Q: I was wondering if you’ve used genetic testing to help choose an antidepressant. I have one patient who is requesting it but hasn’t actually “failed” any antidepressant trials. She doesn’t like the idea of waiting several weeks just to find out something may not work. Cost is not an issue for her and frankly even if I don’t order it I think she’ll probably do it anyways. My understanding is that those tests are really more about how quickly you metabolize drugs which isn’t really the same thing efficacy.How would you go about counseling someone on this?
PL: These tests are not useful for efficacy, despite claims otherwise. Understanding first that “MDD” is a scientifically invalid diagnosis, it doesn’t make sense to look for genes of nonresponse for a scientifically false diagnosis. Obviously, response is low to monoamine agonists (“antidepressants”) in MDD, but this has to do with the diagnosis mixing together many different types of depression, and also with the fact that the drugs are purely symptomatic, not disease-modifying, in their effects. Any symptomatic benefit is thus short-term, and relapse occurs later in many people. This is all called “non-response”, and limited genetic data on short-term symptomatic benefit is then used with this testing to claim prediction of benefit. In the long run, there is no prediction at all, and short term, there would be better prediction by distinguishing depressive phenotypes (pure, mixed, melancholic) as well as diagnostic course (often misdiagosed bipolar illness is present).
In contrast to all this poor utility for efficacy, pharmacogenetic testing can be useful to assess side effects in particular based on pharmacokinetic results for rapid or poor metabolizers. Such testing usually is not worth the money, unless there is high suspicion of rapid (no drug effects at all) or poor (lots of side effects with all drugs) metabolism.