Tamoxifen: Will Less Equal More in Women with Precancerous Breast Disease?

In this issue of Clinical Cancer Research, DeCensi and colleagues provide exploratory analyses of factors predicting success of 3 years of low-dose (5 mg) tamoxifen versus placebo to reduce subsequent ductal or lobular carcinoma in situ (DCIS, LCIS) or invasive cancer in women with a prior biopsy exhibiting intraepithelial neoplasia (IEN; ref. 1). At entry, 70% of the participants had a diagnosis of DCIS, 10% LCIS, and 20% a prior biopsy showing atypical hyperplasia. Approximately 40% of trial entrants were premenopausal and 60% postmenopausal. Precancerous lesions from all women entered were either estrogen and/or progesterone receptor positive (ER⁺ and/or PR⁺) or ER unknown (2). Areas of atypical hyperplasia and LCIS were excised whereas DCIS lesions were treated with excision ± radiation.

At a median follow-up of 5 years, women randomized to low-dose tamoxifen had an approximately 50% reduction in a new DCIS or invasive cancer compared with placebo with no increase in uterine cancer or thromboembolic events. There was an increase in investigator reported hot flashes from baseline, but they were for the most part mild, and quality-of-life indices were no different between the two randomized groups (2).

A level of risk reduction similar to what had historically been observed with full (20 mg) dose tamoxifen with a favorable side effect profile led to the 2019 ASCO Breast Cancer Risk Reduction Update statement that low-dose tamoxifen could be considered as an alternative to full dose in women with IEN (3). Although results in women with IEN were encouraging, there was not sufficient subgroup analysis in the original report to determine the benefit of dose tamoxifen by menopausal status and body mass index (BMI) category. Because they were not included as part of the eligible trial cohort, it was also not possible to determine whether women with risk variables other than a precancerous biopsy would benefit from low-dose tamoxifen. Outcomes with tamoxifen treatment of invasive cancer is predicted by ER and PR expression as well as posttreatment initiation Ki-67 (4, 5). ER and PR are often overexpressed in IEN which may sensitize these lesions to tamoxifen even at lower doses. In the trial reported by DeCensi and colleagues, the median proportion of cells staining for ER and PR were 80% and 70%, respectively, with 10% staining for Ki67 in those lesions tested for these variables (2).

In the current analysis, the authors examined several clinical factors which might have impacted the efficacy of low-dose tamoxifen in their trial. Of these, baseline Ki-67 >10%, perimenopausal or postmenopausal status, low (<16 pg/mL) estradiol levels, and absence of smoking were associated with a more favorable outcome as assessed by time to progression/occurrence of DCIS or invasive cancer. BMI was not a factor (1). Ki-67 has been found to dramatically impact time to progression following a diagnosis of atypical hyperplasia or DCIS (6, 7). In the Mayo Clinic observational series, 14% of women with atypical hyperplasia and Ki-67 of ≥2% developed breast cancer within 10 years but only 3% of those with Ki-67 <2% did so in that time frame (6). Longer follow-up will likely be needed to determine if time to progression for the more slowly proliferating lesions (Ki-67 <10%) will be favorably impacted by low dose tamoxifen in the trial reported by DeCensi and colleagues.

Significant improvement in time to recurrence/progression was demonstrated with low-dose tamoxifen for perimenopausal and postmenopausal but not for premenopausal women with IEN. Tamoxifen-induced increases in gonadotrophin secretion result in augmented ovarian estrogen production in premenopausal women with good ovarian reserve. Increases in systemic estradiol levels of up to 4–10× normal mid-cycle peaks have been documented in premenopausal women with full-dose tamoxifen (8). A total of 5 mg of tamoxifen may not be able to reliably and effectively compete with high levels of endogenous estradiol in premenopausal women. However, a Swiss trial suggests as little as 2.5 mg of tamoxifen may be associated with reduction in mammographic density in premenopausal women which in turn has been associated with reduced breast cancer incidence (9). It is possible that given the relatively limited number of premenopausal women and modest follow-up time that a difference between low-dose tamoxifen and placebo in time to diagnosis of new DCIS or invasive cancer will eventually emerge.

Levels of active tamoxifen metabolites achieved may be particularly important for premenopausal women taking low-dose tamoxifen but were not measured in this trial. Levels of metabolites may in turn be affected by CYP2D6 polymorphisms and/or concomitant use of strong CYP2D6 inhibitors. A prior publication by this group in which metabolites were measured at several dose levels, reported that, at 4 weeks, 8% of women taking the equivalent of 5 mg/day had no detectable tamoxifen metabolites (10). This raises the question of whether active metabolite measurement 2–3 months after low-dose tamoxifen initiation should be considered. Unfortunately, neither threshold nor optimum metabolite blood levels have been unequivocally established although >3.26 nmol/L 4-OH and >9 nmol/L endoxifen concentrations have been suggested (11). No clear relationship was observed between endoxifen levels and breast cancer recurrence in the CYPTAM prospective trial (12). However, most participants switched to an aromatase inhibitor after 2–3 years of tamoxifen making this observation difficult to interpret (12). Without further research it is difficult to use active tamoxifen metabolite levels to guide decisions about low-dose tamoxifen in individual patients, but it might be best to avoid low-dose tamoxifen in women who are known poor CYP2D6 metabolizers.

Obesity impacts more than 40% of U.S. adult women. However, BMI but did not seem to be a factor in the current trial nor did obesity preclude full-dose tamoxifen effectiveness in the NSABP B-14 adjuvant trial although overall survival was worse than for lean women (13).

Good research stimulates more research. We can anticipate future studies comparing compliance, imaging, microbiome, and breast tissue biomarker change in individuals receiving low versus standard doses. Tamoxifen enhances NFκB and breast stem cell activity through a variety of molecular mechanisms which can result in tamoxifen resistance (14, 15). Studies in which low-dose tamoxifen is paired with other well-tolerated agents targeting processes associated with endocrine therapy resistance should be expected. However, at present only a handful of studies with low-dose tamoxifen are reported in clinical trials.gov.

Despite the lingering clinical questions, the favorable side effect profile and demonstrated efficacy of low-dose tamoxifen is likely to lead to more uptake of primary prevention therapy. Should we also offer the low-dose alternative to high-risk premenopausal women and/or women without a precancerous lesion? Since the original report I have primarily offered low-dose tamoxifen to nonobese perimenopausal and postmenopausal women with IEN. I also offer low-dose tamoxifen to women at increased risk of breast cancer but without IEN who decline full dose tamoxifen if they are at least 45 years of age and have signs of early menopause transition. On the basis of the analysis presented in this issue by DeCensi and colleagues, it is also reasonable to extend low-dose tamoxifen to women with IEN and mild obesity. Anecdotally, uptake, compliance, and freedom from bothersome vasomotor symptoms seems better with low dose than for standard dose tamoxifen. In the absence of further research, low-dose tamoxifen should not be considered an alternative prevention treatment for young premenopausal women with substantial ovarian reserve or as endocrine treatment for invasive disease.

No disclosures were reported.