Safety Concerns About the Pfizer Vaccine: The Pre-Licensure Data and Anaphylactoid Reactions in the UK

The short version: Pfizer’s vaccine is currently under review for emergency use authorization in the US (and was recently given the green light from VRBPAC, meaning the FDA will likely approve its EUA shortly), and has been given limited regulatory approval in the UK and Canada. A review of the pre-licensure data paints the portrait of a very effective vaccine that is likely to cause short-lived but undeniably unpleasant side effects, particularly after the second dose. No significant safety concerns were noted. There were also reports that two healthcare workers in the UK who had epinephrine autoinjectors received the vaccine and suffered anaphylactoid reactions. They are recovering well. The UK subsequently declared that anyone with a history of severe allergy avoid this vaccine, which is precautionary and does not make very much sense scientifically. The likely culprit in my view is PEG, a component of the lipid nanoparticle. It is expected that as the vaccine rolls out and more people receive it some adverse reactions will occur. However, this does not mean that YOU will have a reaction or are likely to have one, and in isolation, this does not mean the vaccine is unsafe. Surveillance will continue, but thus far, I cannot find anything concerning about the safety of this vaccine and have every intention of taking it the first moment it is offered to me.

Note: There is a separate post forthcoming discussing the scientific aspects of anaphylaxis.

The Pfizer vaccine for COVID-19 was recently granted the equivalent of emergency use authorization in both the United Kingdom and Canada, which is a massive triumph in fight against the pandemic. This means that regulatory bodies in both nations carefully examined the data and deemed that there was satisfactory evidence to support the safety and efficacy of the vaccines to justify their approval under pre-specified terms. In the case of The UK, the vaccine

does not have a UK marketing authorisation but has been given authorisation for temporary supply by the UK Department of Health and Social Care and the Medicines and Healthcare products Regulatory Agency for active immunization to prevent COVID-19 disease caused by SARS-CoV-2 virus in individuals aged 16 years of age and over.

In the case of Canada, Health Canada states

Pfizer-BioNTech COVID-19 Vaccine was authorized for use in relation to the COVID-19 pandemic, in accordance with section 5 of the Interim Order Respecting the Importation, Sale and Advertising of Drugs for Use in Relation to COVID-19.

Both are the equivalent of the US’s emergency use authorization (EUA). It should be noted that some sources are reporting that the vaccine has been “approved” by these regulatory agencies; however, this is term incorrect. The conventional use of the term “approval” refers to a formal granting of a market authorization under the traditional Biologics License Application (BLA) submission pathway. Under this traditional pathway, the manufacturer submits a data package for the completed development program for a approximately 1 year review by the health agency. Canada, the UK and the US, the health authority has permitted the manufacturer to submit the data as it becomes available under a process called a “rolling review.” Additionally, Health Canada created a temporary submission pathway under the aforementioned interim order to provide manufacturers with a similar EUA regulatory pathway as the UK and US.

Since the issuance of the UK and the Canadian emergency authorization, new information has come to light. Unfortunately some people have seized upon this new information with the intent to undermine vaccine confidence; an unfortunate situation with the potential for serious harms to public health.

The FDA has published the documents submitted to the them and VRBPAC by Pfizer in documents here (an FDA summary) and here (what Pfizer submitted). For the purpose of this review I will focus on the primary safety outcomes, though some other sources have done an excellent job going through and summarizing the findings (Your Local Epidemiologist, the New York Times).

Professor F. Perry Wilson MD MSCE also went through the FDA document live in an excellent video here:

Before I begin, I do want to make one comment regarding the efficacy data. My initial impression of the efficacy outcomes following a single dose (see Pfizer- Figure 13) was that it may be possible to get away with just one dose, which would have a number of advantages:

There are some issues with this thinking, however.

Efficacy for a single dose was measured to be just 52.4% with confidence intervals reaching into the lower bound of 30%, meaning that this does NOT meet the efficacy endpoints set by the FDA (though there are significant limitations to the analysis here because of the very short period between the first and second dose). The interpretation from this immunologically is most likely that the plasmablasts induced by the mRNA vaccine after just one dose die off too quickly, which is likely related to the fact that the antigen in them is not replicating, hence the need for a 2-dose series (I expect if there is a next-generation self-amplifying mRNA vaccine, we may be able to have a one-dose series, but this is conjecture only on my part, and I do not have hard data to support this).

Professor Akiko Iwasaki has a great thread on twitter discussing this here, and Professor F. Perry Wilson also ran the numbers in a thread here.

I think for now the most prudent decision on the basis of these data is a 2-dose regimen of the vaccine while an additional clinical trial is initiated to examine the protection conferred by a single dose.

In the future, including a single-dose arm in the trial design to evaluate the efficacy of a prime-boost vaccine (the first dose primes the immune system, but the second one is responsible for conferring protection) may be a valuable addition to pandemic trial guidelines. This is a broad suggestion that would require a more thorough evaluation of the clinical, ethical and regulatory implications as one would need to carefully consider the impact on various stakeholders (industry, patients, the regulatory agencies, pharmacists, physicians, and healthcare professionals).

For my safety analysis I will focus primarily on the longer (92-page) document from Pfizer (note that a more succinct, but still not-very-brief 53-page FDA reviewer report is also available). Safety evaluations for the vaccine were determined on the primarily basis of approximately 38,000 participants randomized 1:1 to receive either vaccine or placebo with a median of 2 months of follow-up after Dose 2, and an additional 6000 individuals who did not yet complete both doses prior to the cut-off date of November 14, 2020. This sample size should be sufficient to detect, using rough approximations, 1-in-10,000 adverse events, but very rare adverse events are typically found in post-marketing surveillance (post-marketing surveillance of vaccines is the subject of a future post and I will link it here once it is complete). I would also argue that 2 months is an appropriate interval given the totality of information we have regarding the nature of COVID-19 and vaccines as a pharmaceutical. Specifically, concerns about ADE appear to be unfounded and thus waiting for antibody titers to wane further before petitioning for approval in the setting of a pandemic that is still very much out of control seems inappropriate. Additionally, the lifetime of the mRNA itself within the body is approximately a few hours. We should expect that as more people receive the vaccine, rarer adverse events will occur, some of which may be related to the vaccine and some of which may not be (more on this shortly).

Reactogenicity is a term that refers to the propensity for a vaccine to cause common unpleasant side effects, including things like fever, headache, joint pains, etc. It should be made very clear: this vaccine is reactogenic, especially after the second dose, but this DOES NOT mean it is unsafe. Adverse events were more likely to occur in those under 55 years of age than those older, suggesting that they are directly related to the vaccinee’s ability to initiate a robust immune response. In general, despite the reactogenicity, it was rare for individuals to withdraw from the clinical trial.

Results from nonclinical toxicology studies showed that animals did not demonstrate repeat-dose toxicity and all developed robust immune responses to the antigen. In vivo immunogenicity and challenge in rhesus macaques demonstrated protection from disease. Developmental and reproductive toxicity studies in Wistar rats are still under way, per the document.

I’ll skip to the adverse events of interest to most people.

Though there are noticeably more adverse events in the vaccine group than the placebo group, this is what we would expect for a reactogenic vaccine. Here are the important parts: there were similar amounts of life-threatening adverse events in the vaccine and placebo group (actually fewer in the vaccine group) which is reassuring. Though there are many more severe adverse events, I remind people that a severe adverse event refers to those that are Grade 3 per the Guidance for Industry Toxicity Grading Scale for Healthy Adult and Adolescent Volunteers Enrolled in Preventive Vaccine Clinical Trials from the Center for Biologics Evaluation and Research of the FDA which can broadly be summarized as severe enough to interfere with daily activity- which is not the same as dangerous. The document notes that the most frequent unsolicited AEs in the vaccine group were injection site pain (2108 [11.2%]), pyrexia (fever) (1144 [6.1%]), chills (998 [5.3%]), fatigue (1026 [5.5%]), headache (966 [5.1%]), and myalgia (muscle aches) (904 [4.8%]). The solicited AEs, in descending order of frequency, were injection site reactions (84.1%), fatigue (62.9%), headache (55.1%), muscle pain (38.3%), chills (31.9%), joint pain (23.6%), fever (14.2%).

In the course of the study, 6 participants died. 4 were in the placebo group, and 2 were in the vaccine group. The deaths in the vaccine group occurred 63 days after the second dose of the vaccine due to a cardiac arrest 60 days after the second dose (I am going to safely call this one unrelated) and the other is described as being due to arteriosclerosis 3 days after the vaccine’s first dose. The FDA reviewer report has declared both of these to be unrelated to the vaccine. Details on 2 of the 4 deaths in the placebo group indicate they were from a stroke and a heart attack, which is consistent with expectations given the demographics of the trial.

Serious adverse events (SAE; those that either result in death, are life-threatening, require prolonged hospitalization, result in a congenital anomaly, result in prolonged disability or permanent damage) were similar in incidence overall between placebo and vaccine groups. The Pfizer document goes on to state:

Three of the SAEs in the BNT162b2 group and none in the placebo group were assessed by the investigator as related to study intervention: 1 SAE each of shoulder injury related to vaccine administration, ventricular arrhythmia, and lymphadenopathy.

Lymphadenopathy is a swelling of the lymph nodes and is entirely plausible as a consequence of the immune response from the vaccine. Shoulder injury related to vaccine administration (SIRVA) occurs when the vaccine is injected in the bursa of the shoulder or joint space, and is rare (which is also true in this study as of nearly 19,000 vaccine recipients there was a single case), and it causes painful inflammation. SIRVA accounts for the bulk of petitions for vaccine injury compensation, and insofar as vaccines are given via injection, it is not unlikely that vaccines would be the cause of the condition. However, this is readily mitigated through careful consideration of the anatomy of the vaccinee. Regarding the ventricular arrhythmia, the FDA state:

Three SAEs reported in the BNT162 group were considered by the investigator as related to vaccine or vaccine administration: shoulder injury, ventricular arrhythmia, and lymphadenopathy. The investigator and the sponsor thought that the shoulder injury was related to vaccine administration. Two SAEs in the BNT162b2 group and none in the placebo group were considered by the investigator, but not the Sponsor, as related to study vaccination: shoulder injury (n=1), ventricular arrhythmia in a participant with known cardiac conditions (n=1), and lymphadenopathy temporally following vaccination (n=1). In FDA’s opinion following review of the adverse event narratives, two of these events were considered as possibly related to vaccine: shoulder injury possibly related to vaccine administration or to the vaccine itself, and lymphadenopathy involving the axilla contralateral to the vaccine injection site. For lymphadenopathy, the event was temporally associated and biologically plausible.

In other words, I’m not terribly concerned about this vaccine giving anyone a ventricular arrhythmia (literally how would it even do that?).

Lastly, there were 12 cases of appendicitis throughout the trial, 8 in the vaccine group and 4 in the placebo group. These were judged by study investigators to be unrelated to the vaccine, and a statistical analysis further demonstrates that the incidence of appendicitis in the vaccine group is consistent with background rates. The appendix does have some lymphoid tissue and you can occasionally see inflammation provoke appendicitis, but I am skeptical that an IM injection of a vaccine in the arm could provoke appendicitis. At the end of the day though, post-marketing surveillance will make a final judgment, but thus far I do not see this as cause for concern.

So in summary, the Pfizer vaccine’s pre-licensure data demonstrate that the vaccine has a favorable safety profile with predictable adverse events generally as would be expected of a vaccine, and a few unusual events that will need to continue being watched in post-marketing surveillance but nonetheless represent very rare occurrences. The vaccine is reactogenic, but from this analysis, I do not see any cause for true safety concerns (though it would be worthwhile to get the vaccine before some time where you would be able to rest, especially the second dose). The Pfizer document does note that the reactogenic effects tended to be short-lived, but do not really define this further (I would expect no more than 3 days, but I don’t have hard data to base that on; regardless if you get the vaccine and have concerns you should contact your care provider).

That brings us to the response to the case of 2 allergic reactions from the vaccine that occurred in the UK in individuals with a history of severe allergy. A press release from CNN reports that these were both anaphylactoid reactions. This precipitated claims of nefariousness on the part of Pfizer and BioNTech regarding the vaccine, which has been… frustrating. This comes from a claim that individuals with allergies generally were excluded from the clinical trial process, which is a bastardization of the truth. The exclusion criterion in question reads:

History of severe adverse reaction associated with a vaccine and/or severe allergic reaction (eg, anaphylaxis) to any component of the study intervention(s).

In other words, specifically a history of severe allergic reaction to a vaccine, or any component of THIS vaccine (a list of which can be found here in section 6.1; the mRNA itself should not be allergenic). Anaphylaxis to a component of a vaccine is a standard contraindication for ANY vaccine (though anaphylaxis to one vaccine does not guarantee anaphylaxis to another). The CNN article further writes:

“As is common with new vaccines the MHRA [Medicines and Healthcare products Regulatory Agency] have advised on a precautionary basis that people with a significant history of allergic reactions do not receive this vaccination after two people with a history of significant allergic reactions responded adversely yesterday,” said Stephen Powis, the national medical director for NHS England, in a statement. “Both are recovering well.”

This move is very frustrating for several reasons. Dr. Paul Offit put it well I think:

“Certainly, vaccines can cause severe allergic reactions. In the United States, roughly one of every 1.4 million doses of vaccines is complicated by a severe allergic reaction.”

He said that rather than a “blanket recommendation” for people with allergies, “the smarter thing to do would be to try and look at these two patients and see what specific component of the vaccine they were allergic to.”

Offit said people should realize that there are immediate treatments for allergic reaction. “That’s why you’re hanging out in the doctor’s office,” he said, before warning that the reports of allergic reactions “will only serve as yet another way to scare people.”

Allergic reactions have a well-defined protocol for management, especially anaphylactic reactions. This offers a good opportunity to delve into some myths about anaphylaxis.

Firstly, though it is always a medical emergency, fatality from anaphylaxis is very rare (estimated at less than 0.001%), despite a significant number of cases not being treated properly. Do not misconstrue my comment as an attempt to trivialize. Many people are taking the anaphylactoid reaction that these two individuals very unfortunately experienced to be evidence that the vaccine is dangerous. However, it’s important to bear in mind that this represents a very rare event (even if we exclude all other vaccines- and there is some reason to expect that the incidence of allergy from this vaccine would be lower than those in routine use, which I will explain shortly- there were no allergic reactions reported in the entirety of Pfizer’s vaccine group). It should be made clear that people with a history of allergies were NOT excluded from enrolling in the trial, unless those allergies were specific to a component of Pfizer’s vaccine or another vaccine). The chance of death for either of these two individuals was a priori very low from the vaccine- much less than their chances from COVID-19. To be clear, I am not advocating that people who are known to have a severe allergy to a component of the vaccine, get the vaccine (though there are circumstances in which that might be appropriate).

The next point I want to raise is that in general, it’s possible that allergy to Pfizer’s vaccine is less likely than for many routinely used vaccines. This is because Pfizer’s vaccine is not made using common allergens that may be present in other vaccines e.g. milk proteins in DTaP, gelatin (present as as stabilizer in some vaccines), egg proteins (really only relevant for the yellow fever vaccine). There are NO common allergens in the Pfizer vaccine, as shown in that excipient list. Looking through the excipients list, there is one candidate that seems probable as a cause of the allergy: PEG(polyethylene glycol)-2000. PEG linkers are really common pharmaceutical excipients and can be the cause of immediate hypersensitivity reactions, like anaphylaxis, though this is generally considered rare. I expect that the investigation will implicate one of these as the cause, but I do not know in advance and I will await the results. In addition, at the VRBPAC meeting, some additional details were given about the patients in question (7:07:30): these patients were aged 49 and 40, and both had a history of anaphylaxis, one to foods and one to drugs. Further investigation is needed and the MHRA has initiated an investigation under the Yellow Card scheme, and Resus guidelines recommend specialist referral to determine the cause of the anaphylaxis, so I expect we will have clarification soon.

There is still no evidence that the Pfizer/BioNTech mRNA vaccine against COVID-19 is inherently unsafe. It is not risk-free; no intervention is, and as we go forward with the vaccination program, more adverse events will occur, some of which may be serious, some of which may be related to the vaccine (and some which will not be). The vaccine has passed through clinical trials and has been granted emergency authorization by regulatory agencies for use during an unprecedented crisis. It has demonstrated good safety, despite being reactogenic. This occurrence should not be cause for alarm, and it should not discourage people from getting their COVID-19 vaccine. I would still happily roll up my sleeve for Pfizer/BioNTech’s vaccine.


I would like to credit and extend my sincerest gratitude to my dear friend, Andrea Bailey PhD, Senior Regulatory Affairs Specialist, for once more using her prodigious expertise of pharmaceutical regulation to ensure the accuracy of this article and her thorough editing.




Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Edward Nirenberg

I write about vaccines here. You can find me on Twitter @enirenberg and at (where I publish the same content without a paywall)