Liu Kai1, Chen Zhoumiao1, Xu Shaohua1, Chen Zhao1, Li Zhijun1, He Zhengfu1, Cai Xiujun2. 1. Department of Thoracic Surgery, Sir Run Run Shaw Hospital, School of Medicine, Zhejiang University, Hangzhou, 310016, Zhejiang Province, China. 2. Department of surgery, Sir Run Run Shaw Hospital, School of Medicine, Zhejiang University, 3 east qing chun road, Hangzhou, 310016, Zhejiang Province, China. srrsh_cxj@zju.edu.cn.
Abstract
BACKGROUND: Lymph node ratio (LNR) has been suggested to be an effective prognostic tool for stratifying non-small cell lung cancer (NSCLC) cases. In this study, we sought to determine cancer-specific survival (CCS) of NSCLC cases from the SEER registry and used the X-tile method to optimize CCS-based LNR cut-off points for prognostic stratification of node-positive NSCLC. METHODS: CSS and other clinicopathologic variables were retrieved from the SEER registry. Kaplan-Meier methods were used to calculate CSS. The optimal cut-off points for LNR classification were determined by the X-tile approach. Multivariate Cox regression analysis was performed to identify independent risks of CSS. RESULTS: Totally 11,341 lung cancer patients were included. Their median CSS was 22 months (range 0,143). The median LNR was 0.22 (Q1,Q3: 0.11, 0.50). X-tile analysis showed that the optimal LNR cut-off points were 0.28 and 0.81, dividing the cohort into low (LNR1 ≤ 0.28; n = 6580, 58%), middle (0.28 < LNR2 < 0.81; n = 3025, 26.7%), and high (LNR3 > 0.81; n = 1736, 15.3%) subsets. Kaplan-Meier analysis showed that patients with a low LNR had a significantly higher CCS versus patients with middle or high LNR (P < 0.001). Multivariate competing risks regression analysis revealed that LNR was an independent and significant adverse predictor of CSS (LNR2 vs. LNR1: SHR: 1.56, 95%CI: 1.47,1.67, P < 0.001; LNR3 vs. LNR1: SHR: 2.54, 95%CI: 2.30,2.80, P < 0.001). CONCLUSIONS: LNR is an independent prognostic factor of node-positive NSCLC and its optimal cut-off values established using the robust x-tile method effectively define subpopulations of node-positive NSCLC cases, which is important in guiding selection of treatment strategies clinically.
BACKGROUND: Lymph node ratio (LNR) has been suggested to be an effective prognostic tool for stratifying non-small cell lung cancer (NSCLC) cases. In this study, we sought to determine cancer-specific survival (CCS) of NSCLC cases from the SEER registry and used the X-tile method to optimize CCS-based LNR cut-off points for prognostic stratification of node-positive NSCLC. METHODS:CSS and other clinicopathologic variables were retrieved from the SEER registry. Kaplan-Meier methods were used to calculate CSS. The optimal cut-off points for LNR classification were determined by the X-tile approach. Multivariate Cox regression analysis was performed to identify independent risks of CSS. RESULTS: Totally 11,341 lung cancerpatients were included. Their median CSS was 22 months (range 0,143). The median LNR was 0.22 (Q1,Q3: 0.11, 0.50). X-tile analysis showed that the optimal LNR cut-off points were 0.28 and 0.81, dividing the cohort into low (LNR1 ≤ 0.28; n = 6580, 58%), middle (0.28 < LNR2 < 0.81; n = 3025, 26.7%), and high (LNR3 > 0.81; n = 1736, 15.3%) subsets. Kaplan-Meier analysis showed that patients with a low LNR had a significantly higher CCS versus patients with middle or high LNR (P < 0.001). Multivariate competing risks regression analysis revealed that LNR was an independent and significant adverse predictor of CSS (LNR2 vs. LNR1: SHR: 1.56, 95%CI: 1.47,1.67, P < 0.001; LNR3 vs. LNR1: SHR: 2.54, 95%CI: 2.30,2.80, P < 0.001). CONCLUSIONS: LNR is an independent prognostic factor of node-positive NSCLC and its optimal cut-off values established using the robust x-tile method effectively define subpopulations of node-positive NSCLC cases, which is important in guiding selection of treatment strategies clinically.
Authors: Matthew D Taylor; Damien J LaPar; Christopher J Thomas; Matthew Persinger; Edward B Stelow; Benjamin D Kozower; Christine L Lau; David R Jones Journal: Ann Thorac Surg Date: 2013-08-30 Impact factor: 4.330