Introduction
Recently, I was privileged to act as the moderator for a World Drug Safety Congress webinar exploring current considerations for developing and testing COVID-19 vaccine solutions at pandemic speed. Our panel of experts included:
- Dennis Brooks, Global Safety Physician, Infection/Vaccines, AstraZeneca
- Rob Lambkin-Williams, Independent Consultant & Advisor, Virology Consult Ltd.
- ShuPing Yang, Senior Director, Clinical Development, Inovio Inc.
The global spread of COVID-19 and its great impact on the health and wellbeing of our population has imposed a race to develop an effective vaccine at unprecedented speed. A development process that usually spans several years is being compressed into one year – or less. Safety is a vital part of this process, and so we identified and discussed several key factors and considerations around safety in vaccine development. This article provides a synopsis of our insights1.
1. How safety can drive vaccine development
The Brighton Collaboration comprises 500 vaccine experts from 57 countries who standardized guidelines for data collection and analysis, greatly enhancing the comparability of vaccine safety data. Thanks to their work, we can quickly define criteria for case definition and gathering information at a very high-quality level and during the first report. Therefore, it reduces “queries” and the timeframe for vaccine development. As we approach COVID-19, a methodology is available that enables us to shorten the development and trial process. The goal is usually to launch vaccines in the market with only a surrogate endpoint of efficacy, usually the level of antibodies that is protective, achieving the best quality based on our understanding of vaccine safety.
From a pharmacovigilance perspective, we must thoroughly evaluate the adverse effects in the pre-licensure period as much as possible to mitigate the risk when the vaccine is largely utilized. Can we mitigate risk? There is no precedent of such timing for vaccine development. When risk-benefit is evaluated, we ask, “Is the risk-benefit worthwhile when we consider rolling out a vaccine to a vast population?” During clinical trials, we can gather data and analyze safety signal results, which helps us form a risk template.
Is this Mission Impossible? Nobody knows. Fortunately, today we can use software models in the preclinical phase to compare the antigen structure with many human protein expressions to identify possibilities of “mimicry” and auto-immune reactions. It might be a major factor in speeding development. The timing of some SARS-CoV-2 vaccines is scheduled in a very short time, such as the University of Oxford vaccine development process, injecting the initial patient in April of 2020 and working toward broad-scale vaccination in the fall of 2020.
Balancing speed and caution is a challenge in vaccine development. With thousands of people dying of COVID-19 every day, development experts are well-motivated to bring a vaccine to market quickly. On the other hand, vaccines will be used in a healthy population, and should do no harm; this requires a delicate balance between rapid product development and careful safety mitigation. Some new creative ways to identify and mitigate risks during development and launch need to be established, potentially using the classical structure of risk management plan (RMP) and risk evaluation and mitigation strategy (REMS) processes that have been in place for a long time.
A robust safety plan mitigates safety risk during trials and helps during the regulatory approval process. If the safety plan is not well defined, the regulatory agency will have issues, and this will impede the regulatory evaluation process. Potentially, we can leverage safety data from similar vaccine products in past development. MERS, Ebola and related vaccines might share the same DNA backbone, with MERS having added coronavirus relevance. Understanding the safety profile of the DNA backbone allows us as developers to move faster.
Also, safety platforms might be helpful as in current influenza vaccines, where well-known platforms are able to achieve a new flu vaccine every year. In the scientific community, we acknowledge the likelihood of facing ongoing pandemics on a regular basis. Therefore, the established platform might be key for the timing of a viral pandemic. The experience and information accumulated from COVID-19 might be very precious to build safety platforms that will speed future vaccine development using similar successful technology (i.e., mRNA, DNA or Adenovirus vaccine technology).
2. Strategies for benefit-risk evaluation
When we are injecting healthy people with vaccines, we must be confident of a high benefit and low risk for mass populations. As mentioned above, a REMS or RMP process already used for more than ten years can help evaluate vaccine benefit-risk. The objective of a REMS/RMP is to establish surveillance, assessing adverse events and used as a tool to mitigate the risk when a safety signal is detected, such as in an education program for health care providers.. Pre- and post-licensure safety plans are critical components of a clinical development plan. With vaccines, it is classical to have a large number of people (around 60,000) vaccinated to review safety and “real” efficacy (based on the reduction of case disease number in the cohort).
Animal models, particularly nonhuman primate challenge models, can guide clinical safety mitigation and even efficacy assessment, so considering preclinical data is also an important key. Another risk mitigation factor is special populations, for instance, the elderly or pregnant women and those with underlying medical conditions. The safety profiles may be different than those already studied in clinical trials, such as a young and healthy, non-pregnant population. Furthermore, the elderly often demonstrated the need for a different dosage regiment for efficacy (i.e., flu vaccines).
An additional challenge in the actual outbreak is the lack of immunity passport, as we do not know which level of antibody is protective against COVID-19. Further, immunomodulation is a concern, as the disease is expressed in potential waves – leading to pneumonia and cytokine storms in some patients. Experts believe that the virus is not just the problem, but the immune response of the host as well. There are many aspects of COVID-19 we do not yet understand. Those knowledge items are key elements in the benefit-risk tradeoff.
3. Combining CHIM with safety direction

It is a delicate process. Could control of human infection model (CHIM) efficacy studies be combined with vaccine safety evaluation? At some point in our efforts, the vaccine must meet the virus. In a CHIM, we vaccinate people and then deliberately expose them to the virus. We quarantine subjects in a controlled environment to avoid viral spread and to isolate them from other variable factors. Will we get a simple antibody reaction? When the CHIM model is incorporated into the development process, we combine safety and efficacy into the same short time.
We can consider whether a CHIM is ethical and useful for vaccine development when an outbreak is ongoing. Do we prefer to test a vaccine in the real world but also closely monitor the development of CHIMs? Looking at outputs from the CHIM, we ask ourselves whether we can make adaptations. Can we block the cytokine storm observed in some patients? How can we know who will have this reaction?

CHIMs funnel directly into the human viral challenge (HVC) model. This model is being used by organizations to develop vaccine candidates, from target to proof of concept to approval. Early on, we try to understand as many factors as possible, so controlling these factors is helpful as we move from discovery and validation to clinical trial and development.
As we advance through the process, we ask ourselves if we are confident that we know the immune-protection aspect of a vaccine. It may depend on the vaccine, and there are too many false positives or negatives on antibody tests. Also, we do not know the role of cell immunity. Furthermore, we do not yet understand the antibody response in convalescent patients, what level of antibody is protective and for how long. It is not known whether antibodies can be used (immunoglobulins) to save COVID-19 patients. It is a must to evaluate efficacy based on immune reactions, underscoring the need for careful REMS/RMP consideration.
Conclusion
As the scientific community moves through the COVID-19 vaccine development process, our understanding and expertise will increase, positioning us for better and faster response to this outbreak and pandemics of the future. Safety plans, benefit-risk evaluations and models for studying efficacy will play critical roles. Sharing both our questions and the answers we discover can translate into insights that will revolutionize pharmaceutical development and clinical trials across therapeutic areas for years to come.
To listen to the full webinar, please click here
1Note: Panel participants offered their personal insights, not necessarily the opinions of their companies.