An immunotherapy response prediction model derived from proliferative CD4+ T cells and antigen-presenting monocytes in ccRCC.

Most patients with clear cell renal cell carcinoma (ccRCC) have an impaired response to immune checkpoint blockade (ICB) therapy. Few biomarkers can predict responsiveness, and there is insufficient evidence to extend them to ccRCC clinical use. To explore subtypes and signatures of immunocytes with good predictive performance for ICB outcomes in the ccRCC context, we reanalyzed two ccRCC single-cell RNA sequencing (scRNA-seq) datasets from patients receiving ICB treatment. A subtype of proliferative CD4+ T cells and regulatory T cells and a subtype of antigen-presenting monocytes that have good predictive capability and are correlated with ICB outcomes were identified. These findings were corroborated in independent ccRCC ICB pretreatment bulk RNA-seq datasets. By incorporating the cluster-specific marker genes of these three immunocyte subtypes, we developed a prediction model, which reached an AUC of 93% for the CheckMate cohort (172 samples). Our study shows that the ICB response prediction model can serve as a valuable clinical decision-making tool for guiding ICB treatment of ccRCC patients.