Tamoxifen is highly effective in treating estrogen receptor (ER) –positive breast cancer. Meta-analyses of adjuvant tamoxifen treatment for 5 years have shown reduced breast cancer recurrence at 10 years, with reduced breast cancer mortality observed at 15 years(event rate ratio, 0.70; 95% CI, 0.64 to 0.75). 1 Given thattamoxifenmetabolitesbindtothe ER more avidly than the parent drug, tamoxifen recipients are hypothesized to be divisible into a group of excellent metabolizers, who receive benefit with standard dosing, and poor metabolizers, who do not benefit. 2 Accordingly, the development of validated measures for monitoring tamoxifen effectiveness would have great value, particularly among premenopausal women, a group in which tamoxifen is first-line therapy. We highlight the potential relevance of assessing mammographic breast density as a marker of tamoxifen efficacy, particularly among premenopausal women in a clinical setting. Mammographic density is a radiologic measure of the fibroglandular content within the breast and is usually expressed as the percentage of total breast area composed of dense tissue. 3,4 Women with the highest levels of percent density (ie, . 75%) experience a four- to six-fold increased relative risk of developing breast cancer compared with women who have the lowest percentage mammographic density when adjusted for confounders. 3,4 Amassing data suggest that a decline in mammographic density could serve as a much-needed predictor of a favorable response to the selective ER modulator tamoxifen. The results of four retrospective analyses, which include one that evaluated breast cancer–specific mortality among patients with ER-positive disease treated with tamoxifen, indicated that women whose mammographic density declined in the unaffected breast after initiation of tamoxifen therapy had better outcomes (reduced risk of recurrence 5,6 or death from breast cancer 7,8 ). The magnitude of the observed benefit associated with a decline (approximately 10%) in breast density was remarkably similar across studies, despite differences in the methods of mammographic density assessment and analytic approaches, which suggests that these findings are robust. The results from studies that included premenopausalwomen are shown in Table 1 .Ofthe two prior studies that examined tamoxifen-related changes and breast cancer death, 7,8 one excluded premenopausal women, 7 for whom tamoxifen remains as the primary endocrine therapy. Assessment of mammographic density change after tamoxifen treatment, specifically among younger women, merits further study to maximize the potential translational benefit of monitoring mammographic breast density change as a prognostic indicator in the clinic. In our prior retrospective analysis of patients with ER-positive breast cancer (age range, 32 to 87 years) who were diagnosed at Kaiser Permanente Northwest, 8 we compared the change in mammographic density after 1 year of tamoxifen treatment among 97 patients who died of breast cancer with 252 matched patients who did not. Women who were in the highest tertile of mammographic density decline were less likely to die of breast cancer (odds ratio [OR], 0.44; 95% CI, 0.22 to 0.88). 8 To explore the potential utility of assessing mammographic density decline as a measure of effectiveness among premenopausal women who received adjuvant tamoxifen, we performed a subset analysis that examined recurrences and deaths among 136 women with ER-positive breast cancer diagnosed at age 55 years or younger (a surrogate for premenopausal status) who were included in the initial report. Among these women, there were 34 recurrences or deaths at 5 years of follow-up and 50 such events during the whole study period (1990 to 2010). Although this analysis is limited by statistical power, multivariate Cox proportional hazards regression models demonstrated that patients with $ 10% decline in mammographic density after 1 year of tamoxifen therapy had an onsignificant improvement in disease-free survival (DFS) at 5-year follow-up (hazard ratio [HR], 0.78; 95% CI, 0.38 to 1.61);
Read full abstract