Can molecular markers predict when to implement treatment with aromatase inhibitors in invasive breast cancer?

Purpose: Resistance to tamoxifen is linked to overexpression of HER2, and aromatase inhibitors show particular benefit in progesterone receptor (PR) - negative patients. We previously reported reduced survival in patients overexpressing HER1, HER2, and HER3. We now show that both HER1-3 and PR status predicts for early relapse in estrogen receptor (ER) - positive tamoxifen-treated breast cancer patients. Experimental Design: Tissue microarray technology was used to analyze 402 ER-positive tamoxifen-treated patients. Immunohistochemistry using epidermal growth factor receptor, HER2, HER3, HER4, and PR antibodies was done. Kaplan-Meier life table and Cox Regression analysis (log-rank testing of differences in breast cancer - related relapse on tamoxifen) was done. Results: HER1-3 (but not HER4) overexpression predicted for early relapse on tamoxifen (P = 0.0060). PR-negative cases were also significantly more likely to relapse while on tamoxifen (P = 0.017). HER1-3-positive and/or PR-negative patients combined as a "high-risk" group were significantly more likely to relapse on tamoxifen in univariate (P < 0.0001) and Cox's multivariate analysis (P = 0.0069). However, this applied to early relapse on tamoxifen only, as any disease relapse after 3 years of tamoxifen was unrelated to PR/HER status. Conclusions: We show that HER1-3 and PR status can identify time-dependent de novo tamoxifen resistance with risk declining markedly after 3 years of tamoxifen treatment. These results parallel data from the ATAC and Intergroup Exemastane Study trials which suggest that whereas PR-negative patients derive greater benefit from initial aromatase inhibitor treatment, PR status has no effect on response when given as delayed treatment to those disease free on tamoxifen after 3 years.