Chronic hepatitis C and interferon alpha: conventional and cumulative meta-analyses of randomized controlled trials

OBJECTIVE: The purpose of this study was to evaluate the clinical usefulness of surrogate markers of the interferon effect (i.e., alanine aminotransferase levels and serum HCV-RNA status) as predictors of long term response, and to identify the optimal schedule of treatment for patients with chronic hepatitis C by means of meta-analysis. METHODS: Pertinent randomized clinical trials and prospective studies were selected using MEDLINE (1986-1996), a reference list from published articles or reviews. Twenty-six prospective studies reporting data on surrogate markers of interferon response were selected. Thirty-nine trials comparing interferon alpha to no treatment and 25 trials comparing different schedules of interferon were reviewed. Conventional meta-analysis according to the DerSimonian and Laird method was used for the pooling of results. RESULTS: The pooled probability of late relapse among sustained responders with negative serum HCV-RNA 6 months after treatment was very low (8.7%; 95% confidence interval 5.8-11.6%). The overall risk difference between treated and control groups was 16.63% (95% confidence interval 11.95-21.31%) for sustained aminotransferase normalization. Therapy with higher interferon dose compared with standard dose significantly improves the rate of sustained response (pooled risk difference 10.56%, 95% CI 5.47-15.65%). Cumulative meta-analyses suggest that a clear dose-response relationship exists across a wide range of interferon dosages. The multivariate meta-regression model confirms that the total interferon dose is an independent predictor of sustained response and that it seems more important than the length of treatment. CONCLUSIONS: Testing for serum HCV-RNA, 6 months after interferon therapy in sustained biochemical responders, is useful for predicting long term response. The current standard total interferon dose of 234 mega-units is suboptimal. Further trials that directly compare different schedules of treatment are needed.