Comparison of accuracy of different lymph node ratio grouping systems in predicting prognosis of Stage IV colon cancer

Abstract Objective With the improvement of surgical treatment, many studies have questioned the accuracy of the AJCC staging system in that it ignores examined lymph nodes (ELNs). Lymph node ratio (LNR) is supposed to make up for this deficiency. The aim of this study is to compare the prognostic performance of different LNR grouping systems relative to the American Joint Committee on Cancer (AJCC) 8th N staging system in stage IV colon cancer. Methods Approximately 8672 patients undergoing primary surgical resection of stage IV colon cancer were extracted from the Surveillance, Epidemiology, and End Results (SEER) program database. The Harrell’s concordance index (C-index) and Akaike's Information Criterion (AIC) were used to distinguish the prognostic performance of different LNR and AJCC 8th N staging systems. Results Among 8672 colon cancer patients with distant metastasis, 7322 (84.43%) had lymph node involvement, and the median of ELNs was 17. The Group2 staging system of LNR (cutoff points: 1/14; 0.25; 0.50) was superior to the 8th AJCC-N staging system and other LNR grouping systems, with the most considerable discrimination power (C-index, 0.711; AIC, 121419.1), and showed an advantage in any of the ELNs. When models were used as categorical cutoff variables for further clinical application, The second LNR grouping system outperformed the other three schemes with either ELNs less than 12 (C-index, 0.723; AIC, 16735.24), between 12 and 25 (C-index, 0.707; AIC, 76007.79), or more than 25(C-index, 0.701; AIC, 15613.2) with increasing C-index and less AIC value. Conclusion When evaluating different LNR grouping systems and the 8th AJCC-N staging system, regardless of ELNs, the second LNR grouping system (the cut-off value of LNR was 1/14; 0.25; 0.50) showed the best discrimination and goodness of fit in predicting survival in stage IV colon cancer patients.