Influence of endoxifen on mammographic density: results from the KARISMA-Tam trial

Monitoring metabolites of tamoxifen, such as endoxifen, has been suggested as a strategy to ascertain therapeutic effect of tamoxifen therapy, but clinical guidelines are missing. Herein, we aim to investigate the outcome of endoxifen concentrations of low-dose tamoxifen, using change in mammographic breast density as a proxy for therapy response.

In the randomized KARISMA-Tam (Karolinska Mammography project for risk prediction of breast cancer -Intervention Study with Tamoxifen) trial, including 5 doses of tamoxifen, measurements of plasma endoxifen concentrations, determination of CYP2D6 metabolizer status, and mammographic breast density change over the trial period were carried out. Association between endoxifen concentrations and relative mammographic breast density change after 6 months treatment was analyzed using linear regression in a spline model.

A total of 824 women (335 premenopausal, 489 postmenopausal) were included. In analyses of premenopausal women, a spline model described a mammographic breast density decrease, equivalent to the mean (-18.5%) seen in women exposed to 20 mg tamoxifen, at endoxifen concentrations of 2-3 ng/mL. The mammographic breast density decrease reached a nadir at endoxifen levels of 3 ng/mL and did not decrease further at higher endoxifen concentrations. Most intermediate and normal tamoxifen metabolizers (about 90% of all participants) reached an endoxifen concentration of more than 2 ng/mL at tamoxifen doses of 5 and 10 mg. No mammographic breast density decrease was seen in the postmenopausal group.

We have identified a possible window of effect on mammographic breast density at endoxifen concentrations of 2-3 ng/mL in premenopausal women, which corresponds to the doses of 5 and 10 mg tamoxifen. Because mammographic breast density change was used as a surrogate marker for therapy response, results should be confirmed using clinically established outcomes measures.