key: cord-0913744-8z1dim28
authors: Mondi, Annalisa; Lorenzini, Patrizia; Castilletti, Concetta; Gagliardini, Roberta; Lalle, Eleonora; Corpolongo, Angela; Valli, Maria Beatrice; Taglietti, Fabrizio; Cicalini, Stefania; Loiacono, Laura; Di Gennaro, Francesco; D’Offizi, Gianpiero; Palmieri, Fabrizio; Nicastri, Emanuele; Agrati, Chiara; Petrosillo, Nicola; Ippolito, Giuseppe; Vaia, Francesco; Girardi, Enrico; Capobianchi, Maria Rosaria; Antinori, Andrea
title: Risk and predictive factors of prolonged viral RNA shedding in upper respiratory specimens in a large cohort of COVID-19 patients admitted in an Italian Reference Hospital
date: 2021-03-03
journal: Int J Infect Dis
DOI: 10.1016/j.ijid.2021.02.117
sha: 3e847db88779ec0b26d4cdda05b3f9e3034b31e8
doc_id: 913744
cord_uid: 8z1dim28

BACKGROUND: Few data about predictors and outcomes associated with prolonged SARS-CoV-2 RNA shedding (VS) are available. METHODS: Retrospective study including all patients admitted with COVID-19 in an Italian reference hospital for infectious diseases between March 1 and July 1, 2020. Predictors of viral clearance (VC) and prolonged VS from upper respiratory tract were assessed by Poisson regression and logistic regression analyses. The causal relation between duration of VS and probability of clinical outcomes was evaluated through inverse probability weighted Cox model. RESULTS: 536 subjects were included. Median duration of VS from symptoms onset was 18 days (IQR 12-26). The estimated 30-day probability of VC was 70.2% (95%CI:65-75). At multivariable analysis, patients with comorbidities (aIRR = 0.88, p = 0.004), lymphopenia at hospital admission (aIRR = 0.75, p = 0.032) and with moderate/severe respiratory disease (aIRR = 0.42, p < 0.001) had a lower chance of achieving VC. The development of moderate/severe respiratory failure (aOR = 2.65, p = 0.003), a delayed hospital admission after symptoms onset (aOR = 1.18, p < 0.001), having baseline comorbidities (aOR = 1.25, p = 0.019) and D-dimer >1000 ng/mL at admission (aOR = 1.76, p = 0.035) independently predicted prolonged VS. The achievement of VC doubled the chance of clinical recovery (aHR = 2.17, p < 0.001) and reduced the probability of death/mechanical ventilation (aHR = 0.36, p = 0.002). CONCLUSIONS: In this study, severity of respiratory disease, comorbidities, delayed hospital admission and inflammatory markers negatively predicted the achievement of VC, which resulted to be associated to better clinical outcomes. These findings highlight the importance of prompt hospitalization of symptomatic patients, especially in presence of signs of severity or comorbidities.

The emergence and rapid spread of coronavirus disease 2019 (COVID-19) outbreak, caused by the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), has become a global health emergency and one of the greatest challenges of our century. As of February 24 th , 2021, approximately 111 million confirmed cases with more than 2,4 million deaths have been reported worldwide [1] .

Several reports have shown that the shedding profile of SARS-CoV-2 differs from the other coronavirus, whereas is similar to that of influenza virus [2, 3] . Indeed, viral RNA can be detected in upper respiratory tract (URT) few days before symptoms onset, subsequently peaks, within the first week of infection, and then gradually declines over time [2] [3] [4] . According to recent evidences, the median duration of SARS-CoV-2 shedding in URT is 17 days [5] . However, longer durations, up to 83 days, have been described [6] . In several reports, conducted in small cohorts, prolonged viral RNA shedding has been variously associated with male sex [7] , older age [8, 9] , disease severity [7, [9] [10] [11] [12] , delayed hospital admission/therapy start after symptoms onset [7, 13, 14] and comorbidities [14] .

Recently, several studies have questioned the correlation between the duration of viral RNA shedding and the duration of infectivity, showing that live virus can be cultured from respiratory samples for a significant shorter time compared to detection of viral RNA by molecular methods, with a probability lower than 5% after 15 days from symptoms onset [4, 15] . Although these findings have led to update the criteria of deisolation in several international guidelines [16, 17] , strategies based on viral RNA detection are still widely used to guide decisions on infection prevention. Therefore, collection of further data on duration of viral RNA shedding and factors associated with prolonged shedding in larger population could be useful to improve clinical management of COVID-19 patients.

This study aimed to estimate the time to SARS-CoV-2 clearance from upper respiratory tract, to identify predictive factors of both viral clearance and prolonged viral RNA shedding and to explore associations J o u r n a l P r e -p r o o f between viral clearance and clinical outcomes in a large cohort of patients with COVID-19 admitted in an Italian Reference Hospital.

This is an observational, retrospective, single-centre study including all consecutive adult patients with a confirmed diagnosis of SARS-CoV-2 infection, made by detection of viral RNA from a respiratory specimen, who were admitted to the National Institute for Infectious Diseases Lazzaro Spallanzani IRCCS (Rome, Italy) between March 1 and July 1, 2020. Patients in whom the date of symptoms onset was not available and who did not have at least one available follow-up nasopharyngeal (NPS) or throat swab (TS) for SARS-CoV-2 after diagnosis were excluded from the analysis.

Epidemiological, demographic, clinical and laboratory data along with information on treatments and outcomes of all patients with a confirmed COVID-19 diagnosis were collected and recorded using a standardized electronic database (ReCOVery study).

The study was approved by our local Ethics Committee.

In all included patients, diagnosis of SARS-CoV-2 infection was confirmed by the detection of SARS-CoV-2 RNA through real time polymerase chain reaction (RT-PCR) targeting the E and RdRp viral genes on at least one respiratory specimen [18] . Respiratory tract specimens included were NPS, TS, sputum, bronchoaspirate (BAS) and bronchoalveolar lavage (BAL) fluid. Subsequently, during the hospitalization, all patients underwent to follow-up NPS and TS to assess the clearance of viral RNA. The frequency of follow-up NPS or TS, on average every 3 (interquartile range [IQR] 1-4) days, was not completely homogeneous among patients depending on both treating physician judgment and hospital internal protocol. Similarly, the therapeutic management of patients was decided on the basis of hospital internal protocol, guidelines and clinical judgment, according to the best evidence available at that moment.

Viral clearance was defined as negative RT-PCR for SARS-Cov-2 on two consecutive NPS or TS. We considered acceptable a negative viral RNA on a single NPS or TS if it was the last available sample.

Prolonged viral shedding was defined as the detection of SARS-Cov-2 RNA on respiratory specimens for more than 18 days (median duration of viral RNA shedding in our population). We considered as longest shedders those patients in whom the detection of SARS-Cov-2 RNA on respiratory specimens was equal or more than 26 days (75° percentile of viral RNA shedding in our population).

J o u r n a l P r e -p r o o f Furthermore, we defined clinical recovery as being discharged from hospital (including the transfer to a nonacute healthcare facility) or weaning off oxygen in patients who underwent to any kind of oxygen therapy during the hospitalization and who did not need it before the admission.

Finally, the severity of respiratory disease was defined: mild for a ratio of arterial oxygen partial pressure (PaO2) to fractional inspired oxygen (FiO2) between ≤ 300 mmHg and >200 mmHg; moderate for a PaO2/FiO2 ratio between ≤200 mmHg and >100 mmHg and severe for a PaO2/FiO2 ratio ≤100 mmHg [19] .

Continuous and categorical variables were expressed as medians with interquartile ranges and as numbers (%), respectively. The cumulative probability of achieving viral clearance was estimated by Kaplan-Meier curves in the overall population and after stratifying by the severity of respiratory disease developed during hospitalization.

In this latter analysis, the difference of SARS-CoV-2 clearance between different groups was compared by log-rank statistic test. Multivariable Poisson regression analysis was used to assess independent risk factors of viral clearance whereas multivariable logistic regression analysis was used to identify predictive factors of prolonged viral RNA shedding. In both the multivariable analysis, significant factors identified on univariate analysis and other a priori identified confounders (age, sex) were included as covariates in the multivariable models. Causal relation between the duration of viral RNA shedding and the probability of positive (clinical recovery) and negative (invasive mechanical ventilation or death) clinical outcomes was evaluated using inverse probability weighted (IPW) multivariable Cox model.

The following factors were a priori included as covariates in the model: gender, age, number of comorbidities, antiviral therapy (yes versus no), immunomodulatory therapy and PaO2/FiO2 ratio on hospital admission. In all the performed analysis, the date of symptoms onset was considered as the baseline date. All statistical analyses were performed using STATA (version 15.1, College station, Texas, USA). A p-value < 0.05 indicated conventional statistical significance.

Overall, 630 adult subjects were admitted to our centre with a confirmed COVID- 19 Figure 1c) . Interestingly, the probability of viral clearance did not differ after stratifying for less severe degree of respiratory failure (Figure 1b) . Association between viral RNA shedding and clinical outcomes

The potential causal relation between viral clearance achievement and the occurrence of the main clinical outcomes was explored by multivariable IPW Cox model ( Figure 2 

This retrospective study focused on the duration of viral RNA shedding from upper respiratory tract and on the factors associated with both viral clearance and prolonged viral shedding.

In our cohort, the median duration of viral shedding, from symptoms onset to viral clearance, was 18 days.

This finding is in line with the results of a recent systematic review and metanalysis which reported, from the analysis of 43 studies including 3229 subjects, a median length of viral detection from URT of 17 days [5] .

However, the median duration of SARS-CoV-2 shedding in respiratory tract varies widely among the different studies, ranging between 11 and 31 days [6, 7, 10, 13, 20] . This great variability could be partially explained by the heterogeneity, particularly regarding the clinical conditions, of the study populations and the different kind of specimens considered. Not only the median duration but also the maximum persistence of SARS-CoV-2 RNA differs significantly among the studies, with the longest described duration of viral shedding of 83 days, lasting even after symptoms resolution and seroconversion [6] . In this study, we found a maximum length of viral shedding of 73 days, which is one of the longest reported so far. Interestingly, in our cohort, patients with the longest duration of viral RNA shedding were mainly characterized by advanced age, underlying J o u r n a l P r e -p r o o f comorbidities, they were admitted to hospital more than 10 days after illness onset and with a severe clinical presentation of the disease.

In the current study, severity of illness, delayed admission to hospital after symptoms onset, presence of underlying comorbidities and alteration of inflammatory index at hospital admission were identified as independent risk factors of delayed viral clearance and/or prolonged viral shedding.

The association between severity of COVID-19 disease and duration of viral RNA shedding has already been widely reported [7, 9, [10] [11] [12] . Indeed, although several studies have described similar [2, 5, 21] or even longer [5] duration of viral RNA shedding in URT in mild compared to severe cases, the majority of reports found that patients with more severe illness shed viral RNA for a longer period of time [5, 7, 9, [10] [11] [12] 22] . It is worth noting that the definition of severity among these studies is not homogeneous. Particularly, most of them evaluated the differences in viral shedding duration between mild and severe cases [9, 11, 12] . Conversely, Zhou et al.

compared severe versus critical cases [10] and in the study of Xu et al. invasive mechanical ventilation was used as a proxy of illness severity [7] . In our study we used PaO2/FiO2 ratio to define disease severity and, particularly, we found an association between the development of PaO2/FiO2 ratio below or equals to 200, corresponding to a moderate to severe respiratory failure [19] , and a delayed viral clearance/prolonged viral shedding. Interestingly, in a recent study, the correlation between disease severity and longer persistence of SARS-CoV-2 RNA in respiratory tract samples has also been confirmed using virus cultures, which are considered a more reliable index of infectivity compared to molecular methods [15] . The association between illness severity and prolonged viral shedding, which has been previously described also for MERS and influenza virus [23, 24] , might be partially explained by the higher viral loads reported in severe COVID-19 cases compared to mild ones [3, 11, 12] .

Noteworthy, in the current study, the relationship between viral RNA shedding duration and clinical course of COVID-19 appeared to be bidirectional. Indeed, consistently with previous data from other coronavirus and influenza virus infections [24, 25] , we found that clearance of viral RNA from URT almost doubled the odd of clinical recovery and reduced the risk of worse clinical outcomes. The ongoing pro-inflammatory stimulus due to the persistence of SARS-CoV-2 in the respiratory tract and the resulting organ damage may contribute to explain this finding.

Although comorbidities have been identified as one of the main prognostic factors for COVID-19 severity, only few studies have reported an association with duration of viral RNA shedding [14, 26] . Particularly, several conditions as coronary heart disease [14] , hypertension [26] and diabetes [26] have been linked to a longer persistence of SARS-CoV-2 RNA. In our study, we did not find an association between specific comorbidities and persistence of viral RNA but rather we observed that patients with underlying comorbidities were more J o u r n a l P r e -p r o o f likely to have both slower viral clearance and prolonged viral detection with an increased risk for each additional comorbidity.

In our cohort, we found that a deferred admission to hospital after the onset of symptoms hiked the probability of prolonged viral shedding with a risk higher than 15% for each day more. This finding, already reported in previous studies [7, 13, 27, 28] , may suggest a potential effect of antiviral treatments on SARS-CoV-2 clearance, as also demonstrated for influenza A infection [24] . This hypothesis is also supported by the viral dynamics of SARS-CoV-2 characterized by a peak in viral load at the time or shortly before symptoms onset.

However, contrary to influenza A infection, none of the antiviral regimens considered in this study have shown clear advantages in terms of duration of viral shedding so far [5] . Additionally, in our cohort, the use of antiviral therapy was not associated with either the probability of viral clearance or prolonged viral RNA

shedding. An alternative explanation for this result is that the application of general supportive measures alone might have reduced the length of viral RNA shedding. Anyhow, the association between the time from illness onset to hospitalization and the duration of SARS-CoV-2 shedding is a key finding of this study which highlights the importance of a medical strategy based on rapid testing of symptomatic patients and prompt hospitalization and management of the infected ones.

Concerning non-antiviral therapies, the role of corticosteroids in the management of COVID-19 is still a matter of debate. Indeed, although data from recent randomised clinical trials have shown a potential beneficial effect of steroid therapy on survival in severe/critically-ill patients [29, 30] , their impact on viral shedding remain more uncertain, with some evidences indicating a possible association with a delayed viral clearance [31] . In our cohort, we did not observe a clear association between the use of corticosteroids and a prolonged viral shedding, as previously described in both SARS-CoV-2 and other coronavirus infections [27, 32] .

However, further evidences from ongoing clinical trials might better clarify the role of corticosteroids in COVID-19 treatment.

Finally, alteration of pro-inflammatory markers, and particularly lymphopenia and elevated D-dimer, at the time of hospital admission was found to be an independent risk factor of prolonged viral RNA shedding. Given the role of the immune-response in the clinical course of COVID-19, these laboratory index have been already identified as markers of disease severity [10] . However, despite alteration of inflammatory index has been found to be more frequent in patients with prolonged viral shedding in a recent study [13, 33] , an independent association with viral RNA shedding was not previously observed. Further studies are warranted to confirm this result.

Our study is not without limitations. First, due to its observational nature, this study could be prone to bias related to unmeasured confounders. Second, the retrospective collection of medical records may have introduced bias due to the potential inaccurate reporting of data. Third, the estimated duration of viral RNA shedding could have been influenced by the heterogeneity in the frequency of specimen's collection and in the type of respiratory specimen used. Additionally, the lack of quantitative determination of viral load, even thorough an indirect measure as cycle-threshold value, did not allow to draw any conclusion about the potential infectiousness of long-term shedders. Finally, the monocentric nature of the study may limit the generalizability of our findings. Nonetheless, this study has its main strength in the large sample size. Indeed, to the best of our knowledge, this is one of the largest cohorts in which the duration of SARS-CoV-2 RNA

shedding have been investigated. Thus, this study provides significant information on the dynamics of viral RNA shedding and the determinants of delayed viral clearance.

In conclusion, in this large retrospective cohort study, we found that severity of respiratory disease, delayed admission to hospital after the onset of symptoms, presence of underlying comorbidities and alteration of pro-inflammatory markers at the time of hospitalization independently predicted a longer persistence of SARS-CoV-RNA in respiratory tract. Furthermore, our data showed that a more rapid achievement of viral clearance was independently associated to better clinical outcomes and survival. These findings suggest that a medical strategy based on rapid testing of symptomatic subjects and a prompt hospitalization and therapy initiation of the ones with confirmed diagnosis, especially in presence of signs of severity and concomitant medical conditions, could be useful to achieve a more rapid viral clearance and limit worse clinical outcomes.

Moreover, infection control and prevention criteria might consider the clinical course of disease, addressing test-based de-isolation strategies or longer isolation periods for patients with severe/critical COVID-19. 

Abbreviations: aHR, adjusted hazard ratio; CI, confidence interval; IMV, invasive mechanical ventilation * after adjusting for gender, age, number of comorbidities, use of antiviral therapy (yes versus no), use of cytokines inhibitors (yes versus no), use of steroids (yes versus no), PaO2/FiO2 ratio (≤200 mmHg versus >200) during hospitalization (time updated variable). 

Temporal profiles of viral load in posterior oropharyngeal saliva samples and serum antibody responses during infection by SARS-CoV-2: an observational cohort study Lancet Infect

Temporal dynamics in viral shedding and transmissibility of COVID-19

Virological assessment of hospitalized patients with COVID-2019

SARS-CoV-2, SARS-CoV, and MERS-CoV viral load dynamics, duration of viral shedding, and infectiousness: a systematic review and meta-analysis. The Lancet Microbe

Prolonged SARS-CoV-2 RNA shedding: Not a rare phenomenon

Factors Associated With Prolonged Viral RNA Shedding in Patients with Coronavirus Disease 2019 (COVID-19)

Differences of Severe Acute Respiratory Syndrome Coronavirus 2 Shedding Duration in Sputum and Nasopharyngeal Swab Specimens Among Adult Inpatients With Coronavirus Disease

Duration for carrying SARS-CoV-2 in COVID-19 patients

Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study Lancet

Viral dynamics in mild and severe cases of COVID-19

Viral load dynamics and disease severity in patients infected with SARS-CoV-2 in Zhejiang province, China

Factors associated with the duration of viral shedding in adults with COVID-19 outside of Wuhan, China: a retrospective cohort study

Risk Factors for Viral RNA Shedding in COVID-19 Patients

Duration and key determinants of infectious virus shedding in hospitalized patients with coronavirus disease-2019 (COVID-19)

PMID: 33431879; PMCID: PMC7801729. J o u r n a l P r e -p r o o f 16. WHO. Criteria for releasing COVID-19 patients from isolation

Duration of Isolation and Precautions for Adults with COVID-19

Detection of 2019 novel coronavirus (2019-nCoV) by real-time RT-PCR

Acute respiratory distress syndrome: the Berlin Definition

The duration of viral shedding of discharged patients with severe COVID-19

Upper Respiratory Tract Levels of Severe Acute Respiratory Syndrome Coronavirus 2 RNA and Duration of Viral RNA Shedding Do Not Differ Between Patients With Mild and Severe/Critical Coronavirus Disease

Understanding viral shedding of severe acute respiratory coronavirus virus 2 (SARS-CoV-2): Review of current literature

Epub ahead of print

Viral load kinetics of MERS coronavirus infection

Factors Associated with Prolonged Viral Shedding in Patients With Avian Influenza A(H7N9) Virus Infection

Critically Ill Patients With the Middle East Respiratory Syndrome: A Multicenter Retrospective Cohort Study

Profile of RT-PCR for SARS-CoV-2: a preliminary study from 56 COVID-19 patients

Duration of SARS-CoV-2 RNA shedding and factors associated with prolonged viral shedding in patients with COVID-19

Associations of clinical characteristics and antiviral drugs with viral RNA clearance in patients with COVID-19 in Guangzhou, China: a retrospective cohort study

The WHO Rapid Evidence Appraisal for COVID-19 Therapies (REACT) Working Group. Association between administration of systemic corticosteroids and mortality among critically ill patients with COVID-19: a meta-analysis

Dexamethasone in Hospitalized Patients with Covid-19 -Preliminary Report

Corticosteroid use in COVID-19 patients: a systematic review and meta-analysis on clinical outcomes. Crit Care

Corticosteroid therapy for critically ill patients with Middle East respiratory syndrome

Clinical characteristics and predictors of the duration of SARS-CoV-2 viral shedding in 140 healthcare workers

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.Author Contributions: AM was a major contributor in writing the manuscript; PL analysed data; CC, RG, EL, AC, MBV, FT, SC, LL, FDG, GDO, FP, EN, CA, NP, GI, FV, EG, MRC have participated in the research and in the acquisition of data; AA made substantial contributions to the conception of the work and to interpretation of data.All authors read and approved the final manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest FUNDING: this work was supported by Line one -Ricerca Corrente 'Infezioni Emergenti e Riemergenti' and by Progetto COVID 2020 12371675 both funded by Italian Ministry of Health