Coronavirus Research Tracking - 19 March
Vaccines risks, variants and immune resistance, immunological insights, tracking & tracing, AI biases, planning for the next pandemic
This Research Tracker covers reports on blood clot risks from the AstraZeneca vaccine, safety of vaccines for pregnant women, effects and impacts of variants, immunological insights from the past year, a New Zealand tracing case study, and more.
The Research Tracker is prepared by Dr Robert Hickson for the Science Media Centre.
The Science Media Centre is contributing to the COVID-19 Vaccine Media Hub, hosted by Health Desk.
AstraZeneca vaccine and blood clots
Health Desk canvased expert opinions on the association between the AstraZeneca vaccine and blood clots. The conclusion is that it is very unlikely that the AstraZeneca vaccine has caused an abundance of blood clots in people receiving the vaccine, but additional studies are desirable.
In a press statement the European Medicines Agency has declared that "the vaccine is not associated with an increase in the overall risk of blood clots …" Around 20 million people have received the vaccine so far and the EMA has only reviewed 25 cases of blot clots linked to vaccinations.
Derek Lowe’s blog discusses the challenges the AstraZeneca vaccine is facing in Europe. It notes the problems with the earlier Phase 3 trials of this vaccine, and some poor subsequent communications. Lowe also highlights how perceptions and declining public confidence are over-riding the findings of the European Medicines Agency and the WHO.
Are vaccines safe for pregnant women?
A news feature in Nature notes that pregnant women with Covid-19 are at higher risk of hospitalization and severe disease than are women of the same age who are not pregnant. However, it's unclear how widespread Covid-19 infections are in pregnant women. Few pregnant women have been vaccinated yet, so the safety of vaccines for pregnant women is uncertain. Evidence so far indicates that it is rare for the virus to pass from mother to foetus.
A comment in Nature Reviews Immunology notes that data from earlier trials suggest that Covid-19 vaccination is safe for pregnant women when considered in relation to their greater risks if they become infected.
P.1 variant more transmissible
A paper, not yet peer reviewed, reports that transmissibility of the P.1 variant in Manaus, Brazil, is about 2.5 times higher than an earlier circulating version. However, it also estimated that the probability of becoming reinfected with P.1 following an earlier infection was low, about 6.4%.
Increased risk of death for B.1.1.7
A paper published in Nature reports an increased risk of death from the B.1.1.7 variant. This is based on an analysis of community tests and 4,945 deaths. It estimates that the hazard of death with this variant is 61% higher (taking account of testing biases and errors, and a range of socio-economic and demographic factors). This may mean that the absolute risk of death for a 55–69 year-old male increases from 0.6% to 0.9% within 28 days of a positive test.
Variant resistance to immune neutralisation
A letter in the New England Journal of Medicine reports that in tests using pseudoviruses the Moderna vaccine showed no decline in neutralising ability against the B.1.17 variant compared to earlier D614G variants. Neutralising ability was lower but still effective for the P.1 and B.1.427/B.1.429 (California) variants, and for a B.1.1.7+E484K construct. Neutralisation was reduced even more, by a factor of 6.4, for the B.1.351 variant but the pseudovirus was still neutralised.
A study of 99 vaccinated people, published in Cell, reinforced earlier studies that some specific mutations in the receptor binding domain of the spike protein can significantly reduce virus neutralisation. However, it also found that mutations outside of this domain also have important effects.
Unlike the NEJM paper above, this study found that B.1351 effectively escaped being neutralised. This study looked at both the Pfizer and Moderna vaccines. It also used pseudoviruses, so their constructs, and/or other methods may contribute to the difference.
A paper published in the New England Journal of Medicine reports on a clinical trial of the AstraZeneca/Oxford vaccine in South Africa. The vaccine did not demonstrate protection against mild-to-moderate Covid-19 due to the B.1.351 variant. It is uncertain how effective this vaccine is at preventing severe Covid-19 in people infected with B.1.351.
Another paper in Cell found that several monoclonal antibodies that target the N-terminal domain of the spike protein are very potent. Infection experiments with Syrian hamsters found that one of these antibodies substantially reduced viral loads in their lungs. Several of the recent variants have mutations in the N-terminal domain, indicating that selective pressures are operating here as well as on the receptor binding domain.
Estimating level of protection
A paper, not yet peer reviewed, estimates how long protection from vaccines could last. It modelled this using data on in vitro antibody neutralisation from tests of seven vaccines and convalescent patients. Neutralisation level was found to be highly predictive of immune protection, although more standardised data are needed to confirm this.
Decline in neutralisation activity is not linear. There is an initial rapid decline followed by a slower rate of change. The modelling predicts that a vaccine with an initial efficacy of 95% would be expected to maintain 58% efficacy by 250 days.
Such modelling can help inform when to administer booster shots.
Timing of second vaccine dose could be extended
A recently published model, not yet peer reviewed, indicates that delaying giving the second dose from four weeks to up to 9 weeks could maximize vaccine effectiveness and prevent additional infections, hospitalizations, and deaths. For the Pfizer/BioNTech vaccine the model indicated that giving the second shot more than 3 weeks after the first also could reduce hospitalisations and deaths, although probably not infections. These results depend on the efficacy of the first dose not declining before the second dose. More evidence is required on this critical point.
Overly focussed immune response in an immunocompromised patient
A study published in Cell looked at how the virus escaped the immune response in an immunocompromised patient. It found that most of the patient's antibodies targeted one part of the receptor binding domain and that one of several viral mutations could enable it to escape neutralisation. The authors are unsure whether this patient may be very unusual in having an antibody response overly focused on one part of the virus.
Long lasting T cell responses
A paper published in Science Translational Medicine reports that T cell responses to infection increased, while antibody responses declined, over the six months following infection. Some of the viral epitopes that the T cells target may be useful for future vaccines.
A timeline of immunological insights
A review in Nature Reviews Immunology highlights key immunological insights related to SARS-CoV-2 over the past 12 months. One important discovery was that delayed or deficient type I interferon responses are associated with severe Covid-19.
The paper also notes how the pandemic disproportionately affected young investigators, women and under-represented scientists. They were more likely to take on greater responsibilities at home, setting back progress on scientific equity.
Is herd immunity impossible for Covid-19?
A Nature news feature discusses the challenges in achieving herd immunity to the virus. These challenges include scientific, political, behavioural and evolutionary factors.
Radiographic AI not yet clinically useful
An analysis of 62 papers, published in Nature Machine Intelligence, found that image-based machine learning models for diagnosis or prognosis of Covid-19 aren’t currently clinically useful. The models, which analyse chest radiograph or CT images, were found to have underlying biases (in terms of participants, predictors, outcomes or analysis), and/or flawed methods.
New Zealand case study of tracking and tracing
A New Zealand case study, published in Emerging Infectious Diseases, describes how closed-circuit television observations, genomic sequence analyses, and epidemiologic investigations were all used to trace a post-quarantine chain of transmission last September.
An overview of contact tracing and testing methods
An article in The Lancet Microbe summarises some of the methods of contact tracing and testing. It describes the resources that each require and how features of SARS-CoV-2 transmission affect their effectiveness.
Saliva tests work well
A paper, not yet peer reviewed, reports that saliva-based sampling provides an accurate non-invasive means of monitoring antibody responses to SARS-CoV-2 infections and vaccinations, at least out to eight months.
Masks & exercise
Research published in the European Respiratory Journal reports that there are no substantive adverse health or performance effects associated with mask-wearing while exercising. The study involved 12 participants doing cardiopulmonary exercise tests with different types of masks. It found that while respiration was affected at peak exercise, there was only a slight impact on performance.
Global affect of the pandemic on paediatric cancer care
An analysis, published in The Lancet Child & Adolescent Health, found that the pandemic has adversely affected paediatric oncology services worldwide. Disruptions to cancer diagnosis and management have been substantial. The impact has been greatest in low-income and middle-income countries.
Preparing for the next pandemic
A suite of papers in Nature Medicine discuss what needs to be done now to prepare for the next pandemic. The editorial notes that the current pandemic has opened a world of opportunity for rethinking preparations for public health crises.