Clearance of low levels of HCV viremia in the absence of a strong adaptive immune response

Spontaneous clearance of hepatitis C virus (HCV) has frequently been associated with the presence of HCV-specific cellular immunity. However, there had been also reports in chimpanzees demonstrating clearance of HCV-viremia in the absence of significant levels of detectable HCV-specific cellular immune responses. We here report seven asymptomatic acute hepatitis C cases with peak HCV-RNA levels between 300 and 100.000 copies/ml who all cleared HCV-RNA spontaneously. Patients were identified by a systematic screening of 1176 consecutive new incoming offenders in a German young offender institution. Four of the seven patients never developed anti-HCV antibodies and had normal ALT levels throughout follow-up. Transient weak HCV-specific CD4+ T cell responses were detectable in five individuals which did not differ in strength and breadth from age- and sex-matched patients with chronic hepatitis C and long-term recovered patients. In contrast, HCV-specific MHC-class-I-tetramer-positive cells were found in 3 of 4 HLA-A2-positive patients. Thus, these cases highlight that clearance of low levels of HCV viremia is possible in the absence of a strong adaptive immune response which might explain the low seroconversion rate after occupational exposure to HCV.


Background
Spontaneous clearance of acute infection with the hepatitis C virus (HCV) has been frequently associated with a strong HCV-specific cellular immune response [1,2]. Both HCV-specific CD4+ and HCV-specific CD8+ T cell responses were found to be stronger and longer lasting in chimpanzees and humans who cleared HCV as compared to individuals who developed a chronic course of the infection [3][4][5][6][7][8][9][10][11][12][13]. However, HCV-specific cellular immune responses were not always detectable in chimpanzees who cleared the virus spontaneously [14]. In addition, we and others have previously reported clearance of HCV in the absence of seroconversion to an anti-HCV antibody-positive status [15,16]. Subsequently, HCV-specific T cell responses in HCV-exposed anti-HCV-negative individuals have been demonstrated by several groups [17][18][19][20] although it was not known in the latter studies whether these subjects have transiently been viraemic for HCV or not. We had the chance to study prospectively individuals with acute hepatitis C in the setting of a German young offender institution. By systematic HCV-RNA screening of 1176 consecutive new incoming prisoners, we identified 7 HCV-RNA positive individuals who cleared HCV spontaneously during further follow-up. Four of those did never develop anti-HCV-antibodies. HCV-specific T cell responses could be monitored and were found to be rather weak and not significantly different from chronically infected individuals. Thus, these cases highlight that clearance of HCV-RNA may not necessarily correlate with the appearance of acquired immunity, a finding that challenges our current understanding of the immunopathogenesis of HCV-infection.

Patients
Details of the study cohort were presented previously [16,21]. In brief, all new incoming prisoners to the largest German young offender institution for young men (age [16][17][18][19][20][21][22][23][24] were screened for anti-HCV antibodies and HCV-RNA in the year 2002 (n = 1176). The study was approved by the Ethics Committee of the Hannover Medical School including additional blood sampling to study cellular immune responses.
Subjects tested positive for any of the HCV-markers (n = 97; 8.6% of individuals screened) were offered to participate in a follow-up study on the course of HCV-RNA, ALT levels and analysis of cellular immune responses. Followup data and PBMC of seven patients who had cleared HCV spontaneously were available for this study. In four of them, intravenous drug abuse was the most likely source of infection. Three patients reported unsafe sex before imprisonment as the only risk factor possibly being associated with HCV infection. Characteristics of the seven subjects with cleared HCV-infection are shown in table 1. Thirteen patients with chronic HCV infection (HCV-RNApositive at baseline and after 3 and 6 months of followup) and 9 individuals with resolved HCV-infection (anti-HCV-positive/HCV-RNA-negative at baseline and after 3 and 6 months of follow-up) identified in the same screening served as controls for the T cell assays. Characteristics of control patients are shown in tables 2 and 3. Control subjects did not differ in age (20.0 ± 1.4 years and 20.7 ± 1.4 years for patients with persistent infection and individuals being HCV-RNA-negative already at baseline) to the seven study subjects (20.1 ± 0.8 years). All individuals studied were male. None of the subjects had received antiviral therapy for HCV-infection prior to incarceration or after imprisonment.

Serologic and virologic testing
Serum samples were tested for anti-HCV-antibodies using a third generation HCV-ELISA (AxSYM-HCV Version 3.0, Abbott Diagnostics, Wiesbaden, Germany). Antibody reactivity against single HCV antigens was evaluated using the immunoblot assay INNO-LIA HCV III Update (Innogenetics, Ghent, Belgium). HCV-RNA testing was performed by Real-Time PCR (sensitivity 10 2 copies/ml) as previously described [16,21]. For low viremic samples, we always performed a second RNA extraction and applied a second PCR with an alternative primer set and thus positive results in this study were based on at least two independent extractions and two independent PCRs. The HCV-RNA PCR has been validated and is used as a routine

T cell assays (Hannover Lab)
HCV-specific CD4+ T cell responses were investigated by interferon-gamma ELISPOT-assays and proliferation assays as described [22,23]. Antigens (HCV-core, HCV-NS3, HCV-helicase, HCV-NS4, HCV-NS5a, tetanus toxoid) were purchased from Mikrogen (Munich, Germany). Cyropreserved PBMC were used in all experiments. To avoid inter-assay variability, all serial samples follow-up of one individual were tested in the same assay. A stimulation index (SI) of >2.5 in the proliferation assay was considered positive (the mean SI in healthy controls + 3 standard deviations was <2.5 for all proteins tested). In the ELISPOT-assay, a positive result was considered if spot-forming units in the presence of antigen were at least double of spot forming units in the medium control and if at least five spots were detected per well.
Since patients recruited at the young offender institution in Hameln had a high risk of potentially being exposed to HCV due to previous drug consumption and HCV-specific T cell responses have frequently been shown in seronegative HCV-exposed individuals (including our own experience showing induction of HCV-specific T cell responses after needle stick injury), we did not choose individuals from the young offender institution as a "negative" control group but applied identical methods as for the other projects run in the laboratory. T cell responses were investigated in the laboratory in several different cohorts of persistently apparently unexposed HCV-RNA-negative individuals showing frequencies of responses for the different assays of <5% for CD4+ T cell responses [23][24][25].

ELISPOT assays (Vienna Lab)
Cryopreserved PBMC were also shipped to a second lab in Vienna headed by CK. GCP-validated ELISPOT assays are performed in this laboratory on a routine bases for measuring T cell responses in clinical vaccine trials [27]. Assays were performed exactly as previously described [28]. Since not enough PBMC were available to screen for T cell responses with overlapping peptides, a panel of synthetic peptides representing ~30 MHC class I and class II T cell epitopes derived from conserved regions of HCV was used (Table 5). These peptides were identified in a systematic screening with >400 overlapping peptides in PBMC from 65 HCV-recovered individuals (Klade et al., manuscript in preparation). The 30 selected peptides have a cumulative HLA coverage > 85%.

Cases
Out of 97 subjects identified to be anti-HCV-positive and/ or HCV-RNA-positive during the initial screening after incarceration, 90 individuals agreed to participate in the follow-up study and were studied after 3 and 6 months for HCV-RNA and ALT levels. At baseline 71 were HCV-RNApositive and 19 HCV-RNA-negative [16]. Seven of the HCV-RNA-positive patients cleared HCV-RNA spontaneously after 8-31 weeks of follow-up. Characteristics of these seven patients are shown in table 1 and the time course of HCV viremia, ALT levels and T cell responses is shown in figure 1. Only one subject (#80) had biochemical evidence of hepatitis with ALT levels higher than two times upper the limit of normal. None of the individuals reported significant symptoms associated with hepatitis or developed jaundice. Liver function as determined by albumin levels and prothrombine time remained normal in all subjects at any time. Bilirubin levels were minimally elevated once in patient #1 (1.23 mg/dl; normal value 1.1 mg/dl), and at three time points in patient #5 with levels between 1.5 and 2.3 mg/dl. Mean peak HCV-RNA levels were lower in the seven cases with spontaneous HCV clearance than in 62 HCV viraemic patients who did not

Humoral Immunity
Anti-HCV antibodies as determined by a standard 3 rd generation ELISA-assay remained negative throughout follow-up in 4 of the 7 seven cases (Table 1). We also investigated reactivity against single HCV proteins by the INNO-LIA HCV III assay. A faint band against the C2 protein was found once in one of the four anti-HCV negative patients while all anti-HCV-positive patients tested positive for at least four HCV-antigens at all time points (Table  1).

HCV-specific CD4+ T cell responses
Weak transient HCV-specific CD4+ T cell responses were detected against at least one HCV protein either in the interferon-gamma ELISPOT assay or in the proliferation assay in five of the seven patients (figures 1 and 2). None of the 4 patients who remained negative in the anti-HCV ELISA (#1, #5, #13, and #36) had a CD4+ response that was detectable at more than one time-point ( figure 1). Moreover, in all but one case (patient #36, first timepoint investigated) only one protein tested positive in the anti-HCV-negative patients.
Anti-HCV antibodies were already present at baseline in patients #10 and #27 (figure 2). No HCV-specific CD4+ T cell responses were found before or after HCV clearance in these two patients. Subject #80 was the only individual with biochemical evidence of hepatitis with about ten times elevated ALT levels. A robust response against HCVhelicase was found in the proliferation assay at baseline and after clearance of HCV-RNA (SI values of 4.1 and 3.6, respectively). At the second time point, we found also a significant response against NS4 in the proliferation assay as well as a helicase-specific interferon gamma response in the ELISPOT assay detectable (figure 2).
Overall samples from 19 time points were investigated in the 7 patients with spontaneous HCV clearance (2-4 time points per patient as shown in figure 1). At least one HCV protein was reactive in 4 (22%) of the ELISPOT-assays and 3 (16%) of the proliferation assays (table 6). In 7 cases (37%) any of the two CD4 assays yielded at least one positive result. Comparing these results with ELISPOT assays from control subjects from the same cohort with chronic hepatitis C did not show any significant differences in frequency, breadth and strength of CD4+ T cell responses. Interferon gamma-ELISPOT responses were found in 6/24 assays (25%) performed in 13 subjects with chronic HCV viremia. The mean number proteins recognized per assay did also not differ between the two groups (table 6). Five of the 13 chronically infected patients (38%) mounted at least one positive interferon gamma response. The absolute number of spot forming units in the ELISPOT-assay in positive cases was not different in patients who had cleared HCV and in chronic hepatitis C patients (mean SFU 16.8 ± 1.1/10 6 PBMC for patients with clearance and 19.4 ± 4.8/10 6 PBMC for chronic HCV patients). HCV long-term-recovered control patients (anti-HCV positive and HCV-RNA-negative already at baseline) showed ELIS-POT responses in 5/12 assays and proliferative responses in 3/9 assays performed with samples from 9 individuals also with a similar strength in responses (table 6).

IFN-gamma responses against 30 class I and II epitopes (Vienna lab)
Since only a limited number of PBMC was available according to the study protocol approved the institutional ethics committee, it was not possible to apply a full breath screening for T cell responses with overlapping peptides. Thus, we used a panel of 30 synthetic peptides represent- Biochemical, virological and immunological course of acute hepatitis C in asymptomatic patients with spontaneous clearance: patients with spontaneous clearance in the absence of anti-HCV Figure 1 Biochemical, virological and immunological course of acute hepatitis C in asymptomatic patients with spontaneous clearance: patients with spontaneous clearance in the absence of anti-HCV. HCV-RNA levels, ALT levels and T cell responses are shown. Coloured bars show the sum of SI values of the proliferation assay and the sum of specific interferon-gamma spots (spots in the presence of antigen -spots in the medium control) against 5 individual HCV proteins (HCV-core, HCV-NS3, HCV-helicase, HCV-NS4, HCV-NS5a). In addition, the number of individual proteins tested positive at the respective time point is shown for each assay. Interferon gamma ELISPOT responses were also tested against a panel of synthetic peptides representing class I and class II T cell epitopes derived from conserved regions of HCV with a cumulative HLA coverage > 85%. These assays were performed in an independent second laboratory ("Pep-Screen ELISPOT"; Vienna lab). The number of peptides tested positive in this assay is also given.
ing ~30 class I and class II T cell epitopes derived from conserved regions of HCV with a cumulative HLA coverage > 85%. These peptides were tested in an independent second laboratory. Also in this T cell readout most patients tested negative at all time points investigated. Patient #1 tested positive for one HCV-class II peptide at the first time-point investigated and was negative for all peptides on subsequent time points. Patients #5, #10, #13, #27, and #80 tested negative for all peptides at all time-points investigated. Only subject #36 mounted a robust multispecific response against at least 5 peptides after clearance of HCV (figure 3) which were not HLA-A2restricted and thus no tetramer assays could be performed in this study for this individual.
Biochemical, virological and immunological course of acute hepatitis C in asymptomatic patients with spontaneous clearance: patients who developed anti-HCV antibodies Figure 2 Biochemical, virological and immunological course of acute hepatitis C in asymptomatic patients with spontaneous clearance: patients who developed anti-HCV antibodies.

Discussion
We here report seven cases of spontaneous clearance of hepatitis C virus identified through a systematic screening of young male subjects with a high proportion of iv-drug addicts. We believe that the cases are of special interest for several reasons: (i) Four individuals never developed anti-HCV antibodies in a standard 3 rd generation ELISA-assay and also tested negative in an anti-HCV immunoblot assay. Thus, without screening for HCV-RNA these subjects would never have been recognized as being infected with HCV. Similar cases of HCV clearance in the absence of anti-HCV seronconversion have recently been described in four Austral-ian IV-drug addicts [15]. In line with reports on HCVspecific T cell responses in seronegative HCV-exposed persons [17][18][19][20], it is very likely that the true rate of individuals in the general population who had contact with HCV is underestimated, in particular in high risk persons such as iv-drug addicts and maybe also in medical health professionals. Our systematic unbiased screening approach of all new incoming prisoners in a young offender institution [16] supports the conclusion that the percentage of patients who are able to clear HCV viremia without developing anti-HCV antibodies might be much higher than previously thought since the anti-HCV ELISA became positive in only three out seven patients.
(ii) HCV was cleared spontaneously even though the subjects were asymptomatic and six out of seven had no biochemical evidence of hepatitis. On the first view, these findings somewhat contradict with several previous reports suggesting that the likelihood to clear HCV spontaneously is the higher the more symptomatic a patient is and the higher liver transaminases are [29,30]. We have no information how many patients with similar characteristics of exposure have developed chronic infection in our setting since the subjects were identified in a cross-sectional screening approach. At least 65 initially HCV-RNA positive patients had persistent infection although we do not know when these subjects had been infected before incarceration [16]. Moreover, characteristics of our patients differ from most other studies on acute hepatitis C. Our patients were rather young with an age between 16 and 24 and the mode of infection was iv drug abuse in almost all cases. Peak HCV-viral load with levels between 3 × 10 2 and 5 × 10 4 copies/ml was 2-3 logs lower than in patient cohorts on acute hepatitis C that were enrolled in treatment trials [26,31,32]. Nevertheless, for the management of acute hepatitis C patients our study supports the concept to investigate the kinetic of HCV-viremia for 2-3 weeks before an antivirial therapy is considered [33] -at least for individuals with a viremia of < 5 × 10 5 IU/ml. If  HCV-RNA declines, there might be a high chance for spontaneous clearance even in asymptomatic patients.
(iii) HCV-specific CD4+ T cell responses were rather weak or even undetectable and all subjects cleared HCV anyway. Clearance of HCV has been associated with strong and multispecific cellular immune responses in chimpanzees and humans [2]. Our findings are not completely in line with this dogma but are supported by at least one report on chimpanzees where also no correlation of HCVspecific adaptive immunity with HCV clearance was evident [14]. In our study, HCV-specific CD4+ responses were completely undetectable in 2 subjects and rather weak in the remaining 5 individuals despite clearance of HCV. These responses did not differ in strength and specificity from chronically infected control subjects. Of note, identical methods were applied and even the same proteins were used for the in vitro stimulation of T cells in the different assays as in our own previous studies [23][24][25]34,35] and also as in studies from other laboraties. Moreover, the low frequency and strength of T cells in our assays could not be explained by a mismatch of HCV proteins used for stimulation in the in vitro assays with the HCV genotype infecting the patients since all subjects had been infected with HCV genotype 1. In contrast, HCV-specific CD8+ T cell cells were found by tetramer staining in three of the four patients, although the frequencies of tetramer-positive cells were rather low in 2 of them as they did not exceed 0.2% of CD8+ T cells which was much lower as compared to our own experience on symptomatic acute hepatitis C patients [25] and also as compared to other reports on human subjects who cleared HCV spontaneously [8][9][10]12]. Nevertheless, the finding that three out of four HLA-A2 positive patients displayed HCVspecific CD8+ suggests that indeed T cell responses have been primed and that CD8+ T cells may have contributed to clearance of HCV. Importantly, the overall low frequency of HCV-specific T cells in the subjects investigated here was confirmed by T cell testing in a second, independent laboratory using a GCP-validated ELISPOT assay used for investigation of PBMC in clinical trials [28].
What could be the reasons for the lack of an apparent association between HCV-specific CD4+ T cell responses and recovery from HCV? First, our data suggest that low levels of HCV viremia as found in our seven cases may be controlled by innate immune responses without a strong adaptive immunity. Data derived from chimpanzees indeed have demonstrated that early intrahepatic induction of interferon-response genes without significant recruitment of HCV-specific T cells leads to significant reduction of HCV viral load [5]. Moreover, also Post et al. described a rather low strength of T cell responses in their two subjects with HCV clearance in the absence of seroconversion [15]. However, since we had not enough PBMC available to apply an overlapping peptide approach in our patients we might have missed some CD4+ and very likely some CD8+ T cell responses although the second screening with a panel of 30 peptides covering >85% of HLA types should have picked most responses. Nevertheless, the overall strength of responses might have been underestimated in our study. However, the important message is that the strength and frequency of T cell responses in patients with spontaneous clearance was not different from chronically infected or long-term recovered patients.
Also for HIV infection it has been shown that an antigenic threshold for maintaining cytotoxic T cell responses may be required [36]. The required levels of antigenic stimulation may not have been reached in our patients but innate immune responses may have led suppression of HCV replication. We also do not know, if our patients had been exposed to HCV before. Four of the seven patients were IV drug addicts with a duration of abuse of 18-72 months before incarceration. Higher rates of HCV clearance in ivdrug addicts have been reported. However, previous contact to HCV and re-exposure should have led to an expansion of adaptive immune responses which were not detectable in our patients. The lower HCV-RNA levels in our patients may be also explained by the route of exposure since three out of four subjects who did not develop anti-HCV antibodies had reported sexual exposure while almost all chronically infected patients were IV drug addicts.
For five of the seven patients we do not know how long these individuals had been viremic prior to imprisonment and thus we may have missed peak HCV-RNA levels and also some adaptive immune responses occurring prior to imprisonment. However, the main message of this manuscript is that HCV was cleared in patients without the development of anti-HCV antibodies in 4 out of 7 individuals. Moreover, CD4+ T cell responses are usually maintained for some time after clearance if no spontaneous relapse occurs as shown already in 1999 by Gerlach et al. [37].
Homing of T cells to the liver has been described in chimpanzees during acute HCV infection [5]. It is not possible to perform liver biopsies in patients with acute hepatitis C since there is no clinical indication for this invasive procedure. Thus, we can not exclude the presence of stronger intrahepatic HCV-specific T cells in our patients.
One could question if subjects #1, #13 and #36 indeed had acute hepatitis C since HCV-RNA tested positive only once in the absence of anti-HCV and thus the HCV-PCR could have been false-positive. However, a positive HCV-PCR result was always double checked and investigated by both qualitative and quantitative PCR. Moreover, Patient #36 mounted a significant and broad HCV-specific cellular immune response ( figure 1A and 2) supporting contact to the hepatitis C virus. Patient #13 had a significant peak viremia of 7.000 copies/ml making a false positive PCR result due to contamination rather unlikely and this patient also showed HCV-specific CD8+ T cells by tetramer staining.
In conclusion, clearance of HCV-RNA may not necessarily correlate with the appearance of acquired immunity. This finding challenges our current understanding of the immunopathogenesis of HCV-infection and maybe an explanation for the rather low seroconversion rate after occupational exposure to HCV. The data should also be considered in the management of accidental findings of low levels HCV viremia in asymptomatic individuals which may have a higher chance of spontaneous clearance than previously thought.