Medicine and Science

I feel like I gathered a number of interesting items for thought today.

Our lecturer during the first half of class engaged us in a discussion about a published study on the effects of a egfr inhibitor in treating a certain type of cancer. Although the sample size was small, it was compelling in that virtually everyone fitting a certain criteria was cured by the treatment. Then he posed the question whether, if we had 2 untreated patients come in who fit the criteria perfectly, would we treat them too? (Otherwise, the prognosis is 1 year; plus, the drug has virtually no side effects.) From the preliminary results, we feel personally certain that the treatment works-- isn't that what doctors do? Try to make decisions that they believe are for the best interest of their patients? On the other hand, scientifically, we would want to do randomized controls and withhold treatment from one population to prove efficacy first. This means those two specific people may or may not benefit. We are at odds between doing the best for a population overall versus doing the best for individual patients. I had never thought of a dilemma such as this before..

The second half of class, we had a lecture on genome-wide associations with disease. Those of you in these fields probably can guess that Dr. David Altschuler gave the lecture. In short, he is interested in understanding how genetics are linked to common disease (perhaps think of it in relation to why each person looks just a little bit different due to just due to the haplotypes scattered throughout the genome). These kinds of associations are tricky and elusive, and the millions of combinations considered means sample sizes must be in the thousands before changes such as 20% increased chance of diabetes can be shown with a significance of p<10E-10. I think there has been some resistance to his work, since in the reductionist point of view (which pretty much forms the foundation of science) is used to thinking about one gene one disease. Even with treatments such as combination therapy for HIV/AIDS, the targets are specific components chosen by circumstance rather than an understanding consideration of the entire disease pathway. Many feel like his methods and results are intractable.. what can we do with information about thousands of changes, each contributing a miniscule amount to a disease. Nonetheless, this brings me to my second interesting dilemma. There are many things that would be interesting to study, but considerations on the impact of results on career must be taken into account. It's amazing that Dr.Altschuler has found the means to study the things that most scientists consider too vague to tackle. For myself, I am really interested in the issue of robustness, how cells can be stable under environmental or just its own random fluctuations, and I've even found a pretty neat project that is related. In our recent poster session, one faculty member asked me where I wanted to go with my project. Truthfully, I said I didn't know. It's one thing to say, oh look at this cool phenomenon, maybe there's something that specifically makes the cell stay in balance... and entirely a different matter to prove to the scientific community that something that was already pretty good must critically have a factor that makes it just a tiny bit better.