Data Mining And Validation Of CDC20 And MAD2 As Potent Biomarkers In Endometrial Carcinoma

Background:Endometrial carcinoma(EC)is a group of malignant tumors that occur in endometrium,which is one of the three major malignant tumors in female reproductive system.The etiology of EC stay unclear,to date there may be two kind of mechanism.The type I is estrogen-dependent EC,which includes the majority of the EC.And the pathological type of this kind of EC is endometrium endometrial carcinoma(EEA),which differentiates well and possesses estrogen receptor and progesterone receptor.The prognosis of it is relatively good.There are over 50%of this kind of EC exist PTEN gene mutation or inactivation.Type II is estrogen-independent EC,of which pathogenesis have no clear evidence to do with estrogen.There are evidences proving that this kind of EC are correlation with anti-oncogene p16 inactivation,E-cadherin inactivation,HER2/neu overexpression,p53 gene mutation.The pathological of type II EC include serous endometrial adenocarcinoma(SEA),clear cell carcinoma and mucous adenocarcinoma,which differentiate badly and appear negative estrogen receptor and progesterone receptor.The prognosis of them are bad.The prognosis of late stage endometrial carcinoma,especially serous endometrial adenocarcinoma,remains poor worldwide,new prognostic biomarkers are demand.Method:We used data that downloaded from official TCGA database by The Genomic Data Commons(GDC)Data Portal to analyze the data of 23 peritumoral tissues and 534 tumor tissues,and to screened 20860 gene as well as conducted GSEA functions and pathways enrichment analysis.Then,based on the expression and survival data of every single patient,multivariate-COX regressions were conducted.Regression coefficients were then used to build a prognostic model.In this model,the risk-score of every single patient was calculated.Finally,tissue arrays are mined and validated by immunohistochemistry.Result:CDC20 ranked the biggest difference between peritumoral and tumor tissues among all the 20680 genes screened(p<0.0001).MAD2 also showed significant difference between peritumoral and tumor tissues(p<0.0001).Survival analyses suggested that high expression status of these two genes could both result in poor survival outcomes[CDC20:HR = 2.084,95%CI(1.318-3.293),MAD2:HR=1.838,95%CI(1.184-2.852)].In the prognostic model we constructed,patients with high risk scores had worse OS than lower ones[HR=1.713,95%CI(1.17-2.626)].High-risk patients with either endometrioid endometrial adenocarcinoma(EEA)or SEA suffered similar unsatisfied outcomes(p=0.577).Conclusion:In conclusion,CDC20/MAD2 both highly expressed in EC and high level of these two genes could cause unsatisfied prognosis either in EEA and SEA.