The Value Of Circulating Tumor DNA In Prognosis Of Patients With Resectable Esophageal Cancer

We investigated the characteristics of gene mutation and prognosis in patients with resectable esophageal squamous carcinoma(ESCC),and established nomogram models to evaluate lymph node metastasis(LNM),progression-free survival(DFS)and overall surviva l(OS)of ESCC patients.A total of 57 ESCC patients completed 5-year follow-up,and tissue samples,white blood cells,preoperative cell-free DNA(cfDNA)and postoperative cfDNA were detected by next-generation sequencing of 61 genes.Clinicopathological characteristics,mutational features and survival analysis were analyzed.Independent risk factors of LNM were assessed by comparisons among groups and logistic regression analysis.Univariate and multivariate Cox analysis were used to determine the independent prognostic factors of ESCC.Nomogram models were constructed by independent risk factors,and the predictive power of nomogram was evaluated by consistency index,calibration plots and receiver operating characteristic curves.The results indicate that TP53 and PIK3 CA genes were the most predominant driver gene mutations of ESCC.The sensitivity of ctDNA detection was 67.1%(49/7)based on the results of tissue mutation test.Variants with higher mutant allele frequencies(MAF)were more likely to be detected in paired tissue samples,and no statistical differences in the number of variants,MAF or mutation concordance between tissue and ctDNA were found in samples of different stages.Tissue mutation,pre-surgical ctDNA alteration,and expression level of tissue TP53 protein did not affectthe prognosis of ESCC patients.Among ESCC patients who did not receive adjuvant therapy,minimal residual disease(MRD)analysis indicated that postoperative Tumor-MRD+ patients had a higher risk of recurrence.Postoperative ctDNA-MRD+ patients have reduced DFS when at least two ctDNA molecules are taken as MRD criteria.The nomogram model of LNM was established by four factors: vascular lymphatic infiltration,submucosal infiltration depth,drinking history and average MAF of ctDNA variants.The area under the curve(AUC)of the LNM model was 0.84(95%CI:0.74-0.95),the sensitivity,specificity and accuracy of prediction was 82.1%,73.1% and 77.8%,respectively.Tumor location,vascular lymphatic infiltration,adjuvant treatment,number of preoperative ctDNA variants,and mean MAF of ctDNA variants were independent risk factors of ESCC DFS.The AUC of 2-year and 5-year DFS was predicted to be 0.93(95%CI:0.87-0.99)and 0.88(95% CI:0.80-0.97),respectively.The sensitivity and specificity for predicting 2-year recurrence were 95.24%,80.56% and 77.78%,95.24% for predicting5-year recurrence.Tumor site,vascular lymphatic infiltration,adjuvant treatment,number of preoperative ctDNA variants,and mean ctDNA abundance were independent risk factors associated with OS.The predicted AUC values of OS in 2-year and 5-year were 0.93(95%CI: 0.87-0.99)and 0.85(95% CI: 0.76-0.95),respectively.The sensitivity,specificity,and accuracy for predicting 2-year survival were 100%,78.57%,84.21%,and 79.41%,82.61%,80.7% for 5-year survival.Calibration curves demonstrated the consistency between the actual outcome and predicted results in the three nomogram models.The established nomogram models can be considered as a personalized quantitative tool for predicting LNM and prognosis of patients with ESCC,so as to help clinicians make treatment decisions.