31. Computer-aided cytogenomic classification of renal cell carcinoma

Abstract

Renal Cell Carcinoma (RCC) is traditionally classified based on immunohistochemistry and morphology according to WHO criteria. Recent evidence demonstrate that large copy number aberrations (CNAs) can be used to better classify RCCs into distinct subgroups with more accurate prognostic and therapeutic implications. Here we describe our research on using an automated computer classification tool to better classify pathologically challenging RCC cases. We first took 641 cases from The Cancer Genomic Atlas (TCGA) to create 16 distinct clusters using a novel neural network algorithm.