Introduction
The aim of this function is to conduct detailed immunogenicity studies, even in the early stages of development, in order to guide decisions for example on vaccine composition, optimal doses or routes of vaccination. At this stage it is useful to provide a body of evidence to support inclusion of each individual component of the vaccine. Furthermore, data should be gathered on the combinations of the components, for instance to avoid antigen interference. Overreliance on a single model should be avoided, and multiple animal models can be combined within vitro methods. When testing immune responses in inbred mouse strains, for T-cell responses, it is important to test immune responses in more than one strain (e.g., C57BL/6, Balb/c, and C3H), as lack of immunogenicity may simply reflect the absence of MHC restriction in a given strain for the antigen being tested, as MHC restriction is more limited in inbred mouse strains than in humans. This understanding would be critical to choose appropriate mouse strains if subsequent efficacy studies are to be performed in mice.
After an antigen is found to be immunogenic, a range of doses should be evaluated in small animal models of immunogenicity and protection in order to understand the relationship between the two and inform dose selection in an appropriate animal model where it might not be feasible to test multiple dose levels. Furthermore, understanding of the immunology and protection associated with a range of doses and routes in preclinical studies may help guide dose selection in early clinical development and will support potency assay development at later stages. This can be further supported by in-depth mechanistic studies which aim to identify specific and quantifiable immune responses that correlate with preclinical protection. In the later pre-clinical stages, consideration should be given to the immunogenicity assays that are transferable to human trials. It is important tounderstand the limitations of preclinical data due to the absence of validated immune correlates of protection and insufficient clinical efficacy data to select which animal model best predicts vaccine efficacy in humans. Nevertheless, preclinical immunogenicity remains essential throughout the development pathway. Despite the lack of correlates of protection, a TB vaccine will likely have to induce T-cell responses. Especially CD4+ T-cell responses producing predominantly Th1-associated cytokines are required to control TB infection in humans as well as almost all preclinical models. Even though impairment of such responses is strongly correlated with susceptibility to tuberculosis, a protective vaccine immune response likely requires additional factors. These could include factors such as CD4+ T-cells producing IL-17, CD8+T-cells, non-classical T cells, including donor unrestricted T-cells (DURT´s),or B-cell and antibody responses. None of these factors are validated, or otherwise accepted, correlates of protection and as solitary markers they may therefore be of limited value for bringing a vaccine candidate forward. However, a robust understanding of induced immune responses in a variety of preclinical models will aid in refinement of a proposed mode-of-action and selection of appropriate biomarkers in clinical studies.
Vaccine technology specific considerations
Viralvectored vaccines: Pre-existing immunity due to cross-reactivity with related or wild type versions of the vector and/or the source of the heterologous antigen insert needs consideration. If observed, full characterisation of such anti-vector immunity is needed to determine mitigation strategies such as changing the interval between prime and boost. Data on the immune responses to the antigens of the vector are important in estimating the possibility of re-use of the vector virus in another vaccine. These same considerations can be applied for non-viral vectors, such as recombinant bacteria expressing mycobacterial antigens.
Live attenuated mycobacterial vaccines: These are composed of multiple antigens and therefore a ‘representative’ antigen or antigen sub-set e.g. needs to be selected with which to consistently monitor the immune response to the vaccine and differentiate from infected individuals.
Subunit vaccines: Because of the potentially large number of antigen-adjuvant combinations– considering adjuvant type, quantity (dose) and formulation, early immunogenicity studies, such as dose range studies, generally conducted in mice are important for optimisation. See the guidance for stages A and B for examples of assays and readouts.
RNA/DNA based vaccines: A major factor in the immunogenicity of mRNA vaccines is the achieved expression level and antigen localisation. Some of these aspects may be covered by platform-based data, such as luciferase-expressing mRNA to monitor biodistribution. Antigen expression should be evaluated and optimizedin in vitro models before performing elaborate in vivo studies.
TB vaccine target populationconsiderations
The TPP of the vaccine candidate, and in particular the target population, should be a factor in selecting the immunogenicity and efficacy models throughout the different stages of development. Although no single model has a consensus of support within the field, many tools and models are available that together can be leveraged to generate a robust preclinical package. At an early stage, neonatal vaccines should consider their relationship to the standard of care, BCG vaccination. For example, co-administration of BCG with a subunit vaccine candidate may lead to a drastically different immunogenicity testing strategy compared to a candidate aimed at replacing BCG, or one used as a BCG booster at a later age. However, for vaccines for adults and adolescents no reference vaccine is available, though BCG is typically used as it is the best available at this point. Vaccine candidates aimed to prevent disease in adults or adolescents could consider testing immunogenicity in animals pre-exposed to mycobacteria e.g. BCG vaccination, Mtb coinfection or post-treatment models.
- Evaluate immunogenicity
- Compare to benchmark, if applicable
- Evidence of relevant immunogenicity to antigens in at least in vitro or one animal model
- Above baseline and/or benchmark (if applicable) responses to antigens preferred
In Stage A, immunogenicity studies ideally demonstrate that a clear immune response to th etarget antigens can be induced in the same animal model used to demonstrate protection. Ideally this would be combined in one experiment. Mice are often used in early studies but may not be appropriate for all candidates (for example, it would not be appropriate to evaluate vaccines based on CD1-associated lipids, as mice lack most CD1 molecules). There is a large body of literature describing disease and vaccine-induced protection in the most commonly used strains (Balb/C, C57Bl/6 and CB6F1). (Orme and Ordway, 2016, Stylianou et al., 2018 and Aagaard et al., 2011). In vivo studies can be supplemented by in vitro methods, which can help to bridge rodent data to human-based systems. For instance, cytokine release from human PBMCs or celllines expressing particular targeted receptors can help to identify andoptimize adjuvant mode of action, which may act very different in rodent models. Cell-culture models expressing human HLA-variants can also be of great value, for instance by identifying peptides presented on different HLA-classes by e.g. mass spectrometry, after the vaccine is administered in vitro. Furthermore, such relevant antigen presentation can also be investigated by testing recognition by T-cell clones or cell-lines (e.g. Jurkat) transfected with constructs expressing human T-cell receptors.
In the absence of acorrelate of protection, the immune parameters measured should be relevant to immunity to Mtb and to the proposed mechanism of action of the vaccine. Due to the costs of Mtb protection studies, early-stage decisions e.g. on doses, formulations etc, could be based upon immunogenicity, such as the magnitude of antigen-specific T-cells. A response above baseline is considered a minimum requirement but, if similar to others in development, the candidate should show an obvious differentiating characteristic – qualitative or quantitative.
- Confirm and characterise immunogenicity
- Develop hypothesis for immunological mechanism of action; choose model study with appropriate outcomes, sample size, read-outs, data analysis plan, etc.
- Confirmed consistent immune response to antigens in at least one animal model
- Hypothesis for immunological mechanism of action defined; synopsis of study in advanced model available
Immunogenicity studies in Stage B should demonstrate that the candidate is immunogenic in the animal model chosen to confirm protection and non-clinical safety. Studies should aim to link or correlate specific immune responses to protection, if possible, which may include more detailed characterisation of the immune response. The immune response profile identified in Stages A and B should provide a proposed mechanism of action and can be used to assist the design of studies in the NHP (or other advanced) model in Stage C. This would include informing the final lead candidate(s), optimal immunisation protocol (e.g. the timing of prime and boost vaccinations) and identify the most appropriate assays to be performed and samples to be taken. Immunogenicity studies in Stage B might also include dose-finding in order to demonstrate that a relevant immune response is achieved with doses of vaccine that are safe and feasible (for example, for manufacture).
- Confirm immunogenicity against relevant benchmark
- Expand immunogenicity (Th1, Th17, γδ T cells etc.) to explore mechanism in same species used to demonstrate efficacy
- Immune response against relevant benchmark established
- Immune mechanisms and breadth of immune response explored
Taking advantage of the system validity of NHP and the vast range of specific and cross-reactiver eagents, NHP vaccine studies can be deployed for immunogenicity assessments in advanced preclinical stages. It can be opportunistic, however, to consider NHP studies not prior to but in parallel to early clinical phase testing. As in other animal species, it is imperative to define in NHP a primary assay for detecting the administration or the take of the vaccine per se. Depending on the ethical authority, a vaccine response may be required as a formal 'Go-No-go'-criterion before entering into infectiouschallenge. Anticipating the subsequent challenge with Mtb, however, it is advisable to extend the study plan in the vaccination phase with additional immune assays or sample collection for future retrospect analyses. Either driven by specific hypotheses of protective immunity or by unbiased (hypothesis-generating) omics approaches, NHP samples from the post vaccination phase can be expected to contribute to our understanding of vaccine-induced host immunity against TB. Especially if the candidate strategy proofs efficacious in the model, working back from protected phenotypes additional samples and assays shall add towards identifying correlates of protection and towards developing biomarkers, deployable either in future NHP studies (also adding to ethical refinement, reduction or replacement) or in support ofclinical vaccine evaluation. In addition to adaptive immune parameters, NHP are suitable also for innate response assessments, while innate immunity may be trained by whole cell vaccines (in line with the innate training potential of BCG in the clinic) or induced by adjuvant or vector formulation. Especially for subunit vaccination or putative revaccination strategies, macaque colonies that have received standard intradermal BCG immediately after birth are (being) established. Of note, and while alternative administration routes (e.g. mucosalor intravenous) have demonstrated superior efficacy when using standard BCG, such efficacious strategies may serve as positive controls in study designs for candidate vaccine evaluation in NHP. So far, such alternative delivery strategies have provided statistical correlates of antigen-specific Th1/Th17 function, antibodies and cytokine release in the airways as well as early systemic innate immune parameters. Mechanistic studies in NHP suggest that both CD4- and CD8-expressing lymphocyte subsets contribute to protective immunity inmacaques.