These were age at infection, patient gender, HCV genotype, the presence of steatosis, BMI, and inflammatory grading. An interaction term was specified between steatosis and viral genotype, to reflect that the effect of steatosis on fibrosis progression may be specific for different genotypes. The final model was chosen on the basis of the minimum Akaike information criterion. The results summarized in Table 3 show that age at infection was the variable most strongly and positively associated with

fibrosis progression RAD001 in vivo (P < 2E-16). Male gender (P = 0.012), HCV genotype 3 (P = 7.19E-04), and the presence of steatosis (P = 0.012) also resulted in being significantly associated with an accelerated rate of fibrosis progression. Importantly, the two IL28B SNPs CHIR-99021 solubility dmso had no significant effect on the outcome, and thus they were removed from the final model (Table 3). It is

worth pointing out that HCV genotype 2 was associated with a slower rate of fibrosis progression (P = 0.034), as also suggested in a previous study.14 Single-term deletion analysis further confirmed the role of these explanatory variables. Figure 1 shows the remarkable effect of age at infection on rate of fibrosis progression. In contrast, no effect could be attributable to the host IL28B genotype. Notably, no impact on FPR was found, also assuming a dominant model of inheritance for the alleles, rs8099917 G or rs12979860 T, previously associated with treatment failure (data not shown). On the contrary, when these two SNPs were tested for association with therapy outcome, using data from 91 patients with HCV genotype 1 and available treatment information, the effect was readily detectable (see Supporting Information). Actually, rs12979860 genotypes CT and TT versus CC resulted in being associated

with treatment failure (P < 0.01, odds ratio = 3.6, 95% CI = 1.3-10.2), in agreement with previous reports. Finally, we evaluated the role of explanatory variables, excluding individuals that acquired Rebamipide the infection at birth, to exclude possible confounding factors originating from the inclusion of a group of patients that acquired the infection in the setting of an immature immune system. In this subgroup of patients, we obtained the same estimates of the parameters and confirmed the lack of any effect by the two IL28B polymorphisms and the significant role of the nongenetic factors (Supporting Table 1). In summary, the model outlined in Table 3—which includes patient gender, age at infection, viral genotype, and steatosis—explained an estimated 34% of the phenotypic variability, showing that patient age at infection has a major, highly significant role on fibrosis progression, in agreement with other reports.14-16 The estimated effect on fibrosis progression of each additional year at infection was a 2.8% (95% CI = 2.2%-3.4%) increase in FPR.