A human adenovirus C infection-related gene panel for predicting survival and treatment responsiveness in glioma patients.

BACKGROUND: Viruses are critical for the regulation of cancer development as well as therapy. Human adenovirus C (HadVC) has been detected in central nervous system and glioma tissue. The objective of current study was the development of a robust prognostic model based on HadVC infection (HadVCi)-relevant genes.

METHODS: The genome, transcriptome, and virome were systemically analyzed using TCGA dataset for training and two cohorts from CGGA and an immunotherapy trial cohort with 17 patients receiving anti-PD-1 treatment for validation. HadVCi-relevant gene selection from differentially expressed genes (DEGs) between HadVC-infected or non-infected glioma patients using LASSO regression was followed by Cox regression modeling to establish a prognostic HadVCi score. Kaplan-Meier and ROC curve analyses were performed to estimate the predictive capacity of HadVCi score. Chi-square/Spearman/Mann Whitney U test is utilized to identify its correlation to clinicopathological parameters, treatment responsiveness, and immune landscape. TMZ-resistant glioma cells were established and analyzed at the transcriptional level using RNA-seq data.

RESULTS: HadVCi score =(-0.2526673*TRPC6) + (-0.2244276*RNF207) + (-0.0894468*SEC31B) + (-0.0190214* ZCRB1) + (-0.017122*DNPH1) + (0.0495818*CCDC34) + (0.1196349*PURG) + (0.1778997*LILRA5). The score possesses a strong ability to predict overall survival. Further analysis revealed higher HadVCi score is correlated with malignant phenotype, and poorer treatment responsiveness, such as TMZ-based chemotherapy and combined therapies. Additionally, transcriptomic analysis showed malignancy-, stemness-, and radioresistance-related gene activation in HadVCi group, which characterized the poor outcomes and limited sensitivity to standard therapy.

CONCLUSIONS: The HadVCi score may be an effective tool for survival prediction and treatment guidance in patients with glioma.