by Reynald Castaneda in London
Pfizer /BioNTech’s final Phase III efficacy analysis of its Covid-19 vaccine BNT162b2 is likely to show a decrease by around five to ten points from the initial efficacy data of 90% symptomatic protection. A drop would be expected as the trial proceeds from interim to final analysis due to changes in volunteer demographics and waning efficacy.
However, if the final data reports the vaccine is 50% effective, though worthy of an emergency use authorisation (EUA) from the FDA, the 40-point decrease would be a red flag to the vaccine’s durability. Yet, such a dramatic drop is unlikely because the initial data is high enough that the impact of potentially lower efficacy from forthcoming events would be diluted.
Considering the interim analysis is based specifically on volunteers without prior evidence of SARS-CoV-2 infection, there is division as to whether vaccine receivers would need to be tested to screen for prior infection. While a prior infection could impact vaccine efficacy or pose hypersensitivity to more severe symptoms, these concerns remain theoretical in nature.
Initial data from the Phase III BNT162b2 trial (NCT04368728), released via a media release on Monday (9 November), captured the world’s, as well as the market’s, imagination in combatting the ongoing Covid-19 pandemic, triggering 7.7% and 14% share price increases for Pfizer and BioNTech, respectively. BNT162b2 is likely to be granted an EUA by the FDA based on available data along with upcoming safety results, with the minimum bar being 50% efficacy, experts said. Pfizer has a $213.33bn market cap, while BioNTech’s is $26.27bn.
Efficacy decrease to 80%–85% still relevant, though 40-point drop would raise questions
It is encouraging that the interim analysis showed 90% efficacy to prevent symptomatic Covid-19, as it allows room for some degree of reduction once the final data is reported, said Phase III principal investigator Dr Stephen Thomas, chief, Infectious Disease Division, New York Upstate Medical University. The interim analysis occurred with 94 cases and the nearest efficacy threshold, per protocol, was 62.7% for 92 cases, and this has been exceeded, a Pfizer spokesperson said. The next and final efficacy analysis will occur when there are 164 confirmed Covid-19 cases, and the final efficacy may change, the media release stated.
According to statistical modelling based on 164 events, the vaccine’s final efficacy data could decrease to 85% or potentially down to 80%, said Kert Viele, PhD, senior statistical scientist, Berry Consultants, Lexington, Kentucky. Such a reduction would be due to a change in demographics of the trial volunteers who contracted the virus, explained Thomas. While the media release has no patient profile information for the 94 cases, Dr Paul Goepfert, director, Alabama Vaccine Research Clinic, Birmingham, noted it is possible that most events from the interim data were in younger people, considering the ongoing Covid-19 case increase has been largely driven by younger people.
Younger vaccine receivers may experience more efficacy with the vaccine versus older volunteers, Thomas added. Phase III is recruiting participants 16 to 85 years. The spokesperson said the study remains blinded until the final efficacy analysis, including to the split of cases.
Waning vaccine efficacy would contribute to reduced efficacy in the final data, Thomas added. In Phase III, a Covid-19 infection event is counted seven days after the second vaccine administration. Sampling variability could also account for vaccine efficacy percentage reduction, noted Peter Gilbert, PhD, biostatistician, Fred Hutchinson Cancer Research Center, Seattle, Washington.
That said, if efficacy drops from 90% to 50%, which is FDA’s EUA bar, waning vaccine efficacy would be the potential explanation, said Gilbert. There is zero possibility that a 40-point drop would happen due to chance, Viele said. This would underscore the vaccine’s lack of durability, he added.
It is possible that in the first 94 events, the vaccine is 90% effective, and in the forthcoming 70 events, the vaccine would be 60% effective, which would be an average of 70% efficacy, Viele said. But an average of 50% overall would be unlikely, as this means the 70 events would demonstrate the vaccine does not work at all, he explained.
If efficacy goes down to 50% for a certain subgroup, like older people, the vaccine would still have utility given their higher risk of severe disease, noted Mark Jit, PhD, professor, vaccine epidemiology, London School of Hygiene and Tropical Medicine, UK. This news service reported 1 October Pfizer/BioNTech’s trial is designed to report interim data earlier and has potential for a broad label, yet there is uncertainty about how to best use the vaccine given interim data would lack detail.
Prior infection testing could avoid safety concerns but not all expect this requirement
With Monday’s data being based on participants without evidence of prior SARS-CoV-2 infection, it could raise the possibility that vaccine receivers will need to be screened to access the vaccine, noted Dr Maria Elena Bottazzi, associate dean, National School of Tropical Medicine, Baylor College of Medicine, Houston, Texas. There are theoretical vaccine safety concerns lingering, particularly in people who already have antibodies against SARS-CoV-2, said Bottazzi and Dr Scott Halstead, professor, Uniformed University of Health Sciences, Bethesda, Maryland.
Hypersensitivity is an unanswered concern, which could lead to vaccine receivers having more severe manifestations of disease, Halstead explained. This risk of hypersensitivity underscores the need for the vaccine to maintain a 90% efficacy with more data, as such a hypersensitivity reaction would tip the cost-benefit scale negatively, he added.
Yet, Thomas differed, noting it is unlikely for BNT162b2 to require any test before being accessed. There is limited evidence of the potential for hypersensitivity, he added. In fact, other coronaviruses already in the community do not seem to play a part in enhancing Covid-19 potency, he added.
Additionally, in people who have already been infected, the vaccine could potentially be more efficacious as it would boost their pre-existing immunity, Thomas and Goepfert explained. Jit, however, noted there are cases where antibodies generated from a prior infection could reduce vaccine efficacy, but he nonetheless said a test is unlikely, as it would increase the cost in accessing the vaccine.
In the Phase III, a low level of baseline symptoms triggers a test on the day of injection to check for prior infection, Thomas explained. According to the protocols, people with previous clinical or microbiological Covid-19 diagnosis are excluded, though the media release noted the trial will evaluate the vaccine in those who have had prior exposure to SARS-CoV-2.
Antibody tests used to check for prior infection are not accurate, noted Jit. Even if the vaccine receiver has antibodies for SARS-CoV-2, it is still unclear what this means, Halstead said. For example, the antibody could be caused by similar coronaviruses already present in the community, he added.
The trial has more goals to achieve, like finding out if a participant was infected prior to the trial, if the participant was reinfected and if there is different efficacy in seronegative and seropositive people, the spokesperson added.
EUA likely for BNT162b2 but other vaccines shouldn’t be counted out if lower than 90% efficacy
Phase III is continuing to accumulate safety data, and after there is a median of two months’ worth of safety results after the booster shot, the vaccine will be submitted to the FDA for an EUA. BNT162b2 is likely to be authorised given its current data, noted Goepfert, who is a principal investigator in Johnson & Johnson’s Phase III JNJ-78436735 trial.
This news service reported 19 September that any Covid-19 vaccine receiving an EUA from the FDA based on interim analysis data is unlikely to offer granular information about which subpopulations the vaccine is best suited to treat. For example, the 9 November release does not have information if the vaccine can prevent milder or more severe disease as well as what age group the vaccine is best suited for, Bottazzi said.
However, it is likely that the trial’s 164 event target could be reached once Pfizer/BioNTech submit their application, which would provide more colour on vaccine utility, Viele added. Pfizer/BioNTech expects to produce up to 50 million vaccine doses globally in 2020 and up to 1.3 billion doses in 2021, with healthcare workers, older people and people at risk of severe disease likely to be the first vaccine recipients.
As for Moderna’s and J&J’s forthcoming late-phase vaccine data announcements, Pfizer/BioNTech’s 90% efficacy announcement has struck a chord with the public, which could be to the other two asset’s disadvantage, Bottazzi said. If any of these vaccines report 70% efficacy in more specific groups, it could be perceived as being less efficacious even if there is more detail, she said. But it should be communicated to the public that it is hard to compare vaccines owing to difference in technologies and manufacturing nuances, Viele explained.
According to Monday’s media release, a new secondary endpoint evaluating efficacy based on events accrued 14 days after the second dose has been added to allow cross-trial learnings with other vaccines. While welcome, comparisons may still be limited as trials are not powered for secondary endpoints, Bottazzi added. J&J’s single-dose JNJ-78436735 counts an event after 14 days in its Phase III trial (NCT04505722), which is the same with Moderna’s Phase III mRNA-1273 study (NCT04470427). The 14-day window is more important for JNJ-78436735, as it is a single jab.
For Pfizer/BioNTech to also set immunogenicity benchmarks, future data should demonstrate a correlation between neutralizing antibody titers with protection rates, Goepfert suggested. Change of antibody levels in participants is indeed a secondary endpoint.
Reynald Castaneda is an Associate Editor for Clinical Trials Arena parent company GlobalData’s investigative journalism team. A version of this article originally appeared on the Insights module of GlobalData’s Pharmaceutical Intelligence Center. To access more articles like this, visit GlobalData.
