January 27, 2021   6 mins

There’s an old Jewish joke that Woody Allen told in Annie Hall. There are two old people at a restaurant, and one says “Boy, the food at this place is really terrible.” The other one says, “Yeah; and such small portions!”

That is slightly how I feel about reports coming out of the EU saying that 1) the Oxford/AstraZeneca doesn’t work and 2) complaining that they haven’t been given enough of it.

We’re all desperate for the vaccine to arrive, and to release us all from this year-long limbo. The desperation is making us jumpy: people are panicking, convincing themselves that this vaccine doesn’t work, there are side-effects, the rollout is taking too long. It’s not true; the vaccines work, they’re safe, and here in Britain, at least, the rollout is going remarkably smoothly. But the tension is real.

It’s not just in the UK. In the last couple of days, a vaccine trade war has been threatening to break out: AstraZeneca not supplying the EU with all of the vaccine it promised, and the EU threatening to sit on its own supplies of the Pfizer vaccine in response. People in the biotech industry have muttered to me about the possibility of this conflict shutting down supplies of raw materials. Vials, reagents, nucleotide bases and so on are all made in different places, and if countries stop cooperating, it suddenly gets much worse for everyone.

And where all this started was a bitter little row about whether or not the Oxford vaccine works in older people. On Monday, it was reported in the German newspaper Handelsblatt that the Oxford vaccine was only 8% effective in people over 65. 

This is simply false. The 8% figure is biologically implausible and entirely without sourcing, and Oxford University, AstraZeneca and the German health ministry have all said that it is wrong. It’s really important to stress this, because there are already people who should know better out there saying silly things about the vaccine being more risky than the disease. The vaccines work; they work in all age groups; if you get offered one you ought to take it.

There is, it is true, very little clinical data on the Ox/AZ vaccine in older people. That is because they were recruited later to the trial and in smaller numbers. But there is good clinical data in younger age groups. And you can analogise from those younger groups to the older ones by “immunobridging”. 

To do this, the Oxford/AstraZeneca trial looked at other measures: whether the numbers of neutralising antibodies, T-cells, B-cells and so on go up after the dose. You do that in older people as well as younger people. With some diseases and vaccines – notably flu – the immune response is much weaker in older people than in younger people; but with Covid, or at least with the Ox/AZ vaccine, the immune response remains surprisingly consistent.

The trouble is that you don’t know whether these proxy measures really correlate with protection. Often they do, but without clinical data you don’t know.

So you “bridge” from a group in which you have both the clinical data and the immunogenicity data — in this case younger people — to a group in which you only have the immunogenicity data – in this case older people. You assume (with good reason, because it has worked so often in the past) that that will give you a reasonable picture. And that isn’t an 8% efficacy.

My understanding is that the figure is one of, essentially, two things: someone taking a figure from a very wide uncertainty interval; or someone simply reading a number wrongly. The later is what’s being floated by the German health ministry. They apparently suggest that 8% of the total number of participants in the trial were between 56 and 69, and that someone has misinterpreted that as the efficacy level in over-65s.

That seems an implausibly basic mistake to me. It does fit with the numbers in the December Lancet paper: 56-to-69-year-olds made up 974 out of a total of 11,636, or 8.37%, of the total. But as far as I can tell the thinking is literally just “they said 8%, here is an 8% in the paper, are they the same?” It’s not even the right age group. 

Besides, a Handelsblatt journalist denies that’s what’s going on, and the newspaper has doubled down on its claim, quoting a “high-ranking official” saying “It is impossible to mix up the numbers. According to the data we have so far, the effectiveness in people over 60 is less than ten percent.”

My preferred hypothesis is that someone has looked at some clinical data in the over-65s, with a very small number of cases and therefore a very wide confidence interval, seen 8% at the bottom end of that confidence interval, and leaked that. It must be unpublished figures, because the Lancet papers had no “over-65s” group, and besides, they only had five cases in over-55s, which is not really enough to give any sort of estimate. But even so, it doesn’t make sense since the “immunobridging” data lets us be confident that it is reasonably effective in older groups. 

So the 8% figure appears to be bunk. But it’s interesting bunk, because it fits into a wider tension between AstraZeneca and the EU, with the UK either caught in the crossfire or engaged in the conflict itself, depending on who you ask. 

The German newspaper Bild reports that the European Medicines Agency will not approve the use of the AstraZeneca vaccine for the over-65s, out of fears that it will be less effective. And at the same time, there are manufacturing problems at a Belgian factory making the vaccine on license. The EU may get 60% fewer vaccine doses than it asked for; around 31 million instead of around 80 million by March.

(“This vaccine is terrible!” “Yeah, and such small portions.”)

In an interview for La Repubblica, Pascal Soriot, the CEO of AstraZeneca, explains some of this. The production of viral-vector vaccines like the Ox/AZ one is a biological process, like fermentation, not a straightforward chemical engineering. Slightly different processes will end up with different outcomes. The Belgian plant, which makes the batches of vaccine to be taken somewhere else and put into vials, found that its batches were coming out with a lower-than-expected yield of the active ingredient, the adenovirus, in the solution. 

For the record: it would make sense for the EMA not to use the AstraZeneca vaccine in older people, since they have loads of the Pfizer vaccine and good data that it works in older people, and not so much of the AstraZeneca one and no clinical data. The cost-benefit equation is different in the UK, because we have loads of the AZ vaccine and (for the reasons discussed above) good reason to think it will work in older groups. 

But the shortage has led to a row about whether AstraZeneca ought to deliver some of the doses it produces in its UK factory to the EU. At the moment, all the UK-produced doses will remain in the UK; in response the European Commission is tightening controls on exports of the vaccines it does have, such as Pfizer. The Commission says it won’t sue, because that takes time, but they’re not happy. 

Rob Blackie, a strategist who works with biotech firms, thinks that AstraZeneca will have been behind the decision to keep UK-produced doses here: “I don’t think it’s a government decision,” he says. “The implication of lots of things people have said is that the drug companies are in control. My assumption was that it would be strange to do anything so nationalistic, because the problem of pissed-off governments is so big.” But a biotech industry source speculated that there was probably significant pressure from the UK government, because it has bet so heavily on AstraZeneca.

Suggested reading
Is a vaccine trade war brewing?

By Aria Babu

Soriot, in his La Repubblica interview, says that AstraZeneca’s contract with the UK was signed three months before its contract with the EU, and that their contract with the EU simply demands that they make their “best effort” to provide as many doses as possible. But, clearly, emotions are running high – hence the strange German claims about 8% efficacy (Soriot sounds baffled by it) and the simmering threat of a trade war.

It’s not just about the EU stopping exports of Pfizer. If this develops into a full-blown vaccine trade war, then entire supply chains could end up getting disrupted. My biotech source agreed that this is not an implausible outcome. AstraZeneca has a stockpile of the raw materials for its vaccine, but some of those materials keep better than others, so it may become difficult if supply dries up. Even small delays in the vaccine rollout could cost thousands of lives, so this matters.

Avoiding a vaccine trade war, then, seems a really good idea. AstraZeneca says it can produce 300 million doses a year in its Wrexham plant – that’s 25 million a month. Elsewhere it says it should be producing two million a week already. If these numbers are anything like accurate, then it seems unlikely that the main bottleneck in vaccine distribution will be the manufacture. 

According to Soriot, the contract with the British government called for the doses from the Wrexham plant to go to the UK first: “As soon as we have reached a sufficient number of vaccinations in the UK, we will be able to use that site to help Europe as well.” But that implies that the UK government could offer to give some to Europe first; as my biotech industry source says, “In vaccine manufacturing, possession is nine-tenths of the law. If the vaccine vials are sitting in the UK, we can choose to how use them.”

I can understand the UK being wary — vaccines are made in batches, millions of doses at a time, not on a production line; so if one batch goes wrong, you lose a hefty chunk of your supply. But enlightened self-interest seems to suggest that we keep the EU sweet rather than start a war, so if we do have spare vaccine, handing some of it over with a smile on our face seems a wise idea. The food actually isn’t terrible, but it would be good if we can give them larger portions.


Tom Chivers is a science writer. His second book, How to Read Numbers, is out now.

TomChivers