Lymph nodes ratio based nomogram predicts survival of resectable gastric cancer regardless of the number of examined lymph nodes

Abstract

To develop a nomogram to predict the prognosis of gastric cancer patients on the basis of metastatic lymph nodes ratio (mLNR), especially in the patients with total number of examined lymph nodes (TLN) less than 15. The nomogram was constructed based on a retrospective database that included 2,205 patients underwent curative resection in Cancer Center, Sun Yat-sen University (SYSUCC). Resectable gastric cancer (RGC) patients underwent curative resection before December 31, 2008 were assigned as the training set (n=1,470) and those between January 1, 2009 and December 31, 2012 were selected as the internal validation set (n=735). Additional external validations were also performed separately by an independent data set (n=602) from Jiangxi Provincial Cancer Hospital (JXCH) in Jiangxi, China and a data set (n=3,317) from the Surveillance, Epidemiology, and End Results (SEER) database. The Independent risk factors were identified by Multivariate Cox Regression. In the SYSUCC set, TNM (Tumor-node-metastasis) and TRM-based (Tumor-Positive Nodes Ratio-Metastasis) nomograms were constructed respectively. The TNM-based nomogram showed better discrimination than the AJCC-TNM staging system (C-index: 0.73 versus 0.69, p<0.01). When the mLNR was included in the nomogram, the C-index increased to 0.76. Furthermore, the C-index in the TRM-based nomogram was similar between TLN ≥16 (C-index: 0.77) and TLN ≤15 (C-index: 0.75). The discrimination was further ascertained by internal and external validations. We developed and validated a novel TRM-based nomogram that provided more accurate prediction of survival for gastric cancer patients who underwent curative resection, regardless of the number of examined lymph nodes.

Keywords: curative resection; gastric cancer; lymph nodes ratio; nomogram; prognosis.

Figure 1. Impact of mLNS and mLNR staging on gastric cancer-related survival respectively

(A) Overall survival according to AJCC N stage (mLNS), stratified by the number of examined nodes (≤15 and ≥16). (B) Overall survival of node-negative patients with 15 or less retrieved, mLNR (0 to 1/15) and node-negative patients with 16 or more retrieved. (C) Overall survival according to mLNS stage. (D) Overall survival according to mLNR stage. Abbreviation: mLNS, metastatic lymph node stage; mLNR, metastatic lymph node ratio.

Figure 2. Nomogram predicting 1-year, 3-year and 5-year OS for resectable gastric cancer patients after curative resection

Abbreviation: mLNR: metastatic lymph node ratio.

Figure 3. The calibration curves for predicting patients overall survival at 5-year in the training set

(A), validation set with TLN ≤ 15 (B), validation set with TLN ≥ 16 (C), validation set (D) and predicting overall survival at 3-year (E), 5-year (F) in primary cohort. The X-aixs represents the nomogram-predicted survival, and the actual survival is plotted on the Y-axis. The dotted line represents the ideal correlationship between predicted and actual survival.

Figure 4

(A) Overall survival of primary cohort according to the 7th edition AJCC staging system; (B) Distribution of nomogram-predicted 5-year overall survival based on the 7th edition of AJCC staging system.