how can we assess how much of a problem a partial vaccine escape variant like the SA variant B.1351 could be for the UK once we’ve re-opened later this year? (thread)

the problem: whilst we know that vaccines should provide some useful protection from from severe disease with B.1351 and likely transmission as well, relatively small reductions in protection could still lead to hospital overload if widespread transmission restarts

in addition, the UK faces an extra concern as most countries that have vaccinated their population very quickly have used mRNA vaccines which have very high efficacy, whereas here the majority of jabs will be AstraZeneca which is associated with lower antibody titres

so there’s a high chance the U.K. will end up being the test case here, ie we’ll only know whether B.1351 can still spread in a population heavily vaccinated with AstraZeneca when we open up fully domestically and to international travel

now the theory for discussion: is high levels of natural immunity to the original pandemic strain and B.117 a good correlate for high levels of AZ vaccination?

this theory suggests that if B.1351 spreads in a population level hat is close to the herd immunity threshold for the original strains due to prior non-B.1351 infections, you would expect to see the same in the heavily AZ vaccinated UK later this year

this theory is based on the early correlates of protection modelling, the graph below shows that AZ vaccination leads to around half the antibody titres of the average Covid survivor, and with that an associated reduction in efficacy against infection
https://www.medrxiv.org/content/10.1101/2021.03.09.21252641v1

taking into account the natural reduction in antibody titres over time, and assuming that on average people will have been vaccinated in the UK more recently than they were infected in Covid in a comparator country, I’d expect to see similar levels of protection …

plus you’d expect an actual Covid infection to stimulate a strong cell mediated immune response too, which means, if anything, the “T cell protection” argument should be in favour of B.1351 reinfections being less severe than B.1351 infections post AZ

grateful for all thoughts on this, particularly on 1) whether the logic above is sound and 2) whether anyone is aware of countries or regions where we should look out for this pattern emerging? @CovidSerology is good on this and has highlighted Bangladesh already https://twitter.com/declamare/status/1381496736452124675