Abstract: Objective   To explore the expression of microRNA-98 （miR-98） and microRNA-183 （miR-183） in gastric cancer and their clinical significances. Methods  From January 2013 to December 2016， 90 cases of gastric cancer tissues and paired adjacent tissues were collected in our hospital. Quantitative real-time PCR （QPCR） was performed to evaluate the expression of miR-98 and miR-183. The relationships between levels of miR-98 and miR-183 with the clinicopathological parameters （sex， age， lymph node metastasis， differentiation， distant metastasis， TNM staging， depth of infiltration， Lauren typing and tumor size） were analyzed. The receiver operating characteristic curve （ROC） was used to analyze the value of both alone and combined in predicting gastric cancer. Pearson correlation analysis was used to analyze the correlation between miR-98 and miR-183 expression. Results   The expression levels of miR-98 and miR-183 in cancer tissues were 1.567±0.124 and 0.629±0.097， respectively. Compared with adjacent tissues， there were upregulated levels of miR-98 but downregulated levels of miR-183 in gastric cancer tissues （P＜0.01）. Meanwhile， there was a negative correlation between miR-98 and miR-183 （r=-0.2995， P=0.004）. The expression of miR-98 was related to lymph node metastasis and differentiation， while miR-183 expression was related to lymph node metastasis， distant metastasis and tumor size （P＜0.05）. The ROC curve analyses revealed that miR-98 had considerable diagnostic accuracy， yielding an AUC （area under curve） of 0.794 with 61.11 ％ sensitivity and 95.56％ specificity in discriminating gastric cancer tissues. The AUC of miR-183 was 0.824 with the sensitivity and specificity of 76.67％ and 76.67％. The AUC of combination of miR-98 and miR-183 was 0.903 with sensitivity and specificity of 74.44％ and 91.11％. Conclusion   MiR-98 is highly expressed in primary gastric cancer， while miR-183 is abnormally low， and there is a negative correlation between miR-98 and miR-183. Both miRNAs are related to metastasis and differentiation in gastric cancer. These findings indicate that miR-98 and miR-183 might serve as the potential biomarker for the detection of gastric cancer.

Key words: Gastric cancer, MicroRNA-98, MicroRNA-183, Diagnosis