1. Introduction
Although much progress has been made in the field of SARS-CoV-2 diagnostics, to date the dynamics of SARS-CoV-2 replication, shedding and infectivity in humans remain incompletely understood [
1]. Infectivity by any viral pathogen is a very complex process that involves multiple host and viral factors. Different studies on SARS-CoV-2 infection have suggested that the magnitude and duration of viral shedding correlates with biological characteristics of the virus, disease severity, patient age and sex, viral load, stage of the infection and immune status of the infected individual [
1,
2,
3,
4,
5].
SARS-CoV-2 RNA has been detected in different body fluids in addition to nasal swabs during acute infection, including saliva, peripheral blood, ocular secretions, anal swabs, and urine [
1,
2,
3,
4,
5,
6,
7,
8,
9,
10,
11,
12,
13,
14,
15]. A recent meta-analysis found that the mean shedding time of SARS-CoV-2 in the upper respiratory tract, lower respiratory tract, stools, and serum was 17.0, 14.6, 17.2, and 16.6 days, respectively [
5]. However, results from multiple investigations have been very heterogeneous [
8]. Most of these studies used only a single time point to collect samples, included limited sites of biologic specimen collection, and only utilized patients with severe forms of COVID-19 [
5,
9,
10]. As a result, variations in time between symptom onset and sample testing can be a confounding factor when analyzing data from these studies on viral shedding and infectivity. In addition, virus isolation from non-respiratory tract specimens has been unsuccessful in most cases [
1,
9,
10,
11,
12,
13,
14,
15].
To investigate SARS-CoV-2 persistence and replication competence in individuals with mild or severe symptoms, we carried out the present study on 125 COVID-19 cases. The dynamics of infectious virus shedding from multiple biologic sites during acute infection were determined through weekly longitudinal sampling in 123 individuals. We estimated viral clearance rates and potential infectivity at multiple sites, among immunocompetent and immunosuppressed individuals with mild or severe COVID-19 disease. In addition, the shedding dynamics of the B.1.1.28 lineage (ancestral strain) and the Gamma strains of SARS-CoV-2 were compared.
The objective of the study was to investigate the presence and potential infectivity of SARS-CoV-2 in nasopharyngeal swabs, saliva, urine, blood, and feces (anal swab), among immunocompetent and immunocompromised patients, during the acute and convalescence phases of COVID-19, in patients infected with two different virus strains.
3. Results
Ninety-eight immunocompetent patients and twenty-five immunosuppressed individuals were included in the study. Characteristics of the study population is shown in
Table 2. The immunocompetent group were younger (42.7 vs. 57.4 years old,
p < 0.0001), had a higher mean body mass index (27.4 vs. 24.9 kg/m
2,
p = 0.0402) and a higher mean log
10 SARS-CoV-2 viral load (5.6 vs. 4.0,
p = 0.0014) than the immunosuppressed group. The viral genome sequences obtained from all clinical specimens from immunocompetent patients with mild forms of disease belonged to either the B.1.128 lineage (51 cases) or the Gamma lineage (47 cases). The Gamma strain emerged in Brazil in 2020 [
24,
25,
26]. Similarly, among immunosuppressed individuals the viral genome sequences also belonged to either the B.1.128 lineage (14 patients) or to the Gamma variant (11 patients).
None of the included patients were positive for Influenza A and B virus and syncytial respiratory virus (SRV), when we tested nasal swab samples collected at week 1.
SARS-CoV-2 B1 strain detection at different sites in immunocompetent and immunosuppressed patients is show in
Table 3. The highest percentage of virus detection was in nasal samples, 35.8% of samples from immunocompetent subjects and 45.7% from immunosuppressed individuals. Saliva was the second most frequently positive site, 6.0% in immunocompetent subjects and 25.7% in the immunocompromised group. This difference was highly significant (
p = 0.0008). Similarly, urine (
p = 0.0002) and blood (
p = 0.0001) samples from immunosuppressed patients were more frequently positive than from immunocompetent individuals. A small and comparable percentage of anal swabs from both groups were also positive.
SARS-CoV-2 Gamma strain detection at different sites in immunocompetent and immunosuppressed patients is detailed in
Table 4. Similar to the results with the B1 strain, the highest percentage of positive samples were from the nasal cavity in both immunocompetent (42.4%) and immunosuppressed (62.5%) subjects. This difference was significant (
p = 0.0248). Virus was also detected in a higher percentage of samples from saliva, (
p = 0.0001), urine (
p = 0.0001) and blood (
p = 0.0075) from immunosuppressed than from immunocompetent individuals. A small and comparable percentage of anal swabs from both groups were also positive.
The ability to propagate the B1 strain of SARS-CoV-2 in culture from different sites in immunocompetent and immunosuppressed patients is shown in
Table 5. In samples from immunocompetent individuals, virus could be cultured from only 3 of 76 nasal samples (3.9%) and 1 of 9 saliva samples (11.1%). In marked contrast, virus was cultured from 5 of 13 nasal samples (38.5%) and 3 of 8 saliva samples (37.5%) from the immunosuppressed group. This difference in virus cultivation from nasal samples was significant (
p = 0.0014).
Among the 120 samples submitted to virus culture (
Table 5), we observed contamination in 8 samples (1 anal swab and 7 nasal swabs). The contaminated isolates were submitted to two treatments. Initially, they were subjected to microfiltration using a 0.22 µ millipore filter (Minisart
® 17761-ACK—Sartorius Stedim) which eliminates most bacteria [
27]. After this initial step, 1% penicillin-streptomycin is added to the filtered material to eliminate bacterial cells [
28,
29].
Among the 8 cases, after the use of this specific protocol, viral isolation was finally observed in 2 samples (nasal swab samples).
SARS-CoV-2 Gamma strain propagation in culture from samples obtained from different sites is described in
Table 6. Virus was cultured from samples obtained from the nasal cavity of immunosuppressed (6 of 20, 30.0%) but not from immunocompetent (0 of 84) individuals. No virus was obtained from the cultures of saliva, urine, anal and blood from all patients.
The duration of viral shedding at different sites in immunocompetent and immunosuppressed individuals over time with prolonged viral shedding of at least 10 days is shown in
Table 7. Among the immunocompetent patients, viral shedding from different sites varied from 16 days to 50 days and was longest (50 days) in nasal specimens. For immunosuppressed individuals, viral shedding varied from 177 days to 212 days and was also longest for nasal specimens (212 days).
The detection of the B1 and Gamma SARS-C0V-2 strains at different body sites over time is shown in
Table 8. The highest number of positive samples was observed in nasal samples collected at week 1. The percent positive was similar for B1 (73.5%) and Gamma (76.1%) isolates. The second highest percent positive was in week 1 saliva for B1 (21.1%) and in week 1 anal swabs for Gamma (12.8%). The percentage of virus positive in saliva was significantly higher for individuals infected with B1 (21.1%) than in those infected with the Gamma strain (4.5%) (
p = 0.0390). The B1 strain was only detected in 1/51 blood samples collected at week 1 and in none of the samples collected at weeks 2 and 3. In contrast, the Gamma strain was detected in 11.6%, 11.1% and 9.1% of blood samples collected at week 1, 2 and 3, respectively. B1 and Gamma detection in nasal samples (
p < 0.0001), and B1 detection in saliva (
p = 0.0148), progressively decreased over the 5-week testing period.
Analysis of all viral genome sequences obtained from clinical specimens of all patients did not reveal any mutations which would be associated with prolonged viral shedding, viral replication or pathogenicity.
4. Discussion
Among the immunocompetent and immunosuppressed individuals with COVID, a greater number were positive for SARS-CoV-2 in nasopharyngeal swabs than in any other sites tested regardless of the infecting viral strain or the time the sample was collected after infection initiation. At sites other than the nasopharynx—saliva, urine and blood—virus was detected more frequently in immunosuppressed individuals than in those who were immunocompetent. The few positive anal samples were not significantly different between the immunocompetent and immunosuppressed populations. Virus detection from all sites declined over time but persisted longer in immunosuppressed individuals. Viral propagation in culture was also achieved more frequently in samples from the immunosuppressed group. All these findings were consistent with earlier reports [
1,
30,
31,
32]. In immunocompetent patients, considering only samples collected ten or more days after onset of disease, 3.6%, 1.9% and 0.25% of blood, anal and urine samples, respectively, were positive for SARS-CoV-2. In contrast, among immunosuppressed individuals, considering only samples collected ten or more days after onset of disease, SARS-CoV-2 was identified in 41.17%, 5.4% and 22.9% of blood, urine and anal samples, respectively. Our data also confirmed that in immunocompetent individuals with mild COVID-19, successful virus cultivation from clinical samples obtained ten days after the onset of symptoms is an uncommon event, regardless of the viral strain involved [
1,
9,
10,
33] and that immunocompromised patients with COVID-19 are at elevated risk for prolonged viral shedding and persistent replicating capacity [
33,
34,
35]. In our study, despite the frequent finding of SARS-CoV-2 in different biological materials and in some cases for very prolonged periods, viral replication in vitro was identified only in nasopharyngeal and saliva samples. SARS-CoV-2 has been described in different body fluids, by many authors from different parts of the world. Most of these reports, however, included only a small number of prospectively collected samples from only the nasopharynx and saliva. The present study reports findings from multiple sites in a large series of immunocompromised and immunocompetent individuals.
One of our patients, a 40-year-old male who had undergone a prior autologous hematopoietic stem cell transplant due to a diffused large B-cell lymphoma, was found to persistently shed SARS-CoV-2 that could be propagated in culture from nasal swab and saliva samples for more than 196 days. Other studies have identified atypical cases with prolonged shedding of infectious virus for up to 200 days [
33,
34,
35,
36,
37].
Several prior studies have proposed that the persistence of SARS-CoV-2 capable of replication was related to immunosuppression [
38,
39,
40]. As extensively described for numerous viral infections, both innate immunity and T cell-mediated adaptive immune response are essential for the clearance and long-term inhibition of viral infections [
38,
39]. It is also important to acknowledge that studies have associated the persistence of SARS-CoV-2 with the severity of disease, the presence of co-morbidities and use of glucocorticoids [
41,
42,
43,
44]. In our series of immunosuppressed patients, all were hospitalized with severe forms of COVID-19.
Viral evolution of SARS-CoV-2 over time has led to the emergence of numerous variants. Differences in immune evasion, viral loads, and duration of shedding between variants have been described [
1,
5]. It was not surprising, therefore, that in our study differences in viral detection, persistence, and proliferation in culture between the B1 and Gamma strains were observed.
Only a few studies have described temporal changes in SARS-CoV-2 detection [
45,
46,
47]. The virus detection rate in blood in different studies was between 28% and 32% for hospitalized patients. However, patients in intensive care had rates up to 78% [
3,
45,
46,
47]. Our study is unique in that it describes the prospective presence of SARS-CoV-2 in non-hospitalized individuals with only mild forms of COVID-19. In addition, to our knowledge no prior investigation has reported in vitro viral propagation from SARS-CoV-2-positive blood samples.
Review studies have found overall rates of SARS-CoV-2-positive urine samples to be low and levels correlated with severe disease status [
48]. The present study is in agreement with these prior reports. SARS-CoV-2 positive urine samples were identified in 0.25% of immunocompetent COVID-19 patients with mild disease, but in 22.9% of immunosuppressed patients with severe disease. None of the virus-positive samples yielded virus upon cultivation. Similarly, SARS-CoV-2 has been identified in anal swabs [
6,
10,
11,
12]. The dynamics of SARS-CoV-2 shedding at this site has been described as erratic, with the highest viral loads reported during the first weeks after symptom onset and described in severe cases [
5,
6,
10,
11,
12,
41]. In our study, SARS-CoV-2 detection in anal swabs was observed in 10 samples (1.9%) from immunocompetent individuals and in 5.4% of swabs from immunosuppressed individuals. Replicative virus was not obtained from any anal sample. This was in accord with findings from prior studies [
5,
41,
49].
Much progress has been made in understanding the transmission dynamics of SARS-CoV-2 and duration of infectivity. The World Health Organization (WHO) and the Center for Disease Control (CDC) have modified their recommendations in response to data indicating that infectivity decreases to essentially zero after about 10 days from symptom onset in mild to moderately ill patients and after about 15 days in critically ill and immunocompromised patients, with a maximum reported interval thus far of 20 days [
50]. To our knowledge, the present study is unique in comparing both SARS-CoV-2 detection by RT-PCR and by the presence of culturable virus in serially collected samples from different clinical sites, in both immunocompetent and immunosuppressed individuals. We also directly compared the shedding dynamics of two different SARS-CoV-2 strains—B and Gamma—in these individuals. This enabled us to estimate the time period of virus detection and its relationship to capacity for replication in different biological specimens.
It is necessary to acknowledge the limitations of our study. All of our cases occurred during the initial two years of the pandemic, before the widespread circulation of additional viral variants. Therefore, our findings might not be generalizable to current and future SARS-CoV-2 variants. Additionally, with the continuous circulation of SARS-CoV-2 in the community, previous or recurrent infection, vaccination, or a combination of both could alter viral shedding patterns differently from those observed in our study. In addition, while study staff were trained in sample collections it is possible that sample variation in quality could have occurred. However, the consistency in findings from longitudinal sampling suggest that this variation was likely minimal.