Lack of chemopreventive efficacy of metformin in rodent models of urinary bladder, head and neck, and colon/intestine cancer.

Metformin is a biguanide employed in treating type II diabetes. Its potential efficacy for treating cancer has been demonstrated epidemiologically (lower cancer incidence in metformin users compared with users of sulfonylureas or insulin) and mechanistically, primarily in cell culture. Metformin decreases the levels of insulin-like growth factor 1 and secondarily inhibits the mammalian target of rapamycin pathway to exhibit anticancer effects. The current study examined its cancer preventive efficacy in multiple standard in situ arising cancer models. Metformin was administered orally by gavage or in the diet, at human equivalent doses, in numerous cancer models. In the hydroxybutyl(butyl)nitrosamine-induced model of invasive urinary bladder cancer, metformin (50 or 150 mg/kg body weight/day, intragastric) was ineffective despite high urinary concentrations of metformin. Metformin (250 or 500 ppm in diet) failed to decrease the incidence or invasiveness of squamous cell cancer of the tongue in a 4-nitroquinoline-1-(4NQO)-induced model. Finally, in the Min mouse model of gastrointestinal cancer, metformin (400 or 1,200 ppm in diet) was ineffective. Notably, a slight increase in intestinal tumor multiplicity was observed at the higher dose. Therefore, metformin lacked efficacy in multiple standard cancer models in non-diabetic rodents. This lack of efficacy may discourage any large phase clinical cancer trials in non-diabetic individuals in the absence of clear phase-II studies.