Are antibody tests reliable?

Many flooding the market have little federal oversight and may give a false sense of security

Before COVID-19, the nutritionists and dietitians at Lafayette’s Cambiati Wellness Programs focused largely on helping women lose weight and gain energy through diet, vitamins, herbs and cooking classes.

Now they’re selling a $300 antibody blood test, a tool designed to detect whether someone was infected in the past by the respiratory virus that has killed nearly 245,000 worldwide and paralyzed the global economy.

“If people are feeling a lot of anxiety, it makes them feel better,” Rebecca Walker, a nutritionist with a degree in business administration, said of the antibody test. “It’s about peace of mind.”

The test is among dozens flooding the consumer market in the wake of newly loosened U.S. Food and Drug Administration restrictions to boost access to the tool. But are they accurate? And if they are, do they promise us a false sense of security?

Curious about long-ago coughs and fevers, and frustrated by stayat-home restrictions, people are seeking access to a tool that has not been perfected and, so far, has been used mostly by researchers.

Instead of detecting live virus in throat or nasal swabs, these tests reveal evidence of prior infection through markers of immune response in the blood — IGM antibodies, which are made early in an immune response and whose levels usually quickly wane, and IGG antibodies, whose levels rise more slowly after infection but usually persist longer.

The alluring premise of these tests is that the antibodies provide ongoing immunity, so people who have them can return to normal life without fear of sickness or sickening others.

There’s reason for cautious optimism. But we don’t know, because there’s not yet clinical evidence. For now, the promise exceeds reality, despite the marketing of dozens of tests of unknown accuracy and inflated claims.

“What you cannot conclude is that there’s immunity,” said Patrick Hsu, an assistant professor of bioengineering at UC Berkeley. “It’s true in certain cases. But you can’t make that assumption, especially in older people with chronic medical conditions.

“It tells you: Do you have antibodies, and how many? Not whether you’re protected.”

Research is underway to discover if the antibodies actually fend off infection, for how long, and what concentration is needed. In many diseases, such as measles, antibodies provide lifelong immunity, but in others, they

can be short-lived.

That poses a challenge for what Santa Clara County Public Health Officer Dr. Sara Cody calls a “very intriguing and very exciting” tool: “The top-level unanswered question is: Even if you were tested with a test with great characteristics and it can tell you yes, you were infected in the past, nobody knows whether the antibodies provide protective immunity. That’s the real question we have.”

The other big question: How accurate are the tests? A team led by Hsu and Dr. Alex Marson, UC San Francisco microbiologist and scientific director at the Innovative Genomics Institute, is assessing test accuracy on the website Covid Testing Project.

In their recent head-tohead comparison, Hsu and Marson concluded that many of the available tests performed reasonably well.

But others have disturbing false-positive rates, they found. This means that a large proportion of those who test positive may not actually have had the disease.

Stanford Health Care found the quality of commercial tests too uncertain and variable. So when it wanted to start testing its health care workers, it built its own.

So far, only 10 antibody tests for consumers have undergone the rigorous process of getting “emergency use authorization” from the FDA.

About 80 others, including some of unknown accuracy and inflated claims, haven’t been reviewed — but can be legally sold, as long they don’t claim FDA approval and are processed at a state-certified lab. It’s buyer beware.

“They came on the market in very short time with minimal review by the FDA,” said UCSF infectious disease epidemiologist Dr. George Rutherford. “Ideally, we will have some tests that can measure immunity at some point in time. … But if people think that they’re going to be able to go work at a nursing home, in hairdressing or some other sort of intimate personal services — these antibody tests aren’t going to tell them that.”

In the early days of the pandemic, the FDA was criticized for failing to provide enough tests to detect viral infections.

So it has taken a very different approach to antibody testing, with lighter regulation.

Commercial manufacturers can sell the tests after validating the results themselves; they must simply notify the agency and not make claims about federal approval.

Although manufacturers assert that their tests have high sensitivity and specificity, the data has not been independently verified.

At least two of the tests sold in the Bay Area have not yet received federal authorization. The antibody test sold by Arcpoint Labs is made by Diazyme Laboratories Inc. and processed in Jupiter, Florida, by a lab called Access Medical Labs. Prices vary by location; for its Martinez office, the cost is $195; in Monterey, it’s $175.

Arcpoint asserts that its test offers reassurance. On its website, Arcpoint says its test provides “yes” and “no” results: “No Exposure: No protective antibodies” and “Exposed & Recovering: Protective antibodies present.” Promoting its test for workplace use, it urges: “Keep your employees and community safe … the test can help you make informed decisions regarding your staff’s return to work or safe interaction with others.”

Cambiati Wellness Programs would not disclose the manufacturer or processor of its test, saying that is proprietary information. It says its test is awaiting FDA authorization.

Walker said demand has been strong because people are anxious about the virus. “For most people, they say they felt really sick before they were aware of what was happening, or found out late that they were exposed. Or they wonder if they were asymptomatic carriers. We might see a person who wants to go see their preemie grandchild or go back to work.

“We tell them you still have to be cautious. You don’t know what the protective level is,” she said.

Experts say the tests are best used in a research setting, to identify what proportion of the population has been unknowingly infected. According to the U.S. Centers for Disease Control and Prevention, as many as 25% of people who get the virus remain mostly or entirely asymptomatic, and many Americans who had especially bad cases of what they thought was the flu in January or February are now questioning whether they might have had COVID-19.

Recent studies using antibody tests in Santa Clara County, Los Angeles and New York have suggested the possibility that many thousands more people have been exposed to the coronavirus than we’ve yet understood.

If antibodies are proven to be protective, the tests could help get our $21 trillion economy back on track. Scientists, however, fear the public is so anxious to get back to normal before a vaccine is ready that they are willing to accept the antibody tests’ imperfection. Already, some Italian health officials have proposed issuing “certificates” for those with antibodies. The British government has floated a similar idea.

Once we gain a better understanding of immunity and improve testing, “serosurveillance is going to play a major role in … a framework for getting back to normal,” Dr. Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases and a White House adviser, said in a Journal of the American Medical Association livestream.

Two prominent COVID-19 roadmaps — one from former FDA Director Scott Gottlieb and the other from the University of Pennsylvania’s Dr. Ezekiel Emanuel — include widespread antibody testing as a critical step toward reopening society.

But the current testing landscape is filled with uncertainty, said UC Berkley’s Hsu.

“It’s the Wild West right now,” he said.