Federal lab won't publish COVID-19 antibody test accuracy results

A major federal laboratory in Frederick, Md., is validating the accuracy of antibody tests for COVID-19, but it does not plan to release the results of that work to the public because the data might be "commercially sensitive."

After fielding questions from the Star Tribune in recent days, federal officials said late Tuesday that performance data on antibody tests are likely to be released to the public at an unspecified point in the future, by a different federal organization — the Food and Drug Administration.

Critics in public health circles say there's no good reason during a health emergency to delay the release of publicly funded research about the accuracy of screening tools such as COVID-19 antibody tests.

"It should absolutely be made public, and immediately," said former FDA Associate Commissioner Peter Pitts, president of the Center for Medicine in the Public Interest (CMPI) in New York. "Opaqueness only communicates treachery. Openness and honesty communicate truth and accuracy, which is exactly what we want these tests to provide."

The human body naturally makes antibodies to kill the COVID-19 virus, though in most people it takes about two weeks from the onset of symptoms for antibodies to be detectable in blood samples. In theory, the antibodies should provide immunity to getting reinfected with the virus, though it's not yet clear how long immunity lasts or whether it's 100% protective.

Dr. Ruth Lynfield, the state epidemiologist for Minnesota, cautioned Tuesday that it's too early to say someone who tests positive for COVID-19 antibodies can simply disregard social distancing restrictions. The FDA has warned that antibody testing is not definitive, partly because the tests might react to viruses that are similar to the one that causes COVID-19.

Nevertheless, government officials have suggested that widespread antibody testing could be the key to identifying people who could safely return to work and restart state economies.

The FDA, under pressure to allow as many tests as possible to get to Americans quickly, is allowing the makers of antibody tests to market them without the normal process of filing validation data with the agency. That has led to a cascade of more than 100 different tests flooding the market without the usual notifications about accuracy.

The FDA is urging companies to voluntarily request "emergency use authorization" (EUA), which would subject the antibody test to greater scrutiny by regulators. As of Tuesday, the FDA has granted EUA status to eight antibody tests. (The FDA page uses the term serology to describe antibody tests. Most EUA tests are molecular tests that detect snippets of viral genetic material in an active infection.)

In early April, National Cancer Institute (NCI) Principal Deputy Director Dr. Douglas Lowy revealed in a webinar that a group of serology researchers at the Frederick National Laboratory for Cancer Research would pivot from HPV vaccine research to performance testing of COVID-19 tests. The Frederick, Md., lab is sponsored by the NCI, which is part of the National Institutes of Health.

"It's critically important to make sure that the people who are antibody-positive, that these are not false positives," Lowy said in the April 9 call. "The FDA has asked us if we would essentially help to validate the serological tests that have been submitted to the FDA. And that is in progress."

The FDA mentioned the research in an April 18 statement, but declined to elaborate at the time.

In the webinar, Lowy didn't say anything about the work being confidential. But on Monday, an NCI spokeswoman said in an e-mail to the Star Tribune, "Release of results related to specific kits is potentially commercially sensitive info, and therefore NCI is deferring the release of these performance data to FDA."

Late Tuesday, an FDA spokeswoman said the FDA plans to release the data at a future date, though no date was provided. "The FDA will review NIH's results expeditiously, and results from NIH's assessment may complement and inform the determination of whether FDA issues an EUA for a particular test."

Michael Osterholm, director of the Center for Infectious Disease Research and Policy at the University of Minnesota, said he can't understand a decision to let commercial considerations of test makers interfere with the release of test-accuracy data to the public.

"FDA has to take this on. This is their responsibility, not NIH's," Osterholm said. "FDA has to evaluate these, and they have to be clear about what (test results) mean to the public."

CMPI's Pitts said the need to validate antibody tests was a positive sign, because it shows the rapid development of much-needed tests.

But if the performance testing shows some tests are better than others, that is crucial information that should immediately be provided to health care professionals.

"Whoever mandated that this data be kept confidential needs to give it a serious rethinking," Pitts said.

Joe Carlson • 612-673-4779