Which COVID-19 serological assays to choose?

A head-to-head study comparing five serological assays for SARS-CoV-2 antibodies showed that a commercial immunoassay by Siemens and a novel ELISA* test developed by the University of Oxford could achieve the desired sensitivity and specificity without further optimization — providing reassurance for their use.

“To date, few thorough, direct assessments of immunoassay performance on large sample sets have been done,” said the researchers.

Driven by the urgent need to balance demand for serological testing and limited data on assay performance, the typical assessment criteria for regulatory approval for commercial use have been relaxed during the pandemic.

The researchers therefore aimed to facilitate healthcare providers in choosing among the different platforms available for SARS-CoV-2 serological testing by conducting a head-to-head comparison of four commercially available immunoassays (manufactured by Abbott, DiaSorin, Roche, and Siemens) and one new 384-well ELISA that can be scaled up. [Lancet Infect Dis 2020;doi:10.1016/S1473-3099(20)30634-4]

For performance testing, the researchers used 976 blood samples collected years before the pandemic (ie, negative for SARS-CoV-2 antibodies) and 536 samples from patients with COVID-19, collected at least 20 days after symptom onset.

“By running all the assays on the same large panel of blood samples, we showed that the Siemens assay and the Oxford immunoassay both achieved sensitivity and specificity of at least 98 percent on samples taken at least 20 days post symptom onset, in line with the current MHRA** guidance for the regulatory approval of these tests,” reported Dr Nicole Stoesser from the University of Oxford, UK, during the ECCVID 2020 Meeting.  

Sensitivity and specificity were 98.1 percent and 99.1 percent, respectively, for the Siemens assay and 99.1 percent and 99.0 percent, respectively, for the Oxford ELISA, when used exactly according to the manufacturers’ specifications. 

The corresponding values were 92.7 percent and 99.9 percent for the Abbott assay, 95.0 percent and 98.7 percent for the DiaSorin assay, and 97.2 percent and 99.8 percent for the Roche assay.

Nonetheless, all assays tested could achieve the desired sensitivities and specificities of ≥98 percent by adjusting assay threshold or assessing samples collected ≥30 days after onset of symptom.

“Although precise performance metrics varied between immunoassay platforms, all assays that we assessed could be usefully deployed with careful consideration of use case, assay thresholds, and by considering symptom-to-sample timings, thus optimizing the use of available serological testing resource and enabling the most widespread rollout,” pointed out Stoesser and co-authors.

As SARS-CoV-2 antibodies usually only start to appear 5–7 days after infection, the use of such assays is therefore more for detecting the extent of population exposure rather than identifying early acute infection, they explained.

“There is no such thing as a ‘perfect test’, but accurately evaluating how these tests perform can help us understand their limitations and improve how they are used,” said Stoesser. “Importantly, consideration needs to be given to how many false-positive and false-negative results might occur with any given test.”

“This study represents a benchmark for future assessments of serological platforms … [and] should enable refinements of testing strategies and the best use of serological testing resource to benefit individuals and population health,” the researchers stated.

*ELISA: Enzyme-linked immunosorbent assay
**MHRA: UK Medicines and Healthcare products Regulatory Agency