Serologic and clinical correlates of within-host SARS-CoV-2 genetic diversity in a large acute-infection cohort

Within-host evolution of SARS-CoV-2 has been described primarily in the setting of prolonged infection among immunocompromised individuals, but the determinants of viral genetic diversity during acute infection remain incompletely understood. In particular, the relationships among prior immune exposure, vaccination, viral burden, and intra-host viral populations early in infection remain unclear. We examined whether serologic markers of prior immune exposure or vaccination were associated with within-host SARS-CoV-2 genetic diversity during acute infection. We analyzed SARS-CoV-2 whole-genome deep sequencing data from nasal swabs collected at the time of diagnosis from 4200 acutely infected individuals enrolled across a geographically diverse cohort. Intra-host viral genetic diversity was quantified as the number of intra-host single nucleotide variants (iSNVs) detected per sample. Host immune and clinical factors evaluated included receptor-binding domain (RBD)-neutralizing antibody positivity, nucleocapsid antibody positivity, Spike IgG antibody positivity, vaccination status, and immunosuppressed status. Associations between these factors and viral diversity were assessed using multivariable Poisson generalized linear models with a log link, adjustment for overdispersion, an offset for callable genome size, and covariates retained in the base model. At least two iSNVs were detected in viruses from 1977 (47%) participants, indicating measurable within-host genetic diversity during acute infection. RBD-neutralizing antibodies were detected in plasma from 1,598 (38%) participants. In multivariable analyses, RBD-neutralizing antibody positivity was independently associated with higher iSNV counts ( p < 0.001; incidence rate ratio [IRR]=1.24, 95% confidence interval [CI] 1.10–1.41). Similar associations were observed for nucleocapsid antibody positivity ( p < 0.001; IRR = 1.27, 95% CI 1.12–1.44). In contrast, vaccination status, defined as receipt of at least one dose, was associated with lower iSNV counts ( p = 0.043; IRR = 0.86, 95% CI 0.74–1.00), while immunosuppressed status was not significantly associated with iSNV counts ( p = 0.23). Additional factors associated with higher iSNV counts included lower viral loads in nasal mucosa, lower plasma N antigen levels, and longer duration of symptoms prior to diagnosis. Because samples were collected at a single timepoint, these associations do not establish whether antibody positivity contributes to within-host diversity or instead reflects infection timing or prior immune exposure. Serologic markers of prior immune exposure and any-dose vaccination status were associated with measurable differences in intra-host SARS-CoV-2 genetic diversity during acute infection in this large cohort. These findings provide a large descriptive baseline for future longitudinal studies of within-host viral dynamics. Four studies from which we obtained the samples have been registered in ClinicalTrials.gov: ITAC, ID: NCT04546581, Date: 2020–09–08; TICO, ID: NCT04501978, Date: 2020–08–03; TESICO, ID: NCT04843761, Date: 2021–04–09 and ICOS, ID: NCT04385251, Date: 2020–05–08.