Multiple Hypotheses (continued)
It appears to inhibit tumor cell growth through activation of AMP kinase and downstream suppression of mTOR signaling.24,25 Results of multiple studies suggests that metformin may reduce diabetes-related cancer risk and that a dose-response relationship exists, where longer exposure to metformin correlates with lower cancer risk.24-29 Patients with diabetes using metformin monotherapy have lower rates of progression to solid tumor development, particularly colorectal cancer (CRC) and pancreatic tumors, than those treated with sulfonylureas (SUs) or insulin.26
Thiazolidinediones (TZDs) combat hyperglycemia by increasing insulin sensitivity through activation of PPAR-gamma and do not increase endogenous insulin.30 In a recent systematic review of their use for T2DM, TZDs were associated with modest but significant risk reductions in CRC (7%), breast (11%), and lung cancers (9%); results were from only a few studies and should be viewed with caution.31 Pioglitazone, the only TZD readily available in the United States, has been associated with an increased risk for bladder cancer in cohort studies, which was strongly associated with 5 or more years of use.30,32
SUs increase pancreatic beta-cell production of endogenous insulin and thus insulin levels. SUs appear to convey greater risk for developing cancer compared with metformin, but evidence from a meta-analysis of observational studies suggests no effect on overall cancer risk.
A case-controlled study using SUs found no change in breast cancer risk with either short- or long-term use.24,29 However, when compared with metformin, use of SUs was associated with increased cancer mortality in a population-based study from Canada.33
Exogenous insulin is often initiated in T2DM several years after diagnosis, if diabetes cannot be controlled with oral therapies. A number of studies have associated its use with higher risk for some types of cancer, including CRC, pancreatic cancer, and prostate cancer.34,35 In a 2012 meta-analysis, pancreatic cancer risk was three-fold higher among new users of human insulin and elevated to a smaller (61%) but still significant degree among patients using the insulin analog glargine.36 The same analysis found a much higher risk for prostate cancer among new users of glargine, but a lower risk for CRC.36 In a population-based study, T2DM patients using insulin were almost twice as likely as those using metformin to die from cancer-related causes.33 Though hyperinsulinemia presents an appealing hypothesis for the mechanism, patients with diabetes using insulin rarely have elevated levels.37 In the setting of advanced T2DM, where insulin is typically used, multiple additional factors are at play, including long-term exposure to hyperglycemia and related inflammation with uncontrolled diabetes.38 Dysregulation of other systems due to insufficient insulin may play a role as well.37