Trend in Conditional Overall Survival of Esophageal Cancer: A Population-Based SEER Database Analysis

Abstract

To investigate the causes of death in patients with esophageal cancer and report reliable and contemporary estimates of adjusted conditional overall survival (COS). Patients aged 18 years or older diagnosed with esophageal cancer between 1975 and 2016 were identified from the Surveillance Epidemiology End Results (SEER) registry. COS were estimated using an inverse probability weighting method, with an adjustment for age at diagnosis, sex, ethnicity, marital status, tumor grade and stage, tumor location, year of diagnosis, and treatment modalities, which included chemotherapy, radiation therapy, and surgery. Linear trends in adjusted COS over time were tested by fitting a weighted linear regression to the year of diagnosis and time alive after diagnosis, in which the weights were inversely proportional to the variance of the COS. A total of 40 142 confirmed patients were included in the final analysis, in which 19 171 patients were diagnosed with esophageal squamous cell carcinoma, and 20 971 patients were diagnosed with esophageal adenocarcinoma. Esophageal cancer is the leading death in these patients, followed by heart disease (5.2%, n = 1 918), stomach cancer (3.2%, n = 1 191), and other diseases of death (2.9%, n = 881). Significant improvements in adjusted COS were observed as the time alive after diagnosis increases from 0 to 48 months. Notably, 5-year adjusted COS given that patients have already survived 36 months was improved from 0.50 (95% confidence interval [CI]: 0.49 – 0.52) to 0.87 (95% CI: 0.86 – 0.87, p< 0.001) in comparison with the conventional 5-year overall survival. The difference in esophageal cancer-specific adjusted COS after diagnosis were identified on the basis of age at diagnosis, sex, ethnicity, marital status, tumor stage and grade, tumor location, year of diagnosis, and treatment modalities. Moreover, adjusted COS improved from 1975-1979 to 1995-2000, even though none linear trends were observed over time. For patients with esophageal cancer who have survived over time, 5-year adjusted COS improved dramatically, of which other causes of death should be considered. These results are critical to inform patients for making treatment decisions, and to inform physicians for constructing follow-up and surveillance strategies, and to inform health practitioners for providing better service.