Hepatic fat - genetic risk score predicts hepatocellular carcinoma in HCV cirrhotic patients treated with DAAs.

2020 
Background & aims Genetic factors and steatosis predispose to hepatocellular carcinoma (HCC) in chronic hepatitis C virus (HCV) patients, however their impact in cirrhotics cured by direct-acting antivirals (DAA) is still undefined. We assessed the association between a genetic risk score of hepatic fat accumulation (GRS), combining variants in PNPLA3, MBOAT7, TM6SF2, GCKR and HCC in DAA-treated patients. Approach & results We considered 509 consecutive patients with HCV cirrhosis (defined histologically or when liver stiffness ≥12 kPa) treated with DAA. HCC was diagnosed according to international recommendations. GRS was calculated from the weighted impact of single variants on hepatic fat content quantified by H1 -spectrometry in the general population (Dallas Heart Study). During a median follow-up of 43 (3-57) months after DAA start, 36/452 (8%) patients developed de novo HCC, 4-year cumulative probability being 9% (95% CI 7-12%). Male sex (HR 2.54, p=0.02), diabetes (HR 2.39, p=0.01), albumin (HR 0.35, p=0.001) and GRS score >0.597 (HR 2.30, p=0.04) were independent predictors of de novo HCC. In contrast, single genetic risk variants were not useful in stratifying HCC risk. The proportion of patients who developed HCC according to the combination of the independent risk factors ranged from 11% to 67%. HCC recurred in 28/57 (49%) patients with previous history; diabetes and ethnicity were the only independent predictors of HCC recurrence. Conclusions In a large cohort of HCV cirrhotic DAA-treated patients, GRS was associated with de novo HCC independently of classical risk factors including liver disease severity. These data suggest that hepatic fat - lipotoxicity promotes HCC in this setting and may represent a target for chemoprevention. Combination of clinical and genetic predictors may improve HCC risk stratification.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    38
    References
    19
    Citations
    NaN
    KQI
    []