SARS-CoV-2 Serology: Utility and Limits of Different Antigen-Based Tests through the Evaluation and the Comparison of Four Commercial Tests

Introduction: SARS-CoV-2 serology have several indications. Currently, as there are various types available, it is important to master their performance in order to choose the best test for the indication. We evaluated and compared four different commercial serology tests, three of them had the Food and Drug Administration Emergency Use Authorization (FDA-EUA). Our goal was to provide new data to help guide the interpretation and the choice of the serological tests. Methods: Four commercial tests were studied: Elecsys(®) Roche(®) on Cobas(®) (total anti-nucleocapsid (N) antibodies), VIDAS(®) Biomerieux(®) (IgM and IgG anti- receptor binding domain (RBD) antibodies), Mindray(®) (IgM and IgG anti-N and anti-RBD antibodies) and Access(®) Beckman Coulter(®) (IgG anti-RBD antibodies). Two panels were tested: a positive panel (n = 72 sera) obtained from COVID-19-confirmed patients with no vaccination history and a negative panel (n = 119) of pre-pandemic sera. The analytical performances were evaluated and the ROC curve was drawn to assess the manufacturer’s cut-off for each test. Results: A large range of variability between the tests was found. The Mindray(®)IgG and Cobas(®) tests showed the best overall sensitivity, which was equal to 79.2% CI 95% (67.9–87.8). The Cobas(®) test showed the best sensitivity after 14 days of COVID-19 molecular confirmation; which was equal to 85.4% CI 95% (72.2–93.9). The Access(®) test had a lower sensitivity, even after day 14 (55.5% CI 95% (43.4–67.3)). The best specificity was noted for the Cobas(®), VIDAS(®)IgG and Access(®) IgG tests (100% CI 95% (96.9–100)). The IgM tests, VIDAS(®)IgM and Mindray(®)IgM, showed the lowest specificity and sensitivity rates. Overall, only 43 out of 72 sera (59.7%) showed concordant results by all tests. Retained cut-offs for a significantly better sensitivity and accuracy, without significant change in the specificity, were: 0.87 for Vidas(®)IgM (p = 0.01) and 0.14 for Access(®) (p < 10(−4)). The combination of Cobas(®) with Vidas(®) IgM and IgG offered the best accuracy in comparison with all other tests combinations. Conclusion: Although using an FDA-EUA approved serology test, each laboratory should carry out its own evaluation. Tests variability may raise some concerns that seroprevalence studies may vary significantly based on the used serology test.