The FDA, following the advice of its advisory committee, is advising COVID-19 vaccine manufacturers to reformulate future booster shots to include protection against the BA.4 and BA.5 Omicron variants of SARS-CoV-2. While this decision was not surprising and may have seemed perfunctory to many people, it is worth thinking about the wider context that undergirds this decision.
The primary driver of this recommendation is the fact that the spike protein of SARS-CoV-2 has significantly mutated away from its ancestral form, eroding the protection the vaccine affords against infection. This phenomenon became overwhelmingly obvious when the immune-evasive Omicron variants emerged, causing countless infections, including many in those who were vaccinated or had immunity from prior infection. However, though vaccine protection against infection was compromised by Omicron, the ability of the ancestral strain-directed vaccine to stave off what really matters — severe disease, hospitalization, and death — was durable in all except the high-risk.
This divergence between protection against severe disease and protection against infection has led to a question of what the ultimate policy should be with booster vaccination. Targeted boosting to high-risk individuals — for whom protection against severe disease needs to be strengthened — always made sense and was well-supported by the evidence. However, boosting healthy low-risk individuals with an ancestral strain vaccine to transiently prevent mild illness was, in my opinion, never an optimal policy. Booster uptake amongst the high-risk is suboptimal, and I wonder if the universal boosting message caused the message to the high-risk to be diluted.
But might an updated vaccine, targeted at the latest Omicron strains, be valuable as a booster in higher and lower risk groups? Many questions remain unanswered.
Boosting high-risk people with the original vaccine protects against severe disease, raising the question of what added benefit an updated vaccine provides. For low-risk people, who are unlikely to experience severe disease, is an updated vaccine something that will provide enough protection against infection to be considered worthwhile? Is there any further reduction of long-COVID risk above what is already conferred by vaccination and boosting with the original vaccines?
An important factor to consider is the fluidity of the variant march of SARS-CoV-2. Will BA.4 and BA.5 still be relevant in the fall when updated boosters are available or will they go the way of BA.1 and BA.2? Will whatever variants are prevalent in the fall be Omicron descendants? Will future variants be such that protection directed against BA.4 or BA.5 is cross-protective against them?
In the data presented to the FDA advisory committee, evidence showed that the Omicron-based booster vaccine directed against BA.1 led to about a twofold increase in neutralizing antibody titers against BA.1 versus standard boosting. As Paul Offit, MD, and John Moore, PhD, recently wrote, “What do an approximately twofold higher level of neutralizing antibodies mean in practice? Is an Omicron-based booster sufficiently superior to the standard one to justify a switch?”
There is no human clinical data on boosters directed against BA.4 or BA.5 at this time.
An additional question to consider: What benefit does keeping the original version of the virus in boosters provide? Should boosters be bivalent or monovalent? As Offit and Moore also note in their excellent commentary, Pfizer/BioNTech showed improved immunogenicity with a monovalent Omicron-only booster versus a bivalent one that contained the ancestral strain too. There is also some data that using a protein-based vaccine, like the ones developed by Novavax and Sanofi, as a booster after an mRNA primary series vaccination may be more beneficial than mRNA vaccine boosting.
A related matter to consider is what the threshold is to reformulate the whole vaccine primary series versus just reformulating boosters since the ancestral strain is, for all intents and purposes, extinct. It would be more efficient if manufacturers were making just one type of COVID-19 vaccine.
Given that it appears we will be in a continual arms-race with the virus and its spike protein mutations, the value of developing a more broadly protective or universal coronavirus vaccine is clear. Even if the vaccine was unable to target all coronaviruses, or all beta-coronavirus, or all sarbecoviruses but just SARS-CoV-2 and its variants, it would be an improvement over the current situation. Pfizer/BioNTech and the Walter Reed Army Institute for Research (WRAIR) (among others) have made considerable progress on this front. Additionally, nasally-administered vaccines that mimic natural infection, create more immunity in nasal passageways, and potentially block infection more effectively could also become important tools.
In the end, if frequent boosting of the whole population — not just the high-risk — is the policy, updated vaccines may prove marginally more attractive. However, it is unclear to me what level of clinical benefit against infection to expect when compared to ordinary boosters. As Offit and Moore identify, a costly and consequential decision like this should be informed by as much clinical data as possible. But, unfortunately, there is none in humans for BA.4/BA.5 directed boosters.
As has been the case since the advent of COVID boosters, the discussions have been completely unmoored from a discussion of the larger goals with SARS-CoV-2, a virus that will continue to evolve new variants and will likely never be eradicated. Discouragingly, obtaining answers to some of the key questions that I and others have identified is unlikely to be prioritized or integrated into vaccine policy in the near term.
Amesh Adalja, MD, is a senior scholar at the Johns Hopkins Center for Health Security, and a practicing infectious disease, critical care, and emergency physician in Pittsburgh.
Adalja has received honoraria from Sanofi for participating on a COVID-related education steering board.