Current data support that the CYP2D6 genotype is a predictor of response to tamoxifen. These data include outcomes from over 250 patients [1,2] and studies demonstrating that CYP2D6 polymorphisms strongly affect the concentration of active tamoxifen metabolites [3,4]. Although these results are convincing, the numbers are small and more data is necessary before we can recommend testing for premenopausal and perimenopausal women.

Several studies have reported on CYP2D6 genotype and outcomes in tamoxifen-treated hormone positive breast cancer:

Outcome Studies
Additional Studies
Conclusions

Outcomes Studies

A retrospective analysis of a prospective trial by the Pharmacogenetics Research Network (NIH/NIGMS): Consortium on Breast Cancer Pharmacogenomics published in 2005 reported on relapse-free survival, disease-free survival, and overall survival. Patient data was taken from the North Central Cancer Treatment Group adjuvant breast cancer trial 89-30-52 which evaluated the benefit of adding 1 year of fluoxymesterone to 5 years of tamoxifen. CYP2D6 *4 and *6 genotypes were evaluated in 223 patients using paraffin samples and in 17 women using buccal swabs. The concordance between buccal and tumor genotype was 100%. Women with the CYP2D6 *4/*4 genotype (i.e., poor metabolizers) had a shorter relapse-free survival and disease-free survival and a trend toward a worse overall survival. In a multivariate analysis accounting for nodal status and tumor size, patients with CYP2D6 *4/*4 genotype had a worse relapse-free survival and disease-free survival, but this trend did not reach statistical significance. This study lacked concomitant medication history; therefore the effect of CYP2D6 inhibitors on outcomes could not be assessed and likely would have enhanced the outcome.

A follow-up study was presented at The Clinical Science Symposium at ASCO in 2006. Using the NCCTG 89-30-52 data set, a retrospective evaluation of CYP2D6 *4 genotype and CYP2D6 inhibitors on relapse-free survival in patients receiving adjuvant tamoxifen was performed. The following CYP2D6 inhibitors were evaluated: fluoxetine, paroxetine, sertraline, cimetidine, amiodarone, doxepin, tilcopidine, and haloperidol.

Poor metabolizers were defined as CYP2D6 *4/*4 genotype, wild-type/wild-type (normal) receiving a strong CYP2D6 inhibitor, or wild-type/*4 receiving any inhibitor. Intermediate metabolizers (IMs) were defined as CYP2D6 wild-type/*4 and not receiving any CYP2D6 inhibitor or wild-type/wild-type receiving any CYP2D6 inhibitor. Extensive metabolizers (EMs) were defined as wild-type/wild-type and not receiving any CYP2D6 inhibitors. Of 256 eligible patients, 190 were genotyped and medication history taken from 171 patients. All patients were hormone positive. Three patients were on potent inhibitors and 10 on moderate inhibitors (median duration 2-3 years). One hundred and fifteen patients were EMs, 40 were IMs, and 16 were PMs.

Poor metabolizers had a shorter TTR (time to regression) at 10 years' follow-up (p = 0.019) and a shorter relapse-free survival (p=0.009). Two-year relapse-free survival was 98% for EMs, 92% for IMs, and 68% for PMs. There was a similar trend for overall survival but the p value was not significant. Multivariate analysis evaluating size, grade, and nodal status showed that PMs had a statistically significant shorter TTR and disease-free survival compared to IM and EMs. Poor metabolizers and IMs had an immediate peak in recurrence at year 2, whereas EMs had a later peak in recurrence at 4 years. This study demonstrated that CYP2D6 status was an independent predictor of clinical outcome.

Two other studies have not shown that CYP2D6 genotypes predict outcomes in tamoxifen-treated patients.[5,6] Nowell et al. demonstrated that CYP2D6 genotype had no effect on outcome, however the sample size was small (n=162) and did not include medication history, which likely would have contributed to a lack of an effect. In addition, the incidence of the most common CYP2D6 polymorphism (*4/*4) was low and may have not been sufficient to show an effect.

Wegman et al. also did not demonstrate a CYP2D6 effect on outcome in tamoxifen-treated patients. However, in their analysis, hormone-positive wild-type carriers did not derive any benefit from tamoxifen, suggesting a possible weakness in the study. In addition, they did not directly compare response between wild-type and polymorphic carriers and included PMs and IMs in one cohort. They showed that PMs and IMs derived some benefit from 40 mg/day of tamoxifen versus no tamoxifen, suggesting that some tamoxifen was metabolized by the IMs.

Additional Studies

Other studies have demonstrated that CYP2D6 polymorphisms have a significant effect on the concentration of active tamoxifen metabolites, and poor metabolizers produce very little active tamoxifen metabolites compared to wild-type carriers. This data strongly supports the outcomes data that demonstrate the lack of benefit of tamoxifen in patients who carry CYP2D6 polymorphisms [3,4].

Conclusions

The data are in favor of CYP2D6 genotypes as a predictor of response to tamoxifen, but additional studies are necessary in order to recommend testing to all breast cancer patients contemplating tamoxifen. Testing for postmenopausal women is appropriate as AI therapy has been shown to be better than tamoxifen alone for 5 years and is currently being compared to the sequential use of tamoxifen and an AI. Five years of an AI is currently a standard treatment option. Because of the potential harm incurred to patients by increasing breast cancer recurrence risk if they carry a functional CYP2D6 polymorphism, offering CYP2D6 testing to postmenopausal women is responsible and appropriate.