Improving the diagnostic performance for prior COVID-19 with T-SPOT, an interferon-gamma release assay

The precise diagnosis of a prior COVID-19 infection remains challenging. This study aimed to evaluate the efficacy of T-SPOT assays for diagnosing prior SARS-CoV-2 infections by using frozen peripheral blood mononuclear cells (PBMCs) combined with antibody tests.

The study included 122 participants with PCR-confirmed COVID-19 (the positive control cohort) and 67 participants with no evidence of prior infection (the negative control cohort). Antibody testing was conducted using iFlash-SARS-CoV-2 IgG (YHLO, iF_N) and MAGPIX^® assays (Luminex, Lumi_N), which target the nucleocapsid protein. T-SPOT^® Discovery SARS-CoV-2 assays (Oxford Immunotec) were used to detect cell-mediated immune responses against nucleocapsid (Tspot_N) and membrane (Tspot_M) proteins.

Antibody tests had similar sensitivities (if_N: 67.2% and Lumi_N: 64.8%) and specificities (>98.4%). The Tspot_N assay demonstrated comparable performance to the antibody tests, with a sensitivity, specificity, and area under the receiver operating characteristic curve (AUC) of 62.5% (95% confidence interval: 52.0%-72.2%), 98.4% (95% CI: 91.2%-100.0%), and 0.923, respectively. The Tspot_M assay had lower sensitivity (15.3%). The combination of the Tspot_N test and the Lumi_N antibody test significantly improved the sensitivity and AUC to 88.0% and 0.979, respectively (p = 0.012). Net reclassification improvement and integrated discrimination improvement analyses further supported the improved diagnostic performance of the combination assay.

Frozen PBMCs were useful for performing T-SPOT assays. The combination of T-SPOT assays targeting nucleocapsid protein and antibody tests improved the diagnosis of past SARS-CoV-2 infections in vaccinated participants. These findings suggest that integrating cellular and humoral immunity assays can facilitate COVID-19 prevalence studies.