Integration of NRP1, RGS5, and FOXM1 expression, and tumour necrosis, as a postoperative prognostic classifier based on molecular subtypes of clear cell renal cell carcinoma.

Molecular mechanisms of progression of clear cell renal cell carcinoma (ccRCC) have been proven with recent genomic or transcriptional analyses. However, it is still difficult to apply these analyses to daily clinical practice owing to economical and practical issues. Here, we established a pathology-based, postoperative prognostic classification based on the well-validated transcriptional classifier, ClearCode34, in ccRCC. A total of 342 cases with available tissue were identified and randomly allocated into a discovery cohort (n = 138) and a validation cohort (n = 204). Levels of mRNA were quantified using a nCounter Digital Analyzer, and the ccA/ccB subtypes were determined. Histological and immunohistochemistry (IHC) analyses were subsequently performed to establish a pathology-based classification based on the mRNA levels. Finally, the prognostic ability of the new classifier was evaluated in both the discovery and validation cohorts. Of 138 cases in the discovery cohort, 78 (56.5%) and 60 (43.5%) were assigned to the ccA and ccB subtypes, respectively. Proangiogenic genes, neuropilin 1 (NRP1) and regulator of G protein signalling 5 (RGS5), were especially overexpressed in all ccRCC samples and were enriched in ccA-assigned tumours. Histologically, tumour necrosis and the sarcomatoid feature were associated with the ccB subtype. In IHC analyses, expression of NRP1, RGS5, and forkhead box M1 (FOXM1), an epithelial-mesenchymal transition-related factor, significantly correlated with the ccA/ccB subtypes. Combining these three IHC factors and tumour necrosis, we developed the IHC/histology-based classifier, which showed good concordance with the ClearCode34 classifier with an accuracy of 0.80. The established classification significantly stratified relapse-free, cancer-specific, and overall survival rates in both the discovery and validation cohorts. The novel molecular pathology classifier integrating NRP1, RGS5, FOXM1, and tumour necrosis may enable the stratification of oncological outcomes for patients with ccRCC undergoing resection surgery.