OBJECTIVES

The aim of the study was to assess the regression of liver stiffness after successful direct-acting antiviral (DAA) treatment in patients with hepatitis C virus (HCV) monoinfection and HCV/-HIV coinfection. In addition, we aimed to identify factors associated with liver stiffness regression.

METHODS

We studied patients treated with interferon-free DAA regimens with a sustained virological response at week 12 (SVR ) or 24 (SVR ) post-treatment. Liver stiffness was assessed by transient elastography (TE) before the initiation and after the end of treatment (median 12 weeks).

RESULTS

Of 214 enrolled patients, 85 (40%) were HCV monoinfected and 129 (60%) HCV/HIV coinfected. Baseline median TE values were 7.8 kPa [interquartile range (IQR) 5.9-12.0 kPa] in mono-infected patients and 10.7 kPa (IQR 7.8-17.0 kPa) in coinfected patients. Overall, the median TE value decreased from 10.1 to 6.8 kPa (n = 214; P < 0.0001). There was no difference between mono- and coinfected patients (-2.2 versus -3.3 kPa, respectively; P = 0.88), which was verified by an analysis of covariance (ANCOVA) adjusting for baseline TE values. Significant (≥ 30%) regression of liver stiffness was achieved by 45% of patients (54% with baseline TE ≥ 7.1 kPa). In multivariate analysis, a prior HCV treatment was a negative predictor of liver stiffness regression [odds ratio (OR) 0.31; P = 0.001]. A higher baseline TE value was positively associated with achieving a significant regression (OR 1.06; P = 0.02). HIV coinfection status, HCV genotype, age, sex, treatment duration, controlled attenuation parameter value, bilirubin concentration, platelet count and aspartate aminotransferase concentration were not associated with liver stiffness regression.

CONCLUSIONS

Regression of liver stiffness after successful DAA treatment did not differ in patients with HCV monoinfection and those with HCV/HIV coinfection. Half of all patients achieved a significant (≥ 30%) regression. Prior treatment for HCV was a negative predictor for this endpoint, while a higher baseline TE value was positively associated with regression.