Danaher subsidiary Beckman Coulter has joined the ranks of major diagnostics manufacturers with FDA's go-ahead to sell a COVID-19 serology test, disclosing an emergency use authorization Monday for an anti-SARS-CoV-2 antibody test.
Later to receive an emergency use authorization than some rivals, Beckman will need to show its test can match or beat assays from companies such as Abbott, Roche and Siemens Healthineers.
Data quoted from Detroit's Henry Ford Health System’s assessment of Beckman's antibody test said it correctly identified antibodies in 100% of samples taken from COVID-19 patients 14 days after the confirmation of infection by PCR testing. Bernard Cook, division head of chemistry-pathology at Henry Ford, said in Beckman's statement the antibody test’s performance in a “rigorous independent evaluation” led the health system to make the assay the “backbone” of its serology testing program.
The Beckman-generated data shared by FDA in the test’s instructions for use, however, paint a more mixed picture of its performance.
In that study, the test correctly detected antibodies in 96.8% samples taken 15 or more days after a PCR-confirmed diagnosis of COVID-19. The figure, which is called the sensitivity or positive percent agreement (PPA), is lower than those quoted in the data sheets of rival antibody tests.
Serology tests developed by Abbott, Roche and Siemens all achieved a 100% PPA when used to spot antibodies in samples taken 14 or more days after a PCR-confirmed diagnosis. All the tests, including Beckman’s assay, achieved a negative percent agreement of between 99.6% and 99.8%.
If the results represent an accurate comparison of the tests, they suggest the Beckman assay is slightly more likely to deliver false positives than the other products when used to analyze samples taken around two weeks after diagnosis with COVID-19.
Still, other aspects of the data suggest it may still have a unique role to play in the response to COVID-19. Notably, the Beckman test appears to be far better than other assays at detecting antibodies against SARS-CoV-2 in the first week after diagnosis with COVID-19.
Applied to samples taken up to one week after diagnosis, the Beckman test detected antibodies 75.8% of the time. Abbott, Roche and Siemens divided their results up slightly differently in EUA filings, but their comparable figures range from 25.0% to 65.5%. The Beckman test more reliably detected antibodies than the Abbott and Roche assays in samples taken eight to 14 days after diagnosis, too.
Beckman’s test, like the Abbott assay, detects the IgG antibodies that start developing around one week after the onset of symptoms. The Roche and Siemens assays are total antibody tests, meaning they combine the detection of IgG, IgM and IgA. Research suggests IgM antibodies are the first to form but their levels fall to zero over time.
Independent analyses
A clearer picture of the relative strengths of the products may emerge as independent organizations share the findings of their assessments of the assays. FDA has tested around 20 antibody assays as part of its push to remove inaccurate tests from the market, but has not shared evaluations of Abbott, Beckman, Roche and Siemens products. Similarly, the U.K. has initiated a real-world comparison of antibody tests but is yet to publish its findings.
The comparisons publicly available suggest there may be meaningful differences between the performance of the tests, despite the similarity of the results generated by their manufacturers. One study from Medical University of Vienna posted to preprint server medRxiv in early June found tests made by Abbott, DiaSorin and Roche fell well short of their claimed performance, achieving sensitivities ranging from 83.1% to 89.2%.
Used in a population in which 5% of people have antibodies, the sensitivities translate into positive predictive values ranging from 94.8% for Roche’s test to 71.6% for DiaSorin’s assay. The size of the difference led the researchers to state “we must assume that these two assays differ significantly from each other in terms of positive predictability.”
At the start of June, analysts at Jefferies estimated the U.S. will need 60 million antibody tests for the “first COVID-19 season.” Beckman plans to ship 30 million tests a month, although some of the assays will go to markets other than the U.S. Abbott aimed to have capacity for 60 million antibody tests in June. Siemens plans to ramp up to 50 million antibody tests a month. And Roche had a goal to provide “high double-digit millions of tests.”
Abbott, DiaSorin and Roche secured EUAs for their antibody assays around two months before FDA authorized Beckman’s test. Siemens also has a head start of several weeks over Beckman. Still, the Jefferies analysts expect Beckman to carve out a "large share of future antibody testing," noting its high-volume analyzers infrastructure and third or fourth position in the worldwide immunoassay market, according to a June 1 note.
