Vaccine makers and medicine regulators are sketching out plans in case the mutating coronavirus turns vaccine development into a game of cat and mouse.
Just weeks after leading manufacturers secured the first regulatory approvals, mutations in the virus have forced scientists to re-test their Covid-19 vaccines and prepare to tweak their formula should the shots prove less effective.
At the same time, regulators are considering how they could fast-track new approvals and whether they could use the seasonal flu jab as a model to authorise revised versions without requiring long trials.
Scientists are hopeful that the existing crop of vaccines will still prove effective against the highly transmissible new variants identified in the UK and South Africa. Initial laboratory tests at the University of Texas have found that the BioNTech/Pfizer vaccine still works against one of the most concerning mutations in the UK and South African variants.
However, the shot has not yet been tested against all mutations in those variants and with further mutations likely, Andrew Pollard, director of the Oxford Vaccine Group, warned against complacency. "We could well find ourselves in a place where they could evade immune responses in the future."
Andrey Zarur, chief executive of Greenlight Biosciences, a biotech company that is making a messenger RNA vaccine, said it made sense to design new vaccines to "prepare for further drift".
"Then you have a better chance of your new vaccine continuing to protect against new variants," Mr Zarur said.
BioNTech/Pfizer and Moderna, which were the first manufacturers to bring their experimental Covid vaccines to market, are well positioned to respond to mutations. The messenger RNA technology enables the companies to plug in adjusted genetic code in response to any changes in the virus. BioNTech has said it could make a new vaccine in six weeks.
Paul Duprex, director of the Center for Vaccine Research at the University of Pittsburgh, said the "beauty" of mRNA is that it uses "people as factories" to make the protein needed to trigger the immune response.
For other vaccine makers reliant on different technologies the production process can be more time consuming. Dr Pollard said it could take as little as a couple of days to devise a new formulation for the adenovirus-based Oxford/AstraZeneca vaccine, but much longer to produce.
"The thing that would take time is the manufacturing — taking the new seed, putting it in that manufacturing facility and making millions of new doses," he said.
Geoffrey Porges, an analyst at the healthcare focused investment bank SVB Leerink, estimated that the entire process until the new shots were in people's arms would take three to six months for the mRNA vaccines. It would take six to eight months for the adenovirus vector vaccines like Oxford/AstraZeneca and Johnson & Johnson, and up to nine months for protein-based vaccines like those developed by Novavax and Sanofi/GSK, he said.
Sanofi, which is already lagging after its vaccine failed to produce a strong immune response in older adults, told the Financial Times it would take "longer development steps to cover a new virus strain than the mRNA approach, should we need to target a new coronavirus strain".
The vaccines that rely on using the virus in a dead or weakened form, such as those from China's Sinopharm and Sinovac, would also take longer to adapt. Sinopharm and Sinovac did not respond to requests for comment.
Timing will also depend on what the regulators require to approve a rejigged vaccine. The European Medicines Agency said it was already in discussions about what to require if a change is needed. The US Food and Drug Administration said it was still watching to see if the virus mutates in a way that would make the vaccines less effective.
Both regulators told the Financial Times that the flu vaccine, which is tweaked every year to keep up with the most prevalent strains, could provide a model for a fast-tracked approval process.
"We do this every year for influenza vaccines," said Angela Rasmussen, a virologist at Georgetown University. "We don't have to have large clinical trials to make sure that they work, but we've also had influenza vaccines for a lot longer than we've had coronavirus vaccines."
In normal trials a vaccine's efficacy is judged by whether vaccinated trial participants are less likely to get infected than those who have received a placebo. To skip this process scientists would need to have reliable immune system markers they can use to assess whether a person will be able resist the disease. One example, Ms Rassmussen said, was how many antibodies it takes to neutralise the virus.
Peter Hotez, a vaccine specialist at the Baylor College of Medicine, said the FDA should provide more guidance on what studies they would require to approve adjusted vaccines so developers do not have to "reinvent the wheel".
In the long term, he said this problem could be solved by a universal coronavirus vaccine, designed to produce at least partial protection to the whole family of viruses, which also includes Sars and Mers.
The US National Institutes of Health recently invited researchers to apply for funding to create vaccines for coronaviruses with pandemic potential. But this would be a huge challenge, as many scientists have been trying to create a universal flu vaccine for decades. In the past five years, there have been breakthroughs, but two clinical trials failed last year.
Until then, if Covid-19 does mutate in ways that reduce vaccine efficacy those companies working with messenger RNA would have an advantage in an annual vaccination market that could be worth more than $10bn.
"I don't think for a minute that the major companies are wishing for a bad virus," said SVB Leerink's Mr Porges. "But it does fundamentally alter the commercial value of the opportunity."
- Financial Times