Diabetes and Genomics: A New Perspective from Eric Topol, MD
by tony sheng and kelly close
For this issue, diaTribe had the chance to sit down with Eric Topol, MD, of The Scripps Research Institute and Scripps Health in San Diego and ask him about genomics (the study of genes and their functions) and its fascinating implications in diabetes and the future of diabetes therapy. A well-respected cardiologist and an expert in genomics, Dr. Topol says that the current practice of categorizing diabetes into type 1 and type 2 is overly reductionist and misguided. He believes that if we take genomics into account, diabetes can actually be broken into several more subgroups and subcategories than currently used (by our count, at least 12!). He also discusses how the use of genomics could help better predict and prevent diabetes.
diabetes and genomics
diaTribe: It is an honor to speak with you, Dr. Topol. We are very interested in your work studying the genomic basis of diabetes. Please start by telling us a little about how you came to be interested in diabetes and genomics.
Dr. Topol: My father had insulin-dependent diabetes and went blind at the age 49 from retinopathy. I, at that time, was just starting my residency. I said, “Oh, I think I’ll be an endocrinologist—a diabetes expert.” But somehow when I arrived at the University of California, San Francisco, I got excited about cardiology. It was a particularly incredible time in the cardiology world and that lured me to this field, which offered a sense of gratification in helping patients immediately rather than over an extended period.
Switching to genomics, the story is actually quite interesting. There have been many different studies of the genes associated with both type 1 and type 2 diabetes. I think the bottom line is that now we can dissect the disease into multiple components or subtypes. We used to think type 2 diabetes was all the same—simply a result of insulin-resistance. That’s actually not true.
Type 2s are not all the same. In some patients, the problem appears to be in the binding of zinc and insulin that allows for smooth insulin delivery. In others, it’s the transport of insulin, or the release of insulin from the beta cells, and so on. It appears that several different molecular subtypes exist just in type 2 diabetes—patients seem to be the same, but at the genomic level, they couldn’t be more different. This is true in many different diseases, but it’s especially important in diabetes because the disease is so prevalent. For at least a few percent of people with type 2 diabetes, if we could fix the binding of zinc and insulin, it could prove to be a root cause and a solution with formidable impact.
Rare variations in the ability to handle viruses have recently been discovered for type 1 diabetes. In these individuals—accounting for 3-4% of people with type 1 diabetes—if we knew at birth that these patients were genetically predisposed to the disease, we could potentially immunize babies in the first couple of years. This is just the beginning of such an exciting era of prevention.
diaTribe: Wow—that’s 45,000 to 60,000 people in the US alone who could not have diabetes!
Dr. Topol: Exactly. The ultimate goal is real prevention. The simplistic notions that our immune system is attacked in type 1 diabetes and that there’s plenty of insulin but it’s just that patients are resistant in type 2 diabetes—that has to be completely revisited. The way we categorize diabetes today in practice is essentially a reductionist, over-simplified mistake.
Another important idea is to individualize treatment for people with diabetes. The goal is for every single person to understand, ideally, why he or she is predisposed so the disease can be prevented.
individualizing treatment
diaTribe: Could you talk a bit more about how and why we should individualize treatment for people with diabetes and the role of genomics in doing so?
Dr. Topol: If we continue to approach diabetes as we do today, I don’t think we will make much headway. There’s regrettable mismanagement of diabetes because genomics are not accounted for. For example, there are at least 20% of European ancestry people with type 2 diabetes who are prescribed metformin but have no response to the drug because they have a variation of the organic cation transporter 1 (OCT1), the gene that basically is responsible for metformin’s action. Nonetheless, metformin is the number one prescribed drug in type 2 diabetes. Can you imagine, at least a fifth of these people are getting a drug with little or no effect?
Another good example of the use of genomics comes with diagnosing pre-diabetes or diabetes incidence using the fasting blood glucose test. Everyone thinks intuitively that if you have a high fasting glucose then you must be pre-diabetic or diabetic. Well, it turns out there's a significant minority of people who have a high fasting glucose who actually are protected from ever getting diabetes. The gene that is associated with this high glucose but diminished risk is called G6PC2. This finding is striking but the mechanism still needs to be understood—nevertheless, the fact is they have up to a 20% reduced lifetime risk of ever developing diabetes. We’re missing this completely because we’re treating them as if they are all destined to have diabetes. We’re basically committing to a lifelong treatment with drugs, which carry side effect potential and expense they don’t need.
And then, as I mentioned before, there are at least a few percent of patients with type 2 diabetes that become type 2 solely because their insulin can’t bind to zinc. The gene involved is called SLC3OA8. This gene in particular is a very important gene to unravel because if we could just fix that, then we would be able to prevent people from getting diabetes. It’s not a common case but it’s certainly a somewhat important one to recognize.
diaTribe: What percentage of type 2s do you think have that gene?
Dr. Topol: Probably three or four percent.
diaTribe: That’s basically 700,000 people you could help.
Dr. Topol: Exactly. Potentially...
diaTribe: Can you foresee a time when genetics guides treatment in type 2 diabetes? How far off is this?
Dr. Topol: Yes, absolutely. Matching up the right medication at the right dose for the right patient is going to eventually become routine. . . It’s just a matter to time.
genomic scans and the future of diabetes therapy
diaTribe: Is there a practical way that we can give genetic tests to a certain percentage of the population who are at risk for diabetes?
Dr. Topol: Well, I think that’s really where it’s headed. Someday everyone will have the option of having a genomic scan. With the information from the scan, when a patient goes to the doctor with a high fasting blood glucose, the doctor can say, “Oh, well, you have a protective gene. We’re not going to treat you.” Or, “You have this unresponsive gene, so we’re certainly not going to treat you with metformin.” The doctor could basically go down a list of various options.
diaTribe: That’s incredibly exciting! How do you get your genome scanned? We know about 23andMe.
Dr. Topol: It is exciting, but we’re not there yet. 23andMe is one of the testing services. The other two are Navigenics and deCODEme. By getting all your data and identifying all of your key single-nucleotide polymorphisms (SNPs)—small genetic variations that characterize an individual’s genome—you can look up anything you want. You just look up the SNP (using SNPedia, a Wikipedia-style database of SNPs, beyond the output provided by the companies) that you have and then you can determine the known effects. Over time, so many of the things that we don’t know today will gradually get filled in too.
diaTribe: On a practical note, how can patients get tested? Are the facilities in many hospitals? Can it be done by mail order? How should results be interpreted if one’s primary doctor isn’t familiar with the research?
Dr. Topol: The specific diabetes-related tests are not set up for routine use except for TCF7L2, a gene involved in the development of type 1 diabetes, which can be done by deCODEme via a swab of the cheek. So we’re not ready yet, but hopefully we will be in the next couple of years.
diaTribe: And, can you discuss the costs? Some would argue the costs of genome scans are too prohibitive and unrealistic to be widely used.
Dr. Topol: Well, the sum keeps coming down. It’s now down below $1,000. We’ve done a study here in thousands of patients where we were able to get the cost down to just even a couple of hundred dollars. It will keep getting less expensive over time because the chips that are being manufactured to do it are getting cheaper all the time.
Genomic mapping will eventually get us to defining the root cause of the problem in the individual being assessed. Over time, that ability has tremendous promise well beyond diabetes.
diaTribe: This is fascinating. Is there anything that you can say to people who already have diabetes to prevent complications or educate them a little bit more about their risk of complications?
Dr. Topol: Well, that’s a great question. What I’d like to do is a genome-wide study to understand the genes underlying diabetes-related complications. Some people with diabetes never get complications and there have to be new genomic pathways that account for these differences—it’s not just about dieting or glucose management. But that particular study hasn’t yet been done and needs to be done.
It brings to mind the Sonia Sotomayor story. Would diabetes affect her ability to execute her role in the Supreme Court over the long term? That was just the question of the week. But the question comes up all the time: Who is relatively diabetes complication “immune” and who is predisposed like I saw with my father who had every complication you can imagine?
diabetes, complications and genetics – and clinical implications
diaTribe: Dr. Topol, what is the state of our understanding of the genes associated with diabetes?
Dr. Topol: Well, there are about 50 genes now for type 2 diabetes and 20 for type 1 that have been identified. You can partition people by their dysfunctional genes to some extent to start to determine at least the mechanism via which their diabetes comes about. We’re sequencing the whole genome of individuals now, so, over the next couple of years, a lot of the holes in today’s understanding will be filled in with new information.
diaTribe: Why do you think more attention isn’t being paid to genomics and diabetes within the diabetes world?
Dr. Topol: Our program is trying to change this and we need more like-minded programs and clinicians. Unfortunately, most of the work on diabetes in genomics has been published in Nature, Science, and Nature Genetics. People involved with diabetes don’t really focus on these journals.
diaTribe: As we understand it, you’ve identified 12 types of diabetes. Could you tell us more about them? Are certain types more serious than others in terms of complications?
Dr. Topol: Inevitably, there’ll be at least 12 types that can be easily defined, but there’s more work to be done. We still do not understand the link between the disease and the genes that indicate relative susceptibility to, or protection from, complications. That’s a big hole right now.
diaTribe: Got it. And so, what would you say are the most important implications for patients and providers?
Dr. Topol: Well, I mean, it’s a little early to begin changing management and prevention. But I think we’re approaching a very exciting time. Over the next few years, we should see some radical improvement. Not everyone will have a positive benefit, but for those people who we really understand the root cause of their diabetes, we may see some big improvements in their therapy.
diaTribe: Dr. Topol, thank you immensely for sharing your time and insights with us.
