Hepatocellular apoptosis plays a major role in the pathogenesis of chronic hepatitis C. It can be measured noninvasively by determining the circulating levels of cytokeratin-18 fragments. We hypothesized that the effect of antiviral therapy on this parameter will be different in patients with a sustained virological response, relapse (REL) and nonresponse (NR). We quantified cytokeratin-18 fragments in plasma of patients participating in the Swiss Hepatitis C cohort, who received antiviral therapy without stopping because of sides effects. A total of 315 patients were included, 183 with a sustained response, 64 with NR and 68 who relapsed. Mean levels +/-SD of circulating cytokeratin-18 fragments before therapy were 174 +/- 172 U/L for responsders, 188 +/- 145 for nonresponders and 269 +/- 158 U/L for patients who relapsed. The values were significantly higher in the REL group (ANOVA P < 0.006). A sustained response was associated with a significant improvement of the plasma levels (94 +/- 92 U/L, paired test P < 0.000001), whereas there was no improvement in the nonresponder group (183 +/- 158 U/L) and in the relapser group (158 +/- 148 U/L). There was a weak correlation between alanine aminotransferase (ALT) and cytokeratin-18 fragment levels (r(2) = 0.35, P < 0.000001) before therapy but not after therapy and none with hepatitis C virus (HCV) viremia. Successful antiviral therapy results in a significant decrease in circulating levels of cytokeratin-18 fragments arguing for a reduction in hepatocellular apoptosis after clearance of the HCV. Baseline cytokeratin-18 fragment levels are higher in relapsers. Correlations with ALT are weak, suggesting that these two tests measure different but related processes.