High prevalence of undocumented SARS-CoV-2 infections revealed by analysis of nucleocapsid-specific IgG responses in diagnosed and undiagnosed individuals

Abstract

Acute SARS-CoV-2 infections are not always diagnosed; hence an unknown proportion of all infections are not documented. SARS-CoV-2 can induce spike and nucleocapsid protein specific IgG antibodies, which can be detected in seroprevalence studies to identify a previous infection. However, with the introduction of vaccines containing the spike protein it is no longer possible to use spike-IgG as a marker of infection. In many countries marketed vaccines do not include the nucleocapsid protein, allowing the use of nucleocapsid-specific IgG (N-IgG) as a specific marker for previous infection. Importantly however, not all SARS-CoV-2-infected individuals develop detectable N-IgG responses and there are reports of waning of N-IgG titers in previously infected individuals, complicating the use of N-IgG in seroprevalence studies. Here, our aim was to investigate N-IgG as a marker for previous infection. To this end we analyzed a well characterized cohort (n = 2,583; sampled in March, 2022), including 612 participants with a previously diagnosed and documented SARS-CoV-2-infection. We show that 75% (460/612) of the confirmed SARS-CoV-2-infected participants were N-IgG positive, and that the frequency of seropositivity was stable for at least 105 weeks after the latest documented SARS-CoV-2-infection. Among participants with no documented SARS-CoV-2-infection, 32.6% (642/1971) were N-IgG-positive, suggesting a previous infection. Assuming similar frequency of N-IgG-seronegative cases in previously diagnosed and undiagnosed individuals we further estimate that 214 of the 1329 undiagnosed and N-IgG-negative cases had been previously infected, indicating a total infection rate of 56.8% (1,468/2,583), clearly higher than the documented 23.7% rate of infection, in this cohort. In conclusion, our results suggest that while N-IgG is a good marker of previous SARS-CoV-2-infection the large proportion of previously infected N-IgG-negative individuals introduces a risk for underestimations of total level of previously infected individuals in a population. Accounting for this dark number of undiagnosable cases can provide better estimates of total level of infected individuals in a population.

Fig 1. No general antibody waning over time in individuals with a previous confirmed SARS-CoV-2 infection.

Data shown represent N-IgG titers over time after the last documented infection in the 612 confirmed SARS-CoV-2 infected individuals. The N-IgG titers represent Binding Antibody Units (BAU)/ml. Cut-off for a positive response is 11.8 BAU/ml. LOESS curve; the shaded area represents 95% confidence interval.

Fig 2. Flow chart showing study design and results.

Dark number; undocumented and undiagnosable infections.