In Defense of Pharmacogenetics

Author: Nils Reinton
Furst Medical Laboratory, Søren Bullsv. 25, N-1051 Oslo, Norway (nreinton_at_furst.no)

Pharmacogenetics is the analysis of specific genetic markers informing you of how efficiently you metabolize a given drug. When it comes to metabolism of drugs used for treating psychiatric disorders (ranging from mild depression to severe psychosis), three genes are commonly analyzed: Cyp2C9, Cyp2C19 and Cyp2D6. Of course, other things than genetics also influence how you respond to a drug, like compliance, diet and smoking, but individual genetic variation have profound stand-alone effects. If your genetic test shows the presence of clinically relevant genetic variants you will be grouped as a Poor Metabolizer (higher risk for unwanted side effects) or Ultrarapid Metabolizer (not responding to medication or in need of very high dosage). For a patient experiencing adverse events or no response at all when taking his medicine, pharmacogenetics can be a tremendous help in choosing more fitting medications (change dosage or use another (sub)class of drugs). Pharmacogenetics is thus, a vital part of personalized medicine.

Recently a commentary - "A Case Study of Personalized Medicine" by Katsanis et al., was published in Science (subscription may be required), where the use of pharmacogenetics before choosing upon medication was discouraged. The reason for this was the finding that doctors failed to follow up on recommendations based on test-results. I will try to argue against such a negative approach towards pharmacogenetics and attempt to show that pharmacogenetis is in fact a valuable tool, especially for patients on psychoactive drugs who experience adverse events or lack of effect.

Claim 1: Doctors fail to follow medication recommendations based on pharmacogenetic lab-results.

Counter argument: Physicians either not understanding the test-results or ignoring lab results (due to personal conviction of some sort, be it based on clinical experience or not) reflects a general problem not restricted to pharmacogenetics. Health professionals lack of understanding is especially common for genetic tests. One could argue then, that tests that are too difficult for doctors to interpret should not be made available. But, one could easily counter argue that such an approach would be a major obstacle to medical progress. New tests will always need a time-window of learning and enlightenment. To narrow this window, the following message needs to driven home:

A patient that has to try many different drugs to achieve the desired effect without adverse reactions, is mistreated in a costly manner. Pharmacogenetics constitutes a once in a life-time test allowing a targeted approach straight towards the medication most likely to be suitable. When recommendations based on lab-results are implemented in the treatment regime, pharmacogenetic testing is cost effective and in the best interest of the patient.

Claim 2: Physicians seem to be unable to pick the right patients, and pharmacogenetic testing becomes an inefficient (as well as over hyped and expensive) general screening method, also lacking subsequent changes in medication-therapy. That the frequency of mutant alleles in the normal population does not differ from that in a given patient population, supports this notion.

Counter argument: The overall frequency in a population is irrelevant to the patient, as finding his individual response to the drug is what matters. This counter argument however, is an argument favoring screening in any given field of medicine and one may still claim that the overall cost-effectiveness is insufficient. But, in our laboratory, we do not normally get requests for pharmacogenetic testing prior to medication. Physicians using our service send samples from patients already on medication. They have come to their doctor with either adverse events or a lack of clinical effect. In our opinion these are the right patients to test. We have no reason to believe that the physicians resist changing to other medications based on our lab results when the case presented to them is as defined as this. Thus, pharmacogenetics in our hands, is targeted diagnostics rather than general screening. In addition, we have done a small study (a panel of 12 SNP's and 2D6 copynumber variation on 595 patients) to show that the allele frequency in our patient population is either at the high end of normal variation, or above that seen in a normal population (see figure, European country codes in parentheses):


Similar results (see figure) are obtained when looking at frequency of poor and ultra rapid metabolizers (PM and UM respectively).


Consequently, it seems that physicians are able to pick the right patients. And, our results support the notion that pharmacogenetic testing constitutes targeted diagnostics in the best interest of the patient.

Reinton, N.In Defense of Pharmacogenetics. Sciphu.com