A phase III randomized trial of metformin versus placebo on recurrence and survival in early-stage breast cancer (BC) (NCIC Clinical Trials Group MA.32).

Abstract

TPS103 Background: There has been a recent convergence of epidemiologic, clinical and preclinical evidence suggesting metformin (a biguanide used to treat type 2 diabetes) may lower BC risk and improve BC outcomes. This evidence raises the intriguing possibility that metformin may act (i) indirectly by lowering circulating insulin levels (secondary to reduced hepatic gluconeogenesis) leading to reduced insulin/IGF-I receptor mediated PI3K signaling in BC cells or (ii) directly on tumor cells, largely via an AMPK mediated mechanism, leading to mTOR inhibition. Methods: NCIC CTG led, CTEP sponsored, randomized, phase III trial of metformin 850 mg po bid vs placebo bid for 5 years. Eligibility criteria: pTIcNO (and > one: grade 3, ER/PR-, HER2Neu+, lymphovascular invasion, Oncotype DX recurrence score>25, Ki67>14%) or pT2/T3 N0 or pT1-3 N1-3. Standard adjuvant therapy (chemotherapy, hormone, biologics) Adequate organ function and PS Age 18-75 Endpoints: Primary: invasive disease free survival (IDFS). Secondary: overall survival, distant disease free survival, BC free survival, adverse events, new diabetes, cardiovascular hospitalizations, BMI, insulin resistance syndrome attributes (ATP III criteria), Quality of Life, diet and physical activity. Correlative studies: Embedded correlative research examines prognostic factors and targets and predictors of metformin benefit.Variables include: Fasting Insulin, glucose, Homeostasis Model Assessment (HOMA) (baseline, 6 months, end of treatment): genomic/gene expression analysis of blood and tumour, tumor insulin receptor (IR), IGF-IR, tumor molecular markers of target pathways (LKBI, PI3K (STMN1), PKB/Akt (P-PKB/Akt,) mTOR (P-4E-BP1) and IRS-1 (P-IRS-1); Fatigue sub-study with biological correlates (NSABP led). Statistical design: Planned accrual is 3,582 subjects over 3 years with 3 more years follow-up. Target hazard ratio is 0.76 (5 year IDFS of 88.4% vs 85%, power 0.80, 2-tail type 1 error of 0.05). Two interim analyses are planned. Conduct to Date: Study activation: June 2010. Enrollment: 145 subjects. In Nov 2010 a DSMC review supported trial continuation.