Genetic Testing in Diabetes: Separating Hope from Hype

By Kelly Close, Alexander Wolf, and Maxwell Votey

Twitter summary: We sit down with diabetes luminary Dr. Anne Peters to talk genetic testing, #diabetes, + an upcoming revolution in individualized care

When the genetic ancestry company 23andMe was launched in 2007, it came with a bold promise: for $999 (now $99) and a mail-order saliva sample, you could learn the secrets of your genome – your ancestry, your relatives, your ethnic background – essentially information telling you about what makes you, you. Check that out!

Over time, 23andMe (named after the 23 pairs of human chromosomes) focused on an even loftier promise: learning how genetic information (DNA) can affect your long-term health. For instance, do you have a risk for Alzheimer’s or Parkinson’s? 23andMe offered this service directly to consumers – without the supervision from a healthcare professional. It was trailblazing for sure.

But in 2013 the FDA banned the company from marketing these health reports, citing uncertainty over whether the test was safe and accurate for patients. Indeed, many doctors and researchers questioned whether genetic data could truly tell people about disease risk, and whether the results might cause more harm than good. Without a healthcare professional’s interpretation, could patients understand 23andMe’s information?

This debate dovetailed with larger ethical discussions about patient privacy, genetic discrimination, and a whole host of big picture questions about the access to one’s genome – discussions that continue today.

Despite this rocky history, some healthcare professionals have not been deterred. Many continue to use 23andMe to help patients understand disease risk and even to inform therapeutic decisions. While 23andMe can’t provide a full medical analysis (it has been approved for individual diseases, like Bloom's Syndrome), another company – Promethease – has emerged that can read 23andMe's data and give patients the medical information 23andMe can no longer provide. Promethease works by taking the raw data from 23andMe (and other genetic testing companies) and comparing it to genetic findings in peer-reviewed scientific journals. It then generates a report (see a sample report here) outlining what current research indicates various genetic mutations may signify (i.e., your genes indicate you have a 50% higher risk of Alzheimer’s) – in theory anyone (with access to scientific journals) could do this work themselves, but with Promethease it’s fast (~15 minutes) and cheap at only $5 per report. As Promethease itself does not run any lab work or provide people with genetic testing kits, the FDA’s role in regulation is a bit more blurred – though it may certainly still get involved.

Renowned endocrinologist and diaTribe Advisory Board member Dr. Anne Peters (USC Keck School of Medicine) is a regular user of 23andMe; she used it herself (in combination with Promethease) to assess her personal risk of early onset Alzheimer’s disease. Fascinated by the potential of these tests, she began using them in her clinic. diaTribe’s founder, Kelly Close (one of the interviewers), can relate – she has also used 23andMe, before the FDA raised concerns, and was hugely grateful for all the information.

In our interview with Dr. Peters, we discussed how she uses 23andMe (in combination with Promethease) with her patients, her thoughts on the future of genetic testing, and how it could affect personalized medicine. While Dr. Peters has had positive experiences using 23andMe and Promethease, she made one point very clear: it’s absolutely vital to look at this information with the assistance of a healthcare provider who knows how to interpret it (i.e. a genetic counselor). As Dr. Peters said in the interview, "I think people should know what their genes are if they want to, but they also truly need an expert to help interpret the results." If you're interested in getting your 23andMe information for $99, check out this link, and Promethease's $5 service can be accessed here.

Our Interview with Dr. Anne Peters:

AN INTRODUCTION TO GENETIC TESTING

Q: Thanks so much, Dr. Peters, for joining a diaTribe conversation again! Please tell us, first of all, how did you initially learn about 23andMe?

Dr. Anne Peters: A fellow researcher on aging was very interested in personalized medicine: understanding each person’s specific biology and how to treat them based on their genes rather than just treating them. He actually convinced me that I should determine if I had a risk for early Alzheimer’s. I have two parents with Alzheimer’s, and they both developed it at a later age. He said, “Oh, you should just test your genes, do it through 23andMe,” and explained that when used with Promethease, it can actually give you an analysis of the genetic disease risk.

I did it with some hesitation, because I didn’t really want to know if I was getting early Alzheimer’s, but I also did want to know. You know?

The bottom line is it doesn’t appear that I’m at risk for getting early Alzheimer’s, which is a good thing, but doing this to myself made me fascinated about the process. I learned a great deal about my own genes and what the potential for this field is. And I want to say that the most important point here is ‘potential,’ because so much is not known yet.

Q: That's fascinating. What did your own experience with 23andMe teach you?

A: I have never in my life gotten a mosquito bite, and it’s not that mosquitos don’t bite me; I just don’t get a mosquito bite reaction. It turns out that I am genetically programed to have protection from getting a mosquito bite reaction. Who knew?

I also found out that I’m a slow metabolizer of drugs, which I know intuitively because when I take a drug or any kind of medication, I get side effects easily. So, as a doctor and as a human, I was able to say, “These genetic variations in me make sense.”

The other thing I found is that I have almost every known gene for obesity. But, I also have the gene that indicates a strong response to exercise, which means I burn off calories through exercise well. Isn't that amazing? So, there is something about my metabolism that is really good at burning off calories through exercise that overcomes all the obesity genes. To me that is great, because knowing that means that if I actually wanted to lose some weight, all I would have to do was exercise more and not increase my caloric intake. And eventually we may be able to say to patients, “This is the best diet for you because your genes show us that this is how you burn calories best.” Now that is getting closer to personalized medicine.

BRINGING GENETIC TESTING INTO THE CLINIC

Q: How did this experience influence your decision to use it with your patients?

A: It was fascinating because it validated what I knew about myself already. It also showed me how incredibly damaging this could be if you didn’t have medical knowledge to go with what you find! You may have a protective gene and you may have a “bad” gene, and together they’re a balance of who you are. There is also a huge variety of epigenetic influences that impact how genes are expressed.  I’ve since done this testing with other patients, but again, it is only in the context of me explaining their results to them.

I think that is why the FDA prevented 23andMe from giving the medical analysis portion. I’m mixed — I think people should know what their genes are if they want to, but they also truly need an expert to help interpret the results. But the hereditary stuff –  such as if you’re genetically Northern European – is just fun stuff to know.           

Q: We had never heard about Promethease before you filled us in. Can you tell us more about it and how it analyzes 23andMe genetic results?

A: When I tell people about 23andMe, I tell them what it is at this moment in time — it’s a way to look at your genes. It doesn’t look at all of your genes, but it can give you a sense of your ancestry. A lot of people like that, and I like that too because so much of diabetes is inherited. It’s interesting to me to have a sense of where people come from.

But that is not the most interesting part. I tell people that we can put their results into Promethease and then find out much more about their individual genetic disease risks. [How does this happen? As outlined above, patients can send their raw 23andMe data to Promethease, which then creates a report linking the genetic information to known correlated disease risk]. But I also tell them that I don’t want them doing it alone, and that I actually do it for them and then I give them my sense of it. The list of things it tells you can become a little overwhelming, but you could actually get it to give you pie graphs to say, “Yes, you have 4% of the cancer gene, but you have 96% of the not-cancer gene.” It gives you a ratio, so I use it that way. It helps families understand. I like knowing and giving people some sense of things, but again, you’ve got to understand what it doesn’t do.           

What is going to happen in the next five to ten years is huge as we find out which genes determine health outcomes — particularly with regards to drug sensitivity and changing lifestyle. Not everybody has the same genetic make-up – I tell this to patients who are overweight. It’s key to find the targeted approach to each person’s care.

Q: Do you ever refer patients to genetic counselors or someone that can help them further interpret the sequencing?

A: I send couples who are pregnant to pre-natal counseling as needed based on their ancestry and family history. In some cases, I think it could be done with a knowledgable primary care provider, but for any serious issues (for instance, early Alzheimer’s or Huntington’s Disease), I know that some centers have geneticists and therapists/social workers to help with the diagnosis. A geneticist is also needed for anyone with an inherited gene risk, say for breast cancer, or anyone who needs help in understanding their own risks or risk of potential offspring in a high risk setting. I think that these risks should be interpreted with the help of a professional in genetics, and I am not that person. 

The last thing we want is for people to kill themselves out of worry. All genetic tests need to be put into the context of the individual.

GENETIC TESTING AND DIABETES

Q: What about using these results to assess diabetes risk – either type 1 or type 2?

A: When I take my type 1 patients and put them through 23andMe and Promethease, the results don’t tell me anything about diabetes risk in those patients. So, I don’t think that it can tell you type 1 diabetes risk at the moment from this method. I encourage my patients to enroll in Trial Net, if they are eligible.

It can tell you about type 2 risk to some degree, but there are a lot of things you still can’t test for. It was more helpful for drug absorption, drug effects, and obesity risk. And it was also very helpful about Alzheimer’s risk.  However, most type 2 diabetes genes aren’t known, and even the known gene mutations (such as MODY) are picked up by these methods.

Cancer researchers are using this the most and getting the most benefit from it – they are advancing the field. When I talked to the Promethease developer, I was amazed at how little he thought about diabetes. I think it is because we don’t know enough. You can actually know the specific genotype for a cancer, but you can’t for diabetes. We don’t really know what the diabetes genes are, and in many cases they are likely to be multifactorial.

Q: You talked about what may be coming in the next five years, particularly individual sensitivity to certain drugs. Can you tell us more?

A: It's amazing. We may be able to determine who will and who won’t respond to a given drug. Say a drug under development only works in 25% of people. It’s likely that the drug will fall out of development because it will seem to have no effect. But what if we choose only the 25% who are going to be responders? Then the drug could be amazingly effective. Clinically, I am often “guessing” with medication – I can’t always know who will and won’t respond. Some patients get upset that I “experiment” on them, but all it means is that we can’t always predict in advance what the response will be. Each person is a sample size of one when it comes to medicine. 

As a clinician, that would be so helpful for me to know about rates of drug metabolism. I could put a patient on a micro-dose of metformin if they are more likely to have side effects based on the genetic results. My approach to medicine is “everything should be customized.”

I tend to start medications at doses that are less than the usual starting doses. Why not? I do it based on the response to the drug. Imagine knowing that one person might need a drug every other day and another person needs it every day. I think that will help adherence. Many people take a drug and have a bad experience, which may make them inclined not to take it again. However, with knowledge about drug metabolism we may be able to give a patient the right dose of the right pill from the start, and to me that’s exciting – it makes my practice of medicine easier. 

There’s also the aspect of what your body responds best to in terms of diet or exercise. All of that is going to be very helpful. We also may be able to tell who’s at risk for complications with diabetes and who isn’t.

GENETIC TESTING AND THE FDA

Q: What are your views on the FDA decision regarding 23andMe’s ability to provide medial genetic analysis?

A: Until I did it myself, I thought people should have access to their medical genetic information. I still believe people should have access, but the context is critical. If you are going to give people this information, you need to train doctors how to interpret it.

I think that if people go to their regular doctors and say, “What do these genetic results mean?” – doctors can’t currently handle it. Doctors need to know how to read it. This knowledge is so much more advanced than it was even five years ago. We need to educate doctors. I think we could even train diabetes educators to know about this.

There’s a lot of potential here, but we need to train the right people to understand the right parts of the data. We could do a lot.

Q: As usual, Dr. Peters – a mountain of thanks to you for all you do for diaTribe and The diaTribe Foundation. So much learning – we’re so grateful.