select search filters
roundups & rapid reactions
before the headlines
Fiona fox's blog

expert reaction to media reports about the Oxford AstraZeneca vaccine phase 3 trial, about dose and subgroups

Recent media reports have discussed dose and subgroups in the Oxford and AstraZeneca phase 3 vaccine trial.


Prof David Dunn, Senior Trial Statistician for vaccine trials, and Prof Sheena McCormack, Clinical Project Lead for vaccine trials, MRC Clinical Trials Unit at UCL, both said:

“The key result reported by Oxford University on the ChAdOx vaccine administered on two occasions is the overall 70% reduction in the rate of COVID-19.  This level of vaccine efficacy comfortably exceeds the target threshold set by international agencies at the outset of the pandemic.  We need to see the full set of data before judging whether the 70% reduction is consistent across different subgroups, including participants who first received a low dose of the vaccine.  The effectiveness of a vaccine in controlling this epidemic will depend just as much on distribution and uptake as it does on efficacy.  There are practical advantages that allow this vaccine to be distributed quickly once licensed – it is cheaper, only requires fridge storage, and several countries are already building manufacturing capacity.”


Dr Charlie Weller, Head of Vaccines at Wellcome, said:

“It is important to take a step back and remember that these are interim results, so there are still many unanswered questions.  Before drawing full conclusions, we must wait for the trial to finish and for the complete data to be independently and rigorously assessed.

“Safety is the most important consideration for any vaccine trial, which is why independent regulatory bodies like the MHRA and FDA will fully review the data and ensure vaccines are safe and effective before being rolled out to the wider population.  It is very encouraging to see that the University of Oxford/AstraZeneca reported no serious safety events related to the vaccine.

“Months ago, we would have been delighted to see any of the first Covid-19 vaccines reported 70% efficacy.  This would be hugely impactful for public health and highly effective in protecting against serious illness from Covid-19.  We must not lose sight of this.

“To protect the world from Covid-19 we need more than a vaccine: we need vaccination.  Crucially, the University of Oxford/AstraZeneca vaccine can be easily administered in existing healthcare systems around the world, stored at fridge temperature and can use existing delivery mechanisms, which gives it a distinct advantage for swift and successful global rollout.

“It is remarkable that in just 11 months we have multiple vaccines with such impressive interim results.  Producing the billions of doses needed to protect the world is our biggest challenge, so we will need a range of vaccines that work across different populations, age groups and settings.  We are racing the virus, and not each other.”


Dr Penny Ward, Visiting Professor in Pharmaceutical Medicine at King’s College London and Chair of the Education and Standards Committee of the Faculty of Pharmaceutical Medicine, said:

“As luck would have it Professor Pangelos presented at the Faculty of Pharmaceutical Medicine Annual Symposium today, so we were able to gain more insight into the trial and the dose question.  Dosing errors do occasionally occur with biological medicines and it is then better that the dose administered is lower rather than higher than intended.  In this trial the error was spotted early on, disclosed to the regulators and the pathway forward prospectively agreed.

“Many phase III trials with novel medicines would explore multiple dose regimens, particularly when the link between a surrogate marker for disease response (in this case neutralising antibodies to SARS CoV2 virus) and the disease pathology is uncertain.  Low doses in the recently reported PII trial were highly immunogenic in both younger and older people and similar to those induced by higher doses when measured 28 days after the second dose so it is surprising that both a lower and a higher dose regimen was not initially included in the phase III trial.  When PIII studies use multiple different dose regimens is usual to combine doses to explore overall efficacy as well as to then review efficacy in the dose subgroups vs placebo, and then against each of the other dose regimens.

“The data release described an interim assessment of an ongoing trial, so more data will become available as the trial proceeds during the ongoing outbreak(s) in the contributing countries.

“It is usual to combine data across trials in late stage development: this is justified when the outcome measures and analytical approaches are concordant, which they are in the case of AZD1222.  Regulators understand this, particularly the Food and Drug Administration who not only receive the company summary opinion of the results, but also the raw data, which the stats group at the FDA reanalyse to confirm the validity of the company analysis.

“The regulators across multiple regulatory arenas agreed that the target efficacy for prevention of disease was 50% with a lower bound of 30%.  This is based on the success of influenza vaccination in reducing hospitalisation and death from influenza – influenza vaccine does not prevent infection with influenza but it does reduce disease severity and this analogy was then applied to COVID.  In these terms, this vaccine met the required efficacy (exceeded it somewhat) at both dose levels.

“It would be an advantage in many settings to have a vaccine which can slide readily into the established vaccine supply chain, and not to have to invest in large capacity deep freezers as well as having to ship frozen.

“So speaking personally, and acknowledging that AZ could have done more to be transparent on these issues, this says more about AZs communications style, and should not detract from confidence for using this, clearly effective, vaccine.”


Prof Helen Fletcher, Professor of Immunology at the London School of Hygiene & Tropical Medicine, a vaccines expert and UKRI’s Director of International Development, said:

Comment on vaccine dosing in the AZ/Oxford trial

“Every vaccine is unique – you can’t pull something off the shelf and repurpose it – each vaccine has to be developed from scratch so you’re almost back at square one when a new pathogen emerges.  Scaling up the manufacture of a new vaccine from a small amount produced in a research lab to the large amount needed for human trials is hard and it can sometimes takes years.  What’s remarkable is that Oxford University and AstraZeneca have gone from square one to 100 million doses in less than a year.  It’s not surprising if some manufacturing issues were still being ironed out when they started clinical trials but early stage trials are all about safety and the safety data we have seen has been very robust.

Comment on efficacy results

“I’m really looking forward to seeing the data from all of the Phase III vaccine trials.  I work on tuberculosis vaccines and we see variation in vaccine efficacy by several factors including age, previous exposure to mycobacteria, geography, co-infection and co-morbidity.  It would be remarkable to me if similar factors didn’t impact COVID-19 vaccine efficacy.  The important thing is that it’s normal to control for these factors in the design of clinical vaccine trials and in the interpretation of trial data.  This is why it’s important to wait for the full data set.  I also think it’s possible that a lower initial vaccine dose could result in higher vaccine efficacy.  My group has been looking at this with mathematical modellers at LSHTM for several years.  More is not necessarily better when it comes to vaccines and immunotherapies.  You need to stimulate an immune response but if you push too hard you can get hit a negative feedback loop where the immune system seeks to control and dampen down the response.  It’s also possible that a strong immune response to the first vaccine could effectively block an immune response to the second shot of the same vaccine – but we should be able to see evidence of that when the data is published.”


Prof Stephen Evans, Professor of Pharmacoepidemiology, London School of Hygiene & Tropical Medicine, said:

“The data released so far do not give any details about different countries or dosing regimes.  The protocol originally registered in May 2020 gives details of all arms and interventions used.  There have been a series of amendments to those – the idea that no protocol has been published is simply not true.  AstraZeneca published the full protocol after one of the early press releases before the phase 3 data.

“The variety of age groups and dosing regimes is a feature of a number of the vaccine trials and not just the Oxford AstraZeneca vaccine.

“The way the data are put together will have been specified in the protocol and scrutinised very carefully by regulators to ensure that there is no “cherry picking” of the results.

“We have good grounds for trusting that the regulation in this high profile area will be done as carefully or more carefully for these vaccines than for any others in the past.

“Most definitely there are many things that we still need to find out and we will have to wait for full publication.

“The regulators however will have already looked at the lab data, the animal studies and the early phase trials.  They will also have much more data than ever appears in publication.  So they can understand how the trial was conducted and what all the results are.

“In general, and this applies to all the COVID vaccine trials, this is not the ‘normal’ process of authorising a new vaccine – but we are in the middle of a pandemic where there are no vaccines that have been available for this disease.

“What is happening is that the Data and Safety Monitoring Boards (DSMBs) are looking at the data knowing the treatment allocation, which the investigators did not know.  When the trial data indicate that the vaccine is likely to meet the criteria for authorization of use then they have to inform the investigators and the companies.  At that stage because there could be leaks of confidential information which might result in insider trading, the sponsors and investigators have to put out a press release with minimal information because of stock market law.  This of course is not ideal from a scientific point of view but has become necessary in this pandemic.

“In terms of the Oxford vaccine, it is very clear that they have included some elderly people in particular subgroups and may have recruited more over 70s than other trials.

“We do not know the results of these, and speculation about the age distribution in different groups beyond what is clearly detailed in the protocol at is not useful to anyone.

“The choice of an active vaccine to give as the control may be a requirement for ethical reasons.  In locations where that is not required then a placebo may well be used.  This is not expected to have any noticeable effect on the results for efficacy.”


Dr Joy Leahy, Royal Statistical Society Statistical Ambassador, said:

“The Oxford/AstraZeneca vaccine information that we have so far represents an important step in the battle against COVID-19.  The initial results do not appear to be as optimistic as the Pfizer or Moderna vaccine.  However, the Pfizer and Moderna vaccines produced much higher results than expected, and had the Oxford/AstraZeneca vaccine data been released first, I believe they would have met the expectations of the scientific community at the time.

“Everyone, including the scientific community is rightly seeking more data in the form of a peer reviewed paper.  While a press release isn’t the ideal method for scientific analysis, with the vaccine being so highly anticipated, and with financial markets being so sensitive to COVID-19 news right now, it wouldn’t be possible to wait until for a full paper to be prepared and published.  Too many people would have knowledge of the results before they go public, which would increase risk of market malpractice.

“I am personally, eagerly awaiting more data on this vaccine and information on the analysis.  The dosing error does throw up some problems.  Taking any trial, the more different dosing regimens that are in the trial, the more likely it is that any one regimen will turn out to have a particularly high efficacy, just because of chance and random variation.  Trial investigators can’t just pick the best performing regimen and look for regulatory approval for that particular dosing schedule.  If this was the way that medicines were licenced, then we would have a lot of medications being taken that are not the true best performing dose.  Of course, regulators do have statistical methods to account for issues such as these, but as with any statistical methods, there’s only so much that can be done after the fact.

“A major issue in this case is that there was no clinical reason in advance to suspect that the half dose followed by full dose regimen was likely to be more effective, so it does seem like that could increase the likelihood of the high results for this dosing regimen being due to random statistical variation.  There are also some other complications in looking at this subgroup given that they were all in one location and apparently not representative of the full age profile in the trial.  In addition, I would like more information on what subset of the comparator group this regimen was compared to; e.g. was it only compared to the group in the UK who had received the comparator up until that point? As the COVID-19 pandemic has been a rapidly evolving situation, which has been highly dependent to local responses at that point in time, it is very important that like is compared with like at all times.  When doing an analysis on all the patients in the trial across time frames and countries, this should not be an issue as both the vaccine and the comparator have been administered under the same circumstances.

“Even with all these in mind, the available data indicates that the Oxford/AstraZeneca vaccine is quite likely well above 50% effective, which was previously indicated as a threshold for which regulators will be looking.  I for one am feeling much more positive about the prospects for 2021 then I was three weeks ago!”


Prof Eleanor Riley, Professor of Immunology and Infectious Disease at the University of Edinburgh, said:

“These allegations are worrying.  Whether or not they are justified, and whether or not they affect the validity of the data released earlier in the week, they need to be answered clearly and completely.  Trust is at a premium when it comes to vaccines and we must not  do anything that might in anyway undermine that trust.”


Dr Gillies O’Bryan-Tear, Chair, Policy and Communications, Faculty of Pharmaceutical Medicine, said:

“The Oxford/AZ phase 3 interim results have only been communicated in a press release, so it is hard fully to interpret the results of the subgroups until we have the full publication.  That publication is imminent, according to informed sources, so we may only have to wait a few days.

“Based on what AZ and Oxford have said, and also on the study design information available on the publicly available web site,, where researchers post their studies, we can infer that the study was designed originally with only the high dose/high dose vaccine, compared to a control vaccine, Meningitis C in Brazil, the UK and South Africa.  This study continues to enrol, in multiple countries – up to 60,000 – so more data will become available.  Given the efficacy we have seen so far, both in the low dose/high dose group and the high dose/high dose group, it is almost certain that robust efficacy – the exact level remains to be seen – will be shown in both groups, well above the benchmark 50% which was set as the hurdle.  Indeed, the AZ press release stated that there were strongly significant efficacy results in both dose groups (respectively, 90% and 62% in the low dose/high dose group and high dose/high dose group).

“With respect to the regulatory authorities, AZ has said that the regulatory authorities were fully aware of this and were in agreement with the analysis of the subgroups.  They have said that until the full data are reviewed, it is not possible to know why the apparently different results occurred in the two groups.  The immune responses were stronger, in the earlier studies, in the high dose/high dose recipients than in the low dose groups.  Thus the validity of the low dose/high dose group results are unlikely to be questioned, though the eventual efficacy rate may change.  Even if they decide to ignore the results of the low dose/high dose group, the study of the high dose/high dose patients will still be strongly significant, but I think that is unlikely.”


Prof Peter Openshaw, Professor of Experimental Medicine at Imperial College London, said:

“From what I have read, it seems that the calculation of the dose in the UK limb of the trial was out by a factor of 2 and that this resulted in the half dose, followed by full dose limb of the trial in the UK.  It has been reported in the media that this only included cases below the age of 55 – if this is true, it may mean we don’t have any information about this regimen in older adults.  We have to wait for the full data and to see how the regulators view the results of the phase 3 trials.  The US and European regulators might possibly take a different view.  All we have to go on is a limited data release.  The protection from the Oxford AZ vaccine may be less than that from the RNA vaccines, but we need to wait and see.

“It is remarkable that each of the trials that are now reporting shows protection, which we did not know was going to be possible.  We have been wanting vaccines for many diseases for a long time and they haven’t arrived – HIV, TB and malaria being good examples.  The results so far seem to show that it can be done for COVID-19, and that’s very good news indeed.”



All our previous output on this subject can be seen at this weblink:



Declared interests

Prof David Dunn and Prof Sheena McCormack: “We are working with Imperial to assess their COVID-19 vaccine candidate.”

Dr Charlie Weller: “Wellcome co-founded and co-funds CEPI, which has provided support to the development and manufacture of the University of Oxford/AstraZeneca AZD1222 vaccine candidate.”

Dr Penny Ward: “No COIs. I am semi-retired, but I am owner/Director of PWG Consulting (Biopharma) Ltd a consulting firm advising companies on drug and device development.  Until July 2019 I was Chief Medical Officer of Virion Biotherapeutics, which was a company developing broad spectrum RNA therapy for the treatment/prevention of respiratory virus infections.  Between December 2016 and July 2019 I served as Chief Medical Officer of Virion Biotherapeutics Ltd, a company developing antiviral treatments for respiratory viral diseases.  Previous employee of Roche, makers of tocilizumab (anti IL6 antibody) and CMO of Novimmune, makers of empalumab (anti IFN gamma antibody).”

Prof Helen Fletcher: “I’m Trustee of the Jenner Vaccine Foundation.”

Prof Stephen Evans: “No conflicts of interest. I am funded (one day per week) by LSHTM. They get funding from various companies, including Astra Zeneca and GSK but I am not funded by them, I have no involvement in obtaining funding from them and I am not an investigator on any grants obtained from them. I am the statistician to the ‘meta-Data Safety and Monitoring Board’ for CEPI. I am paid for my attendance at those meetings and will be paid expenses for travel if that occurs.”

Dr Joy Leahy: “I have no conflicts of interest to declare.”

Prof Eleanor Riley: “Prof Riley is a member of the UK Vaccines network and the UKRI Covid-19 Taskforce.”

Dr Gillies O’Bryan-Tear: “No conflicts.  Former Head of Vaccine Clinical Development, GSK.  I am a pharmaceutical physician, semi-retired and have worked in a variety of fields, including oncology (cancer) drug development and vaccines development.  I have not been active in vaccine research for many years and have no current interests in or conflict with any of the Covid-19 vaccine research programmes.”

Prof Peter Openshaw: “Peter Openshaw has acted as consultant, panel member or scientific advisor to GSK, Janssen (J&J) and Pfizer.  These have not been specifically on COVID, but involve vaccine and antiviral compounds that might be effective against RSV or influenza.”

None others received.

in this section

filter RoundUps by year

search by tag