Assessing the clinical use of clear cell renal cell carcinoma molecular subtypes identified by RNA expression analysis

Abstract

ObjectivesTo evaluate the clinical use of recently published RNA-based molecular subtyping algorithms. Patients who undergo surgery for clinically localized clear cell renal cell carcinoma can experience very different outcomes, representing a longstanding challenge for the practicing urologist. Two recent publications suggest that molecular subtyping based on the expression of large panels of genes can help clinically localized clear cell renal cell carcinoma prognostication; however, the analyses was not adjusted for routinely collected clinicopathologic indices. Methods and materialsWe obtained level 3 RNA-seq RPKM data and corresponding clinicopathologic features from The Cancer Genome Atlas (TCGA) and assigned patients to the TCGA subtypes as well as to the ccA/ccB subtypes. To determine the prognostic ability of molecular subtyping after adjusting for variables that are collected as routine medical care, we used Cox models and adjusted for the composite Mayo stage, size, grade, and necrosis (SSIGN) score. ResultsBoth the TCGA and the ccA/ccB subtypes are significantly associated with tumor size, category, grade, and presence of necrosis. The association of these subtypes with overall survival is markedly attenuated following adjustment for the composite Mayo SSIGN score. ConclusionsBoth the TCGA and the ccA/ccB subtypes are associated with overall survival after adjusting for the Mayo SSIGN score. However, the effect sizes are similar to what has been reported for single markers, and thus the clinical use and cost-effectiveness of these RNA-based whole-genome signatures are questionable.