Patients with cirrhosis are at risk of hepatocellular carcinoma (HCC) development and, according to current guidelines, should undergo surveillance by ultrasound at six month intervals. Due to the known limitations of surveillance strategies based on ultrasonography, the use of tumor biomarkers, although debated, is common practice in many centers. The aim of the study was to identify the best cut-off value for one of such biomarkers, protein induced by vitamin K absence, or antagonist-II (PIVKA-II). We retrospectively enrolled 1187 patients with liver cirrhosis: 205 with a diagnosis of HCC (median age 67 years, 81.0% males) and 982 without tumor (median age 64 years, 56.2% males). During a median follow-up (FU) of 34.6 (11.4–43.7) months, 118 out of 982 (12.0%) patients developed HCC. Serum PIVKA-II was assessed by chemiluminescence immunoassay on the Lumipulse® G600 II platform (Fujirebio, Tokyo, Japan). In the overall cohort (n = 1187), PIVKA-II showed an area under the curve (AUC) of 0.802 for HCC detection. The best cut-off value that maximized sensitivity was 50 mAU/mL (sensitivity = 80%, specificity = 64%). In the 982 patients without HCC at baseline, PIVKA-II > 50 mAU/mL was associated with an increased risk of HCC development during the FU (HR = 1.74, 95% CI 1.21–2.51; p = 0.003)). In conclusion, the evaluation of serum PIVKA-II showed a good performance for HCC detection; a cut-off value > 50 mAU/mL could be suitable for the surveillance of patients who are at risk of developing HCC.
