Sensitivity of rapid antigen tests for Omicron subvariants of SARS-CoV-2
Abstract
Diagnosis by rapid antigen tests (RATs) is useful for early initiation of antiviral treatment. Because RATs are easy to use, they can be adapted for self-testing. Several kinds of RATs approved for such use by the Japanese regulatory authority are available from drug stores and websites. Most RATs for COVID-19 are based on antibody detection of the SARS-CoV-2 N protein. Since Omicron and its subvariants have accumulated several amino acid substitutions in the N protein, such amino acid changes might affect the sensitivity of RATs. Here, we investigated the sensitivity of seven RATs available in Japan, six of which are approved for public use and one of which is approved for clinical use, for the detection of BA.5, BA.2.75, BF.7, XBB.1, and BQ.1.1, as well as the delta variant (B.1.627.2). All tested RATs detected the delta variant with a detection level between 7500 and 75 000 pfu per test, and all tested RATs showed similar sensitivity to the Omicron variant and its subvariants (BA.5, BA.2.75, BF.7, XBB.1, and BQ.1.1). Human saliva did not reduce the sensitivity of the RATs tested. Espline SARS-CoV-2 N showed the highest sensitivity followed by Inspecter KOWA SARS-CoV-2 and V Trust SARS-CoV-2 Ag. Since the RATs failed to detect low levels of infectious virus, individuals whose specimens contained less infectious virus than the detection limit would be considered negative. Therefore, it is important to note that RATs may miss individuals shedding low levels of infectious virus.
1 INTRODUCTION
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causes an acute respiratory illness known as coronavirus disease 2019 (COVID-19). The WHO reported that more than 750 million cases of COVID-19, including approximately 6.83 million deaths, had occurred as of January 30, 2023 (https://covid19.who.int/). To reduce the burden of SARS-CoV-2, nonpharmaceutical interventions, vaccination, and patient treatment, including specific antivirals, have been employed. If a person with risk factors for severe COVID-19 becomes infected with SARS-CoV-2, early and reliable diagnosis is essential to initiate early treatment with antivirals to prevent severe illness.
The gold standard for COVID-19 diagnosis is reverse transcription-quantitative polymerase chain reaction (RT-qPCR) testing using upper respiratory tract swabs or saliva1 because of this test's high sensitivity and specificity for SARS-CoV-2. However, RT-qPCR requires specialized equipment and several hours to obtain test results. For quick results in local clinics, rapid antigen tests (RATs) for COVID-19 are popular because RATs take just 15–30 min. Because of this convenience, RATs are also useful as home self-tests for symptomatic individuals. Although the sensitivity of RATs is lower than that of RT-qPCR,2-9 RATs with a short turnaround time can help improve diagnosis. To this end, we and others previously compared the sensitivity of RATs and found that most RATs show similar sensitivity against ancestral SARS-CoV-2, SARS-CoV-2 possessing the D614G substitution in its S protein, and the delta variant (lineage B.1.617.2).9-15 RAT performance has also been evaluated for Omicron variant detection, including BA.1 and BA.2.16-22 Several groups found that RATs maintained sensitivity to omicron variants,18, 21 but others reported that RATs performed poorly with these variants.16, 20, 23 The decreased sensitivity could be associated with a lower viral load of Omicron than delta.24 Therefore, sensitivity comparisons of RATs using clinical specimens may be affected by changes in viral properties.
RATs specifically detect the N protein of SARS-CoV-2 by using antibodies against this protein. Therefore, amino acid substitutions in the N protein might affect the sensitivity of RATs. The delta variant possessed the D63G, R203M, and D377Y substitutions in its N protein, whereas most of the Omicron variants share the P13L, del31/33, R203K, and G204R changes in their N protein (Table 1). In addition to these changes, BA.5, BA.2.75, and XBB.1 acquired the S413R substitution, BF.7 had the G30F and S413R substitutions, and BQ.1.1 possessed the E136D and S413R substitutions. It is unclear whether these substitutions in the N protein affect the sensitivity of RATs.
| Variant | Amino acid changes in the N proteina |
|---|---|
| Delta | D63G, R203M, and D377Y |
| BA.5 | P13L, del31/33, R203K, G204R, and S413R |
| BA.2.75 | P13L, del31/33, R203K, G204R, and S413R |
| BF.7 | P13L, G30F, del31/33, R203K, G204R, and S413R |
| XBB.1 | P13L, del31/33, R203K, G204R, and S413R |
| BQ.1.1 | P13L, del31/33, E136D, R203K, G204R, and S413R |
- a Amino acid changes were compared with the prototype Wuhan-Hu-1 (MN908947.3).
Accordingly, here we investigated the sensitivity of RATs approved for general use by the Japanese regulatory authority and available in Japan in December 2022 for their detection of BA.5, BA.2.75, BF.7, XBB.1, and BQ.1.1.
2 METHODS
2.1 Biosafety statements
All experiments with SARS-CoV-2 were performed in biosafety level 3 (BSL3) laboratories at the University of Tokyo, which are approved for such use by the Ministry of Health, Labour and Welfare, Japan.
2.2 Ethics statements
Human saliva samples were collected by following protocols approved by the Research Ethics Review Committee of the Institute of Medical Science, the University of Tokyo. Signed informed consent was obtained from all participants.
2.3 Cells and virus
VeroE6 cells expressing human serine protease TMPRSS2 (VeroE6-TMPRSS2) were maintained in DMEM containing 10% fetal bovine serum (FBS), 1 mg/mL G418, 100 units/mL penicillin, 100 µg/mL streptomycin, and 5 μg/mL plasmocin prophylactic (Invivogen) and incubated at 37°C under 5% CO2. VeroE6-TMPRSS2-T2A-ACE2 cells were cultured in DMEM containing 10% FBS, 100 U/mL penicillin–streptomycin, and 10 µg/mL puromycin. Specimens for the preparation of viral stocks were obtained from deidentified, residual nasal swabs collected in Tokyo, Japan, or Wisconsin, USA. SARS-CoV-2 delta variant hCoV-19/USA/WI-UW-5250/2021 (lineage B.1.617.2), Omicron variant hCoV-19/Japan/TY41-702/2022 (lineage BA.5), Omicron subvariants hCoV-19/Japan/TY41-716/2022 (lineage BA.2.75), hCoV-19/USA/WI-UW-13890/2022 (lineage BF.7), hCoV-19/USA/NY-MSHSPSP-PV73997/2022 (lineage XBB.1), and hCoV-19/Japan/TY41-796/2022 (lineage BQ.1.1) were propagated in VeroE6-TMPRSS2 cells and titrated in VeroE6-TMPRSS2-T2A-ACE2 cells. The whole genome of stock virus was amplified by using a modified ARTIC network protocol in which some primers were replaced or added.25 Viral RNA was extracted by using a QIAamp Viral RNA Mini Kit (QIAGEN). cDNA was synthesized by using a Lunar-Script RT SuperMix Kit (New England BioLabs) and subjected to a multiplexed PCR in two pools using ARTIC-N1 primers v5 and Q5 Hot Start DNA polymerase (New England BioLabs). The DNA libraries for Illumina NGS were prepared from pooled amplicons by using a QIAseq FX DNA Library Kit (QIAGEN) and then analyzed using the iSeq. 100 System in 150-bp paired-end mode and an iSeq. 100 i1 Reagent v2 (300-cycle) kit (Illumina). The reads were assembled by CLC Genomics Workbench (version 23, Qiagen), and the extracted consensus sequences were analyzed by Nextclade (https://clades.nextstrain.org/) to assign the clade.
2.4 RATs
The RATs listed in Table 2 were evaluated according to the procedures described in the manufacturers' instructions, using 75–75 000 plaque-forming unit (pfu) of stock virus in a 50-μL volume. Each virus was diluted using cell culture medium in the presence or absence of human saliva. Cell culture medium containing human saliva was prepared by mixing the cell culture medium (8 mL) with human saliva specimens A, B, or C (0.5–1 mL), which were negative for SARS-CoV-2 by a RAT. Two independent experiments were performed with each dilution.
| No. | RAT name | Manufacturer | Country of origin | Used by | Recommended test sample | Labeling of detection antibody | Time to result (min)a |
|---|---|---|---|---|---|---|---|
| 1 | Espline SARS-CoV-2 (Clinical use) | Fujirebio | Japan | Medical professionals | Nasal or nasopharyngeal swab | Alkaline phosphatase | 10–30 |
| 2 | Espline SARS-CoV-2 N (General use) | Fujirebio | Japan | General public | Nasal swab or saliva | Alkaline phosphatase | 20 |
| 3 | SARS-CoV-2 Rapid Antigen Test | Roche diagnostics | Switzerland | General public | Nasal swab | Colored particle | 15–30 |
| 4 | CLINITEST Rapid COVID-19 Antigen self-test | Siemens Healthcare Diagnostics | Germany | General public | Nasal swab | Gold-colloid | 15–20 |
| 5 | HEALGEN COVID-19 Rapid Antigen self-test | Takara bio | Japan | General public | Nasal swab | Gold-colloid | 15–20 |
| 6 | Inspecter KOWA SARS-CoV-2 | Medical and Biological Laboratories | Japan | General public | Saliva | Gold-colloid | 15–20 |
| 7 | V Trust SARS-CoV-2 Ag | Nipro | Japan | General public | Nasal swab | Gold-colloid | 15 |
- a The time required to obtain the results is based on the individual manufacturer's instructions.
3 RESULTS
3.1 Comparison of RATs
We evaluated one RAT approved for use by medical professionals in clinical settings (#1) and six RATs approved for use by the general public (#2–7) by the Japanese regulatory authority that were available in Japan in December 2022 (Table 2). These seven RATs use a well format; lysis buffer containing the specimen is dropped into the well, and the reaction occurs inside a covered plastic body. The well format is suitable for general use because the human eye can evaluate the results without a machine. Recommended test specimens are nasal swabs and/or saliva for the six RATs for general use (#2–7) and nasal or nasopharyngeal swabs for the RAT for clinical use (#1).
All tested RATs are immunochromatographic tests, meaning that their sensitivity is dependent on the composition of the lysis buffer, the epitopes of the antibodies used, the binding kinetics of the antibodies to the N protein, the volume of specimen used for analysis, and the method used to visualize the immunocomplex of antibody and N protein. We cannot compare the performance of the antibodies in these tests because the manufacturers do not supply information about them. The composition of the lysis buffer is also undisclosed, and it may influence the properties of the individual antibodies. The amount of specimen used for each test varied among the RATs. The highest input ratio (i.e., the amount of sample used for the test as calculated using the formula in the Table 3 legend) was for the CLINITEST Rapid COVID-19 Antigen Self-test (#4) and the HEALGEN COVID-19 Rapid Antigen self-test (#5) at 28.6%, and the lowest was for Espline SARS-CoV-2 (Clinical use) (#1) and Espline SARS-CoV-2 N (General use) (#2) at 8.0% (Table 3). The input ratios for SARS-CoV-2 Rapid Antigen Test (#3), Inspecter KOWA SARS-CoV-2 (#6), and V Trust SARS-CoV-2 Ag (#7) were intermediate at 17.1%, 13.3%, and 17.8%, respectively. The method for visualization of the results also differed among the seven RATs: Espline SARS-CoV-2 (Clinical use) (#1) and Espline SARS-CoV-2 N (General use) (#2) use alkaline phosphatase and its substrate; SARS-CoV-2 Rapid Antigen Test (#3) uses colored particles; and CLINITEST Rapid COVID-19 Antigen Self-test (#4), HEALGEN COVID-19 Rapid Antigen self-test (#5), Inspecter KOWA SARS-CoV-2 (#6), and V Trust SARS-CoV-2 Ag (#7) use gold colloid for visualization of the immune-complex (Table 3). The results are assessed 10–30 min after adding the analyte (Table 2).
| No. | RAT name | Input ratio (%)b | Virus | |||||
|---|---|---|---|---|---|---|---|---|
| Delta | BA.5 | BA.2.75 | BF.7 | XBB.1 | BQ.1.1 | |||
| 1 | Espline SARS-CoV-2 (Clinical use) | 8.0 | 7500c | 7500 | 7500 | 7500d | 7500 | 7500 |
| 2 | Espline SARS-CoV-2 N (General use) | 8.0 | 7500 | 7500 | 7500 | 7500 | 7500 | 7500 |
| 3 | SARS-CoV-2 Rapid Antigen Test | 17.1 | 75 000 | 75 000 | 75 000 | 75 000 | 7500 | 75 000 |
| 4 | CLINITEST Rapid COVID-19 Antigen Self-test | 28.6 | 75 000 | 75 000 | 75 000 | 75 000 | 7500 | 75 000 |
| 5 | HEALGEN COVID-19 Rapid Antigen self-test | 28.6 | 75 000 | 75 000 | 75 000 | 75 000 | 7500 | 75 000 |
| 6 | Inspecter KOWA SARS-CoV-2 | 13.3 | 75 000 | 7500 | 7500 | 75 000 | 7500 | 75 000 |
| 7 | V Trust SARS-CoV-2 Ag | 17.8 | 7500 | 75 000 | 75 000 | 75 000 | 7500 | 75 000 |
- a Variants were examined twice with each rapid antigen test according to the manufacturers' instructions.
- b For all tested rapid antigen tests, 50 μL of test sample was used per test. The samples were mixed with lysis buffer (A). All or part of the lysed sample (B) was subjected to the assay. Input ratios were calculated by using the formula: volume B/(50 μL + volume A) × 100.
- c The number indicates the amount of virus (plaque-forming units) required for a positive result.
- d Boldface type indicates one positive result and one negative result (at the indicated amount of virus used) for the two independent experiments.
3.2 Sensitivity of RATs for the Omicron variants
To compare the sensitivity of the seven RATs, Omicron variants BA.5, BA.2.75, BF.7, XBB.1, and BQ.1.1, together with the delta variant (lineage B.1.617.2), were diluted to 75 000, 7500, 750, and 75 pfu per 50 μL and then examined by using each RAT according to its manufacturer's instructions (Table 3). Espline SARS-CoV-2 (Clinical use) (#1) and Espline SARS-CoV-2 N (General use) (#2) detected all tested isolates at 7500 pfu, although BF.7 at 7500 pfu was detected in only one of the two tests. Inspecter KOWA SARS-CoV-2 (#6) detected 7500 pfu of BA.5, BA.2.75, and XBB.1 and 75 000 pfu of delta, BF.7, and BQ.1.1. V Trust SARS-CoV-2 Ag (#7) detected 7500 pfu of delta and XBB.1 and 75 000 pfu of BA.5, BA.2.75, BF.7, and BQ.1.1. SARS-CoV-2 Rapid Antigen Test (#3), CLINITEST Rapid COVID-19 Antigen Self-test (#4), and HEALGEN COVID-19 Rapid Antigen self-test (#5) detected 7500 pfu of XBB.1 and 75 000 pfu of the other five isolates. These results indicate that the sensitivity of Espline SARS-CoV-2 N (General use) (#2) is relatively high compared with the other RATs for general use.
3.3 Sensitivity of RATs in the presence of human saliva
To compare the sensitivity of the seven RATs in presence of a human specimen, BQ.1.1 diluted with cell culture medium containing human saliva samples that were negative for SARS-CoV-2 was tested using each RAT twice (Table 4). In most cases, the sensitivity of the RATs tested was not affected by the human saliva, with the exception of three cases: the sensitivity of Inspecter KOWA SARS-CoV-2 (#6) in the presence of specimen B, the sensitivity of CLINITEST Rapid COVID-19 Antigen Self-test (#4) in the presence of specimen C, and the sensitivity of HEALGEN COVID-19 Rapid Antigen self-test (#5) in the presence of specimen C. In each of these cases, the sensitivity was 10 times higher than in the absence of human saliva. These results indicate that the sensitivity of the tested RATs was not reduced by human saliva.
| No. | RAT name | Specimen A | Specimen B | Specimen C |
|---|---|---|---|---|
| 1 | Espline SARS-CoV-2 (Clinical use) | 7500a | 7500 | 7500 |
| 2 | Espline SARS-CoV-2 N (General use) | 7500 | 7500 | 7500 |
| 3 | SARS-CoV-2 Rapid Antigen Test | 75 000 | 75 000 | 75 000 |
| 4 | CLINITEST Rapid COVID-19 Antigen Self-test | 75 000 | 75 000 | 7500 |
| 5 | HEALGEN COVID-19 Rapid Antigen self-test | 75 000 | 75 000 | 7500 |
| 6 | Inspecter KOWA SARS-CoV-2 | 75 000 | 7500 | 75 000 |
| 7 | V Trust SARS-CoV-2 Ag | 75 000 | 75 000 | 75 000 |
- Note: BQ.1.1 was diluted in culture medium containing human saliva (specimens A–C) and examined twice with each rapid antigen test.
- a The number indicates the amount of virus (plaque-forming units) required for a positive result.
- b Boldface type indicates one positive result and one negative result (at the indicated amount of virus used) for the two independent experiments.
4 DISCUSSION
We compared the sensitivity for the Omicron variants of six RATs approved for general use and a RAT approved for clinical use by the Japanese regulatory authority and available in Japan. All RATs for general use detected at least 75 000 pfu of all variants tested. The delta variant possessed the D63G, R203M, and D377Y substitutions in the N protein relative to the prototype virus, whereas the Omicron variants tested in this study possess the P13L, G30F, del31/33, E136D, R203K, G204R, or S413R changes. We found that these changes did not reduce the sensitivity of any of the RATs tested. Among the RATs for general use, Espline SARS-CoV-2 N (General use) (#2) showed superior sensitivity to all Omicron variants tested. The sensitivity of Espline SARS-CoV-2 N (General use) (#2) was identical to that of Espline SARS-CoV-2 (Clinical use) (#1), which is approved for clinical use. In most cases, the other five RATs were 10 times less sensitive than Espline SARS-CoV-2 N (General use) (#2). We previously reported that Espline SARS-CoV-2 (Clinical use) (#1) showed superior sensitivity against the delta variant compared with 16 other RATs for clinical use, which showed 10–100-times lower sensitivity than Espline SARS-CoV-2 (Clinical use) (#1).9, 15 The sensitivity of the five RATs (#3–7) for general use is similar to that of these RATs previously tested for clinical use. However, all RATs tested here required more than 7500 pfu of virus, suggesting that individuals whose specimens contained infectious virus at a level lower than the detection limit would be considered negative. Therefore, the use of these RATs for general use requires careful consideration. However, RATs with a high sensitivity are useful for early initiation of antiviral treatment especially in high-risk patients because test results can be obtained immediately. Such use may be considered in the future as antiviral drugs become more widely available.
Our study has several limitations. First, clinical specimens were not tested. Some biological components derived from human or indigenous microflora in clinical specimens may interfere with the detection of the N protein or cause a false positive reaction, resulting in reduced sensitivity or specificity. To mimic clinical specimens, we tested the sensitivity of RATs in the presence of human saliva samples that were negative for SARS-CoV-2 and found that human saliva did not reduce the sensitivity. However, because of the limited number of specimens used, positive or negative factors in the clinical specimens may not have been adequately evaluated. Second, RNA levels based on RT-qPCR were not used to represent the virus amounts. Since RNA levels are correlated with infectious virus titers,15 this would not affect the interpretation of the results within this paper, but might make it difficult to compare our results with those in other papers. Third, prototype isolates such as Wuhan-Hu-1 2019 were not included as a baseline.
In summary, here we demonstrated that six RATs approved for general use and available in Japan show similar sensitivity to RATs used in clinics and that the amino acid substitutions in the N protein of several Omicron variants do not decrease the sensitivity of these RATs. Other variants possessing different amino acid substitutions in the N protein may emerge in the future. Therefore, the sensitivity of RATs should be routinely assessed using prevalent variants to ensure their effectiveness for rapid and accurate diagnosis.
AUTHOR CONTRIBUTIONS
Seiya Yamayoshi and Yoshihiro Kawaoka designed the study. Yuko Sakai-Tagawa and Seiya Yamayoshi performed the experiments. Yuko Sakai-Tagawa, Seiya Yamayoshi, and Yoshihiro Kawaoka analyzed the data. Seiya Yamayoshi and Yoshihiro Kawaoka wrote the manuscript. Peter J. Halfmann, Nancy Wilson, Max Bobholz, William C. Vuyk, Wanting Wei, Hunter Ries, David H. O'Connor Thomas C. Friedrich, Emilia Mia Sordillo, Harm van Bakel, and Viviana Simon provided sequenced residual specimens. All authors reviewed and approved the manuscript.
ACKNOWLEDGMENTS
The authors would like to thank Susan Watson for editing the manuscript. This work was supported by the Japan Agency for Medical Research and Development (AMED) (JP22wm0125002 and JP223fa627001), the National Institutes of Allergy and Infectious Diseases Center for Research on Influenza Pathogenesis and Transmission (CRIPT) (75N93021C00014; HvB, VS, YK), the Centers for Disease Control and Prevention (75D30121C11060, to Drs. O'Connor and Friedrich), and a State of Wisconsin Department of Health Services project (435100-A22-ELCProjE-01, to Drs. O'Connor and Friedrich).
CONFLICTS OF INTEREST STATEMENT
Yoshihiro Kawaoka obtained funds from TAUNS Laboratories, Inc in 2019 to organize the symposium, “Influenza and Other Infections” and collaborative research funds from FUJIFILM Toyama Chemical Co. LTD, Shionogi & Co. LTD, Daiichi Sankyo Pharmaceutical, Otsuka Pharmaceutical, KM Biologics, Kyoritsu Seiyaku, Fuji Rebio; he is also a co-founder of FluGen. The Icahn School of Medicine at Mount Sinai has filed patent applications relating to SARS-CoV-2 serological assays (U.S. Provisional Application Numbers: 62/994,252, 63/018,457, 63/020,503 and 63/024,436) and NDV-based SARS-CoV-2 vaccines (U.S. Provisional Application Number: 63/251,020) which list Viviana Simon as co-inventor.
Open Research
DATA AVAILABILITY STATEMENT
All data supporting the findings of this study are available within the paper and from the corresponding author upon reasonable request. There are no restrictions to obtaining access to the primary data.
