Wikipedia:Reference desk/Archives/Science/2021 July 3

Science desk
< July 2	<< Jun | July | Aug >>	July 4 >

Welcome to the Wikipedia Science Reference Desk Archives
The page you are currently viewing is a transcluded archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

July 3

COVID vaccine mixing

I got my second shot the other day: the Moderna COVID-19 vaccine. My first shot, back in April, had been the Oxford–AstraZeneca COVID-19 vaccine, since that was the only one available at pharmacies in Ontario at the time (though a week later, they were discontinued amidst the ongoing concerns of rare side effects). The nurse who administered my second shot told me that I was in luck, since that combination was "the best". Is that true? I've done some searching, but it seems like most of the comparisons involve O-AZ vs. O-AZ/Pfizer (and the suggestion there is that the mixture is "as good or better" (ref. But what about O-AZ/Moderna? Matt Deres (talk) 14:16, 3 July 2021 (UTC)

Since Pfizer and Moderna are so similar, it's a good guess, but if you want proof, there's none yet. FWIW, the news that the J&J shot doesn't need a booster is based on a study of like 10 people, so I don't believe it yet. One shot of any vaccine is definitely insufficient. 74.64.73.24 (talk) 15:24, 3 July 2021 (UTC)

According to this news report, the German Standing Committee on Vaccination has stated that "current study results" show that the immune response generated after a mixed dose vaccination "is clearly superior."  --Lambiam 17:24, 3 July 2021 (UTC)

Though they did mean compared with two shots of Astra-Zeneca. Here’s the original in German. Cheers ⌘ 19:14, 3 July 2021 (UTC)

This UK study showed good results from a mix of AZ and Pfizer. Alansplodge (talk) 23:37, 3 July 2021 (UTC)

• From the above, the RKI does not reference any scientific study, and the BBC refers to [1]. Notice that the main page of that site says We ask that the media do not continue to contact researchers while this critical work is underway, and to use the media contact form for press enquiries., which makes me suspect they were unhappy with how the research was reported. From there, we get links to two actual publications, [2] and [3]. The former is an interim report on "reactogenicity" (= prevalence and importance of secondary effect) following the second injection, which is higher for a mixed vaccination.

The latter, published less than a month ago, compares all possible schedules of BNT (Pfizer-BioNTech, mRNA vaccine) and ChAd (Astrazeneca, viral vector vaccine). The endpoint is the (geometric) average of antibody (IgG) concentration in patient plasma 28 days after the booster (second shot). My interpretation of the abstract's figures is that ChAd/ChAd fares poorly compared to any scheme with at least one BNT; it is not so much that mixed vaccines are good but that pure ChAd is bad.

Please note that any interpretation rests heavily on the choice of endpoint. It might well be that "one month after booster" is precisely the time at which ChAd/ChAd is the weakest compared to others (daily antibody dosing would proably have been too intrusive to the study participants though). If you think "surely the choice of cutoff does not matter much" I would encourage reading this blog - for instance, check out their analysis of the original Pfizer or AZ trial results. Also, the study participants were all at least 50 years old, probably because of the recruitement period; it would be highly dubious to extend the results to younger populations.

Antibody levels are not a really good endpoint for clinical practice; as the OP's link says, what matters is vaccine efficacy i.e. whether one vaccination scheme really cuts infections or infection severity more than the other; while it is reasonable to expect that higher antibody counts means higher protection the link is far from direct.

Based on what I read, the only plausible argument to mix vaccines is at the population level under a vaccine supply constraint, because maybe two people vaccinated with a BNT+ChAd mixture fare better on average than two people with full BNT and full ChAd. Again, we do not know this: as far as I know nobody has run a mixed-vaccine efficacy study (of course, manufacturers are not going to fund a mixed-vaccine study). Frankly, if I were in a position to make such decisions, I would not greenlight vacccine mixing without data on efficacy, because I am absolutely not convinced the stronger side effects are worth it. Tigraan^{Click here for my talk page ("private" contact)} 10:42, 5 July 2021 (UTC)

Thank you. I suspected it was a bit early to make pronouncements like that. Further details on mixing here, inclduing "Of note is that the order of vaccines made a difference, with an Oxford-AstraZeneca/Pfizer-BioNTech schedule inducing higher antibodies and T-cell responses than Pfizer-BioNTech/Oxford-AstraZeneca, and both of these inducing higher antibodies than the licensed, and highly effective ‘standard’ two-dose Oxford-AstraZeneca schedule. The highest antibody response was seen after the two-dose Pfizer-BioNTech schedule, and the highest T cell response from Oxford-AstraZeneca followed by Pfizer-BioNTech." Matt Deres (talk) 13:51, 5 July 2021 (UTC)