Longitudinal monitoring of liver fibrosis status by transient elastography in chronic hepatitis B patients during long-term entecavir treatment.

The correlation between improvement in longitudinal liver stiffness and fibrosis regression has not been properly evaluated during long-term antiviral therapy in chronic hepatitis B (CHB) patients. In this study, liver stiffness was serially performed by FibroScan ^® every 26 weeks in a prospective cohort of CHB patients receiving entecavir. Results were compared with liver biopsies at baseline and week 78. A total of 120 treatment-naïve CHB patients were analyzed, in which 54 (45%) patients had fibrosis regression at 78 weeks of antiviral therapy. Liver stiffness measurement presented as a rapid-to-slow decline pattern and decreased more significantly in patients with fibrosis regression than those without improvement in fibrosis at week 78 (- 46.4 vs. - 28.6%, P < 0.001). Multivariate analysis revealed that percentage decline of 52-week and 78-week liver stiffness from baseline was independent predictive factors for fibrosis regression (OR = 46.6, P < 0.001; OR = 17.8, P = 0.002, respectively). Moreover, percentage decline of 78-week liver stiffness was moderately predictive of fibrosis regression (AUROC = 0.694, P < 0.001), while the optimal cutoff values were different between non-cirrhosis and cirrhosis patients (38 vs. 45%). Fibrosis regression could be predicted with a high positive predictive value (96%) in non-cirrhosis patients and could be excluded with a high negative predictive value (94%) in cirrhosis patients. In conclusion, serial liver stiffness measurement could be applied for longitudinal monitoring of fibrosis status in CHB patients. Continuous decline of liver stiffness after effective antiviral treatment could partially reflect fibrosis regression at an optimal cutoff value.