Evaluation of three fully-automated SARS-CoV-2 antibody assays

Objectives Serological assays for detection of SARS-CoV-2 antibodies are increasingly used during the COVID-19 pandemic caused by the SARS-Coronavirus-2. Here we evaluated the analytical and clinical performance of three commercially available SARS-CoV-2 antibody assays. Methods A total of 186 samples from 58 patients with PCR-confirmed COVID-19 infection were measured using SARS-CoV-2 antibody assays by Siemens Healthineers, Roche Diagnostics and Euroimmun. Additionally, 123 control samples, including samples collected before December 2019 and samples with potential cross-reactive antibodies were analyzed. Diagnostic specificity, sensitivity, agreement between assays and ROC curve-derived optimized thresholds were determined. Furthermore, intra- and inter-assay precision and the potential impact of interfering substances were investigated. Results SARS-CoV-2 antibody assays by Siemens and Roche showed 100% specificity. The Euroimmun assay had 98 and 100% specificity, when borderline results are considered as positive or negative, respectively. Diagnostic sensitivity for samples collected ≥14 days after PCR-positivity was 97.0, 89.4 and 95.5% using the Siemens, Roche and Euroimmun assay, respectively. Sensitivity of the Roche assay can be increased using an optimized cut-off index (0.095). However, a simultaneous decrease in specificity (98.4%) was observed. Siemens showed 95.8 and 95.5% overall agreement with results of Euroimmun and Roche assay, respectively. Euroimmun and Roche assay exhibited 92.6% overall agreement. Discordant results were observed in three COVID-19 patients and in one COVID-19 patient none of the investigated assays detected antibodies. Conclusions The investigated assays were highly specific and sensitive in detecting SARS-CoV-2 antibodies in samples obtained ≥14 days after PCR-confirmed infection. Discordant results need to be investigated in further studies.