Molecular Stratification of Renal Cancer Reveals Prognostic Biomarkers and Therapeutic Pathways

Keywords: Renal cancer subtypes, Prognostic biomarkers, Therapeutic pathways

TL;DR: We identified two prognostically distinct KIRC subtypes using a 76-gene signature, revealing pathway-level biomarkers and therapeutic targets with strong clinical relevance.

Abstract: Kidney renal clear cell carcinoma (KIRC) is the most common subtype of renal cancer, yet the discovery of robust prognostic biomarkers has been hindered by its profound molecular heterogeneity, complex tumor microenvironment, and metabolic rewiring. Here, we present an integrative transcriptomic analysis of 314 KIRC patients to uncover molecular subtypes and biomarker signatures with clinical relevance. Using an iterative survival-guided feature selection approach, we refined 1,000 highly variable genes into a compact 76-gene signature that enabled unsupervised clustering into two prognostically distinct subgroups. Patients in the high-risk subgroup exhibited significantly poorer overall survival (log-rank p = $4.5 \times 10^{-4}$) and elevated event rates compared to the low-risk group. Differential expression analysis revealed 2,927 subtype-specific genes, of which 70\% demonstrated significant associations with survival in univariate Cox regression. Functional enrichment highlighted convergence on cancer-associated pathways, including TOR signaling, regulation of macroautophagy, and negative regulation of catabolic processes, implicating both canonical oncogenic drivers (e.g., PIK3CA, EIF4EBP2, PRKAA2) and modulators of cellular homeostasis (e.g., UBR1, MTM1). Together, these findings establish a refined prognostic biomarker framework for KIRC, define clinically relevant molecular subtypes, and reveal pathway-level vulnerabilities that may be exploited for therapeutic intervention.

Submission Number: 128