Preclinical testing comes at a relatively early stage in vaccine discovery, as the development of an immune response following antigen administration can only be tested in an organism. If adequate models are available, protection from (or attenuation of) viral infection is the primary milestone. In addition, the immunogenicity of the antigen is tested for its ability to sufficiently induce high antibody titres or a T cell response.

From animals to humans
Encouraging animal data are necessary to start clinical testing in humans. The progression of these trials is similar to any other drug, but the main difference is that vaccines need to be administered to healthy subjects, so the safety risk must be even lower than for a drug that is meant to cure a disease.

Phase I is particularly critical, as this is the first stage that assesses vaccine safety, together with reactogenicity. The vaccine is first administered to a very small number of healthy, naïve, and immunocompetent individuals to determine the safety of the administration. These studies are then extended to determine dose, schedule, and administration route in what are called Phase Ia studies. In Phase Ib, the same vaccine is administered to different population groups to assess the same parameters across a more representative sample of individuals, and to identify adverse reactions in particular groups (for instance, elderly or young children).

In the same studies, immunogenicity is also tested: does the formulation induce seroconversion or epitope-specific T cell responses?

In Phase II, a larger number of volunteers are recruited, generally hundreds or even thousands. The aim is to validate safety in a larger group, to better characterise magnitude and type of immune response, and to use this information to define the administration regime (dose, intervals, boost requirements). If correlates of protection are known, these studies assess whether the vaccine is able to induce them and how long they last, often by comparison with another vaccine or, when this is not available, with a placebo group.

When the pathogen that the vaccine is meant to threat causes infection that can be easily resolved in healthy individuals, human challenges can be carried out. In these studies, healthy volunteers are challenged with the virus after vaccination with the tested vaccine or a control vaccine/placebo, and protection is measured.

Phase III is the largest and final stage that tests the effects of the final formulation, as determined by Phase II. This is carried out on thousands of individuals (the number generally depends on the likelihood of contracting infection) and in field conditions. In some cases, vaccines are directly tested in a ring-fenced manner, where they are administered to close contacts of infected individuals. The principal endpoint of a Phase III trial is vaccine efficacy, measured as the percentage reduction in the incidence of infection amongst vaccinated individuals; however, additional targets can also be measured, including reduced hospitalization. Even after approval, like for any drug, vaccines continue to be monitored and side effects recorded.

Is it all about vaccines?
Though some antigens can be extremely immunogenic, not everything that comes into contact with our immune system causes a large reaction – fortunately! While live attenuated viruses tend to be highly immunogenic, subunit or inactivated vaccines are generally much less potent at inducing an effective immune response. What seems to be missing is an adequate stimulation of the innate immunity, which is then necessary to stimulate antigen-presenting cells (APCs), such as dendritic cells and macrophages, and to boost the maturation of B and T cells. This is when adjuvant comes to the rescue.

Discovered mostly by chance, adjuvants are the unsung heroes of vaccination. Their mode of action is often poorly defined, and while some new adjuvants are currently in clinical trials, these are fewer than what could be expected. One of the reasons is probably the cost of testing, as adjuvants need to be administered in combination with different vaccines and compared with other control adjuvants, largely increasing the size of a trial. Another reason is our limited knowledge of how to fine-tune innate immunity, especially at the population level.

Amongst the adjuvants currently in use, aluminium salts (or simply alum) are the oldest, and they are slowly being replaced by more inert and easier to metabolize formulations. Surprisingly, exactly how alum works is not entirely clear, although it has been shown to induce an inflammatory response in mouse, which is presumably responsible for the recruitment of APCs.

MF59 is a squalene-based oil-in-water emulsion that has been used since the 1990s. Squalene is a fully metabolisable oil also produced in humans during cholesterol synthesis. Its mode of action has been reported to be recruitment and activation of immune cells at the site of vaccination, as well as increased antigen uptake, probably related to cellular stress signals.

About a decade later, a new series of adjuvants was developed by GSK – AS03, AS04, AS01 – based on a detoxified form of LPS called monophosphoryl lipid A (MLA) together with other components such as Saponin (QS-21). Once again, studies suggest increased recruitment and antigen presentation by macrophages in the lymph node draining the site of injection, which is consistent with the adjuvant biodistribution.

Several other formulations and variants are currently in clinical trial, including cationic adjuvant formulations (CAF) where quaternary ammonium surfactants are formulated into liposomes or emulsions; glucopyranosil lipid adjuvants (GLA), made of a TLR4 agonist in squalene-based oil-in-water emulsion; and IC31, a cationic peptide administered together with an oligonucleotide specifically stimulating TLR9. Data emerging from these trials suggest that these formulations have fewer side effects at the site of injection than alum and generate better immunity, although this might not be true for every vaccine.

While the general mode of action appears to be the same (increased recruitment and activation of APC, which in turn potentiate downstream antibody and cytotoxic responses), the right adjuvant-vaccine match appears to be important and requires specific studies.

Altogether, it seems that a lot of progress has been made, but we still have quite a lot to find out, particularly as mouse models don’t accurately recapitulate human immunity. The relatively new science of system vaccinology, through a systematic and unbiased analysis of transcriptomics, proteomics, and metabolomics on human cells and at the population level, is likely to help shed light in these fundamental areas in the near future.