Antigen-specific T lymphocyte proliferation decreases over time in advanced chronic hepatitis C

Abstract

To evaluate T cell immunity in advanced liver disease, antigen-specific lymphoproliferative (LP) responses were prospectively studied in the context of the Hepatitis C Antiviral Long-term Treatment against Cirrhosis trial. Peripheral blood responses to hepatitis C virus (HCV), tetanus and Candida protein antigens were measured at baseline, month 12 (M12), M24, M36 and M48 in 186 patients randomized to either low-dose peginterferon-alfa-2a (PEG-IFN) only or observation. Liver histology was evaluated at baseline, M24 and M48. Patients with cirrhosis (Ishak 5-6) were less likely to have positive LP responses to HCV at baseline than patients with fibrosis (15%vs 29%, P = 0.03) and had lower levels of HCV c100 responses at baseline, M24 and M48 (P = 0.11, P = 0.05, P = 0.02, respectively). For 97 patients with complete longitudinal data, the frequency of positive LP responses to HCV, tetanus and Candida antigens declined over time (P < 0.003), and the slope of this decline was greater in the PEG-IFN treatment group than the observation group (P < 0.02). Lower levels of tetanus LP responses were associated with fibrosis progression and clinical outcomes (P = 0.009). Poorer CD4+ T cell proliferative function was associated with more advanced liver disease in chronic hepatitis C and may be further affected by long-term PEG-IFN treatment.

Trial registration: ClinicalTrials.gov NCT00006164.

© 2011 Blackwell Publishing Ltd.

Figures

Figure 1. Decreased Frequency of Positive Lymphoproliferative Responses During the HALT-C Trial According to Study Visit

Proliferative responses by freshly isolated PBMC to HCV protein antigens, tetanus toxoid and Candida albicans antigens were measured using a 3H-thymidine assay. The threshold for positivity was defined as a stimulation index (S.I.) of ≥4 for all antigens, and an HCV response was considered to be positive if an S.I. of ≥4 was achieved for one or more HCV antigens. All useable assay results available at each study visit were included, and the number of individual assay results at each study visit is shown below the x-axis. 95% confidence intervals are shown.

Figure 2. Lymphoproliferative Responses Decrease Over Time According to Treatment Status

Proliferative responses by PBMC to HCV protein antigens (A), tetanus toxoid (B) and Candida albicans (C) antigens were measured as described in Figure 1. Mean stimulation indices of all useable assay results at each study visit (including assays with S.I.≥ and <4.0) are indicated (D, E and F). Log10 Stimulation Indices (S.I.) were derived by subtracting the log10 mean cpm of unstimulated wells by the log10 mean cpm of stimulated wells. Error bars indicate 95% confidence intervals. Results were categorized according to whether PBMCs were obtained from patients randomized to long-term PEG-IFN treatment (black symbols/solid lines) or to observation (control, open symbols/dashed lines). This analysis was restricted to patients with complete data at all study visits (44 control patients and 53 treatment patients). Fewer PEG-IFN-treated patients had positive or lower level lymphoproliferative responses to HCV protein antigens (A, p=0.002; D, HCV c100, p=0.009), tetanus toxoid (B, p=0.0002; E, p=0.0008), and Candida albicans (C, p=0.02; F, p=0.09).

Figure 3. Lack or Lower Levels of Lymphoproliferation to HCV Antigens are Associated with Cirrhosis

Patients were categorized according to whether they had fibrosis (Ishak 3–4) or cirrhosis (Ishak 5–6), measured at each biopsy study visit. The percentage of patients with positive HCV-specific lymphoproliferative responses (S.I.≥4.0) to any HCV protein antigen (A) or corresponding mean Log10 S.I. to HCV c100 antigen (including assays with S.I.≥ and <4.0) at each study visit is shown (B). Both treated and untreated patients are included at Months 24 and 48. Numbers of patients in each group are indicated. P values are from separate analyses at each study visit. All error bars indicate 95% confidence intervals.

Figure 4. Decreased Levels of Lymphoproliferative Responses to Tetanus are Associated with Clinical and Fibrosis Outcomes

Mean Log10 S.I. (including assays with S.I.≥ and <4.0) and 95% confidence intervals of proliferative responses to HCV c100 (A, D), tetanus toxoid (B, E) and Candida albicans (C, F) antigens are shown. For clarity, results from patients randomized to observation (A, B and C) were separated from those randomized to long-term PEG-IFN treatment (D, E and F). Results from patients with outcomes (open symbols/dotted lines) and those with no outcomes (closed symbols/solid lines) are indicated. This analysis included 93, 79, 67, 65, and 56 control patients and 93, 74, 79, 73, and 70 treated patients at baseline M12, M24, M36, and M48, respectively. The confidence intervals are wide at month 12 because there were few patients in the outcome group at that time. Patients with clinical and fibrosis outcomes had lower lymphoproliferative responses to tetanus toxoid (B and E, p=0.009), but not HCV c100 (A and D, p=0.46) or Candida albicans (C and F, p=0.35).