Metformin Use Is Inversely Associated with Prevalent, but Not Incident Colorectal Adenomas.

Abstract

Chemoprevention for colorectal neoplasia has attracted growing interest, with multiple medications investigated. Metformin may decrease the overall incidence of cancer in patients with diabetes and may decrease the incidence of colorectal cancer. We aimed to determine the impact of metformin use on the behavior of colorectal adenomas in a US veteran population. All patients with at least two high-quality colonoscopies between January 1997 and December 2013 at Veterans Affairs New York Harbor Healthcare System were identified. Outpatient prescription records were used to determine metformin exposure, and colonoscopy findings were recorded. Multivariable logistic regression was used to determine factors associated with adenoma detection on baseline and interval colonoscopy. In total, 1869 patients with two successive colonoscopies (median 4.5years) were included. Four hundred and sixty patients had metformin exposure prior to baseline and/or interval colonoscopy. Overall adenoma detection rate was 59.7% at baseline and 45.9% at interval colonoscopy. On multivariable analysis, metformin use was associated with decreased adenoma prevalence at baseline (OR 0.68; 95% CI 0.51-0.92; p = 0.015). Metformin did not impact adenoma incidence at interval colonoscopy whether prescribed before baseline (OR 1.26; 95% CI 0.60-2.67), after baseline (OR 1.25; 95% CI 0.91-1.72), or before and after baseline (OR 1.14; 95% CI 0.82-1.58). In this retrospective analysis of an average-risk cohort, metformin use was associated with a decreased prevalence of colorectal adenomas at baseline colonoscopy. This inverse association did not persist on interval colonoscopy. Prospective studies are needed to evaluate potential chemoprotective effects of metformin over time.