ABSTRACT
Clinical and epidemiological characteristics of SARS-CoV-2 infection are now widely available, but there are few data on longitudinal serology in large cohorts, particularly from low-and middle-income countries. We established an ongoing prospective cohort of 3840 SARS-CoV-2 RT-PCR positive individuals in the Delhi-National Capital Region of India, to document clinical and immunological characteristics during illness and convalescence. The IgG responses to the receptor binding domain (RBD) and nucleocapsid were assessed at 0-7, 10-28 days and 6-10 weeks after infection. The clinical predictors of seroconversion were identified by multivariable regression analysis. The seroconversion rates in the post-infection windows of 0-7 days, 10-28 days and 6-10 weeks were 46%, 84.7% and 85.3% respectively (n=782). The proportion with a serological response increased with severity of COVID-19 disease. All participants with severe disease, 89.6% with mild to moderate infection and 77.3% of asymptomatic participants had IgG antibodies to the RBD antigen. The threshold values in the nasopharyngeal viral RNA RT-PCR in a subset of asymptomatic and symptomatic seroconverters were comparable (p value: 0.48), with similar results among non-seroconverters (p value: 0.16) (n=169). This is the first report of longitudinal humoral immune responses to SARS-CoV-2 infection over a period of ten weeks from South Asia. The low seropositivity in asymptomatic participants and differences between assays highlight the importance of contextualizing the understanding of population serosurveys. SummaryWe measured anti-SARS-CoV-2 RBD and NC protein IgG in a multi-hospital-based prospective cohort from northern India up to ten weeks post-infection. The lower seroconversion rate among asymptomatic RT-PCR positive participants has public health significance particularly for interpreting community seroprevalence estimates.
ABSTRACT
IgG immunoassays have been developed and used widely for clinical samples and serosurveys for SARS-CoV-2. We compared the performance of three immunoassays, an in-house RBD assay, and two commercial assays, the Diasorin LIAISON SARS-CoV-2 IgG CLIA which detects antibodies against S1/S2 domains of the Spike protein and the Zydus Kavach assay based on inactivated virus using a well-characterized sera-panel. 379 sera/plasma samples from RT-PCR positive individuals >20 days of illness in symptomatic or RT-PCR positivity in asymptomatic individuals and 184 pre-pandemic samples were used. The sensitivity of the assays were 84.7, 82.6 and 75.7 respectively for RBD, LIAISON and Kavach. Kavach and the in-house RBD ELISA showed a specificity of 99.5% and 100%, respectively. The RBD and LIAISON (S1/S2) assays showed high agreement (94.7%;95%CI:92.0,96.6) and were able to correctly identify more positives than Kavach. All three assays are suitable for serosurveillance studies, but in low prevalence sites, estimation of exposure may require adjustment based on our findings.
