University of Minnesota's infectious disease center urges better use of COVID-19 testing

In the rush to deploy as many COVID-19 tests as possible, important nuances are being lost about which kinds of tests are not worth doing, causing waste and potential harm during the pandemic.

Business managers should not be pinning hopes for reopening on antibody testing of their entire workforces, and hospitals should not be doing universal testing of asymptomatic patients, according to a new report from a University of Minnesota policy center. Both would be a waste of testing capacity that could give patients bad data unless there's a specific reason for the testing, such as a local outbreak.

And right now, the report concluded, no one should use antibody testing to issue "immunity passports" because the tests can deliver inaccurate results and no one yet knows how protective antibodies are.

"We're so focused on numbers of people being tested, and we are not focusing on what the accomplishment of that testing is all about," said Michael Osterholm, director of the U's Center for Infectious Disease Research and Policy (CIDRAP), which published the 13-page report this week.

The center hopes to raise awareness about the lack of a cohesive national strategy for diagnosing the pandemic. The report includes a number of cases in which testing would be appropriate within the "smart testing framework" that it advocates.

The medical technology industry has boasted about how fast COVID-19 tests have hit the market. Just six months ago, there was no such thing as a test to detect the SARS-CoV-2 virus or the antibodies the body produces to fight it, because the virus itself hadn't been detected.

As of Thursday, the Food and Drug Administration has received emergency authorization applications for 76 different diagnostic test kits to detect the virus genetically, and another 29 high-complexity diagnostic tests used in certified laboratories like the one at Mayo Clinic.

Meanwhile, more than 100 tests that look for antibodies in blood samples were available just weeks ago, though just 13 are listed as having received emergency-use authorizations today.

Mayo Clinic was among the first health systems to validate an antibody test and has since conducted 7,000 at the request of health system clinicians. Its national reference laboratory also has tested more than 61,000 samples.

New technology continues to hit the market. On May 9, the FDA authorized the first diagnostic antigen test in the U.S., which could deliver reliable positive test results earlier than other forms of testing but cannot be used to rule out an active infection because of its high false-negative rate.

Many biomedical researchers are also focused on the need for a test to detect not just antibodies, but "neutralizing antibodies," which are tiny proteins that block the virus from penetrating into a host cell.

Rochester, Minn.-based Vyriad says it has used its existing cancer-virology technology platform as a springboard to develop the first commercially available neutralizing antibody test that doesn't require a high level of bio­hazard containment.

Dr. Stephen Russell, CEO of Vyriad, said the most immediate use of his company's technology is likely to be for research on vaccine candidates, as the nation works toward the goal of developing 300 million vaccine doses by January under the Trump administration's Operation Warp Speed.

The lab-based test will also be useful in screening donors for convalescent plasma experiments, he said. Vyriad signed its first commercial contract to do neutralizing antibody testing on a commercial basis this week and expects to add more. Since Vyriad can process only 1,000 specimens a day, there are plans to expand to other sites, including Mayo Clinic, though that takes weeks to set up.

Russell explained that antibodies are much smaller than a virus, and they tend to swarm all over the exterior "spike protein" arms that give a corona­virus its distinctive appearance.

"Some of those antibodies are capable of blocking the virus from interacting with its receptor and getting into the target cell. Other antibodies are simply not capable of doing that; they just bind to the protein," Russell said. "They're of no real value in terms of protecting against infection. So the functional test that we have is for antibodies that neutralize, as opposed to antibodies that just bind."

Separately, Mayo Clinic and the University of Minnesota are working on neutralizing antibody research that may generate evidence about the immunity from reinfection after getting COVID-19.

While research with other types of coronaviruses suggests that antibodies provide at least short-term immunity after infection, there's no definitive proof of a similar effect with the virus that causes COVID-19. A Chinese study in monkeys offered some promise, and a Korean study this week reported that people who suffered second infections seemed to kill off the virus.

Mayo, the U, and the state of Minnesota are all planning to use antibody testing on a large scale for population-level research, which can be used to roughly figure the overall prevalence of COVID-19 in a community.

That's a strategy CIDRAP endorses, as long as the research methods are transparent and sampling is done in a way that results can be applied to a larger population. CIDRAP estimates that roughly 5% of Americans have already been infected, with many suffering mild or no symptoms and not ever knowing their status.

Research plans include widespread testing over the next year of samples from blood donors to Memorial Blood Centers, said Dr. Ruth Lynfield, state epidemiologist. But donors are often healthier, so the state also wants antibody testing studies in more diverse populations such as grocery workers, students and health care staff.

"We still don't know enough yet about protective immunity" for people carrying antibodies, she said, "so the way we are using serology is really to look at how much of the population has been exposed."

The U is close to a capacity of 15,000 antibody tests per day with the novel version that it created. Patients in the M Health Fairview system are able to request it, but doctors need to understand the limitations before ordering it, said the U's Dr. Timothy Schacker.

While many antibody tests have been used to check COVID-19 status before patients undergo surgeries or donate plasma as an experimental therapy, some may have been given to patients even though the results offered little therapeutic value, said Elitza Theel, director of Mayo's Infectious Diseases Serology Laboratory.

"There aren't too many — really, any — definitive clinical decisions being made on an antibody result," Theel said. "We really can't make decisions on whether or not someone should use PPE (personal protective equipment) or adhere to social distancing practices based on an antibody result."

That's where the interest in neutralizing antibodies comes in. Finding proof that neutralizing antibodies confer at least some level of protection from reinfection could make a huge difference in nursing homes and moving residents and staff around to protect them, Schacker said.

"If we know they've got neutralizing antibodies, we can cohort them together," he said. "We can start thinking about things like that."

Joe Carlson • 612-673-4779

Jeremy Olson • 612-673-7744