CYP2D6-polymorphism and effect of adjuvant tamoxifen in breast cancer patients

Abstract

Adjuvant tamoxifen at the standard dose of 20 mg daily for five to ten years reduces the risk for relapse and mortality in hormone sensitive breast cancer. The effect however varies and no early marker of poor response is yet available. Varying activation of tamoxifen due to polymorphism in the CYP2D6 gene has been suggested to influence the effect of the treatment, but data are inconsistent. Our previous study in a smaller cohort of tamoxifen treated early breast cancer diagnosed 1998-2000 indicated a poorer prognosis in premenopausal patients with reduced CYP2D6 activity.

The overall aim of this thesis was to investigate various aspects of tamoxifen treatment to facilitate improved personalized endocrine treatment strategies in early breast cancer, with individualized tamoxifen dosing, to improve quality of life, adherence and prognosis.

In study I we investigated the correlation between CYP2D6 genotype and tamoxifen metabolite levels in plasma, focusing on reduced function CYP2D6 variants (n=118). We also explored the relationship between endoxifen levels and adverse effects to tamoxifen. The degree of side effects to tamoxifen appeared to be dependent on endoxifen concentration. We found a distinct correlation between CYP2D6 activity and plasma concentrations of endoxifen. The effect of reduced function variants, in particular CYP2D6*41, on endoxifen formation was greater than anticipated. Markedly reduced endoxifen concentrations were seen in all homozygous carriers of CYP2D6 no function variants and in those with two reduced activity alleles. The fraction of patients with poor tamoxifen activation might thus be larger than expected. This may be important information for future genotype-based tamoxifen dosing. Although the clinical relevance of the proposed target level of endoxifen at around 5.9 ng/mL needs to be evaluated, it is concerning that a third of our study patients had endoxifen concentrations below this level with tamoxifen at the current standard dose. This underlines the importance of further work to define a target concentration of endoxifen for clinical benefit.

In study II we investigated the effect of CYP2D6 activity and other systemic adjuvant therapy on mammographic density (MD) change (n=699) in tamoxifen treated patients. As expected, MD declined during follow up, with a more prominent decrease in the premenopausal subgroup. Other systemic adjuvant treatment did not further extend density decline in this tamoxifen treated cohort. Density reduction appeared to persist after tamoxifen was stopped. Importantly, the previously proposed correlation between CYP2D6 activity and density change in patients with adjuvant tamoxifen could not be confirmed in this cohort with modern complex systemic adjuvant treatment. More data is needed to ascertain whether mammographic density change may be used as a marker of the desired effect of adjuvant tamoxifen.

In study III we compared information from patient records to data from the National Prescribed Drug Register in Sweden on adherence to adjuvant endocrine treatment (n=1235). We also investigated the association between CYP2D6-activity, menopausal status, the patients’ risk for relapse and adherence. Consistency, i.e. agreement, between the two sources of adherence data was good, 86%, when including medication with an aromatase inhibitor (AI) after tamoxifen. In those with at least 4.5 years follow up, adherence to adjuvant tamoxifen was reasonable, 72% and increased to 82%, when including subsequent AIs, based on prescription refill data. Adherence was not found to vary by menopausal status or recurrence risk. Unexpectedly, adherence to tamoxifen was lower in CYP2D6 poor metabolizers, despite data proposing a reduced risk of adverse effects in this group.

In study IV we aimed to validate our previous findings in a larger material in a more modern setting (n=1105), with tamoxifen treated patients operated between 2006-2014, who could also be subject to improved multimodal adjuvant therapy compared to the patients in our older study and to determine if the effect of CYP2D6 genotype is affected by menopausal status. Compared with our previous study, fewer patients, 12%, had a relapse and only 4% died from breast cancer under the 11-year follow-up. In summary, no obvious correlation between poor CYP2D6 activity and a worse prognosis was found in this material, accounting for adherence to tamoxifen and CYP2D6 inhibitors. A correlation between low CYP2D6 activity and a poorer prognosis in premenopausal tamoxifen treated early breast cancer was thus not confirmed. Breast cancer treatment has steadily improved over time. A possible negative effect of poor CYP2D6 activity on clinical outcome in tamoxifen treated patients is therefore likely marginal in a clinical setting with access to multimodal postoperative breast cancer treatment. Although our results do not support CYP2D6 testing for patients with adjuvant tamoxifen in a multimodal clinical setting, we cannot exclude that CYP2D6 genotyping might still be of value in selected groups, such as in in a low resource setting, where many patients, including those at higher risk of relapse, receive tamoxifen monotherapy. Therapeutic drug monitoring of tamoxifen to secure sufficient plasma levels of endoxifen for clinical efficacy and to avoid excess drug exposure associated with severe side effects might also be relevant in the future.

In conclusion, this thesis contributes to the knowledge on CYP2D6 polymorphism and the effect of postoperative tamoxifen in a multimodal setting, the correlation between CYP2D6 genotype and tamoxifen metabolites, which is important for future dose titration studies of tamoxifen, the effect of systemic adjuvant treatment on density change in tamoxifen treated patients as well as adherence to adjuvant endocrine treatment, with focus on tamoxifen.

There is a need for improved management of side effects to tamoxifen treatment, to optimize quality of life and adherence. Therapeutic drug monitoring of tamoxifen might be an approach. More work on predictive markers and early evaluation of response to tamoxifen is warranted.