Longitudinal Comparison of Automated SARS-CoV-2 Serology Assays in Assessing Virus Neutralization Capacity in COVID-19 Convalescent Sera

Abstract

Context.— Serologic tests on automated immunology analyzers are increasingly used to monitor acquired immunity against SARS-CoV-2. The heterogeneity of assays raises concerns about their diagnostic performance and comparability. Objective.— To test sera from formerly infected individuals for SARS-CoV-2 antibodies by using 6 automated serology assays and a pseudoneutralization test (PNT). Design.— Six SARS-CoV-2 serology assays were used to assess 954 samples collected during a 12-month period from 315 COVID-19 convalescents. The tests determined either antibodies against the viral nucleocapsid (anti-NC) or spike protein (anti-S). Two assays did not distinguish between antibody classes, whereas the others selectively measured immunoglobulin G (IgG) antibodies. PNT was used to detect the presence of neutralizing antibodies. Results.— Comparison of qualitative results showed only slight to moderate concordance between the assays (Cohen κ < 0.57). Significant correlations (P < .001) were observed between the antibody titers from all quantitative assays. However, titer changes were not detected equally. A total anti-S assay measured an increase in 128 of 172 cases (74%) of a suitable subset, whereas all IgG anti-S tests reported decreases in at least 118 (69%). Regarding the PNT results, diagnostic sensitivities of 89% or greater were achieved with positive predictive values of at least 93%. In contrast, specificity changed substantially over time, varying from 20% to 100%. Conclusions.— Comparability of serologic SARS-CoV-2 antibody tests is rather poor. Owing to different diagnostic specificities, the tested assays were not equally capable of capturing changes in antibody titers. However, with thoroughly validated cutoffs, IgG-selective anti-S assays are a reliable surrogate test for SARS-CoV-2 neutralizing antibodies in former COVID-19 patients.