key: cord-0907117-0nfddl7h authors: Butterfield, Tiffany R.; Bruce-Mowatt, Alrica; Phillips, Yakima Z.R.; Brown, Nicole; Francis, Keisha; Brown, Jabari; Walker, Jerome P.; McKnight, Neil; Metalor, Kelvin; Taylor, Devon; Bruce, Carl A.; McGrowder, Donovan; Wharfe, Gilian; Sandiford, Simone L.; Thompson, Tamara K.; Anzinger, Joshua J. title: Assessment of Commercial SARS-CoV-2 Antibody Assays, Jamaica date: 2021-02-18 journal: Int J Infect Dis DOI: 10.1016/j.ijid.2021.02.059 sha: 2e391a524b8dbf826413402a92987ae67c47130a doc_id: 907117 cord_uid: 0nfddl7h Background The performance of the Roche Elecsys® Anti-SARS-CoV-2, Abbott Architect SARS-CoV-2 IgM and IgG, Euroimmun SARS-CoV-2 IgA, Euroimmun SARS-CoV-2 IgG ELISA, and Trillium IgG/IgM rapid assays was evaluated in Jamaica. Methods Diagnostic sensitivities of the assays were assessed by testing serum samples from SARS-CoV-2 PCR-confirmed persons and diagnostic specificity was assessed by testing serum samples collected during 2018-2019 from healthy persons and from persons with antibodies to a wide range of viral infections. Results Serum samples collected ≥14 days after onset of symptoms, or an initial SARS-CoV-2 RT-PCR positive test for asymptomatics, showed diagnostic sensitivities ranging from 67.9-75.0% when including all possible disease severities and increased to 90.0-95.0% when examining those with moderate to critical disease. Grouping moderate to critical disease showed a significant association with a SARS-CoV-2 antibody positive result for all assays. Diagnostic specificity ranged from 96.7-100.0%. For all assays examined, SARS-CoV-2 real-time PCR cycle threshold (Ct) values of the initial nasopharyngeal swab sample testing positive were significantly different for samples testing antibody positive versus negative. Conclusions These data from a predominantly African descent Caribbean population shows comparable diagnostic sensitivities and specificities for all testing platforms assessed and limited utility of these tests for persons with asymptomatic and mild infections. The SARS-CoV-2 pandemic has resulted in an unprecedented need for reliable commercial laboratory diagnostics. While SARS-CoV-2 antibody assays have recently become commercially available, performance data have mainly assessed high-income country populations (Van Walle et al., 2020) , with data from populations of mostly African descent lacking. To our knowledge, there has been no published performance assessment of SARS-CoV-2 antibody assays with a predominantly black population. In this study in Jamaica, serum samples were used to assess the diagnostic sensitivity and specificity of the Roche Elecsys ® Anti-SARS-CoV-2, Abbott Architect SARS-CoV-2 IgM and IgG, Euroimmun SARS-CoV-2 IgA and IgG, and Trillium IgG/IgM assays. For diagnostic sensitivity analysis, 42 blood samples collected in tubes without coagulant were obtained from 37 consenting persons (5 persons were sampled at two different time points) testing SARS-CoV-2 real-time PCR positive at the Jamaica National Influenza Centre using the Corman et al method (Corman et al., 2020) . Disease severity was classified according to WHO criteria. Samples were collected 6-103 days after disease onset for symptomatic persons and 20-69 days after a positive real-time PCR test for asymptomatic persons. An additional 17 consenting SARS-CoV-2 real-time PCR positive persons were recruited to compare whole blood and serum for Trillium IgG/IgM rapid test kits. Paired whole blood and serum samples were Table) . Not all residual sera used for specificity analysis was tested across all platforms as sufficient volume was not present for some tests and a limited number of Trillium IgG/IgM test kits were provided for validation. Detection of antibodies was conducted with an Architect i2000SR for the Architect SARS-CoV-2 IgM and IgG assay, a cobas ® 6000 analyzer for the Elecsys ® Anti-SARS-CoV-2 assay, a Thermo Scientific Multiskan FC Microplate Photometer for the Euroimmun SARS-CoV-2 IgA and IgG ELISAs and lateral flow assay rapid test for the Trillium IgG/IgM assay. Each manufacturer's instructions were used for cutoff index values and interpretation of rapid test results. For Euroimmun assays, borderline index values were considered negative to allow for equivocal results to be included for sensitivity and specificity analysis, and to be consistent with other assays that do not include a borderline (or equivocal) interpretation. Previous studies (Buss et al., 2020; Eyre et al., 2020; Péré et al., 2020) of the Architect SARS-CoV-2 IgG assay indicate that samples with a high negative index value result are likely true positive but in this study were considered negative in keeping with the manufacturer's instructions and consistent with Euroimmun borderline results being considered as negative. This study was approved by the UWI Mona Campus Research Ethics Committee (ECP 244 19/20). Table) . Diagnostic sensitivities for samples collected 6-9 days, 10-13 days and ≥14 days after symptom onset were 42.9-71.4%, 85.7-100.0% and 90.0-95.0%, respectively (Fig. 1 ). Samples from asymptomatic and mildly affected persons were only available ≥14 days after onset of symptoms or after an initial SARS-CoV-2 PCR test. Sensitivities ranged from 67.9-75.0% when all disease severities were included (Fig. 1) . Grouping moderate, severe and critical disease showed a significant association with testing antibody positive for all assays: Elecsys ® Anti-SARS-CoV-2 (χ 2 =13.14, p<0.001), Architect SARS-CoV-2 IgM (χ 2 =13.81, p=0.003), Architect SARS-CoV-2 IgG (χ 2 =11.00, p=0.001), Euroimmun SARS-CoV-2 IgA (χ 2 =16.92, p<0.001) and IgG (χ 2 =14.30, p=0.001), and Trillium IgM (χ 2 =6.61, p=0.010) and IgG (χ 2 =11.70, p=0.001). Detection of antibodies was highly congruent between assays (Fig. 2) . Additional participants were recruited to compare whole blood (point of care) and serum (laboratory) for the Trillium IgG/IgM rapid lateral flow assay. Results for Trillium IgG were identical for whole blood and serum; however, Trillium IgM results showed discrepancies when comparing whole blood and serum (Supplemental Figure) . SARS-CoV-2 real-time PCR cycle threshold (Ct) values were compared with the presence of antibodies from persons with samples collected ≥14 days after onset of symptoms or an initial SARS-CoV-2 PCR positive test. For the Elecsys ® Anti-SARS-CoV-2, Architect SARS-CoV-2 IgG, Euroimmun SARS-CoV-2 IgG, and Trillium IgG assays, the Ct value was 23.5 ± 5.7 (mean J o u r n a l P r e -p r o o f ± SD) and 34.6 ± 1.0 for samples testing antibody positive and negative, respectively (p<0.0001). Ct values for SARS-CoV-2 IgM were 23.0 ± 6.2 for samples testing antibody positive and 33.5 ± 2.8 for samples testing antibody negative (p=0.0008), for Euroimmun IgA were 24.0 ± 5.7 for samples testing antibody positive and 31.8 ± 6.8 for samples testing antibody negative (p<0.0001), and for Trillium IgM, 23.0 ± 5.8 for samples testing antibody positive and 33.5 ± 2.8 for samples testing antibody negative (p=0.0003). Our data examining three chemiluminescent assays, two ELISA assays and one rapid test show that the diagnostic sensitivity of these assays for SARS-CoV-2 antibodies is comparable. The similar diagnostic sensitivity and specificity of the Trillium IgM/IgG rapid diagnostic test with chemiluminescent and ELISA assays makes this test suitable for resource-limited laboratories lacking high cost instruments. However, discrepant IgM results between whole blood and serum for the Trillium IgM/IgG rapid diagnostic test warrant additional investigation should results of the IgM component of the test be considered. An accumulating body of evidence indicates that after a SAR-CoV-2 infection antibodies become detectable approximately one week after disease onset (Deeks et al., 2020) . In agreement with these studies, approximately half of the SARS-CoV-2 infected persons in our study had detectable antibodies 6-9 days after onset of symptoms, with most having antibodies ≥10 days after symptom onset. When asymptomatic and mild groups were included in our analysis, sensitivities decreased for all assays, consistent with previous studies [3, 4] . Comparing SARS-CoV-2 antibody results revealed a striking difference in Ct values between persons testing antibody negative or positive. These data are consistent with a recently published study examining SARS-CoV-2-infected asymptomatic contacts and outpatients showing that SARS-CoV-2 PCR Ct values are inversely related to SARS-CoV-2 IgG index values (Wellinghausen et al., 2020) . High SARS-CoV-2 viral loads may cause a more robust production of SARS-CoV-2 antibodies (Zhang et al., 2020) . Our data assessing a Caribbean population of predominantly African descent highlights the limited diagnostic sensitivity of the assays examined for persons with asymptomatic and mild SARS-CoV-2 infections and has important implications for future seroprevalence studies in which a sizable proportion of the SARS-CoV-2-infected population may have experienced no symptoms or mild disease. Disease severity is color coded as follows: green = asymptomatic, blue = mild, orange = moderate, yellow = severe, and red = critical. Conceptualization, Investigation, Writing -Original Draft, Visualization. Alrica Bruce-Mowatt: Conceptualization, Investigation, Writing -Original Draft. Yakima Z. R. Phillips: Investigation. Nicole Brown: Investigation. Keisha Francis: Investigation. Jabari Brown: Investigation. Jerome P. Walker: Conceptualization & Investigation. Neil McKnight: Conceptualization & Investigation Conceptualization & Investigation. Devon Taylor: Conceptualization, Writing -Reviewing & Editing. Carl A. Bruce: Conceptualization, Writing -Reviewing & Editing. Donovan McGrowder: Conceptualization, Writing -Reviewing & Editing Conceptualization, Writing -Reviewing & Editing. Tamara K. Thompson: Conceptualization, Writing -Reviewing & Editing. Joshua J. Anzinger: Conceptualization, Investigation, Writing -Original Draft, Visualization Three-quarters attack rate of SARS-CoV-2 in the Brazilian Amazon during a largely unmitigated epidemic Detection of 2019 novel coronavirus (2019-nCoV) by real-time RT-PCR Antibody tests for identification of current and past infection with SARS-CoV-2 Stringent thresholds for SARS-CoV-2 IgG assays result in under-detection of cases reporting loss of taste/smell Clinical and immunological assessment of asymptomatic SARS-CoV-2 infections Sequential SARS-CoV-2 IgG assays as confirmatory strategy to confirm equivocal results: Hospitalwide antibody screening in 3,569 staff health care workers in Paris Meta-analysis of the clinical performance of commercial SARS-CoV-2 nucleic acid, antigen and antibody tests up to 22 SARS-CoV-2-IgG response is different in COVID-19 outpatients and asymptomatic contact persons Patients with Different Disease Severities in Acute and Convalescent Phases: A 6-Month Follow-Up Study All authors declared no conflict of interest. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. All authors declared no conflict of interest.