Hacker News Reader

Comment by dannyw | Hacker News Reader

You have a dosing error that resulted in some participants getting half the dose initially, and this group turned out to have substantially better resistance (~90%) than the one receiving the full dose (~60%).

This is completely mysterious, in the opposite direction of every other vaccine candidate, and should we be following normal safeguards, there is zero chance this vaccine would be approved.

It also raises concerns of rushedness/negligence. When you get two conflicting measurements of the dosages, the correct response is to test it a third time, not assume the second is correct. This failure mechanism was the same in Boeing.

Not that we need to be: we already have Pfizer and Moderna, that did their trials properly and showed proper results.

View on HN

I disagree with your general conclusion that the vaccine should not be approved as-is. The efficacy results are certainly disappointing compared to the near-miraculous results with the two mRNA vaccines, but 60%+ efficacy is still enough to save many, many lives and bring the world back to normalcy. Remember, the 60% number doesn't mean 60% see improvement, it means of the people who would've gotten COVID without the vaccine, 60% didn't get it at all and 40% got it with significantly less negative effects. In all likelihood it will also greatly reduce the ability to spread the virus.

They have already begun testing the half-full dose as another trial, so if that works better than full-full we may see the approval modified down the line.

> we already have Pfizer and Moderna

Unfortunately production constraints are going to mean that it will take a long time for those two vaccines to be produced in enough quantity to vaccinate most of the rich world. The cost for them and temperature requirements mean that they'd be impractical for most countries even in a few years.

> 60% didn't get it at all and 40% got it with significantly less negative effects

We don't know that. As with the Pfizer and Moderna trials, the endpoint was incidence of disease, and only people with disease symptoms were tested. 60% fewer people experienced symptoms in the treatment group than the control group, but some of those may well still have "gotten it." There were very likely asymptomatic infected people in both groups, and we have no idea how many there were, or how the numbers different between groups. For all we know, just as many people in the treatment group were infected as in the control group, they were capable of spreading infection the whole time, and all the vaccine did was suppress their symptoms. (Again: possibly true of any of the vaccine candidates.)

When the broadest studies indicate that symptomatic cases are ~25x more likely to transmit the virus to people in their household[0] I would say that converting a lot of symptomatic cases to asymptomatic cases is likely pretty consequential.

[0]https://jamanetwork.com/journals/jamanetworkopen/fullarticle...

Are you sure Oxford and AstraZeneca did not test for asymptomatic participants? My understanding is they tested a proportion weekly.

Edit: Nature mentions they did: https://www.nature.com/articles/d41586-020-03504-w

My understanding is that they did test a subset of participants, but that the 62% figure doesn't figure in the asymptomatic cases. They've made it very confusing though, much like how they've explained the half-full dose results.

Relevant section: The Oxford–AstraZeneca team is the only one of the three leading vaccine developers that monitored for asymptomatic infections, by collecting weekly swabs from some participants to determine whether they had the coronavirus but did not become ill.

> I disagree with your general conclusion that the vaccine should not be approved as-is.

Which dosing regime should be recommended in this case?

It isn't actually completely mysterious. If you listen to what immunologists/ & virologists they can explain a number of mechanisms which might be responsible. Obviously I'm not qualified to give a good explanation, but there's plenty of good exposition if you look.

https://www.nature.com/articles/d41586-020-03326-w

- Statistical effects, which might decrease the difference as the trial progresses. - A bigger dose causing an over-effective suppression of the carrier adenovirus, suppressing it's ability to make the COVID spike protein, and hence the production of COVID targetting T cells.

There is also a huge flaw in the testing of the Pfizer and Moderna vaccines: they didn't test everyone. Only people with symptoms (whatever that means) were tested (PCR). Whereas with the Oxford AZ vax all participants were tested, so we can actually get an idea of asymptomatic infection.

> There is also a huge flaw in the testing of the Pfizer and Moderna vaccines: they didn't test everyone. Only people with symptoms (whatever that means) were tested (PCR).

Wow, do you have a source for this? That's quite a stunning difference and really should put the ~~Pfizer~~ BioNTech and Moderna vaccines in doubt if the aim is herd immunity.

Have a look at Table 4 here: https://www.cdc.gov/vaccines/acip/recs/grade/covid-19-modern...

They seem to have preliminary results for asymptomatic testing: 14/14314 asymptomatic individuals were found in the vaccine group v/s 38/14073 in the control group. At least from the preliminary data, the Moderna vaccine nowhere near as effective in stopping asymptomatic infections.

A more detailed overview:

https://theconversation.com/covid-19-vaccine-faq-6-things-to...

You'd have to look at the registered study details for the full protocol. See this BMJ article for references:

BMJ 2020;371:m4058 | doi: 10.1136/bmj.m4058

It doesn't prevent herd immunity if it only prevents the disease (i.e illness) but it makes it very difficult to achieve strong suppression if there is vaccine hesitancy (a polite grab bag term for science denialism, people who watch Fox, loopy muffin brain people, etc) above 20%. Which there will be in many parts of the US and Europe because of years of uncontested bullshit. (Thanks Facebook.)

Reply on news.ycombinator.com