key: cord-296043-jc74soom
authors: Butterfield, T. R.; Bruce-Mowatt, A.; Phillips, Y. Z. R.; Brown, N.; Francis, K.; Brown, J.; Taylor, D.; Bruce, C. A.; McGrowder, D.; Wharfe, G.; Sandiford, S. L.; Thompson, T. K.; Anzinger, J. J.
title: Assessment of Commercial SARS-CoV-2 Antibody Assays, Jamaica
date: 2020-09-29
journal: nan
DOI: 10.1101/2020.09.27.20202655
sha: 
doc_id: 296043
cord_uid: jc74soom

The performance of the Roche Elecsys(R) Anti-SARS-CoV-2, Abbott Architect SARS-CoV-2 IgG, Euroimmun SARS-CoV-2 IgA, Euroimmun SARS-CoV-2 IgG ELISA, and Trillium IgG/IgM rapid assays was evaluated in Jamaica, the largest country of the English-speaking Caribbean. Diagnostic sensitivities of the assays were assessed by testing serum samples from SARS-CoV-2 PCR-confirmed persons. Serum samples collected [≥]14 days after onset of symptoms, or [≥]14 days after 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, assessed by testing serum samples collected during 2018-2019 from healthy persons and from persons with antibodies to a wide range of viral infections, 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. 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.

IgG/IgM rapid assays was evaluated in Jamaica, the largest country of the English-speaking 27

Caribbean. Diagnostic sensitivities of the assays were assessed by testing serum samples from 28 SARS-CoV-2 PCR-confirmed persons. Serum samples collected ³14 days after onset of 29 symptoms, or ³14 days after an initial SARS-CoV-2 RT-PCR positive test for asymptomatics, 30

showed diagnostic sensitivities ranging from 67.9-75.0% when including all possible disease 31 severities and increased to 90.0-95.0% when examining those with moderate to critical disease. 32

Grouping moderate to critical disease showed a significant association with a SARS-CoV-2 33 antibody positive result for all assays. Diagnostic specificity, assessed by testing serum samples 34 collected during 2018-2019 from healthy persons and from persons with antibodies to a wide 35 range of viral infections, ranged from 96.7-100.0%. For all assays examined, SARS-CoV-2 real-36 time PCR cycle threshold (Ct) values of the initial nasopharyngeal swab sample testing positive 37 were significantly different for samples testing antibody positive versus negative. These data 38 from a predominantly African descent Caribbean population shows comparable diagnostic 39 sensitivities and specificities for all testing platforms assessed and limited utility of these tests for 40 persons with asymptomatic and mild infections. 41

The SARS-CoV-2 pandemic has resulted in an unprecedented need for reliable commercial 46 laboratory diagnostics. While SARS-CoV-2 antibody assays have recently become commercially 47 available, performance data have mainly assessed high-income country populations. Assessment 48 of SARS-CoV-2 antibody assay performance in populations of mostly African descent is 49 lacking. To our knowledge, there has been no published performance assessment of SARS-CoV-50 2 antibody assays with a predominantly black population. In this study in Jamaica, serum 51 samples were used to assess the diagnostic sensitivity and specificity of five commercial SARS-52

CoV-2 antibody assays: Roche Elecsys ® Anti-SARS-CoV-2, Abbott Architect SARS-CoV-2 53 preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted September 29, 2020. . https://doi.org/10. 1101 /2020 IgM, anti-HTLV-I/II, HIV Ag/Ab, HBsAg, anti-HCV, convalescent samples from patients 68 seroconverting for influenza A or B virus antibodies, convalescent samples from patients with 69 respiratory disease but without seroconversion for influenza A and B virus antibodies, healthy 70 persons requesting vaccination status, and women seeking routine antenatal care (Supplementary  71   Table) . 72

The detection of antibodies was conducted with an Architect i2000SR for the Architect SARS-74

CoV-2 IgG assay, a cobas ® 6000 analyzer for the Elecsys ® Anti-SARS-CoV-2 assay, a Thermo 75 were reported using proportions and continuous variables reported as mean with standard 88 deviation. Comparison of means was by Welch's t-test while associations between categorical 89 variables were assessed using the χ 2 test. 90 All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Table) . 101 102 Diagnostic sensitivities of the assays ranged from 42.9-71.4% for samples collected 6-9 days 103 after onset of symptoms, 85.7-100.0% for samples collected 10-13 days after onset of symptoms, 104 and 90.0-95.0% for samples collected ≥14 days after onset of symptoms (Fig. 1) . As samples 105 from asymptomatic and mildly affected persons were only available for collections ≥14 days 106 after onset of symptoms or after an initial SARS-CoV-2 PCR test, we examined sensitivities for 107 all disease severities separately. For all assays, sensitivities were lower when asymptomatic and 108 mild infections were included, ranging from 67.9-75.0% (Fig. 1) . When moderate, severe and 109 critical disease was grouped, for each assay there was a significant association between this 110 group and testing antibody positive: Elecsys ® Anti-SARS-CoV-2 (χ 2 =19.03, p=0.001), Architect 111 SARS-CoV-2 IgG (χ 2 =15.72, p=0.003), Euroimmun SARS-CoV-2 IgA (χ 2 =21.11, p=0.007), 112 All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted September 29, 2020. . https://doi.org/10.1101/2020.09.27.20202655 doi: medRxiv preprint 6 Euroimmun SARS-CoV-2 IgG (χ 2 =18.77, p=0.016), and Trillium IgM (χ 2 =11.59, p=0.021) and 113

IgG (χ 2 =17.20, p=0.002). Detection of antibodies was highly congruent between assays (Fig. 2) . 114

The low diagnostic sensitivity across testing platforms for asymptomatic and mild SARS-CoV-2 116 infections led us to question whether a relationship exists between the presence of antibodies and 117 relative SARS-CoV-2 viral load at laboratory diagnosis. SARS-CoV-2 real-time PCR cycle 118 threshold (Ct) values were compared with the presence of antibodies from patients with samples 119 collected ≥14 days after onset of symptoms or an initial SARS-CoV-2 PCR. Ct values were 120 significantly different between samples testing positive and negative for all assays examined 121 (Fig. 3) . For the Elecsys ® Anti-SARS-CoV-2, Architect SARS-CoV-2 IgG, Euroimmun SARS-122

CoV-2 IgG, and Trillium IgG assays, the mean Ct value was 23.5 ± 5.7 for samples testing 123 antibody positive and 34.6 ± 1.0 for samples testing antibody negative. Mean Ct values for 124

Euroimmun IgA were 24.0 ± 5.7 for samples testing antibody positive 31.8 ± 6.8 for samples 125 testing antibody negative, and for Trillium IgM, 23.0 ± 5.8 for samples testing antibody positive 126 and 33.5 ± 2.8 for samples testing antibody negative. 127 128

Our data examining two chemiluminescent assays, two ELISA assays and one rapid test show 130 that the diagnostic sensitivity of these assays for SARS-CoV-2 antibodies is comparable. Each of 131 the tests examined in this study are approved by the FDA and are CE marked, except for the new 132 Trillium IgM/IgG rapid diagnostic test. The similar diagnostic sensitivity and specificity of the 133 Trillium IgM/IgG rapid diagnostic test with chemiluminescent and ELISA assays makes this test 134 All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted September 29, 2020. to production of high levels and quality of SARS-CoV-2 antibodies. 157 All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted September 29, 2020. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 

Negative Positive Borderline All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted September 29, 2020. CoV-2 IgA that tested borderline (open circle) and Trillium IgM that tested negative (open 209 square). Differences between groups was highly significant for all testing platforms. p=<0.0001 210 for Elecsys ® Anti-SARS-CoV-2, Architect SARS-CoV-2 IgG, Euroimmun SARS-CoV-2 IgG, 211 and Trillium IgG. p=<0.0003 for Euroimmun SARS-CoV-2 IgA and Trillium IgM. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted September 29, 2020. . https://doi.org/10. 1101 /2020 

Detection of 177 2019 novel coronavirus (2019-nCoV) by real-time RT-PCR. Eurosurveillance

All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this this version posted September 29, 2020. . https://doi.org/10. 1101 /2020