If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Five to ten percent of all orthotopic liver transplantations (OLTs) for end-stage liver disease are related to hepatitis B virus (HBV) infection.1, 2 Recurrent HBV infection after OLT is a major complication, frequently leading to graft failure or patient death.3 Without prophylaxis, the risk of graft reinfection is about 80%, depending on HBV DNA levels at the time of OLT. The combination therapy with nucleos(t)ide analogs and HBV-specific immunoglobulins (HBIG) has been shown to successfully protect the allograft from HBV reinfection.1, 3–5
The HBV-specific cellular immune response appears to play a key role in the course and outcome of HBV infection as subjects with self-limited HBV infection generally have strong and multispecific cellular immune responses, which can be detected over decades following infection.6–9 In contrast, patients with chronic HBV infection (CHB) show generally weak and narrowly directed virus-specific cellular immunity that may be functionally impaired.10–13 The initiation of Lamivudine therapy in these subjects has been shown to at least transiently increase clusters of differentiation (CD)4 and CD8 T cell–mediated immune responses against the HBV envelope (HBV surface antigen [HBsAg]) and nucleocapsid antigens (HBV core antigen)14–16 Comparably little is known about the role of the virus-specific T-cell responses in the OLT setting, although a deeper understanding of host immunity to HBV in OLT could facilitate the development of novel therapeutic immune interventions.17–19 The present study was thus designed to investigate the HBV-specific cellular immune response in immune-suppressed patients (OLT recipients) compared to patients with CHB (viremic and nonviremic) and self-limited HBV infection. The emerging data may help to optimize HBIG therapy or to identify subjects in whom HBIG treatment could be discontinued.
CHB, chronic hepatitis B infection; EBV, Epstein-Barr virus; ELISpot, enzyme-linked immunospot; HBeAg, hepatitis B e antigen; HBIG, anti-HBs immunoglobulins; HBsAg, hepatitis B surface antigen; HBV, hepatitis B virus; HLA, human leukocyte antigen; IFN, interferon; OLP, overlapping peptides; OLT, orthotopic liver transplantation; PBMC, peripheral blood mononuclear cell; SFC, spot forming cell; TNF, tumor necrosis factor.
PATIENTS AND METHODS
A total of 15 patients with HBV-related OLT were included in this study and compared to 27 patients with chronic and 24 patients with self-limited HBV infection (Table 1). The 15 OLT recipients received transplantation because of decompensated cirrhosis (n = 11) or hepatocellular carcinoma (n = 4). At transplantation, 11 patients had undetectable HBV-DNA (for 4 patients data were not available). Samples were obtained between 12 and 142 months after transplantation from patients without signs of HBV recurrence (median: 42 months post-OLT). All transplant recipients received 10,000 IU HBIG (Hepatect; Biotest, Italy) during the an-hepatic phase of OLT, 4800 IU weekly for the first month post-OLT, and then fixed scheduled intramuscular HBIG administration to maintain anti-HBs titers above 200 IU/mL. They were treated with a calcineurin inhibitor-based immune suppression combined with steroids, which were tapered down within 3 months after transplantation. All OLT subjects were without steroid treatment for at least 9 months (range: 9-90 months, median: 40 months) at the time of study. The maintenance immune suppression treatment was either cyclosporin (n = 10) or FK-506 (n = 5). A total of 3 subjects additionally received mycophenolate mofetil. None of the patients had an episode of rejection requiring anti-CD3 therapy.
Abbreviations: anti-HBc, hepatitis B core antigen antibody; anti-HBs, hepatitis B surface antigen antibody; CHB, chronic hepatitis B infection; HBeAg, hepatitis B e antigen; HBsAg, hepatitis B surface antigen; HBV, hepatitis B virus; IgG, immunoglobulin G; neg, negative; OLT, orthotopic liver transplantation; pos, positive.
The 15 OLT subjects were treated with cyclosporin (n = 10) or FK-506 (n = 5). None of the subjects required anti-CD3 therapy post-OLT. OLT subjects were tested at least 9 months after stopping steroid treatment (range: 9-90 months, median: 40 months).
All CHBs were HBsAg-positive, HBeAg-negative (except 3), and qualitative HBV-DNA–positive. The viral loads in the CHB group with detectable viremia ranged from 850 to 4 × 108 IU/mL (median: 79,341 IU/mL).
The group of subjects with self-limited HBV infection consisted of individuals with consistently positive IgG anti-HBc and anti-HBs for at least 1 year at the time of study, in line with chronic, self-limited infection and not reflecting acute infection subjects.
12 neg, 2 pos
12 neg, 1 pos
HBV DNA (range: IU/mL)
Patients with CHB were all HBsAg-positive, all but 3 were HBV e antigen-negative and qualitative HBV-DNA-positive. CHB patients were divided in 2 groups; 1 with undetectable HBV-DNA (nonviremic CHB) and another with detectable HBV viremia. The viral loads in the latter group ranged from >850 to 4 × 108 IU/mL with a median viral load of 79,341 IU/mL). All subjects with nonviremic CHB were treated with either Lamivudine (100 mg/day) or Adefovir (10 mg/day) at the time of study. A group of HBV-infected patients, followed clinically for nonviral liver diseases (nonalcoholic steatohepatitis, alcoholic and metabolic liver disease) was included as a comparison group with self-limited HBV infection. These subjects were at least 1 year after resolution of acute HBV infection and tested consistently positive for immunoglobulin G anti-hepatitis B core antigen and anti-hepatitis surface antigen antibodies.
All 66 subjects were negative for hepatitis delta coinfection, human immunodeficiency virus and hepatitis C virus and were recruited at the University Hospital in Bologna, Italy, and provided written informed consent prior to enrollment in this study. The Institutional Review Boards of the Ospedale Sant'Orsola and Massachusetts General Hospital approved the study.
Lymphocyte Isolation and Enzyme-Linked Immunospot (ELISpot) Assay
Peripheral blood mononuclear cells (PBMC) were separated from whole blood as described.20 Interferon gamma (IFNγ) ELISpot assays were performed using 200,000 PBMC/well for the detection of HBV-specific responses in the ex vivo assays and 100,000 cells/well for the assessment of HBV responses after peptide-specific in vitro expansion.20 Single overlapping peptides (OLP) were added individually to each well at a final concentration of 4 μg/mL. No peptide was added to 4 wells, which served as negative control wells. Phytohemagglutinin (Remel, Kent, UK) at a concentration of 1.8 μg/mL was added to 1 well as a positive control. Epstein-Barr virus (EBV) peptides were added to 2 wells, as pools of either lytic or latent antigen-derived cytotoxic T lymphocytes epitopes.20 After overnight incubation at 37°C, 5% CO2, ELISpot plates were developed and the number of spots determined using an AID ELISpot Reader (Autoimmun Diagnostika, Germany). Results are expressed as spot forming cells (SFC) per million cells and responses were considered positive when exceeding all of the following criteria: 1) a minimum of 5 spots per well; 2) mean of negative wells plus 3 times the standard deviation of the negative wells; and 3) 3 times the mean of the negative wells.21 This stringent cutoff did not produce any positive responses to nucleocapsid in HBV vaccine recipients.
A set of 77 OLP was synthesized covering the HBV nucleocapsid protein (183 amino acids including the precore and core antigen, 24 OLP) and the envelope protein sequences (389 amino acids, including the large S, preS1 and preS2 antigen, 53 OLP). Peptides were adapted 18-mers overlapping by 10 amino acids22, 23 and were based on the reported HBV genotype D reference strain ayw.24 This reference strain was used for the peptide synthesis since the HBV genotype D is predominant in Italy.25 This was confirmed in 6 out of the 14 viremic CHB patients, all showing genotype D infection (data not shown). The OLP length and overlap were selected to maximize the chance to detect CD4 and CD8 T cell–mediated responses as indicated by our earlier studies comparing different OLP designs.23, 26–28
For in vitro ELISpot assays, each peptide was tested individually in one well in a 96-well ELISpot plate (Millipore, Bedford, MA). In addition, 92 previously defined EBV-derived CD8 T-cell epitopes were included in the ex vivo ELISpot assays. The human leukocyte antigen (HLA) restriction of these EBV epitopes has been described previously and shown to cover all common HLA alleles.20, 29 Peptides were synthesized at the Peptide Synthesis Facility at Massachusetts General Hospital using Fmoc-chemistry.
Intracellular Cytokine Staining
Cells were stimulated as previously described30 with 4 μg/mL of peptide, anti-CD28 and anti-CD49d monoclonal antibodies (each 1 μg/mL) and incubated for 1 hour at 37°C, 5% CO2. A total of 10 μg/mL Brefeldin A (Sigma-Aldrich, St. Louis, MO) and 6 μL Monensin (Golgi-Stop; Beckton-Dickinson Bioscience Franklin Lakes, NJ) were then added for an additional 6 hours before placing the cells at 4°C overnight. One aliquot of cells serving as positive control was stimulated with phytohemagglutinin (1.8 μg/mL); negative control cells were stimulated with anti-CD28/CD49d in the absence of antigen. Cells were stained with anti-CD3 (Pacific blue; BD) anti-CD4 (APC-Cy5.5; Caltag) and anti-CD8 (APC-Cy7; BD) for 30 minutes at 4°C. After washing twice, the cells were fixed and permeabilized (Fix/Perm; Caltag) and then stained with anti-IFNγ (fluorescein isothiocyanate; BD) and anti-tumor necrosis factor (TNF)α (phycoerythrin; BD). Multicolor flow cytometry was performed on a LSR-II instrument (BD). A minimum of 200,000 events were acquired and analyzed using FlowJo software (Tree Star, Inc., Ashland, OR). Responses were considered positive if above a frequency of 0.04% and at least 3 times the frequency of IFNγ-producing cells detected in the unstimulated negative control.
Antigen-Specific T-Cell Expansion
To establish in vitro expanded T-cell lines, cryopreserved PBMC (5-10 × 106 PBMC) were split in 2 aliquots; 1 pulsed with peptide pools containing either HBV nucleocapsid- or envelope-derived OLP (4 μg/mL final concentration per peptide) for 90 minutes, washed, and added to the second aliquot and then cultured together with 10 × 106 irradiated, allogeneic feeder cells. Recombinant interleukin-2 (50 IU/mL) was added on day 3 and twice a week thereafter. After 14 days of culture, cells were washed 3 times in R10 and starved overnight in interleukin-2 free medium before using them in ELISpot and/or intracellular cytokine staining assays.
HBV Serology and Viral Load
HBsAg titer and HBV serological markers (hepatitis B surface antigen antibody, hepatitis B core antigen, HBeAg, hepatitis B e antigen antibody) were measured by chemiluminescent microparticle immunoassay (CMIA; Abbott Laboratories, Rome, Italy). Serum HBV-DNA was quantitatively measured by real-time polymerase chain reaction assay (Affigene HBV Trender; Alfa Wassermann, Bologna, Italy) with a lower detection limit of 850 IU/mL.
Statistical analyses were performed using GraphPad Prism (GraphPad, San Diego, CA). Results are presented as medians. Mann-Whitney test, Kruskal-Wallis test, and Wilcoxon matched pairs tests were used as indicated.
Weak HBV-Specific Immune Responses in Direct Ex Vivo Assays
In order to detect cellular immune responses directly ex vivo, IFNγ-ELISpot assays using thawed PBMC were performed in 43 patients (OLT, n = 11; CHB, n = 21; self-limited HBV, n = 11). Among the 11 transplant recipients, at least 1 OLP response was detected in 4 subjects (36%), similar to the response rates seen in the CHB (9/21; 43%) and the self-limited HBV infection groups (5/11, 45%; Fig. 1A). The median magnitude of detectable HBV-specific immune responses in the 3 groups was also comparable (Fig. 1B). To test whether transplant recipients mounted generally weaker cellular immunity than the comparison groups, EBV-specific T-cell responses were assessed using a pool of previously defined cytotoxic T lymphocytes epitopes.31 Of the 11 OLT patients tested, 10 (91%) showed EBV-specific immune responses, compared to 79% of the CHB group (15/19 tested) and 100% of the patients with self-limited HBV infection (4/4 tested; Fig. 1C; P = 0.1). While the magnitude of the EBV-specific immunity was overall significantly stronger than the detected HBV-specific responses (Fig. 1B and C; P = 0.01), it was not significantly different between the three groups (P = 0.36), indicating that transplant recipients maintain strong cellular immune responses against other viral antigens.
Fewer but Stronger HBV-Specific Immune Responses in Transplant Recipients After In Vitro T-Cell Expansion
To compare the present analyses to published data employing in vitro expanded cells and to determine whether broader HBV-specific immune responses could be detected upon antigen-specific in vitro expansion, the study was expanded to a total of 66 individuals (OLT, n = 15; CHB, n = 27; self-limited HBV, n = 24) for whom short-term T-cell lines were established. As expected, the breadth of responses was greater after in vitro expansion compared to direct ex vivo testing; with the highest frequency of HBV-specific immune responses seen in subjects with self-limited HBV infection compared to OLT or CHB individuals (Fig. 2A). These data are in line with previously reported results showing that the majority of patients who spontaneously recover from acute HBV infection maintain broad virus-specific T-cell response.8, 9 Furthermore, the data demonstrate that OLT recipients mounted virus-specific immune responses of a comparable breadth as chronically infected individuals (P = 0.36).
When the total virus-specific magnitudes of HBV-specific immune responses were compared between the 3 groups, the OLT recipients surprisingly showed stronger responses than subjects with CHB or individuals with self-limited HBV infection (OLT, 405 SFC/106 PBMC; CHB, 175 SFC/106 PBMC; self-limiting, 165 SFC/106 PBMC; Fig. 2B). While these differences did not reach statistical significance (P = 0.17), stratifying the CHB subjects into individuals with detectable (viremic) and undetectable (nonviremic) HBV viral loads, viremic patients showed significantly weaker responses compared to the other groups (Fig. 2C). Thus, while OLT subjects showed the fewest responses among the 3 groups, the magnitude of their responses was significantly higher than in viremic CHB (P = 0.005) and similar to that of subjects with low viremia.
To address whether the focus of the HBV-specific immune response against the nucleocapsid and envelope antigens were different in the different groups of subjects, protein-specific responses were analyzed separately. OLT recipients mounted stronger responses to the nucleocapsid antigen than against envelope and had the strongest nucleocapsid-specific responses of all four groups (Fig. 3A). In particular, their nucleocapsid-specific responses were significantly stronger than those seen in viremic CHB (P = 0.0007) and in subjects with self-limited infection (P = 0.015). No statistically significant differences among the 4 groups were noted for either the magnitude or number of their envelope-specific T-cell responses (Fig. 3B). Together, these data demonstrate that OLT recipients show overall fewer, yet stronger nucleocapsid-specific responses than subjects with self-limited infection and resemble in their response patterns individuals with nonviremic CHB.
HBV-Specific T-Cell Responses in OLT Recipients Are CD8+ T Cell–Mediated and Polyfunctional
To determine whether the detected T-cell responses were mediated by CD4 or CD8 T-cells and to assess their polyfunctionality, 25 positive ELISpot responses for which in vitro expanded cells were still available after the ELISpot screenings were confirmed by intracellular cytokine staining, combining staining for CD4 and CD8 with the detection of IFNγ and TNFα. The 15 responses that were confirmed in cell lines from either OLT recipients or CHB subjects were all mediated by CD8 T-cells and were mostly polyfunctional, as TNFα was simultaneously detected in 78% of all HBV-specific IFNγ-producing cells (data not shown). Testing cells from 4 individuals with self-limited HBV infection identified 2 exclusively CD4 T cell–mediated responses. No other HBV-specific CD4 T cell activity was detected in any other subject, in line with previous reports that have described CD4 T-cell responses to HBV in individuals with self-controlled infection.9, 32
For many viral infections, such as HBV or EBV, resolution of illness is likely mediated by cellular immune responses that provide long-term host control of viremia rather than viral eradication.33, 34 In HBV infection, it has been shown that virus-specific cellular immunity contributes to self-limited disease35 and that suppression of this immune control by chemotherapy or immune suppressive treatment can lead to viral reactivation.17–19 Thus, immune suppressive treatment following solid organ transplantation may place the individual at considerable risk of viral reactivation. Such reactivation could potentially be prevented or better controlled using rationally designed immune-based prophylactic or therapeutic interventions. The present study was designed to assess HBV-specific immunity in OLT subjects and to compare their focus, magnitude, and breadth to responses in individuals with chronic and self-limited HBV infection. The data show a low frequency and magnitude of the HBV-specific immune responses in ex vivo analyses, in line with earlier studies in which the characterization of ex vivo HBV-specific antiviral activity has been described as difficult and unyielding.6, 13, 35 However, the low magnitude of HBV-specific responses in the OLT subjects was likely not related to the immune suppressive treatment, as strong EBV-specific responses were detected in all 4 groups at comparable levels in direct ex vivo tested samples. Immune analyses after in vitro expansion of antigen-specific cells show that individuals with self-limited HBV infection mounted the broadest HBV-specific responses, in line with previous reports describing broad and multispecific immunity in this patient group.8, 35 Surprisingly, the magnitude of the HBV-specific immune response was higher in transplant recipients than in viremic CHB or self-limited infection and was mediated through significantly stronger nucleocapsid, but not envelope-specific, immune responses.
Studies in other viral infections, including HIV, EBV, and Kaposi sarcoma-associated herpesvirus (human herpesvirus-8) have indicated that the magnitude of responses may, at least partly, be driven by antigen availability and viral loads.21, 36–38 The increased focus on nucleocapsid-specific responses in OLT subjects may reflect a similar scenario as the reduced envelope-specific immune reactivity in these otherwise apparently immune-competent individuals could be a consequence of limited envelope antigen availability due to HBIG treatment.39, 40 While this would at least explain the relative shift toward nucleocapsid-specific responses in OLT, the response patterns in individuals with self-limited infection and undetectable viral loads suggests that the HBIG-mediated masking of envelope antigens may not be the sole reason for the diminished frequency and magnitude of these responses. One possible alternative explanation may be that the proliferative capacity of protein-specific responses may differ: Since HBV-specific responses are commonly tested after in vitro culture, the observed response patterns may to a considerable part be a consequence of the cells' ability to proliferate in response to antigenic stimulation in vitro, rather than reflect their true in vivo frequencies. From this point of view, OLT subjects may have conserved nucleocapsid-specific T cells with intact proliferative capacity, but do not effectively expand envelope-specific cells in vitro. Novel, highly sensitive approaches that do not depend on in vitro expansion of specific cells and the inclusion of markers like PD-1 and others, will likely be needed to discern these different possibilities.13, 41 In addition, care must be taken when comparing newer studies such as the present one and a recent report by Boni et al.13 with earlier reports that have been based on the use of defined, mostly HLA-A2 restricted epitopes rather than the more novel use of comprehensive sets of OLP. Potential differences in presentation of the targeted epitope, HLA bias and selection for highly immunogenic epitopes in studies using well-defined epitopes only may not be reflective of data generated in genetically (HLA) more diverse study populations and may result in considerably different interpretation of the available data.
Previous work using recombinant HBV nucleocapsid antigens has documented the presence of proliferative CD4 T-cell responses in OLT recipients with HBV recurrence and liver damage.42 In contrast to the present study, in which synthetic HBV OLP were used, this report measured proliferative T-cell responses using entire recombinant HBV antigens. In addition, our analyses included only patients who did not experience viral recurrence post-OLT and who may thus not have maintained these responses over time if they are indeed antigen-driven. The confirmation of T cell responses by flow cytometry provided little evidence for CD4-mediated immunity in OLT recipients and CHB patients. The absence of CD4 T cell responses in the CHB group is different from a recent study from Boni et al.,13 were CD4 responses were detected in in vitro expanded cells. Although methodological differences (cutoff for positive responses, antigen dose) may explain some of these differences, it is important to note that the latter study was performed in patients with chronic infection with fluctuating HBV-DNA and alanine aminotransferase levels, which is different from our analyses that included only subjects with stable CHB. Thus, more detailed analyses, including comparable patient cohorts and identical antigen sources will be needed to conclusively address a potential association between the proliferative capacity of HBV-specific CD4 and CD8 T-cells and viral control. Such expanded analyses also need to take into consideration that OLT are generally performed across HLA mismatches between organ donor and recipient and thus may induce donor-HLA restricted T-cell responses post-OLT.43, 44 Although the frequency of their detection and their contribution to viral control remains unclear, the assessment of these responses may further help to identify correlates of immune protection in the OLT setting.
We thank Cinzia Fortini and Lilia Testa for their helpful technical assistance. This study was sustained by members of the Bologna Liver Transplant Program and, in particular, Dr. Antonio Pinna, Dr. Gianluca Grazi, and Dr. Giorgio Ercolani.