A person's genetic ancestry can make a big difference in predicting his or her disease risk and potential course of treatment. GT's Christie Rizk recently spoke with UCSF's Esteban González Burchard about his recent study in the New England Journal of Medicine on how inaccurate racial identification affects lung function predictions in African-Americans, and the far-reaching implications of his study for medicine.
Genome Technology: Are patients' ethnic groups being appropriately used as predictors of lung function?
Esteban González Burchard: The point of the study was that self-identified race — or what happens oftentimes when you go to a clinic and some secretary or physician assigns a race to you — isn't accurate. That is less accurate than genetic ancestry testing. That's the clear black-and-white interpretation. In medicine, we use race as a reference standard for many diseases like kidney function or prostate cancer — lots of things. And here we show that these race standards are inaccurate in many ways. Part of the reason is that the demographics of the world are changing — now it's become much more acceptable to have mixed marriages. And it's not just African-Americans. It's most populations in the world. Indians from India are mixed, Asians from Asia are mixed, whites are mixed, and it makes a big difference.
GT: When you are developing a model of lung function and using genetic ancestry, how do you classify people of mixed ethnicity?
EGB: What we're saying is that we would compare you to yourself. We would get this categorical definition of race and take it down to your level — what's good for you as opposed to all people your age, height, race, and things like that. It brings us a step closer to personalized medicine. Ideally we'll have a time — and it's not too far off, maybe 10 or 15 years — where we would be able to take your blood and put it on a chip and say, "These are the drugs that will work good for you or bad for you, these are the clinical risk factors, and this is what you should be compared to."
GT: Would such a model work with current medical practice?
EGB: It's not ready for primetime. The genetic ancestry in lung function is — genetic ancestry testing has gotten cheap enough where we could do it — but risk prediction for disease traits is not ready for primetime. In fact, the FDA put a hold on all those companies that were selling risk-predictions directly to consumers. When you go to 23andMe or Navigenics — say they test you for the BRCA gene — you have to categorize yourself as black, white, or Asian. So even those companies use this gross categorical box, and the risk changes based upon which race you are. Basically we're deconstructing those boxes and saying, "This is what your risk would be based upon your own genetic ancestry." But we have a long way to go. We need to do more studies in racially diverse populations to be able to really home in on what true risks are. ... Even the FDA has to deal with this, and they're trying to get their head around how to deal with this. It goes way beyond lung function.
GT: Should doctors start doing ancestry tests as part of a patient's yearly physical or if they see a patient for the first time?
EGB: I wouldn't go that far, but I would be able to go as far as saying definitely for lung function. It makes a difference. We have arbitrary numbers for when to start oxygen therapy, who's going to pay for oxygen therapy — Medicare versus you — when to get a lung transplant. Those are all based on these lung function numbers, which are all based on race standards. So I can say confidently for lung function testing, yes, I would get genetically tested.