Antibody tests: The promise and pitfalls of using them to reopen the US

As Americans yearn for a future with work, shopping and outings with friends, one concept may become a crucial prerequisite: the widespread use of antibody tests.

Antibody tests, also known as serology tests, are designed to detect the presence of antibodies — proteins that are key elements of the body’s immune response to germs. Antibodies would be present in people who battled the coronavirus and recovered, as well as in people who had been exposed but whose infection was so mild that they didn’t notice they were sick.

If someone has antibodies from exposure to the coronavirus, that person should have immunity to the virus, at least for a period of time. (More on that in a bit.) That means they can go out into the wider community without fear of getting sick themselves. Antibody tests could also be helpful for making staffing decisions in medical units treating coronavirus patients.

Anthony Fauci, a top federal official in the fight against the coronavirus, has floated the idea of "immunity certificates" based on antibody testing. "As we get to the point of considering opening the country, it is very important to understand how much that virus has penetrated society," Fauci said on CNN.

It’s also a key element of Democratic presidential candidate Joe Biden’s plan for reopening the country. And New York Gov. Andrew Cuomo has gone so far as to develop and start implementing an antibody testing system for New York State. "We cannot restart life as we knew it without testing," Cuomo has tweeted.

PolitiFact interviewed several experts about the promise and potential pitfalls of antibody testing. 

The Food and Drug Administration has already granted an "emergency use authorization" for one particular antibody test manufactured by a company called Cellex. The FDA also green-lighted other tests for the market without the usual review process, as long as they are labeled with disclaimers that the agency has not fully reviewed them. 

That enabled more than 90 tests to go to market, some from academic organizations and some from medical companies, according to Politico. The CDC itself is working to produce one. Researchers at Stanford University have produced one test, while Mount Sinai’s Icahn School of Medicine has produced another. 

The key to developing antibody tests is to have an appropriate coronavirus protein, "because that’s what antibodies recognize," said Elizabeth McNally, director of the Center for Genetic Medicine at Northwestern University’s Feinberg School of Medicine.

The tests work by having some viral protein adhered to a surface, McNally said. "A person’s blood sample is mixed with a virus protein and is tested for whether there is an interaction between antibodies in the person’s blood and the antibody on the surface."

The good news is that many of the tests are not overly complicated, often requiring blood from a self-administered finger prick. Some have a turnaround time as rapid as 10 to 15 minutes.

That’s the bad news: The tests so far have not been especially accurate. 

In mid April, the FDA said the National Cancer Institute would start reviewing tests for accuracy. This move was welcomed by medical experts.

The United Kingdom has already pulled some tests for inaccuracy. Many antibody tests now being rolled out "may not be as accurate as we’d like," FDA commissioner Stephen Hahn has acknowledged. Scott Becker, CEO of the Association of Public Health Laboratories, went so far as to label many of the tests "crappy," CNN reported.

"There is very limited data, almost none peer-reviewed, evaluating how well these tests perform," said Elitza S. Theel, director of the infectious diseases serology laboratory at the Mayo Clinic. "So it’s critical that laboratories carefully perform validation and verification studies to ensure that the test they are offering for clinical testing is accurate."

A key metric will be a test’s "sensitivity," which refers to the percentage of positive tests that reflect genuine infections. "Anything less than 99% would mean too many cases in the population are wrongly identified, throwing off population estimates," said Theo Vos, a professor at the University of Washington’s Institute for Health Metrics and Evaluation.

Scientists expect that a past infection will produce some immunity. But they aren’t entirely certain of it, and it remains unclear how long that immunity could last.

Generally speaking, the degree and duration of immunity from an infection varies depending on the germ in question. Immunity from common cold viruses last a few weeks; immunity from chickenpox can last decades. Many infectious diseases fall somewhere in between.

One hopeful sign is that an infection with the virus that causes SARS, which is similar to COVID-19, produces an antibody response for a year, maybe more. That’s not as long as some diseases, but it could buy time until a successful vaccine is produced, which experts say could happen within 12 to 18 months. 

One unknown is whether that immune response from a past coronavirus infection is strong enough to actually preclude a new infection. 

COVID-19 probably gives immunity to "most people, but we don’t know for sure that is the case in every recovered COVID patient," said Angela Rasmussen, associate research scientist at the Center for Infection and Immunity at the Columbia University Mailman School of Public Health. "We don’t know what levels of antibodies confer protection either, so we should not assume that just having antibodies guarantees that you have completely protective immunity. There’s still a lot more to learn."

One practical concern is when to collect samples from individuals. 

"We know it takes over a week in some cases to mount an immune response to the virus and to develop a detectable level of antibodies," Theel said. "The samples need to be collected at a certain time after symptom onset. Using samples collected too soon would lead to negative results simply because the patient hasn’t developed an immune response to the virus yet." 

By the same token, additional testing may be necessary to confirm that someone who is antibody-positive is no longer experiencing an active infection that can get others sick.

Another practical issue is where the test should be administered. 

"We don’t want to have people rushing to a doctor’s office or a hospital since that could actually worsen the spread of the virus," McNally said. "Some companies are producing small individual test kits which will produce a ‘+’ or ‘-’ result. Some of these tests will require a doctor’s order and some may become commercially available."

For instance, McNally’s institution, Northwestern University, is developing a kit that is designed to be sent through the mail. Once returned via mail with a blood drop on specially treated paper, the tested individual can find out their result a few days later through a secure website.

Knowing how many people are immune will be a key element of the decision to reopen the economy. However, undertaking society-wide testing will require coordinated effort.

Already, some studies are under way. The World Health Organization is working on a study of a half-dozen nations, while one effort in the U.S. is already collecting blood samples in such cities as New York, Seattle and Minneapolis. Other studies are occurring in Michigan, Miami and Los Angeles.

One study completed in a hard-hit town in Germany recently found that 14% of those tested had antibodies to coronavirus.

McNally said research studies like these are important because they "allow us to be able to know how the population has developed immunity. We will want to know what percentage of the population has developed antibodies and how this differs across neighborhoods and areas. This is the type of information that will help to guide decisions about bringing people back to work and future safety issues."

Vos recommends focusing on a small number of the most promising tests, then scaling up production. The expected high demand for antibody tests should help push companies to do this quickly, he said. In the meantime, he said, poorly performing tests should be restricted and the FDA should accelerate its regulatory process.

"Things are happening quickly, but without strong oversight there will be a flood of undocumented tests in use, muddying the waters for anyone trying to make sense of what is going on in the community," Vos said.

It’s also important to note that antibody tests are just one part of what’s needed to restart the economy. Experts say it will also require a system of widely available diagnostic tests to determine if someone has an active coronavirus infection, along with "contact tracing," a method of identifying people who were in close proximity to other people who were infected, so that they can be tested and, if necessary, quarantined.

CDC director Robert R. Redfield told NBC News that large-scale antibody testing could fall into place "over the next several months."

McNally told PolitiFact that she expects antibody testing to be ramping up in May.

"We know most people take about four weeks to develop antibodies and immunity, so assuming exposure to the virus occurred in March or April, we really want testing working at scale in May," she said. "I think we’re on target for that."