Association Between Metformin Use and Risk of Prostate Cancer and Its Grade

David Margel; David Urbach; Lorraine L. Lipscombe; Chaim M. Bell; Girish Kulkarni; Peter C. Austin; Neil Fleshner


J Natl Cancer Inst. 2013;105(15):1123-1131. 

In This Article

Abstract and Introduction


Background Metformin is commonly prescribed to treat type 2 diabetes. Recent evidence suggests that it may possess antitumoral properties. The aim of this study was to test the association between metformin use and risk of prostate cancer and its grade among men with diabetes.

Methods Data were obtained from population-based health-care administrative databases in Ontario, Canada. This retrospective cohort study used a nested case–control approach to examine the relationship between metformin exposure and the risk of prostate cancer within a cohort of incident diabetic men aged 66 years or older. We conducted four case–control analyses, defining case subjects as 1) any prostate cancer, 2) high-grade, 3) low-grade, and 4) biopsy-diagnosed. In each analysis, case subjects were matched to five control subjects on age and cohort entry date. Metformin exposure was determined based on prescriptions before cancer diagnosis, and adjusted odds ratios (aOR) were estimated using conditional logistic regression. All statistical tests were two-sided.

Results Within our cohort of 119 315 men with diabetes, there were 5306 case subjects with prostate cancer and 26 530 matched control subjects. Within the cancer case subjects, 1104 had high- grade cancer, 1719 had low-grade cancer, and 3524 had biopsy-diagnosed cancer. There was no association between metformin use and risk of any prostate cancer (aOR = 1.03, 95% confidence interval [CI] = 0.96 to 1.1), high-grade cancer (aOR = 1.13, 95% CI = 0.96 to 1.32), low-grade cancer (aOR = 0.94, 95% CI = 0.82 to 1.06), or biopsy-diagnosed cancer (aOR = 0.98, 95% CI = 0.84 to 1.02).

Conclusions This large study did not find an association between metformin use and risk of prostate cancer among older men with diabetes, regardless of cancer grade or method of diagnosis.


Prostate cancer is the most common male malignancy in the Western world.[1] Despite a marked global discrepancy in age-adjusted mortality rates from prostate cancer, autopsy studies have confirmed that small foci of prostate cancer exist ubiquitously in 42% to 80% of males in their eighth decade.[2,3] Furthermore, histological evidence suggests that microscopic disease may be present in one of three men in their thirties.[4,5] These observations have led to the theory of a late-stage promoter that converts latent prostatic cancer into clinical prostate cancer.[6] It is now believed that prostate cancer is a disease that histologically starts among men in the fourth decade of life.[6] Thus it takes decades for these malignancies to progress into a potentially clinically detectable disease, and the vast majority of these would never be identified.[6] As such, preventing progression of prostate cancer to clinically detectable disease appears to be a realistic goal.

Metformin (1,1-dimethylbiguanide hydrochloride) is the most commonly prescribed oral hypoglycemic agent worldwide. It is used in diabetes because of both excellent tolerability and efficacy in reducing insulin resistance and reducing mortality.[7] The mechanism of metformin action in diabetes involves the inhibition of hepatic gluconeogenesis and the stimulation of glucose uptake in muscle and adipose tissue. At the cellular level this is achieved by metformin activating AMP-activated protein kinase (AMPK), a cellular energy sensor involved in regulating cellular metabolism that is in turn activated by increases in the intracellular AMP/ATP ratio.[8–10]

Metformin has a variety of mechanisms by which cancer progression and carcinogenesis may be retarded. These include direct effects (on the tumor and microenvironment) and indirect effects (on the host that may influence the tumor). In general, both of these effects are thought to be mediated through the AKT–MTOR pathway.[11–13] Mechanisms for pathway activation that are most relevant and common in prostate cancer include loss of tumor suppressor PTEN,[14] mutation of PI3K,[15] or activation of growth factor receptors such as insulin.[16] Results of in vitro and in vivo studies in cancer models have been encouraging.[17–22]

Three previous studies have systematically tested whether metformin is associated with prostate cancer risk. One of these studies demonstrated a decreased risk,[23] a second study demonstrated an association with all antidiabetic medication but not specific to metformin,[24] and the third demonstrated an increased risk of prostate cancer among metformin users.[25] Obviously, no conclusion of the nature of the association of metformin use with prostate cancer is possible at this point.

At present there is no known approved preventive strategy for prostate cancer. In view of the strong in vivo and in vitro evidence of a potential protective effect of metformin,[17–22] as well as support of the concept from other malignancies, especially breast cancer,[20] we felt that a larger population based study was needed. In addition, no study has examined the influence of metformin use on prostate cancer grade. Prostate cancer is a heterogeneous disease, and for any drug to be considered for prostate cancer prevention, it must have the capacity to prevent high-grade prostate cancer. Results from chemoprevention trials with 5-alpha reductase inhibitors demonstrated that a drug may prevent low-grade tumors but may be associated with an increased risk of high-grade disease.[26] Therefore, we sought to investigate the association of metformin use and risk of prostate cancer among diabetic men. Because we believe that prevention of high-grade disease is fundamental, we also examined effects of metformin on the risk of cancer subtypes to test whether effects differ by cancer grade.