Clinically relevant body composition phenotypes are associated with distinct circulating cytokine and metabolomic milieus in epithelial ovarian cancer patients.

Abstract

Introduction: Preclinical evidence suggests that host obesity is associated with tumor progression due to immuno-metabolic dysfunction, but the impact of obesity on immunity and clinical outcomes in patients is poorly understood, with some studies suggesting an obesity paradox. We recently reported that high-adiposity and low-muscle body composition phenotypes are associated with striking increases in epithelial ovarian cancer (EOC) mortality and we observed no evidence of an obesity paradox. However, whether at-risk versus optimal body composition phenotypes are associated with distinct immuno-metabolic milieus remains a fundamental gap in knowledge. Herein, we defined differentially abundant circulating immuno-metabolic biomarkers according to body composition phenotypes in EOC.

Methods: Muscle and adiposity cross-sectional area (cm²) was assessed using CT images from 200 EOC patients in The Body Composition and Epithelial Ovarian Cancer Survival Study at Roswell Park. Adiposity was dichotomized as low versus high; patients with skeletal muscle index (SMI) <38.5 (muscle cm²/height m²) were classified as low SMI (sarcopenia). Joint-exposure phenotypes were categorized as: Fit (normal SMI/low-adiposity), Overweight/Obese (normal SMI/high-adiposity), Sarcopenia/Obese (low SMI/high adiposity), and Sarcopenia/Cachexia (low SMI/low-adiposity). Treatment-naïve serum samples were assessed using Biocrates MxP Quant 500 for targeted metabolomics and commercially available Luminex kits for adipokines and Th1/Th2 cytokines. Limma moderated T-tests were used to identify differentially abundant metabolites and cytokines according to body composition phenotypes.

Results: Patients with 'risk' phenotypes had significantly increased abundance of metabolites and cytokines that were unique according to body composition phenotype. Specifically, the metabolites and cytokines in increased abundance in the at-risk phenotypes are implicated in immune suppression and tumor progression. Conversely, increased abundance of lauric acid, IL-1β, and IL-2 in the Fit phenotype was observed, which have been previously implicated in tumor suppression and anti-tumor immunity.

Conclusion: In this pilot study, we identified several significantly differentially abundant metabolites according to body composition phenotypes, confirming that clinically significant joint-exposure body composition phenotypes are also biologically distinct. Although we observed evidence that at-risk phenotypes were associated with increased abundance of immuno-metabolic biomarkers indicated in immune suppression, additional confirmatory studies focused on defining the link between body composition and immune cell composition and spatial relationships in the EOC tumor microenvironment are warranted.