We were fast: It took only 326 days to create, test, and authorize the first highly effective vaccine against COVID-19. Compared to typical vaccine timelines, where progress is measured in years, the development of COVID-19 vaccines was an incredible accomplishment.

It took Moderna just 65 days from when they received the genetic sequence of the then-novel coronavirus on January 11, 2020, to design the vaccine, demonstrate efficacy in vitro and in animals, and start the very first human trial on March 16. All of this was done on a novel vaccine platform, never before used at scale.

However, it took 270 additional days — just under nine more months — until the vaccine would be approved by the FDA under emergency use authorization for the general population. While the regulatory process comprised several of those weeks in November and December, the majority of the time it took to develop the first effective vaccines was spent on clinical trials. 

To anyone with even the most passing knowledge of biomedical research, this comes as no surprise. Modern clinical trials are incredibly complex. Doses used in trials must be manufactured to sufficient quality and safety standards. Paperwork has to be filed with regulators. Predictions need to be made about which countries will have the most infections, insurance has to be negotiated, and tens of thousands of patients must be recruited. Every adverse event for each of those patients has to be reported in detail and in almost real time. During the COVID-19 pandemic, the clinical trial process worked as well as it ever did and delivered results faster than almost anyone expected. 

Many are wondering if we could have been even faster. Even before the vaccines were approved, some had criticized various regulatory delays — for example, the several weeks it took from when trial results were in to when the vaccines were finally approved. We had the vaccine since the spring; was it really optimal to wait nine months?

The critics rightly highlight the value of speed. In 2020, over 10,000 people were dying with COVID-19 every day. Vaccinating earlier would have helped to prevent some portion of these deaths. In fast-spreading epidemics, the full benefits of vaccination quickly diminish with time. Delivered earlier, they both protect at-risk individuals and help to encourage herd immunity. 

Extreme speed pressures can also, of course, lead to errors — like when AstraZeneca, one of three major developers to bring out a successful vaccine in the first year of the pandemic, accidentally gave a dose that was lower than planned to a large proportion of participants in their trial. While in ordinary times this type of delay would be a minor tragedy (albeit a major business failure), in a pandemic it is a harrowing prospect: What if the intervention that could save millions of lives is delayed by months because of a small and hard-to-detect accident?

One thing that everyone agreed on is that the pandemic proved that we can find cures at breakneck speed. In the next pandemic, lives will depend on whether we can test what we invent fast enough. How fast? CEPI, the Coalition for Epidemic Preparedness Innovations, a foundation that finances independent research projects to develop vaccines against emerging infectious diseases, has set a target to shorten the timeline from a year to 100 days. In order to achieve this goal, we will need a number of breakthroughs not just in the science of vaccine development but also in clinical testing and regulatory approvals. 

Human clinical trials of vaccines begin in Phase 1 by enrolling a small number of subjects, typically fewer than 100, to first establish that the vaccine is safe. In Phase 2, the drug is given to a larger group (typically several hundred) to see whether patients’ immune systems respond to treatment and at what dose. (Outside of pandemics these steps take many years.) Only then does a promising vaccine move to large-scale (Phase 3) trials, usually double-blind, randomized, controlled studies, which can establish efficacy.

Because Phase 3 trials track participants in the real world, there is no way to reliably tell how many of them have actually been exposed to the pathogen that the vaccine is supposed to protect against. This means that to show whether or not it works, trials measure how many people have been infected in each group. To do this, they have to run for a long time or recruit a lot of participants. ¹ The math is very simple: Having twice as many participants will halve the expected length of the trial.

So why didn’t the various COVID vaccine trials simply recruit twice as many participants? First, there are logistical constraints on how quickly people can be vaccinated. It took the COVID trials months to recruit the tens of thousands of participants they required. Presumably, this problem could be alleviated by simply spending more money on a trial. However, the costs of the vaccine trials are already very high. A recently announced Phase 3 trial for a new tuberculosis vaccine, for instance, is expected to cost $550 million. For COVID vaccines, while it is hard to find a cost figure for trials alone, the total costs of R&D were in the billions for each of the successful vaccines. 

In 2020, both vaccine developers and regulators worked hard to speed up the traditional development process. First, developers — Pfizer, Moderna, AstraZeneca — assumed more risks by merging some phases of clinical development into Phase 1/2 or Phase 2/3 trials. Typically this would not be a preference for vaccine makers, in part because a large proportion of vaccines fail at each phase. (The historical average really depends on how one counts. In the last two decades success rates have been relatively high, but it’s reasonable to assume a >50% chance of failure at each stage.) It’s also easier and less risky to write separate protocols for each trial rather than planning all stages at once. Cooperation from country regulators throughout the testing also helped reduce wait times by, for example, allowing companies to skip animal testing. Trials also benefited from a multitude of logistical and manufacturing improvements. ²  

All of this yielded large benefits relative to expectations: in March 2020, most experts regarded an 18-month timeline as optimistic. The 326 days it took Pfizer and BioNTech to deliver their highly effective vaccine was close to optimal given the constraints of the current system; CEPI estimates that the maximum achievable speed was 250-300 days. But while shaving even a month or two off vaccine development timelines could affect millions of lives, this is still some distance off from the 100-day target set by CEPI. So is it possible to get faster than even the very ambitious COVID timelines?

There are several ways we can try. 

Unlike normal trials, which have fixed protocols, adaptive platform trials allow for multiple treatments to be evaluated simultaneously, dynamically adding or removing them from the study in response to what is learned during the trial. Adaptive designs aren’t used widely both because they are complex (not all trials need dynamic elements, especially if they are not under time pressure) and because guidelines provided by medical regulators are insufficient. 

In 2020, simple adaptive elements were used to design stopping rules for COVID vaccine trials — in addition to using joint protocols for multiple phases, they also included interim analyses, allowing them to report results earlier. The reason that Phase 3 trials need to be so large is that only a fraction of their participants get infected. Interim analyses occur after reaching a certain number of infections; if at that point there is overwhelming evidence that the vaccine is effective, the trial can conclude. This helped COVID vaccine trials report their results faster, but it’s hard to say by how much. ³  

Platform trails evaluate many interventions against a common control group. These are more challenging to set up, and it’s difficult to use them to compare effectiveness of different vaccines. However, their obvious benefit is that platform trial participants help study the most promising vaccines without having to start a new trial. For COVID this was not particularly important, because the decision-makers largely bet on the winning horses: Moderna, Pfizer, and AstraZeneca. The lower the chance of success, the larger the benefits from having a system which allows testing many candidates in parallel. Some vaccine trial platform approaches were tried later in the pandemic, and they were also very effective for finding COVID therapeutics, but to make a difference in the next pandemic, we will need to build experience in running this type of trial: to date, only a handful of platform trials have ever been tried.

Human challenge trials (HCTs) are a solution to the problem of waiting for infections to occur: In HCTs, healthy young volunteers are exposed to the virus deliberately in order to study the vaccine’s protective effects. This is routinely done to study the characteristics (e.g., infection process, symptoms, etc.) of various pathogens, such as the flu. In the case of COVID, an HCT could have proved a vaccine’s efficacy in a matter of weeks with only 100-200 participants.

However, there are two major problems with the science of HCTs. First, scientists need to produce an attenuated version of the virus that can safely be given to volunteers — that may take months. (At least, that’s how it has been done historically: A more risky option is to also let people infect each other “naturally,” in a less controlled way.) Second, vaccines are typically less effective in the elderly due to aging of the immune system: the same process which makes the elderly more at risk when exposed. This makes it hard to extrapolate the results from HCTs to people we most need to protect.

In 2020, 40,000 volunteers declared their readiness to participate in an HCT; two of the major COVID vaccine makers explored using HCTs to test their vaccines. ⁴ But there was no real willingness on the part of regulators to approve vaccines based on HCT data, and the idea was never tested.

For many diseases, we know how immune reactions to a vaccine correlate with protection from disease. That is why we can periodically produce a new version of the influenza vaccine without having to recruit tens of thousands of volunteers; it’s sufficient to measure immune response for a handful of subjects in the lab. 

For SARS-CoV-2 no such correlates were firmly established at the time when vaccines were being developed. However, broad scientific efforts are in place to establish or improve correlates of protection (CoPs) not just for coronaviruses but also for other threats. While less preferable, CoPs may offer the first line of defense in future pandemics, at least in some of the main pathogen families.

Another option to ensure faster vaccine coverage of the most vulnerable populations prior to regulatory approval is to use the expanded access framework, often referred to as compassionate use. In the U.S. (and the EU), individuals can request access to some not-yet-approved drugs provided they have a physician’s recommendation. This model was first used in the early 1990s with azidothymidine, the first HIV drug. 

In summer of 2020, some argued for use of COVID vaccines in care homes. Dr. Deborah Birx, the White House Coronavirus Response Coordinator, encouraged vaccine makers to use the compassionate use pathway, though none ultimately did.

Though all of these options were floated in some form, none had an impact on COVID vaccine development. Some were not technologically feasible at that early stage of that pandemic: An attenuated virus would have taken months to develop for use in HCTs, and the vaccine process was too early for correlates of protection to play a role. Some did not seem viable under a decentralized R&D ecosystem: While many regulators are generally supportive of platform trials, it’s hard to imagine the main vaccine producers pooling together to start a joint trial of their vaccines. 

However, in a couple of cases it seems that feasible solutions were simply discouraged by the regulators or vaccine makers because they were seen as too risky. One example is pursuing compassionate use for the elderly in the summer of 2020. Another one is the guidance made in September 2020 by the FDA for the vaccine makers not to submit for emergency use authorization until there were two months of safety data on at least half of trial participants (regardless of the size of the trial, it seems, which makes the reasoning behind the request hard to understand). This in turn made the adaptive elements of trials largely irrelevant, since the trials were then very likely to run until November.

Successful development of vaccines drove profits of tens of billions of dollars for the vaccine makers. Thanks to long-term multibillion vaccine purchase contracts, most of revenue was captured by the first few companies that were able to get their vaccines approved and mobilize large-scale manufacturing. ⁵ But the costs of vaccines were nothing compared to their benefits. By one estimate, in 2020 the social value of a COVID vaccine was over 100 times higher than its cost. ⁶ This gap justified supporting research and mass production of many vaccines in parallel, such as was done in Operation Warp Speed — a form of insurance by increasing the number of “shots on goal.” This was a wise choice. According to another estimate the entire OWS would have paid for itself by shortening the pandemic by 12 hours.

Vaccine approval depends on showing that the product meets the regulator’s stated criteria of individual safety and efficacy. Once a vaccine is authorized for use, we continue to monitor its safety, effectiveness (how well the vaccine performs in the real world as opposed to within a trial), and durability. However, important questions remain. Does the vaccine maintain its response against virus variants? Was the vaccine dose tested in Phase 3 trials optimal? For example, we know that vaccine response changes with age, but due to cost and time pressure the manufacturers will typically choose to only test one dose, the one which they think will work best for an average person. ⁷

This is what potentially happened during the pandemic — and in fact, it may have happened twice. First, in early 2021, the U.K. decided to extend the gap between two doses of the first approved vaccines from the three- to four-week gaps tested in clinical trials to 12 weeks. This allowed more people to get some measure of protection in the first crucial months of the vaccination campaign. A modeling study suggests that this actually reduced COVID mortality in the U.K. in 2021 by about 15%. Many other countries that faced shortages soon followed suit. 

Second, some trial data suggested that it may have been beneficial to lower the doses of Moderna and Pfizer vaccines (which could be done simply by drawing less volume into the syringe): Doing so would reduce the efficacy only a little bit, but allow for much faster vaccinations. This idea was backed by immune response data and subsequently tested in many trials, but not implemented outside of the lab — for the first course of vaccination, at least, since Moderna reduced the dose for booster vaccinations. 

In both cases, those opposed to trying these were reluctant because of the lack of comprehensive clinical trial evidence. And under the current clinical trial framework, it is no wonder that no such data existed: In the race to develop the vaccine, why would a producer spend additional hundreds of millions of dollars to test a slightly less efficacious version of their vaccine? In fact, in the case of AstraZeneca it is likely that it was the initial dosing error that inspired some subsequent trials of lower dose — the patients accidentally given the lower dose did not significantly differ in their immune responses from those given the full dose. While data could subsequently be generated, and in both cases it seemed to confirm feasibility of both extending the gaps and of setting the vaccine dose lower, all of it came in too late. In the end, the experts still had to make uncertain decisions on topics such as lowering doses, extending gaps, or when to deliver boosters without complete clinical trials. 

The success of the COVID vaccines depended on advances to the core science of vaccine making. To make the vaccine development process even faster, we will need to look beyond the lab bench and find out how the science of clinical trials can be improved. We have several options, each of which requires us to also think about incentives and the regulatory framework that influences them. These developments must happen before the next pandemic arrives. None of them alone are guaranteed to reduce the time it takes to create, test, and deliver vaccines to 100 days, but all of them together may increase the likelihood we get there. COVID has proven decisively that it’s economically viable to pay for a wide range of solutions.