Hepatocellular carcinoma incidence and risk stratification in patients with metabolic dysfunction-associated steatotic liver disease on long-term follow-up
G.P. Caviglia,S. Pelusi,A. Liguori,M. Guariglia,C. Rosso,A. Armandi,K. GJini,N. Viceconti,M.L. Abate,L. Miele,L. Valenti,E. Bugianesi
DOI: https://doi.org/10.1016/j.dld.2024.01.019
IF: 5.165
2024-02-01
Digestive and Liver Disease
Abstract:Background Metabolic dysfunction-associated steatotic liver disease (MASLD) is projected to become the leading cause of hepatocellular carcinoma (HCC) worldwide. The identification of novel tools able to stratify the risk of HCC is an unmet medical need. Aim To investigate the performance of non-invasive tests (NITs) and protein induced by vitamin K absence or antagonist-II (PIVKA-II) for HCC risk stratification in MASLD patients on long-term follow-up (FU). Materials and Methods We retrospectively enrolled 366 MASLD patients (age: 58, 49–66 years; males: 198 [54.1%]; T2DM: 168, [45.9%]) with significant liver fibrosis (LSM≥8kPa or F≥2 at liver biopsy); 279 (76.2%) had compensated advanced chronic liver disease (cACLD) (LSM≥10kPa or F≥3 at liver biopsy). NITS (i.e. APRI, FIB-4, and NFS) were calculated at baseline. Serum PIVKA-II was measured by CLEIA (Lumipulse®G600II, Fujirebio). Results During 2.9 (IQR 1.0–4.3) years of FU, 19/366 (5.2%) patients developed HCC (1.49 per 100 person/years). Notably, all the 19 HCC occurred in cACLD patients (n=279) (2.06 per 100 person/years). Most patients had early-stage HCC (BCLC=0/A, n=14). In the whole cohort (n=366), NFS showed the best performance for HCC prediction (C-index=0.81), followed by FIB-4 (C-index=0.78), and by APRI (C-index=0.71); the accuracy of serum PIVKA-II was C-index=0.76. In patients with cACLD (n=279), the predictive performance of NITs was C-index=0.77 for NFS, C-index=0.75 for FIB-4, and C-index=0.68 for APRI, while PIVKA-II showed C-index=0.74. Remarkably, in cACLD patients the combination of NFS+PIVKA-II significantly improved the accuracy for HCC prediction (C-index=0.84), allowing patients to be stratified into 3 risk categories with different HCC incidence: low-risk (0/52; 0%), medium-risk (9/190; 4.7%), and high-risk (10/37; 27.0%) (p<0.001). Conclusion In MASLD patients with significant fibrosis, NFS showed appropriate performance for HCC prediction. However, in patients with more advanced liver disease, serum PIVKA-II may be a useful diagnostic complement to improve HCC prediction and risk stratification.
gastroenterology & hepatology