The SARS-CoV-2 RNA with mild lung lesions lasts longer in non-severe COVID-19 patients: a case series study

Background COVID-19 has become a new infectious disease in the global pandemic, most of which are non-severe patients. It is particularly important to understand the dynamic changes of the whole disease course of non-severe patients from the onset to the follow-up after discharge. Methods On February 1, 2020, 18 cases of non-severe COVID-19 appeared in a hospital in Beijing. We recorded the clinical information and viral dynamics of these patients from the onset of the disease to one month after discharge. Results Eighteen patients (median age 43) were included, including 14 females. Fever (11/18) and cough (8/18) are the most common symptoms. According to the degree of lung inammation, 18 patients were divided into two groups (group A imaging score ≤ 10; group B imaging score > 10). The duration of SARS-CoV-2 positive in group A was signicantly longer than that in group B (the median was 30 and 13, respectively, P = 0.0025). One month after discharge, almost all patients were followed up for IgM antibody disappearance and IgG antibody production. Conclusion In non-severe COVID-19 patients, the positive duration of the SARS-CoV-2 in patients with mild lung injury was longer than that in patients with severe lung injury. The possible mechanism is that the virus-mediated immune system is not fully activated in mild damaged patients.


Introduction
Since December 2019, unexplained pneumonia of unknown origin occurred in Wuhan, China, which was quickly identi ed and named COVID-19 (caused by SARS-CoV-2), and then quickly spread all over the world [1]. As of April 24, 2020, more than 2 million people have been infected and nearly 2 hundred thousand people have died in the world [2].
It has been almost half a year since the onset of the disease, COVID-19 's global control situation is not very satisfactory [3]. However, the source of the virus is still unknown. COVID-19 has a variety of ways of transmission, including respiratory transmission, digestive tract transmission and so on [1,4,5]. The crowd is generally susceptible. According to the de nition of COVID-19 's diagnosis and treatment guidelines, whether in China or other countries and regions of the world, the vast majority of COVID-19 's patients are non-severe patients(Meet any of the following can be identi ed as severe or critical case: nger oxygen saturation ≤ 93%; PaO 2 /Fi0 2 ≤ 300 mmHg; respiratory failure and mechanical ventilation was needed; shock; admitted to ICU, etc.) [6]. The case fatality rate is often low in non-critically ill patients, but it is particularly important to note that these people play an important role in disease control and reducing human-tohuman transmission. According to Sutton D et al. research [7], the proportion of asymptomatic infections is as high as 88%. Therefore, a detailed understanding of the clinical course changes and viral dynamics changes of these mild patients from onset to discharge will help us to understand the development process of COVID-19, which will help us to control the spread and pandemic of COVID-19.
Unfortunately, there are no reports on the detailed course of disease of non-severe COVID-19 patients from the initial stage of illness to 1 month after being cured and discharged from hospital. Therefore, this case series records and analyzes the clinical and follow-up records of 18 COVID-19 patients from the onset of the disease to one month after discharge from a hospital in Beijing, China.

Patients
On February 1, 2020, a SARS-CoV-2 positive case occurred in a hospital in Beijing, followed by an epidemiological investigation by the Beijing Center for Disease Control and Prevention to isolate close contacts. Eighteen con rmed cases have been admitted to Beijing Hui Hospital. This is a case series report that collected 18 COVID-19 patients with mild illness from this hospital (according to China's policy at that time, the rest of the severe patients were transferred to other designated hospitals for centralized treatment).
This study has been approved by the Ethics Committee of Peking University First Hospital (2020-032) and has been registered in Chinese clinical trial registry (ChiCTR2000030096). All patients signed the informed consent form.

Data collection and laboratory examination
After admission, respiratory, serum and fecal samples were collected as far as possible, and the content of SARS-CoV-2 RNA was determined by reverse transcription PCR (RT-PCR). In order to avoid being mistakenly taken as saliva, sputum samples were taken from the respiratory tract samples of the patient after deep cough. What's more, to avoid cross-infection between patients, all patients wear N95 masks in separate wards (especially when taking throat swabs or sputum samples).
RT-PCR was performed with primers and probes for ORF1ab and N genes and a positive reference gene. This result is considered to be positive only when the cycle threshold (Ct) value of the reference gene is < 38 [8][9][10].
Chest imaging score: the main abnormalities of lung parenchyma were ground glass shadow and consolidation. According to the degree of lesion in chest imaging, the score ranges from 0 to 72, where 0 is normal [11][12][13]. According to the severity of pulmonary lesions, the patients were divided into two groups with a cut-off value of 10 points. Two imaging experts were blindly graded, and if the scores were different, the third expert would evaluate and discuss them.
In addition, we also followed up the results of SARS-CoV-2 RNA and antibodies (IgM and IgG) reexamined in the outpatient clinic one month after discharge.

Statistical analysis
Quantitative data and classi ed data are represented by median (IQR) and numbers (percentage), respectively. We used the Mann-Whitney U test or Fisher's exact test to compare differences between patients who mild pulmonary lesion and serious pulmonary lesion. We used Spearman's correlation to assess the relation between viral Ct value and other parameters. Statistical analyses were performed using SPSS ver. 25.0 (SPSS, Chicago, IL, USA). A P value of < 0.05 was considered signi cant. Table 1 shows the clinical characteristics of 18 patients with con rmed non-severe COVID-19: 11 with mild pulmonary lesion (imaging score ≤ 10) and 7 with severe pulmonary lesion (imaging score > 10). The median age of 18 patients enrolled in this study was 43 years old (ranging from 27 to 60 years old), including 14 females. Fever and cough were the most common clinical symptoms, accounting for 61.1% (11/18) and 44.4% (8/18), respectively. Most of the patients were in good health and had no chronic diseases, except that #10 and #15 had high blood pressure and #11 had diabetes. Most patients (13/18) were treated with interferon and thymosin.   In the laboratory examination, the white blood cell (WBC) count was lower than the lower limit of the normal value (reference 3.5-9.5 × 10 9 /L) in 5 cases, nobody is above the normal value. All the patients' lymphocyte count was lower than the normal value (reference 1.1-3.2 × 10 9 /L), but the lymphocyte percentage (reference 20%-50%) was lower than the normal value in only 1 case and higher than the normal value in 2 cases. C-reactive protein (reference < 8 mg/L) was higher than normal in 5 patients. During hospitalization, all patients underwent at least 3 chest X-ray or CT examinations. The median score of the rst image was 4(range 0-24). The rst imaging score of most patients (15/18) was the highest, and only 3 patients (#12, #17 and #18) showed progress.

Relationship between the degree of pulmonary lesion and the duration of virus
As shown in Fig. 1, patients #1 to #11 are patients with mild lung lesions (imaging score ≤ 10, named group A), and patients #12 to #18 are patients with severe lung lesions (imaging score > 10, named group B). The SARS-CoV-2 RNA positive duration of group B was signi cantly longer than group A, and the difference was statistically signi cant (P = 0.0025, Fig. 2A). We noted that all patients in group A were positive for more than 14 days, of which 6 patients were positive for more than 1 month. In contrast, only 3 people in group B lasted more than 14 days, and none of them lasted more than 1 month (Fig. 1A).
Then we discussed the difference of virus duration in different samples. As shown in Figs. 1B and Fig. 2B, there was no signi cant difference in the positive time of throat swabs between the two groups (P = 0.2010). In most sputum samples, the positive time of group A was longer than that of group B, but the difference was not statistically signi cant (Fig. 1C and Fig. 2C, P = 0.6376). It is worth noting that no one in group B was positive, 4 in group A were positive, and 3 were positive for more than 14 days, but there was no signi cant difference between the two groups, in stool sample ( Fig. 1D and Fig. 2D, P = 0.0923).

Correlation between viral load and other indicators
In order to explore the potential relationship between virus Ct value and other indexes, the correlations between ORF1ab gene and N gene and other indexes were analyzed (Fig. 4, Figure S1, Figure S2, Figure S3). We know that the lower the Ct value, the higher the viral load. Figure 4D shows that there is a signi cant positive correlation between the patient's maximum viral load and the percentage of peripheral blood lymphocytes (R = 0.49, P = 0.039). In addition, there was also a good correlation between the rst viral load and the percentage of peripheral blood lymphocytes ( Figure S1D, R = 0.46 P = 0.053). We also analyzed age, WBC, HBG, PLT, etc., and there was no signi cant correlation ( Figure S1, Figure S2, Figure S3).

Follow up results one month after discharge
In order to prevent the virus from returning to positive after discharge, all patients were followed up for one month (Table S1). Most of the patients were followed up twice, and only 4 patients were followed up once. Throat swabs were collected from all patients and all patients were negative. In addition, in order to observe the production of antibodies in patients, both IgM and IgG antibodies were detected. IgM antibodies were positive in patients 10, 12 and 17, but then turned negative. All patients were positive for IgG antibody at the rst reexamination. In the second reexamination, only 2 patients (patients No. 2 and 5) changed from positive to negative. Pulmonary CT also suggested that the lesions of all patients were completely or obviously absorbed. (information not shown)

Discussion
We recorded and followed up the complete pathogenesis of 18 patients with non-severe COVID-19 infection in a hospital in Beijing, and recorded the dynamic changes of a total of 235 virus results. Although they develop the disease together, their clinical processes are quite different. It is particularly noteworthy that in these non-severe patients, the negative conversion of the virus with mild lung lesions is very slow. This plays an important role in preventing and controlling the spread of COVID-19, that is, patients with mild symptoms may be quarantined for longer. The possible mechanisms are shown in Fig. 5 When the virus invades the human body, it mainly binds to the ACE2 receptor on type II alveolar epithelial cells, thus causing damage to the structural integrity of the alveoli. At the same time, the body's innate immune system is activated. At present, the known PRRs mainly includes Toll-like receptor (TLR), Rig-I-like receptor (RLR), Nod-like receptor (NLR), C-type lectin-like receptor and other ways [14][15][16][17][18]. We will focus on TLR, RLR, and RLR.
The activation of TLR occurs mainly in antigen presenting cells, such as dendritic cells (DC), macrophages, monocytes and B cells [19]. After recognizing the viral components, TLR recruits signal transduction molecules containing Toll/IL-1 receptor (TIR), such as myeloid differentiation primary response protein 88 (MyD88) and interferon-β TIR domain junction protein (TRIF), and then MyD88 and TRIF stimulate MAPK and NF-κB pathways, thereby enhancing the production of IFN and pro-in ammatory factors [16,19].
RLR is a family of cytoplasmic receptors, composed of three members, RIG-I, MDA5 (melanoma differentiation associated factor 5), and LGP2 (laboratory of genetics and physiology 2) [20][21][22]. RIG-I recognizes the 5'-triphosphate part of the viral genomic RNA and the double-stranded structure formed by selfannealing at the complementary ends of the viral genome. On the contrary, MDA5 usually detects longer dsRNA sequences. The combination of RIG-1 and MDA5 with viral RNA results in conformational changes, which exposes the CARD domain. Next, MAVS, an antiviral signal transducer located in mitochondria and peroxisome, was recruited to trigger the expression of IFN and proin ammatory cytokines [22].
NLR is another large family of PRR [23,24]. LRR motif detects viral PAMP, to induce structural rearrangement. Subsequently, a variety of signal pathways, including MAPK and NF-κB signal pathways, are activated. At the same time, the assembly of in ammatory bodies is also mediated by members of the NLR family. In ammatory bodies activate in ammatory-related proteases and induce the cleavage of IL-1 β and IL-18 precursors into active forms. The infection of coronavirus is related to the inhibition of IFN synthesis. The ability of virus to regulate type I IFN signal transduction is an important indicator of virus virulence [24].
In the early stage of virus infection, it is mainly innate immunity [25]. This includes the barrier effect of virus invasion site, phagocytosis of phagocytes, antimicrobial substances in body uids (such as complement and lysozyme), interferon and NK cells [26][27][28]. Among them, IFN and NK cells play a more important role. It plays a major immune role in preventing virus invasion, killing and eliminating virus, and terminating infection. Adaptive immunity appears late, including cellular immunity and humoral immunity. Viruses are strictly intracellular parasitic acellular microorganisms. This determines that cellular immunity plays a leading role in the process of eliminating virus infection, but antibodies can neutralize the virus outside the cell, making the virus lose its adhesion and infectivity to host cells. It also plays an important role in the process of anti-infection and spread [27,29].
Interferon cannot directly inactivate the virus, but through binding to the interferon receptor on the surface of the host cell to induce the synthesis of a variety of antiviral proteins, so as to achieve the inhibition of the virus [27]. The main functions of antiviral proteins include degrading viral mRNA, blocking viral transcription and translation, inhibiting viral protein synthesis, terminating viral replication and so on. Virus-infected cells produce and release interferon at the same time as the virus replicates, and quickly induces neighboring cells to produce interferon [27,28]. Therefore, interferon can not only limit the proliferation of the virus in infected cells, but also limit the spread of the virus between cells [27]. IFN-α and IFN-β can not only activate macrophages and NK cells, but also promote the expression of MHC class I antigens in virus infected cells, which is bene cial for CTL to play a lethal role [30]. IFN-γ not only has antiviral effect, but also induces antigen presenting cells to express MHC II antigens, enhances the recognition process of speci c immunity, enhances the killing function of NK cells, macrophages and CTL, and promotes the transformation of Th0 cells into Th1 cells [25,30].
NK cells can kill virus-infected cells directly and non-speci cally without antigen pre-sensitization. The killing effect of NK cells appears quickly, so it plays an important role in the process of immune surveillance and early anti-infective immunity. In addition, after NK cells are activated, they can also release cytokines such as IFN-γ and TNF, and regulate the immune function of the body. [31,32] Adaptive immunity plays an important role in anti-viral infection. CTL is a key adaptive immune effector cell, and its killing effect is limited by MHC class I molecules [33,34]. It kills virus-infected cells by releasing perforin and granzyme or inducing apoptosis of target cells. After killing one target cell, CTL can continue to kill other target cells without damage itself, which re ects the high e ciency of anti-virus in the body. The maturation and activation of CTL need the assistance of CD4 + T cells. IL-2 and IFN-γ can activate and enhance the killing function of CTL. When the infected cells were dissolved by CTL and released virus particles, the neutralization of antibody and the phagocytosis of phagocytes were initiated immediately, and the blocking effect of neutralization antibody and the further ampli cation of phagocytosis and killing effect of conditioning antibody on effector cells were further ampli ed, and the virus particles could be eliminated nally [35,36]. Th cells can exert antiviral effect by secreting cytokines. In addition, humoral immunity also plays an important role in anti-virus and preventing the spread of virus infection [33][34][35][36].
As explained above, innate immunity and adaptive immunity play an important role in ghting viral infection, eliminating the SARS-CoV-2 in different ways.
This also directly or indirectly explains that in non-severe COVID-19 patients, patients with mild or no lung injury, because the immune system has not been fully activated, the in ammatory response is mild, the immune system antivirus is slower, and the virus is carried for a long time.
This study also has some inevitable limitations. First of all, the number of cases included is relatively small, but we have made up for some defects in some aspects through detailed course records and follow-up. Secondly, as COVID-19 is an emerging infectious disease, and medical supplies are relatively scarce, we are unable to quantitatively detect the virus changes of patients every day.

Conclusion
In conclusion, in non-severe COVID-19 patients, the positive duration of the SARS-CoV-2 in patients with mild lung injury was longer than that in patients with severe lung injury. The possible mechanism is that the virus-mediated immune system is not fully activated in mild damaged patients. This has some value in preventing the spread of COVID-19. Availability of data and materials: All data generated or analyzed during this study are included in this published article (and its supplementary information les). The possible mechanism of anti-SARS-CoV-2 of immune system. We only discuss non-severe COVID-19 patients. (A) On the left panel (mild in ammation), the immune system is not fully activated and a small number of immune cells in ltrate, so the virus lasts longer. On the right panel (moderate in ammation), the immune system is fully activated, a large number of in ammatory cells in ltrate, and antiviral immunity is initiated. (B) SARS-CoV-2 invaded, it activated innate immunity and adaptive immunity. Among them, NK cells, dendritic cells and macrophage directly or indirectly inhibit virus replication, or inhibit virus by secreting cytokines. Under the action of antigen presenting cells, speci c immunity is activated, B cells play an antiviral role by producing antibodies, and T cells are also involved in antiviral immunity through direct killing or other ways. With the clearance of the virus, there is also in ammatory in ltration in the lungs of the human body.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.