AZ Will Sideline ‘Lucky Mistake’ To Secure COVID-19 Vaccine Approval
US Unlikely To Accept Data
Executive Summary
A lower first dose boosted the vaccine’s efficacy result – but AstraZeneca has conceded this has to be proven in a separate trial.
AstraZeneca and Oxford University are set to sideline the ‘lucky mistake’ data which boosted the Phase III results of their COVID-19 vaccine earlier this week, after facing criticism about how the study was performed and doubts about the robustness of the interim data.
The partners are now emphasizing the lower 62% efficacy level from their trials in patients who received both high doses – this is less than the 70% overall efficacy that was initially announced, but still a level that they hope can gain rapid approval from the UK and European Union regulators.
That is a rapid unravelling over the course of the week, which began with the partners presenting the interim results from the vaccine, named AZD1222, with some fanfare on 23 November. (Also see "A Stroke Of Luck Boosts Oxford/AZ’s “Vaccine For The World”" - Scrip, 24 Nov, 2020.)
The results were based on pooled data from two separate Phase III trials conducted in the UK and Brazil in a total of 11,636 participants. Included was a subgroup of 2,741 participants who received a low dose in the first of their two injections (now dubbed the low-dose/high-dose regimen) and which produced an efficacy rate of 90%.
These data were as presented as an unexpected highlight of the studies, and when pooled with the standard dose data, lifted the overall efficacy rate from 62% to 70%.
Sir Mene Pangalos
However, since then many external experts have voiced doubts about the robustness of the data, and questioned the clinical trial practices of Oxford, following the revelation that the low-dose regimen was the result of an error – a lucky mistake or “serendipity”, in the words of AstraZeneca’s research chief Sir Mene Pangalos.
AstraZeneca has now confirmed it will investigate the low-dose/high-dose regimen in a separate trial, focusing on the double high dose results in its filing. No further details of this study are yet available.
Pangalos had originally said the company would seek to amend the protocol of its ongoing US trial, but this plan was abandoned as it became clear it would only muddy the waters.
Defending the data and clinical trial protocols, he said the 62% efficacy achieved with the standard high-dose/high-dose regimen meant the vaccine would play a major role in combating the pandemic.
Speaking at the Faculty of Pharmaceutical Medicine (FPM) virtual conference, Pangalos told Scrip that the fact that the trial showed no cases of severe COVID-19 was the most crucial result. “I think this vaccine is clearly highly effective and has 100% efficacy with regard to people getting severely ill, hospitalized or dying,” he said.
“62% protective from developing any single sign of COVID disease is still a highly effective vaccine and I think this vaccine will have a big impact to global health.”
External experts are less convinced. Ian Jones, professor of virology at the University of Reading said: “62% means that four out of 10 people who get the vaccine at this dose will still get infected. That is not acceptable when other vaccines are showing near 100% protection.”
Commenting before the decision to remove the contentious subgroup, Jones said he did not believe the data merited an emergency use authorization in the US and equivalent elsewhere.
“Based on what I have seen, no. I think the dose thing is very muddled. Until it is sorted out the vaccine should be held until such time as the relationships among dose, recipient population and outcome are clear.”
During the FPM conference Pangalos also emphasized the vaccine’s advantages over the mRNA candidates, which require special refrigeration and will have fewer manufactured doses available in 2021. “Remember it is easy to distribute with 4°C cold chain (not -20°C or -80°C), inexpensive and we can manufacture 3 billion doses next year.”
AZD122 is especially important for the global fight against SARS-CoV2, as AZ and Oxford have agreed to supply the vaccine at the not-for-profit price of $4-6 a dose, and to maintain this price after the pandemic in less developed nations.
Sir John Bell
Also speaking at the FPM event was Sir John Bell, regus chair of medicine at Oxford, and who brought the university’s researchers and AstraZeneca together in April.
He defended Oxford’s trial practices and told Scrip: “All the protocols were pre-approved by the regulators, so we weren't cooking this up as we went along.”
He said all of the trial analysis done was pre-specified and emphasized that the 62% efficacy rate was well above the recognised 50% efficacy rate that regulators were looking for.
“I can't imagine any reason why the regulators won't accept that,” he said, though added that the US Food and Drug Administration would only see the full data once it was prepared for peer review, which has now been submitted to The Lancet.
US FDA Will Wait
The FDA looks unlikely to accept these UK and Brazil trial data, even with the contentious subgroup removed, because of remaining issues around trial size and design. Instead it is expected to insist on seeing results from a larger single study conducted as part of Operation Warp Speed.
“The question for us was, will we need the US data to get approval in the US or can we get approval in the US with international data, and it was never clear,” AZ’s CEO Pascal Soriot told Bloomberg. “Now with those results it’s more likely that we will need the US data.”
One consolation is that the company’s US trial could produce its own first interim readout before the end of 2020, likely to be fueled by the record high COVID-19 infection rates in the country.