Prothrombin induced by vitamin K absence or antagonist‐II and alpha foetoprotein to predict development of hepatocellular carcinoma in Caucasian patients with hepatitis C‐related cirrhosis treated with direct‐acting antiviral agents
Elisabetta Degasperi, Riccardo Perbellini, Roberta D’Ambrosio, Sara Uceda Renteria, Ferruccio Ceriotti, Alberto Perego, Corinna Orsini, Marta Borghi, Massimo Iavarone, Mariangela Bruccoleri, Alessandro Rimondi, Annalisa De Silvestri, Angelo Sangiovanni, Pietro Lampertico
Abstract
BACKGROUND: Prothrombin induced by vitamin K absence or antagonist-II (PIVKA-II) and alpha fetoprotein (AFP) are biomarkers for hepatocellular carcinoma (HCC). However, their performance in patients with cirrhosis related to hepatitis C virus (HCV) treated with direct-acting antiviral agents (DAA) is unknown. AIM: To evaluate PIVKA-II and AFP as HCC predictors in DAA-treated patients with HCV-related cirrhosis METHODS: In this single centre study, patients with cirrhosis from chronic HCV infection and with a sustained virological response (SVR) to DAA were tested for PIVKA-II and AFP (Fujirebio, Japan) at the start of DAA treatment (baseline), end of treatment (EOT) and at HCC diagnosis. RESULTS: We included 400 patients with mean age 65 (24-92); 56% were men. From baseline to EOT, PIVKA-II did not change (35 vs 35 mAU/mL, P = 0.43) while AFP significantly decreased (12 vs 6 ng/mL, P < 0.0001). After 52 (3-66) months from baseline, 34 (8.5%) patients developed de novo HCC; median AFP 9 (2-12 868) ng/mL and PIVKA-II 80 (22-1813) mAU/mL. EOT-PIVKA-II (HR 3.05, P < 0.0001) and AFP (HR 2.77, P = 0.001) independently predicted HCC together with diabetes (HR 6.12, P < 0.001) and GGT (HR 1.01, P = 0.03). The 4-year cumulative probability of HCC was 24% vs 2% in patients with EOT-PIVKA-II > or ≤41 mAU/mL (P < 0.0001), and 26% vs 9% for EOT-AFP > or ≤15 ng/mL (P = 0.02). By combining EOT-PIVKA-II and AFP, the 4-year probabilities of HCC were 3% in patients testing negative for both markers, 18% in patients positive for both, and 38% in patients positive for at least one (P < 0.0001). CONCLUSIONS: In patients with HCV-related cirrhosis treated with DAA, PIVKA-II and AFP independently predicted HCC, while their combination improved risk stratification.