Head-to-head validation of six immunoassays for SARS-CoV-2 in hospitalized patients
Rens Zonneveld, Suzanne Jurriaans, Tom van Gool, Jorrit J. Hofstra, T. A. M. Hekker, Pien Defoer, Patricia E. Broekhuizen-van Haaften, Ellen Wentink‐Bonnema, Lynn Boonkamp, Charlotte E. Teunissen, Annemieke C. Heijboer, Frans Martens, Godelieve de Bree, Michèle van Vugt, Robin van Houdt, Janke Schinkel, Menno D. de Jong, Michiel A. van Agtmael, Anna Geke Algera, Frank van Baarle, Diane Bax, Martijn Beudel, Harm Jan Bogaard, Marije K. Bomers, Lieuwe D. J. Bos, Michela Botta, Justin de Brabander, Matthijs C. Brouwer, Sanne de Bruin, Marianna Bugiani, Esther Bulle, Osoul Chouchane, Alex Cloherty, Paul Elbers, Lucas M. Fleuren, Suzanne E. Geerlings, Bart F. Geerts, Teunis B. H. Geijtenbeek, Armand R. J. Girbes, Abraham Goorhuis, Martin P. Grobusch, Florianne Hafkamp, Laura A. Hagens, Jörg Hamann, Vanessa Harris, Robert Hemke, Sabine Hermans, Leo Heunks, Markus W. Hollmann, Janneke Horn, Joppe W. Hovius, Rosaline de Koning, Niels van Mourik, J. Nellen, Frederique Paulus, Edgar J.G. Peters, Tom van der Poll, Benedikt Preckel, Jan M. Prins, Jorinde Raasveld, Tom D. Y. Reijnders, Michiel Schinkel, Marcus J. Schultz, Alex R. Schuurman, Kim Sigaloff, Marry R. Smit, Cornelis Stijnis, Willemke Stilma, Patrick Thoral, Anissa M. Tsonas, Marc van der Valk, Denise P. Veelo, Alexander P. J. Vlaar, Heder de Vries, W. Joost Wiersinga, Dorien Wouters, A. H. Zwinderman, Diederik van de Beek
Abstract
BACKGROUND: Detecting SARS-CoV-2 antibodies may help to diagnose COVID-19. Head-to-head validation of different types of immunoassays in well-characterized cohorts of hospitalized patients remains needed. METHODS: We validated three chemiluminescence immunoassays (CLIAs) (Liaison, Elecsys, and Abbott) and one single molecule array assay (SIMOA) (Quanterix) for automated analyzers, one rapid immunoassay RIA (AllTest), and one ELISA (Wantai) in parallel in first samples from 126 PCR confirmed COVID-19 hospitalized patients and 158 pre-COVID-19 patients. Specificity of the AllTest was also tested in 106 patients with confirmed parasitic and dengue virus infections. Specificity of the Wantai assay was not tested due to limitations in sample volumes. RESULTS: Overall sensitivity in first samples was 70.6 % for the Liaison, 71.4 % for the Elecsys, 75.4 % for the Abbott, 70.6 % for the Quanterix, 77.8 % for the AllTest, and 88.9 % for the Wantai assay, respectively. Sensitivity was between 77.4 % (Liaison) and 94.0 % (Wantai) after 10 dpso. No false positive results were observed for the Elecsys and Abbott assays. Specificity was 91.1 % for the Quanterix, 96.2 % for the Liaison, and 98.1 % for the AllTest assay, respectively. CONCLUSION: We conclude that low sensitivity of all immunoassays limits their use early after onset of illness in diagnosing COVID-19 in hospitalized patients. After 10 dpso, the Wantai ELISA has a relatively high sensitivity, followed by the point-of-care AllTest RIA that compares favorably with automated analyzer immunoassays.