Virological and Clinical Determinants of the Magnitude of Humoral Responses to SARS-CoV-2 in Mild-Symptomatic Individuals

Abstract

Background: Evidence on the determinants of the magnitude of humoral responses and neutralizing titers in individuals with mild COVID-19 is scarce.

Methods: In this cohort study of mild COVID-19 patients, we assessed viral load (VL) by RT-qPCR at two/three time points during acute infection, and anti-SARS-CoV-2 antibodies by ELISA and plasma neutralizing responses using a pseudovirus assay at day 60.

Results: Seventy-one individuals (65% female, median 42 years old) were recruited and grouped into high viral load (VL) >7.5 Log₁₀ copies/mL (n=20), low, VL ≤7.5 Log₁₀ copies/mL (n=22), or as Non-early seroconverters with a positive PCR (n=20), and healthy individuals with a negative PCR (n=9). Individuals with high or low VL showed similar titers of total neutralizing antibodies at day 60, irrespective of maximal VL or viral dynamics. Non-early seroconverters had lower antibody titers on day 60, albeit similar neutralizing activity as the groups with high or low VL. Longer symptom duration and older age were independently associated with increased humoral responses.

Conclusions: In mild SARS-CoV-2-infected individuals, the duration of symptoms and age (but not VL) contribute to higher humoral responses.

Keywords: COVID-19; humoral response; neutralizing antibodies; seroconversion; symptoms; viral load.

Figure 1

Viral load determinations. (A) Maximal values of VL for each individual. Boxes show the median and the 25^th-75^th interquartile range and bars the 10^th-90^th interquartile range. P values correspond to Kruskal-Wallis test with Dunn’s multiple comparisons. (B) VL dynamics in each group (mean ± SEM) with the best fit curve (single exponential decay). (C) Self-reported symptom duration for each individual. Boxes show the median and the 25^th-75^th interquartile range and bars the 10^th-90^th interquartile range. P values correspond to Kruskal-Wallis test with Dunn’s multiple comparisons. (D) Seroconversion (frequency of positive samples for the indicated individual antigens or for any of them) in the different groups assessed by ELISA at day 60 of follow-up.

Figure 2

Neutralizing activity. Individual titers of (A) anti-S1+S2 antibodies, (B) anti-NP antibodies, (C) anti-RBD antibodies, and (D) neutralization. Individual ratios of neutralization titers to (E) anti-S1+S2, (F) anti-RBD, and (G) anti-NP antibodies. In all panels, boxes show the median and the 25^th-75^th interquartile range and bars the 10^th-90^th interquartile range. P values correspond to Kruskal-Wallis test with Dunn’s multiple comparisons. ID, inhibitory dilution; NP, nucleocapsid protein; VL, viral load.

Figure 3

Associations between humoral responses and clinical/demographic characteristics. (A) Heatmap of correlations between total IgG, neutralizing antibody titer, VL, symptom duration, and age for all patients. (B) Neutralization titers in female and male participants (all groups). Bar indicates median values, and the p-value is shown (Mann-Whitney test). Correlation of neutralization titers with (C) VL, (D) Duration of Symptoms, and (E) Age. Individual values are color-coded: green for Non-early seroconverter individuals at diagnosis, orange for individuals with low viral load, and red for individuals with high viral load. Lines indicate linear regression of all values for illustrative purposes. The correlation coefficient and p-value (Spearman correlation test) are shown. (F) The impact of asymptomatic infection was assessed by comparing neutralizing titers between all asymptomatic (grey dots) and all symptomatic patients (red symbols). (G) Frequency of low and high neutralizers (cutoff value 1/250); the Fisher exact test p-value is shown. (H) Correlation between symptom duration and neutralization titer in symptomatic individuals. The correlation coefficient and p-value (Spearman correlation test) is shown. ID, inhibitory dilution; NP, nucleocapsid protein; RBD, receptor-binding domain; VL, viral load.