Thread by @DavidLVBauer

Significant loss of antibody neutralisation vs live B.1.617.2 SARS-CoV-2 variant (-5.8x, akin to B.1.351) in @TheCrick / @UCLHresearch Legacy study of 250 (!) Pfizer-BioNTech 1/2-dose vaccinees, out today @TheLancet, by @dremmacbw, @MaryYiWeiWu et al. thelancet.com/journals/lance… 1/n

Increased age & time since 2nd dose correlated with reduced virus neutralisation across all strains tested. Not a surprise, but given low starting titres vs B.1.617.2, more of a concern to see neutralisation “dropping off”, significantly. Boosters more likely to be needed 2/n

To maximise population coverage, the UK delayed 2nd dose from 3 weeks to ~12 weeks in early 2021. Was a good strategy vs. B.1.1.7, but single-dose vaccine recipients have significantly less ability to neutralise B.1.617.2 -- so strategy now more complicated 3/n

Again, the correlation between increased age and less neutralising ability that we saw with 2-dose recipients is also visible with single-dose recipients. Supports getting 2nd doses in ASAP, esp. in older / vulnerable populations. 4/n

What does this mean for vaccine efficacy? Recent model from Khoury et al @KirbyInstitute gives a good idea of what we can expect... starting from a lower titre (B.1.617.2) means small changes in antibody titre now likely to have larger effect on VE. nature.com/articles/s4159… 5/n

Building better models in future will also require standardisation across labs using the WHO Reference Standard & Reference Panels. (We contributed to its evaluation, so know @TheCrick high-throughput live virus neut assay stacks up pretty well, even at 2000 samples/wk) 6/n

All a result of collaboration betw NHS & UK academic virology labs: swab from @Thushan_deSilva in
🏥 Sheffield @CovidGenomicsUK ➡️ cultured by @wendybarclay11 at 🧪 Imperial ➡️ RNA biked (by me!) to @TheCrick to 🧬 sequence ➡️ dist. by #G2P to 🦠 virologists around UK 🇬🇧 7/n

And critically down to the foresight of @CharlesSwanton @LabGandhi, Bryan Williams @UCLHresearch BRC to set up the SARS-CoV-2 Legacy Study so that researchers would have a cohort "ready to go"

and the amazing team @TheCrick -- from virologists and immunologists to statisticians and database experts -- who made it possible to go from virus to data in less than a week. We'll be putting all the data on Github shortly 9/n

Aside from (1) our lab-based data pointing to reduced neutralisation of B.1.617.2 (delta) by vaccine-induced antibodies, data from PHE suggests it is also (2) more transmissible than Kent, and also (3) more likely to increase hospitalisation risk, see assets.publishing.service.gov.uk/government/upl… 10/n

Our cohort here is close to a "best-case" picture for neutralising antibody titres: relatively young (40y), healthy, and recently-vaccinated. We need to be paying careful attention to other populations w/variants (ethnicity differences? older? immunocompromised?) 11/n

The plus side is that 2 doses give reasonable titres. And there is of course more to immunity than neutralising antibodies (even if that's the clearest correlate of protection we have). 12/n

Summary: vaccine rollout has been a huge success in the UK – slow and steady has almost won the race, B.1.617.2 (delta) variant makes it a bit more uphill as we near the finish line... Two doses as quickly as supply allows, monitor Ab levels, boost those who need. 13/end