Metformin is a bi-guanidine employed in treating Type II diabetes. Epidemiologic studies have suggested that diabetics receiving metformin vs thialidizones or insulin have a lower incidence of cancer. In addition, metformin decreases levels of AMPK and IGF1; encouraging its use in both cancer prevention and therapy settings. Most preclinical studies with metformin have employed cell culture or xenografts. We examined its efficacy in multiple in situ animal cancer models. The preventive effect of Metformin administered by gavage or in the diet, at roughly the human equivalent dose (HED), was determined in four standard animal models: (1) methylnitrosourea (MNU)-induced ER+ mammary cancers; (2) MMTV-Neu ER mammary cancers; (3) hydroxybutyl(butyl)nitrosamine (OH-BBN) induced urinary bladder cancers; and (4) Min mouse model of intestinal tumorigenesis. In the MNU model, Metformin (150 or 50 mg/Kg BW/day) was ineffective as a preventive dose; decreasing neither tumor multiplicity nor latency. Furthermore, both doses increased final tumor weights in treated rats. PK studies at the higher dose of Metformin (150 mg/kg BW/day) showed serum levels slightly higher than humans taking 1.5 g per day. In the MMTV-Neu model, dietary metformin (1200 PPM in diet) decreased the latency of tumor development as well as slightly, but insignificantly, increasing the final tumor multiplicity. Metformin (1200 ppm in diet) in the Min mouse model marginally increased the multiplicity of intestinal adenomas. In contrast, the NSAID naproxen (400 ppm in diet) examined simultaneously in the min mouse strongly inhibited adenoma development. Metformin (150 mg/Kg BW/day) in the OH-BBN induced model of urinary bladder cancer failed to significantly reduce bladder cancers despite relatively high levels of Metformin in the urine. In the MNU model, when rats were placed on a high fat diet to induce a pre-diabetic state, we observed roughly a 20% decrease in tumor multiplicity in the metformin treated rats (P> 0.10). In contrast, tamoxifen and Targretin (an RXR agonist) that were used as positive controls were highly effective in rats on a high fat diet. Metformin didn't decrease tumorigenesis in these standard models, but actually increased tumor formation (incidence, multiplicity, weight) in three of the four models at their HED. These results should particularly discourage any large phase clinical trials in individuals without insulin resistance. The final study certainly raises the question whether the agent might be effective in animals that are diabetic or show insulin resistance

Citation Format: Clinton J. Grubbs, Margie L. Clapper, Joel M. Reid, Vernon E. Steele, Ronald A. Lubet. Metformin promotes tumorigenesis in animal models of cancer prevention. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 2273. doi:10.1158/1538-7445.AM2013-2273