Correlates of protection and viral load trajectories in omicron breakthrough infections in triple vaccinated healthcare workers

Abstract

Vaccination offers protection against severe COVID-19 caused by SARS-CoV-2 omicron but is less effective against infection. Characteristics such as serum antibody titer correlation to protection, viral abundance and clearance of omicron infection in vaccinated individuals are scarce. We present a 4-week twice-weekly SARS-CoV-2 qPCR screening in 368 triple vaccinated healthcare workers. Spike-specific IgG levels, neutralization titers and mucosal spike-specific IgA-levels were determined at study start and qPCR-positive participants were sampled repeatedly for two weeks. 81 (cumulative incidence 22%) BA.1, BA.1.1 and BA.2 infections were detected. High serum antibody titers are shown to be protective against infection (p < 0.01), linked to reduced viral load (p < 0.01) and time to viral clearance (p < 0.05). Pre-omicron SARS-CoV-2 infection is independently associated to increased protection against omicron, largely mediated by mucosal spike specific IgA responses (nested models lr test p = 0.02 and 0.008). Only 10% of infected participants remain asymptomatic through the course of their infection. We demonstrate that high levels of vaccine-induced spike-specific WT antibodies are linked to increased protection against infection and to reduced viral load if infected, and suggest that the additional protection offered by pre-omicron SARS-CoV-2 infection largely is mediated by mucosal spike-specific IgA.

Fig. 1. Overview of study cohort.

Nasal/oropharyngeal/saliva swabs for PCR, mucosal antibody analysis and serum were collected at enrolment (baseline). Participants who tested qPCR positive at enrolment (n = 21) were excluded from all analyses except for estimation of cumulative incidence. Self-administered nasal/oropharyngeal/saliva swabs for qPCR were collected twice weekly for 4 weeks or until positive qPCR test from participants who tested qPCR negative at enrolment (n = 347). Participants who subsequently tested qPCR positive during the screening program (n = 60) were enrolled in an extended program comprising self-administered nasal/oropharyngeal/saliva swabs for qPCR every other day for 15 days.

Fig. 2. SARS-CoV-2 omicron viral load over the first 15 days of infection, symptom relation to Ct values and effects on nadir Ct and duration of qPCR positivity.

a Ct value during the first 15 days of breakthrough infection in all qPCR positive participants (n = 60), and (b) in participants with an asymptomatic course of infection (n = 6). c Ct values in participants who were symptomatic (dark grey) or asymptomatic (light grey) at time of sampling. d Increase in nadir Ct by asymptomatic course of infection, per two-fold increase in WT spike-specific IgG levels and prior SARS-CoV-2 infection and (e) change in days until viral clearance (qPCR neg test) by asymptomatic course of infection, per two-fold increase in WT spike-specific IgG levels and prior SARS-CoV-2 infection. Estimates in d and e are derived from linear regression models and error bars depict 95% confidence interval. Ct Cyclic threshold, qPCR qualitative polymerase chain reaction, pos positive, asympt asymptomatic, S WT serum spike-specific wild type, inf infection. Source data are provided as a Source Data file.

Fig. 3. Effect of antibody levels and prior infection status on risk of infection, and pre-infection antibody titers in participants who subsequently tested qPCR positive or not.

a WT spike-specific serum-IgG at baseline in participants with (n = 144) and without (n = 203) prior SARS-CoV-2 infection. b WT spike-specific serum IgG at baseline in participants that remained qPCR negative (n = 287) and participants that tested qPCR positive (n = 60) during the screening period. c Cumulative incidence over the study period in participants with spike-specific serum IgG levels above/at (n = 87) or below (n = 260) 75^th percentile. Shaded areas depict Standard Error. d Adjusted relative risk of infection among participants above (n = 87) vs below (n = 260) 75^th percentile, per two-fold increase in WT spike-specific serum-IgG level and among those with and without prior infection. Estimates derived from a Poisson regression model and error bars depict 95% confidence interval. e Microneutralization titers against SARS-CoV-2 WT and (f) omicron BA.1 in participants that remained qPCR negative (n = 40) and participants that tested qPCR positive (n = 24) during the screening period. In (a), (b), (e) and (f), Mann-Whitney U test with a two-tailed p-value was performed without adjustment for multiple comparisons. Lines depict geometric mean titer and bars depict 95% confidence interval. S spike, WT Wild type, RR relative risk, CI confidence interval, Ct cyclic threshold, qPCR qualitative polymerase chain reaction, pos positive, neg negative, BAU binding antibody units, Neut microneutralizing titer, ns; p > 0.05. Source data are provided as a Source Data file.

Fig. 4. Post booster WT spike-specific IgG levels with comparisons between subsequent omicron sublineage breakthrough infections.

a Post booster WT spike-specific IgG titers in participants that remained qPCR negative (n = 287) and in participants who tested positive with BA.1 (n = 19), BA1.1 (n = 13) or BA.2 (n = 22) infection during the screening period. b Ct values in qPCR positive participants the first 15 days of breakthrough infection with BA.1, BA.1.1 and BA.2. c Number of symptomatic days for participants who tested positive with BA.1, BA1.1 or BA.2 infection during the screening period. In (a), lines depict geometric mean titer and bars depict 95% confidence interval. In (b) and (c) lines depict median with IQR. Mann-Whitney U test with a two-tailed p-value was performed without adjustment for multiple comparisons. Pos Positive, Neg negative, S spike, WT wild-type, Ct cyclic threshold, BAU binding antibody units, ns; p > 0.05. Source data are provided as a Source Data file.