Abstract P1-14-07: Association between quantitative values of estrogen receptor expression level and pathological complete response in human epidermal growth factor 2-negative breast cancer: Should the clinical definition of triple-negative breast cancer be redefined?

Abstract

Abstract Background: The American Society of Clinical Oncology/College of American Pathologists recommended that the cut-off for negative status of estrogen receptor (ER) should be <1% positively staining cells, although a 10% cut-off has often been used clinically. Prior studies reported that patients with ER ranging from 1% to 9% showed survival outcomes and molecular features similar to those of patients with ER positivity of <1%; however, those studies did not take into account patients' human epidermal growth factor 2 (HER2) status. This means we have yet to clarify the exact clinical definition of triple-negative breast cancer (TNBC) on the basis of response to preoperative chemotherapy. Previous studies reported that hormone receptor–positive tumors were less sensitive to systemic chemotherapies. On the basis of these facts, we hypothesized that in patients with HER2-negative breast cancer ER expression level as a continuous variable has an inverse linear association with pathological complete response (pCR) rate. Our primary objective was to determine whether a quantitative value of ER between 0% and 10% is predictive of pCR rate in HER2-negative patients treated with neoadjuvant chemotherapy. Secondary objective was to find the ideal cut-off value of ER expression. Methods: We included newly diagnosed stage I-III HER2-negative breast cancer patients with available ER (0%≤ER<10%) who were treated with neoadjuvant systemic chemotherapy. ER status was determined by immunohistochemical (IHC) staining; HER2 status was determined by IHC and/or FISH. We used univariate and multivariate logistic regression models to determine the association between baseline variables and pCR. A backward stepwise method was used to select the covariates for the multivariate analysis. Recursive partitioning and regression tree method were used to identify the potential significant cut-off of ER. Results: The analysis included 1155 patients with newly diagnosed HER2-negative invasive breast cancer. The univariate logistic regression analysis showed that ER as a continuous variable was not a statistically significant factor for predicting pCR (ER: OR=0.98, 95%CI: 0.9-1.07, P=0.68). In the multivariate analysis, ER status again was not a significant factor for predicting pCR (OR=0.97, 95%CI 0.9-1.06, P=0.55). ER as a categorical variable, there was no significant difference of the pCR rate between 0<ER<1 and 1≤ER<10 groups (OR=1.27, 95%CI: 0.62-2.62, P=0.52). Among ER> 0 (n=229), the recommended cut-off value of ER was 5.5. However, the odds ratio of pCR rate divided by this value of 5.5 was not significant (ER≤5 vs ER>5; OR 1.94 95%CI 0.54-6.95 P=0.31). Conclusion: Evaluating ER (<10%) as a continuous variable showed no association with pCR rate, and no cut-off of ER was identified with which to stratify patients into groups more or less likely to achieve pCR. A potential meaningful cut-off ER value might exist between 10% and 100% in HER2-negative patients. We will explore whether a meaningful cut-off ER value exists that will change the pCR rate and possibly lead to redefining the clinical definition of TNBC. Citation Format: Fujii T, Kogawa T, Dong W, Moulder S, Litton JK, Tripathy D, Lim B, Shen Y, Ueno NT. Association between quantitative values of estrogen receptor expression level and pathological complete response in human epidermal growth factor 2-negative breast cancer: Should the clinical definition of triple-negative breast cancer be redefined?. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P1-14-07.