Abstract LB326: Inflammatory proteomic signatures of metabolically unhealthy obesity in association with colorectal cancer and adenoma

Abstract Background: Metabolically unhealthy obesity (MUO), marked by high visceral adiposity and unhealthy glycemic traits, is a risk factor for colorectal cancer (CRC) and adenoma (CRA). However, the underlying mechanisms of MUO are not fully understood. We sought to characterize inflammatory proteomic signatures of MUO and their associations with incident CRC and CRA risk. Methods: We analyzed n=639 older adults (63.7±6.1 years), free of major chronic disease from extensively phenotyped VITamin D and OmegA-3 TriaL clinical sub-cohort. We defined four MUO phenotypes from standardized values of imaging-based visceral adipose tissue mass plus one of four glycemic metabolic traits, including fasting glucose, hemoglobin A1c (HbA1c), homeostasis model of insulin resistance (HOMA-IR), and lipoprotein insulin resistance (LPIR; a lipid-based score predictive of incident diabetes). Plasma proteomics were measured by the Olink 384 inflammation panel. We identified corresponding proteomic signatures (i.e. “Inflammotypes”) using elastic net regression models for protein selection and externally validated in the COcoa Supplement and Multivitamin Outcomes Study (N=371, age 69.0±5.2). We then analyzed pre-diagnostic proteomic profiles of cases and matched controls for CRC (665 cases, 665 controls, age 61.4±8.2) and CRA (623 cases, 623 controls, age 55.8±8.1) from the prospective Nurses’ Health Study I and II, and Health Professionals Follow-up Study, matched on age, fasting and date of blood draw, and colonoscopy history (CRA only). We estimated the associations between continuous composite MUO-trait Inflammotypes (per standard deviation [SD]) and odds ratios [OR; 95% confidence intervals [CI]) of CRC and CRA using conditional logistic regression models, adjusting for age, race, smoking, alcohol, diet quality (Alternative Healthy Eating Index score), and physical activity. Results: The Inflammotypes contained 51 to 85 proteins for the four MUO phenotypes and each strongly predicted MUO in the external validation, with correlations ranging from 0.69 to 0.81. The HOMA-IR Inflammotype was associated with greater odds of CRC [OR=1.15; CI=1.02, 1.29] and CRA (OR=1.16; CI=1.03, 1.30). The glucose Inflammotype was not related to CRC, but positively associated with CRA (OR=1.15;CI=1.02, 1.29). The LPIR Inflammotype was associated with CRC among men (OR=1.21; CI=1.00, 1.45) but not women (OR=1.04; CI=0.90, 1.19) and HbA1c Inflammotype was unrelated to either endpoint. Conclusion: We derived novel proteomic Inflammotypes characterizing metabolically unhealthy visceral fat tissue that were positively associated to subsequent risk of CRC and CRA in separate cohorts. These findings support novel inflammatory processes as a potential mechanisms linking MUO and the development of CRC and CRA. Citation Format: Cong Wang, Azam Yazdani, Olga Demler, Yiwen Zhang, Oluwafeyisola Osifala, Mingyang Song, Samia Mora, Jonathan Downie, Andrew T. Chan, Edward Giovannucci, Xuehong Zhang, Aditi Hazra, Meryl LeBoff, Howard D. Sesso, JoAnn E. Manson, Deirdre Tobias. Inflammatory proteomic signatures of metabolically unhealthy obesity in association with colorectal cancer and adenoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_2):Abstract nr LB326.