Antigen tests for Covid-19 are under the spotlight. Both the UK and US governments have plans to use these relatively cheap assays to screen large numbers of people, including those without symptoms, but concerns are mounting that antigen tests’ accuracy among asymptomatic people is insufficient.
Research into one of Abbott Laboratories’ antigen tests, however, gives cause for hope. A study of the Panbio Covid-19 Ag Test, published ahead of peer review, suggests that its accuracy might be acceptable even among asymptomatic people. Abbott declined to comment on whether it intends to seek regulatory authorisation for its antigen tests in the screening setting, but further investigation here is surely warranted.
The researchers, led by Oriol Mitjà of the Fight Aids and Infectious Diseases Foundation in Badalona, Spain, compared Abbott’s Panbio assay with a reference PCR-based test. The tests were run on nasopharyngeal swab samples collected for case confirmation and contact tracing, as well as swabs collected from asymptomatic individuals.
|Data on Abbott's Panbio Ag Covid-19 antigen test|
|Overall trial population||951||91.7%||455||98.9%|
|Source: medRxiv preprint.|
In the overall cohort, the sensitivity and specificity of the test were within the same range as the six antigen tests the FDA has authorised, though its sensitivity was a little lower than Abbott’s BinaxNow antigen test (Balancing the accuracy and cost of antigen testing, October 26, 2020).
BinaxNow is the US-focused sister test to Panbio Ag. They use the same biologics but are in different formats: Panbio Ag uses a small cassette whereas BinaxNow is a card. BinaxNow was authorised by the FDA in August but will not be released in Europe; likewise Abbott will not be seeking FDA authorisation for Panbio Ag.
The crucial point of the Spanish study is that among the subjects with no symptoms, the test’s specificity barely dropped, and it correctly identified nearly 80% of those unwittingly carrying the coronavirus.
It is interesting that the test maintained its high specificity among symptom-free people. The FDA wrote to lab staff and healthcare workers earlier this week, warning of false positives with Covid-19 antigen testing and pointing out that those antigen tests it has authorised are generally indicated only for use in people who have had Covid-19 symptoms. False positives are a consequence of tests with low specificity, and thus unlikely to be a major problem with Panbio Ag.
The test is CE marked and on sale in Europe, but Abbott’s own web page on Panbio Ag clearly states that it is “for patients suspected of current COVID-19 infection”.
Expanding it into the asymptomatic population is by no means an impossible goal, however – even a test with a false-negative rate of 20% would find ready buyers. Abbott was cagey about its plans in this area, but it would surely be unwise to rule out further development of Panbio Ag, and perhaps BinaxNow too, as screening tests for Covid-19.