Viral kinetics of sequential SARS-CoV-2 infections

Study design

Between March 11^th, 2020, and July 28^th, 2022, the NBA conducted regular surveillance for SARS-CoV-2 infection among players, staff, and affiliates as part of an occupational health program. This included frequent viral testing (often daily during high community COVID-19 prevalence) using a variety of platforms, but primarily via nucleic acid amplification tests, as well as clinical assessment including case diagnosis and symptom tracking. To assess viral concentration, RT-qPCR tests were conducted when possible, using anterior nares and oropharyngeal swabs collected by a trained professional and combined into a single viral transport media. Cycle threshold (Ct) values were obtained from the Roche cobas target 1 assay. Ct values were converted to genome equivalents per milliliter using a standard curve.¹⁴ Data on participant age and vaccination status were collected where possible. Viral lineages were assigned using whole-genome sequencing (Supplementary Methods), when feasible. This resulted in a longitudinal dataset of 424,401 SARS-CoV-2 tests with clinical COVID-19 history and demographic information for 3,021 individuals.

During the data collection period, 3,346 infections were identified among the 3,021 individuals. These infections reflected the timing, intensity, and lineage composition of SARS-CoV-2 transmission in the broader United States (Figure 1). Of these infections, we identified 1,989 “well-documented” infections that were sufficiently sampled to infer viral kinetics (Table 1), as defined by at least one RT-qPCR test with cycle threshold (Ct) value under 32 and three tests with Ct values under 40¹¹ One individual had four total infections, and we omitted their third and fourth from the analysis. Of the well-documented infections, 193 were second infections, as evidenced by a previous positive test more than 30 days prior to infection or a clinical history of prior infection (Figure 1, Supplementary Table 1).

• Download figure
• Open in new tab

Figure 1. Onset times of repeat and overall infections in the dataset.

(A): Histogram of first positive test dates for all recorded infections in full dataset (n = 3,346). Colors in both panels correspond to the SARS-CoV-2 variant category (Other/Unspecified: Black; Alpha: Blue; Delta: Red; BA.1/BA.2: Magenta), where Other/Unspecified include all non-Alpha, Delta, and Omicron lineages and any samples that could not be sequenced. (B): Date of the first positive test (points) for well-documented infections (n = 235) in individuals with an initial Alpha, Delta, or Other/Unspecified infection (n = 78 well-documented infections) and a second BA.1/BA.2 infection (n = 157 well-documented infections; see also Supplementary Table 1, group 2). Horizontal lines connect the points that correspond to multiple infections that belong to a single person. Lines that extend off the plot range to the left indicate the absence of a well-documented (though clinically confirmed) first infection, and lines that extend off the plot range to the right indicate the absence of a well-documented (though clinically confirmed) second infection.

Vaccination status (unvaccinated, fully vaccinated, or boosted) was assigned at the time of the first positive test for each infection. Full vaccination corresponded to 14 days following either the second dose of a Pfizer or Moderna vaccine or the first dose of a Johnson and Johnson/Janssen vaccine. A person was considered “boosted” 14 days after an additional Pfizer or Moderna dose following their initial vaccination course.

Estimating viral kinetic parameters

We characterized the viral kinetics of the well-documented infections by fitting a hierarchical piecewise linear model to the viral concentration measurements on a logarithmic scale, following previous methods.^11,14 The model captures the viral proliferation time (i.e., time from first possible detection to peak), peak viral concentration, and clearance time (i.e., time from peak to last possible detection) of acute SARS-CoV-2 infections. The hierarchical structure allowed us to estimate population-level differences in viral kinetic parameters associated with vaccination status, viral lineage, and history of a prior SARS-CoV-2 infection. We adjusted the model fits by the age group of the infected individual (0–29 years, 30–49 years, and 50+ years). Model parameters were fit using a Hamilton Monte Carlo algorithm implemented in R (version 4.1.2) and Stan (version 2.21.3).

Study outcomes

We assessed whether Omicron BA.1/BA.2 viral kinetics differed according to whether the infection was the individual’s first SARS-CoV-2 infection (n = 1,241 well-documented infections) or their second infection (n = 159 well-documented infections; Supplementary Table 1, group 1), stratifying by booster status and adjusting for age. Then, for the subset of individuals with a first infection by a non-Omicron lineage (verified by whole-genome sequencing or an infection onset prior to November 11^th, 2021) and a second infection with BA.1 or BA.2 (n = 235 well-documented infections among 177 individuals; Supplementary Table 1, group 2), we estimated (a) the difference in mean viral kinetics (proliferation time, clearance time, and peak viral load) between first and second infections and (b) the difference in viral kinetics of a second BA.1/BA.2 infection associated with the variant category of the first infection (Alpha, Delta, or other lineages). We note that individuals who underwent two infections may differ from the rest of the population in terms of their exposure patterns, immunity, and/or sampling patterns. Last, using the subset of individuals (n = 58; Supplementary Table 1, group 3) with two well-documented infections, we assessed whether a person’s relative individual-level viral kinetic features (proliferation time, clearance time, or peak viral load) persisted across infections.

Statistical approach

We assessed differences in viral kinetic parameters across category subsets (e.g., by presence/absence of a prior infection or by vaccination status) by subtracting the relevant posterior draws and measuring the posterior probability mass of these differences that sat above/below 0, depending on the scenario. When fewer than 5% of these differenced posterior draws sat above/below zero, we took this as evidence of a significant difference.

To assess relative persistence in individual-level viral kinetic attributes across infections, we measured the Spearman correlation between the rank of the viral kinetic parameter in the first infection and the rank of the same parameter for the same person in the second infection, relative to all other infections caused by the same variant in individuals with two well-documented infections (Supplementary Methods). We used Spearman correlation because, as a rank-based measure, it is less affected by outliers and non-linear relationships between the viral kinetics in the first and second infection. As a sensitivity analysis, we repeated the analysis using the more common Pearson correlation (Supplementary Methods). To assess the statistical significance of the correlations, we conducted permutation tests, randomly assigning first- and second-infection viral kinetic parameters to different individuals and re-calculating the correlation. We then compared the true correlation to the distribution of correlations achieved through 100,000 permutations and concluded a significant difference when the true correlation was higher than 95% of the correlations obtained from permutation.

Study oversight

This work was approved as “research not involving human subjects” by the Yale Institutional Review Board (HIC protocol # 2000028599), as it involved de-identified samples. This work was also designated as “exempt” by the Harvard Institutional Review Board (IRB20-1407).

In BA.1/BA.2 infections, evidence of a previous infection correlates with faster clearance

Omicron BA.1/BA.2 infections in individuals with a history of prior infection (n = 159) featured a faster mean clearance time (4.9 days (4.5, 5.3)) than Omicron BA.1/BA.2 infections in individuals with no prior infection history (n = 1,241; 7.2 days (6.8, 7.5)) (Figure 2A–B; Supplementary Table 1, group 1). Even in individuals who had received a booster vaccine dose, BA.1/BA.2 clearance times were faster in individuals with a history of prior infection (5.1 days (4.5, 5.7)) than in those with no such history (7.1 days (6.7, 7.6); Supplementary Table 2). Omicron BA.1/BA.2 clearance rates were also faster in vaccine-boosted individuals with a history of prior infection (2.6 Ct/day (2.2, 3.0)) than in those without (2.0 Ct/day (1.8, 2.1); Supplementary Table 2). The reduction in second-infection clearance time also resulted in greater symmetry between proliferation and clearance times (ratio of proliferation time to clearance time: 0.7 (0.6, 0.8) for the n = 1,241 BA.1/BA.2 infections with no prior infection history vs. 0.9 (0.7, 1.0) for the n = 159 BA.1/BA.2 infections with prior infection history).

In individuals with multiple infections, second infections were cleared faster than first infections, especially when there was an intervening vaccine dose

We found no significant difference in proliferation time or peak viral concentration between first infections caused by Alpha, Delta, or other/unspecified lineages in individuals with a later Omicron BA.1/BA.2 infection (n = 78) vs. second infections caused by Omicron BA.1 or BA.2 in individuals with a previous Alpha, Delta, or other/unspecified infection (n = 157) (Supplementary Table 1, group 2). However, in this group, clearance times were substantially faster in second infections (5.1 days (4.7, 5.7) vs. 8.8 days (7.8, 9.8) in first infections, Supplementary Table 3). Clearance rates were also substantially faster in second infections within this group (2.4 Ct/day (2.1, 2.7) vs. 1.6 Ct/day (1.3, 1.8) in first infections, Supplementary Table 3). Individuals in this group who had an intervening vaccine dose between their first and second infection had a somewhat faster mean second-infection viral clearance time than those without an intervening dose (5.2 days (4.5, 9.5) with an intervening dose; 6.2 days (4.2, 8.8) without an intervening dose), though these estimates carry substantial uncertainty (Figure 2C–F, Supplementary Table 3). As before, the reduction in second-infection clearance time also resulted in greater symmetry between proliferation and clearance times (ratio of proliferation time to clearance time: 0.4 (0.3, 0.6) for the n = 78 first infections vs. 0.9 (0.7, 1.1) for the n = 157 second infections).

• Download figure
• Open in new tab

Figure 2. Viral kinetics of first vs. second infections.

(A–B): Mean posterior viral trajectory (solid lines) with 95% credible interval (shaded region) for well-documented BA.1/BA.2 infections in individuals where the infection was (A) their first recorded SARS-CoV-2 infection (n = 1,241) or (B) their second recorded SARS-CoV-2 infection (n = 159; see Supplementary Table 1, group 1). (C–F): Mean posterior viral trajectory (solid lines) with 95% credible interval (shaded region) for well-documented infections in individuals who were infected twice during the study period. Panels (C, E) depict the viral kinetics of first Alpha/Delta/Other/Unspecified infections in unvaccinated (C, n = 42) and vaccinated (E, n = 15) individuals who later had a BA.1/BA.2 infection. Panels (D, F) depict the viral kinetics of second BA.1/BA.2 infections in individuals who did not have a vaccine dose between their first and second infection (D, n = 5) and in individuals who did have a vaccine dose between their first and second infection (F, n = 102) as well as a previous Alpha/Delta/Other/Unspecified infection (Supplementary Table 1, group 2, omitting any individuals with unknown vaccination status). In all panels, grey points depict the measured viral concentration for a single test. For each person, the points were shifted horizontally so that the individual’s mean posterior peak viral concentration sits at day 0. Black points and whiskers (A, C, E) depict the mean and 95% credible interval for the proliferation time, peak viral concentration, and clearance time, from left to right, for first infections. These values are repeated in grey on the lower plots (B, D, F) to facilitate comparison with the viral kinetics of second infections.

No evidence that the kinetics of a second infection by BA.1/BA.2 differ according to the first infection’s lineage

We found no evidence the variant category of a previous infection (Alpha vs. Delta vs. other/unspecified) affected the mean viral kinetics of a later Omicron BA.1/BA.2 infection (n = 157, Supplementary Table 1 group 2; Supplementary Table 5). That is, regardless of the lineage of the initial infection, the viral kinetics of second BA.1/BA.2 infections had similar features to one another, though not to the first infection.

An individual’s relative clearance speed is preserved across infections

Among individuals with two well-documented infections (n = 58; Supplementary Table 1, group 3), the ranking of clearance times was roughly conserved across first and second infections; that is, those with a relatively short clearance time in their first infection tended to have a short clearance time in their second infection, and vice versa (Figure 3). The Spearman correlation coefficient between individual-level mean clearance times for first vs. second infections was 0.475 overall (p = 0.0001, permutation test). When stratified by the variant category of the first infection, the Spearman correlation between individual-level mean clearance times for first vs. second infections was 0.783, 0.352, and 0.396 for Alpha, Delta, and infections with other/unspecified lineages, respectively. We found somewhat weaker evidence that a person’s proliferation time rank was conserved across infections, with a Spearman correlation coefficient of 0.267 (p = 0.02, permutation test) for first vs. second infections. We did not find evidence that an individual’s peak viral concentration rank was maintained across infections. These findings were robust to the choice of correlation statistic (Pearson vs. Spearman; Supplementary Table 6).

• Download figure
• Open in new tab

Figure 3. Relative viral kinetic parameters for first vs. second SARS-CoV-2 infections.

(A): Spearman correlation (points) of peak viral load, proliferation time, and clearance time for first vs. second infections among n = 58 individuals with two well-documented infections (Supplementary Table 1, group 3). The height of the grey shaded area is proportional to the probability that the Spearman correlation would fall at a particular value given no correlation between first- and second-infection parameters. (B-D): Scatter plots of the normalized order statistics (points) for the peak viral concentration (B), proliferation time (C), and clearance time (D) for first vs. second infections in the n = 58 individuals with two well-documented infections. Points are colored according to the variant of the first infection (Other/None: black; Alpha: blue; Delta: red); all second infections were caused by BA.1/BA.2. Dashed lines depict the best-fit linear regression to the normalized order statistics when stratified by the variant of the first infection. The solid black line depicts the overall best-fit linear regression to the normalized order statistics combined across all first-infection variants. The grey regions depict the 95% confidence regions for this overall trend line. Order statistics were calculated by ranking the mean posterior values for each parameter (peak viral load, proliferation time, clearance time) for each infection within a variant/cardinality group (e.g., Delta/first infection). Then, these ranks were divided by the maximum rank for that variant/cardinality group, so that the normalized order statistics fall between 0 and 1.