As states in the U.S. move past the initial push for tests to identify active COVID-19 infections, antibody tests are ramping up quickly to aid in disease surveillance and return-to-work screenings. The rush has spurred an explosion in serology tests, many hastily developed and of questionable value. However, as the pandemic enters its third month, some companies are offering high-accuracy tests with validated results.

Since the beginning of April, the FDA has granted emergency use authorization (EUA) to about a dozen antibody tests for COVID-19 and allowed scores more on the market without any regulatory review, raising serious questions about their accuracy and use. Indeed, many of the currently available tests have high rates of false positives, providing people with the false assurance that they were exposed to the novel coronavirus and carry some protection against it, when they do not. That can be dangerous as businesses begin to open up, putting people at risk as they come out of lockdown and begin to mix with others in their community. Even more worrisome is the potential for people with false negative results to unwittingly put others at risk of the disease.

The flood of COVID-19 antibody tests has raised concerns and health care experts and in Congress. On April 24, the House Committee on Oversight and Reform’s Subcommittee on Economic and consumer Policy released preliminary findings showing serious shortcomings in the Trump administration’s handling of serological testing, including a lack of standards and guidelines for serological antibody tests, departure from practices governing molecular tests and failure to take enforcement action against companies marketing shoddy tests.

“Serology testing for coronavirus has the potential to be a critical tool going forward for our nation,” said Chairman Raja Krishnamoorthi (D-Ill.). “Unfortunately, senior FDA and CDC officials admitted that they have not put Americans in the best position to use this tool. I was shocked to learn that we do not have the ability to know whether the available serology tests work and the FDA has no plans to make that a reality.”

On May 4, the FDA revised its EUA policy for COVID-19 testing to require manufacturers of commercial antibody tests to submit validation to the agency within 10 days.

“Often what we saw in much of the early testing … was they didn’t really have the performance they claimed,” Rangajaran Sampath, chief scientific officer of the nonprofit Foundation for Innovative New Diagnostics, told BioWorld. “That led to a lot of concern, as it rightfully should, because you are not adding any value other than saying you’ve done so many tests, which is not meaningful.”

Question of immunity

Adding to the uncertainty is that much is still unknown about SARS-CoV-2, the virus that causes COVID-19. While it is hoped that exposure will trigger an immune response, no one knows how strong it will be or how long it will last. And even if the best scenario proves true and antibodies provide long-term protection, none of the currently available tests tell whether antibodies to COVID-19 are functional – that is, they are capable of neutralizing the virus.

“I have not seen enough science to say any test can tell us whether we are immune to COVID-19. The best they can do is give some level of what immune response our body gave,” Julie Swann, head of the Edward P. Fitts Department of Industrial and Systems Engineering at North Carolina State University and an advisor to the CDC on its response to the 2009 H1N1 flu pandemic, told BioWorld. She worries that if people get antibody tests and don’t know how to correctly interpret them, they may believe they’re immune to the disease and not infectious to others.

She is not alone. The WHO has downplayed the idea that individuals with antibodies to SARS-CoV-2 could receive an “immunity passport,” warning there is no evidence that antibodies protect people from a subsequent infection.

Prevalence matters

Prevalence matters too in determining the effectiveness of antibody testing for disease surveillance, even with the more accurate tests. At this point, the percent of the population that has ever been infected with SARS-CoV-2 is still likely relatively small, meaning many of the positive results will be false positives. “If you have 1% of your population and have a test that’s only 99% specific, that means that when we find a positive, 50% of the time [that] will be a real positive and 50% of the time it won’t,” Deborah Birx, White House coordinator for the COVID-19 response, said in an April 20 press briefing.

“Clearly, one of the challenges we’ve seen is that even the little data that came out of some of the early testing done in high-prevalence settings in China or other places might not translate to immediately rolling it out as country-wide testing in all kinds of prevalence settings,” Sampath said.

William Blair analyst Brian Weinstein agrees. “There is tremendous epidemiologic value that comes from knowing prevalence [of COVID-19], but we would use caution in making decisions about back-to-office or back-to-normal decisions solely on the basis of a positive antibody test,” he told BioWorld.

Some reassuring numbers

Some of the more recent tests approved for emergency use are claiming higher accuracy. Abbott Park, Ill.-based Abbott Laboratories’ SARS-CoV-2 IgG assay has 100% sensitivity and 99.5% specificity when performed 14 days post symptoms, CEO Robert Ford told a recent first-quarter earnings call. The data was based on more than 1,000 samples.

Bio-Rad Laboratories Inc., of Hercules, Calif., launched its SARS-CoV-2 Total Ab blood-based assay at the end of April, claiming 98% sensitivity and 99% specificity. The company claims its test is the only one to track not just immunoglobulin M (IgG) and IgG, but also IgA.

Antibodies develop over a period of time. IgM antibodies typically show up in the six- to eight-day time frame, while IgGs start kicking in around days 12 to 14, indicating a more robust response. IgA plays an important role in the mucous membrane protection and may correlate to disease severity and risk to the lungs.

Each comes into play at a different time during the course of the disease. Looking for all three is expected to improve the results of antibody testing.

Also recently authorized by the FDA is Roche Holding AG’s Elecsys Anti-SARS-CoV-2 antibody test, boasting 100% sensitivity and 99.81% specificity. Specificity of the IgM/IgG immunoassay was validated using 5,272 samples from routine diagnostics, blood donors, as well as a common cold panel and a coronavirus panel test for cross-reactivity to prior infections. The testing also showed no cross-reactivity to antibodies from other viruses such as HIV and hepatitis C.

Sampath said a high-specificity assay is particularly important when trying to gauge infection levels in a community. “The numbers people are talking about are 98% or more specificity,” he said. “In that setting, the sensitivity is less critical. You can get away with a 90% sensitive assay.” But as hospitals and other potentially high-prevalence settings start to look at routine monitoring of worker, sensitivity of 94% or higher will be needed.

To expand access to testing, Quest Diagnostics Inc., of Secaucus, N.J., recently launched a consumer-initiated service through its Questdirect program. People who suspect they were previously infected can request the service and, following a doctor’s assessment of need, have blood drawn at any of the company’s 2,200 patient service centers.

Quest is using the Abbott IgG antibody test and the Euroimmun AG Anti-SARS-CoV-2 ELISA IgG antibody test. Like the Abbott test, Lubeck, Germany-based Euroimmun’s test has high specificity (98.5%-99%), meaning low cross-reactivity, and Quest has conducted its own validation of both tests, Jay Wolgemuth, Quest’s senior vice president and chief medical officer, told BioWorld.

“Since April 21, Quest has performed 325,000 antibody tests [physician-ordered and Questdirect],” Wolgemuth said. “Turnaround time is approximately one to two days after specimen is received at the lab.”

The Quest offering could fill a critical need to increase antibody testing, but with tests that have established high accuracy.

“We have been very concerned about rapid tests and the high false positive rates that many of these have shown to have,” Weinstein said. “The Quest offering should have a lower false positive rate, but people need to be aware that testing positive for antibodies does not guarantee you have them, and even if you do have them, it does not guarantee that you are immune from a second infection. We just do not yet know that definitively.”

Next-generation antibody tests

San Diego-based Abreos Biosciences Inc. is looking to create a multiplex test that goes beyond the simple yes/no COVID-19 antibody results and measures the “specific flavors” of antibodies to provide a more complete picture of the antibodies in a person’s blood.

“What we do is make peptides that are what technically we call antigen mimetics. So we’re able to identify peptides that copy the biochemical shape of the target of the antibodies, in this case the virus, but on a much smaller scale than is typically done,” Bradley Messmer, Abreos’ CEO, told BioWorld. “And we go in and really home in on the precise epitope … that each antibody is sticking, and then we make a biochemical copy of that that can be used to detect those antibodies.”

Last week, the company – which primarily focuses on therapeutic monoclonal antibodies – submitted a grant to the National Institute of Allergy and Infectious Diseases for funding to support its work on a COVID-19 antibody test.

Abreos is initially focused on the IgG antibody, but is considering doing both IgG and IgM. “The differences between IgM, IgG and IgA are at the stalk of the antibody,” Messmer said. By measuring the binding side of the antibody, “when we define one of our peptides, it will work for the IgM, IgG or IgA version of that antibody, just with a different secondary detector.”

The technology could potentially lead to better understanding of who is protected against future infection and who is not.

“Once you see the spectrum of antibodies in different people, you can then say, OK, did anybody not have protection and does that match up with missing a certain antibody flavor or having too much of the other one,” Messmer said.

“This virus is new to humans, so we just don’t know,” he added. “But we know there’s a lot that could go wrong with a simple yes/no answer.”

Recent literature lends weight to Abreos’ approach. A study in the journal Cell suggests that the S1 gene, one of several antigenic epitopes on the SARS-CoV-2 spike protein, may play a role in interactions with receptor and inducing neutralizing antibodies to COVID-19.

“Part of the challenge here was that biology wasn’t really available or executable immediately,” Sampath said.

Meanwhile, scientists at George Mason University in Virginia are developing a saliva test to detect COVID-19 antibodies. Raouf Guirguis, a professor in the College of Science, has created special saliva collection devices being used in the study, which will compare saliva- and blood-based rapid tests for COVID-19 viral protein and antibodies. The university includes one of 13 NIH-sponsored biosafety level 3 biomedical research laboratories, which is equipped to handle COVID-19 tests.

“Antibody tests, while not a perfect predictor of immunity, are a pointer to potential immunity,” N.C. State’s Swann said, adding knowing a large number of people had already been exposed to the disease could help communities plan when and how to begin reopening.

Still, given the high false positive and negative rates of some current tests, “anyone who’s purchasing these tests should be asking for information about their accuracy,” she said.

Editor's note: This article is part of a series of stories assessing the state of the rapidly evolving COVID-19 diagnostics.