Abstract: Objective:To construct a prognostic risk model of long non-coding RNA(lncRNA) associated with microvascular invasion (MVI) of hepatocellular carcinoma (HCC) and screen for key MVI-mRNAs.Methods:MVI-mRNAs were extracted from transcriptome data based on TCGA-LIHC database, prognosis-related MVI-lncRNAs were obtained by correlation analysis and univariate Cox analysis, and a related risk model was constructed by multivariate Cox analysis to screen variables.The risk model was evaluated using Kaplan-Meier analysis, univariate and multivariate Cox analysis, receiver operating characteristic (ROC) curve, and principal component analysis (PCA).Differential analysis of 34 HCC-MVI-mRNAs was performed according to high-and low-risk groups, and GO and KEGG enrichment analysises were performed.RNA sequencing (RNA-seq) and real-time fluorescence quantitative polymerase chain reaction (RT-qPCR) were used to validate MVI-mRNAs in the cancer tissues from HCC patients and paracancerous tissues.Results:9 prognostically relevant MVI-lncRNAs were identified, and a MVI-lncRNA risk model with prognostic value was constructed by 3 lncRNAs highly correlated with MVI: AC129492.1, NRAV, and AC099850.3.The overall survival (OS) of the high-risk group was shorter than that of the low-risk group (P< 0.05).ROC results suggested that risk scores (AUC=0.819, 95% CI: 1.576-2.250) were a more accurate predictor of patient survival than clinical factors.MVI-mRNAs were analyzed for differential expressions by high-and low-risk groups, and 18 key differentially expressed genes (DEGs) were screened.GO and KEGG enrichment analysis showed that the MVI of HCC was mainly related to small molecule catabolism and nucleotide metabolic pathways.The expressions of AASS and GLYAT in HCC were significantly low, while TYMS was highly expressed(P< 0.05).Conclusion:The MVI-lncRNA model established in this study can be used to predict the prognosis of HCC, and MVI-related genes of TYMS, AASS and GLYAT are differentially expressed in HCC tissues.