key: cord-0826636-l2f955pu
authors: Kanjilal, S.; Chalise, S.; Shami Shah, A.; Cheng, C.-A.; Senussi, Y.; Uddin, R.; Thiriveedhi, V.; Cho, H. E.; Carroll, S.; Lemieux, J.; Turbett, S.; Walt, D. R.
title: Performance of three rapid antigen tests against the SARS-CoV-2 Omicron variant
date: 2022-02-19
journal: nan
DOI: 10.1101/2022.02.17.22271142
sha: 6abe121df67daf3c508e4fb25391c1c4db597aaf
doc_id: 826636
cord_uid: l2f955pu

Rapid antigen detection tests (RADTs) for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) are now in widespread use in the United States. RADTs play an important role in maintaining an open society but require periodic reassessment to ensure test performance remains intact as the virus evolves. The nucleocapsid (N) protein is the target for the majority of RADTs and the SARS-CoV-2 Omicron variant has several N protein mutations that are previously uncharacterized. We sought to assess the impact of these mutations by testing 30 Omicron variant samples across a wide range of viral loads on three widely used RADTs: the iHealth COVID-19 Antigen Rapid Test, the ACON Laboratories FlowFlex COVID-19 Antigen Home Test, and the Abbott BinaxNOW COVID-19 Antigen Card, using 30 Delta variant samples as a comparator. We found no change in the analytic sensitivity of all three RADTs for detection of Omicron versus Delta, but noted differences in performance between assays. No RADT was able to detect samples with a cycle threshold (Ct) value of [≥]27.5 for the envelope gene target on the Roche cobas RT-PCR assay. Epidemiologic studies are necessary to correlate these findings with their real-world performance.

Diagnostic testing for infection by severe acute coronavirus syndrome 2 (SARS-CoV-2) remains 41 a cornerstone of efforts to control the coronavirus disease 2019 (COVID-19) pandemic. The 42 reliance on centralized laboratory-based testing has eased with the introduction of rapid antigen 43 tests, which are now available over-the-counter in the US or provided by the federal government. 44

These assays can be self-administered, require little to no equipment and provide results within 45 15 minutes. The ability to test at the point of care with the onset of symptoms or prior to gatherings 46 places them in a key role for maintaining an open society. 47

The SARS-CoV-2 nucleocapsid (N) is the most abundant protein expressed by the virus 1,2 and is 48 the target of the majority of rapid antigen detection tests (RADTs). Detection of the analyte is 49 achieved through recombinant antibodies conjugated to gold nanoparticles that target specific 50 epitopes on the N protein. The antigen-antibody complexes are carried by capillary action to a 51 second set of antibodies which immobilize and concentrate the nanoparticles, making them visible 52 to the naked eye. Mutations in the N protein have been previously described to cause decreases 53 in antigen test sensitivity 3 , therefore periodic reassessment of test performance is necessary as 54 new variants arise. The Omicron variant is characterized by a mutation (P13L) and a deletion 55 (∆31-33) near the N-terminal domain and two mutations adjacent to each other in the linker 56 domain (R203K and G204R) 4 . We sought to characterize the impact of these newly described 57 mutations on the analytic sensitivity of three widely used RADTs for at-home testing: the 58 BinaxNOW COVID-19 Antigen Card (Abbott, Scarborough, Maine), the iHealth COVID-19 59

Antigen Rapid Test (Sunnyvale, California), and the FlowFlex COVID-19 Antigen Home Test 60 (ACON Laboratories, San Diego, California), using their performance versus the Delta variant as 61 a comparator. 62 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Waltham, Massachusetts), which amplifies a set of 6 spike protein mutations that characterize 73 major variants of concern, including the Omicron and Delta variants. One sample was positive for 74 only one target (K417N) due to having very low amounts of nucleic acid, but was assumed to be 75 For each RADT, we mixed the kit-supplied swab with 50 μL of sample for 15 seconds and then 80 followed each assay's instructions for use. Samples underwent one freeze-thaw cycle prior to 81 examination on the iHealth assay and two freeze-thaw cycles for all other assays. Assays were 82 run in duplicate for each RADT and results were evaluated after a 15 minute incubation period by 83 two independent readers blinded to the variant status and Ct value of the sample. Samples were 84 run for all three RADTs over a 2 day period. Chi-squared tests were used to compare the 85 distribution of results by variant for a given RADT and logistic regression was used to estimate 86 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) iHealth COVID-19 Antigen 0/10 (0%) 0/10 (0%)

. CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 19, 2022. ; https://doi.org/10.1101/2022.02.17.22271142 doi: medRxiv preprint between Delta and Omicron variant samples, we note differences in the performance by assay 124 type. These results should be interpreted with caution given our relatively small sample size. We 125 also note a non-significant trend towards decreased detection of the Omicron variant for the 126 Abbott BinaxNOW, which we have noted in a prior study from our group using a different sample 127 showed equivalent sensitivities across a range of dilutions 6 . In contrast, a similar study by Bekliz 134 et al using viral culture and paired clinical samples did find attenuated analytic sensitivity for 135 detection of Omicron by the Panbio test, relative to the Delta variant 9 . The Panbio uses the same 136 N protein epitopes as the BinaxNOW, which is marketed in the United States. The reasons for 137 these differing results are not fully known, but they highlight the need for a repeat study using a 138 larger set of samples with Ct values in the 20 to 30 range to resolve whether the BinaxNOW and 139

Panbio assays have lower analytic sensitivity for Omicron. 140 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted February 19, 2022. ; https://doi.org/10.1101/2022.02.17.22271142 doi: medRxiv preprint

The Architecture of SARS-CoV-2 Transcriptome

A high-resolution temporal atlas of the SARS-CoV-2 translatome and 189 transcriptome

A SARS-CoV-2 Nucleocapsid Variant that Affects Antigen Test 191 Performance

Analytic sensitivity of the Abbott BinaxNOW TM lateral flow 194 immunochromatographic assay for the SARS-CoV-2 Omicron variant

Assessment of the analytical sensitivity of ten lateral flow devices against 197 the SARS-CoV-2 omicron variant

Detection of the omicron variant virus with the Abbott BinaxNow SARS-200 CoV-2 Rapid Antigen Assay

Direct Comparison of SARS Co-V-2 Nasal RT-PCR and Rapid Antigen 203 Test (BinaxNOW(TM)) at a Community Testing Site During an Omicron Surge

Sensitivity of SARS-CoV-2 antigen-detecting rapid tests for Omicron 206 variant

The truth about SARS-CoV-2 cycle threshold values is 208 rarely pure and never simple

Transmission of COVID-19 in 282 clusters in Catalonia, Spain: a cohort 210 study

CoV-2: A Review of Viral, Host, and Environmental Factors

Analytical 215 Sensitivity of the Abbott BinaxNOW COVID-19 Ag Card

The largest clinical study examining the BinaxNOW was performed at a community testing site in 141 San Francisco during a time when rates of test positivity exceeded 40% 8 . The BinaxNOW was 142 able to reliably detect positive samples up to a Ct of 30 for the PCR assay used in this study. 143 While this value is higher than our detection threshold of 27.5 cycles for the E gene of the cobas 144 SARS-CoV-2 RT-PCR assay, this could be explained by differences in reaction efficiencies, 145 thresholding algorithms and a host of other factors as opposed to true differences in viral loads 10 . The iHealth COVID-19 test has been distributed to millions of US citizens through a free 156 distribution program offered by the federal government. To our knowledge, this study is the first 157 to independently evaluate the performance of this assay against the Omicron variant. While both 158 the iHealth and FlowFlex RADTs performed well with our sample set, the iHealth test had a trend 159 towards higher sensitivity and also had the best discrimination between positive and negative is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 19, 2022. ; https://doi.org/10.1101/2022.02.17.22271142 doi: medRxiv preprint overlap at these viral loads is not known, a significant number of people tested during an outbreak 165 may have levels that fall within this range. Further study is necessary to understand whether this 166 variation is reproducible in other contexts. 167The major limitation of our study is the sample size, which limits drawing statistically significant 168 conclusions regarding small differences in test performance. A larger study is warranted to further 169 investigate the differences seen between our RADTs, as even small differences can have a large 170 impact when scaled to the population level. Another limitation is our use of frozen samples in 171 universal transport media rather than direct testing from a patient, but we would not expect there 172 to be a major impact from one to two freeze-thaws on assay performance and the volume of 173 analyte used in each assay was optimized in an earlier study for mimicking real-world 174 performance 12 . A major strength of this study was the ability to compare three different RADTs 175 using identical clinical samples, which allows for a robust comparison of performance. 176In summary, the analytic sensitivity versus Omicron remains stable in our head-to-head 177 comparison of three of the most common RADTs in use in the United States. However, there 178 were differences in inter-assay performance that warrant further study. Our findings will provide 179 a degree of assurance that at-home testing should perform as expected compared to prior waves 180 and also sets a baseline for comparison with future SARS-CoV-2 variants. 181

This study was funded by a grant from the Massachusetts Consortium for Pathogen Readiness. 183

None. 185 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 19, 2022. ; https://doi.org/10.1101/2022.02.17.22271142 doi: medRxiv preprint . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)The copyright holder for this preprint this version posted February 19, 2022. ; https://doi.org/10.1101/2022.02.17.22271142 doi: medRxiv preprint