A radiotherapy resistance-related prognostic signature predicts survival and the immune landscape in rectal cancer

Background: Radiotherapy resistance remains a significant challenge in the management of locally advanced rectal cancer (LARC). To date, no universally applicable and reliable prognostic marker has been established for clinical practice. Thus, reliable biomarkers need to be identified and the molecular mechanisms underlying radiotherapy resistance need to be investigated to improve patient prognosis.

Methods: Multiple independent sources of transcriptomic datasets from The Cancer Genome Atlas (TCGA) (training cohort, n=154) and the Gene Expression Omnibus (GEO), including GSE35452 (containing radiotherapy response and non-response cohorts) and GSE87211 (validation cohort), were systematically integrated. A prognostic signature was developed by identifying radiotherapy-related genes through differential expression analysis and weighted gene co-expression network analysis (WGCNA), followed by least absolute shrinkage and selection operator (LASSO) and multivariate Cox regression analyses. Functional enrichment, immune microenvironment characterization, and single-cell RNA sequencing (scRNA-seq) analyses were subsequently performed to explore the mechanisms underlying radiotherapy resistance.

Results: A four-gene prognostic signature comprising CUTA, IZUMO2, PALB2, and PSCA was established. The signature effectively stratified patients into high- and low-risk groups with distinct overall survival (OS) outcomes (TCGA: P<0.001; GSE87211: P=0.002) and served as an independent prognostic factor [multivariate Cox hazard ratio (HR) =2.532, P<0.001]. The high-risk group exhibited distinct immune environment characteristics and enrichment of epithelial-mesenchymal transition (EMT) pathways compared to the low-risk group. The scRNA-seq analysis revealed that PSCA expression was restricted to an epithelial subpopulation characterized by enhanced cell-cell communication and a more advanced pseudotime trajectory.

Conclusions: This study developed and validated a four-gene signature for survival prediction in rectal cancer (RC). Functional enrichment and immune microenvironment characterization analyses indicated that the signature was associated with tumor heterogeneity and differential treatment responses in RC. The single-cell analysis indicated that PSCA may be a key gene in this process.