| |
The dynamics of T-lymphocyte responses during combination therapy for chronic hepatitis C virus infection
|
| |
| |
HEPATOLOGY Sept 2002;36:743-754.
The findings of this study are that T-cell and CD8 cell response to therapy for HCV does not appear associated with response to therapy.
"......We found, overall, no association between the magnitude or persistence of T-cell responses and sustained viral control and it is possible that the detected T-cell responses are not causally related to viral eradication during therapy, but rather reflect reduction in an excessive viral load that had led to T-cell exhaustion. Other possible mechanisms of enhancement include a direct effect of interferon-alfa on antigen-presenting cells, T cells, or immunomodulatory effects of ribavirin, although from this study it is not possible to further dissect these pathways. However, what is clear from this study is that HCV-specific T-cell responses have not been deleted or completely exhausted during chronic infection and this may be relevant to future therapeutic strategies....."
Abstract
Hepatitis C virus (HCV) readily sets up a persistent infection and is a major cause of liver disease worldwide. Interferon alfa and ribavirin therapy lead to sustained clearance of virus in 31% to 64% of patients with type 1 and non-type 1 genotypes, respectively. It is not clear to what extent these drugs act directly to reduce HCV replication, or indirectly via host immune responses, and what evoked immune responses are associated with clinical outcome. We have examined prospectively 15 patients with chronic HCV infection before, during, and after combination therapy. Quantitative assays for HCV antigen-specific CD4+ and CD8+ T-cell responses, and flow cytometric assays for analysis of the phenotype of T cells, in addition to viral sequencing of core protein, were performed throughout the treatment and follow-up period over 18 months. We found enhancement of proliferative T-cell responses during therapy. Proliferative responses are strikingly heterogeneous in terms of specificity, kinetics, and magnitude. Proliferative responses are often not associated with interferon- release. T-cell responses are rarely sustained irrespective of treatment outcome and this is not due to the evolution of new immune escape variants. T-cell responses tend to peak late in the course of treatment.
In conclusion, combination therapy for HCV has a transient effect on host virus-specific T cells in the blood. Induction of sustained T-cell responses may require additional immune modulation later in therapy.
Hepatitis C virus (HCV) is an important cause of liver disease worldwide, affecting over 170 million people.1 The virus persists in the majority of infected individuals. The mechanisms leading to resolution of primary infection involve, importantly, cellular immune responses. We have previously shown that large expansions of highly activated virus-specific CD8+ T-lymphocyte responses are detectable during acute infection with HCV,2 though functional and genetic studies3,4 highlight the important role of CD4+ T-lymphocyte responses. If these initial mechanisms fail, and chronic infection is established, CD4+ and CD8+ cellular immune responses are generally attenuated.5,6 The reasons for this lack of responsiveness are not understood, although potential mechanisms include T-cell exhaustion or deletion and epitope mutation.
Clinical trials7 have shown that alfa interferon given with ribavirin (combination therapy) results in sustained virologic eradication in 31% to 64% of patients, depending on viral genotype. The mechanism of action of these drugs, and particularly their effects on T-cell responses, are not understood. Both drugs, however, are thought to have immunomodulatory properties in addition to direct antiviral effects: alfa interferon enhances interferon- production of CD4+ T cells in vitro but with little effect on CD8+ T cells,8 and enhances the terminal maturation of professional antigen-presenting cells.9 Ribavirin is a nucleoside analogue, which reduces hepatic inflammation but with little effect on HCV-RNA levels.10 In addition, ribavirin induces a switch from a T-helper cytokine type-2 profile to a T-helper cytokine type-1 profile11 and suppresses interleukin 10 (IL-10) production in vivo.12
Understanding the mechanism of eradication in those patients who clear virus and the mechanism of persistence in those who fail to eradicate virus would potentially allow optimization of current treatment regimens and the development of new antiviral or immunomodulatory agents. If complete deletion or exhaustion of T-cell responses has not occurred, decreasing the HCV load with antiviral therapy may lead to recovery of T-cell responses as in chronic hepatitis B virus infection.13 Alternatively, if antiviral responses are primarily driven by antigenic stimulation, therapy might decrease the levels of specific T cells, as is the case in human immunodeficiency virus-1 infection.14 In this study we evaluated the effect of antiviral therapy on CD4+ and a defined number of CD8+ HCV-specific T-cell responses in a cohort of patients with chronic HCV infection, and correlated these changes with clinical outcome.
RESULTS
Patients attended for 113 of 122 (93%) protocol visits. Each patient was analyzed a mean of 8 times (range, 6-9 times). Four patients made no response to therapy and 3 patients made a transient response to therapy. For the purposes of analyses, these 7 patients have been classified as nonresponders. Eight patients had a sustained virologic response (HCV RNA undetectable by reverse-transcription PCR 6 months after the end of treatment). There was no statistically significant difference in the age, pretreatment viral load, alanine transaminase (ALT) level, or duration of therapy in responders and nonresponders. Two patients (D and F) had ALT levels within the normal range before treatment.
The kinetics of T-cell responses during therapy are strikingly heterogeneous
The HCV-specific antigens targeted, the kinetics, the magnitude, and the breadth of the observed responses during and after therapy were strikingly heterogeneous, although 3 different patterns of T-cell responses emerged: (1) marked enhancement of T-cell proliferative responses against multiple HCV-specific antigens during therapy (3 patients; Fig. 1A); (2) weak T-cell proliferative responses enhanced transiently during therapy (10 patients; Fig. 1B: 4 representative patients); and (3) absent T-cell proliferative responses during and after therapy (2 patients; Fig. 1C).
Proliferative T-cell responses are enhanced during therapy but are sustained infrequently
Analysis of the overall effect of treatment on T-cell responses is complicated by the heterogeneity observed. Nevertheless, we addressed the question of whether overall, therapy influenced T-cell responses and over what time course. Proliferative responses to HCV-specific antigens could be detected before treatment in 5 of 15 patients, all at a low level (mean stimulation index, 5.25, range, 4-8), consistent with previous studies showing lack of CD4 responsiveness in those patients with persistent viremia.4,6 Proliferative responses to test antigens were present in 13 of 15 patients during and/or after therapy, were frequently directed against multiple antigens, and were most commonly directed against core (Fig. 2). Further analysis showed that there was a significant increase in the magnitude of proliferative responses at weeks 12, 36, and the end of treatment time points, but not in the posttreatment time points, compared with pretreatment (Wilcoxon signed rank test, P = .003, .008, and .0001, respectively). The increase at week 24 was of borderline significance (P = .06; Fig. 3A). Although treatment clearly had a significant effect, responses were only sustained (responses detected more than 3 times to the same antigen during therapy and persisting after therapy) in 2 patients (patients A and O; Fig. 1 A).
T-cell responses were enhanced during therapy but did not predict treatment outcome
Although T-cell responses were enhanced during therapy in 13 of 15 patients, this did not predict the effect of treatment. Of the 8 patients with a sustained virologic response to therapy, 2 patients had strong multispecific responses that were sustained after therapy, 5 patients had weak and transient T-cell responses, and 1 patient had no T-cell responses detectable at any time point. Conversely, of the 7 patients who failed to clear virus with treatment, one developed strong but transient multispecific T-cell responses, 5 developed transient T-cell responses including detectable interferon- production by ELISpot to at least some antigens, and one patient had no detectable T-cell responses at any time point.
Overall, there was no difference between the responders and nonresponders in the magnitude of the peak proliferative or interferon- responses during therapy, and the magnitude of the pretreatment or posttreatment proliferative or interferon- responses. We observed that responses frequently emerged late in the course of therapy. The median peak proliferative responses occurred later (34 weeks) during the course of treatment in the responders compared with the nonresponder group (17 weeks). However, therapy was stopped at an earlier time point in treatment nonresponders (mean week 30) compared with treatment responders (mean, week 40) as dictated by clinical practice16 so that late responses in treatment nonresponders may have been missed.
Analysis of CD8+ T-cell responses
CD8+ T-lymphocyte responses were analyzed ex vivo by using a combination of major histocompatibility complex class I peptide tetramers and ELISpot assays in 8 HLA-A2-positive patients (Table 1 ). The number of epitopes studied was limited by the amount of blood ethically obtainable. However, to optimize the chances of detecting CD8+ responses, 4 HLA-A2 tetramers were used that have been shown previously to detect responses in both PCR positive and PCR negative individuals.2,5,22,24,25 Consistent with previous observations,26 responses detected in patients at the initiation of treatment were weak or absent as assessed by ELISpot and tetramer assay. No significant increases in CD8+ T cells were observed either by ELISpot or tetramer analysis. HCV-specific tetramer responses were only detectable in 2 of 8 individuals at any time point. Where a tetramer-positive population was consistently identified, only modest changes in frequency were observed with a slight increase over time, despite the enhancement of CD4+ responses (Fig. 4, patient D).
T-cell responses were enhanced in 13 of 15 patients during therapy, yet 7 of these failed to control virus. A possible reason for this is that escape mutations evolve during therapy within epitopes targeted by T cells, as in HIV infection.28 In 4 patients (B, E, H, and L) who failed therapy, 3 had T-cell responses to core peptides (Fig. 6, table insert bottom).
Discussion By Authors
HCV persistence is a major clinical problem, and current therapies are only partially effective. The importance of cellular immune responses in the control of primary infection is supported by data from several groups.2,3,5,6 It appears that in persistent infection, cellular immune responses in both CD4+ and CD8+ T-cell compartments are attenuated.5,6 Therapy commonly leads to rapid normalization of ALT level and clearance of viral RNA from blood, and is then followed by a period during which virus may or may not re-emerge.7 It is possible that at this point, cellular immune responses play some role in determining outcome, as in acute human immunodeficiency virus infection.29 We addressed the extent to which current therapy modulates cellular immune responses, in particular the dynamics of these responses, and their association with outcome.
A striking finding in this study was the effect of combination therapy in promoting virus-specific proliferative T-cell responses in some patients. These antiviral T-cell responses appear to have been suppressed during chronic HCV infection, yet clearly emerged during combination therapy. These induced responses were markedly heterogeneous in terms of dynamics and specificity but in the majority of patients T-cell responses were weak, tended to emerge and peak during therapy, and were rarely sustained after therapy.
We found, overall, no association between the magnitude or persistence of T-cell responses and sustained viral control and it is possible that the detected T-cell responses are not causally related to viral eradication during therapy, but rather reflect reduction in an excessive viral load that had led to T-cell exhaustion.30,31 Other possible mechanisms of enhancement include a direct effect of interferon-alfa on antigen-presenting cells,9 T cells,8 or immunomodulatory effects of ribavirin,12 although from this study it is not possible to further dissect these pathways. However, what is clear from this study is that HCV-specific T-cell responses have not been deleted or completely exhausted during chronic infection and this may be relevant to future therapeutic strategies (see later).
We were unable to show an enhancement of new responses, or augmentation of existing HCV-specific CD8+ T cells by using a defined number of CD8+ T-cell epitopes. This lack of enhancement was seen with both functional assays (interferon- ELISpot) and direct ex vivo staining by using tetramers. Consistent with this, activation (CD38, HLA-DR) and maturation (CD27, CD28, and perforin) phenotypic markers of the HCV tetramer-positive cells, when present, remained constant throughout therapy. It is simply possible that the limited number of epitopes studied were not influenced by therapy and that an analysis using a panel of peptides covering the whole HCV genome would have revealed new responses,32 or, alternatively, that transient responses as observed in acute disease were missed.2,33 Additionally, mouse models have clearly shown that functional CD4+ T cells are required for maintenance of antiviral CD8+ T-cell function34 and it is possible that the failure to stimulate interferon--producing CD4+ T cells during therapy (see later) contributes to the low numbers of HCV-specific CD8+ T cells that we detected. Other possible explanations include a differential effect of alfa interferon on CD4+ and CD8+ T cells, as has been found in vitro.8 Vertuani et al.35 have recently reported an increase in CD8+ T-cell responses to NS3 1073-1081 as a result of therapy with interferon alfa by using in vitro restimulation, a technique that will detect a small number of antigen-specific precursors. Our analysis, which focused on ex vivo responses, did not find such an association.
Interestingly, patients frequently developed proliferative T-cell responses without associated interferon- production (an example of this is patient C, Fig. 1 A). This was unexpected because one would predict that the ELISpot assay would prove more sensitive and therefore detect responses that were undetectable by proliferation assay. T cells that fail to secrete interferon- have previously been observed in both acute2 and chronic HCV infection,24,36 although the mechanism for this is not clear. Some CD4+ T-cell responses have been shown to switch to IL-10 rather than interferon- production during chronic HCV infection.37,38 It was not possible, because of limitations in cell numbers, to test for IL-10 by ELISpot in this study, but in light of the discrepancy between proliferation and interferon- ELISpot assays, this should be considered in future analyses. Although the ex vivo assays used represent the most sensitive currently available, a further limitation is the potential compartmentalization of T-cell responses within the liver. However, the clearance of viral antigen and normalization of ALT levels on therapy might be expected to diminish any redistribution effects resulting from hepatic inflammation.
We identified 3 patients who developed HCV-specific T-cell responses against core during therapy yet failed to clear virus. Viral sequencing showed that no new mutations arose within the core region during therapy. In one patient (patient B), autologous viral epitopes were found to differ significantly from the prototype peptides used in the analysis before the onset of therapy. However, ELISpot analysis, by using the autologous peptides, showed that these variants were also recognized, excluding an effect of original sin.39 These data strongly suggest that viral persistence in the face of enhanced host immune responses induced by treatment is not caused by the evolution of new immune escape variants, at least within core. This still leaves open the important question as to why virus persists in the face of demonstrable T-cell responses.
Interestingly, we observed that T-cell responses can take many months to emerge and peak after the onset of treatment, despite the fact that HCV-RNA levels become rapidly undetectable in the serum in the majority of patients. The reason for this is not clear but may be due to the fact that treatment has 2 effects on T cells: withdrawal of antigen may reduce antigen-dependent memory T-cell populations, but at the same time may rescue T-cell populations that were impaired by excessive antigen load. Exact dynamics may depend on the degree of T-cell antigen dependence. Whatever the origin, the fact that immunologic responses continue to evolve months into therapy may be relevant to the clinical observation that prolonged antiviral therapy is required, after initial viral clearance, to ensure sustained viral eradication.7 Because patients in general tend to lose T-cell responses at the end of treatment, an important therapeutic window where further augmentation, maintenance, or induction of T-cell responses might prove beneficial may exist. Current vaccine delivery systems that target cellular immune responses, such as DNA prime boost, might be ideal in such a situation (i.e., after viral load is controlled and before treatment is stopped). Interestingly, synergistic activity of ribavirin in combination with enhanced antiviral immune responses to prevent viral persistence has already been shown in vivo by using a murine model.
In summary, combination therapy for HCV has been shown to have a significant but temporary effect on cellular immune responses. Further studies are required to understand how these responses might be further harnessed to improve treatment outcome.
| |
| |
| |
|
|
|
