Effect of Prior Metformin Intake on First Diagnosis of Breast Cancer

ABSTRACT Introduction Metformin has been shown in pre clinical studies to have the ability to reduce tumor growth and exert anti cancer effects. It is being tested as a part of management of breast cancer (BC) in neoadjuvant and adjuvant settings and being entertained as a modality of chemoprevention in BC. We studied the effect of prior metformin intake on patients with first diagnosis of breast cancer. Methods All patients with a diagnosis of BC were identified from the Tumor Registry of our hospital, which serves the underserved population of Chicago, USA. All charts were retrospectively screened for age, tumor grade, node status at diagnosis, onset of diabetes mellitus (DM) prior to cancer diagnosis and intake of metformin including duration. Patients with a diagnosis of DM after cancer, intake of metformin less than 6 months or incomplete medical record were excluded. Difference in means was calculated using the student t-test. Results A total of 1569 patients were screened of which 133 patients were identified to have a diagnosis of DM before breast cancer and met the inclusion criteria. Of these, 95 patients were on metformin for an average 3.57 years (yrs) before BC diagnosis. Thirty-nine patients on metformin were younger than 60 yrs and 56 were older than 60 yrs age. Node positive disease was seen in 51.5% of metformin patients compared to 49% DM patients not on metformin (p = NS). However in patients younger than 60 yrs, node positive disease was seen in 66% of metformin patients compared to 86% non-metformin patients. The young patients on metformin had an average grade of 2.38 compared to non-metformin patients whose average grade was 2.83 (p = 0.02). More patients in young non-metformin group were Her2-neu positive (43%) compared to 23% in young metformin group. In older patients more non-metformin patients had triple negative disease (20%) compared to 9% in metformin group. Conclusions Prior metformin use may have a possible positive effect in patients with initial diagnosis of BC. This effect seems more pronounced in patients younger than 60 yrs of age where metformin intake was associated with a statistically significant reduction in tumor grade with a trend towards reduction in node positive disease and Her2 positivity. Further studies are needed to clarify the potential protective role of metformin in this younger group. Disclosure All authors have declared no conflicts of interest.