Hepatocellular Carcinoma Risk in Hepatitis C Stage-3 Fibrosis After Sustained Virological Response With Direct-acting Antivirals

María Sánchez-Azofra; Inmaculada Fernández; María L. García-Buey; Lourdes Domínguez-Domínguez; Conrado M. Fernández-Rodríguez; Antonio Mancebo; Lucía Bonet; Pablo Ryan; Francisco Gea; Antonio Díaz-Sánchez; Marian García-Mayor; Luz Martín-Carbonero; Pilar Castillo; María L. Manzano; Leticia González-Moreno; Federico Pulido; María L. Gutiérrez; José M. Moreno; Irene M. García-Amengual; Guillermo Cuevas; Antonio Guerrero; Miguel Rivero-Fernández; María E. Portales; María L. Montes; Antonio Olveira


Liver International. 2021;41(12):2885-2891. 

In This Article

Material and Methods

We performed a multicentre, ambispective, observational study of patients from the Gastroenterology/Hepatology and Internal Medicine Departments of 12 Spanish hospitals.

Inclusion criteria: Chronic hepatitis C, baseline stage 3 liver fibrosis and SVR after DAA-based treatment started between January and December 2015. Exclusion criteria: HCC diagnosed before SVR, concomitant liver disease (other than risk factors for fatty liver), cirrhosis and portal hypertension.

Choice of regimen and duration of therapy were at the discretion of the treating physician. Therapy consisted of the standard combinations of simeprevir, paritaprevir, ledipasvir, ombitasvir, daclatasvir, sofosbuvir and dasabuvir with or without ribavirin. Treatment duration ranged from 12 to 24 weeks.

SVR was defined as a negative viral load at 12 weeks after the end of treatment. Stage 3 liver fibrosis was determined in a two-step procedure according to baseline (pretreatment) values. Initially, all patients with TE values ranging from 9.5 to 14.5 kPa were selected.[15,16] Subsequently, we excluded those with a nodular liver surface, splenomegaly (spleen long axis ≥13 cm), ascites or porto-systemic collaterals detected on imaging, thrombocytopenia (<120 × 109/L in HCV monoinfected patients, <100 000 × 109/L in HIV/HCV-coinfected patients) or endoscopic evidence of esophago-gastric varices.

We collected all available information on patient demographics (sex, age), metabolic syndrome and its components, liver test results, HCV genotype, harmful alcohol use (>30 g/d in males, >20 g/d in females), smoking, HIV status, concomitant drugs. Abdominal ultrasound was performed every 6 months, and, if a new focal lesion was detected, HCC was diagnosed in accordance with international guidelines.[9,10]

The study was conducted in accordance with the guidelines of the Declaration of Helsinki and the principles of Good Clinical Practice. All original study protocols were reviewed and approved by the Ethics Committee of La Paz University Hospital (PI-2843/2017).

Statistical Analyses

Qualitative variables were expressed as frequencies and percentages (95% confidence interval [CI], where applicable) and standard quantitative variables as mean and standard deviation (SD), minimum and maximum and median and interquartile range (IQR). Differences in the percentage distribution of qualitative variables were analysed using the chi-square test or Fisher exact test, as appropriate. Continuous variables were compared between groups using the t test or Mann-Whitney test if they were normally distributed. Normality was tested using the Kolmogorov-Smirnov test with a Lilliefors correction. The time to onset of HCC and probability of survival were estimated using the Kaplan-Meier method, and predictors of HCC were analysed using the Cox proportional hazards method. Univariate Cox regression analysis was performed with those variables that showed a statistically significant correlation with the main event in the bivariate analyses or were of special interest in the study. Variables with a P value <.1 or considered clinically relevant were included in the multivariate analysis. Age was analysed in quantitative and qualitative terms (according to the age limits for the highest known risk of HCC), and sex and age were combined in a single variable in the multivariate analysis to reduce the number of covariables given the low number of events. All statistical analyses were performed with SPSS 24.0 IBM (IBM Corp.). Statistical significance was set at P < .05.