PERFORMANCE of SARS-CoV-2 ANTIBODY ASSAYS in A NATIONAL SEROSURVEY in SOUTH AFRICA

ABSTRACT

Background: Accurate and reliable serological assays are essential for epidemiological surveillance of SARS-CoV-2. Several commercial anti-SARS assays are available and use cases for serological testing includes surveillance. However, there is growing evidence of varying performance of SARS-CoV-2 assays dependent of their format. We compare the performance of 3 different assays used in a national serosurvey undertaken between April and June 2021, in South Africa before widescale vaccination roll out. Methods: Venous blood samples from participants ≥12 years were transported under cold chain to a central testing laboratory within 24 hours of collection. Samples were tested for SARS CoV-2 antibodies with the Abbott nucleocapsid (NC)-based Architect anti-SARS CoV-2 chemiluminescent microparticle immunoassay (CMIA), the EuroImmun Spike (S)-based assay and the Roche total IgG NC-based Elecsys Anti-SARS-CoV-2 electrochemiluminescence immunoassay (ECLIA) on the Cobas e411 platform. We compared antibody detection proportions. Results: 8146 participants (median age 40 years, IQR 26-55) 5.6% of whom reported ≥1 SARS-CoV-2 symptom in the preceding 3 months gave a blood sample. Samples were tested on the Abbott assay with different cut-offs-15.5% tested positive at the 1.40 cut-off and 26.8% at the 0.49 lower cut-off. 21.6% of the samples tested positive on the Euroimmun and 39.0% tested positive on the Roche assay (Table). 286 samples were from respondents self-reporting a prior positive PCR test, and among them 149(52.1%), 156(54.6%), and 206(72.3%) were positive on the Abbott (1.40 cut-off), Euroimmun and Roche assays respectively. 116/286(40.6%) of these were positive on all three assays and with 21(7.3%) positive on Roche only. 224/286(78.3%) of those reporting prior PCR test positivity were positive at the lower Abbott cut-off, with 47(16.4%) positive on Abbott only. Conclusion: These samples collected before wide scale vaccination roll out in South Africa show variable performance of these assays with the Roche NC assay detecting more infections that both the Abbott NC assay(0.40 cut-off) and the Euroimmun S assay.This could be reflective of seroreversion previously reported with Abbott and Euroimmun, and the greater sensitivity of Roche assay targeting the more abundant NC as an epitope. Use of direct, double Antigen-sandwich-based assays that are stable and have increased sensitivity over time may be optimal to detect both natural and vaccine-induced immunity in serosurveys.