Prognostic Factors Of Survival And Biological Mechanisms Of Ovarian Cancer Based On Weighted Cox Regression

ObjectiveBased on the SEER database,the prognostic factors of survival in ovarian cancer patients were studied,the significance of tumor stage for the survival of ovarian cancer patients was clarified,and the differences of prognostic factors of survival between early-and advanced-stage ovarian cancer patients was discussed;based on the TCGA database,the differentially expressed genes and the possible differences in molecular biological mechanisms between early-and advanced-stage ovarian cancer were discussed,in order to provide theoretical basis and research directions for early diagnosis and molecular targeted treatment of ovarian cancer patients.MethodsThe data of ovarian cancer patients from 2004 to 2016 in the SEER database was collected retrospectively,the t test and chi-square test were used to compare the clinical characteristics between early-and advanced-stage ovarian cancer patients,the Kaplan-Meier method was used to compare the overall survival rate between early-and advanced-stage ovarian cancer patients,and weighted Cox regression model was used to study the prognostic factors of survival in all patients with ovarian cancer,early-and advanced-stage patients with ovarian cancer.Based on the mRNA expression data of ovarian cancer in the TCGA database,the differentially expressed genes of early-and advanced-stage ovarian cancer were analyzed by edge R package.Gene ontology(GO)analysis and Kyoto encyclopedia of genes and genomes(KEGG)pathway enrichment analysis for the differentially expressed genes were performed in the DAVID 6.8 database.The protein-protein interaction network constructed by differential genes was identified using the STRING database.The top 10 hub genes,which connected most with other genes,was selected.Results1.Multivariate weighted Cox regression showed that tumor stage was the prognostic factors of survival in ovarian cancer patients(AHR:3.078,95%CI:2.926～3.238,P<0.001).Histologic classification(AHR:0.807,95%CI:0.736～0.885,P<0.001),laterality(AHR:1.239,95%CI:1.196～1.283,P<0.001),tumor grade(moderately differentiated:AHR:1.595,95%CI:1.435～1.773,P<0.001;poorly differentiated/undifferentiated:AHR:1.767,95%CI:1.610～1.938,P<0.001),resection surgery(AHR:0.351,95%CI:0.330～0.373,P<0.001),lymphadenectomy(AHR:0.727,95%CI:0.704～0.751,P<0.001)and preoperative serum CA125(AHR:1.437,95%CI:1.333～1.549,P<0.001)were the prognostic factors of survival in ovarian cancer patients.2.There were differences in age,laterality,tumor grade and preoperative serum CA125 between early-and advanced-stage ovarian cancer patients.The 1-,3-,and 5-year survival rates of early-stage ovarian cancer patients were 95.4%,88.2%,and 82.1%,respectively.The 1-,3-,and 5-year survival rates of advanced-stage ovarian cancer patients were 79.5%,51.0%,and 32.9%,respectively.Histologic classification was the prognostic factors of survival in early-stage ovarian cancer patients(AHR:0.713,95%CI:0.606～0.839,P<0.001),but not the prognostic factors of survival in advanced-stage ovarian cancer patients(AHR:0.950,95%CI:0.984～1.055,P=0339).3.309 up-regulated genes and 248 down-regulated genes were found between early-and advanced-stage ovarian cancer.GO analysis showed that the up-regulated differential genes were significantly enriched in GO functions related to extracellular components such as extracellular matrix organization,extracellular region and extracellular matrix structural constituent,while the down-regulated differential genes were significantly enriched in GO functions related to extracellular components such as extracellular region and extracellular space,as well as GO functions related to chemotaxis such as chemokine activity and chemotaxis.KEGG analysis showed that the up-regulated differential genes were significantly enriched in signaling pathways related to energy metabolism such as protein digestion and absorption,while the down-regulated differential genes were significantly enriched in signaling pathways such as cytokine-cytokine receptor interaction.The top 10 hub genes in protein-protein interaction network were GNG11 FPR2、PMCH、PPBP、PF4、AGTR1、CXCL8、GAL、NPY and PYY.ConclusionsEarly-and advanced-stage ovarian cancer had differences in the prognostic factors of survival as well as biological mechanisms,which may provide theoretical basis and research directions for early diagnosis and molecular targeted treatment of ovarian cancer.