Apoptosis-related Genes Postoperative Disease Progression of Esophageal Cancer and the Construction and Validation of Its Prediction Model Based on the TCGA Database Screening

WANG Ronghua; LIAO Zefei; WANG Gengjie; ZHANG Le

doi:10.13730/j.issn.2097-2148.2025.09.007

Abstract：Objective To screen apoptosis-related genes associated with postoperative disease progression of esophageal cancer based on the cancer genome atlas(TC-GA)database,construct a prediction model for postoperative disease progression of esophageal cancer,and validate the value of the predictive model.Methods Gene expression data of esophageal cancer tumor tissues and normal tissues,as well as postoperative disease progression status of patients,were obtained from TCGA database,and apoptosis-related genes with differential expression of esophageal cancer tumor tissues and normal tissues were screened out;the least abso-lute shrinkage and selection operator(LASSO)regression and Cox regression were used to screen key apoptosis-related genes associated with postoperative progression of esophageal cancer,and the disease progression risk model was con-structed,the validity and accuracy of the model were analyzed by Kaplan-Meier survival curve,risk curve and receiver op-erating characteristic(ROC)curve.Finally,47 esophageal cancer patients from the author's hospital were enrolled for validation.Results A total of 81 esophageal cancer tissue samples and 26 normal esophageal tissues were obtained from TCGA database,and 33 differentially expressed apoptosis-related genes in esophageal cancer tumor tissues and normal esophageal tissues were obtained.Twenty-nine apoptosis-related genes related to postoperative disease progression in pa-tients with esophageal cancer were screened out by Cox regression analysis,including[Caspase-1(CASP1),CASP3,CASP4,CASP5,CASP6,CASP8,CASP9,Gasdermin A(GSDMA),GSDMB,GSDMC,GSDMD,GSDME,interleukin 18(IL-18),IL-1B,NLR family CARD domain-containing protein 4(NLRC4),NLR family pyrin domain-containing 1(NLRP1),NLRP2,NLRP3,NLRP7,nucleotide-binding oligomerization domain-containing protein 1(NOD1),phospho-lipase C gamma 1(PLCG1),protein kinase cAMP-activated catalytic subunit alpha(PRKACA),PYD and CARD do-main-containing protein(PYCARD),SR-related CTD-associated factor 11(SCAF11),neutrophil elastase(ELANE),IL-6,NLRP6,tumor necrosis factor(TNF),glutathione peroxidase 4(GPX4)];15 apoptosis-related genes most strongly as-sociated with postoperative disease progression in esophageal cancer patients were screened out using LASSO regression,including CASP3,GSDMB,GSDMC,NLRC4,NLRP3,NLRP7,NOD1,PLCG1,PRKACA,SCAF11,ELANE,IL-6,NLRP6,TNF and GPX4,among these,11 genes(IL-6,CASP3,GSDMB,GSDMC,NLRC4,NLRP3,NLRP7,NOD1,PLCG1,PRKACA and SCAF11)were identified as risk factors,while 4 genes(ELANE,NLRP6,TNF,GPX4)were classified as protective factors;Kaplan-Meier curve analysis revealed that the overall survival time of patients in the high-risk group was significantly shorter than that in the low-risk group(P＜0.05);the ROC curve analysis demonstra-ted that the area under the curve(AUC)for predicting postoperative disease progression risk in esophageal cancer was 0.773(95％CI:0.633-0.879);risk curve analysis indicated that,as the risk score increased,the survival time of pa-tients in the high-risk group decreased more significantly compared to the low-risk group,and the number of death also in-creased.Further validation of the model's predictive efficacy was performed using 47 esophageal cancer patients,and the results showed the AUC of 0.754(95％CI:0.611-0.864).Conclusion The risk prediction model for postoperative dis-ease progression of esophageal cancer constructed based on apoptosis-related genes can effectively predict the risk of post-operative disease progression in esophageal cancer patients,and may be used in clinical practice for identifying high-risk patients with postoperative disease progression of esophageal cancer and guiding their treatment.