When dealing with depression, many patients and doctors grope for months to find the best medication fit. There is potential that advances in genetic analysis will soon allow us to predict an individual's reaction to a drug. This hope is at the heart of a burgeoning new medical field--pharmacogenomics which seeks to fit the right drug to the right person so that adverse effects can be avoided and positive effects fully exploited.
In the field of mental health, several drugs have in recent years been stuck with blackbox suicide warnings because a small percentage of patients have been found to have adverse reactions. There is emerging evidence that the suicidal ideation that occurs in a minority of patients could be genetically predicted. Dr. Roy Perlis published a study in June 2007 suggesting a link between variation in the CREB1 gene and suicidal thoughts among male patients. If the at-risk patients can be identified, doctors can prescribe antidepressants to those who need them without fear of terrible consequences.
The idea that genetic differences can put some patients at risk may seem more intuitive than the idea that there will be differing levels of benefit from a drug, but a detailed examination of depression treatment may elucidate the matter. A large percentage of patients will not respond to the first depression drug they are given, sometimes taking several tries before the right prescription can be found. Most antidepressants also take 4-6 weeks before results begin to manifest (or fail to), making this selection process a time-consuming and difficult one. So it is heartening news that researchers have recently been able to finger specific genetic variants as the potential culprits for the failure rate.
The hope is that eventually doctors will be able to take a DNA sample and then adjust their prescription for all known complicating factors. Once the process of gene sequencing becomes cheaper, or new methods of identifying genetic factors emerge, finding the appropriate depression treatment medication should not have to involve so much trial-and-error.