South Africa's decision not deploy the AstraZeneca vaccine is financially expensive, poor in science, and goes against the government's own assurances that it would take its lead from WHO recommendations, writes Professor Shabir A Madhi, dean of the faculty of health sciences at the University of the Witwatersrand and director of the SA Medical Research Council Vaccines and Infectious Diseases Analytics Research Unit.
The past month has been a roller-coaster ride when it comes to the rollout of COVID-19 vaccines in South Africa. The government has halted the distribution inside South Africa of 1m doses of the AstraZeneca vaccine and is poised to make a potentially historic mistake.
After much criticism early in January 2021 of the virtual absence of any strategy beyond procuring a limited quantity of COVID-19 vaccines through the COVAX facility, the government successfully negotiated procurement of 1.5m doses of the AstraZeneca COVID-19 vaccine through a bilateral agreement with Serum Institute of India.
This was, however, at a price ($5.25) more than double what EU countries ($2.5) had negotiated to pay for the same vaccine (but from a European manufacturer). Within a week surrounding the fanfare of the arrival of one million doses of the AstraZeneca vaccine at a cost of R75m came the disappointing news from a study evaluating this vaccine in South Africa since June 2020 that the vaccine does not protect against mainly mild (usually limited to infection of the upper airways) COVID-19.
The tell-tale signs that the AstraZeneca (and probably all other) COVID-19 vaccines might be less efficacious in protecting against mild-moderate COVID-19 in South Africa was evident from reports that have emerged since mid-January 2021, from local scientists, which indicated that the ability of antibody induced by natural infection by the ancestry SARS-CoV-2 variant had greatly diminished activity against the B. 1.351 variant that emerged in South Africa late in October 2020.
This was a consequence of gene mutations that affected key epitopes of the spike-protein of the virus, which are the components of the protein that most of the current first generation COVID-19 vaccines induce functional antibody against. Consequently, it was expected that even antibody induced by the first-generation COVID-19 vaccines would have diminished activity against the B. 1.351 variant which had undergone substantive changes in key components of target of these vaccines.
Corroborating this were interim results from another study, also conducted in South Africa, on the Novavax COVID-19 vaccine which drew much less publicity than the results of the AstraZeneca vaccine study, despite it being the only vaccine then to show protection against the B. 1351 variant.
Notably in the Novavax study, the attack rate of mild-moderate COVID-19 (probably mainly due to the B. 1351) variant did not differ in the unvaccinated control group between those who had already been infected with the SARS-CoV-2 prototype virus compared to uninfected individuals at the time of study-enrolment.
This provided the first conclusive evidence that the mutations included in the B. 1.351 were clinically significant, and that despite 35-45% of South Africans (especially in densely populated urban areas) likely to have been infected during the course of the first wave by the prototype virus, such infection did not confer any protection against developing mild to moderate COVID-19 due to the B. 1351 variant.
The rude awakening of these results (for some) was that the notion of relying on “herd immunity” evolving by allowing people to be unabatedly naturally infected was doomed not to materialise. Instead, the high force of infection that occurred in South Africa (despite attempts at restrictive and economically devastating lockdowns) probably contributed to the problem of the evolution of a variant that was now resistant to antibody induced by past infection.
Furthermore, the Novavax vaccine results from the South African study showed lower efficacy (49%-60%) against mainly mild-moderate COVID-19 mainly due to the B. 1.351 variant compared with results of the same vaccine in the UK (89% efficacy), where the efficacy was not affected by another variant (B.1.1.7) circulating there.
In studies on other COVID-19 vaccines, including the mRNA vaccines of Biontech/Pfizer and Moderna, for which vaccine efficacy has been reported to be 95% against Covid-19 due to the ancestry virus, laboratory investigations have reported 6.5-8.5 fold reduction in the potency of the antibody induced by those vaccine against the B. 1351 variant relative to that observed for the prototype virus.
For the Johnson and Johnson (J&J) vaccine, which uses a similar technology for design of vaccine as the AstraZeneca vaccine and also induces almost identical functional antibody response after a single and two-dose schedule, the laboratory testing of the antibody activity against the B. 1351 variant have not yet been released. I expect they will be no different to that observed for the AstraZeneca vaccine.
In contrast to the lack of efficacy of the AstraZeneca vaccine against mild-moderate COVID-19 in South Africa, the J&J COVID-19 vaccine results from the South African arm of the study reported a 89% reduced risk of severe COVID-19, also mainly due to the B. 1.351.
Unlike the AstraZeneca vaccine trial in South Africa, which targeted a younger age-group demographic with low prevalence of co-morbidities, a high proportion of participants in the J&J study were individuals at high risk of severe disease — that is, older than 60 and with high prevalence of co-morbidities such as diabetes and hypertension.
Hence, any direct comparison of the efficacy of these two vaccines is misguided and fundamentally flawed. Whether the J&J vaccine indeed protects against the same spectrum of mild illness that is caused by the B. 1.351 variant is yet to be reported or established for any such head-to-head comparison between these two vaccines (or other COVID-19 vaccines) to be meaningful.
So why might there be a difference in how well COVID-19 vaccines work against mild illness compared to severe disease. First, this would not be unique to COVID-19 vaccines. Vaccine against other respiratory pathogens, including seasonal influenza virus, respiratory syncytial virus in infants born to vaccinated women, and pneumococcus consistently report that protection against mild illness (usually of the upper airways) is more challenging to accomplish than prevention of severe disease (usually lower airway disease — pneumonia for example).
The reasons for this may differ between vaccines, including possibly requiring lower concentrations of antibody to protect against severe than mild illness. In addition for COVID-19, challenge studies in vaccinated non-human primates indicate that protection against lower airway infection can be conferred largely independent of vaccine induced antibody, through the stimulation of natural killer cells (also referred to as part of T-cell immunity).
Fortunately, there is a broader repertoire of spike-protein peptides that induce T-cell immunity, most of which are unaffected by even the B. 1.351 spike protein mutations. For the AstraZeneca vaccine, almost none of the B. 1351 mutations are likely to compromise the killer cell responses induced by the vaccine.
Consequently, there remains a strong biologically plausible reason to expect the AstraZeneca vaccine will protect against severe disease due to the B. 1351 variant, likely to a similar magnitude as the J&J vaccine.
It is largely premised on this, that the World Health Organisation in mid-February 2021 recommended that the AstraZeneca vaccine still be rolled out even in countries where the B. 1.351 variant or other similar variants of concern are circulating.
The AstraZeneca vaccine, as well as many other first-generation COVID-19 vaccines, are unlikely to bring about interruption in transmission of SARS-CoV-2 or protect against mild infection where these variants of concern such as the B. 1351 variant are circulating.
This would be compounded if there is sluggish rollout of the vaccine which is drawn out over a year, rather than within a few months. Nevertheless, these first-generation COVID-19 vaccines even in settings such as South Africa still provide the only sustainable option of preventing flooding of our hospitals with severe COVID-19 cases and mitigate COVID-19 deaths once the next resurgence is upon us.
Hence, the decision by South Africa not to deploy the vaccine goes against the spirit of what was previously espoused by the Health Department — that it would take its lead from WHO recommendations.
Furthermore, this decision inadvertently leads itself to choosing between leaving high-risk individuals largely unprotected against being hospitalised and dying of COVID-19, as opposed to rolling out the available AstraZeneca vaccine to those who choose to take the chance of deriving potential protection by being immunised with a safe vaccine that has been established to reduce the risk of severe COVID-19 by 80% even among people older than 80 years in the UK.
The decision by South Africa would have less repercussion if high-risk individuals were likely to be vaccinated with other COVID-19 vaccines for which clinical (and not only extrapolation of laboratory-based data) protection is established against the B. 1351 variant before the next resurgence.
Unfortunately, based on the latest sharing of information on the pipeline of COVID-19 vaccines likely to become available to South Africa, other COVID-19 vaccines, all besides the J&J vaccine which will not have been evaluated clinically against the B. 1351 variant, are only likely to become available from mid-April onward, without any clear indication on what quantities and how soon these may be deployed into the arms of people.
Consequently, the current strategy to try to offload the AstraZeneca vaccine to other African countries where the variant might not have been identified (mostly due to there being little sequencing to actually identify its presence rather than its actual absence) is problematic. This is compounded by the current expiry date of the vaccine being 30 April 2021.
Even were there to be reversal of the decision by government to take its chances with the AstraZeneca vaccine and target it to high-risk individuals who choose to take their chances of being vaccinated rather than remaining unprotected when the resurgence is upon us, getting a million people vaccinated before the end of April would be a mission.
Every additional day of procrastination lends itself to much of the R75m used to procure the vaccine going to waste, while the elderly and other high-risk individuals would certainly remain unprotected as opposed to being offered a fighting chance of being protected against COVID-19 severe disease and death.
Declaration of interests:
National principal investigator of the University of Oxford (AstraZeneca vaccine) and Novavax COVID-19 vaccine trials. The AstraZeneca vaccine study was funded by BMGF and SAMRC, and the Novavax vaccine trail; by BMGF and Novavax. All funding for these studies goes to the Wits Health Consortium of the University of the Witwatersrand.
This article first appeared on BusinessLIVE and is published with the kind permission of the author and BusinessLIVE.
See also from the MedicalBrief archives:
Madhi’s plea to government: Don’t waste the AstraZeneca vaccine