Serological surveillance of SARS-CoV-2: trends and humoral response in a cohort of public health workers (preprint)

ABSTRACT

Background There is considerable debate about the rate of antibody waning after SARS-CoV-2 infection, raising questions around long-term immunity following both natural infection and vaccination. We undertook prospective serosurveillance in a large cohort of healthy adults from the start of the epidemic in England. Methods The serosurveillance cohort included office and laboratory-based staff and healthcare workers in 4 sites in England, who were tested monthly for SARS-CoV-2 spike protein and nucleoprotein IgG between 23rd March and 20th August 2020. Antibody levels from 21 days after a positive test were modelled using mixed effects regression models. Findings In total, 2247 individuals were recruited and 2014 (90%) had 3-5 monthly antibody tests. Overall, 272 (12.1%) of individuals had at least one positive/equivocal spike protein IgG result, with the highest proportion in a hospital site (22%), 14% in London and 2.1% in a rural area. Results were similar for nucleoprotein IgG. Following a positive result, 39/587 (6.6%) tested negative for nucleoprotein IgG and 52/515 (10.1%) for spike protein IgG. Nucleoprotein IgG declined by 6.4% per week (95% CI, 5.5-7.4%; half-life, 75 [95% CI, 66-89] days) and spike protein IgG by 5.8% (95% CI, 5.1-6.6%; half-life, 83 [95% CI, 73-96] days). Conclusions Over the study period SARS-CoV-2 seropositivity was 8-10% overall and up to 21% in clinical healthcare workers. In seropositive individuals, nucleoprotein and spike protein IgG antibodies declined with time after infection and 50% are predicted to fall below the positive test threshold after 6 months.