Preview Mode Links will not work in preview mode

Getting Personal: Omics of the Heart


Jan 29, 2019

Jane Ferguson:  Hello, everyone. Welcome to Episode 23 of Getting Personal, Omics of the Heart, the podcast from Circulation: Genomic and Precision Medicine. It's December 2018. I'm Jane Ferguson. So let's get started.

                             This month I talked to Dr. Merlin Butler from Kansas University Medical Center about an interesting clinical case he described recently in the Journal of Pediatric Genetics, concerning cardiac presentations in a case of classic Ehlers-Danlos syndrome with COL5A1 mutations.

                             Keep listening for that interview, but first, let's talk about the papers in this month's issue of the Journal.

                             Our first paper, entitled "Effects of Genetic Variance Associated With Familial Hypercholesterolemia on LDL Cholesterol Levels and Cardiovascular Outcomes in the Million Veteran Program." Comes from Yan Sun, Peter Wilson and co-authors on behalf of the V.A. Million Veterans Program. They were interested in the relatively between variants in LDLR, APOB and PCSK9, and LDL cholesterol in the general population. Low-frequency variants in these genes have been identified to underlie the greatly elevated LDL cholesterol seen in cases of familial hypercholesterolemia, but the effects of the population level are unknown.

                             Using data from the Million Veterans Program, the team analyzed the associations between putatively pathogenic variants and the maximum recorded LDL cholesterol level, as measured repeatedly over a 15-year period, in over 330,000 participants. They restricted analysis to variants that were present in at least 30 people and found that eight of the 16 variants tested were associated with significantly higher LDL cholesterol. Through phenome-wide association analysis, they found that carriers had a higher likelihood of a diagnosis of hypercholesterolemia or coronary heart disease, but not of other diagnoses. Even though individuals carrying risk variants generally reduce their LDL cholesterol through statin treatment, they still had residual risk, suggesting that even earlier initiation of treatment may be required in individuals with genetic risk of high HDL.

                             Continuing the theme, the next paper comes from Laurens Reeskamp, Merel Hartgers, Kees Hovingh and colleagues from the University of Amsterdam, and is entitled, "A Deep Intronic Variant in LDLR in Familial Hypercholesterolemia: Time to Widen the Scope?" This team had encountered a family with familial hypercholesterolemia, who did not carry a coding mutation in LDLR, APOB or PCSK9, and they wanted to figure out what was causing the elevated LDL cholesterol in this family. They conducted whole-genome sequencing in nine family members, five affected and four unaffected. They found a variant in an intron in LDLR, which resulted in an insertion of 97 nucleotides, leading to a frame shift in premature stop codon in exon 15 of LDLR. They confirmed the disease segregation in a second family, and found a frequency of over 0.2% in additional FH cases without a confirmed mutation.  This study highlights the need to consider more than just exons when looking for causal variants, particularly in families where no coding mutations are identified.

                             Next up, from Kathryn Siewert and Ben Voight from University of Pennsylvania, a paper reporting that "Bivariate Genome-Wide Association Scan Identified 6 Novel Loci Associated With Lipid Levels and Coronary Artery Disease." This paper started with a premise that, because heritable plasma lipids are genetically linked to coronary artery disease, we would have greater power to detect variants contributing to both traits by conducting joint GWAS analysis, rather than independent analyses for lipids or coronary disease, as has been done traditionally. Using data from over 500,000 individuals for CAD and over 180,000 individuals from the Global Lipid Genetics Consortium, they conducted a bivariate GWAS and identified six previously unreported loci that associated with CAD and either triglycerides, LDL cholesterol or total cholesterol. Many of these loci also had signals for effects on gene expression of genes in the region, suggesting that these novel loci may affect lipid levels and CAD risk through modulation of gene expression. Interestingly, for some of the newly-identified loci, there were multiple potential regulatory targets, suggesting that these loci may affect lipids and CAD through separate mechanisms. Overall, for closely-linked traits such as lipids and CAD, this joint GWAS approach gives additional power to detect novel variants.

                             The next article comes from Terry Solomon, John-Bjarne Hansen and colleagues from University of California-San Diego and the Arctic University of Norway. Their paper concerns the "Identification of Common and Rare Genetic Variation Associated With Plasma Protein Levels Using Whole-Exome Sequencing and Mass Spectrometry." They were interested in identifying genetic variants that associate with plasma protein levels, both to understand genetic regulation and to identify potential sources of bias, where a genetic variant affects the assay used to quantify the protein, without necessarily altering biological components of the protein. Using data from 165 participants of the Tromsø Study, they quantified 664 proteins in plasma by tandem mass tag mass spectrometry and genotypes by whole-exome sequencing. They identified 109 proteins or peptides associated with genotype, and of these identified 49 that appeared to be technical artifacts based on genotype data.  Of the rest, many of the genetic variants affected protein level by modulation of RNA, but some appeared to directly affect protein metabolism. Their method of quantifying multiple peptides from each protein and sequencing exons allowed them to identify spurious associations that would often be missed, and highlights the large number of artifacts that could be present in protein quantitative trait locus studies. At the same time, they show that over half of the pQTLs are real, with genetic variants affecting circulating proteins through diverse mechanisms.

                             Our last of the full-length original research articles also applied proteomics. "Proteomic Analysis of the Myocardium in Hypertrophic Obstructive Cardiomyopathy" comes from Caroline Coats, Perry Elliott and coauthors from University College, London. They obtained myocardial samples from 11 patients with hypertrophic cardiomyopathy and measured over 1500 proteins using label-free proteomic analysis. They compared protein expression to six control samples from healthy hearts. They identified 151 proteins that were differentially expressed in HCM hearts, compared with control, and they validated a subset of these using an additional 65 myocardial samples from healthy and diseased subjects. Of eight validated differentially expressed proteins, they represented pathways in metabolism, muscle contraction, calcium regulation and oxidative stress. Of particular interest, they highlighted lumican as a novel disease protein, and showed the potential of proteomics to identify mechanisms underlying HCM.

                             We have two research letters this month, the first from Hisato Suzuki, Kenjiro Kosaki and coauthors from Keio University School of Medicine at Tokyo. It's titled, "Genomic Comparison With Supercentenarians Identifies RNF213 as a Risk Gene for Pulmonary Arterial Hypertension." In this letter, they were interested in identifying genetic variants underlying pulmonary arterial hypertension. They hypothesized that individuals with extremely long lifespan would be less likely to carry potentially pathogenic variants. They performed whole-exome sequencing in 76 individuals with PAH and compared them to 79 supercentenarians who had lived for over 110 years. They report variants in RNF213 and TMEM8A that were present in PAH but not in the controls, suggesting these genes may be important in the pathophysiology of PAH.

                             The second research letter comes from Tessa Barrett, Jeffrey Berger and colleagues from New York University School of Medicine, and is entitled, "Whole-Blood Transcriptome Profiling Identifies Women With Myocardial Infarction With Nonobstructive Coronary Artery Disease: Findings From the American Heart Association Go Red for Women Strategically Focused Research Network." Most of the 750,000 acute MIs occurring in the U.S. each year are caused by obstructive coronary artery disease, but around 15% of the acute MIs occur in individuals whose arteries have less than 50% stenosis and are defined as unobstructed. These individuals are more likely to be female and of higher morbidity and mortality. In this AHSAFRM-funded project, the team sequenced whole-blood RNA from 32 women who presented with an MI with or without CAD, or controls. They report several thousand transcripts differing between groups on conducted pathway analysis, which highlighted several pathways, most notably estrogen signaling. This suggests that estrogen may be a novel component in MIs occurring in the absence of obstructive disease.

                             We also have two clinical letters this month. The first, "Desmoplakin Variant-Associated Arrhythmogenic Cardiomyopathy Presenting as Acute Myocarditis," is brought to us by Kaitlyn Reichl, Chetan Shenoy and colleagues from University of Minnesota Medical School. They report a case of a 24-year-old man presenting with acute myocarditis, who was found to have a pathogenic variant in desmoplakin underlying arrhythmogenic cardiomyopathy, also present in his father and one brother. This case highlights myocarditis as a possible initial presentation of arrhythmogenic cardiomyopathy, which requires cardiac MRI and genetic testing for full evaluation.

                             The second clinical letter comes from Judith Verhagen, Marja Wessels and co-authors from University Medical Center, Rotterdam, and is entitled, "Homozygous Truncating Variant in PKP2 Causes Hypoplastic Left Heart Syndrome." They report on a family with consanguineous parents, where two children were affected with left ventricular hypoplasia, leading to intrauterine death in one child and death at day 19 of life in a second child. Sequencing identified a variant in PKP2, which encodes plakophilin 2. Both parents were heterozygous for the mutation, and their affected children were homozygous for the mutation. This mutation resulted in disorganization of the sarcomere and affected localization of other proteins affecting gap junctions. The case highlights PKP2 variants as causal in hypoplastic left heart syndrome.

                             Dr. Merlin Butler is a professor at Kansas University Medical Center and Director of their Division of Research and Genetics. Dr. Butler joined me to discuss an interesting case of Ehlers-Danlos Syndrome in a father and son, with heart failure in the father. This case is in press in the Journal of Pediatric Genetics, and the prepublication version is available online, published on the 13th of October 2018. We'll tweet out a link to that paper, if you're interested in viewing the full case, but here's Dr. Butler, who joined me to discuss it now.

Dr. Butler:          ... I'm a clinical geneticist here at University of Kansas Medical Center, and I see both adult and pediatric patients, but one of the more common reasons for referral to my adult side clinical genetic services is connective tissue disorders. And that's how we were involved with this particular family, a son and father, that led to my interest in looking at the question about genetics of cardiac transplantation of those patients that present for cardiology services because of heart failure and worked up and ultimately end up as a candidate for transplantation.

                             And that transpired in this particular family, which the patient was a 13-year-old boy who was referred into the clinic because of connective tissue disorder. Actually the primary care wanted to rule out Ehlers-Danlos Syndrome. And so we evaluated the 13-year-old boy in the clinic setting, and then we ordered comprehensive connective tissue and next-generation DNA sequencing panel, and lo and behold, he had a mutation of the classical gene that causes classic Ehlers-Danlos, the collagen 5A1 gene. The gene variant was classified as unknown clinical significance, which is often the case as we know with this technology, next-generation sequencing. Regardless of the condition we're looking at, we find about 10% of time, the panel of tests, the panel of genes that come back that are tested. 10% of the time we find no variants, no spelling errors, no mutations. 10% of the time the results come back from the commercial laboratory ... these are clinic patients, so it's done in commercially-approved laboratories, clinically-approved laboratories ... and we find that about 10% is pathogenic, which means it's disease-causing. The gene variant or mutation has been reported before. There is information in the literature that we know that it causes disease, Ehlers-Danlos, whatever type.

                             About 80% of the time, the results come back as unknown clinical significance, and this is related to connective tissue. You probably order a test in cardiology or any other service and you'll find the same area. Most of the variants come back as unknown. What is meant by that is they haven't been reported previously in the literature, and therefore we don't know ... They may be disease-causing, that particular change, but we don't know that. We as geneticists, we have to then figure out whether that gene variant is a mutation or background noise. So we go through a process by where we try to characterize that particular gene finding to see whether it could be causative in that particular patient we see, or if it looks like it's probably tolerated and is just background noise, and it has really probably no apparent phenotypic change resulting from that particular gene variant.

                             So this particular gene variant that we found, the collagen 5A1, did meet the criteria. We looked for computer programs and silica prediction to see if it was tolerated or damaging. We looked at how common that gene variant is seen in the general population, looking at exact various types of genome databases at the laboratories used to search for that variant in the population that's been serviced by genetic services, to see how rare it is or how common it is. We also check to see if it's a missense change, missense variant that is, one amino acid got switched for a different amino acid. There are five classes of amino acids, so if they stay within the same class, that change one amino acid to the next probably doesn't have much meaning, but if it changes to an entirely different class, like positive to negative, hydrophilic to hydrophobic, that could make a big change at the protein translation level, and therefore impact on protein development and function.

                             And then we looked to see if it's conserved in evolution. The laboratories that we use, they look at approximately 80 different animals, mammals, vertebrates, primates, non-mammal vertebrates, to see if that particular spelling change is conserved throughout evolution. If it is, if C is always that position 205 in the coding sequence of that gene throughout evolution, that means you need to have C at that position, not A, G or T, because that would be conserved and impact that we don't want to change that, because it's conserved through evolution.

                             So those kind of criteria, how common it is in the population, how conserved it is, what the amino acid change might be and what the computer programs predict that change might relate to the function of the protein. So we used those criteria, found this gene variant, although it hadn't been reported before ... well, it hasn't been characterized as pathogenic. In this particular family, 13-year-old son and 55-year-old father, they both had the classical features of classic Ehlers-Danlos, so that gene variant, we know at this point is informative.

Dr. Ferguson:    That's a really helpful introduction to how you go about looking at variants and screening them and picking the ones of most importance. So you had this 13-year-old patient who came in and then you tested the patient, and then did you also test both parents? Other family members?

Dr. Butler:          Well, the mother was no longer in the loop, so the primary care, the pediatrician, referred this 13-year-old boy because of joint laxity. He had experienced multiple spontaneous knee dislocations, beginning around nine years of age. He was 13 when I saw him in clinic. He had a history of knee pain, generalized joint hypermobility, loose skin, excessive bruising and poor scarring. And he had that history coming in, and we certainly could identify those findings on this patient. In fact, we reported this patient in the literature. The title of the paper is "Classic Ehlers-Danlos Syndrome in a Son and Father with a Heart Transplant Performed in the Father," published in Journal of Pediatric Genetics, but during a genetics clinic visit, we assessed a hypermobility Beighton scale, that we used to determine the degree of hypermobility, hyperflexibility, and we recorded a score of eight out of nine. Nine is the maximum number. And what we use as kind of a cut-off, this score is five or more, five out of nine or more, then that would indicate that probably there is some kind of joint issues, connective tissue disorder in the way.

                             He had no heart murmur detected, normal rate and rhythm, but a previous echocardiogram showing he had no valvular problems but he had aortic root dilation. He also had skin marbling, atrophic scars, particularly on the lower leg, and increased pigment secondary to easy bruising. He had asymmetry of the anterior body wall, pretty classical findings that we recognize in Ehlers-Danlos.

Dr. Ferguson:    So the reason we're talking to you about this is actually less related to the son, right? And then related to what you found in the father.

Dr. Butler:          The father, right. So the father was 55 years old when we saw him. So we did testing on the son, based on his examinations, and then we obtained DNA and we found out, had the sequencing. We found he had a gene variant of the collagen 5A1 gene. And the collagen 5A1 codes for collagen, low fibrils protein changes, and that's a classical finding we see in Ehlers-Danlos. So we then, on follow-up, we looked more closely at the father, based on what we found in the child, and the father is 55 years of age and he exhibited similar clinical features seen in his son, including stretchable, thin skin, poor scarring, hypermobile joints, with pain and easy bruising. He had a Beighton score of six out of nine, but due to multiple knee surgeries, we were really not able to able to assess his knee findings.

                             And he had strabismus repair when he was like 12 years of age. He had surgery on his right knee due to frequent dislocations, and had bilateral foot surgeries due to flat feet, pes planus. He had a stroke at 37 years of age, but without hypertension. At 43 years of age he underwent a heart transplant because of heart failure with no known cause, such as infections or anatomical defects or metabolic problems seen. And at 54 years of age he had fusion of the lower vertebrae, correct complications, nerve compression, impacting ambulation. So he had multiple, multiple problems, and we did DNA testing on him. He also had the same gene variant of the collagen 5A1 gene, which causes classic Ehlers-Danlos Syndrome.

Dr. Ferguson:    Yeah, so he essentially had been undiagnosed his entire life, I guess.

Dr. Butler:          In his entire life, he just kind of lived with it. Obviously no one really picked it up because he had multiple, multiple orthopedic surgeries. Of course he had the cardiac transplant because of a very large heart size. They didn't really find out what had taken place with that. They didn't find any reason why he had heart failure. So, because of this connective tissue issue, I began to think more closely about this. Could somehow his cardiac transplantation due to no determined reason why he had heart failure, could that somehow be related to a connective tissue problem, such as classic Ehlers-Danlos?

                             And classic Ehlers-Danlos is fairly common, about one in 20,000 people. As far as our concern in the field of genetics, one in 20,000 is common, because we see rare diseases. So one in 20,000 is common. There's like six different categories of Ehlers-Danlos in classic and hypermobile form, vascular form, but he had the clinical findings, he and his son, and he had mutation of a gene that causes classic Ehlers-Danlos.

                             So the thrust of this communication is, could it be that there may be a group of individuals that are on a heart transplantation service, waiting to be transplanted, that might have a connective tissue disorder, such as Ehlers-Danlos or one of the other connective tissue disorders, that could be an issue and a causation of their cardiac issues? We know that there are around 70 genes being recognized that cause connective tissue, and these numbers increase all the time as we learn more about genetics and the capabilities of testing. There are over 130 recognized genes that are thought to play a role in hereditary cardiomyopathies and there are now thought to be over 230 genes that are commercially available in a comprehensive cardiovascular next-generation DNA panels, and several of those genes are collagen genes.

                             So we know there are hundreds of genes that play a role with cardiac health, I guess. Disturbance of those genes, several of those could be connective tissue. Obviously there's others involved, too ... myopathies and conduction issues. But the question I would have, the focus is, could there be a group that would have a connective tissue? And why is that important? Well, not only do they have issues when it comes to these multiple surgical concerns, but they may have, obviously, concerns that might be related to complications of surgery. We know that connective tissue disorders, they have poor wound healing, scarring and other tissue involvement such as vascular anomalies, aneurysms. So they become ... whether it's for cardiac procedures or whether it's orthopedic, whatever ... they become poor candidates for surgical intervention, for surgical operation procedures, because of the complications of surgery. Connective tissue, poor wound healing, scarring.

                             And because connective tissue is involved in not only the skin, but involves internal organs such as the vessels, where you're concerned about aneurysms and vascular anomalies, that could be playing a role. So there may be more complications related to the surgical procedures than your typical patient who undergoes heart transplantation.

                             So I think that would be important to know, so I would encourage, for the cardiology services, for patients that are on these transplant care and services, to consider a comprehensive genetic DNA analysis to look at connective tissues, as well as other causations of cardiac disease. As I mentioned, there's over 200 different genes been recognized now on comprehensive DNA testing panels related to cardiac and connective tissue problems.

                             So, I would encourage that patients that are on the transplant list, they should undergo a next-generation detailed comprehensive connective and cardiovascular panel ... they're certainly available in several laboratory settings ... that might help lead to not only the diagnosis of the cardiac issues, but might help in medical management and monitoring and the surveillance, as well as the surgical interventions and care following the surgical procedures might be taking place.

                             Frequently have an arterial wall might be a little fragile and obviously clamping during surgical procedures for an extended period of time might cause some trauma, even to a normal artery, let alone an artery that might be disorganized because of connective tissue problem.

Dr. Ferguson:    Yeah.

Dr. Butler:          So those complications might occur as well, too, during the procedure or following the procedure. Even there may not be any aneurysms going in, there might be a weakness of the arterial wall at the clamp site that could lead to an aneurysm following the procedure, so it needs to be monitored.

                             So I'm just bringing these to the medical attention that may or may not be out there, but I want to bring this to ... You know, there have been over 90,000 heart transplants been done since 1983, at least that many, and there's 23 million people worldwide that are affected with congestive heart failure, and that's about 7.5 million people in North America.

Dr. Ferguson:    Yeah.

Dr. Butler:          So it's out there. Some of these genetic conditions are rare, but collectively they're common. Ehlers-Danlos, one in 20,000, is probably considered rare, but yet it still is not rare to the person that has it.

Dr. Ferguson:    Right, right, and maybe enriched in these patient populations. So is this something you think that could be sort of found with more careful physical exams, or do you think that [crosstalk] genomic sequencing is sort of the best way to get at this?

Dr. Butler:          Well, I think that Beighton scale we just mentioned, the hypermobility scale, just to see if there's, you know, if it's pretty common. Most adults can't put their palms on the floor when they're standing up.

Dr. Ferguson:    I certainly can't.

Dr. Butler:          That is usually not gonna happen for multiple reasons. But maybe some of the cardiologists are, but those that aren't, maybe they should consider, just check for hyperflexibility in their adult patient. [crosstalk]

Dr. Ferguson:    Yeah, that seems like an easy [crosstalk] click-and-check, right?

Dr. Butler:          Right. There being loose skins, poor scarring. You can ask the patient, obviously. Easy bruisability and poor scars, and it's pretty obvious in these conditions. I mean, on a physical exam it jumps out at you, particularly the multiple scars and bruising on the lower extremities with the pigmented because of iron deposits. You'll see that pretty clear in the scarring issues.

                             And they'll tell you, too. I mean, the patients, they know. "Oh, yeah. I'm very hyperflexible." So you just ask the question and the patients will tell you. They say yes, and then it might need further testing physically; that is, actually do the exam and see if they have, on this Beighton scale, what the hyperflexibility score looks like. And if it is positive ... what we consider positive is five and above, five out of nine ... then those would candidates for a comprehensive DNA testing, whether it's related to cardiomyopathies, but I think connective tissue collection genes. Like I say, there's roughly 70 of these genes out there now that we test for in the commercial clinical laboratory setting. That should be monitored, as well as adding other genes if need be. So I'd encourage that.

                             Physical examination number one. If it's positive, then check into a DNA panel for these types of disorders. It could help long-term for the care and outcome of the patient.

Dr. Ferguson:    Yeah. I do think that's really important from the patient perspective and then, if more of these cases start being reported, I think it's very interesting also from the research perspective to find out what are the mechanisms that are potentially linking these mutations to cardiac disorders which have- [crosstalk]

Dr. Butler:          That's true, and also realize that a lot of these patients have hypotension, and that can lead to some complications before, during and after surgical intervention, too.

Dr. Ferguson:    Yeah.

Dr. Butler:          So that's important to realize.

Dr. Ferguson:    Very important. Yeah. So thank you for telling us about this interesting case and for raising this. I think it's an important issue and I'm sure a lot of the cardiologists and clinicians listening will start to look out for connective tissue disorders in their own patients.

Dr. Butler:          I think, first thing is just ask questions. Are you hyperflexible? And they'll tell you. It's something that is very obvious to the patient. It will be obvious to the physician once he or she puts their hands on the patient, examine the patient, they realize, "Oh, this patient really is quite hyperflexible, digits and arms and knees and elbows," et cetera, et cetera. But, just ask the question, are they hyperflexible? If they say no, then the connective tissue is lower. It still could be. There could still be some aneurysms, those kind of things going on because there's, like I say, there's 70 genes, and there's six types of Ehlers-Danlos, so there's many other conditions out there that kind of look like an Ehlers-Danlos, but they're not. They may have another gene involving protein that's related to connective tissue, but not in the Ehlers-Danlos group of disorders or genes. Still could play a role. Could be similar. They may [inaudible] aneurysms, and that's important to know before they get into the procedures, too.

Dr. Ferguson:    Yeah, really important, really interesting. Thank you so much for joining us.

                             Thanks, everyone, for listening. And I wish you all the best for the holiday season, and a very happy new year. We're looking forward to bringing you lots more in 2019.

                             This podcast was brought to you by Circulation: Genomic and Precision Medicine, and the American Heart Association Council on Genomic and Precision Medicine. This program is copyright American Heart Association, 2018.