Identification of MCM4 as a Prognostic Marker of Hepatocellular Carcinoma

METHODS: MCM4 expression difference in HCC were analyzed from TCGA and GEO data and verified by real-time PCR and western blot. ROC curve was used to analyze the diagnostic value of MCM4 and AFP. Additionally, the relationship between MCM4 and stage or nodal metastasis status or grade or age in TCGA cohort with HCC was observed from the UALCAN website. The univariate and multivariate Cox and functional analyses were done to explore the prognostic value of MCM4 in TCGA cohort. RESULTS: It was found that MCM4 was significantly highly expressed in HCC tissues from TCGA, GEO, and experimental data. Furthermore, ROC curve analysis showed that MCM4 was superior to be a diagnostic biomarker than AFP from TCGA (AUC(MCM4) = 0.9461, AUC(AFP) = 0.7056) and GEO (GSE19665: AUC(MCM4) = 0.8800, AUC(AFP) = 0.5100; GSE64041 AUC(MCM4) = 0.8038, AUC(AFP) = 0.6304). AUC of MCM4 from real-time PCR result in 60 pairs of HCC and adjacent tissues was 0.7172, demonstrating the prediction value of MCM4. Besides, different expression tendencies of MCM4 among different stages or nodal metastasis status or grade or age were observed from the UALCAN website. In addition, multiROC analysis showed the advantage of MCM4 as a survival prediction at 1, 3, and 5 years with the higher AUC at 0.69 of 1 year, 0.65 of 3 years, and 0.61 of 5 years. It was shown that MCM4 was independently associated with OS in univariate and multivariate Cox analysis. And GSEA displayed that MCM4 was highly enriched in KEGG_CELL_CYCLE signaling pathway following higher correlation positively with CDC6, PLK1, CRC1, and BUB1B in HCC. CONCLUSION: MCM4 might be a potential biomarker in guiding the prognostic status of HCC patients.