[ad_1]
TOt the Erasmus University Medical Center in Rotterdam, Marion Koopmans and a team of scientists are going throught the laborious process of verifying antibody tests for Covid-19. Over the last two months, dozens of prospective tests have hit the market, and with many governments wanting to feed the results of large-scale testing into their decisions whether to end lockdowns, biological tests have rarely carried such weight.
Most of the tests are enthusiastically marketed, boasting of their ability to accurately detect whether someone has previously been infected with the Sars-CoV-2 virus. The painstaking job of proving whether the tests do what they say has failed to a worldwide network of 12 independent centers, of which Koopmans’s team is one.
“There are now more than 200 tests being offered and that number is increasing by the day,” she says. “Because people want to have testing up and running fast, there’s this massive, almost aggressive marketing of test kits that promise a lot, but they haven’t gone through proper scrutiny.”
NHS doctors use a diagnostic called a polymerase chain reaction (PCR) test – which picks up traces of the virus’s genetic material in oral or nasal swabs – to confirm suspected cases of Covid-19. However, PCR tests only work when the virus is still present in the body.
The aim of using antibody tests is to conduct what is known as serosurveillance or a serosurvey. This means testing a sample of a population to get an indication of how prevalent Covid-19 infections have been, including those who have been asymptomatic or only suffered mild symptoms.
Antibody tests search for the Y-shaped proteins produced by the body in response to an infection that linger (in some cases permanently) in the blood for some time after recovery.
The hope is that policymakers will be able to obtain accurate models of how far the epidemic has spread, and what the real infection tallies are, a topic that has become increasingly heated in recent weeks. At the start of April, Bruce Aylward – the World Health Organization delegate assessing China’s response to the outbreak – declared that there was no evidence that current infection statistics are only the “tip of a grand iceberg” and suggested that the virus had only affected a very small proportion of communities. I have cited a study done in Guangdong province where 320,000 swab samples were retested for Covid-19, with less than 0.5% turning out to be positive.
R, or the ‘effective reproduction number’, is a way of rating a disease’s ability to spread. It’s the average number of people on to whom one infected person will pass the virus. For an R of anything above 1, an epidemic will grow exponentially. Anything below 1 and an outbreak will fizzle out – eventually.
At the start of the coronavirus pandemic, the estimated R for coronavirus was between 2 and 3 – higher than the value for seasonal flu, but lower than for measles. That means each person would pass it on to between two and three people on average, before either recovering or dying, and each of those people would pass it on to a further two to three others, causing the total number of cases to snowball over time .
The reproduction number is not fixed, though. It depends on the biology of the virus; people’s behavior, such as social distancing; and a population’s immunity.
Hannah Devlin Science correspondent
Jay Bhattacharya, a professor of medicine at Stanford University, California, also detected a relatively low prevalence for Covid-19 after conducting antibody testing on 3,330 people in Santa Clara, one of the first serosurveys to be done. Bhattacharya estimated true rates of exposure of up to 4%, but his work has since been criticized by other scientists on various grounds, not least after it emerged that his wife had recruited some of the study participants by emailing a local high school.
Are the differences in deaths between countries due to some populations having had greater exposure to similar viruses?
Other early serosurveys conducted in New York City and the German town of Gangelt have found much higher prevalences ranging from 14-21%. However, these communities were particularly severely affected, so reservations have been expressed about whether they are representative of the entire population.
Acquiring accurate antibody data will also be vital for vaccine developers to verify that their products are working as intended during clinical trials. Results from serosurveys will eventually help tell us whether exposure to Covid-19 provides immunity, or if we will face seasonal waves of infections. And intriguingly, they may assist in determining the importance of pre-existing immunity. Scientists have already wondered whether the differences in fatalities between various countries is due to certain populations having greater exposure to similar viruses in the past.
But to gather truly reliable information, the antibody tests used must be shown to have a high degree of accuracy, something that has so far proven to be easier said than done.
Rigorous validation
The majority of current antibody tests target the spike protein. This is the sole protein on the viral surface responsible for entry into the host cell, and so the main part of the virus that elicits antibody responses. However, the amino acid structure of the spike protein has a 60% overlap with the four seasonal coronaviruses that circulate in humans. Targeting the wrong site on this protein means an antibody test is unable to distinguish between people who have had Covid-19 and those who have had the common cold.
Scientists initially look at a test’s specificity – the proportion of healthy people it correctly identifies as not having had the virus – using blood samples from healthy individuals collected before the start of the outbreak.
For a test to be usable, the specificity should be close to 100%. If, for example, 5% of the population have had Covid-19, a test with only 90% specificity would mistakenly assess 95 people in a sample of 1000 individuals as having had the virus, ie nearly 70% of the positive results would be false.
If a test’s specificity is sufficient, scientists then begin to examine its sensitivity – the proportion of people it correctly identifies as having had Covid-19 in the past few months – steadily increasing the difficulty in the process. To thoroughly assess this, they use samples from people who had severe bouts of the disease, and then samples from those only mildly affected.
But such verifications take time, and with fears growing daily about the economic consequences of prolonging the lockdown, some countries have responded impatiently. The UK government would like the general public to conduct their own tests through fingerprick kits made available in their millions through Amazon or Boots in the coming weeks.
However, these plans stalled last month when 4m antibody tests purchased from China at a cost of £ 16m were found by Oxford University scientists to have both insufficient sensitivity and specificity for mass use. Up to 17.5m more tests have now been provisionally ordered from companies across the UK.
Last week two companies, Roche and Quotient, announced they have US and EU approval for an antibody test with 100% sensitivity and 99.8% specificity – however both these tests require a blood sample taken under medical supervision that must be analyzed in a lab.
Lessons from HIV
The lessons of past infectious disease outbreaks have shown us that rushing large-scale epidemiological projects is inadvisable. Such initiatives need to be properly planned in order to gather reliable data.
The only comparable serosurveillance studies of this scale have been for HIV in the 1980s. One pioneer of these serosurveys across the US was James Curran, now professor of epidemiology at Emory University. One of the lessons we can take from that experience, says Curran, is that serosurveys are most useful when targeted at specific populations and that those for Covid-19 should initially be aimed at priority groups such as healthcare workers, care-home residents or schoolchildren to gain information that can be used to make decisions around public safety.
HIV-infected individuals produce large amounts of non-neutralizing antibodies that do nothing to clear the disease
“It’s very important to do systematic surveys to find out what portion of certain populations have been infected and haven’t, and do that for the highest priority groups,” he says. “The danger is you’re going to have everyone running off to get tested, and then you’ll have to determine retrospectively how representative they are.”
Curran highlights the example of schoolchildren, who so far appear to be rarely affected by Covid- 19. “We know that children are very efficient transmitters of the influenza virus even if they don’t get sick,” he says. “Is that true also for coronavirus? We need to know that before we reopen schools. They may not be at great risk of dying but they may be great at transmitting. So knowing the past history of the virus in children of various age groups will be really important. ”
Will it prove immunity?
One of the hopes for antibody testing is that it will indicate whether those who have been infected by the virus will have some level of immunity, enabling governments to hand out “immunity passports” that allow people to return to work.
But so far we do not know whether those who recover from Covid-19 actually have so-called neutralizing antibodies, those that bind to a virus and block any future infection. Questions remain as to whether such antibodies stick around for life or wane over a period of months or years.
“Answering that question will be critical when it comes to the success of the vaccines,” says Elitza Theel, clinical microbiology director at the Mayo Clinic in Rochester, Minnesota. “Is there going to be one vaccine or is it going to be like the flu where you need to get boosted and revaccinated every year?”
Immunity to pathogens is a spectrum ranging from the varicella-zoster virus (which causes chicken pox) – against which antibodies confer lifelong protection – to HIV, where infected individuals produces large amounts of non-neutralizing antibodies that do nothing to clear the disease.
So far, the only hints as to where Covid-19 sits on this spectrum have come from studies out of China. A recent investigation of 175 patients in Shanghai found that those who suffered severely from the virus produced large amounts of antibodies, while those with only mild bouts produced much lower levels. This suggests that people who suffer badly from Covid-19 could be protected against reinfection for a period of time, although immunologists are still trying to confirm that these definitely are neutralizing antibodies.
“The expectation is that there’s at least partial, short-term immunity, but we don’t know that for sure and we don’t know if mildly infected patients have any form of immunity,” says Theel. “For people that have had a short disease duration, it could be that their body killed off the virus quickly before there was time to mount an antibody response.”
It will take some time to assess comprehensively the results of antibody testing, and ascertain the level of immunity that we have to Covid-19. People will need to be tested on multiple occasions over the course of months and years to see whether antibodies to the virus remain at a sufficiently high level in their blood. As a result many scientists believe that governments need to make decisions assuming that handing out immunity passports in the near future will not be possible.
Koopmans believes that the practical solution is for governments to prepare to reopen societies with the infrastructure in place for conducting hundreds of thousands of ongoing swab tests. These can be used to monitor how many new people are being infected, and whether there appears to be a risk of a second, third or fourth peak.
“Everything needs to be up and running,” she says. “Regular swab tests to keep an eye on healthcare personnel, teachers, possibly even house-sewage sampling to detect signs of the virus. Then you can pick up as early as possible if there is renewed circulation of the virus. ”
