Ezra Klein, writing in his blog for the Washington Post, draws attention to the most recent story in the New York Times about the increasingly low cost of comprehensive genetic screening. Klein predicts that in the absence of an individual health care mandate, the access to personal health information via genetic testing will lead to a death spiral for the health care industry as currently constructed as only those with elevated risk of future significant illness will buy comprehensive health care plans.
At the moment, our understanding of the genome remains relatively crude, and our ability to predict future health risks based off of genomic sequencing is limited. But we’re getting better at it. For instance, women in families with a high rate of breast and ovarian cancer can have themselves tested for alterations in the BCRA1 and BCRA2 genes. If they test positive, it means their risk of developing breast or ovarian cancer is significantly higher.
Leaving aside the issue of the individual mandate, I think Klein is jumping the gun a bit on our genetic future. The example he uses, BRCA1 and BRCA2, are not representative of how most of the genes we have decent studies on behave.BRCA1 and BRCA2 are both genes that, in cases where there is a high likelihood of hereditary breast cancer (~5-10% of breast cancer cases), significantly increase the chances of a person developing breast cancer in their future. In other words, these are genes that, in the right conditions, have a large effect. But more than a decade of genome-wide association studies suggests that most genes don’t fall into this category. A 2008 review in Nature gives us the following illustration:
Most of what we have found in the genome thus far are lots and lots of areas of the genome that appear to have small, combinatorial effects on disease outcomes. These are not genes that will, by themselves, significantly elevate one’s risk for a specific outcome or that should change one’s behavior on an issue like long-term health coverage. As we move into the future our knowledge will no doubt improve, but that doesn’t mean we will find more large-effect genes, it will more likely be that we will find greater complexity associated with small-effect genes. Even in the case of BRCA1 and BRCA2, going back to Klein’s example, the elevated risk associated with them is only in cases where there is strong evidence for hereditary breast cancer given a patient’s familial history…in other words, other genetic variants present. This issue reflects a general tendency to underestimate complexity within the genome, a tendency enhanced by the nature of technology-facilitated genetic research that has gradually gone from only being able to identify those characters with the largest effects (i.e. Mendelian traits) to studies with finer resolution able to identify allelic variants with smaller and smaller effects. So the history of our knowledge of the genome is biased towards elements that are actually rare within the genome, while our knowledge of more common variants has lagged.
*****
1. Genome-wide association studies for complex traits: consensus, uncertainty and challenges. Mark I. McCarthy, Gonçalo R. Abecasis, Lon R. Cardon, David B. Goldstein, Julian Little, John P. A. Ioannidis & Joel N. Hirschhorn
Nature Reviews Genetics 9, 356-369 (May 2008). doi:10.1038/nrg2344
