Sep 21, 2017
Jane Ferguson: Hello, and welcome to episode one of Getting Personal: Omics of the Heart, a podcast from the Functional Genomics and Translational Biology Council of the American Heart Association. I'm Jane Ferguson, the current chair of the FGTB Professional Education and Publications Committee. This monthly podcast will bring you up to date with the latest in genomics, other omics technologies, and precision medicine as they relate to cardiovascular and metabolic disease.
In each episode, we'll give you an overview of some of the latest research to be published, and delve deeper into topics of particular interest. Whether you're a clinician, researcher, genetic counselor, or other healthcare or science professional, we hope these podcasts will be informative, and help you stay up to date with the latest developments in this exciting field.
In this episode, my colleague Naveen Pereira talks to Amit Khera about his recent publication with Sek Kathiresan and colleagues in the New England Journal of Medicine entitled Genetic Risk, Adherence to a Healthy Lifestyle, and Coronary Disease, and we highlight a recent AHA scientific statement on the use of genomics. But first, Naveen and I will give you a round up of some interesting papers from the past month.
Naveen Pereira: So Jane, there was this really interesting paper in the American Journal of Medicine whether we can use gene expression signatures along with other clinical covariates to predict the presence or evaluate whether symptoms are suggestive of obstructive coronary artery disease.
Jane Ferguson: Yes. This paper was published online on the 16th of December 2016. It comes from Joseph Ladapo, Mark Monane and colleagues. They carried out this study in 566 patients from the PRESET Registry, which enrolled stable, nonacute adults presenting with typical or atypical symptoms that were suggestive of obstructive coronary disease.
What they did was calculate an age/sex/gene expression score, or ASGES score. They included gene expression, which they measured in a blood sample collected in a PAXgene Tube, and this score ranges from 1 to 40. They've previously validated this, and a score less than or equal to 15 indicates that a symptomatic patient is very unlikely to have obstructive coronary artery disease. The genes they measured include 23 genes that are selectively expressed in circulating neutrophils, NK cells, and B- and T-lymphocytes.
Naveen Pereira: So really, this expression reflects inflammation, and the hypothesis being perhaps these inflammatory markers are very indicative of the presence of obstructive coronary artery disease, or plaque rupture I guess, huh?
Jane Ferguson: Yes, exactly. What they actually found was that individuals with high scores were referred to cardiology or advanced cardiac testing at far greater rates, and then even of subjects with low scores who did undergo additional testing, none of them had any detectable abnormality. Then, in subjects with high scores who did undergo further testing, 14% had abnormal findings. So after a year of followup, 1.2% of patients with an ASGES score below 15 had an adverse event, compared to 4.5% of those with elevated scores.
Naveen Pereira: So a fairly high negative predictive value, huh Jane?
Jane Ferguson: Right. Right, exactly.
Naveen Pereira: Did you find any limitations, Jane, in this study?
Jane Ferguson: There were some. Well firstly, it's worth noting that the score itself, and this test, has been developed by CardioDx, and a number of authors on this manuscript are affiliated with CardioDx. In addition to that, they did not include a control group in this. That certainly is somewhat of a limitation, but the authors say that this is probably still useful, and it may have some clinical utility in guiding decision making for patients with obstructive CAD. However, whether or not this is actually true will probably require some additional testing.
Naveen Pereira: So quite a foray into using this perhaps in the emergency room or in hospital. So I guess our audience should look out for this in the American Journal of Medicine, December 2016.
Jane Ferguson: Yeah.
Naveen Pereira: So there was another paper that we kind of thought was interesting, Jane, from the European Heart Journal.
Jane Ferguson: Yes, exactly. This comes from Jozef Bartunek, and Andre Terzic, and their colleagues, and they were reporting this on behalf of the CHART Program.
Naveen Pereira: So this was published on January 15, 2017.
Jane Ferguson: Yeah. This was a prospective, randomized, double-blind, sham-controlled trial, which was the Congestive Heart Failure Cardiopoietic Regenerative Therapy, or CHART-1 trial. In this trial, they were aiming to test safety and efficacy of delivery of cardiopoietic cells. They recruited subjects who had symptomatic ischaemic heart failure, and they consented to bone marrow harvest and mesenchymal stem cell expansion. They ended up randomizing 315 subjects.
They received cardiopoietic cells delivered endomyocardially by a retention catheter, or either a sham procedure. The outcome that they were looking at was a hierarchical score, which is assessed 39 weeks post-procedure. That score comprised all-cause mortality, the number of worsening heart failure events, the Minnesota Living with Heart Failure Questionnaire, a difference in the six minute walk test, change in left ventricular end-systolic volume, and change in left ventricular ejection fraction. So it was interesting. They found a neutral effect on the primary end point, but they did find some evidence of benefit in subgroup analyses, which were based on baseline heart failure severity.
Naveen Pereira: But and this was not designed to show efficacy, because it was primarily a safety trial. Is that right, Jane?
Jane Ferguson: Yes, exactly.
Naveen Pereira: Right.
Jane Ferguson: Overall, they found that there were no indications for concern regarding safety, so I think they've shown that certainly this is a technique that is safe and is well-tolerated, and I think it's really quite exciting. Future studies that are adequately powered, particularly looking at subgroups of individuals, may actually identify patient populations that would derive particular benefit from cardiopoietic cell therapy.
Naveen Pereira: Fascinating, so it'll be interesting to see what the Phase III clinical trial will show. Overall, a new foray into regenerative medicine.
Jane Ferguson: Yeah, yeah. Really interesting.
Naveen Pereira: Hi everybody. My name is Naveen Pereira. I'm from the department of cardiovascular diseases at Mayo Clinic in Rochester, and on behalf of the Functional Genomics and Translational Biology Council of the American Heart Association it gives me great pleasure to interview Amit Khera. We are going to be discussing this very exciting paper that was published in The New England Journal of Medicine on November 13, 2016, titled Genetic Risk, Adherence to a Healthy Lifestyle, and Coronary Disease. Amit, welcome. We are so glad you could make it. We really appreciate you doing this for us.
Amit Khera: Naveen, thank you so much for having me. It's a real pleasure. By way of introduction, as you said my name is Amit Khera, I recently joined as a staff cardiologist at Massachusetts General Hospital in Boston where I see both general cardiology patients and also work in the prevention center. But one of the things I've noticed is that many of us have heard a lot about precision medicine, and how we can incorporate genetics into some of our clinical decision making, or risk stratification. So I've really been working with Sek Kathiresan at both Mass General and the Broad Institute to get training in both genetics to complement some of the clinical medicine aspects in order to help us get at some of those questions.
Naveen Pereira: Fantastic. Amit, what got you interested in genomics?
Amit Khera: Sure. Well, you know for a complex disease like coronary artery disease, and risk of a heart attack, we've really known for a long time, like since the 1960s that there is a familiar pattern, meaning that if your brother or your father had a heart attack at a young age, your risk of having one is increased by almost a factor of two. It's really been only recently that the technology has allowed us to get at those questions, and really isolate the exact genetic determinant.
So really in the last 10 years, we've been able to identify a large number of variants that influence an individual's risk of coronary artery disease. So it really an opportunity to be in a place where the technology was coming along, where we have discovered all these variants, clinical medicine of course has come a long way since the 1960s as well. So the idea was to really put these two bodies of work together, and see what we could come up with.
Naveen Pereira: Yeah. This is very exciting. Amit, I completely agree with you. In our clinical practice we see patients with strong familial history of coronary artery disease, so certainly inheritance has been suspected for some time, and in fact genome-wide association studies have been done to identify loci for coronary artery disease.
As you know, the effect size of these individual variants have been small. And so groups have got together to form genetic risk scores, where they take kind of an aggregate of the effect of these individual variants, and we think this is more helpful. And this is what you did for your paper, so can you describe to us a little bit about how you derived the genetic risk score that you applied in this great paper?
Amit Khera: Sure. The first aspect of our paper involved proving basically that we could quantify someone's genetic risk for having a heart attack. So in order to do that, as you said, we took advantage of a number of previously published genome-wide association studies. There are about 50 genetic variants all across our genome and different chromosomes that we know are strongly linked to coronary artery disease from a statistical standpoint, but actually might only have a very modest impact on coronary disease.
So let's say any individual could have a maximum of 100 risk variants. Now, some people might have inherited just by chance 80 variants, and other people might have inherited only 20. So we basically genotyped, meaning measured all 50 genetic variants in a large number of people, and then we said, "Those who are in the top quintile," meaning the top 20% of the genetic risk score, we're going to say "those people are at high genetic risk." And by contrast, if you're in the lowest quintile, we said, "Okay, those people are at low genetic risk."
Naveen Pereira: Right.
Amit Khera: Then the question became okay, well does that categorization actually predict your risk of having a heart attack? So in order to do that, we analyzed over 50,000 individuals from three different prospective cohort studies, and what we found actually was that if you compare the high genetic risk to the low genetic risk people, their risk for having a coronary event over prospective follow up was increased by about 91%, meaning almost two fold.
Naveen Pereira: Wow. Oh, that's amazing. So using the genetic risk score, you could almost predict a doubling of the risk for coronary events. That's fantastic. Can you describe these populations briefly, Amit? Who are these people that you applied the genetic risk score to?
Amit Khera: Sure. So we took advantage of three prospective cohort studies. The first was a Atherosclerosis Risk in Communities study, and that was a community based population of about 8,000 people. The second was the Women’s Genome Health Study, over 20,000 women who were originally recruited as part of a randomized control trial, and the third was the Malmö Diet and Cancer Study, which again had more than 20,000 individuals.
The really nice thing about these studies was that they were asked questions in a similar way, and they were followed ... in each case, participants were followed for about 20 years. So we really had a long time to observe what happened to these folks over time.
Naveen Pereira: So these are really longitudinal cohorts, not specifically disease oriented cohorts, but just community based, Amit?
Amit Khera: That's exactly right, and in fact none of the individuals had coronary disease at baseline. They were all disease free-
Naveen Pereira: I see.
Amit Khera: ... and then we followed them over 20 years to see who developed the coronary artery disease and who did not.
Naveen Pereira: So this is really applicable to the general population.
Amit Khera: I do believe that these risk estimates would for sure hold true.
Naveen Pereira: Okay, wonderful. So Amit, you know you have the genetic risk score for coronary artery disease, and you have some great longitudinally followed population based cohorts, and you were studying a specific phenotype, so can you describe to us the phenotype?
Amit Khera: Well, the primary outcome phenotype was incident coronary events, and those were all adjudicated by different committees, but it basically involved individuals who had either a new heart attack or myocardial infarction, they had to have one of their vessels either stented or bypassed via revascularization, or in fact it was determined that they died from coronary artery disease. So that was the outcome which we were trying to predict.
Naveen Pereira: Amit, let's get straight into it. What did you find?
Amit Khera: So as a preventive cardiologist, I often see patients in my clinic who come to me and they say, "You know, almost everyone in my family has had a heart attack." Oftentimes at a very young age, and in some cases that can lead to almost a sense of determinism, where they feel like maybe they are unable to control their fate. So our primary question was a really a pretty simple one, which is to what extent can a healthy lifestyle offset someone's genetic risk of coronary disease.
So as I mentioned, we had a way of quantifying someone's genetic risk, and then we next said, "Okay, we want to quantify someone's lifestyle risk." So for that we kept it pretty simple. We had four criteria of what makes up a healthy lifestyle. First, no current smoking. Second, avoiding obesity. Three, regular exercise, and fourth, adhering to a healthy diet. And we said, "Okay, if you have at least three out of those four," we gave you a pass and said "you had a favorable lifestyle." Now if you had only zero to one out of those four, you had an unfavorable lifestyle.
One of the interesting things was that actually the genetic risk and the lifestyle risk actually were totally independent. There was no association for example between those who had high genetic risk and what their lifestyle was. So it really reinforced longstanding views that genetics and lifestyle are really independent axes of someone's individual level of risk. Now-
Naveen Pereira: So both, Amit, both could contribute to your individual risk for coronary artery disease?
Amit Khera: Exactly. As I mentioned, the high genetic risk versus low genetic risk, there was about a two fold difference in risk, and we saw an almost identical pattern versus a favorable lifestyle versus an unfavorable lifestyle. There was about a two fold risk [inaudible 00:17:08].
Naveen Pereira: Interesting.
Amit Khera: Then that got us to the next question, which is to say if we analyze only those at high genetic risk so everyone had a similarly increased degree of genetic risk, to what extent could that risk be offset by a favorable lifestyle? This really gets back to the questions and the conversations we have with our patients who have a family history all the time. What we found there I think was a nice message, was that if you are at high genetic risk, you could actually decrease your risk by about 50% if you adhered to a favorable lifestyle, as compared to those with an unfavorable lifestyle.
So for example, when we looked at it in absolute terms, in terms of a 10 year risk of having a coronary event, in one of the cohorts, those with a high genetic risk but an unfavorable lifestyle had about an 11% chance of having a coronary event, versus if you had the same high genetic risk but a favorable lifestyle, your risk was only about 5%. And we saw that, a very consistent pattern across all the cohorts and all the categories of genetic risk, that those who had a favorable lifestyle ... the risk was decreased by about 50% in those with a favorable lifestyle.
Naveen Pereira: So that's fascinating, Amit. When physicians see a patient who have a really strong history of coronary artery disease within the family, and come up to you and say, "Doc, am I destined to have a heart attack?" You know, now with the availability of genotyping, with direct to consumer testing, people can find out their genetic risk. So they may not necessarily be doomed. Their fate is not predetermined. What you're suggesting is that fate can be modifiable.
Amit Khera: Right. I think certainly for coronary disease your DNA is not your destiny, at least for these common variants. I think we provide evidence that really lifestyle factors powerfully modify your risk, really regardless of your genetic risk profile.
Naveen Pereira: So Amit, can we make any recommendations based on the results of your paper?
Amit Khera: Well, I think ... The American Heart Association has really endorsed these four lifestyle criteria as a way of improving the population's health in the population as a whole, and I think actually that our results actually support that. Which is to say that this really supports the fact that these healthy lifestyle parameters are critically important for everyone, and I think that's a good starting point.
Genomic medicine is actually in its early days, but really what we hope to do is first to identify individuals, a subset of the population, who are at increased risk for a disease like heart disease, and I think we've shown that we can actually do that reasonably well. Like 20% of the population has a double risk. And the second part is actually to disclose this risk to both the patients and their providers in a way that's meaningful. And third, is actually demonstrate that we can actually implement the therapy to mitigate this increased risk.
So I think we, in this paper, we provided evidence that a healthy lifestyle can mitigate that risk. Papers from our group, both published and some in press, have actually demonstrated that taking a statin can also powerfully modify this increased risk. And you might imagine that there may be other interventions that ... especially if an intervention has increased risk, you really may want to target it to those people who actually ... if a medicine has increased side effects, you may want to target it to those at the highest risk. I think that, you know, this polygenic risk score does provide at least one way of stratifying people into those high risk groups.
Naveen Pereira: Yeah. Amit, really impressive results, 50% relative risk reduction in a high genetic risk population. You make a compelling argument. Obviously, however, this is not a prospective randomized clinical trial. It's really hard to do these. You had the advantage of well designed cohorts to study this in a cross sectional way. We don't know how these behaviors change. So these are some of the limitations, but the results are quite compelling, and contribute to the literature. Any other comments, Amit? Anything else that we should take home here?
Amit Khera: No, I think as you said, there are some limitations. I think our really goal was to lay the foundation for future efforts where we really think about what the optimal way is for genetic information to be integrated into routine clinical practice to help prevent disease, and that's really what our group is planning on focusing on for the future years.
Naveen Pereira: We look forward to hearing more exciting results from your laboratory, Amit. It's been a pleasure. We should end I guess with a quote from William Shakespeare, "It's not in the stars to hold our destiny but in ourselves." Correct?
Amit Khera: Thank you very much. Sounds great.
Naveen Pereira: Thanks, Amit.
Jane Ferguson: So as we just heard from Naveen and Amit, the combination of genetic risk and modifiable lifestyle parameters are crucial in determining CAD risk. A recent AHA statement from the FGTB Council focused on this topic. The statement, entitled Nutrigenomics, the Microbiome, and Gene-Environment Interactions: New Directions in Cardiovascular Disease Research, Prevention, and Treatment, focused on how dietary and genetic contributors to disease have been studied in the past, and how emerging omics technologies can be used to rapidly advance these fields.
Genomics, transcriptomics, metabolomics, proteomics, lipidomics, epigenetic profiling, and metagenomic characterization of the microbiome can all be used alone or in combination to better understand mechanisms underlying gene-environment contributions to disease. While the ultimate goal would be the development of improved therapeutic options, including personalized and precision approaches, a considerable amount of research remains to be done before this goal can be clinically implemented. You can read this statement in the June 2016 issue of Circulation: Cardiovascular Genetics.
Naveen Pereira: So, Jane, this has been an exciting first podcast. I really look forward to doing more with you.
Jane Ferguson: Yeah. I think this is great, so thank you everyone for listening, and happy heart month. We will look forward to bringing you a podcast again next month.
Naveen Pereira: Thank you.