Exploration of BMI and circulating metabolic factors as predictors of metformin benefit in CCTG MA.32.

BACKGROUND: In the MA.32 randomized adjuvant breast cancer (bc) trial, metformin (vs placebo) did not impact invasive disease free (IDFS) or overall survival (OS) in estrogen/progesterone receptor (ER/PgR) positive or negative bc; exploratory analyses suggested a benefit in HER2 positive bc. We investigated whether body mass index (BMI) and obesity-associated blood variables predicted metformin benefit in immunohistochemically defined bc subtypes [luminal (ER/PgR positive, HER2 negative), triple negative (TNBC; ER, PgR, HER2 negative), HER2 positive]. METHODS: 3649 non-diabetic patients with high risk T1-3, N0-3 M0 bc were randomized. Baseline fasting plasma was assayed for insulin, glucose, leptin, hsCRP; Homeostasis Model Assessment (HOMA) was calculated. For each bc subtype and each outcome [distant recurrence free survival (DRFS), OS, IDFS], Cox models examined interactions of Body Mass Index (BMI) and blood variables with metformin vs placebo outcomes. RESULTS: Mean age was 51.1 to 53.0 years; mean BMI 27.3 to 27.5 kg/m. 2 Most cancers were T2, N0 or N1 and grade 2 to 3; 2104 (57.7%) were luminal, 925 (25.3%) TN and 620 (17.0%) HER2 positive. Median follow-up 95.9 months. In luminal bc, significant interactions were identified for leptin, insulin and HOMA on DRFS and in TNBC a significant interaction was identified for glucose on DRFS, with potential adverse effects of metformin at lower levels of each variable. In those with HER2 positive bc, no variable predicted metformin benefit. CONCLUSIONS: BMI and blood variables did not identify subgroups with luminal or triple negative bc who benefitted from metformin nor those with HER2 positive bc who did not benefit.