Xiaojie Wang1, Shenghui Huang1, Xingrong Lu1, Ying Huang2, Pan Chi1. 1. Department of Colorectal Surgery, Union Hospital, Fujian Medical University, 29 Xin-Quan Road, Fuzhou, 350001, Fujian, People's Republic of China. 2. Department of Colorectal Surgery, Union Hospital, Fujian Medical University, 29 Xin-Quan Road, Fuzhou, 350001, Fujian, People's Republic of China. hy3169@sina.com.
Abstract
BACKGROUND: To define the incidence of gastroepiploic lymph node (GLN) metastasis in patients with cancer of the transverse colon, including the hepatic flexure, and to identify the preoperative predictors of GLN involvement in a large-volume center in China. METHODS: This retrospective monocentric cross-sectional study respected the STROBE statement. Of 3208 consecutive patients who underwent colon cancer resection, a total of 371 patients with cancer of the transverse colon including the hepatic flexure who underwent complete mesocolic excision and GLN resection in our center were retrospectively reviewed between November 2010 and November 2017. Logistic regression was performed to identify predictors of GLN metastasis. Endoscopic obstruction was defined as a luminal obstruction of the colon severe enough to prevent the colonoscope from passing beyond the tumor regardless of the presenting symptoms. RESULTS: The GLN involvement rate was 4.0 (2.0-6.1)%. Patients who had GLN involvement had a significantly higher rate of endoscopic obstruction (P = 0.030), higher rate of signet ring adenocarcinoma or lymphovascular invasion (P < 0.05), higher preoperative CEA level (P = 0.037), more advanced pN stage (P < 0.001) and more advanced M stage (P = 0.003) than the patients without GLN involvement. ROC curve analyses showed that the cutoff value for CEA was 17.0 ng/ml (46.7% sensitivity, 84.3% specificity, P = 0.037) for the prediction of GLN metastasis. Multivariate analysis showed that endoscopic obstruction, signet ring adenocarcinoma, a CEA level ≥17 ng/ml and M1 stage were independently correlated with the GLN metastasis. CONCLUSION: The incidence rate of GLN metastasis was low. To the best of our knowledge, the present study was the first to evaluate the preoperative predictors of GLN metastasis. Combinations of predictive factors may be useful for stratifying patients at high risk of GLN metastasis.
BACKGROUND: To define the incidence of gastroepiploic lymph node (GLN) metastasis in patients with cancer of the transverse colon, including the hepatic flexure, and to identify the preoperative predictors of GLN involvement in a large-volume center in China. METHODS: This retrospective monocentric cross-sectional study respected the STROBE statement. Of 3208 consecutive patients who underwent colon cancer resection, a total of 371 patients with cancer of the transverse colon including the hepatic flexure who underwent complete mesocolic excision and GLN resection in our center were retrospectively reviewed between November 2010 and November 2017. Logistic regression was performed to identify predictors of GLN metastasis. Endoscopic obstruction was defined as a luminal obstruction of the colon severe enough to prevent the colonoscope from passing beyond the tumor regardless of the presenting symptoms. RESULTS: The GLN involvement rate was 4.0 (2.0-6.1)%. Patients who had GLN involvement had a significantly higher rate of endoscopic obstruction (P = 0.030), higher rate of signet ring adenocarcinoma or lymphovascular invasion (P < 0.05), higher preoperative CEA level (P = 0.037), more advanced pN stage (P < 0.001) and more advanced M stage (P = 0.003) than the patients without GLN involvement. ROC curve analyses showed that the cutoff value for CEA was 17.0 ng/ml (46.7% sensitivity, 84.3% specificity, P = 0.037) for the prediction of GLN metastasis. Multivariate analysis showed that endoscopic obstruction, signet ring adenocarcinoma, a CEA level ≥17 ng/ml and M1 stage were independently correlated with the GLN metastasis. CONCLUSION: The incidence rate of GLN metastasis was low. To the best of our knowledge, the present study was the first to evaluate the preoperative predictors of GLN metastasis. Combinations of predictive factors may be useful for stratifying patients at high risk of GLN metastasis.