FDA Clinical Pharmacology Advisory Committee Meeting, October 18th DNA.


Statement of David A Flockhart MD, PhD

Professor of Medicine, Genetics and Pharmacology

Indiana University School of Medicine


Thank you for the opportunity to address the committee. I would like to highlight the implications of the data on tamoxifen pharmacogenomics that our group and others have generated over the last 8 years. Our work to date has been presented within the formal presentations so I will instead focus on the implications of these data for patients with breast cancer, for the field of clinical pharmacogenetics in general, and for the utility of CYP2D6 testing in clinical practice. While these implications are not the direct subject of this deliberation, I feel strongly that some appeciation of them is important in order that the committee be fully informed as it deliberates on the questions before it.

Patients with Breast Cancer

First, for patients with breast cancer the pharmacogenetic testing in this context offers a means of making a drug selection, not a means of changing a drug dose. This is different from what the committee has considered before, and I would submit to you that this makes it an easier decision. Postmenopausal women who are the majority of patients with breast cancer in this country have available to them a number of possible choices for the endocrine treatment of breast cancer. The choices would include tamoxifen or any of the three available aromatase inhibitors, letrozole (Femara™),  anastrozole(Arimidex™) or exemestane (Aromasin ™) . It follows that the risks inherent in a pharmacogenetically – guided approach are less, since whatever the eventual treatment, it would be with what is currently an FDA-approved drug and regimen. Since this test involves change of drug, rather than a dose effect, this test may be useful for organizations such as Medicare and pharmacy-benefit management companies who wish to use new tools to reduce the cost of care by using effective generic medications such as tamoxifen in place of more expensive drugs that are still on patent like the aromatase inhibitors.

Secondly, this testing would be aimed at improving efficacy, not at reducing toxicity. The committee has considered the cases of TPMT genotype as a predictor of drug toxicity in childhood leukemia, of the UGT1A1 test as a predictor of the toxicity of irinotecan in colon cancer and most recently, of CYP2C9 and VKOR testing to predict the toxicity of warfarin. All of these tests would result in a change in dose aimed at reducing adverse effects. These precedents are all important, but this is the first situation where we are examining a pharmacogenetic test with the potential to sustain and improve the efficacy of an important and life-saving therapy by either not offering tamoxifen to patients who are genetic CYP2D6 poor metabolizers, or of specifically using it in patients who have rapid or ultrarapid CYP2D6 genotypes, for whom tamoxifen might be superior to AIs. This last point requires further research, but the possibility that there may be postmenopausal women who would be more appropriately treated with tamoxifen than an aromatase inhibitor is real.

The Field of Personalized Medicine.

This testing has implications for the broad field of personalized medicine, but it is notable that breast cancer has arguably been more personalized that any other therapeutic area. The use of estrogen and progesterone receptor testing as markers of drug efficacy, of HER-2 as a marker of Herceptin’s effects and of the use of panels of genes such as the Oncotype DX™ to predict response to chemotherapy mark this field as one that has pioneered personalized approaches to the treatment of our patients. That said, there is no test of any kind based on germline DNA that is widely used in the treatment of breast cancer and this would be the first. These data highlight the importance of integrating germline DNA polymorphisms into the armamentarium of predictors of drug response to treatment.

Cytochrome P450 2D6

The activity of this enzyme has arguably been studied with more diligence than any other genetically polymorphic phenotype related to drug therapy. As of October 2006, there are more than 2,800 papers describing various aspects of its genetics and substrate specificity in humans since it discovery in Europe by the groups of Michel Eichelbaum and Robert Smith in the late 1970s. The enzyme metabolizes a large number of clinically important drugs, including most beta-blockers and antidepressants. To date there are a number of drug labels that describe pharmacokinetic changes resulting from this polymorphism for drugs such as the tricyclic antidepressants, atomoxetine and codeine, but there are none where a CYP2D6 test is actually recommended for any population or subpopulation. Although it is arguable that there are large pharmacokinetic effects of the CYP2D6 polymorphism, this is the first situation where the data indicate an important pharmacodynamic outcome has been altered. Intensive research on the pharmacodynamic outcomes of treatment with other CYP2D6 substrates is being carried out and the availability of these tests will simplify and enhance the transit of this research into clinical practice.