Abstract 3006: Metabolomic signatures of metabolic disturbance and inflammation in relation to colorectal cancer risk

Abstract

Abstract Background: Metabolic disturbance and inflammation may explain observed associations between higher body mass index (BMI) and increased risk of colorectal cancer (CRC); however, the underlying mechanisms are not fully understood. Objectives: We characterized individual plasma metabolites and metabolomic signatures of metabolic disturbance and inflammation and evaluated their association with prospective CRC risk within the Nurses’ Health Study and the Health Professionals Follow-up Study. Methods: Among 686 colorectal cancer cases and 686 age-matched controls, we used reduced rank regression of markers of metabolic disturbance (BMI, waist circumference, C-peptide, and adiponectin) or inflammation (BMI, C-reactive protein, interleukin-6, and tumor necrosis factor receptor superfamily member 1B) with cross-sectional measures of 353 plasma metabolites to develop a Y-score for metabolic disturbance and Y-score for inflammation among men and women separately. We then used elastic net regression to derive a signature of metabolites, and multiple linear regression to identify individual metabolites, associated with each Y-score. We evaluated the association of individual metabolites and the metabolomic signatures with odds of CRC using conditional logistic regression adjusted for other CRC risk factors. Results: The metabolomic signature of metabolic disturbance consisted of 41 metabolites selected via elastic net regression in men and 72 in women; the metabolomic signature of inflammation consisted of 68 metabolites in men and 119 in women. The metabolic disturbance metabolomic signatures captured, on average, 36% of variation in markers of metabolic disturbance in women and 35% in men; the inflammation signature captured 35% of variation in inflammatory markers in women and 26% in men. The metabolomic signature of metabolic disturbance was associated with increased odds of CRC (odds ratio (OR) comparing highest to lowest quartile = 1.63; 95% confidence interval (CI), 0.92, 2.91; Ptrend = 0.31) and the metabolomic signature of inflammation was associated with increased odds of CRC (OR = 2.01; 95% CI, 1.14, 3.57; Ptrend = 0.008) among men. Neither signature was associated with CRC among women. Of the metabolites associated with metabolic disturbance and inflammation, 13 metabolites were also associated with CRC: 4 metabolites classified as uremic toxins (2 purine nucleosides and 2 amino acid derivates); 3 sphingolipids; 3 glycerophospholipids; 2 sterols related to cholesterol homeostasis; and 3-ureidopropanoate, a uracil metabolism substrate. Conclusion: We identified plasma metabolomic signatures and individual metabolites associated with metabolic disturbance, inflammation, and CRC risk, highlighting pathways such as protein metabolism and lipid homeostasis that may relate adiposity-related metabolic disturbance and inflammation to CRC development. Citation Format: Alaina M. Bever, Dong Hang, Amit D. Joshi, Connor M. Geraghty, Dong Hoon Lee, Fred K. Tabung, Shuji Ogino, Jeffrey A. Meyerhardt, Andrew T. Chan, Edward L. Giovannucci, A. Heather Eliassen, Liming Liang, Meir J. Stampfer, Mingyang Song. Metabolomic signatures of metabolic disturbance and inflammation in relation to colorectal cancer risk [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 3006.