Predictive Biomarkers of Lymph Node Metastasis in Early Gastric Cancer: A Reference of Clinicopathological Characteristics, Protein Expression, Epstein-Barr Virus Status, and Microsatellite Instability

Abstract

Background/Aim: Predicting lymph node metastasis (LNM) in early gastric cancer (EGC) is crucial for making treatment decisions. This study aimed to confirm risk factors for LNM and identify novel auxiliary biomarkers for predicting LNM in EGC. Patients and Methods: We established a training set, comprising 63 patients with LNM-EGC and 274 patients with non-LNM EGC, and a test set, comprising 19 patients with LNM-EGC and 146 non-LNM EGC. Immunohistochemistry for lymphangiogenic and related pathway components (VEGF-C, TGF-β1, SMAD2/3, VEGF-D, pSTAT3, E-cadherin, CD44, c-MET, YAP, and HER2), in situ hybridization for Epstein-Barr virus-encoded small RNAs, and multiplex PCR for microsatellite instability were conducted. Results: In the training set, Lauren’s diffuse/mixed classification, stromal desmoplasia, submucosal invasion ≥500 μm, lymphatic invasion, and high VEGF-C and SMAD2/3 expression were independent risk factors for LNM (p<0.05). A large tumor size, mixed histology, submucosal invasion, perineural invasion, and ulceration were determined as risk factors using univariate analysis (p<0.05). The tumor cutoff size for predicting LNM was 2.65 cm, based on a ROC analysis. The test set study verified that stromal desmoplasia, submucosal invasion, and high VEGF-C expression were independent risk factors for LNM (p<0.05). Moreover, mixed histology, lymphatic invasion, ulceration, and high SMAD 2/3 expression were identified as additional risk factors using univariate analysis (p<0.05). Conclusion: Stromal desmoplasia, submucosal invasion, and high VEGF-C expression are potential biomarkers for LNM in EGC. VEGF-C expression might serve as an adjunct biomarker for predicting LNM on forceps-biopsy tissue at initial diagnosis.