UK SMC BRIEFING + EXPERT REACTION: Phase 3 Oxford/AstraZeneca COVID vaccine trial results

This is more good news for an exit strategy from the COVID-19 pandemic, that we have another apparently effective vaccine, vectored on a chimpanzee adenovirus. The best dosing schedule was a lower dose followed by a higher dose. With some vaccines we do see a blunting of the immune response with subsequent doses, and there could be a dose dependent effect here that creates optimal boosting when the first dose is lower in antigen content.

In other good news, the published phase 2 study of the AZ vaccine last week showed excellent immunogenicity in older people - although they excluded the “older” elderly and people who were very frail. The second dose seemed important in boosting and creating a more sustained antibody response in older people. The AZ vaccine has regular cold storage conditions (2-8 degrees C) which makes cold chain management more feasible, especially for remote and rural areas.
 
In terms of the 70% efficacy, this is derived from combining the data from two different dosing schedules. I would imagine they will select the more immunogenic schedule, so we would expect efficacy of 90% with that schedule. I would not put too much stock on the comparative efficacy between the different vaccines (Pfizer, Moderna, Astra Zeneca) which have provided efficacy estimates at this stage, because they would reflect very short term follow up - at the most, the only available data at this stage would be efficacy at 1-3 months post-dose 1, and no trial would have data beyond 6 months post-vaccination, because the earliest phase 3 trials commenced in April.

The more important question is, what is the efficacy at 12 months, 24 months and 36 months? We also need to know that efficacy endpoints are the same when comparing vaccines - for example, PCR confirmed infection, serologically confirmed infection, hospitalisation or death. The most important endpoints are prevention of hospitalisation and death. That will be a more important endpoint for assessing the vaccines.

Further, we do need to see the results of the phase 3 trials published in the peer reviewed literature before making valid comparisons.  Based on the preliminary, non-peer reviewed data, all three vaccines with their optimal dosing are >90% efficacious, which is great news.
 
There will be differences between the many vaccine candidates in efficacy and safety - and both of these will need to be considered. However, there are 7 billion people in the world, and all vaccines with good efficacy will be needed to ensure everyone in the world is vaccinated.

In the future, having multiple vaccines on the market is good for the global outlook of controlling COIVID-19.  In some people, especially those at high risk, it is even possible that they may not respond to one vaccine, but do respond to another.

From a marketing perspective each company highlights their competitive advantages, but whilst this vaccine does not need storage at -80°C, it does need storage in a fridge so may create challenges for remote and rural communities, but would certainly be easier than a vaccine that requires ultra low temperature storage and transportation.

Another positive announcement relating to vaccine safety and efficacy from late phase trials is welcome news, particularly given this is a different platform to the mRNA vaccines of Pfizer/BioNTech and Moderna that we have been hearing about mostly of late.

Enthusiasm from this announcement, as with the others in recent times, needs to be measured however given the full data is not yet available and the results are yet to be published in a peer reviewed journal.

This vaccine is a viral vectored vaccine and clearly has some favourable properties in that it does not require the extreme cold chain of the mRNA vaccines and can be manufactured at a modest cost. It is also additionally exciting given Australia has an agreement to obtain approximately 34 million doses of this vaccine (as compared to only 10 million of the Pfizer vaccine) and a significant proportion of that allocation will be manufactured in Australia.

It is interesting to note that two different dosing regimes were used in this study with two full doses administered one month apart demonstrating 62% efficacy and a half dose followed by a full dose showing an efficacy of 90%, thus the overall reports of 70% which, while not as high as the mRNA vaccines, is still relatively high.

Unfortunately the half followed by full dose arm of the trial is reported to have only thus far included 2741 volunteers so more information will be required to confirm such high efficacy of this regime.

Having said that, the data safety monitoring board determined that the primary endpoint of the study had been achieved and noted that no serious safety events were seen and that it was well tolerated. Hence, this vaccine is likely to be available for use relatively soon which is obviously great news.

It’s very encouraging to see that in the higher efficacy dose regimen, a combination of a half dose followed by a full dose resulted in 90 per cent efficacy against the disease. 

At a time where demand for vaccines far outstrips supply, using one and a half doses rather than two means Australia’s supply of the Oxford-AstraZeneca candidate will go much further.  

The interim data shows the vaccine reduced asymptomatic infections, and there were no severe cases or hospitalisations among those who received the vaccine. 

It’s also important to see that the vaccine demonstrated a strong immune response in older people, a key high-risk group in this pandemic. 

CSIRO is pleased to have been able to play a role in the validation of this vaccine through pre-clinical testing at our Australian Centre for Disease Preparedness earlier this year. 

As part of CSIRO’s preclinical trial of Oxford-AstraZeneca's candidate, our scientists evaluated the efficacy of one versus two doses as well as administration of the vaccine via a nasal delivery and/or an intramuscular injection. 

This is hopeful news as biotechnology company CSL started manufacturing the vaccine candidate in Victoria on 9 November 2020.  

Phase III trials are conducted in large populations in current disease outbreak sites where the virus is in high circulation. While it’s important to remember these are provisional results that haven’t been peer reviewed, these interim results bring us another step closer to defeating the novel coronavirus and seeing a safe and efficacious vaccine delivered in Australia and around the world.

This is another report of interim analyses disseminated by media release on the progress of a vaccine trial, this time the Oxford/AstraZeneca vaccine. As distinct from the novel mRNA vaccines that have been the subject of media releases over the past few weeks, this vaccine is based on an adenovirus viral vector that carries the genetic material for the SARS-CoV-2 spike protein.
 
The results are intriguing in that two different dosing regimens were assessed. Both dosing regimens involved the administration of two doses of the vaccine at least a month apart, but curiously, the one that involved a half dose followed by a full dose showed 90% efficacy and the other dosing schedule which involved the administration of two full doses showed 62% efficacy. The media release reports an average efficacy of 70% combining the results of both of these dosing schedules which they correctly equate with the vaccine providing good protection against COVID-19, and further, they indicate that there were no safety issues.
 
This represents another positive report to go with previous ones over the past few weeks. The advantage of this vaccine over the ones based on mRNA is that this one can be stored at ‘fridge temperature’ which provides a huge advantage in terms of vaccine delivery. We await the full analysis of the findings in a peer reviewed journal.

The positive news is that the Oxford trial data is at least the fourth human trial to report efficacy following two US studies on MRNA vaccine and Russian study on adenovirus vector vaccine, showing effective vaccines against COVID-19 are indeed feasible. However, whereas the two MRNA vaccines showed greater than 90% estimates of efficacy the Oxford adenovirus vaccine achieved only an overall point estimate of 70% effectiveness in preventing symptomatic disease and the main vaccine group had just a 62% estimated effectiveness in preventing symptomatic disease. Allowing for the statistical uncertainty around these point estimate values, this means the Oxford vaccine may have barely exceeded the 50% minimum effectiveness threshold set by WHO and other bodies for a COVID-19 vaccine.  This will only become clear when the actual data and statistics on the trial outcomes is made available.
 
Clearly the Oxford and Astra Zeneca teams are putting a brave face on this result, but overall it is a disappointing outcome when compared to the much greater apparent success of the MRNA vaccines. This outcome may have been predicted from the Oxford monkey challenge data where their vaccine only provided partial protection versus much more robust protection seen in monkey models with other vaccines including both nucleic acid and protein based vaccines.
 
There are many potential problems with regulatory approval of a vaccine with such a low protection rate. First, it may provide a false sense of confidence to people receiving the vaccine, who may think themselves protected and thereby engage in riskier behaviour increasing their risk of virus exposure. Second, it significantly increases the risk that the vaccine will itself drive the development of immune-escape virus mutants, accelerating loss of vaccine protection. Third, with a low starting level of immunity, any partial protection in some subjects is more likely to rapidly wane leaving vaccinated subjects again vulnerable to infection.
 
While the Oxford and Astra Zeneca teams have drawn comparisons of their vaccine efficacy to that of seasonal influenza vaccines, this is not an accurate comparison as the effectiveness of seasonal influenza vaccines is measured over a period of 12 months whereas the current Oxford effectiveness data covers a much shorter period that in many subjects is just a few months. Furthermore, influenza vaccines are not a good comparator due to the large number of diverse influenza strains these vaccines need to protect against and also the variable level of pre-existing immunity in much of the population. A more accurate comparison would be to vaccines against virus diseases such as measles where typically protection rates of at least 90% are expected. On such a comparison, the Oxford vaccine compares very unfavourably.
 
In respect to the subgroup analysis in the smaller group that had an estimated effectiveness of 90% this may just be a statistical anomaly, and it would be important to know if this was a prespecified primary endpoint and the confidence intervals around this point estimate which would be expected to be large given the small group size.  Before a vaccine could be approved based on such a small subgroup analysis a further new Phase 3 trial would need to specifically test this alternative vaccine strategy to confirm this is a true difference.