Prognostic factors of radical gastrectomy for stage III gastric cancer and predictive value of metastatic lymph node ratio for prognosis: a report of 995 cases

Abstract

Objective To investigate the prognostic factors of radical gastrectomy for stage Ⅲ gastric cancer and predictive value of metastatic lymph node ratio for prognosis. Methods The retrospective case-control study was conducted. The clinicopathological data of 995 patients with stage Ⅲ gastric cancer who were admitted to the Zhongshan Hospital of Fudan University between January 2003 and December 2014 were collected. There were 690 males and 305 females, aged from 20 to 75 years, with an average age of 61 years. After clinical staging according to results of preoperative accessory examinations, patients with early gastric cancer underwent D1+ lymphadenectomy, patients with advanced gastric cancer underwent D2 lymphadenectomy and patients with serosa invasion underwent D2+ lymphadenectomy. Observation indicators: (1) treatment situations; (2) follow-up and survival situations; (3) prognostic factors analysis after operation; (4) stratified analysis: ① stratified analysis of tumor pathological N staging; ② stratified analysis of number of lymph node dissected; ③ stratified analysis of tumor pathological TNM staging; (5) receiver operating characteristic (ROC) curve. Patients were followed up using outpatient examination and telephone interview to detect postoperative survival up to January 2016. The overall survival time was from the operation data to last follow-up or time of death. Measurement data with normal distribution were represented as Mean±SD. The survival rate and curve were respectively calculated and drawn by the Kaplan-Meier method and Log-rank test was used for survival analysis. The COX proportional hazard model was used for univariate and multivariate analysis. The ROC curve and area under curve (AUC) were used to check the accuracy of number of positive lymph nodes and metastatic lymph node ratio for prognosis. Comparison of the AUC was analyzed by the Z test. Results (1) Treatment situations: of 995 patients underging gastrectomy, 677 underwent distal gastrectomy, 85 underwent proximal gastrectomy, 233 underwent total gastrectomy. There were 117 undergoing D1+ lymphadenectomy and 878 undergoing D2 lymphadenectomy or D2+ lymphadenectomy. The number of lymph node dissected, number of positive lymph nodes, metastatic lymph node ratio were 27±12, 10±9 and 0.41±0.28, respectively. (2) Follow-up and survival situations: 995 patients were followed up for (35±32)months. During the follow-up, the 1-, 3-, 5-year overall survival rates were 77.9%, 47.8%, 36.2%. (3) Prognostic factors analysis after operation: results of univariate analysis showed that sex, tumor histological type, vascular embolus, degree of tumor invasion, tumor pathological N staging, number of lymph node dissected, metastatic lymph node ratio, tumor pathological TNM staging were related factors affecting prognosis of radical gastrectomy for stage Ⅲ gasteric cancer (hazard ratio=0.817, 1.486, 1.268, 2.173, 1.957, 1.737, 3.357, 2.169, 95% confidence interval: 0.686-0.973, 1.059-2.086, 1.074-1.497, 1.195-3.954, 1.480-2.588, 1.390-2.170, 2.476-4.602, 1.740-2.704, P 0.4, showing a statistically significant difference (χ2=121.300, P 0.4 (χ2=7.580, 65.320, P 0.4 (χ2=80.806, P 0.1. The postoperative 1-, 3-, 5-year overall survival rates were 92.2%, 68.6%, 59.1% in the 108 patients with metastatic lymph node ratio ≤0.1 and 79.2%, 32.8%, 21.9% in the 24 patients with metastatic lymph node ratio >0.1, respectively, showing a statistically significant difference (χ2=14.499, P 0.2. The postoperative 1-, 3-, 5-year overall survival rates were 92.3%, 73.8%, 61.0% in the 138 patients with metastatic lymph node ratio ≤0.2 and 76.5%, 40.1%, 22.2% in the 127 patients with metastatic lymph node ratio >0.2, respectively, showing a statistically significant difference (χ2=42.536, P 0.4. The postoperative 1-, 3-, 5-year overall survival rates were 88.5%, 62.8%, 47.0% in the 194 patients with metastatic lymph node ratio ≤0.4 and 64.3%, 29.8%, 21.0% in the 404 patients with metastatic lymph node ratio >0.4, respectively, showing a statistically significant difference (χ2=51.860, P 0.4. The postoperative 1-, 3-, 5-year overall survival rates were 78.2%, 40.0%, 28.6% in the 44 patients with metastatic lymph node ratio ≤0.4 and 78.1%, 18.7%, 12.9% in the 73 patients with metastatic lymph node ratio>0.4, respectively, showing a statistically significant difference (χ2=4.727, P 0.4, respectively, showing a statistically significant difference (χ2=6.132, P 0.4, respectively, showing a statistically significant difference (χ2=23.699, P 0.4, respectively, showing a statistically significant difference (χ2=36.215, P<0.05) . (5) ROC curve: ROC curves of postoperative overall survival rates in patients with stage Ⅲ gastric cancer were drawn using the number of positive lymph nodes and metastatic lymph node ratio, of which AUC were 0.619 (95% confidence interval: 0.588-0.649) and 0.706 (95% confidence interval : 0.677-0.734), showing a statistically significant difference (Z=8.842, P<0.05). Conclusions Sex, tumor histological type, tumor pathological N staging, number of lymph node dissected, metastatic lymph node ratio, tumor pathological TNM staging are independent factors affecting prognosis of radical gastrectomy for stage Ⅲ gastric cancer. There is different prognosis of patients with different metastatic lymph node ratios in the subgroup of the same tumor pathological TNM staging, number of lymph node dissected, tumor pathological TNM staging. Compared with tumor pathological N staging, metastatic lymph node ratio has a more accurate predictive value for prognosis. Key words: Gastric neoplasms; Gastrectomy; Radical resection; Lymphadenectomy; Metastatic lymph node ratio; N staging; Prognosis