Performance evaluation of Roche and Abbott Panbio multiplex SARS-CoV-2 and influenza A/B rapid antigen tests

To the Editor, The clinical presentation of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the virus responsible for the ongoing coronavirus disease 2019 (COVID-19) pandemic, is similar to that of influenza illness and distinguishing the two infections based on symptom combinations is challenging.1Wu Z. Harrich D. Li Z. et al.The unique features of SARS-CoV-2 transmission: comparison with SARS-CoV, MERS-CoV and 2009 H1N1 pandemic influenza virus.Rev Med Virol. 2021; 31e2171Crossref Scopus (44) Google Scholar Currently, real-time polymerase chain reaction (RT-PCR) assays are the primary means of distinguishing these infections. In the Southern Hemisphere, including Australia, the re-opening of international borders in early 2022 led to the first influenza season in 2 years. The peak of the influenza season in Australia in 2022 was predicted to coincide with continuing widespread transmission of SARS-CoV-2 (predominantly the Omicron variant) on a background of low immunity to influenza due to lack of both vaccination and natural infection in the preceding 2 years. Respiratory sample collection for infection diagnosis can be uncomfortable and testing fatigue is a recognised consequence of the SARS-CoV-2 pandemic.2Mathers J. Poyner C. Thompson D. et al.Exploration of the uptake of asymptomatic COVID-19 lateral flow testing in Birmingham, UK: survey and qualitative research.BMJ Open. 2022; 12e056606Crossref PubMed Scopus (4) Google Scholar Therefore, simultaneous testing for multiple pathogens on a single sample helps mitigate these challenges. Rapid antigen testing (RAT) for SARS-CoV-2 has been in widespread use in Australia, including home self-testing. At the time of our study, no multiplex RATs that include simultaneous testing for influenza and SARS-CoV-2 had been approved for use in Australia by the Therapeutics Goods Administration. There is currently a paucity of peer reviewed published performance data for multiplex SARS-CoV-2/influenza RATs. A study of combo RAT QuickNavi-Flu+COVID19 Ag (Denka Co Ltd, Japan) published in June 2022 for SARS-CoV-2 detection in 1,510 nasopharyngeal samples, found the sensitivity and specificity of the antigen test were 80.9% and 99.8%, respectively.3Takeuchi Y. Akashi Y. Kiyasu Y. et al.A prospective evaluation of diagnostic performance of a combo rapid antigen test QuickNavi-Flu+COVID19 Ag.J Infect Chemother. 2022; 28: 840-843Abstract Full Text Full Text PDF PubMed Scopus (7) Google Scholar However, during the study period, influenza viruses were not detected by QuickNavi-Flu+COVID19 Ag or reference real-time RT-PCR and therefore the influenza component of this combo antigen test was not able to be evaluated. Two manufacturer sponsored trials of Panbio COVID-19/Flu A&B Rapid Panel are registered on ClinicalTrials.gov; however, results for these are currently pending.4US National Library of Medicine, National Institute of HealthClinical evaluation of the Panbio™ COVID-19/Flu A&B rapid panel.20 Jul 2022https://clinicaltrials.gov/ct2/show/NCT05163730Google Scholar Accordingly, we sought to evaluate the analytical performance characteristics of two multiplex SARS-CoV-2/influenza RATs, including performance with the locally dominant SARS-CoV-2 and influenza strains circulating in Victoria, Australia. We performed an analytical evaluation of the SARS-CoV-2/FluA and Flu B combination Rapid Antigen Test (Roche, Australia) and Panbio COVID-19/Flu A&B Rapid Test Cassette (Abbott, Australia). To evaluate the analytical sensitivity of each multiplex antigen kit we used a dilution series of representative isolates for influenza A, influenza B, and irradiated SARS-CoV-2.5Deerain J. Druce J. Tran T. et al.Assessment of the analytical sensitivity of 10 lateral flow devices against the SARS-CoV-2 Omicron variant.J Clin Microbiol. 2022; 60024799-21Crossref PubMed Scopus (45) Google Scholar Omicron BA.2 was the chosen SARS-CoV-2 variant as this was dominant in circulation in Victoria at the time of testing. Influenza A H3N2 (A/Darwin/726/2019) and H1N1 (A/Victoria/2455/2019), and influenza B (B/Victoria/28/2020) vaccine isolates were kindly provided by the World Health Organization Collaborating Centre for Reference and Research on Influenza (WHO CCRI). Each stock of SARS-CoV-2, influenza A and influenza B virus was serially diluted 10-fold in viral transport media (VTM) from 10−1 to 10−5, equivalent to a range of viral loads between 8 log10 copies/mL to 4 log10 copies/mL. Briefly, each virus dilution was mixed with kit-supplied assay buffer in equal volume. The number of drops applied to each test cassette and incubation time were followed as per the manufacturer's instructions for use. Each dilution was tested in quadruplicate and limit of detection (LOD) was determined by the last dilution where all four replicates returned a positive result. Results were interpreted by at least two independent readers, with a third reader used if any discrepant results were obtained. To assess specificity each kit was tested against a panel of quantified non-SARS-CoV-2 viruses previously prepared from VIDRL's collection. Briefly, duplicate samples were prepared for respiratory syncytial virus (RSV), human rhinovirus (HRV), adenovirus 3 (hAdV3), seasonal human coronaviruses 229E and OC43, human herpes simplex virus 1 (HSV1), cytomegalovirus (CMV) and parainfluenza virus 3. Each isolate was diluted 1/100 in VTM equivalent to a cycle threshold (Ct) value of ≤25 by an in-house respiratory RT-PCR assay.6Deerain J.M. Tran T. Batty M. et al.Assessment of twenty-two SARS-CoV-2 rapid antigen tests against SARS-CoV-2: a laboratory evaluation study.MedRxiv. 2021; 16 Dec (https://doi.org/10.1101/2021.12.15.21267691)Google Scholar To simulate performance of each kit for detecting SARS-CoV-2 and influenza coinfection, different combinations of BA.2, influenza A H3N2 and influenza B were prepared. H3N2 (A/Darwin/726/2019) was selected as this was the major influenza A strain circulating in Australia at the time of the study. Combinations were prepared at a 1:1 ratio of high copies/mL (‘strong’) and low copies/mL (‘weak’) and applied to each cassette. A weak dilution (++) was defined as the last dilution where all four replicates tested positive for a given virus. For BA.2 and A/H3N2 this equated to 7.1 and 7.5 log10 copies/mL respectively, and 8.6 log10 copies/mL for B/VIC. A strong dilution (+++) was defined as the dilution that preceded, equating to 6.1 and 6.5 log10 copies/mL for BA.2 and A/H3N2, and 8.6 for B/VIC respectively. An ‘all strong’ sample was also prepared by adding all three viruses at a 1:1:1 ratio for the strongest dilutions described above. Although these combinations may not all reflect common viral loads in clinical coinfections, combinations were tested to check for any perturbation to test performance that might arise from strongly positive samples. Samples were mixed, applied to the test cassette, and incubated as per the manufacturer's instructions. Mean Ct values representing viral load for each virus were calculated from triplicate real-time RT-PCR assays targeting the SARS-CoV-2 nucleocapsid (N) gene influenza A/H1N1 pdm09 Hemagglutinin (HA) gene, A/H3N2 Matrix protein 2 (M2) and influenza B nucleoprotein (NP) gene (Table 1). All assays were performed using the 2X Precision FAST qPCR Master Mix and amplification conditions of 95oC (2 min); ×45 95°C (5 s), 60°C (30 s). Viral RNA copies/mL were quantified by droplet digital PCR (ddPCR) using the same primer sets and performed as previously.6Deerain J.M. Tran T. Batty M. et al.Assessment of twenty-two SARS-CoV-2 rapid antigen tests against SARS-CoV-2: a laboratory evaluation study.MedRxiv. 2021; 16 Dec (https://doi.org/10.1101/2021.12.15.21267691)Google ScholarTable 1Primer and probe sequences for influenza and SARS-CoV-2 real-time PCR (RT-PCR) and droplet digital PCR (ddPCR) assaysForwardReverseProbeSARS-CoV-2 NCACATTGGCACCCGCAATCGAGGAACGAGAAGAGGCTTGFAM-ACTTCCTCAAGGAACAACATTGCCAA/H1N1 pdm09 HAGGAAAGAAATGCTGGATCTGGTAACCCTTGGGTGTCTGACAAGTT6FAM-CAGTCCACGATTGCAAT-MGBA/H3N2 HAGCAACAGGRATGMGRAATGCATTTTCTATGAAACCYGCKATTGATATGCCTCTAGTTTGTTTCTCTB/VIC NPATGGATACAAGTCCTTATCAACTCTGCTTCATTAARACGCTCGAAGAGHEX–ccaTctTctTcaTcctccact– BHQ1HA, hemagglutinin gene; N, nucleocapsid gene; NP, nucleoprotein. Open table in a new tab HA, hemagglutinin gene; N, nucleocapsid gene; NP, nucleoprotein. Overall, both multiplex kits were able to consistently detect BA.2, H1N1, H3N2 and B/VIC isolates at dilutions equivalent to viral loads between 7.1 and 7.6 log10 copies/mL (Fig. 1A), equivalent to cycle threshold (Ct) values of 17.3–19.5 on the in-house derived RT-PCR assay. BA.2, H1N1 and H3N2 were consistently detectable by both kits at 6.1–6.5 log10 copies/mL (Ct values 24.1–26.0). None of the kits detected the respective virus at Ct values of 29.5 (BA.2), 27.5 (A/H1N1), 29.4 (A/H3N2) or 27.7 (B/VIC). For specificity testing, reactivity was not observed against individual dilutions of the various non-SARS-CoV-2 or influenza viruses (Table 2), resulting in a specificity of 100% [95% confidence interval (CI) 62.8–100].Table 2Specificity panel of non-SARS-CoV-2 or influenza respiratory virusesRSVHRVhCoV229EhCoVOC43hAdV3HSV1CMVHPIV3Log10 copies/mL (Ct value)6.2 (25.0)6.7 (23.0)6.6 (25.0)7.6 (19.0)7.9 (19.0)6.6 (23.0)5.7 (24.0)7.3 (23.0)RocheNDNDNDNDNDNDNDNDPanBioNDNDNDNDNDNDNDNDCMV, cytomegalovirus; Ct, cycle threshold; hAdV3, human adenovirus 3; hCoV, human coronavirus; HPIV3, human parainfluenza virus 3; HRV, human rhinovirus; HSV1, human herpes simplex virus 1; ND, not detected; RSV, respiratory syncytial virus. Open table in a new tab CMV, cytomegalovirus; Ct, cycle threshold; hAdV3, human adenovirus 3; hCoV, human coronavirus; HPIV3, human parainfluenza virus 3; HRV, human rhinovirus; HSV1, human herpes simplex virus 1; ND, not detected; RSV, respiratory syncytial virus. Simulated coinfections, tested by combining strong and weak dilutions of each virus isolate, returned results as expected (Fig. 1B) i.e., a strong detection for isolates included in high viral load, and a faint band for isolates with lower viral load. No inhibition of detection was observed for isolates that were included in low quantity. Interestingly, however, a faint false-reactive band was observed for influenza A in the influenza B/VIC (+++)/BA.2 (++) combination, but was not observed in the reciprocal sample, nor for any other combination. Here we present the findings of the first analytical assessment of two commercial SARS-CoV-2 and influenza multiplex RATs against SARS-CoV-2 BA.2 Omicron, influenza A and influenza B. The removal of global travel restrictions in 2022 saw a significant increase in the transmission of influenza and other respiratory pathogens that had largely been absent throughout the COVID-19 pandemic. Early detection to appropriately isolate and treat symptomatic individuals presenting with influenza-like illness (ILI) is important to ongoing management of respiratory infections and protection of healthcare infrastructure. Both kits were able to detect all four viruses between 7.1 and 7.6 log10 copies/mL. BA.2 and influenza A H1N1 and H3N2 were also detected at between 6.1–6.7 log10 copies/mL. The Panbio multiplex kit was also able to detect influenza B/VIC at 6.6 log10 copies/mL. Similar to concentrations used in our study, viral loads for infected individuals typically range between 5 and 10 log10 copies/mL for SARS-CoV-2, depending on sampling site,5Deerain J. Druce J. Tran T. et al.Assessment of the analytical sensitivity of 10 lateral flow devices against the SARS-CoV-2 Omicron variant.J Clin Microbiol. 2022; 60024799-21Crossref PubMed Scopus (45) Google Scholar, 6Deerain J.M. Tran T. Batty M. et al.Assessment of twenty-two SARS-CoV-2 rapid antigen tests against SARS-CoV-2: a laboratory evaluation study.MedRxiv. 2021; 16 Dec (https://doi.org/10.1101/2021.12.15.21267691)Google Scholar, 7Pan Y. Zhang D. Yang P. et al.Viral load of SARS-CoV-2 in clinical samples.Lancet Infect Dis. 2020; 20: 411-412Abstract Full Text Full Text PDF PubMed Scopus (1056) Google Scholar and 4–8 log10 copies for influenza A and influenza B.8Hijano D.R. Brazelton de Cardenas J. Maron G. et al.Clinical correlation of influenza and respiratory syncytial virus load measured by digital PCR.PLoS One. 2019; 14e0220908Crossref PubMed Scopus (31) Google Scholar Because viral load may be a determinant of infectivity for individuals infected with influenza or SARS-CoV-2, understanding assay performance is necessary for consideration of their use in population testing. Further, as global travel recommenced other respiratory pathogens like RSV and adenovirus have re-emerged, contributing to the number of presentations with ILI. Thus, assay specificity is also an important feature of RATs, particularly when considering simultaneous detection of pathogens in a single specimen. Both kits were challenged with a panel of ‘distractor’ viruses and did not show any cross reactivity. The 2022 winter ‘flu season’ in Australia coincided with increasing SARS-CoV-2 Omicron cases that placed growing pressures on the healthcare systems. More than 200,000 influenza cases and more than 10 million SARS-CoV-2 cases were reported by 30 August 2022.9https://www1.health.gov.au/internet/main/publishing.nsf/Content/cda-surveil-ozflu-flucurr.htm#currentGoogle Scholar,10World Health Organization COVID-19 response.2022https://covid19.who.int/region/wpro/country/auGoogle Scholar In our setting, influenza H3 A/Darwin/726/2019-like is the dominant strain, as is B/Victoria A/Austria/1359417/2021-like.9https://www1.health.gov.au/internet/main/publishing.nsf/Content/cda-surveil-ozflu-flucurr.htm#currentGoogle Scholar Co-infections either with influenza A and B, or SARS-CoV-2 and influenza have also been reported and, although presentations are low compared to mono-infections, additional challenges may arise that complicate the treatment and prognosis of at-risk populations.11Alosaimi B. Naeem A. Hamed M.E. et al.Influenza co-infection associated with severity and mortality in COVID-19 patients.Virology. 2021; 18: 127Crossref Scopus (46) Google Scholar,12Dadashi M. Khaleghnejad S. Abedi Elkhichi P. et al.COVID-19 and Influenza Co-infection: a systematic review and meta-analysis.Front Med (Lausanne). 2021; 8681469Google Scholar Indeed, we showed that analytical testing of SARS-CoV-2 (BA.2 Omicron) and influenza A/H3N2 and B/VIC in different combinations can be co-detected in a single specimen and suggests a likely usefulness in a real world setting. RATs for SARS-CoV-2 have played a significant role in the early detection of SARS-CoV-2 for rapid isolation and treatment of infected individuals and are currently the primary means of community diagnosis in Australia. In contrast to standalone SARS-CoV-2 RAT, many of the current generation multiplex kits require nasopharyngeal rather than deep nasal sampling, likely due to the traditional requirement of this sampling method for influenza and incorporating nasal sampling of the multiplex RAT would likely increase community usage. Nevertheless, it is anticipated that multiplex kits could play a vital role in subsequent winter seasons for detecting influenza and SARS-CoV-2 infection including during the soon anticipated northern hemisphere winter. Both multiplex RAT devices studied detect BA.2 and influenza A and B isolates at viral loads observed in clinical infection and are able to detect combination SARS-CoV-2 and influenza simultaneously in a single specimen. Data in this study reflect analytical testing and monitoring of RAT performance in a clinical setting and are recommended to inform appropriate ongoing deployment. The authors state that there are no conflicts of interest to disclose.