Safety, reactogenicity, and immunogenicity of MTBVAC in infants: a phase 2a randomised, double-blind, dose-defining trial in a TB endemic setting

Abstract

Background: Safer and more effective tuberculosis (TB) vaccines than Bacille Calmette Guérin (BCG) are needed. We evaluated the safety, reactogenicity, and immunogenicity of three dose levels of the live-attenuated Mycobacterium tuberculosis (Mtb) vaccine, MTBVAC, compared to BCG, in South African infants.

Methods: Healthy, HIV-unexposed, BCG-naïve infants were randomised to receive a single intradermal dose of BCG (2.5 × 10⁵ CFU, n = 24) or MTBVAC (2.5 × 10⁴, 2.5 × 10⁵, or 2.5 × 10⁶ CFU, each n = 25). Safety endpoints were solicited systemic, solicited injection site, and unsolicited adverse events (AE), and serious AE (SAE). Immunogenicity was measured using interferon-γ release assay (IGRA) and whole blood intracellular cytokine staining assay. Follow-up was 12 months post-vaccination.

Findings: Ninety-nine infants were enrolled between 18 February 2019 and 08 March 2021. Seventy-eight infants experienced reactogenicity AE (all mild except one grade 2 erythema). Induration, swelling, and erythema were more frequent as MTBVAC dose increased. All reactogenicity events were less frequent in infants receiving MTBVAC 2.5 × 10⁵ CFU compared with BCG. Twelve infants (three BCG and nine MTBVAC recipients) experienced 14 vaccine-unrelated SAE, including one death due to bronchopneumonia (MTBVAC recipient). Eight infants were treated for unconfirmed pulmonary TB (four BCG and four MTBVAC 2.5 × 10⁴ CFU recipients); one BCG recipient was treated for unconfirmed TB meningitis. MTBVAC was immunogenic at all 3 doses, inducing predominantly Th1-cytokine-expressing CD4 T cells, which peaked at Day 56. The 2.5 × 10⁵ and 2.5 × 10⁶ CFU MTBVAC doses induced similar response magnitudes and were more immunogenic than BCG. Day 56 IGRA conversion was observed in 61 (87.4%) infants receiving any MTBVAC dose, but only 28 (42.4%) remained positive by Day 365.

Interpretation: MTBVAC appeared safe, well-tolerated, and immunogenic at doses between 2.5 × 10⁴ and 2.5 × 10⁶ CFU in South African infants. The 2.5 × 10⁵ CFU MTBVAC dose, being less reactogenic and more immunogenic than BCG, was selected for a multi-centre, phase 3 trial.

Funding: This trial was funded by the European and Developing Countries Clinical Trials Partnership (EDCTP).

Keywords: BCG; Infant; Mycobacterium tuberculosis; Tuberculosis; Vaccine.

Fig. 1

Consort diagram. CFU, colony forming units.

Fig. 2

Comparison of reactogenicity events at the injection site between cohorts and vaccination regimens. Graphs depict risk ratios with 95% confidence intervals, and P-values (Exact's test) which represent the compared proportions of infants who presented with the listed local adverse events between (a) BCG and the different MTBVAC doses, (b) the lowest MTBVAC dose (2.5 × 10⁴ CFU) and the other MTBVAC doses, and (c) the mid MTBVAC dose (2.5 × 10⁵ CFU) and the highest MTBVAC dose (2.5 × 10⁶ CFU). Shifts to the right indicate a greater risk of adverse reactions in infants who received the test vaccine dose (below each comparator) compared to infants who received the comparator vaccine (in bold, above). Number of infants per group are presented in Tables 3 and 4.

Fig. 3

Kinetics of cytokine-expressing MTBVAC-specific T cell responses in newborns vaccinated with MTBVAC or BCG. Frequencies of MTBVAC-specific CD4 T cells measured after MTBVAC stimulation by whole blood intracellular cytokine staining assay in MTBVAC or BCG vaccinees. (a) Individual longitudinal trajectories of CD4 T cell responses, categorized by treatment arm and dose at the indicated time points. Each line represents one participant. The top panel shows frequencies of CD4 T cell responses expressing any combination of IFN-γ IL-2, TNF, IL-17 A, or IL-22 (Any cytokine-expressing CD4 T cells) and the bottom panel shows polyfunctional IFN-γ+IL-2+TNF+CD4 T cells. (b) Median longitudinal trajectories of any cytokine-expressing CD4 T cells (top) and polyfunctional Th1 T cells (bottom) are shown. Error bars denote IQRs. (c) Frequencies of CD4 T cell responses expressing any combination of IFN-γ, IL-2, TNF, IL-17 A, or IL-22 (top row) and of polyfunctional IFN-γ+IL-2+TNF+CD4 T cells (bottom row) in newborns. Horizontal lines represent medians, boxes the IQR, and whiskers denote 1.5 times the IQR from Q1 and Q3; dots show outliers beyond the whiskers. P-values were obtained using a linear mixed model, and adjusted using Holm's multiplicity method. Adjustments were made within each visit. (d) Median longitudinal trajectories of IL-17 or IL-22 expressing MTBVAC-specific CD4 T cells, or any cytokine-expressing MTBVAC-specific CD8 T cells (right). Numbers of infants included in the immunogenicity dataset are presented in the Consort diagram in Fig. 1.

Fig. 4

IFN-γ responses measured by QFT in newborns vaccinated with MTBVAC or BCG. IFN-γ concentrations (IU/mL) measured by QFT-TB Gold Plus assay, categorised by timepoint post-vaccination. The higher IFN-γ concentration between TB1 and TB2, after subtraction of the Nil condition, is shown. The dotted horizontal lines indicate the positivity cut-off (0.35 IU/mL) of the assay. Horizontal lines represent medians, boxes the IQR, and whiskers the range. The percent and number of QFN-positive participants at each time point for each cohort are indicated below each graph. Orange triangles represent QFT results of infants diagnosed with TB at TB investigation or at the closest timepoints to TB investigation when it occurred between scheduled visits. Numbers of infants included in the immunogenicity dataset are presented in the Consort diagram in Fig. 1.