Karen Titus
June 2020—The arrival of a pandemic has shown—among many, many other things—that anyone who talks about it typically starts by saying, “This is a pandemic.” The next sentence tends to be, “It’s a completely different situation,” whether the focus is grocery shopping, exercising (or not), voting, or practicing medicine.
Pointing to the pandemic is a polite way of saying, “All bets are off.” For many, it’s been a springboard to innovation and breakthroughs, even in the midst of considerable anguish.
For clinical laboratories, however, much has felt unsettling, especially when the conversation turns to serology testing for SARS-CoV-2.
It’s a topic stuffed to overflowing with interest, enthusiasm—and, early on, antibody tests themselves. That also meant months, possibly stretching into summer, in which laboratories faced more questions than answers, at a time when the demand for serology tests seems insatiable (and somewhat perplexing).
“The laboratory is in the situation of having access to serologic assays, and not having any idea of what the performance characteristics are,” says Angela Caliendo, MD, PhD, professor and executive vice chair, Department of Medicine, Warren Alpert Medical School, Brown University.
Elitza Theel, PhD, director of the infectious diseases serology laboratory and associate professor, laboratory medicine and pathology, Mayo Clinic, is even more succinct. “We had a test before we knew how to use it,” says Dr. Theel, who spoke to CAP TODAY the first week in May.
And, given the situation (“This is a pandemic”), everything is bound to keep changing. Quickly. “Boy, you’re not kidding,” says Dr. Caliendo.
[dropcap]S[/dropcap]erology testing has become the blank tile in the game of Scrabble. Those well beyond the lab walls, and even medicine, are asking antibody tests to turn into whatever seems needed: a badge of immunity, a reason to reopen economies, a sign of disease recovery rates. It’s the focus of consumer ads and social media discourse. It’s even becoming a campaign issue.
And as with diagnostic testing in the early months of COVID-19 (see “A lab world embroiled in pandemic,” CAP TODAY, May 2020), these outsized expectations have bumped up against the limitations that always accompany lab tests.
In this case, this has meant having to build knowledge almost from scratch, like the bread so many began baking at home when the pandemic hit.
“There’s so much about this virus we don’t understand,” says Dr. Caliendo, who spoke to CAP TODAY the first week in May. That has made it difficult to answer that most fundamental question: Why are we using this test? Serology testing is helpful on a population basis, she says. It’s unclear how helpful it is right now—emphasis hers—on an individual basis.
“Right now serology has very limited utility,” agrees John Waugh, system vice president, Pathology and Laboratory Medicine, Henry Ford Health System, also speaking in early May. “There are no guidelines on how to use the information.”
Dr. Theel notes that while it’s hard to know what questions, precisely, will be relevant from month to month, questions will not be in short supply anytime soon. Do antibodies indicate protective immunity against reinfection? That’s the biggest issue right now, she says, given its implications for returning to work and vaccine efficacy.
“I think there are two questions,” Dr. Theel says. “Do we develop a protective immune response, and then, two, how long does that response last?” Studies are underway to look at the first question; obviously the second will require more longitudinal monitoring—months and years.
More immediately, Dr. Theel adds, antibody tests have played an essential role in screening of potential convalescent plasma donors. And as health care slowly regains its land legs—as people seek (or are encouraged to seek) routine care, proceed with surgeries, etc.—Mayo has begun screening patients prior to procedures, with both molecular testing and serology. A negative PCR test is a green light for the procedure; if positive, it’s delayed. The serology results are used for risk stratification.
Similar discussions are happening at Henry Ford, says Waugh, as surgeries that were elective eight weeks ago become more urgent. If patients are going to be tested, shouldn’t the health care workers who interact with them be tested as well? What about patients on dialysis, in skilled nursing facilities, and in home health care? And the personnel taking care of them? Patients in hospice are being tested—should family members taking care of them be tested as well? “That’s where our capacity limits creak right now,” Waugh says. At the same time, large companies (General Motors, for example) are asking to use testing to create safer workplaces as they reopen.
At ARUP Laboratories, Patricia Slev, D(ABCC), has seen sustained demand for antibody testing. “Many institutions are trying to determine how to offer this,” says Dr. Slev, section chief, Immunology Division, and medical director of core immunology, microbial immunology, and the serologic hepatitis and retrovirus laboratory.
One clear application is for seroprevalence in communities and regions, she says. There are also a number of ongoing studies to monitor frontline workers. ARUP is involved in both types of studies, says Dr. Slev, who is also associate professor of pathology, University of Utah.
And clinically? “That’s a good question at this point,” says Dr. Slev, who spoke to CAP TODAY in mid-May. The lack of information about immunity—the ifs, whethers, how longs—also means it’s a difficult one. “We cannot use this assay to suggest that someone is immune. We cannot use the result from this assay to guide PPE or social distancing policy.”
It’s also unclear how long a person is contagious. “We know that a PCR test stays positive for some time, often beyond when a person is contagious,” says Gregory Storch, MD, the Ruth L. Siteman professor of pediatrics, Washington University School of Medicine, and former director of the microbiology laboratory at St. Louis Children’s Hospital, pointing to studies that correlate PCR with culture. Is it possible that antibody development signifies the end of contagiousness? “That’s an open question that will need to be investigated.” If the answer is yes, that might signal another use for the antibody test.
News reports to the contrary, serologic testing is filled with nuances. When a person mounts an antibody response to an infection, “they make a whole repertoire of antibodies, which may have different functional significance,” says Dr. Storch. One test might measure protective antibodies; another might measure those that reflect infection but do not confer protection. “So once again, this is something that is going to have to be looked at over time by careful clinical studies, and it may separate out some of the antibody tests.”
[dropcap]A[/dropcap]s with everything related to the virus, knowledge is also rapidly evolving. Dr. Slev predicted some answers might be forthcoming as early as June. Moreover, she says, there’s much about the virus that can be inferred from SARS and MERS. “But it does seem to have some unique properties, and we want to make sure—before we act on some of these results—we know what they mean.”
Dr. Theel agrees. “Once we better understand the meaning of an antibody-positive result, I think antibody testing will become much more clinically relevant, and the importance of serologic tests is going to increase dramatically. And we can all make better decisions at that point.”
Even as there has been a call to arms to begin widespread antibody testing, labs have been reluctant to pick them up. When serology tests first started appearing on the market—a time that reasonably feels like ancient history—“Everybody grabbed onto it as, This is how we’re going to restart the economy, this is how we’re going to get people back to work,” says Dr. Theel. “But on the laboratory side, I would say we were and continue to be very cautious.”
Dr. Storch has watched the unfolding saga with growing concern.
“There’s a lot of rhetoric,” says Dr. Storch, who spoke to CAP TODAY the first week in May.
The hope that a positive antibody test result indicates immunity is weighed down by a big question mark. Researchers lack direct evidence that the presence of antibodies to SARS-CoV-2 in blood corresponds to a state of immunity. The expectation is reasonably sound, since it’s true of most viruses. “But not for all,” Dr. Storch warns.
Results might also help guide the behavior of those who may be at less risk. “But I hasten to say we’re not there yet,” Dr. Storch says.
Serologic assays can also help determine distribution of infection in the population. Since some people infected with the virus are asymptomatic, a serologic test is the best means for determining seroprevalence. Here the data are emerging more rapidly, he says. (Not that the data have been published, he acknowledges.) Based on results from a variety of settings, Dr. Storch says, “It appears that in most regions the seroprevalence is in the ballpark of five to 15 percent.”
This is significant for several reasons, he says. First, it confirms infections despite a lack of specific diagnosis, giving insight into the extent of those who were asymptomatic. On the flip side, “It indicates there still is a very large pool of people out there who have not yet had the virus, and therefore are susceptible.” And, ominously, “It strengthens the concern that there may be additional waves of COVID-19.”
[dropcap]A[/dropcap]t the same time that demand for antibody tests was unleashed, labs were taking stock of what they didn’t know, starting with which tests to use. This normally routine process—choosing tests is what labs do, after all—soon turned into a task that could make even the most experienced lab director feel queasy.
Mayo began offering IgG-based serologic assays the first part of April. At the time, no tests had emergency use authorization from the FDA—the agency wasn’t requiring it. Dr. Theel says she and her colleagues had to sort through a virtual pile of assays available at the time, choosing what they felt were the best possible and offering them clinically after a thorough validation.
That has since changed (change the one predictable thing about SARS-CoV-2) with the arrival of EUA-approved assays, many of them “with better performance characteristics than what we found in the earlier ones,” says Dr. Theel. “So we are transitioning to those.”
As of early May, her lab had validated or evaluated serologic assays from 10 different manufacturers, some well established, others newer to the market. Performance characteristics varied, to put it mildly.
“One thing we found,” she says, “is for some assays there’s a difference in sensitivity depending on disease severity. So for all the tests we evaluated, after about 14 days to symptom onset, all hospitalized patients were IgG seropositive regardless of test method or test kit.” That differed for patients with mild illness—some assays had a sensitivity that was much lower in so-called outpatient cases. It bears watching, says Dr. Theel: Do all patients with mild disease truly develop an antibody response? And is sensitivity different based on assay?
There also has been preliminary work, primarily from preprint studies (requiring cautious interpretation), to suggest older individuals produced higher levels of neutralizing antibodies compared with younger ones, Dr. Theel says, which might point to patient-specific differences that could affect antibody results. “It would be interesting to look at,” she says, then adds, perhaps unnecessarily, “It’s not something that, frankly, I’ve had the time to pursue further.”
If choosing a new test for a new virus in the middle of a pandemic is hard, choosing from a lot of them is even harder.
Early on, laboratories encountered an abundance of serology tests, which arrived on the market with minimal FDA oversight.
If the early restrictions on diagnostic testing kept the reins tight, the agency’s approach to antibody testing was more like dismantling the stalls, flinging open the barn door, and burning down the fences on the farm. “It does seem that the pendulum swung completely in the opposite direction,” Dr. Storch says. “When it came to the antibody tests, they took an extremely permissive and unusual approach.” Allowing tests to come on the market with nothing more than notification was “a disservice to the medical community,” he says.
Though the FDA initially waived EUA requirements, some companies decided to seek EUA anyway. And in early May the FDA issued further clarification and guidance.
But in the interim, that made things even harder for those doing the testing, observers say.
Freeing companies from the usual task of validation simply pushed that task onto labs. It’s an enormous job, says Dr. Caliendo. At the time she spoke to CAP TODAY, she and colleagues were designing a validation study to be used by the state of Rhode Island to validate a number of assays for possible use. She estimates more than a thousand specimens would be tested for the specificity study, including those collected pre-pandemic. “That’s a lot of work for a laboratory to do.”
And until the FDA revised its policy, more than 100 tests landed on the market, Dr. Storch says, “allowing for chaos.”
Adds Dr. Caliendo: “I am honestly shocked at the number of rapid tests out there. People just put them on the market and started selling them without validation data. So we have no idea if they work.”
Robert Christenson, PhD, professor of pathology and professor of medical and research technology, University of Maryland School of Medicine, dove into the chaos early. He says he began looking for a good antibody assay for SARS-CoV-2 around December. “There were no good options. There were some really dodgy assays coming out.”
He contacted eight companies, all with production outside the United States, and his lab began by using a lateral flow device. Because of supply line challenges, he decided it made more sense to develop a test in-house. The LDT is modeled after the CDC/NIH anti-SARS-CoV-2 assay that targets the spike protein antigen, and the lab is pursuing an EUA, says Dr. Christenson, who is also medical director of core laboratories and point-of-care services at the University of Maryland Medical Center.
Once the lab was well down that road, Ortho Clinical Diagnostics began offering an assay that also used the spike protein antigen. “So we also bought into that one,” he says.
Armed with two antibody assays, the laboratory plans to offer testing to all employees in the university’s health system, which has 28,000 to 30,000 employees, says Dr. Christenson, who spoke with CAP TODAY in mid-May. That should provide useful insight into the prevalence in the Baltimore area of SARS-CoV-2 infection among health care workers who want to learn if they’ve been exposed to the virus, he says.
As for the lateral flow assay? “It seemed pretty good,” he says. “But it was really expensive. My recommendation was to ask the company to send us 250 kits, and we’d validate the assay. But it was a sellers market and they said, ‘No—you either take the whole shipment or we’re going somewhere else.’” The device was later recalled.
Dr. Christenson says he now has two “dynamite” tests and plans to run both for the time being to improve the positive predictive value and to do comparison studies.
ARUP currently offers IgG testing, performed on two platforms: Abbott, which targets the nucleocapsid phosphoprotein, and Euroimmun, which targets the spike glycoprotein.
In the new normal, labs don’t have a lot of data on which to make their decisions. “I can say for the two platforms we tested, we didn’t see any significant differences, despite the fact that they use different antigens,” says Dr. Slev. Absent a robust literature, she says, “We have more to learn. I certainly cannot recommend one assay over another, at this point.”
[dropcap]H[/dropcap]aving two assays also offers peace of mind regarding the supply chain, which was often absent when laboratories were ramping up molecular diagnostic tests.
Though many of those problems stemmed from shortages of nasal swabs and transport media, which are not an issue with serology, concerns linger.
ARUP chose its tests because it wanted accurate, high-throughput assays, says Dr. Slev. But it also offers two tests to avoid a repeat of supply chain problems.
The initial antibody test free-for-all did nothing to ease anyone’s jitters. Dr. Christenson recalls the difficulties with some of the first serology assays he looked at. “They’d say, ‘OK, we can have this to you by next Monday.’ And then next Monday would come, and the following Monday would come, and—nothing. That’s why we just went ahead and bit the bullet and developed our own ELISA modeled after the CDC and NIH. The pipeline was ridiculous.”
Dr. Christenson says he doesn’t fear shortages, “now that we have our own pipeline to offer the test. I think Ortho has proved to have a good pipeline too.” He was expecting a large shipment from the company later in the week and wondered if it would come through. (It did.) The memories of earlier, PCR-related supply anxieties persist.
Waugh sounds less confident about the serology supply chain, and it has nothing to do with smaller companies. Rather, it’s the larger companies that give him pause. Typically, he notes, launching a new test would entail a slower, staged rollout. “We typically see launches in Europe, maybe Asia, and then the U.S.” That gives companies time to work out problems and develop a new supply chain for the U.S. market, pending FDA approval, which itself was a fairly drawn-out process.
But this is a pandemic. “It’s very different. They’re all launching quickly in a world market, essentially at the same time,” Waugh says. “Most of these companies typically make supplies in only one or two places globally. So we know there is going to be allocation of supplies to countries, and allocation to hotspots. And as we saw with PCR testing, we certainly have no reason to doubt that the federal agencies will try to steer supplies to commercial labs.
“I think we will watch this movie again,” he says.
In the meantime, he was planning to run two methods at the Henry Ford system by the end of May. One is from Beckman Coulter, an IgG assay targeting the spike protein. The other is a combination IgG/IgM antibody test from Roche. Like other labs, “We want to be able to compare performance characteristics of each and look for discordant data. We expect they’ll match quite well, but there could be differences, just because they’re built differently.”
When Henry Ford began looking at antibody tests, Waugh told his colleagues they didn’t need to pick the winning horse immediately, but they did have several criteria, including a company with a stable, well-established supply chain. “Somebody we’ve worked with, somebody who can deliver on time. Because of the experiences we had with PCR testing,” he says. Their first criterion is also telling: It had to be very well positioned for EUA approval if not already approved. Other criteria: It must demonstrate at least 98 to 99 percent sensitivity and focus on IgG, which is the superior predictor of long-term immunity. “And it must be aligned,” Waugh says, “with our existing test platforms and automation.”
[dropcap]T[/dropcap]here’s another reason the demand for antibody testing runs high: People simply want it. January and February now seem filled with cases that were retroactively diagnosed by patients themselves, who are sure their devastating flu-like symptoms came from COVID-19. Everybody wants to know whether they’ve been infected, says Dr. Christenson. “I don’t know what they’re going to do with that information.”
Dina Greene, PhD, technical director, Kaiser Washington Laboratories, says the pull is strong. “We want to believe we can’t get reinfected. But that’s not the way viruses always work.”
Nevertheless, “There are some people who are pushing this so aggressively,” she continues. “And I can’t understand why.” A positive test may not indicate whether an infected person will develop long-term lung sequelae. It may not mean a reduced risk of reinfection. It may not indicate the clinical course if reinfection does occur. “There are a lot of things that might be clinically important,” says Dr. Greene, speaking in early May, “but nobody actually knows what type of information they’re getting right now.”
She compares it to ancestry DNA testing. “You’re not looking for something targeted. You may find out things you don’t want to know.” And until answers come into focus—which could happen quickly, she acknowledges—“it’s dangerous to use these tests as certificates of immunity. If you go back to work cautiously, fine. But don’t go back to work because an antibody test told you you have antibodies against one of these antigens.”
Antibody testing has a role in public health and research, Dr. Greene says. “But until we understand immunity, I don’t think there’s a clinical role for it. And that’s getting messy right now.” She worries about widespread promotion of antibody testing as a clinical tool. “We always say, ‘If you don’t know what to do with a result, don’t order the test.’ But now people who’ve definitely been quoted saying that are saying, ‘Order serology. Let’s test everyone and give them all the results.’”
Further proof that a pandemic does upend nearly everything, Dr. Greene says, “I’m definitely not on the conservative side of the street for . . . anything. But I’m standing on the conservative side of the street right now for this.”
“I might be wrong,” she adds. “But we don’t know that yet.”
[dropcap]T[/dropcap]he lack of data has, paradoxically, been matched by a glut of information. As with the antibody tests themselves, some of it looks good, some of it doesn’t, and some of it is merely puzzling. In the past, researchers declined to talk about their findings until it was published in a peer-reviewed journal. That now seems like a quaint habit from a dusty era, like using finger bowls.
Dr. Slev calls it “a whole new world.”
“We’re so bombarded with information all the time, from so many angles, constantly,” she continues, making it difficult to discern what has been peer reviewed, where something has been published, and how studies have been performed. “Sometimes things are in preprint and have not been peer reviewed, primarily because there is such a need to understand this virus.”
“We all want to get papers out quickly,” agrees Dr. Greene. “The public and the scientific community are so hungry for information.” She says there’s plenty of credible work emerging. “We’re just not having enough time to digest it all, assimilate it, and credibly think about it. What are its limitations? What’s still missing? What else do we need to understand before we can use the results of the work to modify clinical practice?”
Dr. Theel notes many of the studies are based on LDTs, rather than the commercially available tests most labs might use.
“And a lot of data people have been using has been word of mouth,” she says. Indeed, during a Compass Group call in May, participants at one point began discussing information they’d heard about from a YouTube press conference.
“It has been a very surreal time,” Dr. Christenson says. Now that larger manufacturers are stepping in with reliable tests, he hopes evidence-based information will follow, although that too will take time. “Then we’ll know what we don’t know about these tests,” he says. At least that’s a start.
In the meantime, preprint papers and online publications are everywhere, with labs trying to sort through them all, not only for themselves but also for their clinical colleagues. That includes revisiting the basics of laboratory medicine.
Dr. Caliendo has spent much of her time talking to house staff and clinicians, “just laying out in very clear terms how the tests work, what we understand about the tests, what we understand about the disease. They need to understand the specimen type, the timing of the collection, the biology of the infection—all these will affect testing performance.”
The lack of information and an early glut of tests have made the lab’s job even harder, Waugh says. “I’ve said this to my own colleagues here: The next worry is that a lot of the public will assume that testing is all the same quality.” That leads to an ill-conceived definition of “best”—whatever is fast and cheap. “What we’re seeing, and what I think will bear out here, is that the fastest, cheapest tests have greater degrees of cross-reactivity with other coronaviruses,” he says. “There will be bad decisions made by people who opt for fast and cheap.”
Warns Dr. Storch: “A bad test can be worse than no tests.”
Ideally people would look at performance characteristics, as determined by rigorous FDA reviews and a robust, peer-reviewed literature. But this is a pandemic. “That process is just not in place right now,” says Waugh. With the drumbeat demands for testing—more tests! more tests!—“we don’t even care if they’re good tests or bad tests. Just more tests,” Waugh says.
[dropcap]L[/dropcap]abs are also having to steer more public conversations around serology testing. Some efforts have been more successful than others. As the lab tries to learn on the fly, the world is tapping its feet.
The peanut gallery is huge. “We’re dealing with an extremely high-profile disease and public health problem, which puts the decisions made by the laboratory directors under more scrutiny than they’ve ever been subjected to in the past,” says Dr. Storch.
When Dr. Caliendo hears national leaders talk casually and overconfidently about the meaning and availability of tests, she has one response: “They should just stop.” Until that day arrives, however, she and others will continue to talk to others outside the medical community who badly need to understand the implications of antibody testing, particularly on local and state levels.
“The voice of the experienced laboratorian or clinician” is so important to make available to the public, Dr. Caliendo says. “I have done more interviewing with the media about lab testing for COVID than I have done about lab testing for anything else over the last 20 years of my career.”
Are they listening? There’s reason for optimism, she says. She and colleagues have been releasing video talks on COVID-19 and SARS-CoV-2, and she’s been told viewership is high. She was also named co-chair of a state testing and validation task force. “Our governor is very smart, and she listens. She wants to see science and data. It’s so refreshing.”
Dr. Slev reports a similarly good experience with Utah’s governor and public health officials. “We have been reached out to, we have communicated with them, and they have been receptive and understanding about laboratory testing.”
If there’s a need to educate clinicians and government leaders, the same goes for the public and the media. Again, some of this work involves unraveling misinformation. Even if some of the early antibody tests are now in retreat, the poor perceptions they created have continued to mill about, like the peasants in a Chekhov play.
Amid the nonstop questions, discussions, rhetoric, and opinions about SARS-CoV-2 serology testing, other aspects of laboratory medicine have gone dark.
Dr. Theel says a colleague recently asked her a question about Chlamydia. It stopped her in her tracks. “It was mind-blowing,” she says, “that I could think about anything other than COVID.”
Karen Titus is CAP TODAY contributing editor and co-managing editor.