Most of these trials found that the updated vaccines — based not only on Omicron, but also on older variants, including Beta — performed a bit better in terms of this measure than the original vaccines did. “This is a clearly superior booster,” Stephen Hoge, the president of pharmaceutical firm Moderna, told investors on 8 June when touting such results from the company’s BA.1-based bivalent vaccine.

To try to make sense of results like Moderna’s, a team led by Deborah Cromer, a mathematical modeller at the University of New South Wales (UNSW) in Sydney, Australia, collected all of the updated vaccine-trial results the researchers could find, as well as studies looking at the efficacy of fourth doses of the original vaccine.

Both vaccine types sent antibody levels skyrocketing, but the updated versions did so to levels that were, on average, 1.5-fold higher than those of older vaccines based on the original SARS-CoV-2 sequence. “We’re not talking about a step change,” Cromer says.

Fewer hospitalizations

Studies suggest that higher levels of neutralizing antibodies equate to better protection against COVID-19. But it wasn’t clear from the updated vaccine trials how much more effective they might be.

To determine this, Cromer’s team applied a model that she, UNSW immunologist Miles Davenport and their colleagues developed relating efficacy of the original COVID-19 vaccines to antibody levels4. The model found that most of the benefits provided by updated vaccines are conveyed by an extra dose of any vaccine.

For instance, in a population where half of people are already protected against a symptomatic SARS-CoV-2 infection through previous vaccination or infection, an updated vaccine booster bumped protection up to 90%, compared with 86% protection provided by an extra dose of the original vaccine. For protection against severe disease, however, the difference in protection was less than 1% (see ‘How COVID boosters measure up’).

At a population level, updated vaccines could make sense. Cromer’s team estimated that, for every 1,000 people, a booster campaign based on updated vaccines would result in 8 fewer hospitalizations, on average, than one based on older vaccines. “If that translates to hospital beds saved and severe cases averted, that might be a sufficient level to warrant the recommendation for a variant-modified booster,” she says.

The relative benefits of variant-based boosters could grow stronger if pre-existing immunity is suddenly lowered by the appearance of a new variant (see ‘Immunity matters’). This occurred in December 2021 after the emergence of Omicron — and might happen again. In this scenario, an Omicron-based vaccine could provide much better protection than older ancestral vaccines, says Cromer.

Immunity matters. Charts showing where booster jabs can improve immunity.

Long-term benefit

When the US booster campaign begins, it set to use different Omicron vaccines from the one approved by the United Kingdom. In June, an advisory committee to the US Food and Drug Administration (FDA) asked companies to develop bivalent vaccines that were based on the original strain and the BA.4 and BA.5 coronavirus variants — which have identical spike protein sequences — instead of the bivalent BA.1 vaccine that had been trialled by Moderna, Pfizer–BioNTech and others. The hope was that, by better matching circulating strains, the vaccines would prove more effective.

But Cromer’s analysis suggests the differences might be paltry. Even updated vaccines based on the Beta and Delta variants should protect against BA.4 and BA.5 infections nearly as well as vaccines based on those variants. Similarly, bivalent vaccines that included the original vaccine looked no more effective than vaccines based solely on a newer variant.

For these reasons, the FDA’s decision to spurn a BA.1-based booster probably wasn’t worth it, says Cromer, particularly as SARS-CoV-2 continues to evolve. “It doesn’t seem to suggest that it’s going to give a dramatic improvement in the effectiveness of the booster vaccine to have that slight change.”

Dean Follmann, a statistician at the US National Institute of Allergy and Infectious Diseases in Bethesda, Maryland, says that even the marginal benefits of a vaccine based on BA.4 and BA.5 might be enough to justify their roll-out. “It’s probably somewhat better. A lot better — probably not.” Moreover, he says the main message of the analysis should be that any COVID-19 booster is a good one.

Other scientists question the decision to pursue variant boosters when the benefit is so small. Moore worries that people are being misled into thinking updated vaccines are vastly more effective than existing vaccines, and might then expose themselves to greater infection risks.

Paul Offit, a vaccine scientist at the Children’s Hospital of Philadelphia in Pennsylvania who was one of two members of an FDA advisory committee to vote against updating COVID-19 vaccines, says that Cromer’s analysis underscores his scepticism. “If this vaccine is no more effective than the current vaccines, then why distribute it?” He says. “You’re going to have very little impact on the incidence of severe disease.”

In the long run, it probably makes sense to develop variant-based vaccines, says Cromer, but the idea that they need to closely match circulating viral strains is unrealistic — and counterproductive when highly effective vaccines are already available. “The most important vaccine booster is the one that you actually get.”