Background: Few risk models have been provided to predict long-term prognosis after esophagectomy. This study investigated the reliability of a risk calculator as well as classification and regression trees analysis for predicting long-term prognosis after esophagectomy for esophageal cancer. Methods: The study enrolled 438 patients who underwent esophagectomy at Keio University Hospital, Tokyo, Japan, between July 2000 and June 2016. Patients who underwent R0 or R1 resection or esophagectomy with combined resection of other organs were included. The authors investigated the usefulness of a risk model for 30-day mortality and operative mortality described in their previous report for predicting long-term prognosis after esophagectomy. Results: The 438 patients (377 men and 61 women) in this study had a 5-year overall survival (OS) rate of 62.8% and a disease-free survival rate of 54.3%. The OS was higher for the patients with 30-day mortality risk model values lower than 0.675% than for those with values higher than 0.675% (p < 0.001). The cutoff values for prediction were shown to be significant risk factors in the multivariate analysis. The risk calculator was validated by comparing the cutoff values with Harrell’s C-index values of clinical stage. For overall risk, the C-index of operative mortality was 0.697, and the C-index of cStage was 0.671. Conclusions: The risk calculator was useful for predicting recurrence and death after esophagectomy. Furthermore, because the C-index of the risk model for operative mortality was higher than for clinical tumor-node-metastasis stage, this risk-scoring system may be more useful clinically.
ASJC Scopus subject areas