Current definitions of response to interferon (IFN) depend upon decreasing the serum alanine aminotransferase (ALT) level to normal during therapy. Although this complete response is often associated with loss of detectable hepatitis C virus (HCV) in serum, the frequent occurrence of relapse indicates that eradication of the virus is not achieved. Thus, consideration of more meaningful end points for IFN response is necessary. To begin to define a more effective end point, we recently tested serial serum samples from 15 patients with three different biochemical response patterns (five patients, no response; five, complete response with relapse; five, complete and sustained response) for HCV-RNA by both reverse transcription polymerase chain reaction (RT-PCR) and branched DNA signal amplification assay. In the IFN nonrespondes, HCV-RNA was detected in serum and liver at completion of IFN therapy in all five patients. Similarly, in the complete responders followed by relapse, HCV-RNA was still detected in serum and liver at the end of IFN therapy in four and five patients, respectively, despite complete normalization of serum ALT. In the complete and sustained responders, HCV-RNA was detected in serum and liver at completion of IFN therapy in only one and no patients, respectively. However, all five sustained biochemical responders had virologic relapse detected at one or more time points within 6 months after cessation of IFN therapy despite persistently normal ALT. HCV-RNA remained detectable in serum at the pretreatment level without biochemical abnormalities for a range of 1 1/4 to 3 years.(ABSTRACT TRUNCATED AT 250 WORDS)