Predictable factors for lymph node metastasis in early gastric cancer analysis of clinicopathologic factors and biological markers

Predicting lymph node metastasis (LNM) accurately is very important to decide treatment strategies preoperatively. The aim of this study was to explore risk factors that predict the presence of LNM in early gastric cancer (EGC). A total of 230 patients with EGC who underwent curative gastrectomy with lymph adenectomy at Xinhua Hospital from January 2006 to July 2014 were retrospectively reviewed. We studied the relationship between clinicopathological factors, biological markers (p53, ki67, nm23, vascular endothelial growth factor (VEGF), epidermal growth factor receptor (EGFR), E-cadherin (E-cad), beta-catenin (b-catenin), glutathione S-transferase (GST), and topoisomerase II (Topo II)), and LNM of EGC patients by chi-square test and logistic regression analysis. Meta-analyses were further conducted to review the effects of the proteins (P53, ki67, E-cad, and b-catenin) on LNM in ECG patients. LNM was detected in 42 (18.3 %) of 230 patients. Incidences of LNM was distinct in different tumor size (p = 0.044), depth of submucosal invasion (p < 0.0001), and P53 overexpression (p = 0.004). Multivariate analysis further indentified that large tumor size (aeyen20 mm, odds ratio (OR) = 2.168, p = 0.041), submucosa (OR = 4.000, p = 0.0005), and P53 overexpression (OR = 3.010, p = 0.022) were independent risk factors of LNM in EGC patients. The meta-analysis revealed a significantly statistical association of P53, ki67, and b-catenin with an increased risk of LNM in EGC patients (P53, OR = 1.81, p = 0.017; ki67, OR = 2.53, p = 0.0003; b-catenin, OR = 0.53, p = 0.01). Tumor size (aeyen20 mm), the depth of invasion (submucosa), and P53 overexpression may be helpful predictors of LNM in EGC patients. Furthermore, the results of meta-analysis revealed that P53, ki67 overexpression, and abnormal expression of b-catenin may be associated with LNM in EGC. The results need further validation in single large studies.