A longitudinal comparison of spike and nucleocapsid SARS-CoV-2 antibody responses in a tertiary hospitals laboratory workers with validation of DBS specimen analysis.

There is a requirement for easily accessible, high throughput serological testing as part of the SARS-CoV-2 pandemic response. Whilst of limited diagnostic use in an acute individual setting, its use on a population level is key to informing a coherent public health response. As experience of commercial assays increases, so too does knowledge of their precision and limitations. Here we present our experience of these systems thus far. We perform a spot sero-prevalence study amongst staff in a tertiary hospitals clinical microbiology laboratory, before undertaking validation of DBS serological testing as an alternate specimen for analysis. Finally, we characterise the spike and nucleocapsid antibody response over 160 days post a positive PCR test in nine non-hospitalised staff members. Amongst a cohort of 195 staff, 17 tested positive for SARS-CoV-2 antibodies (8.7%). Self-reporting of SARS-CoV2 infection (P=<0.0001) and testing of a household contact (P = 0.027) were significant variables amongst the positive and negative sub-groups. Testing of 28 matched serum and DBS samples demonstrated 96% accuracy between the sample types. A differential rate of decline of SARS-CoV-2 antibodies against nucleocapsid or spike protein was observed. At 4 months post a positive PCR test 7/9 (78%) individuals had detectable antibodies against spike protein, but only 2/9 (22%) had detectable antibodies against nucleocapsid protein. This study reveals a broad agreement amongst commercial platforms tested and suggests the use of DBS as an alternate specimen option to enable widespread population testing for SARS-CoV-2 antibodies. These results suggest potential limitations of these platforms in estimating historical infection. By setting this temporal point of reference for this cohort of non-patient facing laboratory staff, future exposure risks and mitigation strategies can be evaluated more fully.