With Donald Trump applying pressure to release a vaccine, possibly before any vaccine completes Phase 3 trials, it’s worth looking at what that means, and the possible consequences other than a complete loss of public faith in the FDA, CDC, and the vaccination program.
Before they reach the public, most drugs and vaccines go through a lot more than three phases. There are pre-clinical studies done to determine whether the basic idea behind a treatment can be expected to work. There are cell studies, done in a Petri dish or equivalent, to see if the drug appears to be effective outside a living body. If those first two steps look promising, a treatment can move on to animal studies, often to several animal studies, in everything from mice to primates. Still, it’s understood that even trials conducted among humanity’s closest relatives don’t necessarily reflect the safety or efficacy of a drug in people. That’s why there are at least three phases of human testing. Sometimes more. And there are very, very good reasons not to take shortcuts.
A minority of drugs start with what’s known as a “Phase 0” study, specifically created to speed up the approval process. These studies involve only a few people, and give researchers a chance to test some aspects of basic safety using very low doses of a drug; doses intentionally lower than those expected to be effective. In the case of COVID-19, there were some novel technologies involved. For example, both Moderna and Pfizer/BioNTech are using memory RNA vaccines with previously untried delivery systems, so some very light dosage tests were administered.
If everything is still going well, it’s on to Phase 1, where people in the first group are given similar low doses and watched carefully for their reactions—safety is the goal. These studies are usually done in waves, with the final set of volunteers receiving a level that’s close to the expected dose in general use. Still, the number of people involved in a Phase 1 trial can be expected to number only in the dozens, as these trials carry a substantial risk of adverse reaction.
It’s important to recognize that while the results of a Phase 1 trial are focused on safety, they do not answer the question “is this drug safe to use in the general population?” The entire focus of the safety review at this stage is limited to “is it safe enough to proceed to testing this drug on a larger group of people?” That’s it. The small number of volunteers in Phase 1 testing can deliver a definitive “no” on the safety of a drug or treatment, but they cannot deliver a definitive “yes.”
If researchers and administrators feel good about the safety results in the Phase 1 trial, they can begin Phase 2. This is the phase that tries to answer a very basic question: Does the treatment work against the disease? The size of the Phase 2 trial is generally larger than Phase 1. In many drugs, this phase is still restricted to around 100 individuals, but with vaccines it’s generally larger. The Oxford vaccine being manufactured by AstraZeneca, most often the focus of “October surprise” stories, had just over 1,000.
When manufacturers are in a hurry, as is the case in trying to come up with a vaccine for COVID-19, the line between Phase 1 and Phase 2 can get a little blurry, with researchers looking at not just safety but efficacy in Phase 1 volunteers, and Phase 2 volunteers receiving dosage levels not seen in Phase 1. Both of those things happened in various COVID-19 trials.
However it’s divided up, if Phase 1 trials show that the drug is basically safe enough to test, and the Phase 2 trials show some encouraging results in terms of dealing with the disease or condition being treated, a Phase 3 trial can be authorized. Again, this doesn’t prove that the drug is widely effective, or that it’s safe for the general public. It only says that it’s safe enough to move to Phase 3, and that there were enough encouraging results in Phase 2 to warrant additional study.
Phase 3 is where the size of the trial can go up dramatically. The Oxford/AstroZeneca vaccine is expected to be tested by over 90,000 people—30,000 of them in the United States—with 2/3 of those volunteers actually getting the vaccine, and the other 1/3 receiving a placebo injection. The testing for Oxford got underway in July, as did that for Moderna. Pfizer/BioNTech is close behind. Three Chinese vaccines are also in Phase 3 trials.
Just like Phase 1 and Phase 2, the Phase 3 trial results will come in waves. Sometime in September, there should be releases from both Oxford and Moderna to address safety from the first wave of Phase 3 patients. Those results are also likely to provide information on the level of antibodies produced. However, they are unlikely to have enough data at that point to give any measure of efficacy in the sense of “did this vaccine effectively prevent the volunteer from being infected by the SARS-CoV-2 virus?”
Still … this is the scary point. Because the pre-print release of this initial set of data could very well be the point at which the FDA, under pressure from Trump, issues an Emergency Use Authorization allowing AstroZeneca, and possibly Moderna, to proceed with distributing their vaccine to first responders, healthcare workers, and those at greatest risk of infection.
To see why it would be very bad for the FDA to make that decision requires looking no further than the FDA. Because in 2017 that agency released a report on 22 trials—including 5 vaccines—where the outcome of full Phase 3 testing was very different than the results suggested by Phase 2.
In the case of seventeen trials, the problem was simple enough; what had appeared to be an effective treatment in small-scale testing, did not turn out to be effective when tested more broadly. The reasons for this were numerous. They included some unintentional selection bias in establishing volunteers for Phase 2, some poor procedures in conducting test or analyzing results, and some hopeful readings of outcomes shading toward deliberate overstatement of benefits.
In the other five trials, the problem wasn’t just that the treatment was ineffective, it was also unsafe. This included trials of a vaccine against Staph. Extensive Phase 3 trials found cases of serious side effects, drug interactions, effects that were amplified by the presence of other medical conditions … things that simply did not turn up when testing with smaller numbers of volunteers. Sometimes these safety issues can be overcome. Sometimes they can even be justified when talking about a treatment given only to people whose outcome is very dire. However, it’s very hard to justify significant safety issues in a vaccine that could be distributed in billions of doses.
And while right now Phase 2 trials on all three of the vaccines most likely to appear in the U.S. by the end of the year are showing that volunteers develop good levels of antibodies, it’s still unclear what level of antibodies is needed to protect against COVID-19, or even how strong the relationship is between the antibodies and immunity. As the FDA noted when reporting on these trials that flunked Phase 3 … “As several of these case studies illustrate, promising biomarker data in phase 2 do not necessarily translate into effective product performance. Biomarker data were promising in phase 2 testing in products targeting conditions ranging from heart disease (aliskiren, darapladib, torcetrapib) to Staph infection (V710 vaccine), and from AD (semagacestat) to herpes infection (HSV-2 vaccine). These experimental products were not proven effective when tested in phase 3 trials.”
It seems all but certain that exposure to COVID-19 provides a period of immunity from being reinfected by the virus. The rare instances in which patients have become infected a second time are most notable exactly because of how rare they are. But we don’t know the best way to measure that immunity, or how the antibodies currently being examined in vaccine trials really relate to immunity, especially over time.
Hurrying this process means exposing millions of people to unwarranted risks, either in the form of safety or efficacy concerns. For the sake of everyone being vaccinated, and for the huge benefit in public opinion necessary for a successful vaccination campaign, it’s vital that the testing of any vaccine be thorough and reviewed before widespread distribution begins.
