Performance of an automated chemiluminescent immunoassay for SARS-COV-2 IgM and head-to-head comparison of Abbott and Roche COVID-19 antibody assays

CS, Lau; SP, Hoo; YL, Liang; SK, Phua; TC, Aw

INTRODUCTION: We evaluated the performance of the new Abbott SARS-CoV-2 IgM assay on the Architect immunoassay analyser and compared it to the Architect IgG/Roche Cobas total antibody assays in both SARS-CoV-2 RT-PCR positive cases and healthy controls. METHOD: 200 healthy control samples and 48 individuals with other antibody-positive disorders (18 hepatitis/18 dengue/11 ANA/1 dsDNA) served to assess for potential cross-reactivity. Anonymised residual leftover sera positive for SARS-CoV-2 on RT-PCR were recruited as cases (N ​= ​133). The sensitivity/specificity/cross-reactivity of the Architect IgM assay were assessed. Concordance between the 3 assays were also analysed. RESULTS: There was no cross-reactivity with controls and other antibody positive samples. The Architect IgM assay was 100% specific (95% CI 98.5 to 100) and sensitivity was 77.8% (95% CI 60.8 to 89.9) ≥14 days post-first positive RT-PCR (POS). Sensitivity of the combined Architect IgM and IgG results (30.8%) was significantly better than the Cobas total antibodies (15.4%) in early disease (p ​= ​0.04). While the Architect IgM assay had moderate agreement with the Cobas total antibody result (Cohen’s kappa 0.72), a combined Architect IgM and IgG result had better agreement (Cohen’s kappa 0.83). CONCLUSION: The Architect IgM assay has good specificity and no cross-reactivity with other antibody positive cases. A combined Architect IgM and IgG result has better sensitivity than the individual assays for early COVID-19. The Architect IgM assay is not comparable to the Cobas total antibody assay, but the Architect IgM and IgG combined result has good agreement with the Cobas assay.