Abstract 2319: Anti-diabetes medication use and risk of colorectal polyps: a colonoscopy-based study using natural language processing

Abstract Purpose: Type-2 diabetes (T2D) has been previously associated with an increased risk of colorectal polyps, but the effect of anti-diabetes medication use on polyp risk remains unclear. Methods: Using pathology reports from the University of Utah (UU) Enterprise Data Warehouse (EDW), we developed a rule-based natural language processing (NLP) pipeline to extract colorectal polyp diagnoses and features (site, shape, number, size) on 15, 679 patients who underwent a colonoscopy at the UU Gastroenterology clinic from 2011-2020. The NLP pipeline, validated by manual abstraction of 350 pathology reports, demonstrated excellent performance (accuracy 91%). We identified 6, 556 T2D patients using ICD codes and extracted patient characteristics from the EDW, including age, sex, race, smoking, BMI, and active anti-diabetes medication prescriptions (insulin, metformin, other). Adjusted odds ratios (OR) and 95% confidence limits (95%CI) were calculated for medication use using multivariable polytomous logistic regression models. Results: A total of 73% of patients with T2D used anti-diabetes medications and were, on average, 62 years old, 76% White, and 50% female, with a mean BMI of 33 kg/m2. Patients taking insulin had a statistically significant increased risk of colorectal polyps compared to patients not taking anti-diabetes medications or those who were taking medications other than insulin [OR(95% CI)= 1.44(1.22, 1.70) and 1.36(1.17, 1.60)], respectively. Metformin use was not associated with an increased risk of polyps when compared to patients not on anti-diabetes medications [OR(95% CI)= 1.16(1.00, 1.34)]. When compared to patients taking medications other than metformin, those on metformin had a decreased risk for polyps [OR(95% CI)= 0.63(0.54, 0.74)]. When evaluated by polyp subtype, insulin use was associated with an increased risk for adenoma and serrated polyps [OR(95%CI)= 1.28(1.08, 1.53) and 1.76(1.44, 2.16), respectively], compared to no anti-diabetes medication use. Patients taking metformin had an increased risk for serrated polyps [OR95%CI= 1.36(1.11, 1.67)], compared to non-users of metformin. When evaluating polyp severity, risks were higher for patients with advanced pathology (advanced adenoma and sessile serrated polyps) compared to those with non-advanced pathology. Conclusion: Insulin use was associated with an increased risk, while metformin use was associated with a decreased risk for colorectal polyps, compared to patients taking other medications. These results are consistent with some published literature for colorectal cancer, suggesting that these pathways may be acting at the pre-neoplasia stage rather than during neoplastic conversion from polyps to colorectal cancer. Further research is needed to understand how T2D and its treatment contribute to colorectal neoplasia. Citation Format: Jessica van Onselen, Ryzen Benson, Stephanie Richardson, Maci Winn, Candace Winterton, Bejamin Krick, Svenja Pauleck, Ainhoa Gomez-Lumbreras, Polly Newcomb, Neli Ulrich, John Inadomi, Sheetal Hardikar. Anti-diabetes medication use and risk of colorectal polyps: a colonoscopy-based study using natural language processing [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 2319.