In this study, the clearance rate and factors influencing spontaneous HCV clearance were assessed in a prospective cohort of retrospectively identified DU with acute HCV, regardless of their clinical presentation at the time of acute HCV infection. To our knowledge, this is one the largest longitudinal studies on factors associated with spontaneous HCV clearance in individuals with drug-use-related acute HCV infection. The rate of spontaneous HCV clearance was 33.0%. Our main finding is that women with the favorable genotype for rs12989760 were more likely to clear HCV (OR 6.62, 95% 2.69–26.13), whereas females with the unfavorable genotype were as likely as men with the favorable and unfavorable genotype to clear HCV.
The rate of clearance we found is higher than observed in studies among acute clinical cases (reviewed by Micallef 
), but may be underestimated in injecting DU. Many injecting DU experience repeated exposure to HCV and HCV re-infection after their initial HCV seroconversion, due to continuing risk behavior 
. Such re-infection after clearance might result in persistent or recurrent HCV viremia, leading to an underestimation of the clearance rate. However, in this cohort of DU with acute HCV, we did not find an association between ongoing risk behavior and reduced rates of viral clearance shortly after the initial HCV infection.
Studies have suggested that individuals presenting with clinical symptoms after exposure to HCV after needle-stick injury or presenting at an outpatient clinic are more likely to spontaneously resolve acute HCV infection 
. Self-reported fever was associated with HCV viral clearance in this cohort of DU, other clinical symptoms were not.
The association between HCV clearance and absence of HIV was borderline. HIV infection has been associated with loss of viral control of HCV, as evidenced by a higher HCV viral load in HIV co-infected individuals 
. Since HCV is more efficiently transmitted by an infected needle stick than HIV, HCV usually precedes or coincides with HIV infection in DU. Therefore, the number of participants with HIV at HCV seroconversion was small, limiting the power to detect an effect. In addition, therefore all HIV co-infected DU in our study retained high CD4 counts at the point of HCV infection. This might explain why HCV viral load did not significantly differ between co-infected and mono-infected individuals in our study. As in line with our findings that early HIV infection already lowers the HCV clearance rate. We and others have shown that acute HIV co-infection hampers the beneficial HCV-specific CD4+
T-cell responses targeting non-structural proteins in DU 
Interestingly, chronic HBV was borderline significantly associated with HCV clearance in the multivariate model, while cases were limited. Patients with spontaneous viral clearance of chronic HCV after HBV-superinfection have been described 
, and cross-sectional studies have shown that HBsAg-positive HIV-infected HCV-seropositive individuals are more likely to be HCV-RNA-negative than HBsAg-negative HIV-infected individuals 
. Although the effect on liver disease is unknown, chronic HBV infection seems to favor clearance of acute HCV infection 
. Further studies on viral interference including the role of HBV, which may modify HCV replication are warranted.
Having defined viral clearance as two consecutive HCV-RNA-negative visits after anti-HCV seroconversion, we defined HCV viral persistence as the continuous presence of HCV RNA at one or two of these visits, regardless of HCV strain present. Therefore, we did not distinguish between viral persistence of one strain and reinfection by another. Furthermore, although HCV clearance is believed to take place within the first 6 months after acute infection, evidence shows that clearance might take much longer 
. Since we included HCV RNA measurements only in the first 2 years after HCV seroconversion, we recognize that multiple measurements in a longer time span would be necessary to evaluate possible late clearance and its predictors. In our cohort, after a median follow-up after HCV seroconversion of 14.6 years (IQR 7.9–19.6), only 5 out of 71 (7.0%) individuals who did not clear HCV spontaneously within the first 2 years after HCV seroconversion were HCV RNA-negative at the last study visit, without HCV treatment in the meantime, before November 2005 or the penultimate visit preceding death, indicating that late clearance might occur, but is not very frequent (data not shown).
Our main finding is the potential interaction between the favourable genotype of rs12989760 and women in spontaneous clearance of HCV. However, since our study population is relatively small, this finding needs to be confirmed in larger studies among HCV seroconverters. A possible interaction between IL28B and sex in HCV fibrosis progression has been described previously by Falletti et al. 
This retrospective study among 629 cHCV infected patients investigated the role of IL28B on the histological outcome of cHCV infection. One of their findings is that males carrying the favourable C-allele for rs12979860 had a significant increased risk (OR 2.06) for an Ishak staging score >2 as compared to females with the favourable C-allele (reference category), while participants carrying the TT genotype also had an increased risk (OR 2.37) for an Ishak score>2, irrespective of sex.
The potential interaction between female gender and Il28B might be explained by the involvement of toll like receptor 7 (TLR7), a receptor that is involved in recognition of viral products (single stranded RNA) and activation of innate immunity 
. Stimulation of TLR7 in peripheral blood mononuclear cells (PBMC) from women results in significantly higher IFN-α responses as compared to males 
. The TLR7 activation signal is transduced via MyD88 to the IL-1R-associated kinase 1/4 complex that activates interferon regulatory factor 7 (IRF7). Recently, it has been demonstrated that IL28B (IFN-λ) is mainly controlled by IRF7 
. Interestingly, IFN-λ mediates its antiviral activity through the activation of JAK-STAT pathways, similar to IFN-α which is still the major treatment modality for cHCV infection, thereby inducing interferon stimulating genes (ISG) that suppress viral activity. Marcello et al showed, in vitro, that co treatment with both IFN-α and IFN-λ enhanced the antiviral activity, suggestive of a synergistic interaction 
. Whether increased reactivity upon TLR7 stimulation results in both increased IFN-α and IFN-λ responses needs to be investigated.
Next to the relatively small sample size, our study is limited by the fact that variables on behaviour and symptoms are self-reported, including initiation of injecting drug use for those that entered our cohort as anti-HCV positive cases. We believed that we could minimize this bias by only including IDU who started injecting within two years before inclusion into the study. In an earlier report, we have shown that approximately 50% of IDU in the ACS become infected with HCV within two years after initiating injecting drug use and self-reports are valid 
. In addition, in a sensitivity analysis including only HCV seroconverters with a small interval between last anti-HCV negative and first anti-HCV positive test, results were comparable.
In conclusion, women with the favorable CC genotype for rs12979860 have the greatest likelihood to spontaneously resolve HCV. The decision to start HCV treatment might be postponed in this group, if not coinfected with HIV. Spontaneous clearance of HCV seems to be primarily driven by host genetic factors and presence of coinfection. The possible synergistic interaction between female sex and the favorable genotype warrants confirmation in larger studies and warrants further study into the immunological and virological mechanisms explaining this finding.