| Objective: There is no consensus on the optimal timing of postoperative radiotherapy(PORT)for locally advanced esophageal squamous cell carcinoma(ESCC).We aimed to determine whether the timing of PORT affects the long-term prognosis of ESCC.At the same time,all independent risk factors are integrated to establish nomograms for predicting long-term prognosis of patients with esophageal cancer who received R0 resection,and the models were verified.Methods: We retrospectively analyzed the treatment and survival of 351 locally advanced ESCC patients who underwent radical surgery and PORT in our hospital from June 2006 to June 2016.Receiver operating characteristic curves were used to estimate the optimal cutoff point of the time interval between surgery and PORT,and then patients were divided into the early PORT group and the delayed PORT group.Differences in Overall survival(OS)and progression-free survival(PFS)between groups were assessed using log-rank tests and demonstrated using Kaplan-Meier curves.The clinicopathological and treatment-related factors collected were subjected to univariate Cox analysis.Those factors with p < 0.10 in the univariate Cox analysis were then incorporated into the multivariate Cox analysis to identify independent predictors of survival.All independent predictors were included in the nomogram to predict the OS and PFS of patients with locally advanced ESCC who received radical resection(R0 resection)and PORT.Further,the validity and accuracy of the models were verified by C-index,AUC model and calibration curve.Results: The median follow-up was 53 months(range: 3–179 months).The interval between surgery and PORT significantly influenced both of the PFS and OS.The 5-year PFS rates in the early and delayed PORT groups were 41.0% and 58.0%,respectively.And the 5-year OS rates in the early and delayed PORT groups were44.6% and 58.6%,respectively.For patients with locally advanced ESCC who received R0 resection and PORT,multivariate Cox analysis demonstrated that the independent risk factors associated with PFS include sex,number of positive lymph nodes,number of lymph nodes removed,LNR,tumor length,surgical approach,adjuvant chemotherapy,radiation dose and time interval.Besides,pathological N classification was also independent risk factor related to OS.Compared to early PORT,PORT at >48 days after surgery was associated with better OS(Hazard ratio [HR]:1.406,p = 0.037)and PFS(HR: 1.475,p = 0.018).In the chemotherapy subgroup,incorporation of chemotherapy timing into the analysis suggested that 2–4chemotherapy cycles followed by PORT was the optimal treatment schedule as compared to 0–1 chemotherapy cycle followed by PORT and concurrent chemoradiotherapy(5-year PFS: 65.9% vs.51.0% vs.50.1%;p = 0.049).The nomograms for OS and PFS were superior to the TNM classification(concordance indices: 0.721 vs.0.626 and 0.716 vs.0.610,respectively),and were well-calibrated.Conclusion: The study has shown that sex,number of positive lymph nodes,number of lymph nodes removed,LNR,tumor length,surgical approach,adjuvant chemotherapy,radiation dose,the time interval between surgery and PORT are the independent risk factors related to PFS of patients with locally advanced ESCC who received R0 resection and PORT.Besides the factors listing above,pathological N classification was also independent risk factor related to OS.Delayed PORT(>48days)provided better survival benefit than early PORT,PORT following chemotherapy might lead to the best survival rate.In addition,we established new nomogram models based on all independent risk factors for clinical prediction of PFS and OS,which will help to guide the individualized treatment of patients. |
PDF Full Text Request