The candidate TB vaccine, MVA85A, induces highly durable Th1 responses

Abstract

Background: Vaccination against tuberculosis (TB) should provide long-term protective immunity against Mycobacterium tuberculosis (M.tb). The current TB vaccine, Bacille Calmette-Guerin (BCG), protects against disseminated childhood TB, but protection against lung TB in adolescents and adults is variable and mostly poor. One potential reason for the limited durability of protection may be waning of immunity through gradual attrition of BCG-induced T cells. We determined if a MVA85A viral-vector boost could enhance the durability of mycobacteria-specific T cell responses above those induced by BCG alone.

Methods: We describe a long-term follow-up study of persons previously vaccinated with MVA85A. We performed a medical history and clinical examination, a tuberculin skin test and measured vaccine-specific T cell responses in persons previously enrolled as adults, adolescents, children or infants into three different Phase II trials, between 2005 and 2011.

Results: Of 252 potential participants, 183 (72.6%) consented and completed the study visit. Vaccine-induced Ag85A-specific CD4+ T cell responses were remarkably persistent in healthy, HIV-uninfected adults, adolescents, children and infants, up to 6 years after MVA85A vaccination. Specific CD4+ T cells expressed surface markers consistent with either CD45RA-CCR7+ central memory or CD45RA-CCR7- effector memory T cells. Similarly durable Ag85A-specific CD4+ T cell responses were detected in HIV-infected persons who were on successful antiretroviral therapy when MVA85A was administered. By contrast, Ag85A-specific CD4+ T cell frequencies in untreated MVA85A-vaccinated HIV-infected persons were mostly undetectable 3-5 years after vaccination.

Conclusion: MVA85A induces remarkably durable T cell responses in immunocompetent persons. However, results from a recent phase IIb trial of MVA85A, conducted in infants from the same geographic area and study population, showed no vaccine efficacy, suggesting that these durable T cell responses do not enhance BCG-induced protection against TB in infants.

Figure 1. Longitudinal tracking of Ag85A-specific IFN-γ ELISPOT responses in subjects vaccinated with MVA85A 3 to 6 years ago.

Red data points denote measurements from individuals who became infected with M.tb since completion of the original clinical trial. M.tb infection was defined as conversion to a positive IFN-γ ELISpot assay response to ESAT-6/CFP-10. Frequencies of IFN-γ spot forming cells (SFC) detected before vaccination and at the last time point after MVA85A vaccination were compared using the Wilcoxon signed rank test. Blue p-values denote comparison of data from all individuals, while red p-values denote comparison of M.tb-uninfected (negative responders to ESAT-6/CFP-10) individuals only. (A–D) Longitudinal tracking of Ag85A-specific T cell responses in adults (A, n = 17), adolescents (B, n = 9), children (C, n = 15) and infants (D, n = 27), who received 5×10⁷ plaque forming units (pfu) of MVA85A. (E) Comparison of Ag85A-specific IFN-γ ELISPOT responses in subjects who were vaccinated more than 3 years ago with different doses of MVA85A as infants, or who received the placebo vaccine, Prevenar (n = 24 for Group 1, n = 27 for Group 2; n = 24 for Group 3 and n = 23 for the placebo Group). The overall effect was calculated using the Kruskal-Wallis test, while responses in each dose group were compared to the placebo group using the Mann-Whitney U test. Median and IQR IFN-γ ELISPOT response values are shown in Table S1.

Figure 2. Characterization of Ag85A-specific CD4 T cell Th1 cytokine expression patterns after long-term follow-up, 3–5 years after MVA85A vaccination.

(A) Representative gating of CD4 T cell expression of IFN-γ, IL-2, TNF-α and/or IL-17 in blood left unstimulated, or stimulated with BCG or Ag85A peptide pool, collected 1,187 days post-vaccination from a single participant who was 5 months old when originally vaccinated. (B–E) Frequencies of cytokine-expressing Ag85A or BCG-speciific CD4 T cells were measured by whole blood intracellular cytokine staining assay in adolescents (B), children (C) and infants (D and E). Cytokine expression patterns in adolescents and children who acquired M.tb-infection and those who remained uninfected were not different (data not shown); all individuals, irrespective of M.tb-infection status are shown in B (n = 9) and C (n = 16). Cytokine expression patterns in all infants (D and E, n = 15, 11, 12 and 11 for groups 1, 2, 3 and placebo respectively) were also not significantly different. For all box and whisker plots, horizontal lines represent medians, boxes represent the IQR and whiskers represent the range for each group of participants.

Figure 3. Memory phenotype of highly persistent Ag85A-specific CD4 T cells.

(A) Representative flow cytometry overlay plots, from a MVA85A-vaccinated child, showing CD45RA and CCR7 co-expression patterns by the entire CD4 T cell subset (blue) and Ag85A-specific CD4 T cells co-expressing IFN-γ, TNF-α and IL-2 (black dots) after in vitro stimulation with Ag85A peptides, live BCG or PHA. (B–D) Proportions of cytokine-expressing Ag85A-specific CD4 T cells expressing the indicated combination of CD45RA and CCR7, measured 3–5 years after MVA85A vaccination in adolescents (B, n = 9), children (C, n = 16) and infants (D, n = 15, 11 and 12 for groups 1, 2 and 3, respectively). (E) Proportions of cytokine-expressing BCG-specific CD4 T cells expressing the indicated combination of CD45RA and CCR7 in infants (n = 15, 11 and 12 for groups 1, 2 and 3, respectively). For all box and whisker plots, horizontal lines represent medians, boxes represent the IQR and whiskers represent the range for each group of participants.

Figure 4. Longitudinal tracking of Ag85A-specific IFN-γ ELISPOT responses in M.tb-infected and/or HIV-infected subjects, who were vaccinated with MVA85A 2 to 6 years ago as part of the TB011 trial .

(A) M.tb-infected, HIV-negative individuals (n = 11); (B) M.tb-uninfected, HIV-infected individuals (n = 8); (C) M.tb and HIV-co-infected individuals (n = 9); (D) HIV-infected individuals on successful ART (irrespective of M.tb-infection status, n = 9). Frequencies of IFN-γ spot forming cells (SFC) detected before vaccination and at the last time point after MVA85A vaccination were compared using the Wilcoxon signed rank test. Median and IQR IFN-γ ELISPOT response values are shown in Table S1.

Figure 5. Early MVA85A-induced Ag85A-specific T cell responses predict the level of persisting Ag85A-specific T cell responses.

(A) Representative Spearman correlation analysis of the frequencies of Ag85A-specific T cells detected by IFN-γ ELISpot assay on day 84 post-vaccination and more than 3 years post-vaccination in infants who received 5×10⁷ pfu of MVA85A. (B–D) Plots showing the Spearman r values obtained from correlation analyses between the frequencies of Ag85A-specific T cell responses detected after long-term follow-up (more than 1000 days post-vaccination), and those detected before MVA85A vaccination (day 0), or at 7, 28, 84 and 168 days MVA85A vaccination. Blue bars denote Spearman r values with a p-value above 0.05, orange bars denote p-values between 0.05 and 0.01 and red bars denote p-values below 0.01. (B) Infants who received 2.5×10⁷ pfu of MVA85A; (C) Infants who received 5×10⁷ pfu of MVA85A; (D) Infants who received 1×10⁸ pfu of MVA85A.