Background: Metformin use has been linked to pathologic complete response (pCR) following neoadjuvant chemotherapy for several malignancies. We aimed to investigate the association of diabetes mellitus (DM) and metformin use with pCR in breast cancer.
Materials and methods: All breast cancer patients who received neoadjuvant chemotherapy during June 2013-October 2016 at two academic medical centers were identified. A retrospective cohort study evaluated patients who did and did not achieve pCR. Multivariable logistic regression identified independent predictors of pCR, specifically looking at metformin use and DM.
Results: The study group included 351 breast cancer patients, with 90 (25.6%) achieving pCR after neoadjuvant chemotherapy. The rate of DM did not differ between those with and without pCR, nor did the rate of metformin use. Multivariable logistic regression identified HER2-positive tumors and smaller preoperative tumor size as predictors of pCR. The estrogen receptor (ER) positivity was associated with an absence of pCR. Importantly, neither DM nor metformin use was predictive of pCR.
Conclusions: This study by the two institutions supports previous data of tumor-related factors known to be associated with pCR; however, the current analysis found neither DM nor metformin to be independently associated with pCR. Thus, additional prospective study is warranted prior to validating metformin as an antitumor agent.
Keywords: Breast cancer; Diabetes mellitus; Metformin; Neoadjuvant chemotherapy; Pathologic complete response.
Copyright © 2019 Elsevier Inc. All rights reserved.