Abstract Background: Recent reports suggest a link between aberrant expression of microRNAs (miRNAs, miR) and various cancers, including hepatocellular carcinoma (HCC). We examined whether measurement of serum levels of 3 miRNAs (miR16, miR195, and miR199a) could be used to differentiate HCC from HCV-related chronic liver disease (CLD). Method: The 3 miRNAs were measured using real-time PCR-based assays in sera from 104 patients with HCC, 59 patients with CLDs, and 67 normal control subjects. Small nuclear RNA U6 was used as an internal control. Cutoff values were determined by visual inspection of miRNA distribution curves. Serum levels of the conventional HCC markers alpha-fetoprotein (AFP), AFP-L3, and des-gamma-carboxyprothrombin (DCP) were also measured, with cutoffs based on established reference ranges. Results: Median serum levels of miR16, miR195, and miR199a were significantly lower in patients with HCC than in CLD and control subjects (all p < 0.01). As a single marker, miR199a had the highest sensitivity for detecting HCC, followed by AFP, DCP, miR16, AFP-L3%, and miR195 (Table). This pattern was unchanged when analysis was limited to HCC cases with tumors <3 cm (Table). The combination of AFP, AFP-L3, DCP, miR199a, and miR16 identified 97 of 104 (93.3%) HCC cases, including 90.7% of those with tumors <3 cm. In cases considered negative with all 3 conventional HCC markers, the 3 miRNAs identified 19 of 26 patients with HCC (73% sensitivity) but also incorrectly identified 9 of 52 CLD patients as having HCC (83% specificity). Conclusions: The combination of AFP, AFP-L3, DCP, miR16, and miR199a yielded greater sensitivity for HCC than any other individual marker or combination evaluated. Use of miRNA as a second line of testing for cases considered negative by with conventional HCC markers holds potential to improve sensitivity and should be explored in larger, prospective studies.Comparison of microRNA markers and conventional markers to differentiate HCC (n=104) from CLD (n=59) CutoffSensitivitySpecificityPPVaNPVAFP20 ng/mL59.188.113.398.6AFP-L310%36.298.339.898.0DCP7.5 ng/mL46.7100.0100.0b98.4miR199a10 (delta CT)62.893.222.398.7miR19515 (delta CT)21.9100.0100.0b97.6miR168 (delta CT)45.788.110.698.1AFP, AFP-L3, DCPCombinationd75.288.116.499.1AFP, AFP-L3, DCP, miR199aCombinationd92.481.413.399.7AFP, AFP-L3, DCP, miR199a, 195Combinationd92.481.413.399.7AFP, AFP-L3, DCP, miR199a, 195, 16Combinationd93.371.29.199.7HCC with small tumor <3cm AFP20 ng/mL37.288.18.897.8AFP-L310%4.798.37.897.1DCP7.5 ng/mL20.9100.0100.0c97.6miR199a10 (delta CT)72.193.224.799.1miR19515 (delta CT)18.6100.0100.0c97.5miR168 (delta CT)32.691.510.697.8AFP, AFP-L3, DCPCombinationd51.288.111.898.3AFP, AFP-L3, DCP, miR199aCombinationd88.481.412.899.6AFP, AFP-L3, DCP, miR199a, 195Combinationd88.481.412.899.6AFP, AFP-L3, DCP, miR199a, 195, 16Combinationd90.771.28.999.6Delta CT, difference in cycle threshold (CT) between target and internal control (small nuclear RNA U6) amplicon.a 3% prevalence was used to calculate PPV.b 64% prevalence was used to calculate PPV when specificity is 100%.c 45.7% prevalence was used to calculate PPV when specificity is 100%.d At least one of the listed analytes was above the cutoff value. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 2723.