Jinlin Cao1, Ping Yuan1, Honghai Ma1, Peng Ye1, Yiqing Wang1, Xiaoshuai Yuan1, Feichao Bao1, Wang Lv1, Jian Hu2. 1. Department of Thoracic Surgery, The First Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, China. 2. Department of Thoracic Surgery, The First Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, China. Electronic address: hujian_med@163.com.
Abstract
BACKGROUND: The number of lymph node metastases (LNMs) and the positive lymph node ratio (LNR) have been shown to be significant independent prognostic factors in predicting overall survival in patients with esophageal cancer. Our aim was to evaluate whether a novel prognostic indicator-the log odds of positive lymph nodes (LODDS)-predicts survival in esophageal cancer. METHODS: Patients who underwent esophagectomy for esophageal cancer between 1988 and 2007 were identified from the Surveillance, Epidemiology, and End Results (SEER) database of 18 registries, and a Chinese patient cohort was subjected to validation. The prognostic efficacy of LNM, LNR, and LODDS was compared. Prognostic performance was measured using Harrell's C-index statistic, Schemper's proportion of explained variation, and the Akaike information criterion (AIC). RESULTS: A total of 4,123 patients in the SEER database and 134 patients in the Chinese cohort met our criteria in this study. LODDS gave a better prognostic performance than either LNM or LNR in both the SEER database and the Chinese cohort. Multivariate analyses showed significant values for LNM, LNR, and LODDS as prognostic factors (p < 0.001). However, only LODDS was a significant prognostic factor regardless of the number of lymph nodes retrieved (p = 0.677). Furthermore, after stratification of patients with no nodes involved or all nodes involved, the values of LODDS still distinguished the heterogeneity efficiently. CONCLUSIONS: LODDS predicts survival more accurately than either LNM or LNR in patients undergoing resection for esophageal cancer, especially for patients with an insufficient number of lymph nodes retrieved.
BACKGROUND: The number of lymph node metastases (LNMs) and the positive lymph node ratio (LNR) have been shown to be significant independent prognostic factors in predicting overall survival in patients with esophageal cancer. Our aim was to evaluate whether a novel prognostic indicator-the log odds of positive lymph nodes (LODDS)-predicts survival in esophageal cancer. METHODS:Patients who underwent esophagectomy for esophageal cancer between 1988 and 2007 were identified from the Surveillance, Epidemiology, and End Results (SEER) database of 18 registries, and a Chinese patient cohort was subjected to validation. The prognostic efficacy of LNM, LNR, and LODDS was compared. Prognostic performance was measured using Harrell's C-index statistic, Schemper's proportion of explained variation, and the Akaike information criterion (AIC). RESULTS: A total of 4,123 patients in the SEER database and 134 patients in the Chinese cohort met our criteria in this study. LODDS gave a better prognostic performance than either LNM or LNR in both the SEER database and the Chinese cohort. Multivariate analyses showed significant values for LNM, LNR, and LODDS as prognostic factors (p < 0.001). However, only LODDS was a significant prognostic factor regardless of the number of lymph nodes retrieved (p = 0.677). Furthermore, after stratification of patients with no nodes involved or all nodes involved, the values of LODDS still distinguished the heterogeneity efficiently. CONCLUSIONS:LODDS predicts survival more accurately than either LNM or LNR in patients undergoing resection for esophageal cancer, especially for patients with an insufficient number of lymph nodes retrieved.