Persistent SARS-CoV-2 replication in severe COVID-19

Background: The diagnosis of SARS-CoV-2 infection is based on viral RNA detection by real-time RT-PCR (rRT-PCR) in respiratory samples. This detection can remain positive for weeks without implying virus viability. Methods: We have performed cell culture to assess viral replication in 106 respiratory samples rRT-PCR positive for SARS-CoV-2 from 105 patients with COVID-19. Fifty were samples from 50 patients with mild forms of COVID-19 who did not require hospital admission. Fifty-six samples were obtained from 55 hospitalized patients with severe pneumonia. Samples were obtained at different time points covering the time from clinical diagnosis to the follow up during hospital care. Results: In 49 samples (49/106, 46.2%) a cytopathic effect (CPE) was detected in cell culture. Our study demonstrates that while in patients with mild COVID-19, viral viability is maintained in fact up to 10 days in patients with severe COVID-19 the virus can remain viable for up to 32 days after the onset of symptoms. Patients with severe COVID-19 as compared with mild cases, presented infective virus in a significantly higher proportion in samples with moderate or low viral load (Ct value > 26): 22/46 (47.8%) versus 7/38 (18.4%), (p <0.01), respectively. Conclusions: Persistent SARS-CoV-2 replication could be demonstrated in severe COVID-19 cases for periods up to 32 days after the onset of symptoms and even at high Ct values. COVID-19 severity is a more determining factor for viral viability than the time elapsed since the onset of symptoms or the Ct value obtained in the RT-PCR assay.

3 Background COVID-19 is an acute respiratory tract infection caused by a new human coronavirus, SARS-CoV-2, that emerged in Wuhan, China, in late 2019 1,2 . On January 31st the first case of COVID-19 was detected in Spain, an imported case from Germany in Canary Islands, and thereafter on February 25th the first case was detected in Madrid 3 . The first case of COVID-19 was confirmed at the Hospital Universitario 12 de Octubre on March 1st, a large teaching hospital with 1200 beds, covering an area over 400,000 inhabitants in southern Madrid.
The main test for the diagnosis of COVID-19 is the detection by molecular diagnostic techniques (real-time RT-PCR [rRT-PCR]) of the SARS-CoV-2 virus in respiratory samples 4 . Detection of virus RNA is very sensitive to diagnose COVID-19 disease during the acute phase, since active replication appears to take place in the upper airway 5 . The kinetics of SARS-CoV-2 replication and detection by rRT-PCR in nasopharyngeal specimens has been well described in mild cases where viral replication is detectable up to the 8 th day postsymptoms 5,6 . Detection of viral RNA by rRT-PCR is more persistent and does not necessarily indicate viable virus forms 7 . The infection cycle of SARS-CoV-2 in severe cases of pneumonia requiring hospitalization and respiratory support is much less known. It has been suggested that the viability of the virus in respiratory samples is related to the time elapsed from the onset of symptoms and to the cycle threshold (Ct) value obtained in the amplification assay 5,8 . This value is used as a relative measure of quantification. In this study we have compared detection by rRT-PCR and the infectivity of SARS-CoV-2 in respiratory samples from mild COVID-19 cases to those from severe hospitalized cases of bilateral pneumonia.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 12, 2020.

Microbiological methods
Nasopharyngeal samples were collected with flocked swabs in UTM™ viral transport medium (Copan Diagnostics, Brescia, Italy). Bronchial aspirates were diluted 1:1 in UTM™ viral transport medium upon arrival at the laboratory.
Previously published rRT-PCR protocol to detect E gene 4,5 was adapted for processing on the automated molecular diagnostic platform Panther Fusion, using its open access functionality 9 . Five hundred µl of each sample was transferred to a specimen lysis tube and loaded directly onto the Panther Fusion System. RNA extraction and amplification reagents (Open Access RNA/DNA enzyme cartridges, Extraction Reagent-S and Internal Control-S) were provided by Hologic (San Diego, CA, USA). Primers and probe for E gene were provided by TIB MolBiol (Berlin, Germany). The Ct value obtained in this assay was used as a measure of relative quantification throughout the study.
For cell culture, an aliquot (250 µl) of the residual sample was decontaminated using gentamicin and amphotericin B, and inoculated in 24-well plates onto . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted June 12, 2020. . https://doi.org/10.1101/2020.06.10.20127837 doi: medRxiv preprint Vero E6 cells and cultured in Medium 199 supplemented with L-glutamine and 10% of fetal bovine serum. Plates were incubated in a CO2 5% atmosphere for 5 days. The development of CPE was examined daily. SARS-CoV-2 CPE specificity was confirmed by immunofluorescence (shell-vial technique) with a human serum with a high titer of anti-RBD IgG as primary antibody, and a FITC-labeled anti-human IgG as secondary antibody 10 . Additionally, upon CPE observation, culture supernatants were collected from each well, and rRT-PCR performed and confirmed to be positive at least 3 Ct lower than that of the original sample. All procedures related to cell culture were performed at a BLS3 facility.

Data analysis
Demographic data, COVID-19 severity, symptom time to test (STT), Ct values and CPE detection were recorded and analyzed. Quantitative variables were described using median and interquartile range (IQR) and compared by Mann-Whitney U test. Categorical variables were expressed by relative frequency and compared by Fisher's exact test. Statistics were performed on Graph Pad Prism V8 software. P-values <0.05 were considered to be statistically significant.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted June 12, 2020.  is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted June 12, 2020.

Correlation between virus viability and viral load
Samples with higher viral loads (Ct value < 25) in both groups of patients showed viable virus in a percentage higher than 90%. However, it is notable . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Discussion
The use of rRT-PCR as a diagnostic and follow-up method for SARS-CoV-2 infection has led to infer misleading information regarding the duration of infectivity of patients, the possibility of reactivation or even reinfection that has not been demonstrated so far 7 . Although rRT-PCR is the gold standard as a diagnostic method, it is less useful as a follow-up technique, since samples from patients who have overcome both mild and severe SARS-CoV-2 infections can retain detectable viral RNA for variable periods of time 5,9,13 Furthermore, the quantification of the virus is not well solved in respiratory samples, since their quality has a significant influence on the results.
The assessment of SARS-CoV-2 viability will allow to establish criteria for the isolation of the patients, evaluation of the efficacy of the treatments used, and even prognostic value.
In severe cases of COVID-19 a hyper-inflammatory process triggered by SARS-CoV-2, typically during the second week of disease, and the subsequent expression of high levels of cytokines is invoked to explain the severe pulmonary damage leading to respiratory failure. This has led to therapeutics recommendations on using antivirals as soon as possible and manage the severe consequences of an uncontrolled immune-driven response, during the second and third week, with anti-inflammatory therapy such as steroids or IL-6 receptor antagonist monoclonal antibodies 11 . The contribution of maintained viral replication to the pathogeny of lung tissue damage and respiratory failure, along to the antiviral countermeasures, should be further explored.
Here we show a high rate (55,3%) and persistence (STT up to day 32) of positive culture in samples from severe COVI-19 cases, indicating that replication is preserved in respiratory samples in a much higher proportion than previously reported.
Our study presents the highest rate of positive culture of SARS-CoV-2 of those published to date 5,8,12 . A recent study has shown prolonged viral shedding in patients with severe COVID-19 and this fact appears to be closely related to a high viral load and to a low neutralizing antibody response 12 . Although we have observed a significant positive correlation between viral load and the presence . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted June 12, 2020. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted June 12, 2020. . https://doi.org/10.1101/2020.06.10.20127837 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 12, 2020. . https://doi.org/10.1101/2020.06.10.20127837 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 12, 2020. . https://doi.org/10.1101/2020.06.10.20127837 doi: medRxiv preprint