Abstract
BACKGROUND: Hepatocellular carcinoma (HCC) posed significant health problems and deaths. There are various challenges in the management of HCC, including the late detection or diagnosis. The ongoing diagnostic method in HCC also hinders the detection on the early stages of the disease, thus biomarkers need to be explored further for HCC detection. Serum alpha fetoprotein (AFP) and Midkine (MDK) are two proteins which might be the biomarker of choice in the detection of HCC. This meta-analysis aims to analyze the accuracy of Midkine and AFP in the detection of HCC. METHODS: The systematic review and meta-analysis was conducted by adhering to the Preferred Reporting System for Systematic Review and Meta-Analysis (PRISMA) guidelines. We conduct literature screening and selection followed by quality assessment from various databases such as PubMed, MEDLINE, SpringerLink, ProQuest, EBSCOhost, Cochrane, and EMBASE. The included studies were then extracted and analyzed cumulatively using MedCalc and MetaDTA with forest plot and ROC curve as outcome. RESULTS: 12 studies were included in this study. The AFP biomarker yields sensitivity value of 62.5% (97.5% CI 0.442 - 0.778) and specificity value of 95% (97.5% CI 0.842 - 0.986), while the Midkine biomarker denotes sensitivity value of 91.6% (97.5% CI 0.83 - 0.961) and specificity value of 82.2% (97.5% CI 0.83 - 0.96). CONCLUSIONS: Both AFP and MDK are proven to be a good diagnostic tool or biomarker in the detection of HCC. The use of both in combination should provide high quality diagnostic marker for HCC suspected patients. Further studies on this should be conducted.