Transcriptomic profiling to identify cancer–immune meta-programs in penile squamous cell carcinoma.

Abstract

e17035 Background: Penile squamous cell carcinoma (PSCC) is a rare malignancy with substantial heterogeneity in tumor-intrinsic biology & the immune microenvironment. While genomic studies have identified recurrent alterations, how cancer & immune programs jointly shape PSCC ecosystems & outcomes remain poorly defined. Methods: We profiled 67 PSCC tumors using the nCounter PanCancer Immune Profiling Panel, interrogating cancer-relevant genes & curated immune-response pathways. The cohort included 29 HPV-positive tumors (43.3%) & was balanced across stage (33 stage I–II; 34 stage III–IV). Highly variable genes (HVGs) were identified, followed by unsupervised clustering, pathway-based analyses using 61 curated canonical gene sets, & survival analyses for recurrence-free survival (RFS) & overall survival (OS). To model joint cancer–immune ecosystems, non-negative matrix factorization (NMF) was applied to derive cancer–immune meta-programs with program-level signals interpreted using independent single-cell RNA sequencing data. Results: HVG analysis revealed transcriptional variability spanning inflammatory & myeloid-associated genes, interferon & antigen-presentation genes, tumor-intrinsic epithelial & cancer-testis antigens. Unsupervised expression-based clustering identified four major transcriptional clusters that were not strongly associated with any clinical covariates. Pathway-level analyses demonstrated distinct transcriptional programs associated with stage, HPV status and smoking history. Survival analyses identified Th17 signaling & canonical PI3K signaling associated with RFS in univariable analyses, canonical PI3K signaling & cytotoxic cell signatures remaining associated with RFS in multivariable models. NMF resolved this heterogeneous landscape into three orthogonal cancer–immune meta-programs that captured coordinated transcriptional structure across tumors. Program 1 was enriched for interferon-α & complement pathways, Program 2 for interferon-γ signaling, & Program 3 for E2F targets & epithelial–mesenchymal transition. These meta-programs showed clear prognostic stratification with Program 2 associated with favorable RFS and OS, whereas Program 3 associated with adverse outcomes. Integration with single-cell PSCC data demonstrated preferential enrichment of Program 1 in myeloid compartments, Program 2 across B cells, myeloid cells, and subsets of T cells, & Program 3 predominantly within epithelial, fibroblast, and endothelial compartments. Conclusions: Transcriptomic analysis in PSCC identifies coordinated cancer–immune meta-programs that define biologically distinct tumor ecosystems & stratify clinical outcomes beyond conventional clinical features supporting a promising framework for understanding penile cancer molecular heterogeneity, motivating ecosystem-informed biomarker & therapeutic strategies.