Why did the Pfizer young kids trial fail?
On Friday, Pfizer announced results of their Phase II/III trial in children 6 months to under 5 years of age. There were two big wins and one massive set back:
Wins: Vaccine was safe, so no severe events occurred. The vaccine dosage worked for kids between 6 months and under 2 years.
Failure: Vaccine dosage didn’t work for 2- to under 5-year-old population
Because of the failure, the process for authorization was stopped and Pfizer announced a Plan B.
As a mom, I was shocked and heartbroken. My girls were so close to getting their shot. We’ve waited so long and really needed a win.
As an epidemiologist, I’m not entirely surprised because 1 in 5 pediatric clinical trials fails (here, here). But I do have a ton of questions for Pfizer and really hope they clarify a few things publicly.
Pediatric vaccine trial
In May 2021, Pfizer’s clinical trials for the 6 months to under-5 population began. This was almost a year after the adult clinical trials began, which is a typical “age de-escalation” design: We test adults before we test more vulnerable populations like children.
Pfizer also overlapped phases. The scientists didn’t wait for Phase I to complete before moving onto Phase II. And didn’t wait for Phase II to end before moving onto Phase III. This isn’t the first time this has been done. In fact, this type of design even had a name before the pandemic: seamless and adaptive design. We basically save a ton of time by removing the “white space” between phases. For example, when we combine Phase II and Phase III, scientists can evaluate “dose selection” (which is typically determined in Phase IIb) and “confirmation” (typically Phase III) in one trial.
According to clinicaltrial.org, the official database of all clinical trials, Pfizer conducted a Phase I “dose finding” study where they tested several different dosages: 3 µg, 10 µg, 20 µg, and 30 µg (adult dosage) among a few dozen healthy children. Their ultimate goal was to find the “sweet spot”—the lowest dosage possible causing the lowest rate of side effects, but also enough dosage to generate an immune response.
Pharmaceutical companies have different approaches to finding the sweet spot for vaccines. For example, while Moderna has been trying to push the red line (what’s the maximum dosage without having too much risk?), Pfizer has had the opposite approach (what’s the minimum dosage with just enough benefit). As you can imagine, each approach comes with advantages and disadvantages. For example, Moderna is now facing the possibility of no authorization due to myocarditis concerns (i.e. too much risk) because of their high dosage, while Pfizer is facing the possibility of no authorization due to ineffectiveness (i.e. not enough benefit).
Thirty seven percent of clinical trials fail at this point—Phase I. But we know Pfizer’s clinical trial didn’t fail because they moved onto Phase II/III. So, this tells me that they successfully determined the “sweet spot” of 3 µg.
The next phase of the clinical trials is where a much larger group—a couple thousand kids—is given this “sweet spot” dosage (3 µg). By increasing the number of children, scientists confirm the dosage, but also have more “power” to detect rare side effects and increase the generalizability of the vaccine impact by including a wide range of ages, races/ethnicities, children with comorbidities, etc.
At this stage Pfizer also evaluated the vaccine by randomizing kids to the vaccine or a placebo. This allows scientists to evaluate a pre-determined “primary outcome”. In other words, what do we want the vaccine to change? In adult studies, Pfizer’s primary outcome was efficacy: How much does the COVID19 vaccine reduce risk of disease compared to a placebo?
With the pediatric trials, the primary outcome was not efficacy. This is because this was an age de-escalation study. We already know this vaccine biotechnology and RNA formula works; we just need to ensure a smaller dose works as well for younger kids as the full dosage does for older kids/adults. So the pediatric primary outcome is “immunobridging”: Does the smaller dosage (3 µg) mount the same immune response among under 5 year olds compared to the full dosage (30 µg) among 16-25 year olds?
Last Friday we got the answer: No. While the smaller dosage worked for the 6- to 24-month-old population, it did not work for 2- to under 5-year-old population.
What went wrong?
So obviously the 3 µg dosage was incorrect for 2- to under 5-year olds. But why wasn’t this detected in Phase I “dosage finding” stage? I don’t know. No one knows, because this data has not been publicly released or published. I have many questions, which include:
Did Pfizer test another dosage in addition to 3 µg in the Phase II/III trial? Why or why not? If they did, what were those results? If they did not, then what were the Phase I results that made them confident a 3 µg dosage would work?
Was the Phase I age distribution skewed? Because the vaccine dosage worked for 6- to 24-month-olds but not for 2 to under 5 year olds, I’d be curious to see if the sampling was skewed younger in Phase I. If so, then I could see how the suboptimal effectiveness was revealed when this dosage was expanded to a broader, more generalizable group.
What variant did they test? The ever-evolving landscape of this virus makes clinical trials very difficult. May 2021 is when Phase I/II started. This was before Delta. So the pediatric dosage was probably based on the original Wuhan strain and it probably did work. When the clinical trial was moved into the Phase II/III, Delta emerged, and we know Delta reduces neutralizing antibodies by 6-fold. In that case, the vaccine dosage may not have worked.
What was the antibody response? From the Pfizer announcement we know the antibody response was not comparable to the 16-25 year olds’ response. But that doesn’t mean there wasn’t a response at all. So, what was the response? And antibodies aren’t the only story; T cells were likely mounted. Was efficacy also tested in Phase II/III? This may help make the case for their Plan B (see below) and we should start vaccinating the two dose 3 µg series now.
What happens now?
Whatever the reason, though, the clinical trial failed. Our kids under age 5 won’t be getting a vaccine as soon as we had hoped. Because this is an age de-escalation design, they can’t just skip over toddlers and start vaccinating infants either.
So, Pfizer is going to Plan B. The same kids from the Phase II/III trial are going to get a third dose (3 µg) 2 months after the second dose. There are pros and cons to this approach. From Pfizer’s perspective, this is the easiest route— they don’t have to start another Phase III all over again with a different dose. That means they don’t have to recruit more people, it won’t be as expensive, and it won’t take as much time. They will lose money by not having a kids vaccine soon. It’s clear that this 3 µg is the incorrect dosage for toddlers, so I don’t like this messy approach.
A third dose may work, but what if it doesn’t? Starting with another dose would be the most rigorous and scientific path. And I hope they are doing this concurrently so we can compare in the end: How do 2 doses of 10 µg compared to 3 doses of 3 µg for toddlers? But it’s not clear that Pfizer has plans for this. Getting the correct dosage would also prevent kids having to get three shots, which would ease implementation thereafter.
I certainly don’t have answers and was just as surprised as the rest of the country that this clinical trial failed. Needless to say, it’s still going to be awhile until there are vaccines for our little ones. As a mom, I am incredibly frustrated. As a scientist, I am also very frustrated by Pfizer’s lack of transparency. But I weirdly find comfort in the fact that the clinical trials are robust enough to find errors like they should. This is just another example of the realities of science unfolding before our eyes. And sometimes that science is not all good news.
“Your Local Epidemiologist (YLE)” is written by Dr. Katelyn Jetelina, MPH PhD— an epidemiologist, biostatistician, professor, researcher, wife, and mom of two little girls. During the day she has a research lab and teaches graduate-level courses, but at night she writes this newsletter. Her main goal is to “translate” the ever-evolving public health science so that people will be well equipped to make evidence-based decisions. This newsletter is free thanks to the generous support of fellow YLE community members. To support the effort, please subscribe here: