To the Editor:

We have read with great interest the paper by Love et al [1] about the relationship between HER2/neu expression and response to adjuvant endocrine therapy in premenopausal women with breast cancer. Whereas HER2/neu and estrogen receptor (ER) are believed to be important cell survival and cell death factors in human breast cancer, if and how they interact to confer resistance to hormone therapy is still in debate. Several observations are consistent with a major role for c-erbB2 in the development of endocrine resistance, considering also the HER2/neu acquired expression during hormonal therapy, either by clone selection or by phenotype modification caused by tamoxifen [2,3]. However, the results reported by Love et al [1] suggest a discrepancy between the clinical and preclinical scenario. Although HER2/neu is confirmed to be an independent prognostic indicator of poor overall survival, surprisingly it seems to predict a better response to tamoxifen and ovarian ablation in ER-positive premenopausal early breast cancer patients. We are grateful to the authors for their pioneering observation, even though we think that some comments are needed.

The authors made a secondary analysis from previous data obtained prospectively in a randomized trial designed to investigate the impact of adjuvant endocrine manipulations compared with observation and delayed endocrine interventions at recurrence in premenopausal women with breast cancer, regardless of hormone receptor and HER2/neu status [4]. Of 709 original patients, only 282 are considered, and differently collected and outcome variables are compared in these subgroups [1]. Even though Love’s study lacks power [5], the P values, regarding HER2/neu overexpression and response to adjuvant oophorectomy and tamoxifen, are statistically significant at univariate analysis, suggesting a quantitative interaction at multivariate analysis. However, in the statistical methods section, the authors do not mention any likelihood of false-positive results caused by repeated analyses. The reader is not informed about the use of any corrective methods; thus, significant P values could run the risk of only being borderline, if one takes into consideration the items of multiple significant testing and the small number of patients studied. Love et al previously stated [4] that women with recurrent breast cancer were treated with oophorectomy plus tamoxifen or oophorectomy or tamoxifen alone in a rate of 23%, 1%, and 52% respectively, whereas 19% of these patients did not receive any hormonal treatment. However, exploring interactions between HER2/neu and treatment status [1], the authors do not report the hormone-receptor and HER2/neu status of women whose disease recurred. Whether patients with disease recurrence are considered or not in the control group, it is noteworthy that at univariate analysis differences in disease-free survival for HER2/neu-negative patients are robustly significant with respect to endocrine treatment, whereas these results become statistically weak when their overall survival is examined. It might be interesting if authors could speculate on these findings; that is, report on HER2/neu expression in patients dying as a result of causes other than breast cancer. In addition, although the results of Love’s study are consistent in showing greater benefit to the HER2/neu-positive subgroup given adjuvant treatment, the available data are not sufficiently long-term to draw any definitive conclusion.

Finally, Love et al do not point to the role of the luteal phase of the menstrual cycle with respect to the efficacy of surgical oophorectomy plus tamoxifen in premenopausal women with early breast cancer [6]. It would be interesting to test whether timing of initiation of endocrine manipulations during a given menstrual cycle phase might exert a positive or negative effect in clinical outcomes for breast cancer patients who overexpress HER2/neu. The multivariate analysis in the 177 ER-negative HER2/neu patient groups failed to demonstrate any treatment effect or interaction with HER2/neu status. These findings could be discussed in the light of the current evidence that ER-negative tumors frequently show low c-erbB2 levels (defined as the mean value minus one standard deviation) and that hypoexpression appears to be worse than overexpression for disease-free survival in breast cancer patients [7].

Data about the role of HER2/neu in selecting endocrine therapy are virtually nonexistent and, before considering this marker as being pivotal in drawing treatment strategy, additional trials are needed. Nevertheless, the results reported by Love et al are exciting and suggest new approaches for treatment of premenopausal early breast cancer patients with ER-positive and HER2/neu-overexpressing tumors.

Authors’ Disclosures of Potential Conflicts of Interest

The authors indicated no potential conflicts of interest.