Analyzing the Pfizer COVID-19 vaccine data in 5 to 11 year-olds

Pfizer vaccine 5 to 11 years old

In October 2021, Pfizer submitted a formal request to the US FDA for Emergency Use Authorization (EUA) of their BNT162b2 (Comirnaty) COVID-19 vaccine in children 5 to 11 years old. Supporting this request, the company submitted data from studies in this age group. As with any other COVID-19 vaccine, the request is reviewed by the FDA’s Vaccines and Related Biological Products Advisory Committee in order to make a recommendation for the FDA to grant authorization. Let’s take a detailed look at the data that was submitted for regulatory review, along with the FDA’s briefing document.

Why vaccinate children 5 to 11 years old?

Many vocal antivaccine proponents will state that “kids only get a cold from COVID”; however, this is not true.

Unfortunately, with the emergence of the Delta variant we have seen increasing rates of COVID-19 in children. In August and September, infections and hospitalizations due to COVID-19 rose significantly in children aged 5 to 11 compared to earlier in the pandemic:

Source: CDC

This rise is most likely driven primarily by increasing transmission in children, however there is also some preliminary (unpublished) evidence linking Delta variant infections with more severe illness.

All this is without considering the other complications that can occur with COVID-19 infection; such as: heart inflammation (that’s right, it’s more frequent in COVID infection than with the vaccine), multisystem inflammatory syndrome (MIS-C), or post-COVID condition (i.e. “long COVID”).

They probably all had pre-existing conditions

Many of the same antivaccine pundits will also say that severe infections “only happen in kids with pre-existing conditions”. Beyond the fact that this statement is cruel and absolutely ridiculous (children with asthma don’t all of a sudden deserve to be hospitalized or die?), it’s also blatantly false.

If you look at the data presented by the CDC’s COVID-NET surveillance program, almost half (46%) of the children in the United States who were hospitalized for severe COVID-19 had no risk factors or conditions putting them at increased risk:


What trials were done to evaluate Pfizer’s COVID-19 vaccine in children?

The studies to evaluate Pfizer’ COVID-19 vaccine included children aged 6 months to 11 years. However, Pfizer separated this population into 3 distinct groups:

  • 6 months to less than 2 years old
  • 2 years to less than 5 years old
  • 5 years to 11 years old

Pfizer began with the oldest group (i.e. 5 to 11) and then continued with the younger groups after safety and tolerability was demonstrated in the 5 to 11 group. The trials started with an initial (small) phase 1 trial, and then proceeded to a much larger phase 2/3 trial.


How many children in the 5-to-11-year-old age group were recruited in the studies?

Phase 1

In the initial phase 1 trial, there was a total of 49 children randomly allocated to 1 of 3 different doses: 10 µg (µg=microgram), 20 µg, and 30 µg.

This study was to look at the different doses in order to select the dose which gives the best effectiveness, while also limiting side effects.

Phase 2/3

After the dose was selected in the initial phase 1 study, they began recruiting children for the larger phase 2/3 study. This study randomly allocated 1518 children to receive the vaccine, and 750 to receive placebo (i.e. a 2 to 1 ratio).

At the FDA’s request, Pfizer doubled the amount of children in the study to ensure that safety was evaluated more thoroughly. So, an additional 1591 children were randomly allocated to receive a vaccine and 788 to receive placebo. For the analysis of safety, this gives us a total population of 4647 children: 3109 with a vaccine and 1538 with placebo.


How did Pfizer settle on a dose of 10 µg for 5-to-11-year-olds?

In the phase 1 study, different doses were evaluated. The first evaluation was for safety; where they looked for both local side effects (such as pain/redness at the site of injection) and systemic side effects (such as chills, headache, fatigue).

What was observed was that the amount of side effects increased along with the dose used (i.e. 30 µg had more side effects than 20 µg; which had more side effects than 10 µg):

Chart showing the various systemic side effects observed with the different doses
(Source: Pfizer FDA submission documents)

The second evaluation looked at the immune response generated with the different doses. Because the 30 µg dose was associated with many more side effects, it was not evaluated for immune response. This means the choice was now down to 10 µg or 20 µg.

No significant difference was seen in the immune response from the 10 µg dose or the 20 µg dose. Because they had the same impact on immune response, along with the fact that 10 µg had less side effects, the 10 µg dose was chosen. This represents 1/3 (one third) of the dose used in those 12 years and older.


How did the phase 2/3 study evaluate the vaccine?

The study was a randomized double-blind placebo-controlled trial.
Let’s first explain exactly what that term means:

Randomized

As I explained in a previous post dealing with ivermectin, randomization is the act of randomly allocating participants of a trial into the different treatment groups. The reason this is done randomly is to make sure that the groups are as similar as possible.

If participants are not randomly allocated, there tend to be differences that will emerge between groups. These differences can become very problematic in certain situations; for example:

Imagine a trial evaluating a blood thinner in which one group has a much higher bleeding risk at baseline than the other. This will, from the get-go, set up one group to be more likely to have bleeding side effects than the other (which could drastically impact the results of the study).

Double-blind

We all have biases; and no one is immune to them. Because we all have biases, it’s best to try to remove any bias that you can from a clinical trial. Double-blind means that both the participants and the investigators analyzing the final results don’t know which treatment a participant received. This information is kept by a separate group of people involved with the trial (and they don’t have a role in evaluating the results).

As I detailed in my previous post about confirmation bias, we humans have a subconscious tendency to trust something that agrees with our preconceived views more than something that refutes them. So, a double-blind trial takes confirmation bias (and other biases) completely out of the equation.

Placebo-controlled

A placebo is the most common way for trials to maintain blinding of participants and investigators. A placebo in the setting of the Pfizer vaccine trials is just a saline (i.e. salt water) injection.

As I explained in my previous post on homeopathy, the placebo effect is a real phenomenon. Because of this, having a placebo group in a study allows us to compare those with a potential placebo effect against the actual intervention being studied. This way, we can ensure that the intervention being studied is providing participants with more than just a placebo effect.

Study process

With regards to a COVID-19 vaccine trial, investigators monitor for 2 specific outcomes: side effects and cases of COVID-19.

After participants are given either the vaccine or the placebo, the study will continue to follow them for a predetermined period of time and then accumulate the results to see the rates of COVID-19 infection in the different groups. On top of this, they will also be following participants to see if they develop any side effects.


What side effects were seen?

The analysis of side effects with vaccine trials are usually subdivided into two categories: solicited and unsolicited side effects.

Unsolicited side effects are any side effects that are reported by the participant, without any prompting.

For solicited side effects, participants are given a diary to fill out that questions them specifically on the presence of certain side effects (such as pain/redness at the injection site, fatigue, headache, chills, etc). The good thing with this type of analysis is that nothing is missed. However, the rate of solicited side effects almost always ends up being higher; this is due to a bias that is introduced with questioning.

Do you notice how often you blink?
Well you do now, because I brought your attention to it.
*Manual blinking mode activated*

This is how solicited side effect reporting can sometimes skew results towards having more side effects.

For example, if you directly ask someone if they feel tired after receiving a vaccine, they are more likely to report it than if you didn’t ask. This is probably why almost 25% of placebo recipients reported fatigue after their 2nd dose.

Solicited side effects

The most common side effects observed in the group that had to fill out the diary are listed below. So that I’m not accused of being biased, I cited the rates for the 2nd dose. (These were generally the highest; so the rates for 1st doses will usually be lower)

I also decided to list the rate in the placebo group, so that you have an idea of the difference while also appreciating the placebo effect:

Side effectVaccine
Group
Placebo
Group
Injection site pain71%29.5%
Injection site redness18.5%5.4%
Injection site swelling15.3%2.7%
Fatigue39.4%24.3%
Headache28%18.6%
Chills9.8%4.3%
Fever6.5%1.2%
Muscle pain11.7%7.4%
Joint pain5.2%3.6%

Generally speaking, the rate of local side effects was slightly higher in the 5 to 11 group than was seen with older children in other studies. However, the rates of systemic side effects were the opposite (i.e. lower than what was seen with older children in other studies).

These side effects were predominantly mild to moderate, and they generally only lasted 1 to 2 days.

Unsolicited side effects

Other side effects that were reported infrequently were swollen lymph nodes (less than 1%) and allergic-type reactions, such as rashes (less than 1%).

While some serious adverse events were reported in the study, none were related to the use of the vaccine. For example, one of the serious adverse events was a child who inadvertently swallowed a penny (I’d say that it’s safe to assume the vaccine didn’t cause that).

Was there any heart inflammation reported in the study?

There were zero cases of heart inflammation reported in the study. However, it’s important to note that with the rate observed in older groups, the size of this study is unlikely to show any cases.

It’s expected that the rate of heart inflammation will be lower in this age group, the reasoning behind this hypothesis is:

  • Children 12 to 15 years old have less risk of heart inflammation than the 16-19 age group; so it is presumed that younger children will also be less at risk
  • Hormones may impact the risk (so would be expected to be lower in this age group)
  • 1/3 of the dose is used in this age group, which should reduce the risk even more

It’s also essential to remember, as I’ve already mentioned, the risk of heart inflammation is much higher with COVID-19 infection than it is with vaccines.

How long were they followed after receiving the vaccines?

The initial group of 1518 children who received at least 1 dose of the vaccine was followed for, on average, between 2 and 3 months after their last dose (95% of them had between 2 and 3 months of follow-up data).

In the additional safety group of 1591 children who received at least 1 dose of the vaccine, the majority (71%) had at least 2 weeks of follow-up data after their last dose.

As a reminder, side effects from any type of vaccine would generally not be expected beyond 6-8 weeks after the last dose.


Was the vaccine effective in the 5-to-11-year-old age group?

The next step in the analysis of the study is to evaluate if the vaccine helped reduce the risk of COVID-19 infection.

Initial immune response data

A small random subset of this study was first compared with a similar randomly chosen group from the study in 16-to-25-year-olds.

When comparing these two groups, no differences were observed in the immune response to the vaccine.

They also did an additional analysis of the immune response against the Delta variant, and saw that the vaccine produced a good immune response that should be capable of fighting the Delta variant.

Actual clinical data

While the immune response data above is great, what we really want to see is actual results in the children being studied.

During the trial, a total of 19 confirmed cases of COVID-19 were observed:

Vaccine group Placebo group
3 cases

(out of a group of 1450 children who received both doses)
16 cases

(out of a group of 736 children who received both doses)

These results translate to a Vaccine Efficacy (VE) estimate of 90.7%

For those who are more visual, here is a graph representing the rates of COVID-19 in both the vaccine group (blue line) and the placebo group (red line):

Vertical axis represents COVID case incidence and the horizontal axis represents days since 1st dose
(Source: Pfizer submission to the FDA)

Most cases occurred in July/August 2021 (which is when the Delta variant was starting the 4th wave in the United States). In my opinion, this also helps to reinforce that the vaccine should work very well against the Delta variant.

Why were COVID-19 infection rates so low in the study?

Some will note the small numbers of cases in the studied population. There are probably multiple reasons for this, but it’s all assumptions.

One main factor is possibly that these children would be the most likely to have parents that are going to be extra cautious (since parents who don’t take COVID-19 as seriously would be less likely to consent to enrolling their children in this study).


What does the risk/benefit analysis look like for the 5-to-11-year-old age group?

On top of the analysis of the study results, the FDA briefing document also shows a transparent review of the risks and benefits of the vaccine in the 5-11 age group.

In this review, they took into account many different factors:

  • Cases, hospitalizations, ICU admissions, and deaths caused by COVID-19 infection in this age group
  • Cases, hospitalizations, ICU admissions, and deaths related to heart inflammation secondary to the vaccine
    • The database they used for the rate also included unconfirmed cases of heart inflammation secondary to vaccines (potentially overestimating the risk)
  • They assumed a vaccine effectiveness of 70% against all infection types and 80% against hospitalizations
  • They also looked at different transmission rates in the United States
    • September 11, 2021
    • Peak of the fourth wave in the US (August 2021)
    • Lowest level of transmission in the US (June 2021)

For all scenarios except the very low transmission scenario, the benefits were overwhelmingly greater than the risks. In the other scenario (the one in which transmission rates would be at lower levels than they currently are), the potential benefits seemed to not outweigh the potential risks. However, there are a few potential problems with this analysis:

  1. It’s not clear if they also took into account the risks of heart inflammation or multisystem inflammatory syndrome (MIS-C) from COVID-19 infection in children
    • If not, this would underestimate the potential benefits
  2. The database where they got their risk of death in children shows a 3x lower risk of death than other databases
    • This could also underestimate the potential benefits
  3. Hospitalizations for heart inflammation secondary to COVID-19 vaccines are not always equivalent to a hospitalization for COVID-19 infection. Sometimes, hospitalizations for vaccine induced heart inflammation are “precautionary” and generally much shorter/less severe.
    • So it may not always be appropriate to compare the two
  4. As I already mentioned, the effectiveness listed in their analysis seems low
    • This could also underestimate the potential benefits
  5. This also doesn’t take into account the lower dose being used (1/3 of the dose used in those 12 and over)
    • The current hypothesis is that side effects would be reduced.

All this being said, I understand that the FDA has to be transparent with this. I just would have preferred a slightly better approach. But, even with all these potential concerns, all but one of the scenarios showed that the benefits of the vaccines clearly outweighed the risks.

Plus, it’s also important to realize that these scenarios are for a finite length of time; so they don’t take into account the reality that it’s very unlikely that we will eradicate SARS-CoV-2 (virus that causes COVID-19).


The Bottom Line

The data submitted by Pfizer to the FDA shows that the benefits outweigh the risks when it comes to COVID-19 vaccination in children 5 to 11 years old.

As you know, the FDA eventually approved the request for Emergency Use Authorization for the Pfizer vaccine in children 5 to 11 years old.


Main references:

Pfizer. Vaccines and Related Biological Products Advisory Committee October 26, 2021 Meeting Documents

FDA Briefing Document: EUA amendment request for Pfizer-BioNTech COVID-19 Vaccine for use in children 5 through 11 years of age

*Other references linked throughout*

Dan Landry

Daniel (Dan) Landry, founder of Rxplanation.com, is an infectious diseases pharmacist at the Dr-Georges-L.-Dumont University Hospital Centre in Moncton, NB, Canada.

Recent Posts