Construction and validation of a regulatory T cells-based classification of renal cell carcinoma: an integrated bioinformatic analysis and clinical cohort study

Renal cell carcinoma (RCC), exhibiting remarkable heterogeneity, can be highly infiltrated by regulatory T cells (Tregs). However, the relationship between Treg and the heterogeneity of RCC remains to be explored.

We acquired single-cell RNA-seq profiles and 537 bulk RNA-seq profiles of TCGA-KIRC cohort. Through clustering, monocle2 pseudotime and prognostic analyses, we identified Treg states-related prognostic genes (TSRPGs), then constructing the RCC Treg states-related prognostic classification (RCC-TSC). We also explored its prognostic significance and multi-omics landmarks. Additionally, we utilized correlation analysis to establish regulatory networks, and predicted candidate inhibitors. More importantly, in Xinhua cohort of 370 patients with kidney neoplasm, we used immunohistochemical (IHC) staining for classification, then employing statistical analyses including Chi-square tests and multivariate Cox proportional hazards regression analysis to explore its clinical relevance.

We defined 44 TSRPGs in four different monocle states, and identified high immune infiltration RCC (HIRC, LAG3+, Mki67+) as the highly exhausted subtype with the worst prognosis in RCC-TSC (p < 0.001). BATF-LAG3-immune cells axis might be its underlying metastasis-related mechanism. Immunotherapy and inhibitors including sunitinib potentially conferred best therapeutic effects for HIRC. Furthermore, we successfully validated HIRC subtype as an independent prognostic factor within the Xinhua cohort (OS, HR = 16.68, 95% CI = 1.88–148.1, p = 0.011; PFS, HR = 4.43, 95% CI = 1.55–12.6, p = 0.005).

Through integrated bioinformatics analysis and a large-sample retrospective clinical study, we successfully established RCC-TSC and a diagnostic kit, which could stratify RCC patients with different prognosis and to guide personalized treatment.