Prognostic evaluation of esophageal cancer patients with stages I-III

Abstract

Purpose: The purpose of this study was to investigate the impact of clinicopathological factors and treatments on the overall survival (OS) and esophageal cancer-specific survival (ECSS) of stages I-III esophageal cancer (EC) patients and to establish a prognostic visual nomogram. Methods: We collected clinical data of patients diagnosed with stages I-III EC without receiving chemotherapy from 2004 to 2014 from the Surveillance, Epidemiology, and End Results (SEER) database. Prognoses were analyzed using the R language software, and nomograms were obtained according to the visual processing logistic regression model, which was verified using the Harrell C-index, receiver operating characteristic (ROC) curve, and calibration curve. Results: A total of 4,305 patients were selected, mostly white males. Most patients were over 60 years old and old age predicted poor prognosis. EC, primarily adenocarcinoma, occurred mostly in the lower third of the esophagus. About half of the patients had T1 (58.00%) and grade II (50.41%) cancer. Of all the patients, 2,448 was treated with surgery and the majority (n = 1,476; 64.85%) of these patients had stage I EC. Stages I-III patients underwent surgery had significantly better OS and ECSS, and endoscopic therapy was associated with the best outcome amongst all the surgical methods. 3.67% of the patients received radiotherapy, predominantly postoperative radiotherapy (2.69%). Older age, squamous cell carcinoma, overlapping lesion of the esophagus, and grades II and III were high-risk factors for poor OS and ECSS for stage I patients, whereas endoscopic therapy, esophagectomy, and esophagectomy with gastrectomy were low-risk factors. Stage II patients with older age, male sex, T3, N1, and grades II and III had shorter OS and ECSS, but patients with any surgical treatment had significantly longer OS and ECSS. T4, N1, and grade III correlated negatively with OS and ECSS in stage III patients, and any surgical treatment correlated positively with longer OS and ECSS. The OS and ECSS rates of stages I-III EC patients with a total score of more than 150 points in the nomogram were both only 40% after 3 years and 30% after 5 years. The C-index, ROC curve, and calibration curve indicated that the nomograms established in this study were suitable to assess patient prognosis. Conclusion: The nomogram established in this study is an effective clinical tool to predict the prognosis of stages I-III EC patients without chemotherapy.