Transcriptomic profiling of psychological stress-related genes uncovers prognostic subtypes in colorectal cancer.

BACKGROUND

Colorectal cancer (CRC) represents a major malignancy of the gastrointestinal tract with substantial global morbidity and mortality. Mounting evidence underscores the pivotal role of chronic psychological stress in the pathogenesis of gastrointestinal cancers. Psychological stress-related genes have been implicated in modulating tumor progression and represent promising prognostic biomarkers. Nevertheless, their precise prognostic significance in CRC remains incompletely characterized.

METHODS

We initially identified prognostic genes through univariate Cox regression analysis of The Cancer Genome Atlas (TCGA) dataset. An optimal prognostic signature was subsequently constructed using least absolute shrinkage and selection operator (LASSO) regression and multivariate Cox analysis. Predictive accuracy was assessed via Kaplan-Meier survival analysis, receiver operating characteristic (ROC) curve analysis, and calibration curves. Furthermore, immune microenvironment characterization and chemotherapeutic response profiling were performed across risk-stratified subgroups to elucidate underlying biological mechanisms.

RESULTS

Of 374 psychological stress-related genes examined, 62 demonstrated significant associations with overall survival (OS). The resulting 11-gene signature achieved area under the curve (AUC) values of 0.711, 0.768, and 0.800 for 1-, 3-, and 5-year OS prediction in the TCGA training cohort, respectively. External validation in three GEO cohorts showed attenuated but reproducible prognostic stratification, with AUCs ranging from approximately 0.56 to 0.63. High-risk patients exhibited significantly reduced survival probability. Computational analyses suggested a higher estimated M2 macrophage fraction and reduced predicted sensitivity to oxaliplatin.

CONCLUSION

We developed and externally evaluated an 11-gene transcriptomic signature related to stress-associated tumor biology in CRC. The signature stratified survival in retrospective cohorts, although external discrimination was moderate. This signature should be regarded as an exploratory research tool rather than a clinically validated assay. Computational analyses generated hypotheses regarding immune contexture and oxaliplatin sensitivity, and UBP1 emerged as a candidate tumor-suppressive regulator. Prospective multicenter validation and mechanistic studies are required before clinical application.