Olesya Mironchuk, Andrew L. Chang, Farzaneh Rahmani, Kaitlyn Portell, Elena Nunez, Zack Nigogosyan, Da Ma, Karteek Popuri, Vincent Tze Yang Chow, Mirza Faisal Beg, Jingqin Luo, and Joseph E. Ippolito
Abstract Clinically, the body mass index remains the most frequently used metric of overall obesity, although it is flawed by its inability to account for different adipose (i.e., visceral, subcutaneous, and inter/intramuscular) compartments, as well as muscle mass. Numerous prior studies have demonstrated linkages between specific adipose or muscle compartments to outcomes of multiple diseases. Although there are no universally accepted standards for body composition measurement, many studies use a single slice at the L3 vertebral level. In this study, we use computed tomography (CT) studies from patients in The Cancer Genome Atlas (TCGA) to compare current L3-based techniques with volumetric techniques, demonstrating potential limitations with level-based approaches for assessing outcomes. In addition, we identify gene expression signatures in normal kidney that correlate with fat and muscle body composition traits that can be used to predict sex-specific outcomes in renal cell carcinoma.