Preclinical Research: The Kingmaker of Vaccine Development

Vaccine development often only hits the headlines when there’s a breakthrough clinical trial or final approval. Behind every successful vaccine, however, is a lesser-known stage of research that quietly lays the groundwork: preclinical studies. Don’t be fooled: these early investigations are essential for answering crucial questions which underpin the future of the vaccine. Is the vaccine safe? Does it trigger the right kind of immune response? Could it actually prevent illness in humans?

Using a combination of in vitro experiments, computer simulations, and animal testing, researchers evaluate vaccine candidates and filter out those that aren’t suitable for further trials, particularly those involving human subjects. Here, we’ll give an overview of the questions preclinical vaccine testing seeks to answer, as well of some of the techniques used to investigate the safety and efficacy of vaccine candidates.

The final boss of many vaccine development programmes is the clinical trial, where the vaccine is pitted against a control in human subjects to prove it fulfils its goals. Before the traditional three-stage clinical trial can commence, however, they must overcome the mini-boss of demonstrating that it is safe and effective enough for further investigations to be ethical and worth the investment.

The requirement for preclinical testing is far from unique to vaccines. All novel drug products must demonstrate a base level of safety and efficacy to begin more comprehensive clinical trials. For now, however, we’re going to focus on the key components of preclinical testing from the perspective of vaccine development. There are several questions which are critical for determining whether a vaccine can progress:

Is the vaccine safe?

Preclinical research is important for determining whether the vaccine is likely to cause frequent and/or dangerous side effects before it is tested in human subjects. If there are indications that these might be a concern – if test animals exhibit serious illness after being dosed with the vaccine, for example – then it would be unethical to test the vaccine further in humans.

Does the vaccine elicit an immune response?

To be confident that a vaccine provides protection, we must first demonstrate its ability to activate the immune system: its immunogenicity. This commonly takes the form of an antibody or T-cell response, with the specifics dependent on the best method of interrupting the natural progression of the disease.

Does the vaccine afford protection?

Once we are confident that the vaccine generates an immune response, we must show that this response is protective against the disease. An important aspect of this stage is establishing correlates of protection. These are markers which indicate that the subject has protection from the vaccine, and can be used to help determine whether the vaccine would be effective in humans based on data from non-human subjects.

What is the optimal dose of the vaccine?

It is also important to establish the dose of vaccine which stimulates the strongest possible immune response while still being safe. Different formulations of the vaccine are also often tested at this stage. For example, the inclusion of an adjuvant – an ingredient which increases the immune response to the vaccine – can be tested during preclinical experiments.

How does the vaccine work?

While the fact that the vaccine is safe and effective may be sufficient to provide a good medical intervention, the business of why it is so can be crucial to future development of the vaccine. While animal models are chosen to have immune systems similar to those of humans, for example, it may be the case that the mechanism of action of the vaccine in those models does not afford the same protection in humans. An understanding of the mechanism can inform updates to the formulation and/or dosing of the vaccine to overcome these challenges.

In Vitro vs In Vivo Potency Assays: Modern Vaccine Batch Release

Choosing a Path: Natural History Studies for Vaccine Development

What is a Safety Population in a Clinical Trial?

Historically, preclinical research for vaccines was almost exclusively performed using non-human animal subjects. This is still common today, but such studies can be supplemented and even superseded using in vitro experiments and computer simulations. Let’s examine the costs and benefits of each of these approaches, and the different aspects of the vaccine performance they can highlight.

Animal models are used frequently in preclinical vaccine research. Early studies might use small animals, such as mice or guinea pigs, while studies later in the development process might focus instead on more human-like animals such as primates. We’ve previously covered some of the ethical challenges associated with using in vivo experiments in the context of vaccine batch release: most of these hold here too. Steps should be taken to only use animal subjects where necessary, minimise their usage where it is, and ensure that the suffering those subjects experience is reduced as far as is possible. These are the three Rs principles of animal use in research: Replacement, Reduction, and Refinement.

Nevertheless, animal models can provide an extremely powerful experimental platform for preclinical vaccine research. By examining the vaccine’s performance in a living organism, with all its complex interacting systems, researchers can begin to predict likely vaccine performance in humans. In particular, the presence of a full immune system allows the immunogenicity of the vaccine to be established. Gathering safety information is also an important part of animal studies as any adverse reactions to the vaccine in the subjects can be recorded and investigated. Indeed, most vaccines cannot be approved without some form of safety information based on animal studies.

Another capability associated with animal studies is the challenge study. These involve deliberately exposing both vaccinated and unvaccinated subjects to the disease and observing whether there is a systematic difference in outcomes for the two groups. This not only helps establish whether the vaccine offers protection against the disease but also allows for the investigation of correlates of protection. If certain markers are measured both in subjects which are vaccinated and unvaccinated, then these can be correlated to the observed outcomes. These can then be translated into a relationship between those markers and the degree to which a subject is protected by the vaccine.

While there are clear benefits of testing a vaccine in a living organism, no animal model is a perfect analogue for the human body. That means that there is no guarantee that results from animal models will be observed when the vaccine is used in humans: it is possible that the vaccine will not be as protective or induce side effects not observed in non-human subjects. Research into a vaccine for HIV is one such example: vaccines which provided protection for an analogue virus in monkeys were found to not be protective when tested in humans. There is also no guarantee that safety outcomes will translate from animal models to humans. These challenges fall alongside the usual time and resource intensiveness associated with in vivo experiments – lab space, housing and feeding animals, etc.

While in vitro experiments are performed in simplified systems, they can nevertheless provide useful information about the safety and efficacy of a vaccine, particularly early in its development. An example of this is high throughput screening of vaccine candidates. In vitro experiments usually require just hours to reach a result, as opposed to weeks or even months for in vivo studies. That means that several candidates could be tested simultaneously and winnowed down to the best performing options, which can then be taken forward to animal trials. For example, a range of possible antigens from a target virus could be tested in an immunobinding assay to determine which induced the strongest response.

In vitro experiments are crucial for understanding the mechanism by which the vaccine generates an immune response in the body. This is where the simplified nature of the systems being investigated is a benefit: a living organism has a vast array of complex and interacting systems which can obscure the exact effect a vaccine has on the immune system. In an in vitro study, by contrast, the key components can be isolated. A vaccine candidate can be introduced to a sample of serum, for example, and the response of proteins in that sample can be examined. This is also an opportunity to test different adjuvants to determine which boosts the immune response the most.

Many in vitro experiments use lab-grown cell culture models. These include white blood cell cultures which can be “vaccinated” in vitro to directly observe the immune response produced. These experiments can also provide safety information by testing for any direct toxicity to the cells. If there are indications of cell damage or death as a result of exposure to the vaccine, this could be a sign that the formulation is likely to cause side effects.

More recently, preclinical studies have started to employ organoids and 3D tissue models to test vaccine candidates. These are miniature or otherwise simplified versions of real organ structures which are grown in a lab from donated cells. One example might be a lymph node environment grown entirely in vitro. This can then be exposed to a vaccine, and the response observed. This added complexity serves as a bridge between cell culture studies and full in vivo experiments, and, since donated human cells are used, might even better predict human responses than animal models.

21^st century medical research is regularly supplemented by digital analysis, computer simulations, and, more recently, artificial intelligence. Preclinical vaccine research is no exception. As computation has become more efficient and less expensive, in silico techniques have been applied to an ever-increasing part of the vaccine development process to better target future research.

One example of how modern technology has been used to augment vaccine development is reverse vaccinology. A key stage of early vaccine development is to select an antigen to be targeted by the immune response generated by a vaccine. These are proteins expressed on the surface of a pathogen, such as the spike protein of a COVID-19 virus, which used by the immune system to identify and eliminate pathogens. Traditionally, selecting an optimal antigen would involve culturing live pathogen and identifying antigen targets using assays, which is a time consuming and inefficient process. Reverse vaccinology, by contrast, utilises modern genomics and simulation techniques to predict which antigens would make good targets for a particular pathogen. This can dramatically reduce labwork by identifying only the antigens with the greatest potential for further investigation, compressing work which could take years to months or even weeks.

Similarly, researchers can use immunological simulations to predict the immune response to a vaccine. While the immune system is far too intricate to be simulated in full complexity even using the most powerful modern computers, even these simplified predictions can be used to guide vaccine research and select the most promising candidate for further investigation. Immunological simulations can be used to predict whether a response to a vaccine would be predominantly T-cell or antibody mediated, identify biomarkers which could prove useful as correlates of protection, and whether any pathways are likely to be overstimulated and lead to side effects. While further laboratory investigation of predictions of simulations will always be required, this labwork can be streamlined to focus on the most promising lines of inquiry using in silico techniques.

Artificial intelligence is also proving increasingly useful as a tool for vaccine research. AI models can be used to efficiently identify antigens or suggest adjuvants for a vaccine both alongside more traditional simulations and labwork. As with so many areas of modern life, there is huge interest in the power that AI holds, and the potential for augmenting human-led vaccine research with AI tools is almost limitless.

While no vaccine will be designed and tested using only in silico modelling and simulation, such research serves as an extremely effective filter and accelerator for preclinical vaccine research. By identifying targets and predicting immune responses more efficiently than could be done using traditional labwork, effective use of in silico research allows the timescale – and, therefore, cost – of vaccine development to be reduced. It also has the potential to allow targets and pathways which would never have been noticed by human researchers alone to be identified, allowing for improved safety and efficacy over what would have been otherwise possible.

One of the main objectives of preclinical is to decide whether a vaccine candidate is worth the time and expense of further investigation. If a vaccine does not demonstrate that it provides sufficient protection in preclinical trials, or if there are significant safety concerns, then it may be pointless or even unethical to undertake further trials.

While this screening is important, it is far from the only outcome of preclinical studies. Evidence gathered in preclinical research is used to shape later stages of the clinical trial process. As previously mentioned, one aspect of this are correlates of protection. These biomarkers can be used to inform the endpoints measured in future trials.

Alongside the correlates of protection, another aspect of the clinical trial process which is informed by preclinical research is the safety monitoring plan. While subjects of clinical trials are carefully monitored for any adverse events, particular attention can be paid to any events which were common in preclinical studies. For example, if an animal study found that several subjects experience injection-site reactions, then this can be flagged as an adverse event to be aware of in the safety monitoring plan for later trials. Trial design can also be adjusted to minimise the effect or frequency of these events.

Further, preclinical animal trials often inform the dosing regimen used in human trials. Typically, the first dose used in humans is a small fraction of the maximal safe dose in animal studies. This is then escalated until a protective response is then detected. The number of doses for complete vaccination can also be informed by animal studies. Some vaccines, such as mRNA COVID-19 vaccines, are more effective with two doses separated by a few weeks or months. This prime-boost regime is tested in preclinical animal studies before deployment in human trials.

Preclinical research is one of the most important phases in vaccine development. It gives scientists the opportunity to test vaccine candidates for their safety, efficacy, and feasibility before moving towards human trials. These early studies help researchers avoid costly mistakes, refine their approaches, and, ultimately, choose the most promising vaccine candidates to carry forward.

As science and technology continue to advance, preclinical tools are becoming even more powerful, including in vitro and in silico experiments and simulations. All the while, animal models remain a crucial if controversial aspect of preclinical research. And while no model can perfectly predict how a vaccine will perform in humans, this phase remains essential for building a strong, evidence-based foundation for clinical success.

In the end, every safe and effective vaccine we have today owes part of its success to the quiet, careful work that happened behind the scenes, long before the first clinical trial began.