SARS-CoV-2 testing options that screen for active illness or past infection have expanded rapidly over the last few months. As institutions craft guidelines on optimizing polymerase chain reaction (PCR) diagnostic tests and serology antibody tests, some labs have been developing their own testing and result reporting algorithms.
Since the U.S. Food and Drug Administration’s (FDA) first emergency use authorization of antibody tests in March, developers have responded in high numbers, prompting the FDA, other federal entities, and organizations to issue comprehensive guidance on the uses for serologic testing. AACC, in its recommendations on SARS-CoV-2 antibody testing, covers the utility, performance, and limitations of these assays, and what they should and shouldn’t be used for in managing COVID-19, the illness caused by the new coronavirus. “Serology testing is complementary to molecular diagnostic testing in managing the COVID-19 pandemic. It may play an important role in assessing the prevalence of the disease and may support epidemiological efforts such as contact tracing while research into anti-viral therapies and vaccines continues,” AACC stated.
AACC stressed that antibody tests should not be used as a primary method to diagnose acute infection or exclude SARS-CoV-2 infection when a patient presents with symptoms. In other guidance, TriCore Reference Laboratories stated that serology testing should be used to indicate infection following a negative PCR test in symptomatic patients who later develop the illness, and to identify previously infected people for convalescent plasma donation.
The American Medical Association further clarifies the limitations of serology tests, including the risk for false positives. SARS-CoV-2 has low disease prevalence, which means that false positives will appear among high performing tests. “Take, for example, a community of 100 individuals with a disease prevalence of 5%. If a serological test with a specificity of 95% was used in this population, it would be expected to return 5% false positives, so 5 out of the population of 100,” AMA explained.
The quality of some serology tests has been questioned. FDA responded by setting 90% sensitivity and 95% specificity performance characteristics for approved tests and added a new section to clarify which tests should no longer be distributed. As part of an independent performance validation study with the National Institutes of Health’s National Cancer Institute, FDA also published test performance data on several antibody test kits on open.fda.gov.
Other government entities have weighed in on SARS-CoV-2 testing. The White House, in its blueprint on testing platforms, advocates for using newer technologies such as antigen and nucleic acid tests at the point-of-care to boost testing capacity, and using antibody tests to calculate community spread of SARS-CoV-2. Guidelines from the U.S. Centers for Disease Control and Prevention (CDC) offer three strategies to improve positive predictive value in SARS-CoV-2 testing:
- Choose a test with a very high specificity (99.5% or greater) to yield a high positive predictive value in populations with prevalence >5%.
- Focus testing on individuals who likely have SARS-CoV-2 antibodies, including those who may have had COVID-19-like illness.
- Use an orthogonal testing algorithm that administers a second test to individuals who initially test positive.
“Effective orthogonal algorithms are generally based on testing a patient sample with two tests, each with unique design characteristics (e.g., antigens or formats),” CDC indicated. Notably, the agency stated that for populations with low SARS-CoV-2 prevalence (<5%), a test with 90% sensitivity and 95% specificity will have a positive predictive value of just 49%. This means that less than half who test positive will actually have COVID-19 antibodies.In a consolidated testing strategy that includes high-risk populations and critical infrastructure sectors, CDC describes five categories of people for SARS-CoV-2 testing.
Some individual labs have developed their own strategies and algorithms on testing. In a Q&A page, Quest Diagnostics says it only uses serology tests whose clinical performance ranges from 90% to 100% and has a specificity of approximately 99% to 100%.
Extensive testing protocols from Boston University Medical Center offer instruction on testing, specimen collection, and testing considerations based on hospital location and at discharge, and testing considerations for individuals with certain conditions, such as immunocompromised patients receiving biologic treatments or those returning to outpatient hemodialysis.
Flavio F. Alcantara, MD, PhD, laboratory director of IACS (Santos Institute for Clinical Analysis), a private independent medical laboratory in Santos, Brazil, shared on AACC Artery an algorithm his lab developed for SARS-CoV-2 testing, outlining testing protocols for symptomatic and asymptomatic patients, and serology assay choices.
For symptomatic patients, the algorithm recommends RT-PCR during the first 7 days of symptoms, PCR and serology (IgM + IgG) within 7 to 14 days of symptoms, and serology after 14 days once a patient’s quarantine ends, preferably using an assay that detects IgG anti-nucleocapsid antibodies. “Nucleocapsid is the most abundant protein in the virus and likely to generate robust antibody responses,” wrote Alcantara. His lab was still working out the serology part of the algorithm, considering a two-step process, with anti-nucleocapsid antibodies as a screening assay and anti-spike (RBD) antibodies as confirmatory assay.
The algorithm was based on a literature review and personal experience, Alcantara told CLN Stat. So far, it’s been working well, “although this is an opinion and not a conclusion based on any study,” he added. The goal is to release an improved version soon. “We always report the type of antigen used/the antibody detected. Personally, I believe anti-S (RBD) antibodies will soon be important for the monitoring of vaccine responses, therefore, I suggest attention is paid to the type of antibodies detected,” said Alcantara.
Alcantara’s lab has sent e-flyers describing the algorithm and the virus and antibody profile curves to its colleagues.
Looking forward, FDA just approved the first single antigen assay to detect active infection. Depending on the sensitivity of the assay, an antigen test may be a real game changer, Alcantara said. Most PCR assays for SARS-CoV-2 have a limitation of detection (LoD) of <100 copies/mL. “If the antigen test has a LoD not too far from that, say <1.000 copies/mL, it could potentially allow on-the-spot detection of patients in the contagious phase,” he said. While important for triaging of patients in hospitals, the antigen test may have other uses. “It will certainly be more affordable, and if mass produced and delivered, might be used even out of the clinical laboratory context, in drugstores and pharmacies,” Alcantara said.