Ben Huang1, Chen Chen1, Mengdong Ni1, Shaobo Mo1, Guoxiang Cai2, Sanjun Cai3. 1. Department of Colorectal Surgery, Fudan University Shanghai Cancer Center, Shanghai 20032, People's Republic of China. 2. Department of Colorectal Surgery, Fudan University Shanghai Cancer Center, Shanghai 20032, People's Republic of China. Electronic address: gxcai@fudan.edu.cn. 3. Department of Colorectal Surgery, Fudan University Shanghai Cancer Center, Shanghai 20032, People's Republic of China. Electronic address: csjfuscc@163.com.
Abstract
BACKGROUND: Lymph node ratio (LNR) is considered a better staging system than N stage in rectal cancer. More recently, log odds of positive lymph nodes (LODDS) was identified as a novel prognostic classifier in many malignancies. Accordingly, our study aims to compare the predictive ability of LODDS with LNR for cancer-specific survival (CSS) in patients with stage III rectal cancer. METHODS: We analyzed a subpopulation of the Surveillance, Epidemiology and End Results (SEER) database containing patients with stage III rectal cancer. The patients were categorized into four groups (LNR1 to 4) according to the LNR cut-off values 0.25, 0.50 and 0.75. The patients were divided into five groups (LODDS1 to LODDS5) according to the LODDS cut-off values of LODDS -1, 0, 1 and 2. Univariate and multivariate analyses using the Cox proportional hazards model were performed to analyze the risk factors for survival outcome. RESULTS: A total of 17,632 patients were included from the SEER database. Patients with LNR4 could be further divided into LODDS4 and LODDS5 and had a 5-year CSS of 39.1% and 23.3%, respectively (p < 0.001). A multivariate analysis without the inclusion of LODDS showed that the LNR was an independent prognostic factor (HR 2.254, 95% confidence interval (CI) 2.034-2.497, p < 0.001). However, after adjusting for LODDS, the LNR was no longer associated with CSS (HR 0.709, 95% CI 0.481-1.045, p = 0.083), and LODDS was identified as an independent prognostic factor (HR 1.303, 95% CI 1.197-1.419, p < 0.001). CONCLUSION: The prognostic value of LNR can be confounded by LODDS. In stage III rectal cancer patients, LODDS has superior discrimination power over LNR and can more accurately evaluate CSS.
BACKGROUND: Lymph node ratio (LNR) is considered a better staging system than N stage in rectal cancer. More recently, log odds of positive lymph nodes (LODDS) was identified as a novel prognostic classifier in many malignancies. Accordingly, our study aims to compare the predictive ability of LODDS with LNR for cancer-specific survival (CSS) in patients with stage III rectal cancer. METHODS: We analyzed a subpopulation of the Surveillance, Epidemiology and End Results (SEER) database containing patients with stage III rectal cancer. The patients were categorized into four groups (LNR1 to 4) according to the LNR cut-off values 0.25, 0.50 and 0.75. The patients were divided into five groups (LODDS1 to LODDS5) according to the LODDS cut-off values of LODDS -1, 0, 1 and 2. Univariate and multivariate analyses using the Cox proportional hazards model were performed to analyze the risk factors for survival outcome. RESULTS: A total of 17,632 patients were included from the SEER database. Patients with LNR4 could be further divided into LODDS4 and LODDS5 and had a 5-year CSS of 39.1% and 23.3%, respectively (p < 0.001). A multivariate analysis without the inclusion of LODDS showed that the LNR was an independent prognostic factor (HR 2.254, 95% confidence interval (CI) 2.034-2.497, p < 0.001). However, after adjusting for LODDS, the LNR was no longer associated with CSS (HR 0.709, 95% CI 0.481-1.045, p = 0.083), and LODDS was identified as an independent prognostic factor (HR 1.303, 95% CI 1.197-1.419, p < 0.001). CONCLUSION: The prognostic value of LNR can be confounded by LODDS. In stage III rectal cancerpatients, LODDS has superior discrimination power over LNR and can more accurately evaluate CSS.