Medscape recently produced a fascinating interview with Dr. Francis Collins, Director of the National Intitutes of Health that focuses on genomics but ranges broadly over several areas, including medical education and preventative medicine. It's not that long and there is also a text format if you don't want to watch the interview. Highly recommended. This a thought-provoking interview.
I think Dr. Collins is overly optimistic about the impact of genomics on day-to-day patient care. Pharmacogenomics, for one example he uses, has not made any impact thus far in my experience. While it is a tool, perhaps, for demonstrating why a patient may not be responding to a particular drug, the use of pharmacogenomic testing before initiating treatment has not been prospectively studied to my knowledge in any trials, for any drug. This leaves aside the further studies regarding cost-benefit analyses as well as the issue of payment/reimbursement. Right now, it seems cheaper and tried to put a patient on a standard dose and see what happens.
Another example is the idea of genomic sequencing of cancers. It sounds good because it provides a "personalized" profile of a cancer. What about tumor heterogeneity? Do we sequence the primary tumor or metastasis? Which part of the primary tumor? Which metastasis? Finally, how does one compare "unique" tumor genotypes and develop groups to statistically compare treatment responses? At some point, patients have to be lumped together (and thus muddying up personalized differences) in order to find appropriate guidelines for treatment and provide some prognostic information.
Lots more to think about!