Performance estimation of two in-house ELISA assays for COVID-19 surveillance through the combined detection of anti-SARS-CoV-2 IgA, IgM, and IgG immunoglobulin isotypes

Abstract

The main objective of this study was to estimate the performance, under local epidemiological conditions, of two in-house ELISA assays for the combined detection of anti-SARS-CoV-2 IgA, IgM, and IgG immunoglobulins. A total of 94 serum samples were used for the assessment, where 44 corresponded to sera collected before the pandemic (free of SARS-CoV-2 antibodies), and 50 sera were collected from confirmed COVID-19 patients admitted to the main public hospital in the city of Valdivia, southern Chile. The Nucleocapsid (Np) and the receptor-binding domain (RBD) proteins were separately used as antigens (Np and RBD ELISA, respectively) to assess their diagnostic performance. A receiver operating characteristic (ROC) analysis was performed to estimate the optical density (OD) cut-off that maximized the sensitivity (Se) and specificity (Sp) of the ELISA assays. Np ELISA had a mean Se of 94% (95% CI = 83.5–98.8%) and a mean Sp of 100% (95% CI = 92.0–100%), with an OD 450 nm positive cut-off value of 0.88. On the other hand, RBD ELISA presented a mean Se of 96% (95% CI = 86.3–99.5%) and a mean Sp of 90% (95% CI = 78.3–97.5%), with an OD 450 nm positive cut off value of 0.996. Non-significant differences were observed between the Se distributions of Np and RBD ELISAs, but the latter presented a significant lower Sp than Np ELISA. In parallel, collected sera were also analyzed using a commercial lateral flow chromatographic immunoassay (LFCI), to compare the performance of the in-house ELISA assays against a commercial test. The LFCI had a mean sensitivity of 94% (95% CI = 87.4–100%) and a mean specificity of 100% (95% CI = 100–100%). When compared to Np ELISA, non-significant differences were observed on the performance distributions. Conversely, RBD ELISA had a significant lower Sp than the LFCI. Although, Np ELISA presented a similar performance to the commercial test, this was 2.5 times cheaper than the LFCI assay (labor cost not considered). Thus, the in-house Np ELISA could be a suitable alternative tool, in resource limited environments, for the surveillance of SARS-CoV-2 infection, supporting further epidemiological studies.