Evaluation of the new 2.0 version of the Roche SARS-CoV-2 Rapid Antigen Test

We evaluated the new 2.0 version of the Roche Diagnostics SARS-CoV-2 Rapid Antigen Test (RAT 2.0) for the detection of SARS-CoV-2. Our evaluation material comprised of nasopharyngeal samples of 140 persons positive for SARS-CoV-2 nucleic acid amplification test (NAAT) and of 100 persons negative for SARS-CoV-2 NAAT. The sensitivity limit of the RAT 2.0 was further estimated with the additional selected samples of 27 persons with high NAAT cycle threshold (Ct) value representing low viral load. For the detection of possible cross-reactions in the RAT 2.0, routine respiratory samples positive for influenza A (N = 5), respiratory syncytial virus (RSV) (N = 4), or combined RSV and human coronavirus OC43 (N = 1) were included in the study material. The overall sensitivity of the RAT 2.0 was 92.1% and specificity 100%. When evaluating the samples with NAAT Ct value ≤ 30, the sensitivity was 97.0%. All samples for cross-reactivity testing containing other viruses instead of SARS-CoV-2 remained negative in RAT 2.0. According to our findings, this RAT 2.0 offers a reliable tool for the diagnostics of acute COVID-19 in this pandemic environment.

We evaluated the new 2.0 version of the Roche Diagnostics SARS-CoV-2 Rapid Antigen Test (RAT 2.0) for the detection of SARS-CoV-2. Our evaluation material comprised of nasopharyngeal samples of 140 persons positive for SARS-CoV-2 nucleic acid amplification test (NAAT) and of 100 persons negative for SARS-CoV-2 NAAT. The sensitivity limit of the RAT 2.0 was further estimated with the additional selected samples of 27 persons with high NAAT cycle threshold (Ct) value representing low viral load. For the detection of possible cross-reactions in the RAT 2.0, routine respiratory samples positive for influenza A (N = 5), respiratory syncytial virus (RSV) (N = 4), or combined RSV and human coronavirus OC43 (N = 1) were included in the study material. The overall sensitivity of the RAT 2.0 was 92.1% and specificity 100%. When evaluating the samples with NAAT Ct value ≤ 30, the sensitivity was 97.0%. All samples for cross-reactivity testing containing other viruses instead of SARS-CoV-2 remained negative in RAT 2.0. According to our findings, this RAT 2.0 offers a reliable tool for the diagnostics of acute COVID-19 in this pandemic environment.
Roche Diagnostics is one of the leading SARS-CoV-2 rapid antigen tests (RATs) distributors worldwide. On October 2022, they launched a new 2.0 version of their RAT, i.e., SARS-CoV-2 Rapid Antigen Test 2.0 (RAT 2.0; Roche Diagnostics, Mannheim, Germany) manufactured by SD Biosensor (Suwon, Republic of Korea). The former version of this test has been included in the common list of COVID-19 RATs that are considered mutually appropriate for use in context of the situations by the European Commission's (EC) Health Security Committee (European Commission, 2022). EC has also widely purchased these RATs to the European Union member states via Emergency Support Instrument (European Commission, 2020) According to the distributor's package insert, the sensitivity and specificity of the new RAT 2.0 version were ≥99% compared to nucleic acid amplification test (NAAT) (Roche Diagnostics, 2022). The sensitivity was 100% in the samples with the NAAT Ct values ≤30. However, -to our knowledge -, no independent evaluations have not yet been published and the test is not in the EC Health Security Committee's list. Since the older test will be shortly replaced with this 2.0 version, there is urgent need for a its performance assessment.
Our evaluation material for the determination of sensitivity and specificity of the RAT 2.0 comprised of 240 nasopharyngeal samples collected from 240 unique adult persons on the 14th and 16th of December 2022. Of them, 140 persons [aged 49 years (median, range 19-95 years); 47 males] were positive for SARS-CoV-2 NAAT and 100 persons were negative. The samples had been collected to 2 ml VACU-ETTE® Virus Stabilization Tubes (VST; Greiner Bio-One GmbH, Kremsmünster, Austria), and the primary COVID-19 diagnostics were based on Cobas® SARS-CoV-2 (Roche Diagnostics International AG, Rotkreuz, Switzerland) NAAT assay detecting target sequences of E and ORF1a/b genes. The viral load of the positive samples was represented by the mean of two targets' cycle threshold (Ct) values. All positive samples were reactive for both targets. Furthermore, to estimate the sensitivity limit of the RAT 2.0 in relation to the high NAAT Ct values (Ct >30) representing low viral load, the additional selected samples of 27 persons [aged 46 years (median, range 24-97 years); 10 males] with the NAAT Ct values of 25.0-35.3 (all reactive for both targets) were analyzed. All persons studied had admitted to Fimlab Laboratories for routine SARS-CoV-2 NAAT testing.
The residual samples after SARS-CoV-2 NAAT were stored at − 4 • C and put up for the SARS-CoV-2 Rapid Antigen Test 2.0 intended for nasopharyngeal samples and for professional use only (RAT 2.0; Roche Diagnostics International AG, Rotkreuz, Switzerland) within 48 h after sampling. The volume of 350 µl VACUETTE® VST sample was added into the RAT 2.0 extraction buffer tube, and after that the test was performed according to manufacturer's instructions.
The study was based on a standard clinical validation procedure from the residual SARS-CoV-2 samples of the test intended for clinical use in the laboratory, and the approval of the ethical committee was not required.
Of the 140 SARS-CoV-2 NAAT positive evaluation samples, 129 were positive and 11 were negative with RAT 2.0 (Table 1.). All NAAT negative samples were also negative with RAT 2.0. Thus, the sensitivity of the RAT 2.0 was 92.1% (confidence interval, CI 86.4-96.0%) and specificity 100% (CI 96.4-100%). When evaluating the samples with NAAT Ct value ≤30, the sensitivity was 97.0% (CI 92.4-99.2%). The Ct values of the four RAT 2.0 false negative samples varied from 27.7 to 29.9. Of the samples with Ct value >30, only one was RAT 2.0 positive (Ct value 30.5) and seven remained false negative. All samples for crossreactivity testing containing other viruses instead of SARS-CoV-2 were negative in RAT 2.0.
When the additional selected samples with low SARS-CoV-2 NAAT mean Ct values (Ct 25.0-35.3) were analyzed separately, of those with Ct value ≤30 (N = 16) only four gave a false negative result in RAT 2.0, while of those with Ct value >30 (N = 11) all remained false negative.
We have earlier evaluated the old Roche Diagnostics SARS-CoV-2 RAT version with the same study setting, and we can observe that the RAT version 2.0 seems to be even more sensitive and specific than the old version, as shown in Table 1 (Flinck et al., 2022). The new and the old RAT versions were not compared in parallel in this study due to scanty residual samples, which is a limitation of this study. However, the Cobas® NAAT Ct levels of the SARS-CoV-2 positive evaluation samples as well as the proportion of the samples with high Ct value >30 [21.4, median (range 14.6-34.6) and 8/140 (5.7%), respectively] in this study seemed to be quite comparable to those of the adults in the former evaluation [21.7, median (range 15.2-35.6) and 5/63 (7.9%), respectively]. The present evaluation was not instructions for use (IFU) compliant, since it was not performed from the sample type recommended by the manufacturer, i.e., not directly from the nasopharyngeal swab on-site but from the VST sample, since the additional sampling was not possible in this laboratory evaluation. In the old test version, preparing a sample from certain VSTs was allowed.
The World Health Organization (WHO) has published the minimum performance criteria of ≥80% sensitivity and ≥97% specificity for the SARS-CoV-2 antigen tests (WHO, 2021). European Centre for Disease Prevention and Control (ECDC) agrees with these WHO criteria, but also supports the use of higher performance tests with ≥90% sensitivity and ≥97 specificity compared to NAAT (ECDC, 2021). According to our findings, Roche SARS-CoV-2 RAT 2.0 fulfils these quite challenging criteria of WHO and ECDC for SARS-CoV-2 antigen tests. Furthermore, most of the false negative RAT samples in our study gave NAAT Ct values above 30. According to WHO, infectiousness is associated with high viral loads resulting in NAAT Ct values below 25-30 (WHO, 2021). Thus, we can conclude that this RAT 2.0 offers at least equal or even better performance compared to the old version of the test for the diagnostics of acute COVID-19 in this pandemic environment. However, additional studies are required to confirm the test efficacy for screening of SARS-CoV-2, especially in asymptomatic individuals and in children.

Funding
This research was supported by the Tuberculosis Foundation, Tampere, Finland (number 411010, 2020). *The cycle threshold (Ct) values of the nucleic acid amplification test (NAAT) positive samples were calculated as the mean of two targets (E and ORFa/b) of Roche Cobas® SARS-CoV-2 NAAT. All NAAT positive samples were reactive for both targets. **Only the adult (age ≥15 years) cases (N = 154) were included from the data earlier published by Flinck et al. (2022). When evaluating the samples with NAAT Ct values ≤30, only the results obtained by Cobas® SARS-CoV-2 NAAT were included (N = 63, of which Ct value was ≤30 in 58); CI, 95% confidence interval.