The variables shaping FDA’s guidance for new SARS-CoV-2 strains
The flu is only a starting point, COVID-19 brings different factors into play
The flu is only a starting point. COVID-19 brings different factors into play.
While FDA has made it clear that efficiency is a priority in adapting authorized COVID-19 products to new viral variants, questions remain about what circumstances would require a new vaccine launch, the surrogate markers it would rely on, and the epidemiological data needed to guide design of new products and trials.
The decades-long practice of serially updating influenza vaccines provides a useful template. But COVID-19 vaccines face some different headwinds and tailwinds — while there is less understanding of the virus and a weaker surveillance network, the pandemic response is deploying newer vaccine modalities that are quicker and easier to change and scale.
On Thursday, Acting FDA Commissioner Janet Woodcock said the agency will publish guidance documents “in the next few weeks” for ongoing development of authorized COVID-19 vaccines, therapeutics and diagnostics to address new viral variants.
In her statement and in the briefing that followed it, Woodcock formalized the message FDA has been telegraphing for weeks: trials for authorized products modified to address escape variants will be smaller and more streamlined than the pivotal studies that supported their initial emergency use authorization (EUA).
“We do not believe there will be a need to start at square one with any of these products,” Woodcock said in the statement. “We do not want to create obstacles to getting these tools to the frontlines.”
Woodcock said the need for product updates will hinge on two factors: the prevalence of the concerning variants, and the degree to which they compromise the original product’s efficacy. But she did not specify a threshold for either parameter.
The agency thinks escape variants haven’t yet crossed either of those lines in the U.S. But that picture could soon change as more data on variant-focused efficacy come online, and as the U.S. scrambles to build up its pathogen sequencing surveillance infrastructure, which has lagged behind other countries such as the U.K.
Vaccines face a far bigger challenge than therapeutics and diagnostics because they require massive studies to determine efficacy directly, and indirect correlates of protection are still under investigation.
“We do not believe there will be a need to start at square one with any of these products.”
Woodcock said the criteria for assessing the performance of vaccine updates will be based on “what we know at that point in time about immune correlates of protection.” Last month, former scientific adviser and co-leader of Operation Warp Speed Moncef Slaoui told BioCentury that correlates of protection validated by NIH and FDA could be available as soon as mid-February.
Yet it’s possible that new variants’ interactions with the immune system could differ enough from the original strain that correlates of protection established for the latter no longer apply.
Other challenges include the need to align product profiles and trial designs with shifting variables such as whether a new variant is co-existing with or overtaking the original strain, and what fraction of people are already exposed or vaccinated, versus immunologically naïve.
Woodcock said FDA has been working closely with manufacturers on the threat posed by new variants, and that the guidance recommendations are unlikely to come as surprise. Publishing the guidances will will get FDA’s thinking down on paper, provide uniform sets of advice, and create the opportunity for public comment to refine the agency’s thinking, she said.
FDA also plans to publish the scenario planning analyses it uses to anticipate how products and supply chains could be impacted by different courses the pandemic could take.
Woodcock thinks it would be “very wise,” to hold advisory committee meetings on reviews of products updated to address new variants, with the goal of “maximal public transparency.”
“We can’t afford confusion by the public,” she said.
Correlates are coming
Woodcock stopped short of unveiling specifics from the guidances, but regulatory experts who spoke with BioCentury believe the guide for vaccine developers could call for a combination of immunogenicity and real-world data.
Former FDA Commissioner Mark McClellan, who is director of the Duke-Margolis Center for Health Policy, said the first step to understanding whether product updates are necessary is determining whether the sera of people vaccinated with authorized vaccines maintain neutralization potency against new variants.
Such neutralization data has already begun to emerge for vaccines from Moderna Inc. (NASDAQ:MRNA) and Pfizer Inc. (NYSE:PFE)/BioNTech SE (NASDAQ:BNTX). Johnson & Johnson (NYSE:JNJ) and Novavax Inc. (NASDAQ:NVAX), which on Thursday each announced regulatory submissions for their Phase III candidates, have shared protection data for some new variants, but have not yet published neutralization assays.
The picture so far is that authorized and late-stage vaccines are moderately limited in their ability to protect against emerging variants, particularly the B.1.351 (501Y.V2) strain first detected in South Africa, but these vaccines are still likely to be effective at preventing severe disease, hospitalization and death.
McClellan thinks the large amount of safety data collected in sponsors’ pivotal trials provide confidence that there will not be unexpected adverse events with updated products.
He said there is also “really good data” on what markers of response are associated with protection from the original strain, and thinks these soon-anticipated correlates could be used to establish efficacy for a modified vaccine via a “several hundred person trial.”
If reliable immune correlates emerge, they could be used analogously to surrogate endpoints for accelerated approval, which are considered “reasonably likely” to predict clinical benefit, and require follow-up with confirmatory studies.
But modifying authorized vaccines to address new variants might just require showing that the new products induce similar immune signatures to the originals, which is a lower bar to clear than finding markers that correlate with protection, said Karen Midthun, principal of drug and biological products at FDA regulatory consulting firm Greenleaf Health and former director of FDA’s Center for Biologics Evaluation and Research (CBER).
“The less impactful and prevalent the new variants are, the more FDA may want to do longer studies.”
She said some vaccine categories, such as hepatitis B vaccines, have established threshold antibody levels known to be required for protection, which can be used as the basis for full approval. Others like pertussis vaccines have no established correlates of protection, but the immune signatures they induce have been used in bridging studies, for example to expand a vaccine’s label to include new target populations.
McClellan thinks there will be a post-EUA surveillance process that follows vaccinated individuals in real-world settings to determine whether they remain protected from serious infections.
He believes the level of evidence required to authorize modified vaccines will be calibrated based on the urgency of the situation.
“The less impactful and prevalent the new variants are, the more FDA may want to do longer studies,” he said. “We may have more time to watch how the variants are evolving and be able to do more clinical testing on modified versions of the vaccines.”
Influenza vaccines provide a regulatory precedent for the speedy interplay between pathogen surveillance and new vaccine approval, with some key differences.
Each year, the WHO convenes a meeting to discuss pathogen surveillance data from an international network of public health agencies. Analysis from that meeting is discussed by an FDA advisory committee every March, which makes recommendations about what strains should be included in the next set of vaccines.
Manufacturers typically submit a CMC supplement, but no new clinical data. Often they already have a vaccine that matches the new strains on the shelf, and the major bottleneck is the time required to grow the stocks of virus that form the basis for these inactivated vaccines.
In the case of COVID-19, there are no decades of learning, or established surveillance and review networks, to lean on.
But the mRNA and adenoviral vector vaccines leading the pack can be altered and synthesized on the order of weeks, not months, enabling a more rapid testing and manufacturing cycle. While protein-based vaccines such as Novavax’s generally take longer to manufacture than nucleic acid-based ones, the company believes it will be able to speedily produce a next-generation product because of the small amount of protein antigen required by the company’s platform.
Something added, something new
Whether first-generation vaccines will require a variant-targeted booster or a multivalent replacement will depend on epidemiological studies, said Midthun.
“Is the new variant taking over? If so, it might make sense to have a bivalent vaccine that includes both the old and the new variant, or you might move toward a monovalent vaccine based on the new variant,” Midthun said. She added that it’s important to respond to variants beyond those spreading in one’s home country. “You want a vaccine that is going to work globally, not geo-specifically.”
Companies have already begun staking out positions based on the data at hand.
On Jan. 25, Moderna announced it would take a two-pronged approach to addressing new variants: the company will test the protective effects of a third dose of its currently authorized mRNA vaccine, as well as a booster against the South African variant.
And on Friday, CureVac N.V. (NASDAQ:CVAC) and GlaxoSmithKline plc (LSE:GSK; NYSE:GSK) announced a collaboration to develop multivalent vaccines against multiple variants, guided by pathogen surveillance data from the Variant Vaccine Expert Advisory Group of the U.K. Vaccines Task Force.
Another question is how to best layer variant-targeted vaccines on top of existing natural or vaccine-generated immunity, and whether trials should be designed to test modified vaccines in individuals with different exposure or vaccination histories.
Woodcock said some insight into the interaction between natural and vaccine-induced immunity is emerging based on preliminary data from people who were vaccinated and later discovered to have been seropositive.
She said multivalent vaccines or multiple different kinds of boosters may be required to address varying gaps in immunity. “We will at least float those ideas in the guidance. It will require thought and input.”