key: cord-0941574-qpi5ol6g
authors: Hagbom, M.; Carmona-Vicente, N.; Sharma, S.; Olsson, H.; Jamtberg, M.; Nilsdotter-Augustinsson, A.; Sjowall, J.; Nordgren, J.
title: Evaluation of SARS-CoV-2 rapid antigen diagnostic tests for saliva samples
date: 2021-05-16
journal: nan
DOI: 10.1101/2021.05.14.21257100
sha: 34c6a945101caa5508286ba2348cd4fea062d8a0
doc_id: 941574
cord_uid: qpi5ol6g

Background: The COVID-19 pandemic has highlighted the need for rapid, cost effective and easy-to-use diagnostic tools for SARS-CoV-2 rapid antigen detection (RAD) for use in point of care settings or as self-tests, to limit disease transmission. Using saliva samples would further greatly facilitate sample collection, diagnostic feasibility, and mass screening. Objective: We tested two rapid antigen immunochromatographic tests designed for detection of SARS-CoV-2 in saliva: Rapid Response COVID-19 Antigen Rapid Test Cassette for oral fluids (Rapid Response) and DIAGNOS COVID-19 Antigen Saliva Test (DIAGNOS). Evaluation of detection limit was performed with purified SARS-CoV-2 nucleocapsid protein and titrated live SARS-CoV-2 virus and compared to Abbott Panbio COVID-19 Ag Rapid Test (Panbio) designed for nasopharyngeal samples. Sensitivity and specificity were further evaluated on RT-qPCR positive and negative saliva samples from individuals hospitalized with COVID-19 (n=34); and asymptomatic health care personnel (n=20). Results: The limit of detection of the saliva test from DIAGNOS was comparable with the Panbio test and showed higher sensitivity than Rapid Response for both nucleocapsid protein and diluted live viruses. DIAGNOS and Rapid Response further detected seven (47%) and five (33%), respectively, of the 15 RT-qPCR positive saliva samples in individuals hospitalized with COVID-19. Of the 39 RT-qPCR negative samples, all were negative with both tests (specificity 100%; 95% c.i. 0.91-1.00). Only one of the RT-qPCR positive saliva samples (Ct 21.6) contained infectious virus as determined by cell culture and was also positive using the saliva RADs. Conclusion: The results show that the DIAGNOS test exhibit a similar limit of detection as the Panbio RAD and may be an important and easy-to-use saliva RAD complement to detect infectious individuals.

There is an urgent need for rapid and easy to use diagnostics for SARS-CoV-2 to limit disease 40 transmission during the ongoing pandemic. The diagnostic gold standard, reverse 41 transcription real-time PCR (RT-qPCR) have a high specificity and sensitivity, but have 42 limitations with regards to time, cost, and logistics (1, 2). 43

To overcome these issues several rapid antigen diagnostic tests (RADs) have rapidly been 44 developed for diagnosis of SARS-CoV-2. There are now a multitude of such RADs available (3). 45

At least a few have demonstrated good sensitivity and specificity in comparison to . Moreover, previous studies have shown that RADs can exhibit high sensitivity in 47 detecting samples containing infectious virus, indicating a high sensitivity to detect contagious 48 individuals (7, 8) . The vast majority of RADs today are designed for nasopharyngeal samples 49

(3-6). Using saliva instead of nasopharyngeal swabs has several advantages. It is a noninvasive 50 technique, easy to self-collect with little discomfort, does not require specialized health care 51 personnel and thus reduced risk for the user (9). Today, there are few RADs designed for saliva 52 available that have been thoroughly validated. 53

In this study, we have evaluated saliva RADs from two suppliers: Rapid Response™ COVID-19 54 Antigen Rapid Test Cassette for oral fluids (Rapid Response) and DIAGNOS™ COVID-19 Antigen 55 Saliva Test (DIAGNOS). Both RADs are immunochromatographic assays detecting the SARS-56

CoV-2 nucleocapsid protein in saliva without any specialized instruments. The tests were first 57 evaluated using SARS-CoV-2 nucleocapsid protein and live SARS-CoV-2 virus titrated in saliva 58 or test kit buffer and compared to Abbott Panbio™ COVID-19 Ag Rapid Test (Panbio) designed 59 for nasopharyngeal samples. Finally, the tests were evaluated on saliva samples from 60 individuals hospitalized with COVID-19 and asymptomatic health care personnel and 61 . CC-BY-ND 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 May 16, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 compared to RT-qPCR. We further correlated the sensitivity of the RADs with regards RT-qPCR 62

Ct-values and infectivity in cell culture. 63 . CC-BY-ND 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 May 16, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 

Saliva samples were obtained from COVID-19 hospitalized individuals in the COVID-19 cohort 67 of Vrinnevi hospital, Norrköping, Sweden. Saliva samples were taken by a nurse and after 68 being transported to the laboratory the same day, aliquoted and stored at -80°C until analysis. 69

In total 34 saliva samples were used in this study, collected five to 30 days post symptom 70 onset. Saliva samples from asymptomatic healthcare workers (n=20) were also analyzed. 71

The two antigen saliva tests used in the study were Rapid Response™ COVID-19 Antigen Rapid 72 To test the sensitivity and detection limit of the nucleocapsid protein (NCp), purified NCp 77 (Nordic Biosite, Sweden, Code: OOEF01087) at a concentration ranging between 500 ng to 78 5 pg, with 10-fold dilutions was tested with both kits. Each kit buffer was used for making the 79 dilution series. One drop of kit buffer was added, followed by 50 µL of sample containing the 80 protein and two more drops of buffer. Both kits were tested at the same time to make it 81 possible to directly compare the readings. After 15 minutes of incubation, two persons, 82 independently of each other, made the readings which were documented. 83 84

Saliva negative for SARS-CoV-2 were used to make dilutions of cell cultured infectious SARS-86

CoV-2. A 10-fold dilution series starting from 1:10 to 1:1.000.000, were prepared and used to 87 . CC-BY-ND 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) of each other, read all test sticks to determine positivity or negativity. Positivity was further 98 categorized into three strengths: "+++" where the intensity of the test band was stronger than 99 the control band, "++" where the test band intensity was similar to the control band and "+" 100 where the test band intensity was weaker. There was no difference in observations made by 101 the two persons. Photos were taken for documentation. 102 103

All SARS-CoV-2 PCR positive saliva samples were cultured on Vero E6 cells. Vero E6 cells were 105 cultured in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% fetal calf 106 serum (FCS) and gentamycin. Cells were seeded as monolayer in 48-well plates and at time of 107 confluency they were infected with the saliva sample in a biosafety level 3 (BSL3) laboratory, 108 essentially as described (7). Before infection the cells were washed two times with DMEM 109 supplemented with 2% FCS and gentamycin. 20 L of saliva samples were diluted (DMEM 110 supplemented with 2% FCS and gentamycin) to a total volume of 350 L in a 48 well with cells. 111 . CC-BY-ND 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 May 16, 2021. ; https://doi.org/10.1101/2021.05.14.21257100 doi: medRxiv preprint Samples were blind passaged two times on Vero E6 cells for 3 days each. All cultured saliva 112 samples were tested after passage two for presence of SARS-CoV-2 using the Panbio antigen 113 test and RT-qPCR, to investigate virus replication. 114 115 RNA extraction and in-house RT-qPCR for envelope and RdRp genes 116 Viral RNA was extracted using QIAmp Viral RNA kit (Qiagen, Hilden, Germany) according to 117 manufactures instructions with the exception that 10 times less sample volume was used, due 118 to need of saliva for both the two RADs as well as cell culture. RNA from the COVID-19 cohort 119 saliva samples (n=54), saliva containing titration of SARS-CoV-2 virus and cultured saliva 120 samples on Vero E6 cells were analyzed using a qPCR for the envelope and RdRp genes. The 121 primers and probes for the envelope gene were E_Sarbeco_P1, E_Sarbeco_F and E_Sarbeco_R 122 from (10). The primers for RdRp were RdRpF: 5'-GTC ATG TGT GGC GGT TCA CT-'3 and RdRpR 123 5'-AAA CAC TAT TAG CAT AAG CAG TTG-'3 , modified from (11), and probe RdRp_Pi 5' AGG 124 TGG AAC CTC ATC AGG AGA TGC '3 from (11). RT-qPCR was performed on CFX96 (Biorad) using 125 iTaq Universal Probes Supermix (Biorad) with following cycling conditions: reverse 126 transcription at 46°C for 30 min; followed by initial denaturation at 95°C for 3 min, followed 127 by 45 cycles of 95°C for 5 seconds and 56°C for 1 minute (RdRp primers) or 58°C for 30 seconds 128 . CC-BY-ND 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 May 16, 2021. ; https://doi.org/10.1101/2021.05.14.21257100 doi: medRxiv preprint

In this study we have evaluated two SARS-CoV-2 RADs designed for use in saliva samples, 137 enabling noninvasive, easy sampling. Firstly, we investigated the detection limit of the assays 138 comparing to Panbio, a widely used RAD for nasopharyngeal samples, using the SARS-CoV-2 139

NCp protein in a range concentration from 500 ng to 5 pg. The Rapid Response test detected 140 50 pg, and the DIAGNOS test as well as the Panbio detected 5 pg. These detection limits are 141 in the range of conventional ELISAs (12), which is a standard method of protein detection. We 142 subsequently performed two different sets of virus detection tests, using dilution series of 143 laboratory cultivated SARS-CoV-2 diluted in i) saliva samples, or in ii) test kit buffer, the latter 144 enabling a direct comparison with the Panbio RAD. The DIAGNOS and Rapid Response 145 detected the same virus amount as the Panbio using SARS-CoV-2 diluted in test kit buffer (data 146 not shown), whereas the DIAGNOS test exhibited a lower detection limit for virus diluted in 147 saliva compared to the Rapid Response test (Figure 1) . 148

We subsequently tested the saliva samples collected from patients hospitalized for COVID-19 150 using the RADs (Table 1) (Figure 1 ). In addition to saliva samples of hospitalized patients, saliva samples from 155 asymptomatic health care workers were also included (n=20), all of which were negative by 156 . CC-BY-ND 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 May 16, 2021. ; https://doi.org/10.1101/2021.05.14.21257100 doi: medRxiv preprint

In total, all RT-qPCR negative saliva samples (n=39), 20 samples from the health care workers 158 and 19 from patients hospitalized with COVID-19, were negative with both tests (specificity 159 100%; 95% c.i. 0.91-1.00). lower diagnostic accuracy in some studies, while higher in others (9, 13, (16) (17) (18) ;. with one 181 . CC-BY-ND 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 May 16, 2021. ; https://doi.org/10.1101/2021.05.14.21257100 doi: medRxiv preprint study reporting a higher concordance to nasopharyngeal samples early after symptom onset 182 (18). In this study, no major differences were observed between RT-qPCR positivity with 183 regards to days post symptom onset, with positive saliva having a median of 10 and a negative 184 saliva a median of 11 days (Table 1 ). The evaluated saliva RADs, however, were more likely to 185 yield a positive result if the saliva sample was collected early during the infection (median 6 186 days) (Table 1) . Of the RT-qPCR positive saliva samples collected within a week of symptom 187 onset (n=7), five (71%) and four (57%) were positive with DIAGNOS and Rapid Response, 188

respectively. This was also associated with viral load, with RT-qPCR positive saliva samples 189 collected within one week having a median Ct-value of 25.7 (n=7) compared to a Ct-value of 190 31.5 for RT-qPCR samples collected after one week (n=8). Nevertheless, it is important to 191 consider that onset of symptoms is based on self-reporting and may not be specifically 192 accurate to This study has other limitations. One major limitation is that no saliva samples were available 194 from onset of symptoms as i) saliva are not routinely used for SARS-CoV-2 diagnostics and ii) 195 patient recruited in the COVID-19 cohort had been ill for several days before seeking hospital 196 care. Samples were taken in a range of 5-30 days, with a median of 10.5 DPS and thus do not 197 reflect the intended purpose of the RADs, i.e., early after onset of symptoms. Moreover, as 198 we had limited amount of saliva for both the two RADs as well as for the RT-qPCR and cell 199 culture, we had to use less saliva for RNA extraction than is recommended by instructions of 200 the QIAGEN viral extraction kit, thus likely lowering the sensitivity of the RT-qPCR and 201 overestimating the Ct-values. Finally, a relatively low amount of saliva samples was analyzed, 202 thus warranting careful interpretation of the results. 203 204 . CC-BY-ND 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 May 16, 2021.

To conclude, using SARS-CoV-2 NCp protein and titrated live SARS-CoV-2, our results show 205 that the rapid saliva antigen test DIAGNOS had similar limit of detection as Panbio, a widely 206 used RAD developed for nasopharyngeal samples. DIAGNOS further exhibited a 47% 207 sensitivity on RT-qPCR positive saliva from COVID-19 hospitalized patients. Sensitivity was 208 however higher on samples collected early after symptom onset, more in line with the 209 intended use of the RADs, corresponding also to a higher viral load. The Rapid Response test 210 showed a higher limit of detection as well as lower sensitivity of RT-qPCR positive saliva 211 samples in hospitalized patients compared to DIAGNOS. The overall results suggest that the 212 DIAGNOS saliva antigen test may be a good and easy-to-use complement and possible self-213 test to be applied for SARS-CoV-2 detection. 214 215 Acknowledgements 216

We thank Annette Gustafsson for valuable help with study coordination and patient sampling. 217

We also thank Melissa Govender and Francis Hopkins for help with sample organization and 218

Lennart Svensson for critical reading of the manuscript. This study was supported with a 219 coronavirus ALF grant (30320005) CC-BY-ND 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 May 16, 2021. ; https://doi.org/10.1101/2021.05.14.21257100 doi: medRxiv preprint Table 1 . CC-BY-ND 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 explosion of new coronavirus tests that could help to end the 228 pandemic

Molecular diagnostic technologies for COVID-19: Limitations and 230 challenges

2021. 233 Comparison of seven commercial SARS-CoV-2 rapid point-of-care antigen tests: a 234 single-centre laboratory evaluation study

Evaluation of the Panbio Covid-238 19 rapid antigen detection test device for the screening of patients with Covid-19

Rapid, point-of-care antigen and molecular-based tests for diagnosis of 244 SARS-CoV-2 infection

Comparison of seven commercial 247 SARS-CoV-2 rapid Point-of-Care Antigen tests. medRxiv

SARS-CoV-2 rapid antigen test: high sensitivity to detect infectious virus

Navarro 253 D. 2021. Field evaluation of a rapid antigen test

Device) for COVID-19 diagnosis in primary healthcare centres

Diagnostic Salivary Tests for SARS-CoV-2

Detection of 2019 novel coronavirus (2019-nCoV) 262 by real-time RT-PCR

Norder 265 H. 2021. Surveillance of wastewater revealed peaks of SARS-CoV-2 preceding those of 266 hospitalized patients with COVID-19

Predicting detection limits of 268 enzyme-linked immunosorbent assay (ELISA) and bioanalytical techniques in general

Accuracy of rapid antigen detection test for nasopharyngeal swab 273 specimens and saliva samples in comparison with RT-PCR and viral culture for SARS-274

CoV-2 detection

Viral RNA load as determined by cell culture as a management tool for 277 discharge of SARS-CoV-2 patients from infectious disease wards

Lung transplantation for COVID-19-associated acute respiratory distress 283 syndrome in a PCR-positive patient

Saliva or

Nasopharyngeal Swab Specimens for Detection of SARS-CoV-2

Self-Collected Saliva by Quantitative Reverse Transcription-PCR (RT-qPCR), Direct RT-296 qPCR, Reverse Transcription-Loop-Mediated Isothermal Amplification

Antigen Test To Diagnose COVID-19

Maria 299 Cattelan A, Plebani M. 2021. Salivary SARS-CoV-2 antigen rapid detection: A 300 prospective cohort study