key: cord-0815578-nh17jbue
authors: Correa, I. A.; Faffe, D. S.; Galiez, R. d. M.; Goncalves, C. C. A.; Maia, R.; da Silva, G. P.; Moreira, F. R. R.; Mariani, D.; Campos, M. F.; Carvalho, I. L.; de Souza, M. R. M.; Cunha, M. S.; Nascimento, E. R. d. S.; Ribeiro, L. d. J.; da Cruz, T. F. C.; Policarpo, C.; Gonzales, L.; Rodgers, M. A.; Berg, M.; Vijesurier, R.; Cloherty, G. A.; HAckett, J.; Ferreira Junior, O. C.; Castineiras, T. M. P. P.; TANURI, A.; Costa, L.
title: A SARS-CoV-2 negative antigen rapid diagnostic in RT-qPCR positive samples correlates with a low likelihood of infectious viruses in the nasopharynx.
date: 2022-03-20
journal: nan
DOI: 10.1101/2022.03.17.22272008
sha: ca18e281681113febab86aeca94279ac8b677654
doc_id: 815578
cord_uid: nh17jbue

SARS-CoV-2 transmission occurs even among fully vaccinated individuals; thus, prompt identification of infected patients is central to control viral circulation. Antigen rapid diagnostic tests (Ag-RDT) are highly specific, but sensitivity is variable. Discordant RT- qPCR vs Ag-RDT results are reported, raising the question of whether negative Ag-RDT in positive RT-qPCR samples could imply the absence of infectious viruses. To study the relationship between a negative Ag-RDT results with virological, molecular, and serological parameters, we selected a cross sectional and a follow-up dataset and analyzed virus culture, subgenomic RNA quantification, and sequencing to determine infectious viruses and mutations. We demonstrated that a positive SARS-CoV-2 Ag-RDT result correlates with the presence of infectious virus in nasopharyngeal samples. A decrease in sgRNA detection together with an expected increase in detectable anti-S and anti-N IgGs was verified in negative Ag-RDT / positive RT-qPCR samples. The data clearly demonstrates the less likelihood of a negative Ag-RDT sample to harbor infectious SARS-CoV-2 and consequently with a lower transmissible potential.

We selected samples from August-2020 to September-2021 in two study cohorts: a 137 cross-sectional cohort composed of patients that, at the time of diagnosis, presented an 138 antigen test that was further confirmed (antigen concordant) or not ( For viral isolation Vero E6 or Vero-hA/T were plated in six well plates 24 hours before 157 inoculation to achieve 70% of confluence overnight. Cells were infected with 250μL of 158 VTM diluted in DMEM plus 300 U/mL penicillin and 300 μ g/mL of streptomycin, with 159 no addition of FBS for 1 h for viral adsorption. After the adsorption period, 10% FBS-160 supplemented DMEM with 300 U/mL penicillin and 300 μ g/mL streptomycin was 161 . CC-BY-NC-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 March 20, 2022. ; https://doi.org/10.1101/2022.03.17.22272008 doi: medRxiv preprint added to the inoculum and the cells were incubated at 37 °C and 5% CO 2 for 72 h 162 (passage #1). The supernatants of passage #1 were collected and 250 μ L was used to 163 infect fresh cell cultures as previously described, and incubated for another 72h 164 (passage #2), and the remaining supernatant was stored at -80°C for RNA extraction 165 and RT-qPCR for SARS-CoV-2 detection. Passage #2 supernatants were also collected 166 and stored at -80°C for viral titration, RNA extraction and RT-qPCR for SARS-CoV-2 167 detection. For both cell passages, 100 µL of the collected supernatant was used to 168 perform the Panbio™ COVID-19 Ag assay. For viral titration, 10-fold dilutions of each 169 sample from passage #2 were used to infect Vero E6 cells. After 1 h of virus adsorption, 170 media was replaced by fresh DMEM with 1% FBS, 100u/mL penicillin, 100 µg/mL 171 streptomycin and 1.4% carboxymethylcellulose (Sigma Aldrich) followed by incubation 172 for 4 days at 37 °C with 5% CO 2 . Then, cells were fixed with 10% formaldehyde and 173 stained with 1% crystal violet in 20% methanol for plaque visualization and . CC-BY-NC-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) to cDNA with SuperScript IV (ThermoFisher, USA) and whole genome amplification 209 was performed with the ARTIC SARS-CoV-2 V3 primer panel and the Q5 hotstart 210 . CC-BY-NC-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) . CC-BY-NC-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. Microtitre plates (Immulon 2 HB) were coated with and trimeric spike protein at 240 4μg/mL or nucleocapsid protein at 1 ug/mL in 100 mM sodium carbonate-bicarbonate 241 buffer (pH 9.6) and incubated overnight at 4 o C. Excess protein was removed by 242 washing five times with PBS + 0,05% Tween-20 (PBST, Sigma) and unbound sites 243 blocked using 5% BSA in PBST. Samples were added at 1/50 dilution in PBST + 2% 244 BSA, followed by 1 hour at 37 o C. The plates were washed five times with PBST, and 245 polyclonal anti-human IgG antibody conjugated to HRP (Promega) added for one hour 246 at room temperature. Plates were washed five times with PBST followed by addition of 247 the chromogenic substrate TMB (Sigma) for 10 minutes and reaction stopped with 1N 248 sulphuric acid. Absorbance was read at 450 nm with an ELISA microplate reader 249 (Biochrom Asys). A positive control specimen (from a known COVID-19 patient) in 250 simplicate and a negative control (pre-epidemic plasma sample) in triplicate were added 251 to every assay plate for validation and cut-off determination. Results were expressed as 252 a reaction of sample optical density value divided by assay cutoff. 253 254

The data from the cohort was acquired using the KoBoCollect online/offline web-based 256 form (Available at: https://www.kobotoolbox.org). The data was extracted as a dataset 257 based on XLSForms and merged with the laboratory data. Differences between 258 concordant and discordant groups were assessed by Mann-Whitney non-parametric U-259 test. Multivariate model based on logistic regression was applied for multivariate 260 . CC-BY-NC-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. Table 1) , 309 suggesting a relatively higher abundance of infectious viruses present in concordant 310 . CC-BY-NC-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 March 20, 2022. ; https://doi.org/10.1101/2022.03.17.22272008 doi: medRxiv preprint samples. When the sample C t is considered as an indicator of virus isolation success, the 311 higher probability of virus isolation significantly occurred for both concordant and 312 discordant samples in the low C t range (10-20). This probability dropped to less than 313 25% with C t values greater than 26 ( Figure 4C ). When analyzing discordant and 314 concordant samples separately for virus isolation success in the same C t range, the 315 probability of virus isolation was below 10% for the former and above 50% for the 316 latter (p = 0.002). 317

The absence of viral particles in nasopharynges still positive for total viral RNA could 318 be due in major part to the presence of viral subgenomic RNA (sgRNA). To verify the 319 influence of sgRNA on the C t value, we performed in parallel target-specific RT-qPCR 320 for genomic and N-sgRNA in a few concordant and discordant samples (Supplementary 321 Figure 3A ). We observed that discordant samples presented higher C t values for both 322 genomic and N-sgRNA than concordant samples (Supplementary Figure 3B 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 March 20, 2022. ; https://doi.org/10.1101/2022.03.17.22272008 doi: medRxiv preprint following sample collection varied from 6 to 32 days. On day 0, the sample C t ranged 361 from 12-26 followed by an overall increase in the C t value for viral RNA over time 362 (Table 1 and Figure 6 ). We observed that N-sgRNA values also increased and antigen 363 positive results decreased as time passed since symptom onset (Table 1) . Therefore, a 364 prolonged SARS-CoV-2 infection increases the likelihood of a negative Ag-RDT result 365 without impacting detection by RT-qPCR. Importantly, out of 5 follow-up samples that 366 were still positive for the Antigen RDT, 3 were collected early after symptom onset (6-8 367 days We demonstrated that a prolonged infection period decreases the likelihood of viral 375 particle detection by the Panbio™ Ag-RDT which in our data set is overall correlated 376 with lower viral loads in nasopharyngeal samples as well as a lower level of viral 377 replication/infectivity. However, a few discordant samples had viral loads in the range 378 of 18-25 which predicted a substantial amount of viral replication and the presence of 379 infectious viruses. Then, we asked whether the establishment of the anti-SARS-CoV-2 380 antibody response in those individuals would contribute to a decrease in detection by 381 the Ag-RDT. The presence of serum IgG anti-S and anti-N antibodies was measured in 382 initial and followed-up samples. We observed that 92.59% and 100% of samples 383 collected 6 to 14, and >14 days after diagnosis, respectively, had anti-N IgG. 96.29% 384 and 83.33% of samples collected 6 to 14, and >14 days after diagnosis, respectively, 385 . CC-BY-NC-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 March 20, 2022. ; https://doi.org/10.1101/2022.03.17.22272008 doi: medRxiv preprint had an anti-S IgG response (Table 1) . Amongst initial diagnostic specimens, 23.33% 386 already had anti-N IgG and 43.33% had anti-S IgG responses (Table 1) Since the COVID-19 pandemic remains a public health challenge even after the rapid 395 development and increased access to available vaccines, it is important to ensure testing 396 capacity with fast and reliable methods to guide patient isolation to curb virus 397 transmission. 398

Our study cohort was composed of two datasets: a cross-sectional, where the Panbio™ 399 COVID-19 Ag test was only performed in symptomatic patients, and a follow-up 400 dataset in which at least two samples were collected from the same patient for the Ag-401 RDT with a minimum of 6 days between sample collection. For both datasets, the viral 402 isolation rate was higher for the concordant samples (56.52 -66.7%) when compared to 403 the discordant ones (10.53 -10%) even for the same C t range (14-29 versus 18-29). For 404 discordant samples, viruses could not be cultured when C t values were higher than 30, 405 confirming previous reports of low or absence of infectious viruses at this C t range 406 (29,30). We observed that the virus isolation rate is not altered by different viral variants 407 or sample collection. In different epidemiological weeks, RT-qPCR results indicated 408 higher C t s for discordant samples for both genomic and sgRNA compared to concordant 409 . CC-BY-NC-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. was also correlated to a C t <30 in our RT-qPCR test, accompanied by the absence of 481 measurable IgG against the S and N viral proteins and less than 7 days after symptom 482 onset. Our data also supports the proposal to identify patients using Ag-RDT testing in 483 the general population to substantially reduce the number of infected individuals with 484 higher risk of viral transmission while avoiding unnecessary individual isolation. 485 . CC-BY-NC-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 March 20, 2022. . CC-BY-NC-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 March 20, 2022. ; https://doi.org/10.1101/2022.03.17.22272008 doi: medRxiv preprint 1

CoV-2 infection: Interim guidance. 531 https://www.who.int/publications/i/item/antigen-detection-in-the-diagnosis-of-532

sars-cov-2infection-using-rapid-immunoassays. CC-BY-NC-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 March 20, 2022. ; https://doi.org/10.1101/2022.03.17.22272008 doi: medRxiv preprint

cohort analysis using individualised data. EBioMedicine

630 Effectiveness of the Comirnaty (BNT162b2, BioNTech/Pfizer) vaccine in 631 preventing SARS-CoV-2 infection among healthcare workers

Veneto region

Safety and Efficacy of the BNT162b2 mRNA Covid-19 Vaccine

Kalkstein 638 N, et al. Protection of BNT162b2 Vaccine Booster against Covid-19 in Israel

New COVID-19 Cases and Hospitalizations Among Adults, by 642 Vaccination Status

Outbreak 645 of SARS-CoV-2 Infections, Including COVID-19 Vaccine Breakthrough 646 Infections

MMWR Morb Mortal Wkly Rep

Detection 650 of SARS-CoV-2 variants by Abbott molecular, antigen, and serological tests. 651 medRxiv

Analytical comparison of nine SARS-CoV-2 antigen-detecting rapid diagnostic 654 tests for emerging SARS-CoV-2 variants

CoV-2 Nucleocapsid Variant that Affects Antigen Test Performance. J 658 Clin Virol

Rapid SARS-CoV-2 antigen 660 detection potentiates early diagnosis of COVID-19 disease

Figure 1. Schematic representation of the UFRJ-CTD cohort during

Mildly symptomatic individuals tested 765 for SARS-CoV-2 diagnosis at the Center for COVID-19 Diagnosis of Federal 766 University of Rio de Janeiro from