Using Family History to Estimate Health RisksFeb 21, 2014
Geneticist Dr. Lisa Cannon-Albright explains how knowing your genealogy can give you a head start in understanding your risk for developing disease. She also describes her work building a genealogy database incorporating Veterans Administration records, and how it may help identify soldiers more likely to suffer from post-traumatic stress disorder, substance abuse, and other complex disorders.
Announcer: Examining the latest research and telling you about the greatest breakthroughs. The Science and Research Show, is on the scope.
Interviewer: Geneticist Dr. Lisa Cannon-Albright, is perhaps best known for finding that the genes BRCA1 and BRCA2 cause breast cancer. Now she's taking a step back and developing tools that can be used by anyone to evaluate their risk for common diseases. Dr. Cannon-Albright, what motivates you to take on this project?
Dr. Cannon-Albright: We've been very successful. We identified BRCA1, and BRCA2, and P16. P16 is a melanoma gene. Those genes don't explain all familial cancers of those types by any means. So to me, the question still arises, what about all the other people who are at risk for these cancers and other diseases, but we haven't found their gene yet that explains their family.
People say someday you'll go to the mall, and give them a drop of blood, and get your profile. I think that's a long way off. So what I see is, let's use family history, which is one of the best indicators of risk for many, many common disorders. Use that to classify people according to risk. So number one, we can do it today. Number two, it doesn't cost you anything. It's your family history. You might have to call your brother and ask him about your mother, etc., but it's available to everybody.
Interviewer: Don't physicians take family histories as it is?
Dr. Cannon-Albright: If you look at the literature, very often people who say they're studying family history have a simple question. "Do you have a family history of colon cancer?" That's not the way to take a family history.
The better studies maybe say, "Do you have any first-degree relatives affected with colon cancer? And what we've discovered, it's not just your first-degree relatives. Your second-degree relatives can affect your risk estimate, and even, to a lesser degree, your third-degree relatives can affect it.
Interviewer: So this is what you're developing, some sort of tool for keeping track?
Dr. Cannon-Albright: Well, you could call the paper a tool. It has quick, easy tables for clinicians or even patients to look at. And we've started on the breast cancer and lung cancer analysis.
Interviewer: Have you done proof of concept? I mean, do you know that this works?
Dr. Cannon-Albright: We're so lucky in Utah, because we have this genealogy, and this it's linked to a tumor registry. And the tumor registry goes back to the early '60s and later. So we can study all the cases of colon cancer, people who have genealogy, which is basically more than half the people in Utah. And so we use that to create our models.
We went back to create our models. We went back in time in the database and just locked the data 20 years ago, and said, "Okay, here's what all these people's risk of colon cancer is today." And then we went forward 20 years in time in the database and said, "So did the people whom we thought would get colon cancer actually get colon cancer?"
It's a population-based tool. It's not going to do that, but it is going to find the people who are at the highest risk, and that's a very small portion of the population. And it can tell those people the correct screening to get. You know, it saves the healthcare system and individuals money, and it saves people time, and it makes sure the people at highest risk get screened with the highest dedication to finding cancer.
Interviewer: I suppose the caveats are not everybody knows what their third-degree relatives health was.
Dr. Cannon-Albright: Yes, lots of caveats. Because first of all, probably only 20-30% of cancers are familial. So an only child of a small family, or people who were adopted, or for whatever reason separated from their family's medical knowledge, wouldn't have that opportunity. You know, but the nice thing is there are normal screening guidelines. So everybody would automatically fall to the baseline, which is, if you need to be screened for colorectal cancer, it starts at age 50.
Interviewer: So what would you recommend to people who might be interested in knowing what their risks are for different diseases?
Dr. Cannon-Albright: I actually got genotyped by 23andMe, which is one of these one of many companies that will give you your kind of personalized genetic picture.
Dr. Cannon-Albright: And I helped them out with a project. They were collecting some data, and so my colleague and I, they just said, "Oh, you can have the test, gee. Just spit in this tube." And so we did it, and we both just kind of laughed at it, because it tells me that I'm at low risk for several pretty serious diseases, breast cancer and stroke, that I've already had! So, hello!
But that's been my feeling all along. We don't have a handle yet on just simply using genetic markers to try to assess risk. It's way too early in the game. I'd say the most important thing would be keep an accurate genealogy and keep an accurate medical history for your family, as far out as you can go. And make sure that your children recognize that that's a part of your legacy, and it stays in a safe, and they're going to take over when you're not here, and in my mind, it's really your duty to carry this around in your head and act on it appropriately. So talk to your doctor about it and be proactive about it.
Interviewer: What is your vision for the future?
Dr. Cannon-Albright: My view is someday a national resource. I mean, how wonderful would it be if you could go to a resource and find yourself, your family, maybe add some genealogy, or add in your medical data, and we'll be able to track and estimate risk. So that's kind of my view, and I'm sort of working on something like that with the VA right now. It's a genealogy of the United States. It's already got 38 million individuals in it. We're going to link it to the VHA patient population. We tried to link 11 million vets, and we linked about 5% of them, half-a-million. But if you have a half-a-million people with medical data that you've linked to a genealogy, already imagine how powerful it is.
Interviewer: So you have this database. So I guess I'm a little unclear on how you intend to use it.
Dr. Cannon-Albright: So for instance, the VA population has some phenotypes that are pretty uncommon in the rest of the population. If I wanted to study Gulf War illness, or post-traumatic stress syndrome, I'd have a really hard time finding high-risk pedigrees. Because that environmental exposure of having been to the gulf or of having had some traumatic event is so rare that you could study giant pedigrees and not have very many events.
So using this resource, I can do it the opposite way. So I find all the veterans without identifying anybody. Just using ID numbers, we can find all the veterans who have, for instance, a diagnosis of Gulf War illness, and them link them to the genealogy and see what clusters we find, clusters meaning pedigrees, sets of related people who have Gulf War illness.
And now, again, we've developed tools to test whether that pedigree has more cases of Gulf War illness among the veterans in that family than you would have expected. And so maybe you take that into account when you decided who you're going to deploy to the Gulf War, or who you're going to put in combat versus not.
And then, hopefully like I say, one day, when we study those families, if we actually find the variance that is responsible, then maybe we'll understand what's happening, and possibly we could develop treatments that would reduce the morbidity.
Interviewer: Could you even use those risk models to find people who might be more likely to commit suicide, or undergo drug abuse, or something like that, which I would think is higher in that population?
Dr. Cannon-Albright: Absolutely. And again, I have this terrible bias of believing that everything has a genetic predisposition. And I'm only calling it a bias to be polite, because in my head, it's the truth. But yeah, so there is no phenotype that I think should be ignored, especially, like you say, pretty significant things like suicide and harmful addictions. Wow, if we could find out there was a predisposition, and we could actually do something about it, how powerful would that be?
Announcer: Interesting. Informative. And all in the name of better health. This is The Scope Health Sciences Radio."