Tu1509 PRE-PROCEDURAL FEATURES FOR PREDICTING THE RISK OF LYMPH NODE METASTASIS IN GASTRIC CANCER

Abstract

Endoscopic submucosal dissection (ESD) is a curative method for some cases of early gastric cancer (EGC). However, there are still some patients who underwent additive gastrectomy after ESD because of lymph node metastasis (LNM) risk. If an accurate predicting for the risk of LNM before ESD is possible, appropriate treatment strategy could be established at the timing of decision for ESD. We aimed to investigate pre-procedural features for predicting the risk of LNM in EGC. We collected the data of 889 patients who underwent ESD for EGC. Patients with non-curative resection due to LNM, positive vertical margin, lymphovascular invasion, or submucosal invasion in undifferentiated-type and submucosal invasion over 500μm in differentiated-type were assigned as non-curative resection group (n=143). Among curative resection patients, undifferentiated-type limited to mucosa and differentiated-type with submucosal invasion limited to 500μm, were assigned as a control group (n=746). We analyzed endoscopic, radiologic, pathologic, single patient classifier (SPC) genes features about the risk of LNM. Radiologic feature is fold thickening or perigastric lymph node enlargement on CT. SPC genes (GZMB WARS, SFRP4) were known to be associated with prognosis of gastric cancer. A larger size (>20mm, P=0.002), elevated morphology (I&IIa, P=0.006), the presence of tumor erythema (P<0.001), white mucosal change (P=0.005), fold convergence (P<0.001) and abnormal finding at CT (P=0.002) were associated with the risk of LNM. Moderate differentiated adenocarcinoma (OR=2.07, P=0.003), poorly differentiated adenocarcinoma (OR=4.08, P<0.001) were increased the risk of LNM than well differentiated adenocarcinoma. 11 patients (1.2%) identified LNM after additive surgery. Almost patients (9/11) with LNM showed intermediate to high risk based on SPC genes. Several pre-procedural features were associated with the risk of LNM. Therefore, considering these features in addition to existing ESD criteria may be helpful in providing patients with the best treatment option.