RAPS is closely monitoring developments in the Coronavirus (COVID-19) outbreak. See our public safety page for the latest updates.

Regulatory Focus™ > News Articles > 2020 > 10 > FDA’s COVID-19 vaccine adcomm raises questions as first readouts loom

FDA’s COVID-19 vaccine adcomm raises questions as first readouts loom

Posted 23 October 2020 | By Michael Mezher 

FDA’s COVID-19 vaccine adcomm raises questions as first readouts loom

With several COVID-19 vaccine candidates nearing Phase 3 readouts and the prospect that one or more could receive emergency use authorization (EUA) before the end of the year, advisors to the US Food and Drug Administration (FDA) met on Thursday to deliberate the standards the agency has set for authorization or licensure of a vaccine.
 
In a nine-hour public meeting meant to provide transparency and build public confidence in the process surrounding COVID-19 vaccine development and review, members of FDA’s Vaccines and Related Biological Products Advisory Committee (VRBPAC) were tasked with reviewing FDA’s guidance, considering ethical issues surrounding the unblinding of Phase 3 clinical trials if a vaccine is authorized under an EUA and recommending studies needed to evaluate vaccines for safety and efficacy in the postmarket setting.
 
The much-awaited meeting, which was scheduled in August, follows the issuance of FDA’s guidance on COVID-19 vaccine development and licensure and the recent tussle between the agency and the White House over stricter guidelines for issuing an EUA for a vaccine. (RELATED: FDA issues COVID-19 vaccine guidance, setting 50% effectiveness threshold, Regulatory Focus 30 June 2020; FDA issues COVID-19 vaccine EUA guidance after clash with White House, Regulatory Focus 6 October 2020).
 
EUA vs BLA
 
Two FDA officials gave presentations and took questions on the differences in FDA’s expectations for chemistry, manufacturing and controls (CMC) and clinical data for an EUA or licensure.
 
For an EUA or a biologics license application (BLA), Jerry Weir, director of the Division of Viral Products within the Office of Vaccines Research and Review (OVRR), said that, “Any CMC data that will not be available at the time of licensure, or at the time of EUA issuance, must be discussed with FDA in advance, sufficiently justified, and judged to have minimal impact on product quality.”
 
Weir emphasized that, “The CMC information and data that we would expect and that would be needed to support the use of a COVID-19 vaccine under an EUA are generally similar to that needed for licensure.”
 
While BLA issuance requires a preapproval inspection, Weir said that FDA expects to see information about a company’s manufacturing sites and their good manufacturing practice (GMP) compliance status at the time of an EUA submission.
 
“What we are expecting to do is that we will have GMP compliance assessed using site visits and other submitted information to ensure that the products and the manufacturing facilities are GMP compliant,” he said.
 
Weir added that the appropriate quality specifications established for all drug product lots used under EUA and testing results would need to be submitted to the agency at the time of distribution following issuance of an EUA. Regarding lot release data, Weir said, “Even though the FDA regulations for lot release do not apply to investigational drugs, we expect to obtain essentially the same information in other ways.”
 
Doran Fink, deputy director of the Division of Vaccines and Related Products Applications within OVRR, gave an overview of FDA’s expectations for safety and efficacy data from Phase 3 clinical trials. In its guidance, the agency has set a 50% threshold for efficacy with a 30% adjusted lower confidence interval bound, along with a pre-licensure safety database of at least 3,000 participants who received the vaccine. Different follow-up durations are specified for licensure and authorization.
 
Fink said that FDA could potentially be flexible with its recommendation that there be at least five cases of severe COVID-19 in the placebo group of a study when considering issuing an EUA.
 
“In cases where the vaccine efficacy point estimate and lower bound are both exceptionally high and there are no cases in the vaccine group, fewer than five cases may be acceptable,” he said, noting that CBER may require post-licensure studies to address known or potential serious risks identified during review.
 
One of the central issues at play if FDA were to license or authorize a vaccine for COVID-19 is how the availability of a vaccine would impact ongoing blinded trials.
 
Fink stressed that the issuance of an EUA for a COVID-19 vaccine in and of itself should not warrant the unblinding of Phase 3 trials, emphasizing that a product authorized under an EUA is still considered to be investigational.
 
“Given these considerations, a discussion of the conditions and the timing that would make unblinding of an ongoing clinical trial imperative deserves careful thought and attention, as do considerations of the possible mechanisms that could be used to replace loss of such follow-up,” he said, noting that EUA requests should include a strategy to ensure follow-up in ongoing trials and to handle loss of follow-up due to withdrawal of patients.
 
Some vaccine developers, including Pfizer, have argued that it is their ethical responsibility to inform study participants about the availability of a vaccine under EUA. In comments submitted to the public docket for the advisory committee meeting, Peter Honig, senior vice president, global regulatory affairs at Pfizer, wrote that were the company’s vaccine to receive an EUA, “We would propose to amend our ongoing study to allow cross-over of eligible placebo subjects to the active vaccine arm if they wish to do so.” Pfizer, whose BNT162b2 candidate began Phase 2/3 trials in July, has said it could seek an EUA for its vaccine next month.
 
On the other hand, Janssen, whose Phase 3 trial for its JNJ-78436735 candidate trailed Pfizer’s by two months, warned in its comments that unblinding trials “could have serious ramifications.”
 
“Such a scenario could thus jeopardize longitudinal follow-up for both safety and efficacy, which in turn can result in serious concerns about the validity and scientific integrity of the affected studies,” wrote Janssen’s Head of Global Regulatory Affairs Karin Van Baelen.
 
Responding to a question about the potential for an expanded access protocol to provide access to an investigational COVID-19 vaccine instead of an EUA, Fink explained that expanded access falls under FDA’s regulations for investigational new drug regulations and that an expanded access protocol on the same scale as what is being envisioned for an EUA would merit similar considerations for benefit-risk.
 
“Among many other things, those regulations require use of an institutional review board and also obtaining informed consent from recipients … Operationally speaking, an expanded access protocol would add some complexity, and that is why emergency use authorization is being considered primarily as the mechanism for addressing the public health emergency,” Fink said.
 
Committee members’ concerns
 
In the final hours of the meeting, advisory committee members were asked to deliberate FDA’s guidance and planned approach to bringing COVID-19 vaccines to market.
 
In turning the meeting back to the committee members, OVRR Director Marion Gruber said that FDA is interested to know if it struck the right balance between safety and efficacy and expediting access to a vaccine.
 
The committee’s sole consumer representative, Sheldon Toubman, staff attorney at the New Haven Legal Assistance Association, said he is concerned about the use of COVID-19 of any severity as the primary endpoint for most of the Phase 3 trials that are underway. Toubman also asked whether the median two-month follow-up duration for an EUA is sufficient. “Isn’t two months a little short?” he asked, pointing to the World Health Organization’s recommendation for three months follow-up for its emergency use listing procedure.
 
Fink replied that FDA’s intent is, “To strike a balance between getting information on the most clinically significant outcomes of COVID-19 and how a vaccine might be able to prevent those outcomes versus being able to make an impact on the pandemic in as reasonable an amount of time as possible based on good data.”
 
Fink noted that vaccine developers are free to choose what they consider to be the most relevant endpoints for their products, and it is FDA’s job to evaluate whether the vaccines are effective for that specific indication. “Other bodies, such as [the Advisory Committee on Immunization Practices] ACIP determine whether the vaccine should be used in certain situations.”
 
“That being said, when we do make our benefit-risk determination for EUA or for licensure, we do expect to have data to inform whether the vaccine is or may be effective against more severe disease,” Fink added.
 
Later in the meeting, Toubman stressed that he does not think an EUA is the appropriate vehicle for delivering a vaccine more broadly. “Public trust equals success, lack of trust means no success, that seems pretty clear,” he said, “Where that leads me to is a conclusion that EUA probably should not be used here.” Toubman noted that many Americans, rightly or wrongly, believe that a vaccine is being rushed for political reasons.
 
“That is the perception, and so anything that sounds like ‘emergency use authorization’ it sounds like it’s being done rushed and it’s not the full review. So, even if EUA standards were similar to full licensure, it doesn’t sound good to the public and what it sounds like matters,” he said.
 
When asked about demographic requirements in clinical trials by Steven Pergam, medical director for infection prevention at the Seattle Cancer Care Alliance, Fink explained that FDA has never “had requirements for demographic composition of data to support licensure of a vaccine, and I think it would be very difficult to outline such requirements for EUA.”
 
But Fink said that FDA would look carefully at demographic representation and decide to authorize or approve a vaccine for the populations and uses that are supported by the data.
 
Some gaps in demographic data “would not necessarily in and of themselves result in a restriction,” Fink said, noting that FDA “would have to think about whether it makes sense from a scientific basis to be concerned that there is some difference based on differences in demography,” and that age is the most common example where efficacy is not assumed between different groups.
 
Other committee members raised various issues they felt were not adequately addressed by FDA’s guidance.
 
Hayley Altman-Gans, professor of pediatrics at Stanford University Medical Center, said she is concerned that certain safety issues, such as immune thrombotic events, are being overlooked.
 
Kathryn Holmes, professor emerita at the University of Colorado School of Medicine, said she would like to see Phase 3 studies using infection as an endpoint rather than COVID-19 of any severity. “That is the rubric which would prevent spread through the community most effectively, and that is what would protect our elderly as well,” she said.
 
While Fink agreed that sterilizing immunity is the gold standard of protection, he said it is not always achievable. “In our June 2020 guidance we did make a recommendation that prevention of infection should be evaluated, if not as a primary endpoint, then as a secondary endpoint, and that endpoint could be evaluated using either serological methods … or through periodic sampling using virologic methods,” he said.
 
David Wentworth, chief of the Virology Surveillance and Diagnosis Branch, Influenza Division at the Centers for Disease Control and Prevention, questioned FDA’s blanket follow-up duration recommendations and asked whether the duration should be different depending on the specific platform being used.
 
Luigi Notarangelo, chief, Laboratory of Clinical Immunology and Microbiology at the National Institute of Allergy and Infectious Diseases, also took issue with the efficacy endpoints being used.
 
“I have problems with the standardization of efficacy … I do appreciate that it is very important to standardize efficacy across multiple trials, multiple platforms. But the problem is that these efficacy measures that are included in the document, they have two problems. First of all, they really are biased, skewed towards mild disease … The other problem with those efficacy measures is most of them are really subjective,” he said. Michael Kurilla, director of the Division of Clinical Innovation at the National Center for Advancing Translation Sciences at the National Institutes of Health, concurred, saying he has “concerns about the utility of a 50% reduction in symptomatic disease when we don’t really have any evidence that these vaccines are going to induce sterilizing immunity.”
 
In response, Philip Krause, deputy director of OVRR, said, “This also has to be thought about in terms of the frequency of each of these possible endpoints, so if the endpoint of the trials is severe disease, the trials may need to be almost 10 times as big and those trials would be infeasible and we would never get a vaccine. If the endpoints are infection, that can, with some additional work, be a feasible endpoint, but the science is not there to do that right now.”
 
Krause added that, “There simply does not exist an example in vaccinology of vaccines that are effective against mild disease that are not more effective against severe disease. So a 50% effective vaccine against mild disease is very likely to be greater than 50% effective against severe disease.”
 
Acting VRBPAC Chair Arnold Monto, professor of epidemiology at the University of Michigan School of Public Health, agreed, saying, “What we have to do is keep focusing on EUAs versus BLAs—formal licensure—and not really try to talk about sterilizing immunity and other things which are not part of standard vaccine licensure. Most of our vaccines are licensed to prevent laboratory confirmed disease … and we rarely get into looking at a definition of serious disease.”
 
Notarangelo said he has concerns about immunobridging from adults to children and recommended that children not be considered for the vaccine until there is more evidence.
 
Gans, however, disagreed, saying immunobridging “isn’t something we should completely take off the table.”
 
Notarangelo also said he thinks manufacturers should be inspected prior to receiving an EUA. “I don’t really understand the reason why the manufacturing facilities are not inspected. I think that is something that could be done, it could be done even ahead of time, I think it would provide some additional trust into the process,” he said.
 
In terms of safety follow-up, Amanda Cohn, chief medical officer at the CDC’s National Center for Immunizations and Respiratory Diseases, said she is less concerned with adverse events identified in the two-month period required for an EUA than she is with rare, serious adverse events that could occur later. “This is why our safety surveillance post-authorization needs to be so strong and effective so that we do identify potentially more rare adverse events than you would identify in a trial of 30,000 individuals,” Cohn said.
 
Cohn also raised the issue of waning immunity and wondered if the two-month follow-up duration could mask a vaccine that is only temporarily effective. “Very rarely do we look at [vaccine efficacy] VE so shortly after completing the series,” she said.
 
Paul Offit, professor of pediatrics at The Children’s Hospital of Philadelphia, said he is disappointed that FDA is unable to direct vaccine developers to study their vaccines in specific demographic groups.
 
“It is disappointing, I think that given that this is a vaccine that is being paid for by the public, I mean [Biomedical Advanced Research and Development Authority] BARDA is public money, that the FDA can’t direct this vaccine to make sure we are testing it in groups like those who are at greatest risk, if you look at various racial or ethnic backgrounds, health problems or age,” he said.
 
On the issue of unblinding Phase 3 trials, Krause explained that trial protocols generally do not address crossover, “So there has not been any promise to the people in the trial that they will be eligible to receive a vaccine when it becomes available.”
 
Krause also pointed out that, “Placebo recipients otherwise likely wouldn’t be the first in line to get a vaccine. Normally you would think the first in line, even as a vaccine became available, would be those who are at greatest risk … If anything, the average trial recipient might actually be at a lower priority than certain other people who might be in line to get a vaccine.”
 
Notarangelo and Archana Chatterjee, dean of Chicago Medical School, emphasized the importance of continuing blinded studies, with Chatterjee pointing out that it will also take months before the vaccine can be more widely distributed, allowing more time for the trials to accumulate data.
 
Michael Nelson, president of the American Board of Allergy and Immunology, raised several issues that had not been thoroughly discussed in the meeting, including the need to study vaccines in some of the populations that have been excluded from clinical trials and the need to generate data on vaccine coadministration.
 
Wrapping up the meeting, Gruber said what she heard from the committee is that FDA’s guidance and approach to safety and effectiveness, “Are right on the money and there is real buy-in for that.”
 
Gruber noted some of the concerns raised, including “the importance of making sure that minorities are included in clinical studies.” With regard to looking at different endpoints, Gruber said that would be a consideration for the agency “if we have new vaccines entering clinical studies; it may be a little bit difficult for those who are already in Phase 3.”
 
FDA

Regulatory Focus newsletters

All the biggest regulatory news and happenings.

Subscribe