Abstract 5035: MicroRNA expression profiles classify renal cell carcinoma subtypes

Abstract

Abstract Background: Renal cell carcinoma (RCC) accounts for about 2-4% of all human malignancies. Hereby, the entity of papillary RCC (pRCC) is the second most common entity with about 10-15% of all RCC. According to histomorphological characteristics pRCCs can be further subdivided in two distinct subtypes. Type 2 pRCCs are correlated with a worse clinical outcome than type 1 pRCCs. In order to characterize these RCC subtypes we studied microRNA (miRNA) expression profiles of primary pRCCs type 1 and 2 as well as clear cell RCC (ccRCC). Material and Methods: We performed comparative miRNA expression profiling using Affymetrix miRNA microarrays. Ten to fifteen samples were analyzed for each subtype containing matched pairs of microdissected malignant and nonmalignant tissue. Hierarchical clustering was used to identify characteristic miRNA expression signatures. The expression of selected miRNAs was validated using quantitative PCR. Results: Hierarchical clustering was not able to identify miRNA expression signatures characteristic for all RCC subtypes, highlighting the heterogeneity among RCC tumors. But when stratified according to histological subtypes, tumor and nonmalignant tissue samples could be separated from each other. Interestingly, pRCC tumors of type 1 and 2 displayed a characteristic miRNA expression profile making it possible to separate these two tumor subtypes. A differential expression of miRNAs miR-200c, mir-143, miR-145, miR-210 and miR-10b could be confirmed by qRT-PCR in a cohort of pRCC samples (p<0.05; paired t-test). Applying a binary logistic regression model, a combination of only three of these miRNA miRNAs was sufficient to correctly determine the subtype of all pRCC tumor samples. Conclusion: Our findings demonstrate that miRNA expression in RCC is heterogeneous but there are characteristic miRNA expression profiles within RCC subtypes. In pRCC, specific miRNAs may reflect a molecular correlate to the histological differences between the two subtypes, which in turn are correlated to the clinical outcome. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 5035. doi:1538-7445.AM2012-5035