Body composition and adiposity-associated thromboinflammation in colorectal cancer prognosis

SUMMARY Obesity increases the risk of numerous cancers including colorectal cancer (CRC). With the growing global epidemic of obesity, a greater understanding of the obesity cancer-link, including its underlying biology and its adverse influence on patient outcomes is critically needed. Our ability to treat patients optimally has been

hampered, in part, by the traditional use of body mass index (BMI) to define obesity. BMI fails to take into account important dimensions of body composition, such as fat distribution (e.g., visceral obesity) and muscle mass (e.g., sarcopenia). A clinically applicable approach to characterizing obesity and body composition

combined with an understanding of CRC-obesity biology will advance our ability to tailor treatments and interventions to disrupt the adverse impacts of obesity on treatment toxicity and patient outcomes. We propose an interdisciplinary study that addresses both gaps in knowledge by (1) application of a cutting-edge artificial

intelligence (AI)-powered, fully automated CT scan-based platform to generate accurate, individualized volumetric measures of adipose tissue compartments/skeletal muscle mass and (2) testing the role of thromboinflammation, a novel biologic pathway to investigate in the context of obesity-CRC outcomes.

Thromboinflammation, a state characterized by increased platelet activation and systemic and tumor-specific sequelae, plays a role in metastasis and has recently, but separately, been associated with obesity and cancer; however, it has never before been investigated as a biological interface and potential target of

intervention in the obesity-CRC link. Our study leverages a unique resource of longitudinal CT images, clinical data, tumor, plasma, and urine from a population of N=1,200 stage I-III CRC patients enrolled in the NCI-funded ColoCare Study. In Aim 1, we will apply a novel AI-powered fully automated radiologic 3D body segmentation software to generate robust

volumetric adiposity and muscle mass profiles and examine associations with treatment-related toxicity and clinical outcomes. In Aim 2, we will examine thromboinflammation biomarkers in blood, urine, and tumor in relation to visceral obesity and survival outcomes and test whether this pathway mediates Aim 1 associations.

Together, we will discern differences by race and other factors (e.g., age, sex, stage, and tumor site) The proposed project takes a powerful and rigorous team science approach to address key questions on the impact of body composition and adiposity-promoted thromboinflammation on CRC treatment-related

toxicities and survival. Use of a single, large, study patient population with an exceptional set of uniformly available CT scans, data, and biospecimens is cost-effective and maximally informative. Our study will not only yield clinically meaningful prognostic measures of body composition and thromboinflammation but will also

provide insights into obesity-CRC biology with the potential to identify novel personalized treatment strategies.