Research On The Application Value Of Non-invasive Detection In Endometrial Cancer Screening And Conservation Efficacy Predictio

Objectives1.To explore the value of combined BHLHE22 and CDO1 gene methylation detection in screening endometrial carcinoma(EC)for women with EC related high risk factors.2.To investigate the possibility of urine metabolomics as a non-invasive method which can predict the therapeutic efficacy of fertility-sparing treatment for young women with EC and atypical endometrial hyperplasia(AH).Methods1.Exfoliated cells were collected from women with EC related high risk factors in several medical centers.A diagnostic model was constructed by detecting the methylation levels of BHLHE22 and CDO1 genes.The EC screening efficacy of this diagnostic model was assessed.2.Urine samples of young women with EC or AH who received fertility-sparing treatment were collected before treatment and at each timing of efficacy evaluation during the treatment.Urine metabolomics analysis was performed to identify differential urinary metabolites and enriched pathways in different groups.A predictive model was built to predict the therapeutic reaction and its accuracy was assessed.Results1.A total of 1194 cervical scrapings of women with EC related high risk factors from 7 medical centers were collected.Median age of all patients was 44 years old and 62.6%of them had abnormal uterine bleeding.The endometrial histological results were divided into EC(n=128),AH(n=45),benign lesion(n=569),and normal endometrium(n=452).CT values of BHLHE22 and CDO1 gene methylation detection reveals significant differences between the normal endometrium vs.EC,benign lesion vs.EC,and AH vs.EC groups(P<0.05).Receiver operating characteristic(ROC)curve analysis showed that the diagnostic model based on BHLHE22 and CDO1 gene methylation levels had good performance for EC screening,with an area under the curve(AUC)of 0.93(95%CI:0.910.96),sensitivity of 90.6%,specificity of 88.6%,and accuracy of 88.9%.2.A total of 165 urine samples of 129 EC and AH patients undergoing fertility-sparing treatment were collected,including prior treatment(PT,n=50),partial response(PR,n=47),complete response(CR,n=59),stable disease(SD,n=7),and progressive disease(PD,n=2).Urine metabolomics analysis revealed distinct separation trends for urinary metabolites among PT vs.PR,PT vs.CR,and PR vs.CR groups,and weak separation trends among PT vs.PD,PT vs.SD.Further identification of differential urinary metabolites and enriched pathways among PT vs.PR,PT vs.CR,and PR vs.CR groups revealed the highest number of differential urinary metabolites in PT vs.PR group,followed by PT vs.CR group,and the least in PR vs.CR group.The main enriched pathways were related to amino acids,fatty acids,phospholipids,arachidonic acid,retinol,and steroid hormone biosynthesis.Among these,steroid hormone biosynthesis,retinol metabolism,and lysine degradation pathways were shared among these three groups.The therapeutic prediction models built using the metabolites with higher AUC values among the three groups showed good performance,with AUC values of 0.99(95%CI:0.99-1.00),0.91(95%CI:0.89-0.93),and 0.92(95%CI:0.90-0.94),respectively.Conclusions1.The screening model based on combined detection of BHLHE22 and CDO1 gene methylation can serve as the tool for non-invasive early screening of EC,this model has a good screening and diagnostic performance of EC.2.Distinct urine metabolomics characteristics were found at different timing during feritility-sparing treatment for young women with EC/AH.The therapeutic efficacy evaluation models constructed by differential urinary metabolites have good performance,and show promise as a useful,non-invasive tool to evaluate therapeutic efficacy before endometrium biopsy.