However, it is clear that these two variables are insufficient Recent studies performed in Japan have suggested to distinguish long-term from nonresponders or transient rethat hepatitis C virus (HCV) genome heterogeneity sponders precisely. To make this distinction, several studies might be taken as a predictive virological parameter of from Japan, based on single-strand conformation polymorresponse to interferon alfa (IFN-a) treatment. However, phism (SSCP) and/or sequencing procedures, have suggested there is presently no information on the impact of this that the degree of HCV genome complexity might influence virological parameter in patients from Western counthe efficacy of IFN-a therapy. [16][17][18][19][20][21] HCV is an RNA virus that tries infected by different HCV genotypes. We have inshows marked genetic variability, especially in some domains vestigated this issue by using amplification of HCV E2 of the envelope protein. [22][23][24] This high rate of mutation achypervariable region 1, followed by single-strand conforcounts for the circulation in a given infected patient of a popumation polymorphism assay (PCR-SSCP). We have studlation of HCV RNA molecules, referred to as quasispecies, and ied 95 French patients infected with various genotypes this heterogeneity can be rapidly evaluated by polymerase and treated with IFN-a-2b. We analyzed the impact of chain reaction (PCR) followed by SSCP or heteroduplex gelthe following parameters by univariate and multivariate shift analysis. 22,25,26 analyses: HCV heterogeneity, HCV genotype, viral load, However, none of these studies has analyzed whether the and liver histology in response to therapy. Age ú40 years complexity of the HCV genome is really an important pre-(P õ .01), viral load ú35 1 10 5 Eq/mL (P õ .01), genotype dictive parameter of IFN-a response in a large series of pa-1 (P õ .01), and a number of SSCP bands ú3 (P õ .001) tients from Western countries infected by different HCV gewere significantly associated with nonresponse or renotypes. In particular, it is not presently known whether this lapse; cirrhosis was associated with nonresponse. In correlation holds in subjects infected with HCV genotypes multivariate analysis, three variables were indepenother than 1b. We have therefore investigated this issue in dently associated with the absence of long-term rea series of French patients treated with IFN-a-2b for HCVsponse: SSCP bands ú3, genotype 1, and viral load ú35 related chronic active hepatitis. ; thus, most subjects will show either months (3 times per week) at a dose of 3 MU for 57 patients and for nonresponse or relapse after therapy.2 Therefore, the identifi-the remaining 38 patients at a dose of 6 MU for 6 months followed cation of prognostic factors that would allow the distinction by 3 MU for another 6 months. There was no significant difference between the groups before IFN administration according to age (44.2 between long-term responders and nonresponders or relapsers importance of virological parameters: both HCV genotype and...