Avoiding the Subject: On Pfizer Vaccine Quality Control, FDA Says Less than European Counterpart

On December 10, as the Food and Drug Administration (FDA) considered issuing its first emergency authorization for a vaccine to combat the pandemic, the agency asked a committee of outside advisors to vote on this question:

“Based on the totality of scientific evidence available, do the benefits of the Pfizer-BioNTech COVID-19 vaccine outweigh its risks for use in individuals 16 years of age and older?”

The FDA’s wording of the question was read aloud, debated, repeated, and left unchanged, and the official minutes of the meeting record that the vote of 17 to 4 with one abstention was “based on the totality of scientific evidence available.”

The next day, the FDA issued the authorization.

But for the FDA, what was the totality of the evidence?

How much did it know, and what if anything was it leaving unsaid?

Ten days after the FDA authorized the vaccine for use in the United States, the agency’s European Union counterpart greenlighted the shot’s use in Europe. The European Medicines Agency (EMA) set a markedly different example of disclosure on a subject the FDA addressed only minimally in public: quality control in the manufacture of the vaccine.

The European regulator laid out its findings about quality control in a 140-page report. It disclosed concerns that had arisen during the course of its review—some of which at one point had been classified as “Major Objections”—and explained how they had been addressed. It emphasized the limits of what it knew, and it said additional data was needed.

That included information on key characteristics of the vaccine.

“The characterisation and control of active substance and finished product are limited in relation to critical quality attributes and impurities,” the EMA said.

In the realm of biotech, characterization is the process of analyzing something to identify and understand its properties. The term can encompass specifications for a product, methods used to measure whether the product conforms to the specifications, and the documentation of that information in material submitted to a regulator.

One of the issues the European regulator flagged: the possibility that aberrations in the vaccine could lead to “unwanted immunological events.”

The EMA said the risk of that happening was “considered low.”

The European regulator concluded that issues it spotlighted shouldn’t stand in the way.

“The benefits to public health of the immediate availability outweigh the risks inherent in the fact that additional data are still required,” it said.

“Given the emergency situation, it is considered that the identified uncertainties can be addressed post-authorisation,” it said.

In its December 21 assessment report, the European regulator presented information not to be found in the briefing paper the FDA gave members of the FDA’s vaccines advisory committee before its December 10 public meeting, not discussed at the meeting, and not disclosed in the public version of the memorandum in which the FDA explained its assessment of the vaccine.

The EMA’s public assessment described impurities that had been observed, analyzed, and ultimately found acceptable. It discussed inconsistencies between batches, another issue ultimately addressed to the EMA’s satisfaction.

Much of the EMA’s account pertained to the presence in the vaccine of irregular forms of the vaccine’s key component—a genetic material called mRNA—and levels of “RNA integrity.”

The report also noted the detection of something anomalous involving lipids, a fatty substance used to coat the mRNA. It said lipid-related impurities had been identified in recently produced finished product batches. It also said “particulate matter” had been observed in the vaccine.

FDA advisory committee members interviewed by the Project On Government Oversight (POGO) said they were unaware of that background information on the vaccine.

In granting its authorization, the European regulator dictated a plan to address remaining uncertainties over the ensuing months. It imposed what it called “specific obligations” on the manufacturer.

The required follow-ups included, for example, introducing an appropriate strategy for controlling lipid-related impurities, further analyzing RNA integrity, and providing more information on the method used to assess the vaccine’s potency.

“The quality of this medicinal product, submitted in the emergency context of the current (COVID-19) pandemic, is considered to be sufficiently consistent and acceptable subject to the specific obligations abovementioned,” the EMA summarized.

It is unclear whether the FDA was dealing with more information, less information, or the same information as its European counterpart when it granted its emergency use authorization (EUA).

What is clear is that it said less publicly.

It is unclear whether the FDA was dealing with more information, less information, or the same information as its European counterpart when it granted its emergency use authorization.

There are indications that the FDA was in contact with the EMA about quality control issues as early as November and that the two regulatory agencies were sharing information.

The FDA did not clear up the picture for this story. Responding to questions about its transparency—or lack thereof—the FDA was less than transparent.

In a March 12 email to POGO, an FDA spokesperson seemed to say the FDA chose not to discuss those issues during the advisory committee meeting. However, the email could be read different ways.

In a February 22 email to POGO, an FDA spokesperson granted this much: “It is important to understand that FDA routinely engages in discussions with its international regulatory counterparts on scientific issues, and these engagements also occur outside of a declared pandemic.”

The FDA’s authorization of the vaccine “followed an open and transparent review process that included input from independent scientific and public health experts and a thorough evaluation by the agency’s career scientists to ensure this vaccine met FDA’s rigorous, scientific standards for safety, effectiveness, and manufacturing quality needed to support emergency use authorization,” the spokesperson added.

“The potential to undermine public trust”

With this article, POGO is not second-guessing regulators’ conclusion that the benefits of the vaccine outweigh the risks. Nor are we saying that people should refrain from taking the vaccine, or that the European regulator’s disclosures are cause for alarm.

Rather, we are highlighting the contrast between the European Medicines Agency’s public assessment of quality control and the FDA’s relative silence on the subject.

The contrast raises questions about the FDA’s openness—a precious quality at a time when the government is urging people to get vaccinated as part of a national strategy to stem the devastating pandemic.

It also provides another lens through which to view the FDA’s rush to greenlight the vaccine under pressure from then-President Donald Trump.

The obvious answer is they should have disclosed what they knew.

Robert M. Kaplan, an emeritus professor of health services and medicine at UCLA

“The obvious answer is they should have disclosed what they knew. I think that goes without saying,” said Robert M. Kaplan, an emeritus professor of health services and medicine at UCLA who previously served as associate director of the National Institutes of Health and as chief science officer at the federal Agency for Health Care Research and Quality.

“It’s hard for me to imagine the FDA didn’t know about it,” Kaplan said, referring to the information the EMA ultimately spotlighted.

If the FDA concluded that the knowns and unknowns of the quality control issues were nothing to worry about, the agency should have shared that information with its outside advisors, Kaplan said. Then, advisory committee members could have studied it. If they agreed with the FDA, they could have said in an open forum, “We’ve looked at it carefully and we’re not concerned about it,” Kaplan said.

Disclosing information could reassure the public, Kaplan said. However, not disclosing it “has the potential to undermine public trust,” he said.

Kaplan, who has testified at FDA hearings on vaccines for COVID-19 as a member of the public and is now on the faculty of Stanford’s Clinical Excellence Research Center, made those comments in interviews with POGO.

“I would have wanted to have known”

Dr. Patrick S. Moore, who participated in the December 10 advisory committee meeting as a temporary committee member, said that, as a general principle, the FDA should be open with vaccine information.

“You have to be completely transparent with all the information. You have to describe all the potential benefits, adverse consequences, fears, and expectations,” Moore, a professor and medical researcher at the University of Pittsburgh who specializes in microbiology and molecular genetics, said in an interview in early February.

“All the issues have to be legitimately discussed, they have to be evaluated by experts, and the experts have to be uncorrupted,” Moore said. “And they have to try to present the information as best they can to people that will actually be making the decision, ‘Do I want to take this vaccine?’”

“By not giving full presentation of facts,” the FDA will undermine confidence in the vaccination campaign, Moore said.

If the FDA knew about the issues discussed in the EMA report, and with the understanding that they had been addressed to the European regulator’s satisfaction, would there have been any value in the FDA sharing that background with its advisory committee?

By not giving full presentation of facts, the FDA will undermine confidence in the vaccination campaign, Moore said.

“Certainly, I would certainly think that that would be important,” Moore told POGO.

Informing committee members would enable them to give the FDA advice on the subject, he said. If there was reason to think there was a problem, it was “a problem to be solved, not something, not a problem to be hidden,” he said.

But Moore said he was “not terribly concerned about” the information and did not think it would have led the committee to a different conclusion.

“Let me say that from the data that I’ve seen, there is no question that this vaccine is as safe as it can be … and that its benefit many, many, many, many, many-fold outweighs any risk that I’m aware of,” Moore said.

Another person who served on the advisory committee had this reaction to information in the EMA report: “I would have wanted to have known it” before voting. That person spoke to POGO on condition of anonymity.

A third person who served on the committee, Dr. Mark H. Sawyer, said, “If there is some evidence that it should be a concern, then, sure, I’d like to know that.”

But, reacting to excerpts from the European regulator’s report, Sawyer said observations that mRNA was showing up in the vaccine in “truncated” form did not raise concerns in his mind about the safety or efficacy of the vaccine.

“I do have confidence that the FDA is aware of those issues and would have brought them forward if there was any reason for them to raise concern,” said Sawyer, a professor of clinical pediatrics at the UC San Diego School of Medicine and an infectious disease specialist at Rady Children’s Hospital.

Given the limited amount of time in the advisory committee meeting, “we could not possibly cover every single detail related to the production of the vaccine,” said Sawyer, who is also a former member of the Centers for Disease Control and Prevention Advisory Committee on Immunization Practices and a current member of a group appointed to advise the California state government on COVID-19 vaccines.

As for lipid-related impurities, that subject “falls in, to me, in the larger category of uncertainty about mRNA vaccines generally and whether there would turn out to be a safety problem related to them,” Sawyer said in early February.

Sawyer said it was his understanding that “the only way we were going to be able to really determine if that was an issue is to study patients who received the vaccine or individuals who received the vaccine … not from the clinical trials but from the larger population after the EUA, which is an ongoing process.”

The government is tracking safety and efficacy data on millions of members of the general public who have received COVID-19 vaccines.

According to a recent FDA advisory committee briefing, as of February 13, a program through which healthcare organizations provide data had found no statistically significant increased risks for any of about 20 health conditions that program was specifically tracking. The list included conditions such as Bell’s palsy, a form of facial drooping or paralysis; anaphylaxis, a potentially fatal allergic reaction; and Guillain-Barré syndrome, an autoimmune disease in which the body’s immune system attacks part of the nervous system.

Avoiding the Topic

The FDA said it was being transparent.

“Today’s meeting … continues FDA’s commitment to an expedited review process that is transparent, scientifically sound and data driven,” Dr. Doran Fink, deputy director for clinical review in an FDA division that evaluates vaccines, said toward the opening of the December 10 advisory committee meeting.

FDA officials walked the committee through extensive clinical trial data on the safety and effectiveness of the Pfizer-BioNTech vaccine.

But, as the committee considered whether to endorse emergency authorization of the still-experimental product, quality control received only perfunctory discussion. The FDA seemed determined to avoid the topic.

“FDA has reviewed the manufacturing quality and consistency data for the Pfizer-BioNTech vaccine and found it adequate to support emergency use authorization of the vaccine,” Fink, one of the lead FDA briefers, told the committee.

“This will not be discussed in detail further at this meeting,” Fink added.

Later in the meeting, when a committee member asked a question about manufacturing quality, Fink said little more.

“We are not intending to discuss details of the manufacturing process, many of which are proprietary, during today’s meeting. But as Dr. Gruber”—Marion Gruber, director of the FDA office that reviews vaccines—“and I discussed in response to a previous question, our review of the manufacturing information is an ongoing process and will continue even after the vaccine is authorized,” Fink said.

The FDA’s Unresponsive Response

Was the FDA somehow unaware of the data that caught the European Medicines Agency’s attention? Did it know and choose not to say anything publicly about the information? Did it consider the matters resolved, unimportant, or merely less-than-disqualifying by the standards of an emergency authorization?

POGO asked the FDA extensive questions for this story.

On the transparency of its authorization process, the FDA was less than transparent.

Asked in February about matters covered in the EMA report, the FDA didn’t say whether it knew about them when it authorized the vaccine, what it thought about them, whether it had discussed them with Pfizer or BioNTech, whether it had discussed them with the EMA, or why it didn’t say anything about them to the advisory committee or the public.

In a February 22 email to POGO, the FDA pointed to the “decision memo” explaining the basis for its emergency use authorization.

The 57-page FDA memorandum includes a section on manufacturing which features several redactions and, in its public form, says nothing about any quality control concerns or anomalies on the road to authorization.

Much of the wording is obtuse.

The document says information was “requested from the Sponsor”—meaning Pfizer.

“The analytical procedures developed and used for the release and stability monitoring” of the active ingredient and end product “include tests to ensure their identity, purity, quality, and potency,” the memo says. “The assays”—in other words, tests—“are appropriate and acceptable” for quality control, the memo says.

“The summaries of the qualification results demonstrate precision, accuracy, sensitivity, specificity, and reproducibility for each evaluated analytical assay, indicating that they are suitable for the intended use,” the memo says.

Got it?

The manufacturing process “underwent changes during vaccine development,” it says. According to the memo, the changes at least partly involved increasing batch size to meet commercial need. Other information about the changes is blocked out.

A “comprehensive analytical comparability assessment has been performed” using “submitted data,” the memo says. What was being compared is blocked out. A “more comprehensive comparability assessment” involving additional lots from multiple production sites was ongoing, the memo says.

Notations in the text indicate that the FDA categorized the redactions as protecting trade secrets or other confidential commercial information.

The memo also addresses the vaccine’s stability—its shelf-life and susceptibility to deteriorating.

“All available stability data … support the emergency deployment of the Pfizer-BioNTech COVID-19 Vaccine,” the memo says.

In the February 22 email to POGO, an FDA spokesperson who requested anonymity added that the “benefit-risk assessment” of the vaccine “remains favorable.”

Sequel to the FDA’s Unresponsive Response

In March, POGO inquired again as part of an effort to give Fink, who spoke for the FDA at the advisory committee meeting, an opportunity to comment personally. Our first question:

Why did Fink and the FDA decline to discuss manufacturing quality and consistency data in further detail, even when members of the advisory committee asked related questions?

In a written response on March 12, FDA spokesperson Alison Hunt said, “FDA makes the determination of the topics to bring before an advisory committee.”

“As you note, the matter was not discussed during the advisory committee meeting as the agency reviewed the information and found it adequate to support an EUA for the vaccine,” Hunt added, referring to an emergency use authorization.

We also asked why Fink and the FDA didn’t discuss with the advisory committee and the public issues that were later discussed in the EMA report, such as differences between batches, concerns about RNA integrity, the presence of impurities and “particulate” matter, and uncertainties the EMA said could be addressed post-authorization.

In answer to that question, the FDA spokesperson referred us to the preceding answer, saying, “Please see the response to question 1 above.”

In other words, the answer in which she said the FDA reviewed information and found it adequate.

Hunt also pointed to an FDA guidance document from 2008 describing the agency’s “thinking” at the time about information given to advisory committee members. The document said the FDA considered manufacturing quality control information exempt from public disclosure under the Freedom of Information Act. The document said the FDA would instruct advisory committee members not to discuss such information during open meetings.

The 2008 document raises the question: What else has the FDA been withholding from the public about quality control over the years?

If the FDA is following the 2008 document, it appears that, as a standard practice, the agency could be keeping the public in the dark about problems that have the potential to affect public health and safety.

However, far from declaring the subject of manufacturing quality out of bounds for advisory committees or public discussion, the FDA teed it up. It devoted part of an October advisory committee meeting to explaining at a general level how the agency would approach the subject with respect to COVID-19 vaccines.

“Chemistry, manufacturing, and controls (CMC) and facility information and data are critical to ensure the quality of vaccines and the consistency of vaccine manufacture,” the slide deck for the FDA briefing began.

The European regulator did not think its report on the Pfizer-BioNTech vaccine was exposing any trade secrets. It said it had already scrubbed those. The EMA assessment quoted in this story is marked, “as adopted…with all information of a commercially confidential nature deleted.”

Sharing with Regulators

In response to questions for this story, Pfizer spokesperson Steven L. Danehy said the vaccine “is manufactured under the strictest of control measures to ensure consistent identity, quality, purity and potency for optimal patient safety and efficacy in every one of the nearly 100 million doses produced to date.”

“Vaccines, by their nature, have characterization profiles that include some level of variability. Our vaccine is consistently manufactured with the same characterization profile,” the Pfizer spokesperson said. “We proactively and transparently share with regulators to their full satisfaction our learnings in this regard prior to an authorization or approval,” he added.

Like the FDA, Pfizer was less than transparent in its response to questions from POGO.

For example, we noted that, on January 15, Pfizer and BioNTech announced that “certain modifications of production processes” were required and that as a result deliveries from a plant in Belgium would “experience a temporary reduction.” We asked if the modifications had anything to do with the manufacturing quality issues discussed in the December 21 EMA public assessment report or with any other quality control problems.

Pfizer did not directly answer that question.

Backdrop: Under Pressure, a Pattern

The FDA’s unwillingness to discuss quality control in detail is part of a pattern of conduct POGO has spotlighted in past articles.

In October and December meetings about coronavirus vaccines, the FDA glossed over advisory committee members’ warnings, curbed and corralled discussion, and gave the impression that it was going through the motions of consulting outside experts to project the appearance of a thorough and objective process. Some panelists who raised questions and expressed concerns later disappeared from the committee lineup.

All of that unfolded against the backdrop of Trump explicitly pressuring the FDA to greenlight the vaccine and trying to take credit for it.

“We’re just days away from authorization from the FDA, and we’re pushing them hard, at which point we will immediately begin mass distribution,” Trump said on December 8.

On December 9, senior FDA official Peter Marks publicly stated that an FDA decision on whether to authorize the Pfizer-BioNTech vaccine could come “days to a week” after the Vaccines and Related Biological Products Advisory Committee met on December 10 and made its recommendations, S&P Global Market Intelligence reported.

But the authorization was issued faster than that.

On December 11, the day after the advisory committee endorsed the vaccine, Trump directed an early morning tweet at Dr. Stephen Hahn, who was then head of the FDA.

“Get the dam vaccines out NOW, Dr. Hahn @SteveFDA. Stop playing games and start saving lives!!!”

Also on December 11, Trump’s chief of staff told Hahn to be prepared to submit his resignation if the agency did not clear the vaccine by the end of the day, the Washington Post and other news outlets reported, citing unnamed sources.

Hahn reportedly disputed those accounts.

Nonetheless, late on December 11, in a politically and emotionally charged milestone in the nation’s struggle against the pandemic, the FDA issued the authorization.

Now, President Joe Biden and public health authorities are urging people to take the vaccine for their own protection and for the benefit of all. One of the challenges they face is a trust gap—especially among Black Americans.

In a recent survey, researchers at Carnegie Mellon University found that 29% of Black adults had not received a vaccine for COVID-19 and were hesitant to get one, compared to 20% of white adults.

“We all know there is some history—there’s some hesitancy about taking this vaccine. We all know there’s a history in this country of having subjected certain communities to terrible medical abuses in the past,” Biden said on February 19 at a Pfizer factory in Michigan. “But if there’s one message to cut through to everyone in this country, it’s this: The vaccines are safe. Please, for yourself, your family, your community, this country, take the vaccine when it’s your turn and available. That’s how to beat this pandemic.”


At the hearing the FDA held in October, medical professionals urged the agency to be open with the public as it reviewed coronavirus vaccines.

Speakers warned that any missteps in the FDA’s handling of the vaccines or any unpleasant surprises down the road risked jeopardizing trust in vaccines generally and in the government’s ongoing, overall efforts to rally the public in response to the coronavirus emergency.

“More than ever we really need to ensure that the vaccine process is transparent,” Dr. Annabelle de St. Maurice, a professor of pediatric infectious disease at UCLA who has published on vaccine hesitancy and worked at the Centers for Disease Control and Prevention, said at the October 22 FDA meeting.

More than ever we really need to ensure that the vaccine process is transparent.

Dr. Annabelle de St. Maurice, a professor of pediatric infectious disease at UCLA

“We need to get this right to maintain vaccine confidence for future generations,” she said.

Researchers also explained how racial and ethnic minority populations’ experiences had left them mistrustful of the government, the health care system, and COVID-19 vaccines. In a presentation to the advisory committee, researchers quoted people who had expressed their views in so-called listening sessions.

“I don’t think the FDA can be trusted to keep people safe,” one person said.

“I would not like you to sell me but show me and tell me, educate me,” another said.

“They want to use us, and I don’t want to keep getting used,” a third said. “We’re not going to be guinea pigs again.”

A Closer Look at the EMA Disclosures

POGO sought information for this story from the European Medicines Agency. Reached by phone, Veronika Jekerle, head of pharmaceutical quality at the EMA, declined to comment and referred POGO to the agency’s press office. The press office did not respond to inquiries.

The EMA’s report described inconsistencies between batches produced for use in the clinical trial using one process and batches produced for potential commercial distribution using a different process. (It refers to them as “Process 1” and “Process 2,” respectively.)

The finding was potentially important because, if the vaccine distributed to the general public differed significantly from the vaccine tested in the clinical trial, the results of the clinical trial might not be predictive of the outcomes the general public would experience.

The pioneering nature of the vaccine made the ability to produce it well on a large scale all the more worthy of public scrutiny.

The Pfizer-BioNTech vaccine uses a new technology based on genetic material known as mRNA, or messenger RNA. The mRNA is given a lipid coating to protect it from breaking down prematurely. Once inside a person’s cells, the mRNA is supposed to produce copies of the virus’s distinctive spike protein and thereby stimulate a specific immune response—resistance to the coronavirus.

The EMA’s December 21 assessment report refers to different forms or “species” of mRNA detected in the vaccine.

“Truncated and modified RNA are present as impurities,” the report says.

Some level of truncated RNA was to be expected, the EMA said. The EMA report shows that the European regulator was considering how much variation would be acceptable.

“Data demonstrates the presence of truncated/modified forms of mRNA at somewhat higher levels in the batches manufactured with the commercial process as compared to material used in clinical trials,” the EMA report says.

“In comparability studies, a decrease in RNA integrity was observed for the initial Process 2 batches compared to Process 1 batches,” it says.

The report explains that the EMA went back and forth with the manufacturer, asking for more information when it found answers lacking.

After adjustments were made, “the percentage of RNA integrity has increased to levels more consistent with Process 1 batches,” the report says.

The EMA’s assessment represented progress from earlier in its review, when, it said, some concerns about manufacturing rose to the level of “Major Objections.” Those concerns included differences between batches.

Yet the report indicates that the EMA’s concerns were not all put to rest.

The truncated and modified forms of mRNA “are not sufficiently characterised,” the report says.

Among the uncertainties: whether the different forms of mRNA would cause the body to produce proteins other than the intended spike protein, and whether to any significant degree those would cause the body’s immune system to react in bad ways.

The amount of any such proteins “is expected to be too low to elicit an immune response of biological relevance,” the report says.

At least in theory.

“Considering the low dose of mRNA … the impurities are not considered a safety issue based on general toxicological principles. However, when present in the cell there is a possibility that aberrant proteins will be expressed with possibilities for unwanted immunological events,” the report says.

“The risk of this occurring is considered low,” it adds.

The report makes similar observations about what it calls lipid-related impurities.

“Lipid related impurities were observed in recently produced finished product batches. Based on the low dose … it is considered that the amounts of these impurities are too low to be of toxicological significance.”

Where the FDA was silent on these points, the EMA decided they were worth mentioning.

The report explains that the EMA authorized the vaccine subject to a list of follow-up quality control studies and other steps for which it spelled out a schedule of deadlines stretching into July.

The report calls those “specific obligations to address the identified quality development issues that may have a potential impact on the safe and effective use of the medicinal product, and which therefore are needed to achieve comprehensive pharmaceutical (quality) data and controls for the product.”

The report says subjects that should be evaluated included any generation of certain proteins that might “cause an autoimmune process.”

In autoimmune diseases, the body’s immune system runs amok and attacks healthy cells.

“To the point of autoimmune responses or autoimmunity problems related to the vaccine, that was and is an issue with all vaccines, and particularly these vaccines because they’re a new platform,” Sawyer, the UC San Diego professor and FDA advisory committee participant, told POGO. “And I expect there was no way that we were going to know whether that was an issue short of the release of vaccine and ongoing safety monitoring to see if that actually occurred.”

With respect to the Pfizer-BioNTech vaccine’s “active substance,” the EMA report says, “Overall, the information provided is satisfactory. However, certain information is still pending due to the very short time frame of product development and will either be updated in the dossier imminently or further addressed in specific obligations and other post-approval measures.”

More than a week after the FDA had authorized the vaccine in the United States, the EMA report said additional analysis of truncated mRNA and other impurities “should be provided to support that they are not impacting clinical performance in terms of safety and/or efficacy.”


Did the FDA have access to the data its European counterpart was analyzing?

It appears that it should have, and there are indications that it did.

The FDA had a responsibility to evaluate quality control. As FDA official Doran Fink said in a December 10 presentation, the agency’s expectations for COVID-19 vaccines receiving emergency use authorizations included “data to demonstrate manufacturing quality and consistency.” Guidance the FDA issued for manufacturers in October indicated that such authorizations were intended for vaccines “for which there is adequate manufacturing information to ensure … quality and consistency.”

As noted above, Pfizer said it “proactively and transparently” shares information with regulators “prior to an authorization or approval.”

Though the European Medicines Agency issued its assessment 10 days after the FDA authorized the vaccine, the timeline of the EMA’s review suggests that the European regulator’s focus on quality control began much earlier.

BioNTech began submitting documentation to the EMA in early October “as part of a rolling review,” the European regulator’s report on the Pfizer-BioNTech vaccine says. A timeline in the report also includes an entry saying that, on November 6, “The applicant submitted documentation as part of a rolling review on quality data to support the marketing authorisation application.”

Reaching a key milestone in the process, BioNTech submitted its application to the European regulator on November 30. That was more than a week before the FDA authorized the vaccine. The application to the European regulator was composed of, among other elements, “complete quality data,” the report said.

Meanwhile, on December 9, the EMA announced that it had been the subject of a cyberattack. The EMA later said hacked documents had been leaked on the internet, including “internal/confidential email correspondence dating from November, relating to evaluation processes for COVID-19 vaccines.”

Subsequently, the French publication Le Monde published an article about what it described as confidential documents that had been stolen from the EMA and posted on the dark web. Le Monde said the documents focused mainly on the evaluation of the Pfizer-BioNTech vaccine. Le Monde said the EMA told it that leaked emails reflected discussions that had taken place.

Le Monde cited EMA emails from November discussing communications between the EMA and the FDA about manufacturing and quality control issues.

In a January 25 update on the cyberattack, the EMA noted that some of the hacked documents had been picked up by media outlets.

“Whilst individual emails are authentic, data from different users were selected and aggregated, screenshots from multiple folders and mailboxes have been created and additional titles were added by the perpetrators in a way which could undermine trust in vaccines,” the EMA said.

POGO received copies of documents that match November messages described by Le Monde. POGO asked the EMA questions about them, but the EMA did not respond.

POGO also asked the FDA about apparently hacked documents. The FDA did not answer those questions.

But, as noted above, an FDA spokesperson said the agency “routinely engages in discussions with its international regulatory counterparts on scientific issues.”

The EMA said something similar in a January news release.

“EMA is in constant dialogue with … other regulators across the network and internationally,” it said.

It would be surprising if the FDA was unaware of issues that threatened to hold up authorization of the vaccine in Europe—issues not aired publicly during the FDA’s review.


Since the Pfizer-BioNTech vaccine was authorized, the manufacturing process and controls have continued to evolve, an EMA document shows.

The document refers to a series of developments, such as “Change in the manufacturing process of the finished or intermediate product” and “Replacement or addition of a manufacturing site.”

In the race to get people vaccinated, Pfizer has been ramping up production.

Fink, deputy director for clinical review in the FDA’s Division of Vaccines and Related Products Applications, said at the December 10 meeting that many manufacturing details are proprietary.

Commercially proprietary information can include, for example, the formula for Coke or the secret recipe for Colonel Sanders’s Kentucky Fried Chicken.

In an interview with POGO, Dr. Aaron S. Kesselheim, a professor at Harvard Medical School who serves on a different FDA advisory committee, said the quality control issues described in the EMA’s December assessment of the Pfizer-BioNTech vaccine don’t seem to fall in that category.

They seem “like something that might be relevant to an advisory committee member or a physician or a patient,” he said.

That doesn’t mean people should be worried about the safety of the vaccine or that the information the EMA reported is cause for concern, said Kesselheim, who directs a program on regulation, therapeutics, and the law at Brigham and Women’s Hospital.

But “providing more insight into the FDA’s inner workings where they don’t conflict with confidential business information … would be useful and helpful” to improve trust in the FDA and the approval process, he said.

The FDA could have taken the advisory committee meeting into closed session to discuss any truly confidential manufacturing information, such as trade secrets. Last fall, when it explained how it planned to review vaccines for COVID-19, the FDA said it might. However, the FDA concluded the daylong December 10 meeting devoted to the Pfizer-BioNTech vaccine without going into closed session.

The FDA also could have given the committee a confidential briefing paper covering information it considered exempt from disclosure under the Freedom of Information Act. But, the FDA told POGO, it did not.

Dodging the Question

The bar for an FDA emergency use authorization is lower than the bar for FDA approval or licensure. But, as a practical matter, in the case of the first COVID-19 vaccine to receive the FDA’s blessing, the emergency action is almost tantamount to approval. It provides all the authority needed to immunize everyone age 16 and up, potentially for the duration of the current public health emergency.

Diana Zuckerman, president of the nonprofit National Center for Health Research, zeroed in on FDA statements that manufacturing quality and consistency data was adequate to support an emergency use authorization.

“My gut reaction is that it’s damning with faint praise to say something is adequate for an EUA, for 2 reasons,” Zuckerman said by email.

“Saying something is adequate is a low bar—shouldn’t the standard be that evidence is persuasive? Or at least ‘good’?” Zuckerman wrote. “The FDA’s standards for an EUA are acknowledged to be quite a bit lower than the standards for approval, so being adequate for an EUA is an especially low bar,” she added.

Authorization isn’t the last word. The system includes subsequent lines of defense for the public.

As with other vaccines, the FDA-approved instructions to people administering the shots call for them be on the lookout for signs of trouble.

“Inspect the vaccine in the vial,” the instructions say. “The vaccine will be an off-white suspension. Do not use if vaccine is discolored or contains particulate matter.”

The instructions count on people administering the vaccine to strike a careful balance. “Gently invert the vial . . . 10 times to mix,” they say. The instructions also say, “Do not shake.” That direction is underlined.

POGO asked the FDA to explain the reason for that instruction and what could happen if the vaccine is shaken before it’s injected. The FDA declined to answer and said to ask Pfizer.

Pfizer’s Danehy said the company included the instruction “to maintain the integrity of the vaccine.” The company wouldn’t say what could result if the vaccine is shaken.

To detect any adverse reactions that show up later, the government is tracking reports of side effects and medical data on people who have received the vaccine.

Further, under the emergency use authorization, Pfizer must perform ongoing quality control tests and submit “Certificates of Analysis” for each lot “at least 48 hours prior to vaccine distribution.”

Whether the FDA has been performing tests of its own is another matter—and a question to which the FDA wouldn’t give POGO a straight answer.

Under a system called “lot release,” the manufacturer of a vaccine or other biological product “generally must submit samples of the product from the lot in question in order to permit the Agency to perform confirmatory testing,” the FDA website says. The FDA can require the manufacturer to wait for a green light from the agency before distributing any lot.

That’s for products that have received FDA licensure—the term for FDA approval when it comes to biological products such as vaccines.

But emergency use authorizations for COVID-19 vaccines “do not require lot release,” an FDA spokesperson said in an email to POGO.

In other words, the FDA could depend on test results submitted by the manufacturer.

For clarity, we requested a “yes” or “no” answer to the follow-up question, “Is it accurate to say that the FDA itself has not tested samples of the vaccines granted emergency use authorizations for COVID-19?”

On this point, too, the FDA was less than transparent.

“In general, FDA testing strategies, including testing frequency of samples are non-public information,” the FDA spokesperson wrote.