Work

442 Characterization of transcriptomic programs and rare somatic mutations affecting survival and immune checkpoint inhibitor (ICI) response in clear-cell renal cell carcinoma (ccRCC)

Year: 2026

Type: conference-abstract

Abstract: Background Although ICI therapy is a mainstay of ccRCC management, molecular determinants of resistance to ICIs in ccRCC remain incompletely understood. Prior work has linked common genomic alterations and transcriptomic subsets to modality-specific therapy response and survival, but these analyses do not capture the effects of rare mutations and the interplay of transcriptional programs within tumors. Through large-scale harmonized genomic analyses of over 2000 ccRCC tumors with detailed clinical annotations, we investigated whether low-frequency somatic alterations and network-level gene programs can elucidate novel biomarkers for ICI.Methods Uniform processing and quality control of ccRCC samples from 12 cohorts (including 8 with ICI-treated patients) yielded 2026 bulk RNA sequencing profiles and 1809 tumor-normal whole-exome sequencing pairs. Significantly mutated genes (SMGs) were identified using MutSig2CV, MutPanning and OncodriveFML. RNA-seq data were batch-corrected using ComBat before ssGSEA pathway enrichment analysis and CIBERSORTx-based immune deconvolution. Co-expression network inference identified nine mutually co-expressed gene modules ( figure 1A), whose activity we scored per sample with UCell. Associations between mutations, immune composition, module scores, and progression-free survival (PFS) were tested across ICI and tyrosine kinase inhibitor (TKI) therapies (figure 1B-C).Results We identified 118 high-confidence SMGs, including canonical ccRCC alterations like VHL (68%), PBRM1 (39%), and SETD2 (19%), and 72 rare alterations occurring in <2% of tumors. These SMGs converged on chromatin remodeling, transcriptional regulation, and cell-cycle pathways. We recapitulated known associations between genomic variants, survival outcomes, and inferred immune composition. Rare (0.8%) MLH1-mutant tumors showed increased TMB/MSI and a trend toward improved PFS with ICI but worse outcomes with TKI. Conversely, alterations in B2M (0.6%) and STK11 (0.4%), which have been linked to ICI resistance in other malignancies, exhibited immune-related transcriptional changes, with STK11 loss notably linked to shorter PFS on immunotherapy. The network modules also captured known ccRCC biology, including proliferative, fibroblast, angiogenesis, and immunity programs ( figure 1A); the former two were associated with shorter PFS and specifically enriched in progressive disease for both ICI and TKI compared to responders and stable disease patients (figure 1B-C). Finally, while the angiogenesis module was specifically predictive of TKI response, the lymphoid and myeloid modules were predictive of longer PFS only for ICI-based regimens (figure 1C).Conclusions This large clinico-genomic analysis of ccRCC identified clinically relevant rare mutations and novel, interpretable transcriptional modules linking molecular features to tumor biology and modality-specific outcomes. Through this, our analyses enable finer prognostication and suggest candidates for biomarker development and precision oncology.Abstract 442 Figure 1 Interpretable gene programs discovered by gene co-expression network analysis. (A) Module annotations and key genes; (B) Scaled module scores across groups; red tiles indicate significant Wilcoxon tests for response vs. resistance; (C) Cox regression log-hazard ratios for PFS vs. module scores, age, sex and sarcomatoid histology

Source: Journal for ImmunoTherapy of Cancer

DOI: 10.1136/jitc-2026-sitc2026.0442