Reduced sensitivity of antibody tests after omicron infection

In the past years, a huge variety of serological assays have been developed to detect antibodies binding to the nucleocapsid or the spike protein of SARS-CoV-2. The majority of these tests were developed and validated with sera from individuals infected with pre-omicron virus variants while highly mutated SARS-CoV-2 variants have since emerged. Omicron, the most-recently described variant of concern (VOC) and its subvariants contain highly-mutated spike proteins that escape immunity induced by pre-omicron variant infection or vaccination.1Rössler A Riepler L Bante D von Laer D Kimpel J SARS-CoV-2 omicron variant neutralization in serum from vaccinated and convalescent persons.N Engl J Med. 2022; 386: 698-700Crossref PubMed Scopus (249) Google Scholar, 2Wratil PR Stern M Priller A Willmann A Almanzar G Vogel E et al.Three exposures to the spike protein of SARS-CoV-2 by either infection or vaccination elicit superior neutralizing immunity to all variants of concern.Nat Med. 2022; 28: 496-503Crossref PubMed Scopus (129) Google Scholar, 3Yu J Collier AY Rowe M Mardas F Ventura JD Wan H et al.Neutralization of the SARS-CoV-2 omicron BA.1 and BA.2 variants.N Engl J Med. 2022; 386: 1579-1580Crossref PubMed Scopus (197) Google Scholar, 4Sheward DJ Kim C Ehling RA et al.Neutralisation sensitivity of the SARS-CoV-2 omicron (B.1.1.529) variant: a cross-sectional study.Lancet Infect Dis. 2022; 22: 813-820Summary Full Text Full Text PDF PubMed Scopus (36) Google Scholar Moreover, individuals who are unvaccinated and omicron infection convalescent have been shown to barely cross-neutralise pre-omicron VOCs, indicating that omicron is a new virus serotype.5Rössler A Knabl L von Laer D Kimpel J Neutralization profile after recovery from SARS-CoV-2 omicron infection.N Engl J Med. 2022; 386: 1764-1766Crossref PubMed Scopus (55) Google Scholar We therefore hypothesised that serological assays developed for detection of antibody responses against the spike protein of wild-type-like strains, might not give reliable results when testing plasma samples from individuals who are omicron infection convalescent. We obtained plasma samples from individuals who were previously SARS-CoV-2 naive after confirmed infection with ancestral SARS-CoV-2 (n=9), omicron BA.1 (n=8), or BA.2 variant (n=7; appendix p 5) and did four commercially available serological tests (appendix pp 3–4) to assess titres of binding antibodies against SARS-CoV-2 nucleocapsid by Roche (Rotkreuz, Switzerland) or receptor-binding domain (RBD) by Roche and by Abbott (Abbott Park, IL, USA), and binding IgG antibodies against subunit-1 of spike protein by Euroimmun (Lübeck, Germany). Moreover, we determined titres of neutralising antibodies against replication-competent SARS-CoV-2 virus isolates (D614G, omicron BA.1, and omicron BA.2) in a focus-forming neutralisation assay and against replication-defective vesicular stomatitis virus (VSV) pseudotyped with the spike protein of Wuhan-1 SARS-CoV-2 (VSV-WT), as previously described. All individuals developed binding antibodies to nucleocapsid and, with one exception in the omicron BA.1 group, also developed neutralising antibodies against the variant they were exposed to, indicating seroconversion (table; appendix p 6). As described previously, we found limited cross-neutralisation.1Rössler A Riepler L Bante D von Laer D Kimpel J SARS-CoV-2 omicron variant neutralization in serum from vaccinated and convalescent persons.N Engl J Med. 2022; 386: 698-700Crossref PubMed Scopus (249) Google Scholar, 5Rössler A Knabl L von Laer D Kimpel J Neutralization profile after recovery from SARS-CoV-2 omicron infection.N Engl J Med. 2022; 386: 1764-1766Crossref PubMed Scopus (55) Google Scholar Moreover, although all individuals who were ancestral virus convalescent were positive in all three spike protein-specific serological tests, only 57% of analysed individuals who were omicron BA.2 convalescent tested positive in Abbott and Euroimmun spike protein assays. The anti-spike protein RBD ELISA (Roche) detected antibodies in all eight BA.2 convalescent sera. However, in the omicron BA.1 study cohort, only five (62%) of eight sera tested positive in the Roche anti-RBD ELISA, three (38%) of eight in the Abbott anti-spike protein ELISA, and two (25%) of eight sera in the Euroimmun anti-spike protein ELISA. This finding might be explained by the greater number of mutations occurring in the spike protein of the omicron BA.1 variant than in the omicron BA.2 variant.3Yu J Collier AY Rowe M Mardas F Ventura JD Wan H et al.Neutralization of the SARS-CoV-2 omicron BA.1 and BA.2 variants.N Engl J Med. 2022; 386: 1579-1580Crossref PubMed Scopus (197) Google ScholarTableAntibody positivity in plasma samples of individuals who were ancestral virus infection convalescent and omicron BA.1 or BA.2 infection convalescent, determined by virus neutralisation assays and four commercially available serological testsVirus neutralisationAnti-nucleocapsid protein antibodies (Roche)Anti-RBD antibodies (including IgG; Roche)Anti-RBD IgG (Abbott)Anti-S1 IgG (Euroimmun)VSV-WTD614GOmicron BA.1Omicron BA.2n positive (%)95% CIn positive (%)95% CIn positive (%)95% CIn positive (%)95% CIn positive (%)95% CIn positive (%)95% CIn positive (%)95% CIn positive (%)95% CIAncestral virus (N=9 samples)9 (100%)66–1009 (100%)66–1002 (22%)5–564 (44%)19–73NA*Ancestral virus convalescent samples not analysed for nucleocapsid-specific antibodies using the Roche assay because of the low sample volume. However, all nine samples used here were in an earlier study positive positive for nucleocapsid-specific antibodies using an assay by Abbott.NA*Ancestral virus convalescent samples not analysed for nucleocapsid-specific antibodies using the Roche assay because of the low sample volume. However, all nine samples used here were in an earlier study positive positive for nucleocapsid-specific antibodies using an assay by Abbott.9 (100%)66–1009 (100%)66–1009 (100%)66–100Omicron BA.1 (N=8 samples)00–3700–377 (88%)51–1004 (50%)22–788 (100%)63–1005 (62%)30–873 (38%)13–702 (25%)6–60Omicron BA.2 (N=7 samples)1 (14%)0·5–534 (57%)25–842 (29%)8–657 (100%)60–1007 (100%)60–1007 (100%)60–1004 (57%)25–844 (57%)25–84Data are number of positive results (%) and 95% CIs. 95% CIs of positive rates were calculated by modified Wald method using a GraphPad online tool. VSV=vesicular stomatitis virus. WT=wild type. RBD=receptor-binding domain. NA=not analysed.* Ancestral virus convalescent samples not analysed for nucleocapsid-specific antibodies using the Roche assay because of the low sample volume. However, all nine samples used here were in an earlier study positive positive for nucleocapsid-specific antibodies using an assay by Abbott. Open table in a new tab Data are number of positive results (%) and 95% CIs. 95% CIs of positive rates were calculated by modified Wald method using a GraphPad online tool. VSV=vesicular stomatitis virus. WT=wild type. RBD=receptor-binding domain. NA=not analysed. In accordance with the limited cross-neutralisation of VSV-WT and SARS-CoV-2 D614G, we found a reduced sensitivity of commercially available serological assays in the omicron BA.1 or BA.2 convalescent study cohorts; however, antibodies against the nucleocapsid protein were detected in all participants.5Rössler A Knabl L von Laer D Kimpel J Neutralization profile after recovery from SARS-CoV-2 omicron infection.N Engl J Med. 2022; 386: 1764-1766Crossref PubMed Scopus (55) Google Scholar Our data indicate that serological assays using recombinant ancestral spike protein domains might underestimate the true antibody titres of individuals who are omicron infection convalescent, which at least partially escape serological detection. Past infection with omicron in individuals who were previously naive and unvaccinated might even be overlooked when only using spike-protein-specific serological assays. Because omicron subvariants have become dominant worldwide, sensitivity adjustment or establishment of omicron-specific serological assays should be considered for valid antibody analysis in the future. We declare no competing interests. AR and LK contributed equally to the study. Download .pdf (.28 MB) Help with pdf files Supplementary appendix