A Novel Prognostic Risk Model Based on Oxidative Stress to Predict Survival and Improve Treatment Strategies in Stomach Adenocarcinoma.

Abstract

Background: Stomach adenocarcinoma (STAD) is the fifth most common tumor worldwide, imposing a significant disease burden on populations, particularly in Asia. Oxidative stress is well-known to play an essential role in the occurrence and progression of malignancies. Our study aimed to construct a prediction model by exploring the correlation between oxidative stress-related genes and the prognosis of patients with STAD.

Method: STAD data from TCGA were used to identify the differentially expressed oxidative stress-related genes (OSGs), with data from GEO serving as the validation cohort. Univariate Cox and LASSO regression analyses were performed to select prognosis-related genes for the risk model, which was then integrated with clinical features into a nomogram. The physiological functions and pathways of these identified genes were explored using GO and KEGG analyses. After evaluating the prediction value of the prognostic model in the GEO cohort, drug sensitivity and immune infiltration were comprehensively analyzed using R. Expression levels of the prognostic genes were verified by quantitative real-time PCR in gastric cancer and paired normal tissues.

Results: Cox regression and LASSO regression analysis identified SERPINE1, VHL, CD36, NOS3, ANXA5, ADCYAP1, POLRMT and GPX3 as the signature genes from 160 differentially expressed OSGs. Both Kaplan-Meier survival analysis and ROC curve at 5 years in the TCGA and the GEO cohort exhibited great predictive ability of the prognostic model, with the AUC >0.7 in TCGA. Validated as an independent risk factor, the model was integrated with clinicopathological variables (including age, stage, and gender) to build a nomogram for more accurate risk stratification. Moreover, therapy sensitivity analysis between the low- and high-risk categories showed that those who scored higher would benefit more from BEZ235, Dasatinib, Pazopanib, and Saracatinib. Meanwhile, differences in the tumor environment, immune infiltration and response to immunotherapy between the two groups were noted. Finally, qRT-PCR validated the differential expression of these genes in STAD and paired normal tissues.

Conclusion: Our study has effectively established an oxidative stress-related prognostic model, providing a promising tool for personalized clinical strategies and improved STAD patient outcomes.