Performance of the Roche/Snibe electrochemiluminescent anti-SARS-COV-2 spike assays compared to the Roche/Abbott IgG nucleocapsid and Abbott IgM spike assays

Introduction: We evaluated the Roche Elecsys Anti-SARS-CoV-2 and Snibe SARS-CoV-2 S-RBD IgG spike chemiluminescent immunoassays and compared them to existing Roche/Abbott nucleocapsid and Abbott IgM spike assays. Methods: We enrolled 184 SARS-CoV-2 RT-PCR positive samples and 215 controls (172 pre-pandemic, and 43 cross-reactivity) to evaluate the Roche spike antibody (anti-SARS-CoV-2-S) assay. For the Snibe evaluation, we included 119 RT-PCR positive samples and 249 controls (200 pre-pandemice, 49 cross-reactivity). 98 cases had been tested on three spike assays (Roche total antibody, Snibe IgG and Abbott IgM). Results: The Roche anti-SARS-CoV-2-S assay had a CV of 0.5% (0.82U/mL) and 2.3% (8.72U/mL) and was linear from 1.16 to 240U/mL. The Snibe assay was linear from 6.43 to 77.7AU/mL, CV of 5.5% (0.43AU/mL) and 8.8% (0.18AU/mL). The Snibe spike assay was significantly more sensitive than the Abbott IgG assay at 0-6 days POS (35.2% vs 3.6%, mean difference 29.6%, 95% CI 17.5 to 41.8, p < 0.0001). Optimized LORs significantly improved the sensitivity of the Roche spike (48.1%-56.7%) and both nucleocapsid assays (Roche 43.3%-65.5%, Abbott 3.6%-18.5%) in early disease. Conclusion: Although both spike assays showed higher sensitivity than their nucleocapsid counterparts, lower, optimized LORs provided the most significant improvements to sensitivity.