopathic is a word we use when we have no idea what's going on and we're not really interested in finding out what's going on. So people have said, when somebody calls a disease idiopathic, it's sort of idiotic on the part of the doctor and pathetic on the part of the patient and I sort of agree with that as well. So let me go through this case and the main message is don't say idiopathic. And as I'll discuss, there's this problem that is emerging as we get better and better at sequencing, we find things and we don't know what to do with them in the genome. And we'll come back to that in a bit. So the story I want to frame this a 32 year old man who comes in with a three week history of gradually increasing shortness of breath and ankle swelling. His cardiogram shows, as is typical in many of these cases, left bundle branch block and sinus rhythm otherwise. And an echocardiogram is done and it shows four chamber dilatation and thinning. I don't have the movie, but there's no regional wall motion abnormality. And as is sometimes the case in this particular patient, there's an obvious, very large clot at the left ventricular apex, you can see that indicated by the arrows. So these patients are susceptible to serious arrhythmias, they clearly present with heart failure as this patient did and they're susceptible to the thromboembolic complications through the mechanism that you can see on this slide. So when we see somebody who has what we call dilated cardiomyopathy, that's what this is. We divide them traditionally into ischemic and non-ischemic causes. Ischemic cardiomyopathy is probably the commonest cause of cardiomyopathy. It's in patients who generally are known to have advanced coronary disease or who have some condition, usually diabetes, that prevents them from having symptoms of coronary occlusions and they present after having had several infarcts. With ischemic cardiomyopathy, the hallmark of that is regional wall motion abnormalities, often in the territory of the major coronary arteries. This man is too young for that, has no history for that and he has non-ischemic cardiomyopathy and the common causes are listed here. The commonest cause in the United States is almost certainly burned out hypertension, which this man does not have. It's traditional to say when you don't know the mechanism to say that maybe this patient has had a recent viral infection and it's affected the heart, so viral myocarditis or post viral myocarditis. A number of other causes, excessive alcohol intake in some patients. HIV can do this, there's a syndrome of partum cardiomyopathy, which may have an overlap with the genetic causes that I'll talk about in a moment. A number of drugs and toxins can do this, notably certain anti cancer agents can result in dilated cardiomyopathy well, after therapy. There's this idiopathic term, which again, I told you I despise. And then usually on these lists at least until recently, there was this other category of familial. Often the bottom of the list, often in lighter type. Often in smaller type, just like here. So here’s this patient, and if you take a family history in a patient like this, it's not unusual and this happened in this patient, to find other people who have what appears to be symptoms that could be consistent with dilated cardiomyopathy. So in this particular case, there’s a great uncle who died suddenly at the age of 25, and there's a distant cousin, or I guess that's a cousin, who had a heart transplant. So those are the kinds of things that people with this particular disease end up having. It turns out that if you do linkage analysis in these families, like large families, like this one or you look at candidate genes, you can now identify dozens of genetic variance that track with dilated cardiomyopathy in families. So there are clear genetic causes for the familial forms, and there's a real strong sense that some of those overlap with the other forms. So perhaps, the alcoholic cardiomyopathy only manifests itself In a patient who has one of these mutations, the peripartum cardiomyopathy, the same idea. So on this slide, you can see in the middle is dilated cardiomyopathy. And there are many, many genes that are attached to dilated cardiomyopathy, but there are other cardiovascular diseases on here. Notably, hypertrophic cardiomyopathy, and you could see that there's a separate set of genes in which mutations cause hypertrophic cardiomyopathy, but some of them there's an overlap. So, some patients will have either dilated or hypertrophic cardiomyopathy. There's this other disease called arrhythmogenic ventricular dysplasia. And again, there's this overlap syndrome. There are diseases of ion channels, I've talked about some of those in the long QT syndrome, and one of those disease genes is the sodium channel gene, SCN5A. Mutations in SCN5A cause a diverse set of diseases, including in some unusual instances, dilated cardiomyopathy and then there's neuromuscular diseases like muscular dystrophy. One of the major causes of death in patients with muscular dystrophy is dilated cardiomyopathy. So you can see there is this overlapped set of syndromes, so some of the genes will produced dilated cardiomyopathy alone, others will produce a variety of phenotypes and we don't understand why some people have one phenotype in and not the other. But it's pretty clear that there's lots and lots of genetic causes across multiple families of genes. So the families of genes are shown here. The three cells in the middle are three generic cardiomyocytes connected end to end as they typically are. And in panel A, you can see the T tubular structure and genes that control calcium cycling, for instance, IRYRY02. Mutations in those will cause arrhythmias and sometimes cardiomyopathy, genes that contribute to the structure of the sarcomere in panels B and C. The contractile apparatus in these cells, mutations in them will cause hypertrophic cardiomyopathy or dilated cardiomyopathy. In panel D at the upper right, you can see blown up the cell to cell connection and it turns out that this disease called arrhythmogenic right ventricular cardiomyopathy and more commonly arrhythmogenic cardiomyopathy arises from problems in the proteins that connect two cells to each other. And there are sort of four or five of those where mutations can cause the disease, then there is a class of genes that are expressed at the nuclear envelope protein, notably a gene called lamin A/C. And one of the manifestations of mutations in lamin A/C is cardiomyopathy, there are a number of others. And then finally, there's mutations in genes that connect the cardiomyocyte to the extracellular matrix and those are also causes of cardiomyopathy. So many, many different causes through many molecular mechanisms. And the hope, of course is by identifying the molecular mechanisms, we'll be on the way to specific drug therapies for each of these subtypes. So, it becomes interesting and important to start to classify these by mutation type and mutation function. So here's the problem, that's the framework, but I wanted to use the problem of dilated cardiomyopathy to illustrate an interesting emerging problem in our field. There are many large cohorts of patients that have been followed for many, many years in the United States. Two of them are shown on this slide. The Framingham Heart Study, which is centered in Framingham, Massachusetts just outside Boston, and the Jackson Heart Study Center in Jackson, Mississippi. Framingham's a cohort that started out with about 6,000 Caucasians in Framingham and is now into its third generation. So about 5 or 6,000 people per generation, it's been going on for decades. Jackson Heart Study's a little younger, focused primarily on African Americans. So here's an interesting experiment that was done in these two resources. 3,600 subjects, African American and Caucasian, had 8 disease genes for hypertrophic cardiomyopathy or dilated cardiomyopathy sequenced. Using next generation sequencing that I've talked about before and 11% of them have rare non-synonymous variance in those genes. So, one in 10 people caries a variant in one of those eight genes. Now the problem is we all recognize that some of those variants are benign, and there are computer based methods and other methods to try to tease out whether they're benign or malignant. So when you use the strictest of criteria, 0.6% are likely pathogenic. So that reduces the number of people who may carry a real pathogenic diseased gene to a very, very small number. So here's the kicker. If you look at 22 subjects who are discovered incidentally in these communities to carry rare, probably pathogenic mutations in these eight disease genes, four of them have evidence of cardiomyopathy. That means that there are 18 out of the 22 who genetic testing has identified as having a disease. They don't have the clinical disease, but they have all the genetics around the disease, including a high likelihood of pathogenic mutation. And yet they don't have the disease. And what's more interesting is that if you look at those patients over long periods of time, because remember these are community studies. So you can follow them for many years. Those patients that carry those variance appear to have an increase cardiovascular risk with an odds ratio that's modest around two, but it still suggests this. So maybe they don't have dilated cardiomyopathy in its full blown form, but maybe they have a little bit of that and a little bit of that is enough to impair survival. So that's one sort of question that we're going to be left with as we start to sequence lots and lots of people and here's another part. So I've showed you this slide before and if you'll look in the corner, the commonest genes that cause dilated cardiomyopathy or the commonest genes in which mutations cause dilated cardiomyopathy are TNNT2, MYH7, MYH6 and then this gene called TTN. And notice TTN causes 25% of all cases of dilated cardiomyopathy. Well, where does that evidence come from? So TTN stands for Titin, T-I-T-I-N. I often think it should be called Titan, T-I-T-A-N, because it is the largest gene in the genome. It has 325 exons and encodes an extraordinarily large protein. And the function of that protein is shown here. It's part of the apparatus that holds together the sarcomere. You see those two spring like structures that are labeled Titins. So that's what Titin does, it sort of holds the whole mess together. And because it is so large, it became pretty daunting to think about how to screen that gene for variants that might be associated with dilated cardiomyopathy. It's just technically a real challenge to do that. But with the advent of next generation sequencing, it became possible to ask the question, could variance in Titin be associated with dilated cardiomyopathy? Notice I didn't say cause, I said be associated with. So here are the data. When you look at a group of patients with dilated cardiomyopathy, about 25% of them have variants in Titin that you would predict will disrupt function. So these aren't just missense variants that change a single amino acid. These are variants that disrupt splicing, or these are variants that result in premature truncation of the protein. [COUGH] And when you look at a control group of patients with hypertrophic cardiomyopathy, a different disease we think, it's a little bit of overlap, those kinds of variants are very rare. What's interesting Is if you look at just a group of normal controls, people off the street. There is a frequency of about 3% of these predicted to be highly deleterious mutations. Now this doesn't count the missense mutations, the snip in Titin, because Titin is so big and we carry snips every 1,000 or 2,000 nucleotides in our DNA. Every single person on average has a missense mutation in Titin. So obviously, most of those don't do anything to function. But the real question is not whether Titin causes dilated cardiomyopathy, I think these data argue pretty strongly that having a mutation, a rare variant in Titin that's predicted to disrupt function, really carries a very high risk of being associated with dilated cardiomyopathy. But 3% of the population out there, walking around on the street without any evidence of the disease, also have those kinds of variance. So let me just close by saying, first of all, my categorization has changed and I put familial at the top of the list, not at the bottom. I put it in bigger type not littler type and I put it in bold, because I think it's one of the leading causes of non-ischemic cardiomyopathy. When we see a patient who is young, who presents like this, who hasn't had a real recent galloping viral illness, who doesn't have HIV, who doesn't have anything else and often they don't. They have familial or genetic dilated cardiomyopathy, unless we can find some other cause. So that's the first lesson and the second lesson I think is that we all have rare variants. And those rare variants sometimes would be predicted by every computer algorithm you can could up with to actually disrupt function of that gene. And why it is that we actually have lots and lots of variants and in fact don't have diseases is really, really interesting question. But as we come to an age where people are going to have genome sequenced at birth or when they start a job or whatever, we're going to discover that people have lots of variants. And the challenge to the field will be to figure out which of those variants absolutely need to be followed up. Which of those variants might be important and which of the variants, and we hope that most of them, can be ignored. >> [APPLAUSE]