Development of an Autologous Human Dendritic Cell Vaccine against Mycobacterium tuberculosis in Patients with Extensively Drug-Resistant Tuberculosis

Abstract

Introduction: Extensively drug-resistant tuberculosis (XDR-TB), and resistance beyond XDR-TB (often untreatable), is an increasing public health concern globally. Drug resistance has outpaced the drug development pipeline. Therefore, alternative immunotherapeutic approaches are urgently needed. In this proof-of-concept study, and based on precedents in breast and prostate cancer, we aimed to develop an autologous therapeutic dendritic cell (DC) vaccine by evaluating two TB antigen-specific multi-peptide pools in combination with different adjuvants (such a cellular vaccine would require ex-vivo manipulation in a clean room and reinfusion back into the patient). Vaccine efficacy was evaluated using an in vitro mycobacterial containment model. Methods: DCs were derived from monocytes isolated from the peripheral blood of patients with XDR-TB (n=30) and participants with presumed latent TB infection (LTBI; n=15). DCs were matured with a differential combination of cytokines and pattern-recognition receptor agonists (maturation cocktail) together with different TB antigen combinations. The complete cocktail contained interferon-, interferon-, CD40L, IL-1 and TLR-3, TLR-7 and TLR-8 agonists, whilst in the limited cocktail the TLR agonists were absent. Two M.tb-specific multi peptide pools suited to GMP-grade vaccine manufacture were evaluated: (i) an immunodominant peptide pool (ESAT6 + CFP10 + Ag85B + TB10.4; referred to as ECAT) and (ii) a PE/PPE peptide pool. PPD and the lysate of a clinical M.tb strain (HN878), though not amenable to GMP-grade vaccine development, served as controls representing the entire antigenic repertoire of M.tb. Thus, there were four major comparator groups (unstimulated DCs, limited cocktail-only stimulated DCs, antigen-only stimulated DCs, and antigen + complete cocktail stimulated DCs). As the two latter groups were interrogated using the two multi-peptide pools and the two broad-spectrum antigen controls, a total of 10 different experimental groups were generated (see overview figure 3.2). DCs were assessed for the expression of key maturation markers using flow cytometry and the secretion of Th1-polarising cytokines by ELISA. The ability of DC-primed peripheral blood mononuclear cells (PBMCs) to restrict the growth of M.tb-infected monocyte derived-macrophages was evaluated using an in vitro validated mycobacterial containment assay. Results: In patients with XDR-TB, DCs matured with any M.tb-antigen + complete cocktail, compared to DCs matured with M.tb-antigen only, showed significantly higher upregulation of CD80, CD83, CD86, and CCR7 (p< 0.001 for all comparisons), and higher secreted levels of IL12p70 (0.67 versus 0.01 ng/mL per 106 cells; p< 0.001). A similar pattern was seen in the containment experiments: mycobacterial stasis within the XDR-TB group was significantly better with antigen + complete cocktail versus antigen alone (p≤0.0002 for PE/PPE and PPD), and the limited cocktail did not show this effect. Furthermore, PE/PPE + complete cocktail matured DCs achieved a higher magnitude of mycobacterial containment compared to ECAT + complete cocktail-matured DCs (50%, IQR:39-75, versus 46%, IQR: 15-62, p= 0.02). Using PPD and the HN878 lysate did not improve the containment effect. Furthermore, the improved containment effect of the PE/PPE + complete cocktail, versus ECAT + complete cocktail, was only seen in the XDR-TB and not in the LTBI group. Conclusion: In patients with XDR-TB, an effector response primed by PE/PPE antigen and complete cocktail-matured DCs was able to better restrict the growth of M.tb in vitro. These data indicate proof-of-concept feasibility to generate a DC-based immunotherapeutic intervention for therapeutically destitute patients with DR-TB. Further mechanistic studies and future phase 1 and 2 human clinical studies are warranted.