Clinical and therapeutic implications of hepatitis C virus compartmentalization

Background & Aims: Blood mononuclear cells (BMCs) frequently are infected by hepatitis C virus (HCV) variants that are not found in plasma. The influence of this compartmentalization on the natural and therapeutic outcome of hepatitis C is unknown. Methods: We studied 119 patients with previously untreated chronic HCV infection. Sixty-five of these patients started first-line treatment with pegylated interferon-alfa and ribavirin after enrollment in the study. The internal ribosomal entry site (IRES) of HCV RNA was amplified and compared between plasma and BMCs by means of single-strand conformational polymorphism (SSCP) analysis, line-probe assay, and cloning sequencing. Results: The IRES SSCP patterns differed between plasma and BMCs in 54 (48%) of 113 assessable patients. Twenty-seven (24%) of these patients were co-infected by 2 HCV types or subtypes, only 1 of which was detectable in BMCs (n = 25) or in plasma (n = 2). SSCP-defined compartmentalization was more frequent in former drug users than in others (35/56 [60%] vs 19/56 [34%]; P < .01), and less frequent in patients with genotype I HCV in plasma (26/73 [24%] vs 28/40 [65%]; P < .01). The only variables that were independently predictive of a sustained virologic response were SSCP-defined compartmentalization (25/31 vs 10/32; P = .0001) and genotype 2 or 3 infection of BMCs (22/31 vs 8/34; P = .002). Conclusions: A significant proportion of patients with hepatitis C are co-infected by 2 or more HCV variants with distinct IRES sequences and distinct cellular tropism. This compartmentalization is a strong independent predictor of treatment efficacy.