Application Of Maternal Plasma Cell-free DNA Sequencing In Non-invasive Prediction Of Macrosomia

Background and ObjectivesFetal macrosomia is common pregnancy complication,and it occurred in about 7%of all pregnancy in China.It is associated with the increased risk of several short-term adverse outcomes,including increased chance of cesarean delivery or vaginal midwifery,shoulder dystocia,brachial plexus injury,perinatal asphyxia and neonatal mortality.Furthermore,a large body of literatures have reported the associations between macrosomia and long-term health risks,such as adult obesity,diabetes,cardiovascular disease and cancers.Therefore,the early diagnosis might contribute to successfully pregnancy management.Prevalence of fetal macrosomia is increasing globally over the past several decades,however,there is no reliable methods by now.Cell-free DNA(cfDNA)has been first reported in lupus patients by Tan in 1966.In 1997,cell-free fetal DNA(cffDNA)has been first discovered by Lo in the maternal circulating blood.As merely approximately 10-15%of total cfDNA was derived from placenta with the rest majorly derived from maternal myeloid/lymphoid cells lineages.Several studies showed that cfDNA fragments are nucleosome-bound DNA fragments,preferentially survive digestion and are released into the circulation.Moreover,nucleosome have distinct distribution characteristics near the transcription start sites(TSS),which can reflect the gene expression level.To strengthen pregnancy management and reduce the incidence of cesarean section,we developed a non-invasive prediction classifier of fetal macrosomia by the whole-genome nucleosome footprint profiling of maternal plasma cell-free DNA based on low-coverage next generation sequencing and bioinformatics.MethodsIn this nested case-control study,the participants were enrolled at three independent hospitals,including Nanfang Hospital,The Third Affiliated Hospital of Sun Yat-sen University,and Cangzhou People's Hospital.All plasma samples was collected at 12+0-27+6 weeks of gestation,which were performed low-coverage whole-genome sequencing of cfDNA.At the discovery stage,we selected 47 cases of fetal macrosomia and 47 gestational age-matched controls.Whole-genome sequencing of cfDNA was then performed on each of these samples.Nucleosome footprint patterns at primary transcription start sites(pTSS)were compared between groups,and a p-value was calculated using the Wilcoxon rank sum test.P-value was then adjusted to the false discovery rate(FDR)using the Benjamini-Hochberg procedure.Gene transcripts with FDR<0.05 and fold change?1.3 were considered to have significant differential coverages at the pTSS.At the training stage,a stepwise method for feature selection and logistic regression analysis were used to select the genes combinations to construct classifiers.The robustness of these classifiers was assessed using the leave-one-out cross validation method.Receiver operating characteristic(ROC)analysis was used to evaluate the performance of each classifier,including area under curve(AUC),accuracy,sensitivity,and specificity.The classifiers which performed well and displayed the largest AUC in the training cohort were chosen as the optimal classifiers for each pregnancy complication.In validation stage,the performance of classifier was then further validated using four independent validation cohorts,including one internal cohort,two external cohorts and all subjects.ResultsIn this study,810 participants were included,consisted by 162 pregnancies who developed macrosomia(macrosomia group,n=162)and 648 healthy controls,in the proportion of 1:4 matching the gestational weeks of blood collection with macrosomia group(control group,n=648).Total of 1086 significantly different TSSs regions were found between macrosomia and healthy controls(p<0.05),with 575 up-regulated genes promotors and 511 down-regulated genes promotors.A"macrosomia classifier"(C MA-A)classifier include 12-genes(SMC3,MASTL,CREM,C1QTNF12,MLXIP,MAP3K9,IGSF6,APC2,GPM6A,TMEM128,NIPBL,TMEM184A)with an AUC of 0.766(95%CI:0.678-0.854)were developed.CA had an AUC of 0.817(95%CI:0.689-0.945),0.791(95%CI:0.721-0.861),0.762(95%CI:0.675-0.848)and 0.779(95%CI:0.736-0.823),in the internal validation cohort,two external validation cohorts,and all subjects,respectively.ConclusionIn summary,this study suggest that nucleosome footprint profiling based classifiers provide high predictive capabilities for predicting fetal macrosomia at early gestational age.The techniques required for low-coverage DNA sequencing without additional tests,are easily applicable to routinely NIPT data,with clinical application prospect;it may contribute to successfully pregnancy management and choice of appropriate delivery mode.