Viral Dynamics of Omicron and Delta Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Variants With Implications for Timing of Release from Isolation: A Longitudinal Cohort Study

Abstract

Individuals with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection have been told to self-isolate to avoid transmission to others. Early in the pandemic, Centers for Disease Control and Prevention (CDC) guidelines for isolation were based on estimates of the duration of infectivity, with early studies showing rare culture positivity and transmission beyond 10 days in immunocompetent hosts [1]. In January 2022, CDC guidelines shifted to recommend 5 days of isolation from symptom onset or positive test if asymptomatic, followed by an additional 5 days of strict mask-wearing anytime there is contact with other people [2]. This newer guidance was based on SARS-CoV-2 transmission studies that showed that most transmission occurs early in the course of infection [3, 4]; the increased awareness of the mental health, economic, and social impacts of prolonged isolation [5]; and that only 25%–30% of cases truly isolate for a full 10 days [6].

A proposed strategy to limit time in isolation of SARS-CoV-2–infected individuals has been to use antigen rapid diagnostic testing (RDT) as a proxy for infectiousness [2]. A test-to-return program in a Massachusetts school district found a positivity rate of 35% when performed on day 6 after symptom onset or positive test, concluding that such a program helps identify those safest to return to the classroom prior to 10 days [7]. Modeling efforts have further suggested that antigen RDT could be used to reduce the self-isolation period while preventing ongoing disease transmission [8].

New SARS-CoV-2 variants have shown increased transmissibility, potentially impacting best practices for isolation guidelines. Notably, the Omicron variant may be up to 3 times more transmissible than the Delta variant [9], with a shorter incubation period [10] but lower viral loads at diagnosis [11] than previously circulating SARS-CoV-2 lineages. Early SARS-CoV-2 viral challenge data with wild-type virus in seronegative, healthy volunteers showed, on average, nasal culture clearance by day 7 with culture positivity up to 12 days from SARS-CoV-2 exposure [12]. The length of time that vaccinated individuals, both primary series and boosted, can transmit virus and the kinetic difference by variant remain unclear.

When to start the clock on the isolation period has proven to be an additional complication when implementing the CDC guidelines. Schools and other large organizations may use day of positive test, while the CDC notes that day 0 is the day of symptom onset and only the day of positive test for cases that remain asymptomatic. For large-scale screening programs that use repeat SARS-CoV-2 testing at regular intervals for early case identification, the complications from operationalizing the use of the start of symptoms can be prohibitive. However, for those who develop symptoms after testing, this practice may shorten the isolation period and allow individuals to go back into conjugate living situations where strict mask-wearing is difficult.

We aimed to document within-host viral dynamics of 2 recent SARS-CoV-2 variants, Delta and Omicron, during the period when most individuals would leave isolation based on CDC guidelines. We recruited participants from a university campus that had a multifaceted surveillance testing and coronavirus disease 2019 (COVID-19) control program [13]. We compared infection with Omicron vs Delta, the vaccination and booster status of the infected individual, and the difference from test date and symptom onset date on detectable virus and culture positivity.

METHODS

The SARS-CoV-2 Viral Dynamics Post-vaccination Study is an observational cohort study that, beginning in November 2021, has been enrolling Boston University (BU) students, faculty, and staff. Participants are enrolled after a verbal assent or electronic self-consent process following diagnosis with SARS-CoV-2 by polymerase chain reaction (PCR) testing as a part of the BU SARS-CoV-2 screening program that includes regular testing 1–2 times per week. Participants answer an initial clinical questionnaire and thereafter self-perform daily symptom screens and self-collect anterior nasal swabs, with electronic medical record and diagnostic specimen laboratory data abstraction performed by study staff.

Following study enrollment, participants can also opt into an antigen RDT substudy. Those who provide verbal assent are provided with 3 Abbott BinaxNOW kits to self-perform according to the manufacturer’s directions on days 4–6 from diagnosis. Participants are asked to upload a photograph of the test strip at the time it is read into our REDCap database (version 12.1.1, Vanderbilt University), where study staff then interpret the results.

Diagnostic isolates are tested with reverse-transcription quantitative PCR (RT-qPCR) using a 2-target SARS-CoV-2 assay with RNase P control [1] as a part of the BU SARS-CoV-2 surveillance program at the BU Clinical Testing Laboratory [3]. Positive samples are then transferred to the National Emerging Infectious Diseases Laboratories, where viral genomes are sequenced using a modified ARTIC primer–based protocol for amplification and the Illumina platform for sequencing [14]. For participants for whom diagnostic isolate sequencing was not yet available (N = 32), variant determination was made based on the timing of the diagnostic isolate: all infections after 28 December 2021 were considered Omicron because BU sequencing surveillance activities detected <1% Delta circulating [11]. Study isolates are collected in viral transport media and frozen following collection. Samples are then analyzed in 2 ways. The presence of SARS-CoV-2 viral nucleic acid in the sample is determined by SARS-CoV-2 CDC N1 RT-PCR and 200 µL of sample are incubated with Caco2 cells for 4 days. At the end of culturing, cells are fixed, and the presence or absence of SARS-CoV-2 viral replication is determined through indirect immunofluorescence microscopy using an antibody that recognizes the SARS-CoV-2 N protein. Cultures are determined to be positive for viral growth if they show a cytoplasm-restricted fluorescence signal above background.

In this analysis, we compared demographic and clinical characteristics of participants infected with the Delta and Omicron SARS-CoV-2 variants. Fully vaccinated individuals had completed an initial vaccine series with either a single vector or 2 messenger RNA (mRNA) vaccines listed for emergency use by the World Health Organization (WHO), with the exception of 1 participant who received a non-WHO–listed vaccine and was subsequently boosted with an mRNA vaccine. We compared culture positivity beyond day 5, time to culture conversion, and duration of positive cultures when calculated from date of diagnostic test to symptom onset. Culture conversion was defined as the day of first negative culture result with no subsequent positive cultures. For the purposes of all cycle threshold (Ct) comparisons, we used the N1 genes from both assays. We plotted scatter plots and fitted mixed-effect models to assess the relationship between Ct values and time since diagnosis and symptom onset and by variant and vaccination status. Curve fits were performed using generalized additive mixed models with random intercept and a cubic B-spline with 4 knots to incorporate individual participant’s Ct trajectories. To look at time to culture conversion for the whole cohort and by variant and vaccination status, we generated Kaplan–Meier plots where the event was defined as the first negative SARS-CoV-2 culture with no subsequent positives. Heat-treated baseline diagnostic isolates were not culturable and were presumed positive for the purposes of this analysis. When the first recorded negative culture was preceded by missed tests, we used interval censoring with the earliest point of the interval being the time of the last positive test or diagnosis if there were no positive cultures recorded. Additionally, we compared median Ct values at diagnosis by variant and vaccine status using Wilcox nonparametric tests. Finally, for the antigen RDT substudy, we calculated the sensitivity and specificity of the Abbott BinaxNow compared with culture positivity at days 4–6 from diagnosis and plotted Ct values by day from diagnosis and RDT result. Analyses were conducted with R (version 4.0).

The BU Charles River Campus Institutional Review Board (IRB) and the BU Medical Campus IRB approved the study.

RESULTS

Of 92 SARS-CoV-2 RT-PCR–positive participants enrolled in the study, 17 (18.5%) were infected with the SARS-CoV-2 Delta variant and 75 (81.5%) with Omicron (Table 1). Most participants (N = 65, 69.1%) had symptoms at the time of diagnosis. While a greater proportion of individuals infected with Omicron had received a booster vaccine compared with those infected with Delta (42.7%, n = 32 vs 5.9%; n = 1; P = 0.004; Table 1), all participants had completed the initial SARS-CoV-2 vaccine series (Table 1). The median number of days from most recent vaccination to infection was 245 (interquartile range [IQR], 202–282) for fully vaccinated, not boosted participants and 42 (IQR, 30–55) for fully vaccinated, boosted participants (Supplementary Figure 1).

Overall, 84% of participants from diagnostic test and 71% of participants from symptom onset had culture-converted (no growth) by day 6 (Table 2). Overall, 48% (N = 44) of participants never had a culture-positive research isolate. Next, we considered the impact of calculating the isolation period as days from diagnosis (common in large institutional settings such as universities) compared with days from symptom onset (CDC guidance) [2] on viral dynamics at release from isolation. In this cohort, almost half of the participants were diagnosed after symptom onset (N = 38, 41.3%), while 10 (10.9%) tested positive while presymptomatic. Use of time from symptom onset rather than time from diagnosis shifted the axis right for those diagnosed after symptom onset and left for presymptomatic participants. In comparison to within-host viral load decay over time from diagnosis (Figure 1A), the mixed-effect model of within-host viral load decay from symptom onset (Figure 1B) shows a plateau in the presymptomatic period (days –3 to 0) followed by a more gradual viral decline. Similarly, adjusting the viral culture results from days since diagnosis (Figure 1C) to days from symptom onset (Figure 1D) reveals a considerable drop in culture positivity following day 4 with an associated reduction in the maximum duration of culture positivity of 3 days (from 15 days to 12 days; Table 2).

Figure 1.

Progression of polymerase chain reaction N1 cycle threshold values from diagnosis (A) and symptom onset (B), and culture growth from diagnosis (C) and symptom onset (D). Abbreviation: Ct, cycle threshold.

Open in new tabDownload slide

There were no significant differences in time to viral culture conversion between Delta and Omicron variants, even when stratified based on whether participants had received a COVID-19 booster vaccination or had only completed the initial vaccine series (Figure 2B). The trends for within-host viral load decay over time were similar in those who completed the initial vaccine series and were infected with the Delta variant (Figure 3B) and the Omicron variant (Figure 3C). However, for participants who had received a COVID-19 booster vaccination, there was a trend toward slower with-in host viral load decay (Figure 3D), though daily culture conversion rates were similar (Figure 2B) and RT-PCR Ct was higher in those boosted with Omicron infection at diagnosis (median, 27.5) than in those fully vaccinated with either Delta (median, 18.8; P = 0.004) or Omicron infection (median, 26.4; P = 0.36; Supplementary Figure 2).

Figure 2.

Kaplan–Meier curves indicating days from diagnosis to negative viral culture for all participants (A) and by severe acute respiratory syndrome coronavirus 2 variant and vaccination status (B).

Open in new tabDownload slide

Figure 3.

Polymerase chain reaction N1 cycle threshold at days since symptom onset for all participants (A); Delta-infected, fully vaccinated participants (B); Omicron-infected, fully vaccinated participants (C); and Omicron-infected, boosted participants (D). Abbreviation: Ct, cycle threshold.

Open in new tabDownload slide

For the subset of 14 participants in the RDT substudy, the sensitivity and specificity of the Abbott BinaxNOW test were 100% and 86%, respectively, at day 4 through day 6 from diagnosis when compared with SARS-CoV-2 culture (N = 32 isolates; Table 3), yielding a negative predictive value of 100% and a positive predictive value of 50%. RT-PCR Ct values were higher in those who tested negative on the RDT (Supplementary Figure 3).

DISCUSSION

Our findings suggest that the majority of young, healthy, vaccinated adults have a limited infectious period for SARS-CoV-2 based on culturable virus and rapid viral decay. We found that beyond 5 days from symptom onset, only 17% remained culture-positive. Our median time to Omicron culture conversion was 2 days (IQR, 1–5) for boosted participants with Omicron; 3 days (IQR, 1–5) for vaccinated, unboosted participants with Omicron; and 3 days (IQR, 1–6.5) for participants with Delta. This is notably earlier than the median 6 days to culture conversion reported in an older cohort with more medical comorbidities [15]. Our results are similar to data from a National Basketball Association cohort that showed that 40% with cycle thresholds of less than 30 at day 5, though this study did not examine culture positivity [16].

Whether we calculated days to culture clearance from date of diagnosis or from symptom onset led to differences in the likelihood that individuals released from isolation would still be infectious and how quickly some were released. Reliance on time from diagnosis for determining duration of isolation kept 41% of people in isolation longer than necessary (using culture conversion as a measure of noninfectiousness). Additionally, the 10 (11%) who were presymptomatic at the time of diagnosis were released from isolation before they had reached the CDC-recommended 5-day point, although only 2 (20%) remained culture-positive at that point, which was similar to the overall cohort fraction. It is much easier programmatically for large institutions with serial testing programs to rely on time from diagnosis, a number they capture through these programs, rather than symptom development, which requires additional follow-up for those without symptoms and honest reporting of those diagnosed. It is reassuring to see that only 17% of our cohort failed to culture-convert by day 6; however, it will be key to continue to reinforce strict masking for days 6–10 from symptom onset, especially in scenarios with asymptomatic screening strategies where symptoms may develop after testing.

We found no major differences in culture conversion or viral load decay between the Delta and Omicron variants. This is consistent with previous data that showed no major differences in Omicron infection duration when compared with Delta [16] or time to culture or PCR conversion between the 2 variants [15]. Similar to the findings of Boucau et al [15], we found no significant difference in culture or PCR conversion in boosted participants when compared with those who were fully vaccinated. This suggests that the population-level protection provided by COVID-19 booster vaccination doses against Omicron [17, 18] result from protection against infection rather than altered viral kinetics in vaccinated individuals, causing reduced transmission or more rapid clearance.

Among participants who used an antigen RDT in days 5–7 following SARS-CoV-2 diagnosis, RDT had perfect negative predictive value and sensitivity when compared with culture. While we used a single type of antigen RDT in our substudy, the Abbott BinaxNOW has been shown to successfully detect the Omicron variant [19]. Our findings are similar to those from early reports in other settings that suggest 30%–55% positive antigen RDT between day 5 and day 10 of illness [7, 20, 21]. One study that used a different antigen test showed higher antigen RDT positivity (75%) but similar culture positivity (35%) at day 6 [20]. Given the similar negative predictive value, the work by Cosimi et al [22] highlights possible differing performance characteristics at the end of infection across tests. Our preliminary findings suggest that a young, healthy, fully vaccinated individual can be reassured with negative antigen RDT that they are unlikely to be shedding viable virus toward the end of their infection. However, reliance on antigen RDT for release from isolation would also result in prolongation of isolation for those who do not have culturable virus, potentially up to 50% of those with positive antigen tests.

Our study is limited in its generalizability as most participants were young and healthy and all had completed at least the initial COVID-19 vaccination series. Additionally, many participants had recently received booster vaccination when Omicron began to circulate, possibly resulting in higher neutralizing antibody titers than those further out from their last vaccine. We were also limited by the delay from diagnosis to enrollment in the study as many participants moved into isolation housing, resulting in 48% of participants not having a culture-positive research isolate. The diagnostic isolates were heat-treated and therefore not available for culturing. Additionally, our work was likely not well powered to determine differences between Delta and Omicron variants stratified on COVID-19 vaccination status. While our study is strengthened by our daily sampling and culture data, it is not clear how well culture positivity in the laboratory correlates to transmissibility of SARS-CoV-2 at the end of an infection, nor did we rule out reinfection in individuals with persistently positive isolates. More work needs to be done to understand why certain immunocompetent individuals have prolonged culture positivity and how to predict which cases will remain at risk of transmitting beyond the isolation period. Regardless, it is clear that even in a young, healthy cohort with presumably optimal response to initial and booster vaccination, an overwhelming majority of both Delta and Omicron variant SARS-CoV-2 infections culture-convert by day 6. Our work provides further support to the guidelines for strict masking beyond the initial 5-day isolation period for SARS-CoV-2 infections to help prevent transmission from the minority of cases who remain culture-positive.

Supplementary Data

Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Acknowledgments. The authors acknowledge the Boston University (BU) students, faculty, and staff who have been impacted by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic.

Disclaimer. The contents presented here are solely the responsibility of the authors and do not necessarily represent the official views of the National Institutes of Health (NIH), BU, or the Massachusetts Consortium on Pathogen Readiness (MassCPR).

Financial support. This project was supported by the MassCPR and Boston University. T. C. B., L. F. W., and Z. Z. are supported by the NIH (NIAID K23 grant AI152930-01A1 and R35GM141821). This publication was additionally supported by the National Center for Advancing Translational Sciences, NIH, through the Boston University Clinical & Translational Science Institute (BU-CTSI) grant 1UL1TR001430.

References

Author notes