Skip to main content

Clinical Trials

The Oxford/AstraZeneca Vaccine Efficacy Data

We now have a complete writeup of the efficacy data from the Oxford/AZ vaccine effort in The Lancet, and I’m glad to see it. There have been a number of questions about this candidate and its effects in the clinic, so the chance to get a closer look is welcome.

This is of course the “ChAdOx” vaccine, with the first part of that acronym standing for “chimpanzee adenovirus”. And it’s the first viral-vector candidate that has reported out. J&J is coming soon with their adenovirus-26 vector, perhaps even sooner than expected due to the terrible number of coronavirus cases in the US, and Russia’s Gamelaya Institute has their one-shot-of-adenovirus-5 and one-shot-of-adenovirus-26 vaccine as well, for which data are unfortunately very sparse. This paper is from the UK, South Africa, and Brazil trials (others are ongoing), with a total of over 23,000 participants. The portioning of those into treatment and control groups, though, is where some of the arguing starts. These trials are notable for using an approved meningitis vaccine in the control group in the UK trial, rather than saline (as indeed the US trial of this vaccine is doing). The Brazil arm did that for the first shot, with saline for the second, and the South African trial used only saline controls. Famously (by now) a subset of patients in the UK trial got a half dose of vaccine first, followed by a full dose as booster, which seems to have been an out-and-out mistake in handling and formulating the vaccine. Update: here’s the New York Times on the various mistakes that have been associated with this one. Among those is a further difficulty with the half-dose vaccine group – some of them got their booster shots much later than others, for unexplained reasons.

In the larger cohort who got the two regular doses, vaccine efficacy was 62% (with a 95% confidence interval of 41% to 75%). There were 27 cases in the 4440 treatment patients, compared to 71 cases out of 4455 patients in the controls. Meanwhile, in the half-dose-first group, efficacy comes out at 90% (with a 95% confidence interval of 67% to 97%). Both of those 95% CIs are a bit of a spread. In this group, 3 out of 1367 patients in the treatment group came down with coronavirus, compared to 30 out of 1374 in the controls. That control-group infection rate is definitely higher than what was seen in the two-full-dose group, which makes you wonder if it’s running randomly high (which would make the half-dose group look better than it really is) or if the larger full-dose group was running randomly low (which would make it look worse than it really is, but remember, with its larger sample size there’s correspondingly somewhat less room to believe that it was that far out of whack). We need to add in to these calculations the news that the half-dose group included no patients older than 55, and to wonder what effect that had on the numbers, too. The paper reports a combined overall efficacy of 70.45 (95% CI between 54.8% and 80.6%), but how much you trust that one comes down to how different you think these two groups are.

Basically, the more you believe the 90% number is a statistical fluke, the more you then believe that the real efficacy of the vaccine is probably in the 60-per-cent range. On the other hand, if you buy into the idea that there’s something really different and useful about giving everyone a half dose at first, you can be somewhat more optimistic. That belief is optimistic in itself, but it’s not completely nuts, either. Remember, one of the things about a viral vector is that you raise antibodies not only to the protein that whose genetic message you’re delivering, but to the viral vector itself. It’s not impossible that a lower dose the first time made subsequent antibody neutralization of the second dose less of a problem. But you’d need to prove that. And to prove, with intention, that a half-first-dose really does work better to start with.

A bright spot in the data is that (counting from 21 days after the first shot) there were overall ten hospitalizations for coronavirus (two severe, one leading to death), but none at all in any of the treatment groups. This would be a good point to note that the coronavirus cases totaled in all these trials were based on symptoms, which were then confirmed by RT-PCR swab assays. The UK trial also had regular swab assays taken to try to get data on asymptomatic infections, which is something that wasn’t done in either the Pfizer/BionTech or Moderna trials. The numbers aren’t huge, and they aren’t tight, but they’re all we have. There were 24 asymptomatic patients in the low-dose-first group, for a vaccine efficacy of 58.9% (95% confidence interval from 1 to 82.9%, and that’s what I mean by “not tight”), and 45 patients in the two-standard-dose group (efficacy of 3.8%, unfortunately). So at least for this vaccine, the efficacy at preventing asymptomatic cases is notably lower than that seem for symptomatic ones, and that may well be true for all of them. The differences in those asymptomatic numbers, though, also argue that the low-dose vaccination regimen was indeed different from the two-standard-dose one, don’t they?

Now to safety. The top line numbers look reasonable, with the usual reactogenic effects in the treatment groups, and with very similar numbers of adverse events in the treatment and control groups. The part that makes you wonder a bit, though, is the case of transverse myelitis (which made quite a bit of news at the time, at least once the details rather grudgingly emerged). This was two weeks after the second vaccination, and considered by the monitoring committee as likely to be related to the vaccine. It turns out that there were two more, but that news isn’t as bad as it sounds. One of these was in a participant who (as it turned out) had undiagnosed multiple sclerosis, so that’s less likely to be related to the vaccination. The other was in one of the meningitis control arms. It sounds like there was some arguing about whether that was vaccine-related or not (it was 68 days after the shot), but it’s certainly not related to the coronavirus vaccine, anyway. There still is that one case, and transverse myelitis is one of those neurological syndromes that can indeed be associated with autoimmune function (and thus possibly with vaccination). So this will bear close watching as this vaccine gets rolled out.

Which brings up the question of when that might be. As I understand it AstraZeneca is even now wrestling with the problem of whether to run another trial using that low/high dosing protocol. If the different results seen here are real, that could be a significant improvement, both in overall efficacy and in using up less vaccine, but there’s also a real chance of chasing after a mirage. If you go with the numbers as they are, though, it would seem that this vaccine is likely to be provably inferior to the Pfizer/BioNTech and Moderna mRNA ones. That doesn’t mean that it’s not good enough to be useful in the pandemic – just that there are others that could be even more useful, if you could deploy them instead. As it stands, there’s a real risk of having some patients (or indeed whole countries) feel as if they’re being dealt a second-class vaccine if this one is their only option.

Naturally, the same questions apply here as were mentioned in the Pfizer article that I posted earlier today – duration of action, effect on viral transmission, and so on. We’re going to have to wait and collect more data to be able to say anything about these, for any of the vaccines. But the data today make me quite curious about the coming J&J data. That’s another adenovirus vector, as mentioned, and there’s at least a possibility that we will then have a good head-to-head comparison between mRNA vaccines and adenovirus vector vaccines for the same pathogen. That’s a pretty strange situation – when did we ever have multiple simultaneous vaccines for the same disease? And there’s the Novavax recombinant protein one coming after that. No, this is a unique situation. Let’s hope it stays unique!

Update: it looks like India’s regulatory authorities have already declined to issue an accelerated approval for the Oxford/AZ candidate, which may be a sign of trouble to come. . .

74 comments on “The Oxford/AstraZeneca Vaccine Efficacy Data”

  1. myst_05 says:

    Is there any logical reason NOT to approve the Oxford vaccine RIGHT NOW and start saving lives? Since its near-100% effective against hospitalisations (which is the most important issue, really), it could IMMEDIATELY start saving lives all around the world. The 7-day average for deaths around the world is now around 11k/day, an absolutely staggering number. Delaying an effective vaccine for even a single day will result in hundreds of avoidable deaths.

    If someone doesn’t want an “inferior” vaccine, let them wait. Personally I’d much rather get the Oxford vaccine and experience some mild COVID symptoms than get nothing and potentially go to the hospital.

    1. FoodScientist says:

      Derek, for the attenuated/ inactivated vaccines; how are they producing enough of the virus? You’d think cell cultures would take up a lot of space/ not produce enough. Or do they just throw a bunch in a vat and they survive long enough for the virus to replicate?

      The thought of using bats or pigs to culture it is a little “ewwy”.

      1. Thomas says:

        I think there is a cell line where the vaccine-virus can grow. But it doesn’t grow in normal humans.
        It all seems to happen in vats.

        1. K sidhant says:

          indian covishield is doing great job across sub continent.

    2. Random Med Chemist says:

      Your comment on hospitalizations I don’t believe is fully supported by the data. The small number of observed infections could make the lack of hospitalizations a fluke. I believe (and I defer to those better with the biostatistics than I) that you would need a much larger and longer study to be able to confidently make that claim.

      1. Chris Phillips says:

        I don’t think 0 hospitalisations versus 10 is likely to be a fluke.

        On those figures I reckon the expected value of “efficacy against hospitalisation” is 90.1%, with a confidence interval of 60.6%-99.8%.

        Of course, as there were 0 hospitalisations among those vaccined, the data would also be consistent with a higher efficacy than that, but we’d need higher numbers to know.

        1. Wilber Deck says:

          Naive statistical question: shoulldn’t the confidence interval include the data point (100% efficacy)?

          1. Chris Phillips says:

            No, because the confidence interval is the middle 95% range of efficacies in the estimated probability distribution. If it’s a continuous distribution, there has to be a 5% range above it at the top, which will include total efficacy. If there were an a priori reason to expect the vaccine could be absolutely 100% effective with non-zero probability (rather than say 99.99%), it could lie in the confidence interval, but not otherwise.

  2. Fob Snow says:

    Derek, another thing that confuses the Oxford results is that the second shot was received at very different times by different participants and in different arms of the trial. Still at the end of the day, there were no hospitalizations or severe disease etc among those who received the vaccine – solving the primary issue we have with Covid-19.

    While the US regulators will probably not approve it, the vaccine will save lives of most covid patients regardless of whether the trial efficacy is 60% or 90%. Millions of doses of Astrazeneca’s product are already available, there is no need to let more people die in the name of generating more data.

    1. Graydon says:

      We have a trial that says that the vaccine successfully suppresses symptoms.
      We’ve got bunches of data that says asymptomatic cases experience damage.
      Do we have reason to believe that the vaccine prevents damage?

      Because if it doesn’t prevent damage and other vaccines do — and it can be noted that nigh-everyone is carefully not testing for this by evaluating on symptoms, so we don’t know — then those vaccines are strongly preferable.

      1. P says:

        What do you mean by asymptomatic patients experiencing damage?

        1. Graydon says:

          A fairly large percentage of COVID-19 cases present no symptoms. This does NOT mean the disease doesn’t cause damage; for example,

          A vaccine that prevents acute cases but doesn’t prevent either transmission or asymptomatic infection is useful, in the same sense wearing a mask is useful, but would not reliably prevent the disease from harming the portion of the vaccinated population where the vaccine has prevented acute cases of the disease. (It wouldn’t be advancing the overall goal of extirpating the disease, either; the opposite, if anything, as vaccinated people relaxed other precautions.)

          1. David Baker says:

            That article talks about lasting effects in people who did experience symptoms but who have recovered from acute disease. That’s very different from “damage experienced by asymptomatic people.”

          2. Barry says:

            Many who are asymptomatic (but can still be diagnosed by PCR and by thoracic X-ray) do suffer lasting damage/debility

          3. theasdgamer says:


            Lol @ “Many who are asymptomatic (but can still be diagnosed by PCR and by thoracic X-ray) do suffer lasting damage/debility”

            From your referenced paper: “Long-term follow-up analyses are clearly required to determine whether the reticulation represents irreversible fibrosis, thus clarifying the transient or permanent nature of CT changes noted in asymptomatic COVID-19 survivors.”

      2. Joey says:

        The Pfizer vaccine study has not produced any information about whether it prevents asymptomatic cases either.

        I’d take the Phizer vaccine after it was approved, and continue to wear a mask if I were worried about either asymptomatically spreading the virus, or about whatever risk there may be of “damage” from an asymptomatic case.

  3. Noni Mausa says:

    Another factor is the shipping/storage protocol and the cost of the Oxford/AZ vaccine. Even if somewhat less effective, it ships and stores at normal fridge temperature, and will be one tenth the cost of Pfizer, IIRC. There are situations where this will be crucial.

    1. Joseph says:

      Pfizer’s storage conditions are admittedly nontrivial. Moderna, on the other hand stores and ships at -20, and the existence of the freezer aisle in your grocery store is proof of how robust that supply chain is.

      1. Ian Malone says:

        Even the -70C cold chain doesn’t seem to be that problematic while we’re in the “mass-rollout” phase. It can be stored at fridge temperatures for a while right before being administered, in the UK mobile units are starting to vaccinate people in care homes. I’m not sure exactly how it’s structured, but the system is operating around hospitals as distribution centres

        Likely more of an issue once we get to the point of wanting people to come to GPs to get vaccinated, but it seems likely we’ll have vaccination centres in the UK at some point.

        1. A Nonny Mouse says:

          My son has just been asked to help with vaccination (medical student, 5th year) but has refused as they are paying below minimum wage!

          So much for everyone mucking in (told him what we thought..).

          1. No, he is quite right. It is skilled work. We all know that, in the event, much of it will be parcelled out to comparatively unskilled (hastily/inadequately trained) workers. We also know that the efficacy will suffer — ?marginally, ?severely — from unskilled administration, but that effect will never be measured. Signposting the work as low-value work, by trying to underpay, will lead to more unskilled administration and lesser efficacy.

      2. Not-an-epidemiologist says:

        Sure, but Moderna by themselves can’t produce enough vaccine to save the world (unless you’re willing to wait multiple years). Not everyone is going to be able to receive the Pfizer or Moderna vaccines, and the AZ candidate is way cheaper to boot.

  4. Blaine White, M.D. says:

    How about the first full dose also raised a Cytotoxic CD8 T-cell response to the vector? Then on the booster the T-cells killed the host cells before they could make much more immunizing Spike?

  5. Philip Ammar says:

    I just talked to a recruiter today for the US AZ/Oxford vaccine. I asked whether they were testing the full/full dose or the half/full dose in this trial, but since this its still the original trial they are enrolling for, it will be the full/full. I was a little disappointed, but I think its still worth getting this fully enrolled to get better data on it. If I do end up with the vaccine, at least it seems like it is somewhat protective against serious illness.

  6. With increasing doses, experimental dose-response curves for immunization show an ascending limb rising to a peak (immunization) and a descending limb (immunological tolerance). The descending limb is as specific as the ascending limb. It means that the very cells that are going to response to lower dosages are being destroyed (or inhibited). So the cells may not be present to response to a second dose (or to the pathogen which the protocol hopes to target).

  7. An Old Chemist says:

    AstraZeneca COVID-19 vaccine data hint at single-dose efficacy, effect on asymptomatic cases (FierceBiotech, 12-09-2020)

  8. Bash says:

    If the AZ vaccine prevents hospitalization and/or severe covid, and basically relegates anyone who takes it to getting a cold at most if they get covid, then why is this a bad thing? I would have thought this would be a great victory?
    What is crushing societies is overwhelmed hospitals. If ChAdOx eliminates that and death, what is the problem?
    AZ didn’t help themselves by running a sloppy trial

    1. Dick King says:

      If the vaccine prevents all serious infections but does not prevent much asymptomatic transmission, that would be a contribution, and you can argue that vaccine refusers get what they deserve, but some people cannot take the vaccine for individual medical reasons.


    2. Rudolf says:

      Actually the trials were set up and run by the academic scientists at Oxford to generate and publish the most comprehensive data. With less experience than Big Pharma they were less focussed on preparing a Regulatory Package.
      Their commercial collaborators at AZ are tearing their hair out! but have to work with what they have.

  9. Barry says:

    If I were designing, or authorizing, or running such a Clinical trial, I would want to collect all possible data. That means not only self-reported symptoms but also serological assay (circulating IgG) and PCR assay (infection, symptomatic or not) and maybe gavage (looking for IgA on the mucosa) if the subjects would allow it. I’m incredulous that Pfizer and Moderna would have skipped these. More likely, they reported symptomatic cases only because they wanted to present the clearest narrative to justify FDA approval. Given the urgency, FDA was disinclined to ask for more.

    1. Ilya says:

      Pfizer and Moderna protocols do not require PCR testing of all patients, only those with COVID-19 symptoms are being tested.

  10. Aniket Naik says:

    Dr. Lowe,

    The news about India’s regulatory body declining early review of Oxford vaccine is fake news, the ministry of health has issued a clarification regarding the same, they say no such decision is yet made.

  11. daksya says:

    Given that the data indicates that LD/SD regime is *not* worse than the SD/SD regime, is there any harm in dosing with LD/SD anyway and hope for the best?

  12. theasdgamer says:

    Here’s the dope on covid hospitalizations and covid testing by hospitals, per two people who work for/in several hospitals.

    First, people who go to hospitals for non-procedural tests like blood draws, ct-scans, and ultrasounds aren’t tested for covid.

    Second, people who go to hospitals to get elective surgeries and procedures ARE tested for covid as a matter of admitting procedure. This is by far the largest cohort of patients.

    Third, people who go to hospitals for emergent care are tested for fever. If they have a fever, then they are tested for covid. Also, a physician may decide to test a patient for covid based on history, even when asymptomatic.

    1. Some idiot says:

      Thanks! Good to have some clear data.


      One question: which country/countries does this data relate to? I would guess that there _may_ be significant country to country variation, but it just a guess…!

      1. theasdgamer says:

        Botswana, lol.

  13. sPh says:

    If we do not learn in 60-90 days that on realizing the discrepancy in dosing and dose/response Oxford/AZ immediately opened an additional trial of 10,000 with the 30/100 dosing and with full support of their government, well, Gaia help the human race if a real emergency arises (really deadly pandemic, killer asteroid, etc). The lack of understanding of how to act in true existential crisis situations we have seen with COVID19 is more than a bit disturbing.

  14. cynical1 says:

    Okay, so there were three cases of transverse myelitis (one supposedly now being rediagnosed to MS by a “team of neurologists”) in a trial with 23,878 participants where they unblinded about half of them (11,636) for this analysis. 2 in the treatment arm and one with the other control vaccine. According to Wikipedia, the incidence of TM is 4.6 per 1 million per year. So, they had two absolutely confirmed cases in a cohort size which should have had virtually no chance of even having one. Riddle me that. That’s some serious bad luck. I’d like to see if that patient in the meningitis control arm magically has antibodies to the spike protein. (Call me suspicious.)

    The other thing is that this is based on interim analysis of roughly half of the participants enrolled in the trial. Assuming they unblinded those half in the same proportion, that means there is another half that will be unblinded in the future with some percentage getting the 1/2 does full dose regimen. When they combine those numbers will it have the statistical power to show that is superior?

  15. Mike Dee says:

    Note that the UK government will run a trial in January where they will combine one dose of the BioNTech BNT162b2 vaccine and one dose of the Oxford ChAdOx1 nCoV-19 vaccine as per the report in the Guardian linked below. It was not mentioned which vaccine will be the first, and which the second dose. This has potentially a number of positive effects:

    – Optimal immune system activation by a viral vector
    – No issues with immunity to the viral vector as only one dose is given
    – If two doses of ChAdOx1 nCoV-19 turn out not to be effective, but the combination with BNT162b2 works out fine, this allows double the number of people to get vaccinated

    2nd question: If I recall correctly, both BioNTech and Moderna modified the spike protein RNA so that the spike protein stays locked in the pre-fusion state. In contrast, again if I recall correctly, the Oxford vaccine uses unmodified RNA. Could this result in the protein flipping to the fusion state, making an immune response less effective to the actual virus?

    1. Riah says:

      potential for original antigenic sin possibly

      1. Mike Dee says:

        The RNA is identical though between BNT162b2 and ChAdOx1 nCoV-19 (except for the locking in the prefusion state, which is not relevant for the surface of the protein) so I don’t think that’s an issue here.

        1. Marko says:

          The RNA can’t be identical if one of them is engineered for the “locked” orientation. The coding for the modified amino acids is in the RNA sequence. And the “surface” of the spike is vastly different between the prefusion and postfusion forms , presenting a multitude of different epitopes for immune recognition depending on the form. See :

  16. Martin says:

    Here’s a question I’ve been wondering about:

    Is there any significant risk in receiving two different vaccines? Why not give people Oxford/AZ as soon as possible, then months or a year later when supply is available, give them Pfizer or Moderna?

    I’d rather have a decent vaccine sooner as long as I can still get the best one later…

    1. Mariner says:

      You’d suppose that the Sputnik V vaccine would provide some evidence in the risk of taking two different vaccines. After all, that’s what it fundamentally is.

      The question then would be how trustworthy the data is. The Russian government seems keen to use it as a propaganda tool – vaccine nationalism and all that stuff. Not that others aren’t doing the same, of course.

    2. Mike Dee says:

      The UK government will trial this next month. See my comment above.

      1. Mariner says:

        Yes, I’ve read that.

        My point is that, theoretically, at least, the Sputnik V trials carried out so far already test two different vaccines being used in tandem.

        Obviously not quite the same as the differing technology approaches of the AZ/Pfizer trial mooted.

  17. Chris Phillips says:

    “That control-group infection rate is definitely higher than what was seen in the two-full-dose group, which makes you wonder …”

    But the half-full group were reportedly vaccinated in April, before the full-full group. If that’s correct, I think a higher infection rate would be expected, as the half-full group would have caught the latter part of the UK’s first wave.

    “We need to add in to these calculations the news that the half-dose group included no patients older than 55, and to wonder what effect that had on the numbers, too.”

    They separated out the over-55s in the UK trial so that they could compare those up to 55. It seems there were just too few over-55s in the trial to have much effect on anything. There was only one symptomatic COVID-19 case in an over-55 in the whole UK vaccine arm.

  18. Academic Med Chemist says:


    The difference in infection rates in the placebo groups for the half vs full dose groups may be an issue of time. I think I read somewhere that the half doses were given first, and you’d expect patients that have been in the trial longer to have more diagnosed cases.

  19. Doug Feinberg says:

    I have no background in science or statistics. With that disclaimer I begin: Will the vaccine be 45% effective? Ninety five percent effective? It seems like both percentages are plausible given the data.

    However, I’m most interested in the lack of any patients having to be hospitalized for having COVID 19. Assuming that those statistics accurately represent a very low risk of bad illness for those vaccinated, then that takes away most of COVID 19’s danger. If COVID 19 truly was no more dangerous than the flu, then we could stop most of social distancing and mask wearing

    1. a s says:

      Getting a flu is still quite unpleasant and bad for productivity. Do people just think it’s mild because they’ve confused it with a cold?

      COVID is more infectious than the flu and mild cases can leave you with Long COVID even if they don’t hospitalize you, so you’d still want to take precautions. Even if we return to normal like Asian countries have, I hope we’ll remember to wear masks like they do.

      1. DataWatcher says:

        Sorry, but the specter of living the rest of our lives in a world of “faceless people” who can’t smile at one another or share empathy through facial expressions represents a dystopian nightmare that I find unbearable.

        Not being able to see one another’s faces in everyday interactions and encounters is extremely alienating — it’s “emotional distancing” just the same way as staying six feet from other people is so-called “social distancing” — and I think it makes people feel even more lonely and isolated (even if we’re not always consciously aware of it). “Every face tells a story” is a lot more than just an expression: One of the deepest, most meaningful, and essential aspects of our life is the visceral awareness that we’re surrounded by human stories all the time, and we “read” these stories through seeing people’s faces.

        We certainly know that a significant component of human communication — including expressing and sharing empathy — derives from facial expressions. Yes, “the eyes are the window to the soul,” as the saying goes, but only in conjunction with the rest of the face; they’re only part of the whole picture. A human face cannot be bisected and compartmentalized that easily. And it’s not just the individuals whom we know, or with whom we interact on a regular basis (in other words, “recognition” is only part of the story), but also the faces of passers-by, people sitting on the bus or the train, someone handing us a cup of coffee from behind a counter — the list goes on. It’s how we connect, part of what makes us know that we’re all members of the human family, the human community, the “stories” and lives all around us. Without this. an inextricable part of our humanity withers and dies.

        And, of course, in the context of COVID, the masks send another, more damaging message: They are barriers we erect between ourselves and others, because we fear one another’s physical presence, as well as our own, as potentially toxic. They are a constant, if by now almost subliminal, reminder that Jean-Paul Sartre’s axiom, “Hell is other people,” has reached its apotheosis: Disease is other people; pestilence is other people; contagion is other people (as well as, potentially, ourselves). This takes an additional toll on our social, mental, and emotional well-being.

        People can stay “distanced” from each other, not hug or touch each other, and not be able to see each other’s faces, for just so long. Human beings are social animals; it’s part of our makeup. We are bred for “social gathering,” not “social distancing.”

        A society of “people without faces” avoiding one another, quite literally, like the plague, is a stressed (and lonely) society. Without the glow of empathy that comes from “face-to-face” encounters and communication (there’s a reason we use that term!), the glow of our humanity is extinguished, as well. Not all toxins are physical; nor is all death. It’s a very harrowing thing to consider that we could succeed in keeping our bodies alive, even as our souls wither and die.

        1. Fortranis says:

          OMG Shut up.

          I could care less about seeing a person’s face in the street. I’d rather not be dead.

  20. Ben says:

    “Basically, the more you believe the 90% number is a statistical fluke, the more you then believe that the real efficacy of the vaccine is probably in the 60-per-cent range.”

    I think this is the wrong way to think about the data. Even if you believed the 1/2 + 1 dose in no better than the 1+1, there’s no reason you should throw away the data from the 1/2 + 1 study. You should take the combined result, which suggests ~70% effectiveness.

  21. Tony M says:

    Actually No. If you look at the LD/SD results for 18-55 years, you get quite a significant p value for the 90% efficiency value (95CI 67.3% to 97.0%). The combined result is not only less efficient but also less certain 70.4% (95% CI 54.8% to 80.6%). The SD/SD has a efficiency rate of 62.1% (95% CI 41.0% to 75.7%). They really need more data. However, you could consider approving the LD/SD just for the 18-55 Year group while gathering the extra data. (Refer tables in Lancelot report on pages 7 and 8 ).


    1. Duncan says:

      I was wondering about that. Is there any reason that everyone in a given population has to get the same vaccine? I think that there are several different flu vaccines each year aren’t there?

      So if (say) the Pfizer one works effectively in older people we could use that one on the older populations and if the AZ one is effective in under 55s then we use it in that population. Or does mass vaccination not work that way?

      I suppose that also there is a broader question now around whether to start ‘writing off’ some projects and to now focus energy and resources – perhaps we are not at that point yet, but we have to think about it. I think that we always knew that there would be some winners and some would fall by the wayside?

    2. Ben says:

      Sorry, was this in response to me? I didn’t make any claim as to whether the 90% should be trusted or not. I’m merely saying that if you believe the 90% number is a statistical fluke then it’s probably because you have a strong prior belief that the LD/SD treatment should be no better than the SD/SD treatment. If you think they should be about equally effective (given the data, it seems tough at this point to believe the SD/SD treatment is significantly more effective), and therefore believe the 90% is a statistical fluke, then I think you should be looking at the combined results (~70% effective) and not just the SD/SD results (~62% effective). I can’t see a reason why you’d want to throw away the LD/SD results, unless you believed the treatments were significantly different, in which case you likely wouldn’t “believe the 90% number is a statistical fluke”.

      Whether or not you believe the 90% number seems highly dependent on your prior beliefs about how likely the LD/SD treatment is to be more effective than the SD/SD treatment (clearly if a study showed that eating skittles was 90% effective, we would dismiss that as a statistical fluke, since our prior belief is that it’s so unlikely to be true), and I’m not nearly educated enough in this field to have a strong opinion on that.

      1. Andrew Kewley says:

        I don’t believe the 90% efficacy claim, because I have prior beliefs about how much of a mistake it is to cherry pick results in clinical trials. (after reading thousands of phase 2 and 3 trial reports in journals over the last decade.)

        Fact is, the 90% result had a very wide confidence intervals that overlapped with the 62% result. There were also significant population differences (no people over 55) for the 90% result.

        In total, I do not believe the 90% result is generalisable.

  22. aairfccha says:

    Meanwhile, a vaccine candidate developed at the University of Queensland has a bit of a problem – it generates false positive HIV tests – and has essentially been dropped.

    1. aairfccha says:

      Also bad news for the GSK/Sanofi candidate, at the moment it’s not effective enough in older people.

      1. JS says:

        Well, if the vaccine works well in younger people and prevents transmission (big unknown), then it would still seem to be useful. Might a test for that be worthwhile given how far along they are?

    2. Duncan says:

      The way I read that article is that it sounds like the Australian government wants now to put money, energy and resources into the successful vaccines rather than funding more projects – I suspect that is a sign of things to come now. And, of course perhaps that’s no bad thing. I think we all knew that governments were not going to fund dozens of very expensive and risky projects indefinitely. I guess if you have Sanofi and GSK resource levels then you can carry on despite setbacks but that’s not everyone!

      Terrible for the people at Queensland given that it sounds as if they felt the product worked. Presumably the false positive HIV tests would mean that HIV screening would then be rendered useless? Implications for blood donation too?

  23. Fernando Pablo Espósito says:

    “Undiagnosed multiple sclerosis”. MS is diagnosed by an exclusion criteria and you have to have at least two neurological episodes to start considering you have that disease. At least for the RR type.

    It looks fishy.

  24. Dylan says:

    These adenoviral vaccines have been on my mind for the last few weeks. We are calling this a vaccine, conflating it with traditional attenuated vaccines, but really, the adenoviral vectors in particular are a _DNA-based gene therapy_ that we’re going to prophylactically inject into 4.5 billion ++ people. On it’s face, this is pretty wild, but the more I think about the risks, the more confused I get here..

    How is nobody talking about the risk of insertional mutagenesis for all of the adenoviral vaccines? There is growing evidence that ‘episomal’ vectors can integrate in places that appear to be bad news for the risk of oncogenesis (AAV in particular, granted, a slightly different viral vector – see recent trials from 2020: And empirically, I’m sure anyone selecting for stable transformants with a plasmid would know that ‘episomal’ vectors don’t have a habit of staying that way.

    AstraZeneca’s regulatory filings with the EMA are particularly egregious – “it is known that insertion does not happen with adenoviral vectors, and therefore, no safety studies are required to assess this risk, the risk of germ line transmissibility, or carcinogenic risks [sic]”.

    I have searched for hours to find the evidence of an absence of risk (pertaining to insertional mutagenesis) – but all I can find is that there is rather, an absence of evidence on this topic. The EMA documents cite a review that has no link to primary findings that demonstrate this fact. I would imagine that I must be missing something.

    Can anyone point me in the right direction? What am I missing here? Where is the solid evidence that adenoviral vectors don’t integrate?

    1. DaveS says:

      I too have searched literature databases for many hours, unsuccessfully seeking evidence that a mutagenic/ carcinogenic risk has been thoroughly excluded for adenovirus vector vaccines.

      Unfortunately my concerns were further enhanced by the following reference, to which the Abstract concludes “We provide evidence that the unusual transforming activities of the adenoviral oncoproteins may be due to their mutagenic potential. Our results strongly support the possibility that even tumors [sic] that lack any detectable virus-specific molecules can be of viral origin, which could have a significant impact on the use of adenoviral vectors for gene therapy”
      DOI: 10.1128/JVI.75.7.3089-3094.2001

      I must stress that the above paper (published in 2001) is not discussing the ChadOx1 strain. But the suggestion that some strains of adenovirus can be carcinogenic surely means that we need explicit evidence that this risk does not apply to the strains used in adenovirus vector vaccines?

      Like Dylan, I am left thinking, what am I missing? Am I using a poor literature search strategy and overlooking key relevant papers? Am I misunderstanding something?

      1. Derek Lowe says:

        Keep in mind that the proteins that this reference talks about (E1a and E1b) are necessary for adenovirus replication. The adenovirus vaccines have these deleted to make them replication-incompetent. A lot has taken place in this field since this twenty-year-old publication.

  25. exGlaxoid says:

    I see that the UK has already approved the AstraZeneca Vaccine. Also, they have advocated giving the first dose (of both vaccines) to as many people as possible, and then giving the second dose one a slower schedule, given that more than 3 or 4 weeks should still work fine, and one shot of either vaccine appears to give a pretty good protection, so giving more people some protection should likely keep the most people from getting seriously ill or hospitalized, compared to doing both shots exactly at the 3 or 4 week time. I have to say that seems quite reasonable to me.

  26. Ann Scott says:

    Sorry, a very basic question. If a vaccine is, say, 60% effective, does that mean everyone who receives it is 60% more likely to be protected? Or does it mean that 40% of recipients are not protected at all? If the latter, how would you know if you were one of the unlucky 40%? Are there follow up tests to find out who is unprotected and why?

    1. Derek Lowe says:

      Not a bad question at all. Vaccine efficacy is measured by comparing the vaccinated group with unvaccinated controls – if both groups have the same number of infections, the VE is 0%, and if there are flat zero infections in the vaccinated group, the VE is (theoretically) 100%. Statistics will put confidence intervals around both of those, of course, for a given number of patients and cases, but that’s the idea.

      So a vaccine that is 60% effective means that out of an unvaccinated group of (X thousand) people where you would have expected to see 100 cases, you will only have 40 if you vaccinated everyone. So the final numbers are as if you vaccinated 60% of them with a 100% vaccine and left the rest to fend for themselves unvaccinated – but in the real world, what it means is that people had a whole range of immune responses to the vaccine, but only 60% of those on average were enough to fight off the virus in the conditions the trial participants were living under. That last part is one of the reasons you want tens of thousands of trial participants, to average out that number across all sorts of living conditions and types of patients.

  27. Bjorn says:

    @Dylan – thank you for flagging this. I have been looking for the same, anything on the likelihood of chromosomal integration of Adenovectors.

    There is one paper that I did find – it seems that host integration does happen in vivo, although gene inactivation is calculated to be at a low rate compared to the occurrence of random mutations. However the authors highlight additional mechanisms through which these insertions could pose a risk.

    Chromosomal integration of adenoviral vector DNA in vivo

    I am equally stunned nobody is talking about this particular issue.

    1. sgcox says:

      In that study they used vector for murine FAH which is highly homologous to human gene. Rare integration events observed were due to recombination. No such worry for vaccines caring viral antigen.
      We all get infected constantly by adenoviruses who invade our cells and replicate there. No problems with potential integration have been observed and reported after billions and billions of infections.

      1. Bjorn says:

        Thank you for that helpful context.

      2. Davide says:

        Dear sgcox,

        I thank you for your previous response. I had the same doubt of Dylan and Bjorn. I also thank you in advance for the response to my questions.

        Is it correct to compare what happens with wild-type adenoviruses and what happens with a modified one? Don’t they contain some part of the genome useful for the apopthosis of the host cell, which would make diseappear any effect of a possible insertion? Isn’t such part of the genome removed from the inoculated vector used in the vaccine?

        1. sgcox says:

          Dear Davide,
          Adenovirus vaccines, J&J, Oxford/AZ. Sputnik, etc. all have essential E1/E2 genes removed. It stops virus particle multiplying and killing the host cell. It simply produces its payload and quietly disappear. Transfected cells stay alive and present Spike to immune system. Basically the same mechanism as mRNA vaccine, just a different way to deliver Spike RNA instruction inside cells.

          p.s. Vaccine is produced in lab/factory by specifically engineered cells complemented with E1 and E2. Deficient adenovirus vector can multiply in these cell lines but cannot in normal cells and does not magically acquire missing genetic instruction. And yes, it has been rigorously tested in hundreds of labs and companies by now.

  28. sgcox says:

    Hi Davide,
    I tried to answer your question yesterday but my comment disappeared 🙁
    Derek already mentioned above that adenovirus vaccines are engineered to not able to replicate in human cells (deletion of essential E1/E2). Once it transfect human cells it simply produces its payload, Spike protein which is exposed on cell surface and alert immune system. No viral particles are produced and no cell lysis or apoptosis.

    Vaccine is produced in vitro is cell lines made to produce with missing viral proteins and thus suitable for vaccine production. Very common and now not new technique used for multiple purposes in biotechnology.

Leave a Reply

Your email address will not be published. Required fields are marked *

Time limit is exhausted. Please reload CAPTCHA.

This site uses Akismet to reduce spam. Learn how your comment data is processed.