Evolution of antibodies against SARS-CoV-2 over seven months: experience of the Nationwide Seroprevalence ENE-COVID Study in Spain

Abstract Objectives To analyse temporal trends in SARS-CoV-2 anti-nucleocapsid IgG throughout the four rounds of the nationwide seroepidemiologic study ENE-COVID (April-November 2020), and to compare the fourth-round results of two immunoassays detecting antibodies against nucleocapsid and to S protein receptor-binding domain (RBD). Methods A chemiluminescent microparticle immunoassay (CMIA) was offered to all participants in the first three rounds (Abbott; anti-nucleocapsid IgG). In the fourth round we offered this test and a chemiluminescence immunoassay (CLIA) (Beckman; anti-RBD IgG) to i) a randomly selected sub-cohort, ii) participants who were IgG-positive in any of the three first rounds; and iii) participants who were IgG-positive in the fourth round by point-of-care immunochromatography. Results Immunoassays involving 10,153 participants (82.2% of people invited to donate samples) were performed in the fourth round. A total of 2595 participants (35.1% of participants with immunoassay results in the four rounds) were positive for anti-nucleocapsid IgG in at least one round. Anti-nucleocapsid IgG became undetectable in 43.3% of participants with positive first-round results. Pneumonia was more frequent in participants with anti-nucleocapsid IgG in all four rounds (11.2%) than those in which IgG became undetectable (2.4%). In fourth round, anti-nucleocapsid and anti-RBD IgG were detected in 5.5% and 5.4% participants of the randomly selected sub-cohort, and in 26.6% and 25.9% participants with at least one previous positive result, respectively. Agreement between techniques was 90.3% (kappa: 0.72). Conclusions The response of IgG to SARS-CoV-2 is heterogeneous and conditioned by infection severity. A substantial proportion of the SARS-CoV-2 infected population may have negative serologic results in the post-infection months.