Italian 
English Antibody identification in COVID-19 pandemic: a comparison between immunochemiluminescence and immunochromatography methods
AUTORI
1UOC Laboratorio Analisi, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, 20122 Milano
2UOC Dermatologia, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, 20122 Milano
ABSTRACT
Antibody identification in COVID-19 pandemic: a comparison between immunochemiluminescence and immunochromatography methods
Introduction: in the fight against the COVID-19 pandemic, the determination of the serum antibodies against SARS-CoV-2 is highly relevant, although the reliability of the results delivered is sometimes questionable. The aim of this paper is to evaluate the performances of a rapid immunochromatography test for IgG and IgM antibodies, comparing them with an immunochemiluminescence method.
Methods: we analyzed 357 sera for the presence of IgG anti-SARS-CoV-2 spike proteins S1/S2 with an automated immunochemiluminescent test (DiaSorin®) and the presence of IgG and IgM anti-SARS-CoV-2 nucleocapside protein with an immunochromatography method (LEPU®) based on lateral flow technology.
Results: with Diasorin® method, 248 subjects resulted to be negative and 109 positives, whereas LEPU® test was positive (IgM+ and/or IgG+) in 98 subjects. The overall concordance between LEPU® and DiaSorin®, was 94.1% (95% CI 91.0-96.2). Cohen’s kappa was 0.86 (95% CI 0.80-0.92), indicating good agreement. 21 out of 357 (5.9%) samples had a discordant result and were re-analyzed with a third method (Roche Diagnostics Electrochemiluminescence®): 4 out of 5 DiaSorin® negative/LEPU® positive samples were confirmed as negative by Roche®; conversely, among the 16 DiaSorin® positive/LEPU® negative samples, 5 were confirmed as positive by Roche®, 6 as negative and 5 were not retested due to insufficient sample volume.
Conclusions: despite the methods were designed to detect different antibodies an overall high agreement between techniques was found. Discrepant results were found and were likely due to different antigen targets recognized by methods. The observation that only 6 out of 11 DiaSorin® positive samples were not confirmed by ROCHE®, supports the antigen-dependent hypothesis.