Coronavirus Research Tracking - 19 June

Progress on drug trials, risks from pre-existing medical conditions, and the dynamics of infection.

This week’s highlights look at a mixed bag - two promising drug trials, the influence of pre-existing health conditions, and a range of studies examining aspects of transmission dynamics.

The Research Tracker is prepared by Dr Robert Hickson for the Science Media Centre. As this is a new service, please don’t hesitate to provide feedback.

A steroid may reduce deaths for Covid-19 patients on ventilators or receiving oxygen

The steroid dexamethasone shows promise in treating severe cases of Covid-19. In a randomised trial ⅓ fewer patients on ventilators who received the drug died (the risk of death declined from 40% to 28%). For those receiving oxygen there were 20% fewer deaths (reducing mortality risk from 25% to 20%), compared with patients not receiving the drug. 

However, full details of the trial have yet to be published. The steroid is typically used as an anti-inflammatory to treat asthma and rheumatoid arthritis. The researchers think the steroid helps moderate the immune response. 

Steroids are not recommended during the early stages of infection because they can reduce the ability of the immune system to fight infection. Do not self-prescribe steroids to prevent or treat Covid-19. In this trial the drug was administered later, and only for those with severe symptoms.

Some clinicians though, wary of the recent paper retractions in The Lancet and New England Journal of Medicine, want to see more data from the study before using the drug to treat severe cases of Covid-19.

Some US researchers have also commented that the mortality rates for patients on ventilators and oxygen in the UK trial are higher than seen for patients in some US hospitals US. The effectiveness of the steroid may, therefore, vary between hospitals and health systems.

The anti-viral drug Remdesivir may help prevent lung disease

In a study published in Nature, six rhesus macaques infected with SARS-CoV-2 did not develop lung disease when treated early with remdesivir (a general antiviral drug), whereas macaques not given the drug did. Treatment did not reduce the levels of virus in the upper respiratory tract, so they may still be infective during treatment. 

Macaques aren’t a perfect model for this disease, and further testing is required. The study, though, builds on an earlier human trial published in the New England Journal of Medicine that also demonstrated that the drug (used alongside other treatments) can help in the recovery of Covid-19 patients with lower respiratory tract infections.

Unlike dexamethasone there is no evidence at this stage that remdesivir by itself reduces mortality.

Three health conditions significantly increase the risk of hospitalisation and death from Covid-19

The CDC, analysing reports from US States and Territories, found that Covid-19 patients with pre-existing cardiovascular diseases, diabetes or lung disease were 12 times more likely to die, and six times more likely to be hospitalised than others infected with the virus. Those 80 years or older were the most likely to die, regardless of whether they had pre-existing health conditions.

Their analyses also show that people describing themselves as Hispanics, African Americans, and Native Americans have been disproportionately affected by Covid-19.

Pre-existing conditions generally increase risks

A study just published in The Lancet estimates that about 20% (1.7 billion people) of the world population have an increased risk of developing severe Covid-19 if they were infected. They base this on the proportion expected to have underlying health conditions.

Risk estimates vary from country to country. They provide an Excel spreadsheet so individual countries can be examined. For NZ, their estimate is that 24% of the population has a heightened risk of developing severe symptoms. However, there is a high level of uncertainty associated with their calculations, which don’t take account of other risk factors such as socioeconomic deprivation.

Estimating asymptomatic cases is difficult

Several recent analyses come up with different estimates for the level of asymptomatic infections (people who never show Covid-19 symptoms but have the virus).

In a pre-print paper a Swiss team analysed 28 earlier reports on virus testing from around the world. They calculated that overall about 15% of infections did not result in symptoms, although testing processes and methods varied between studies.

The proportion of pre-symptomatic cases varied considerably across these studies. However, this paper also analysed four papers that modelled early infections in China and Singapore, and estimated that 40-60% of virus transmission may occur from pre-symptomatic cases.

A separate review of 16 reports (some of which were the same as the Swiss group’s) estimated that 40-45% of infections were asymptomatic. This was published in the Annals of Internal Medicine

However, this review has been criticised for having less rigorous criteria for selection of studies and for the definition of asymptomatic cases. All these authors, though, agree that further research into asymptomatic cases, and pre-symptomatic spread, is needed.

The CDC analysis described above calculated that 40% of cases reported to them were asymptomatic. This will under-represent asymptomatic infections since many asymptomatic cases will not be tested. 

One of the better data sets for asymptomatic infections at the moment is from the cruise ship Diamond Princess, where many of the passengers and crew were tested. This provided an estimate of 17.9% of cases being asymptomatic. The confined circumstances of this case, and the demographics of the passengers, means that the Diamond Princess case can’t be used as a general predictor.  

The 17.9% level is already being challenged. A recent modelling study (not yet peer reviewed) of the Diamond Princess estimated that 70-78% of infections were asymptomatic.

The authors of this study suggest reasons for the different estimates is that previous calculations relied on diagnosed cases, and that PCR tests may be less reliable for asymptomatic infections. On the other hand, their model may be flawed. Clearly further research is necessary.

A small study suggest that asymptomatic patients may have weaker immune responses to Covid-19

A study of 37 asymptomatic people in China, published this week in Nature Medicine, indicated that they had a weaker immune response to the infection compared to those showing symptoms. This was measured by comparing Immunoglobulin G, neutralising antibody and cytokine levels.

The immune response to infection was more difficult to detect in asymptomatic patients 2-3 months after infection, suggesting that immunological testing may not identify everyone who has been infected. Asymptomatic patients also shed the virus for longer (a median 19 days compared to 14 days), although it is unknown if this increases transmission risk. Further research of asymptomatic infections is essential.

Infection risk is greatest in the early stages of infection

A Taiwanese study published in JAMA Internal Medicine found that transmission of the virus appears highest just before or within six days of symptoms appearing. 

This is supported by a laboratory study, to be published in Clinical Infectious Diseases, that infected cell lines with 90 viral samples taken from infected people. It found that infection was successful only for samples collected within 8 days of symptoms appearing, or from samples where viral levels were high.

Three factors are common to clusters of cases

Analysis of 61 Japanese Covid-19 clusters supports the view that transmissions occur most frequently when “the 3 C’s” apply - closed spaces with poor ventilation, crowded places, and/or close-contact settings. Hospitals and aged care facilities were the most common places for clusters. Other clusters were linked to restaurants and bars, workplaces, and music events. Focusing control efforts on such settings may be the most effective strategy.

In Japan people aged 20-39 appear to be the most common sources of infection for the non-healthcare clusters. Based on case data 40% of these may also have been due to asymptomatic or pre-symptomatic infections. The study will be published in Emerging Infectious Diseases.

Super-spreaders are also important

Super-spreading, where relatively few individuals are responsible for most of the transmission, occurs with infectious diseases. There are several reports of super-spreading events with SARS-CoV-2.

  • Analysis of the relatively limited spread of SARS-CoV-2 in Hong Kong found about 20% of cases were responsible for 80% of all transmissions there. This paper has not yet been published. 

  • A modelling study, currently being reviewed, predicted that about 80% of transmission cases may be caused by about 10% of those infected. 

Super-spreaders can’t be identified beforehand, so the effective control strategies need to focus on the “3C” settings, in addition to requiring good personal hygiene and physical distancing. A good summary of infectivity and super-spreading research is provided in a Medscape article (registration required).

A database of super-spreader events is available as a Google docs file.

Put down the toilet lid before flushing

A study, published in Physics of Fluids, modelled the fluid dynamics of toilet flushing. The model indicates that viral particles may rise up to one metre above the toilet seat following flushing, and remain in the air for over one minute, increasing the risk of infection. Where toilets have lids, the authors recommend putting them down before flushing. They also suggest ways to redesign toilets to reduce the risks of aerosol formation from flushing.

Other SMC resources