Genomics Is Changing Management of Chronic Diseases

An Expert Interview With Bradley Patay

Jacquelyn K. Beals, PhD

March 16, 2012

March 16, 2012 — Editor's note: Chronic diseases have been considered an area of healthcare in which advances in genomic medicine might ultimately yield the most benefits for public health. Diseases such as diabetes, heart disease, cancer, and many pediatric conditions fall under the heading of chronic disease. Each of these lends support to the belief that exposure to risk factors, even before birth, alters gene expression in ways that favor the development of chronic disease.

In an interview with Medscape Medical News, Bradley A. Patay, MD, assistant professor at Scripps Translational Science Institute, and head of the section of internal medicine at the Scripps Clinic in La Jolla, California, discusses highlights from the Future of Genomic Medicine (FGM) V, particularly from his perspective as a full-time clinician. FGM V was held March 1 and 2 in La Jolla and was sponsored by the Scripps Translational Science Institute.

Medscape: How did FGM V go?

Dr. Patay: What really hit me this year was the personal stories that the Beery family shared, and then some of the case studies about different families involved with certain diseases. Speaking as a physician, it's ultimately about changing people's lives. When you have the Beery family up there talking about their diagnostic odyssey, and then having a really good outcome, it cements to me the potential for this kind of technology.

The Beery children have a neurotransmitter issue that's extremely rare. It is estimated that only about 50 families in the world have it. What this underscores is that rare diseases, medallion-type diseases, may be rare, but when you aggregate them they make up a large problem, both for the families and for the cost to the American economy. One of the issues is how much diagnostic testing physicians like me will do trying to find an answer when it's a potentially life-threatening or serious problem. In the future, genomic information will hopefully be integrated much earlier into the diagnostic strategy than is done today.

I'm a full-time clinician.... National guidelines are good, but we know that people who write the national guidelines have to make them very easy to implement, so some of the effectiveness for individualized patients might be lost. If you could personalize guidelines to make them more specific to the patient, maybe you could improve outcomes. What we hope to do at Scripps, what I've been working on, is modeling disease based on people's data, using algorithms to restratify and predict when people might have issues.

Medscape: What was covered in the presentation on type 1 diabetes?

Dr. Patay: We know that about 50% of the genomic risk of type 1 diabetes involves the part of chromosome 6 where the HLA genes are, but half comes from other parts of the genome.... We have an environment that can potentially increase or decrease the prevalence of disease — type 2 diabetes is a classic example, and you could even argue type 1 is. Then we have this DNA or code or software that runs to tell the body how to operate. A certain combination of "software" might lead to type 1 diabetes in this environment.

For type 2 diabetes, maybe in starvation environments where there's not a lot of food, that genome might potentially increase or benefit your health. What we're learning is that the pathway to type 1 diabetes isn't just one route; there are multiple pathways that can get you to type 1. If we can alter some of those pathways, we may be able to alleviate or prevent type 1 diabetes.

Medscape: Which topics seemed to resonate with the audience?

Dr. Patay: People were interested in the need for algorithms for integration of genomic information into clinical care. The problem right now is that data acquisition is becoming cheaper and cheaper across healthcare, across finance, across marketing. So when you gather data, the ability to data mine is increasing exponentially. But unless we have ways of dealing with that data, doctors are going to drown in it.

For example, if we have to look at 50 different little areas of the genome in type 1 diabetes and try to figure out what the potential risk is and how to act on that, a doctor's going to drown in the data. What we really need is to develop algorithms that will help us make the decisions and provide more effective care.

Medscape: In your own clinical practice, have you seen changes already or are they still just anticipated?

Dr. Patay: I think the first changes are really in the oncology space, in rare diseases, and with certain drugs. In terms of drugs, at the Scripps Clinic we're now testing people for clopidogrel sensitivity in processing that drug, the metabolism, with CYP450 2C19. If you're a nonresponder, we're going to change the way we give the medication or the dosage, or switch to a different drug to help improve outcome and prevent restenosis of the stent. That's a rare outcome for many people but it happens; it makes a difference if you look at a population. Some of these outcomes are so rare that it doesn't actually resonate with an active clinical cardiologist, but if you look at the population level, it makes more sense.

In terms of cancer, I think it's still largely in the research realm, but it's moving forward. An example is Eric Topol's IDIOM project. The Idiopathic Diseases of Man study is similar to what many other institutions are carrying out. Basically, for men, women, or children who have rare disorders we don't know about, we can now apply genomic technology to see if we can find an answer. It's going to help some people who have reclassification of disease.

Medscape: I see you're connected with the College of Genomic Medicine.

Dr. Patay: I think what's needed to drive education in genomic medicine is partnership with national organizations and foundations, essentially all the stakeholders. A lot of good education is being done. The National Human Genome Research Institute has a lot of good information on their Web site, and there are a few others. But people need to be able to embed algorithms into clinical care on a daily basis. The problem is the electronic health records. They're so overwhelmed now with meaningful use, and many other things that need to get done right, that I think genomics is lower on the priority list.

Maybe that's fair, because what really galvanizes physicians is whether information is going to affect decisions and improve outcomes. That's a tough thing to prove in genomic medicine. It's going to take a long time and a lot of money, and it might only affect a few people. Part of what we're struggling with is how to create trials that show care, and effective care, maybe not the traditional study design with randomized controlled clinical trials.

At the end of the day, it is the patients' stories that really affect you the most, more than a bioassay on a certain drug. For example, seeing this child prone on the ground, face down, unable to get up, and then with genomic information they give this child a drug and she is dancing with her sister. That's what I'll remember 4 or 5 years from now, not the 50 loci that were found with type 1 diabetes. For certain people, genomic medicine makes a huge difference.

Dr. Patay reports being an advisor to Invivoscribe.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.