Digestive Involvement in SARS-CoV-2 Infection: A Retrospective Multi-center Study

Background Coronavirus disease 2019 (COVID-19) is an emerged infection raised wide concerns for the pneumonia and respiratory manifestations. Also, digestive complications are frequently observed in COVID-19 patients but the significance remains undetermined. Methods A retrospective analysis of alimentary symptoms, liver dysfunctions and other clinical parameters of 514 hospitalized COVID-19 patients (282 mild, 162 severe and 70 critical cases) admitted to the 3 designated medical units of Wuhan Union Hospital from Jan 20 to Feb 29, 2020 was performed. Results

It is well established that most patients with COVID-19 have fever along with respiratory symptoms, such as dry cough and dyspnea [1,6]. However, there is increasing evidence that, in addition to typical respiratory symptom, COVID-19 can also present with gastrointestinal symptoms (especially for diarrhea) and liver injury [7,8]. Moreover, it was also reported that some patients present initially with gastrointestinal symptoms [9]. This may lead to misdiagnosis in the clinic, and the digestive involvement of COVID-19 should be recognized and valued.
SARS-CoV-2 is an accepted respiratory transmitting virus that spread via droplets [10], also through close contact with mucous membranes in the eyes, mouth or nose [11,12]. Apart from the oropharyngeal swab, nasopharyngeal swab and sputum, the viral nucleic acid could also be detected in saliva and feces [13,14]. Moreover, in clinical practice, we found stool samples test from  were positive while respiratory samples are negative. Importantly, angiotensin converting enzyme 2 (ACE2), identified as the cell entry receptor of SARS-CoV-2 [15,16], were most remarkable expression on the surface of small intestinal epithelial cell [15,17]. In summary, SARS-CoV-2 may be present in the gut, leading to the alimentary damage and symptoms, and possible fecal-oral transmission.
In this retrospective study, we describe the clinical features of hospitalized COVID-19 patients with variable severity, that is the mild, severe and critical patients. We focused on the alimentary symptoms and liver injury in COVID-19 patients and attempted to find if there is any significant difference in the patients with or without digestive manifestations, to provide evidence for the digestive involvement in the clinical prevention and management of SARS-CoV-2 infection.

Study Design and Participants
The protocol of this study was seen in Figure 1. The clinical records of 521 consecutive patients fulfilled with the diagnostic criteria of COVID-19 according to World Health Organization interim guidance [18], who admitted to the 3 designated COVID-19 care medical institutions affiliated to

Study Definitions
The confirmed case of COVID-19 is defined as positive SARS-CoV-2 nucleic acid in oropharyngeal or nasopharyngeal swab specimens via real-time fluorescence reverse transcription-polymerase chain reaction (RT-PCR). Only laboratory-confirmed cases were included in the analysis [18,19]. The mild cases refer to mild patient with fever, respiratory and/or other symptoms, and typical imaging findings of viral pneumonia. The severe cases had to meet one or more of the following conditions: 1) with shortness breath and respiratory frequency ≥30 per minute; 2) blood oxygen saturation ≤93% in resting condition; 3) arterial partial pressure of oxygen (PaO 2 )/ fraction of inspired oxygen (FiO 2 ) Particularly, symptoms of fever, cough, expectoration, fatigue, shortness of breath, chest tightness, muscle soreness, poor appetite, nausea, vomit, diarrhea and abdominal pain were recorded in detail. immunomodulation and respiratory support) and outcomes data were further obtained. The collected data were reviewed and checked by 4 trained researchers and attending clinicians independently.

Statistical Analysis
Categorical variables were described as number (%) and compared with χ² test, or Fisher's exact test when the data were limited. Continuous data were expressed as median (IQR) and compared via the Kruskal-Wallis test followed by a Dunn's post-test. A 2-sided α 0.05 was considered statistically significant. All statistical analyses were performed using the R software, version 3.5.1 (R Foundation for Statistical Computing). The statistical graphs were created via the GraphPad Prism package, version 8 (GraphPad Software, LLC).

Baseline Characteristics
The demographic and clinical characteristics were shown in Table 1. A total of 514 hospitalized patients with confirmed SARS-CoV-2 infection including 282 mild, 162 severe and 70 critical cases were included in the present analysis. The media age was 54 years, ranged from 23 to 96 years, and older in the severe and critical suffers. Totally, 271 (53.1%) patients were male, and the proportion of male was much higher than female in the critical ones (70.0%). The common symptoms were fever (87.6%), cough (70.2%), fatigue (58.4%), chest tightness (52.0%), shortness of breath (44.2%), expectoration (29.8%), muscle soreness (19.8%) and pharyngodynia (2.9%). Compared with mild patients, more serious fever (92.6% and 90.0% respectively) with higher maximal temperature, chest tightness and shortness of breath were observed in the severe and critical patients. Hypertension (43.2%), diabetes (19.3%) and cardiovascular disease (12.7%) were the most common comorbidity.
Especially, hypertension was associated with disease severity that higher in severe (60.5%) and critical (55.7%) patients. The severe to critical patients exhibited more serious infection and inflammation with higher white blood cell and neutrophil, declined lymphocyte, enhanced levels of CRP, ESR, procalcitonin and ferritin (all P<0.001). Note: Continuous variables are presented as median (25th-75th percentiles), and categorical variables are presented as count (percentage).

Alimentary symptoms and liver function injury
A series of alimentary symptoms including poor appetite (50.2%), diarrhea (25.5%), nausea (16.3%), vomit (11.9%) and abdominal pain (3.3%) presented in COVID-19 patients ( Table 2). Diarrhea was quite common gastrointestinal symptom with higher morbidity in the severe and critical patients (32.1% and 27.1% respectively) than the mild patients (21.3%). The diarrhea was mainly watery in nature without mucus or blood, and abdominal pain was rarely observed. The severity of diarrhea varied among patients, most patients had a few loose or watery stools but some patients had up to 20 bowel movements per day. About 13.2% patients developed diarrhea in the first 3 days after the onset of symptoms, and 63.2% diarrhea was occurred in the previous 10 days ( Figure S1). There were 31 patients (6.03%) presented with diarrhea in the absence of respiratory symptoms, including 23 mild cases and 8 severe cases ( Figure 2). Moreover, 9 patients (1.75%) had only diarrhea without fever and respiratory symptoms. Liver injury was also frequently observed in patients with COVID-19 (Table 2), with increased ALT (43.3%), AST (36.7%) and decreased albumin (80.9%), but less increased total bilirubin (10.9%) and direct bilirubin (14.2%), which were more common in the severe to critical patients. It indicated a major damage of the hepatocytes and synthetic function, but not the function of bile secretion, in COVID-19 associated liver injury. However, most of the cases were mild injury. Those with relatively severe liver damage (ALT≥80 U/L) only accounted for 7.4% of the total patients, and 27.5% of the critical suffers. The liver injury was associated with higher inflammatory activity, such as increased white blood cell and neutrophil counts but declined lymphocyte counts, and elevated levels of CRP, ESR, procalcitonin and ferritin (Table S2).

Clinical characteristics of patients with and without diarrhea
There were 131 patients (25.49%, including 60 mild and 71 severe/critical cases) presented with diarrhea, and 383 patients (74.51%, including 222 mild and 161 severe/critical cases) without diarrhea ( Figure 2). In the mild cases, patients with diarrhea had only slightly increased neutrophil (P=0.021) and decreased lymphocyte (P=0.011), but had no significant difference in multiorgan functions, compared with those absent of diarrhea (Table 3). In the severe/critical cases, there was increased WBC (P=0.030) and neutrophil (P=0.010) and trend to decreased lymphocyte (P=0.082), as well as higher level of inflammatory ferritin in patients with diarrhea relative to those without diarrhea. Moreover, diarrhea evidently companied with the more serious liver injury with increased activities of serum ALT, AST and LDH, and raised levels of total and direct bilirubin (Table 3). Continuous variables are presented as median (25th-75th percentiles), and categorical variables are presented as count (percentage).

Clinical characteristics of diarrhea patients with and without respiratory symptoms
In the patients with diarrhea, 100 subjects (19.45%, including 37 mild and 63 severe/critical cases) presented with respiratory symptoms and 31 subjects (6.03%, including 23 mild and 8 severe cases) absent of respiratory symptoms (Table S1). As it shown, there were higher proportion of mild cases in patients with diarrhea only (74.2%) or with respiratory symptoms only (58.0%) than those with both diarrhea and respiratory symptoms (37.0%) ( Figure 2). Notably, the patients with diarrhea in the absence of respiratory symptoms were observed less systemic inflammatory activity, including the lower levels of CRP, ESR and Ferritin, relative to diarrhea combined with respiratory symptoms (Table   4).  Discussion COVID-19 is generally considered a lower respiratory disease via droplet transmission. However, the current study suggested that gut and liver involvement was common in SARS-CoV-2 infection, which may carry significant implications on the pathogenesis and clinical management of COVID-19.
Alimentary symptoms were frequently observed in patients with COVID-19. Diarrhea was the typical gastrointestinal complain that troubled 25.5% of the total COVID-19 patients. This rate was a little higher than early reported 3.8% to 22.2% in 2019-nCoV infection [2][3][4]20]. Similarly, SARS-CoV caused 20.3% diarrhea on presentation, and 38.4% patients developed symptoms of diarrhea during the course of illness [21], also middle east respiratory syndrome coronavirus (MERS-CoV) led to 26.0% diarrhea and 21.0% vomiting as for its enteric tropism [22]. About 13.2% patients developed diarrhea in the first 3 days after the onset of symptoms, and 63.2% diarrhea was occurred in the previous 10 days. Thus, diarrhea is more likely to develop in the early phase of infection, and the initial pathophysiology of COVID-19. All supported the gut involvement and diarrhea in the pathogenesis of novel SARS-CoV-2 infection.
Moreover, diarrhea was more common in the severe and critical patients with higher morbidity (32.1% and 27.1% respectively). This was in line with the observation that higher incidence of diarrhea in the severe or ICU cases than non-severe COVID-19 patients [3,4]. And in the severe/critical subgroup, COVID-19 patients with diarrhea were reported more apparent systemic inflammatory activity and liver injury compared with those without diarrhea. It further indicated that comorbidity of diarrhea may accompany with more serious illness in the severe/critical COVID-19 suffers.
However, the severity of diarrhea, mainly frequency of defecation, was not apparently associated with hypoxemia (oxygen requirement) and the severity based on respiratory dysfunction of these patients.
Therefore, the gastrointestinal tract and the lung may react independently and differently to SARS-CoV-2 infection. It may argue that diarrhea is related to the use of drugs in some cases, including mainly antibacterial drugs (moxifloxacin, cephalosporins and azithromycin) and antiviral drugs (abidol, oseltamivir, ribavirin and lopinavir/ritonavir). However, there no significant difference in the prehospital medications between the patients with and without diarrhea. Moreover, the use of antibiotics in the mild cases was rare, but the occurrence of diarrhea also very high (21.3%). In addition, lots of patients presented with diarrhea that had no prior medication history. So, the medication was inadequate to interpret the development of diarrhea, and other potential culprits for diarrhea such as viral load in the gut, intestinal dysbiosis and combined infections are necessary to future determine.
Interestingly, there were 31 patients (6.03%) presented with diarrhea in the absence of respiratory symptoms, and 9 patients (1.75%) of which had only diarrhea without fever and respiratory complaints. It further supported the view of "gut type" in the clinical management of COVID-19.
Generally, these patients were mild illness and normal CT findings in the lung, and revealed less systemic inflammatory activity relative to the patients with both diarrhea and respiratory symptoms.
It thereby may be easily misdiagnosed, especially for patients with diarrhea and gastrointestinal symptoms as the initial appearance. Identification of this part of patients with a gut-type may play a critical role in the prevention and management of the epidemic situation.
The mechanism underline intestinal injury and diarrhea in SARS-CoV-2 infection was undiscovered.
However, more and more evidences indicated the enteric tropism of SARS-CoV-2. It has been detected the viral nucleic acid in feces [14]. Swallowing sputum and saliva may be an important source of enteric virus [23,24]. It makes the fecal-oral transmission possible that might be a major transmission route of SARS-CoV-2 [14,23]. Most important, ACE2 is widely expressed on the surface of intestinal epithelial cells [15,16], which is the known entry target attacked by SARS-CoV-2 [16]. All above favored the direct damage of SARS-CoV-2 on the gut mucosa, leading to diarrhea and gastrointestinal symptoms. The intestinal colonization of SARS-CoV-2 and the viral load may be associated with the severity of diarrhea. However, the mechanism of viral colonization and histological changes of the gut on SARS-CoV-2 infection are necessary to revealed.
Apart from alimentary symptoms, the liver injury was also frequently observed in patients with COVID-19 [8]. Generally, mild liver dysfunction was common on SARS-CoV-2 infection. Moreover, the major abnormalities were hepatocyte damage with increased activities of ALT and AST, and declined synthetic function with low levels of pre-albumin and albumin, while the function of bile secretion was less involved. However, ACE2 is negative in hepatocytes and sinusoidal endothelium, but high expressed on the bile ducts [15,25]. Therefore, the liver injury was less likely to be caused by directly attack of SARS-CoV-2. The severity of liver injury and the proportion of serious liver injury were much higher in the severe to critical COVID-19 patients. As we shown, liver injury was associated with higher inflammatory activity, such as increased WBC and neutrophil counts but reduced lymphocyte counts, and elevated levels of CRP, ESR, procalcitonin and ferritin. SARS-CoV-2 activity induced systemic inflammation and immunoreaction may be one cause of the liver dysfunction. Other factors such as drugs and hypoxia injury are likely to participate in the process of hepatic damage under SARS-CoV-2 infection, but it lacks of sufficient evidences.
The present study has several limitations. Firstly, this was a retrospective cohort study, and unrecognized biases cannot be excluded. Secondly, medications may cause diarrhea in some patients although we try to rule out potential drug-induced diarrhea. This should be avoided in the future prospective researches. Finally, only a small part of patients had tested the viral nucleic acid in the stool specimens, and preliminary evidence indicated its correlations with diarrhea. Further research is vital to determine how the dynamic changes and load of fecal virus on the severity and duration of diarrhea, and potential fecal-oral transmission.

Conclusions
In summary, our data favored the digestive involvement in the process of COVID-19. The enteric tropism of SARS-CoV-2 should be the major culprit for diarrhea and other alimentary symptoms, also makes the fecal-oral transmission possible. There may be a "gut type" of COVID-19 that is significant for the understanding and clinical management of this emerging infection.

Ethics approval and consent to participate
The study was approved by the institutional review board prior to data collection. All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2008. Informed consent was waived due to retrospective study.

Author contributions
Haitao Shang collected medical records data, analyzed the data and drafted the manuscript; Chao Huang and Yuhu Chen helped for data collection; Shengyan Zhang helped for data analysis; Pengcheng Yang and Gaichao Hong supported for data entry and sorting; Lei Zhang and Xiaohua Hou designed and supervised the study and revised the manuscript.

Funding
This study was supported by grants of an emergency science and technology projects of novel coronavirus pneumonia from the Science and Technology Department of Hubei Province, China (2020FCA014), and the projects of international cooperation and exchanges from the National Natural Science Foundation of China (NSFC), China (No. 81720108006).

Availability of data and material
The raw data generated and analyzed in the current study are not publicly available due to appropriate protection of patient personal information but are available from the corresponding author on reasonable request.

Figure 1
The study protocol.

Figure 2
Diarrhea as the common gastrointestinal symptom in COVID-19 patients.

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download. Table S1.docx Table S2.docx