Aim 1: Determine associations among body mass index (BMI), body composition variables, and tumor subtype. The amount and quality of muscle, visceral adipose tissue, and subcutaneous adipose tissue will be derived from existing pre-surgical CT scans; and ccA/ccB subtypes will be derived from archived tumor specimens using NanoString.
Aim 2. Identify body composition variables independently associated with ccRCC prognosis after accounting for tumor subtype. Multivariable models adjusting for demographics, BMI, comorbidities, and known prognostic factors will estimate how body composition variables are associated with disease progression, cancer-specific, and overall survival and whether these associations are modified or mediated by tumor subtype.
Aim 3. Identify novel pathways through which body composition may affect prognosis.