Our aim was to identify microRNA (miRNA) signatures that can distinguish the different RCC subtypes accurately.
A total of 94 different subtype cases were analysed. miRNA microarray analysis was performed on fresh frozen tissues of three common RCC subtypes (clear cell, chromophobe, and papillary) and on oncocytoma. Results were validated on the original as well as on an independent set of tumours, using quantitative reverse transcription-polymerase chain reaction (qRT-PCR) analysis with miRNA-specific primers.
Microarray data were analysed by standard approaches. Relative expression for qRT-PCR was determined using the ΔΔCT method, and expression values were normalised to small nucleolar RNA, C/D box 44 (SNORD44, formerly RNU44). Experiments were done in triplicate, and an average was calculated. Fold change was expressed as a log2 value. The top-scoring pairs classifier identified operational decision rules for distinguishing between different RCC subtypes and was robust under cross-validation.
We developed a classification system that can distinguish the different RCC subtypes using unique miRNA signatures in a maximum of four steps. The system has a sensitivity of 97 % in distinguishing normal from RCC, 100 % for clear cell RCC (ccRCC) subtype, 97 % for papillary RCC (pRCC) subtype, and 100 % accuracy in distinguishing oncocytoma from chromophobe RCC (chRCC) subtype. This system was cross-validated and showed an accuracy of about 90 % . The oncogenesis of ccRCC is more closely related to pRCC, whereas chRCC is comparable with oncocytoma. We also developed a binary classification system that can distinguish between two individual subtypes.
MiRNA expression patterns can distinguish between RCC subtypes.