Effect of low-level HCV viraemia at week 24 on HCV treatment response in genotype 1 patients.

BACKGROUND We examined the detection of low-level viraemia at week 24 as a predictor of sustained virological response (SVR) and viral relapse/breakthrough, and the agreement between the Roche Cobas TaqMan™ HCV RNA assay (TaqMan) and Roche Cobas(®) Amplicor HCV qualitative assay (Amplicor; both Roche Molecular Diagnostics, Pleasanton, CA, USA) for detection of low-level viraemia. METHODS A total of 871 treatment-naive HCV genotype 1 patients participating in an induction-dose pegylated interferon therapy study had virological responses assessed using TaqMan. A total of 151 patients with HCV RNA levels ≤500 IU/ml had samples tested in parallel using the Amplicor and TaqMan assays. RESULTS SVR was significantly lower and relapse/breakthrough significantly higher in patients with low-level residual viraemia at week 24 compared with those who had undetectable viraemia: SVR was 72%, 29% and 14% (P<0.0001) and relapse/breakthrough 28%, 71% and 86% (P<0.0001) in patients with viraemia that was undetectable, detectable <15 IU/ml and detectable 15-<50 IU/ml, respectively, at week 24. The negative predictive value (NPV) for a week-24 virological response for SVR was 86%, 90% and 90% using TaqMan cutoffs of undetectable, <15 IU/ml and <50 IU/ml, respectively. The percentage agreement between Amplicor and TaqMan was similarly high for TaqMan cutoffs of 50 IU/ml and 15 IU/ml, but lower for undetectable viraemia (83%, 83% and 70%, respectively). CONCLUSIONS These data emphasize the importance of achieving undetectable HCV RNA during pegylated interferon therapy to maximize SVR; however, the current 24-week stopping rule of undetectable HCV RNA appears too stringent when using sensitive PCR assays given the observed lower NPV for SVR using the TaqMan undetectable cutoff. Our data also suggest that a TaqMan <15 IU/ml result is comparable to an Amplicor-negative result (that is, below the assay cutoff value) when monitoring viral response.