Article Text
Abstract
Introduction Chronic hepatitis C virus infection (HCV) is a common cause of cirrhosis and end-stage liver disease. Pegylated interferon (PEG-IFN) and ribavarin (RBV) is currently the treatment of choice for genotype 3 (G3) HCV resulting in a sustained virological response (SVR) in 70–80%. Advanced fibrosis is known to be associated with failure of antiviral therapy. Increasingly, liver stiffness measurement (LSM) is being used to non-invasively assess fibrosis. However, it is not known whether LSM predicts response to antiviral therapy and whether there are predictive cut-offs. Our aim was to assess whether baseline LSM can predict SVR in HCV G3 patients treated with PEG-IFN+RBV.
Methods Retrospective review of outcomes in naive patients with HCV G3 treated with PEG-IFN+RBV in our clinic from Jan 2007 to Oct 2011. Post transplant and co-infected patients were excluded. Patients with a valid LSM within 1 year of starting treatment who completed > 12wks and recorded outcome of treatment were included in the LSM analysis.
Results 144 patients (mean age 40±10 years, 56% male, 16% cirrhotic, and 42% high viral load) received PEG-IFN+RBV for HCV in the study period. 92% completed > 12 wks treatment. 92 (64%) of patients had a valid LSM (median 6.5kPa; 3.5kPa to 39.1kPa). 24% had a LSM > 10.6kPa consistent with advanced fibrosis. The overall SVR rate was 68%. 11% were lost to follow up and the outcome unknown. LSM was significantly associated with SVR (p = 0.001). The AUROC for LSM in predicting treatment response was 0.74 (95% CI 0.58–0.90). The optimum cut-off to predict non-SVR was 10.6kPa (69% sensitivity, 85% specificity). 90% with LSM ≤ 10.6kPa achieved SVR versus 47% with LSM > 10.6kPa (p < 0.001). All patients with low viral load (< 600,000 IU/ML) and LSM < 10.6 kPa who had > 12 wks treatment achieved SVR (n = 33).
Conclusion Fibrosis assessed non-invasively with LSM can help predict response to antiviral therapy in patients with HCV G3. LSM (> or < 10.6kPa) could be factored into treatment algorithms to determine the optimum treatment course lengths.
Disclosure of Interest None Declared.