Various liver and gastrointestinal involvements occur in patients with coronavirus disease 2019 (COVID-19) at variable prevalence. Most studies report mild liver function disturbances correlated with COVID-19 severity, though liver failure is unusual.

To study liver and gastrointestinal dysfunctions in Egyptian patients with COVID-19 and their relation to disease outcomes

This multicentre cohort study was conducted on 547 Egyptian patients from April 15, 2020 to July 29, 2020. Consecutive polymerase chain reaction-confirmed COVID-19 cases were included from four quarantine hospitals affiliated to the Egyptian ministry of health. Demographic information, laboratory characteristics, treatments, fibrosis-4 (FIB-4) index, COVID-19 severity, and outcomes were recorded and compared according to the degree of liver enzyme elevation and the presence of gastrointestinal symptoms. Follow-ups were conducted until discharge or death. Regression analyses were performed to determine the independent factors affecting mortality.

This study included 547 patients, of whom 53 (9.68%) died during hospitalization and 1 was discharged upon his request. Patients’ mean age was 45.04 ± 17.61 years, and 21.98% had severe or critical COVID-19. Alanine aminotransferase (ALT) and aspartate aminotransferase (AST) were available for 430 and 428 patients, respectively. In total, 26% and 32% of patients had elevated ALT and AST, respectively. Significant liver injury with ALT or AST elevation exceeding 3-fold was recorded in 21 (4.91%) and 16 (3.73%) patients, respectively. Male gender, smoking, hypertension, chronic hepatitis C, and lung involvement were associated with elevated AST or ALT. AST was elevated in 50% of patients over 60-years-old. FIB-4 was significantly higher in patients admitted to the intensive care unit (ICU), those with more severe COVID-19, and non-survivors. The independent variables affecting outcome were supplementary vitamin C intake (1 g daily capsules) [odds ratio (OR): 0.05, 95% confidence interval (CI): 0.008–0.337]; lung consolidation (OR: 4.540, 95%CI: 1.155–17.840); ICU admission (OR: 25.032, 95%CI: 7.110–88.128); and FIB-4 score > 3.25 (OR: 10.393, 95%CI: 2.459-43.925). Among 60 (13.98%) patients with gastrointestinal symptoms, 52 (86.67%) had diarrhoea. Patients with gastrointestinal symptoms were predominantly females with higher body mass index, and 50 (83.40%) patients had non-severe COVID-19.

Few Egyptian patients with COVID-19 developed a significant liver injury. The independent variables affecting mortality were supplementary vitamin C intake, lung consolidation, ICU admission, and FIB-4 score.

Hepatic and gastrointestinal (GI) involvement occurs in 16% to 78% of patients infected with the novel coronavirus, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)[1-3]. In some of these populations, liver enzyme elevation exceeds five times the upper limit of normal (ULN)[4]. Most studies report mild liver function disturbances correlated with coronavirus disease 2019 (COVID-19) severity, though liver failure is unusual[5]. Hepatic and GI involvement can occur with and without pulmonary manifestations of COVID-19[1]. Different underlying mechanisms may contribute to liver injury related to COVID-19, including a direct viral effect, as the virus enters through the angiotensin-converting enzyme-2 receptor on cholangiocytes. Other indirect pathways include sepsis, drug-related hepatic injury, uncontrolled immune reactions, and cytokine storm[6].

Diarrhoea is the most frequently recorded GI symptom reported with COVID-19, ranging from 3% to 30%, followed by anorexia, nausea, vomiting, and abdominal pain[7]. SARS-CoV-2 RNA has been detected in stool specimens from patients with polymerase chain reaction-confirmed COVID-19 in respiratory samples. Those patients showed extended viral shedding in stool up to 11.2 d after viral eradication from respiratory specimens, suggesting faecal-oral transmission and warranting subsequent precautions[8]. The possible underlying mechanisms for GI involvement are the direct viral invasion of cells in the GI tract via the angiotensin-converting enzyme-2 receptor on gastric and duodenal glandular cells and proximal and distal enterocytes[8]. This viral invasion disrupts absorption and intestinal secretions, which activates the enteric nervous system and causes diarrhoea[9]. Other indirect mechanisms include antibiotic-associated diarrhoea, indirect inflammatory damage, and the ‘‘gut-lung axis” theory of immune-mediated effects on the respiratory tract by disturbed digestive tract flora[1,10]. This study aimed to determine the prevalence and extent of liver and GI derangements in Egyptian patients with COVID-19 infection and their relation to COVID-19 disease outcomes.

This study included 547 COVID-19 patients hospitalized in four quarantine hospitals affiliated to the Egyptian Ministry of Health and Population (see SupplementaryTables 1 and 2 for baseline clinical, laboratory, imaging, treatment, and outcome features of the studied cohort). Of the 547 patients, 53 (9.68%) died during hospitalization, 493 (90.46%) recovered and were discharged, and 1 was discharged upon his request. The most common symptoms were fever (54.66%), cough (52.47%), dyspnoea (32.54%), and fatigue (30.71%), and 118 (21.57%) patients were asymptomatic. The mean age of the studied patients was 45.04 ± 17.61 years (the median age interquartile range) was 45.00 (30.00:60.00) years and 300 (54.84%) were male. Diabetes was the most common comorbidity (24.86%), followed by hypertension (23.77%) and coronary artery disease (4.20%). Chronic liver disease was reported in 18 (3.29%) patients: 14 had chronic hepatitis C (of whom 7 had liver cirrhosis), 2 had chronic hepatitis B, and 2 had fatty liver disease. We confirmed their liver condition using their most recent abdominal ultrasound study before hospitalization for COVID-19. Regarding COVID-19, 427 (78.02%) had mild and moderate disease, 120 (21.98%) had severe or critical disease, and 122 (22.34%) were admitted to the ICU during their hospitalization. Flow chart of the study cohort is illustrated in Figure 1.

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Flow chart of the study cohort. COVID-19: Coronavirus disease 2019; ALT: Alanine aminotransferase.

In our study, ALT and AST were available for 430 and 428 patients, respectively, and their other liver function tests and inflammatory markers were also available. Most patients had normal AST (291; 67.99%) and normal ALT levels (318; 74%). We divided patients as follows: Those with liver enzyme levels within the normal range, 1–2 fold over the ULN, 2–3 fold over the ULN, or >3 fold over the ULN (Tables 1-4). Among patients who required ICU admission, 48.50% had elevated AST and 35.60% had elevated ALT. On admission, FIB-4 was significantly higher in patients admitted to the ICU, those with more severe COVID-19 disease, and non-survivors (SupplementaryTable 3).

We found that male gender, cigarette smoking, systemic hypertension, chronic hepatitis C, and abnormal chest CT findings were associated with significant elevation of baseline AST and ALT. Also, AST levels were elevated in patients older than 60 years, and diabetes mellitus was associated with elevation of baseline ALT. Elevated AST was also associated with elevated white blood cell count, absolute neutrophils count, neutrophils/lymphocytes ratio, ALT, gamma-glutamyl transferase (GGT), D-dimer, ferritin, and CRP and negatively correlated with absolute lymphocyte count and serum albumin. Elevated ALT was associated with elevated white blood cell count, Absolute Neutrophils count, neutrophils/lymphocytes ratio, AST, alkaline phosphatase (ALP), GGT, D-dimer, ferritin, and CRP and negatively correlated with absolute lymphocyte count and serum albumin (Tables 1-4).

No patient under 18-years-old showed elevated ALT levels, and only 1 had less than a 2-fold elevation of AST. All patients under 18 recovered and were discharged. Elevated AST levels were observed in 32 (18.4%) patients with mild disease and 23 (44.20%) with the critical disease, whereas elevated ALT levels were observed in 33 (19%) patients with mild disease and 32 (61.5%) with the critical disease (SupplementaryTable 3).

Univariate analysis revealed several factors associated with survival (age < 60 years; no fever; no dyspnoea; normal findings on chest CT) and several associated with increased mortality (age > 60 years; leucocytosis; increased neutrophil/lymphocyte ratio; elevated AST, CRP, D-dimer or ferritin; baseline renal function impairment with elevations 1–2 above ULN) (Table 5).

Patients were followed up till discharge or death. On discharge, patients showed significant improvement in haemoglobin, gamma-glutamyl transferase, serum albumin, ferritin, CRP, and lactate dehydrogenase levels with mild elevations in ALT and AST levels (Supplementary Table 4).

Multiple stepwise logistic regression analyses were conducted to identify significant predictors for outcomes of death and recovery. Independent variables included age, asymptomatic status, fever, dry cough, dyspnoea, sore throat, steroid use, supplementary vitamin C intake, azithromycin intake, subcutaneous heparin, chest CT findings, ICU admission, FIB-4 score, creatinine level, and urea level. The model was significant [χ² (144)], and the P-value was < 0.001, meaning the independent variable could explain the change in the dependent variable by up to 90.8%. Significant predictors in our model included treatment with supplementary vitamin C (1 g daily capsules according to the Egyptian Ministry of Health protocol[13]) [odds ratio (OR): 0.05, 95% confidence interval (CI): 0.008–0.337]; lung consolidation (OR: 4.540, 95%CI: 1.155–17.840); ICU admission (OR: 25.032, 95%CI: 7.110–88.128), and FIB-4 score > 3.25 (OR: 10.393, 95%CI: 2.459-43.925). The most significant predictor was ICU admission (Table 6).

Within our cohort, 60 (13.98%) patients presented with GI symptoms in addition to respiratory symptoms. They were predominantly females with significantly higher body mass index. Diarrhoea was the most common symptom, affecting 52 (86.67%) patients. Headache and sore throat were more frequent in patients with GI symptoms compared to those without GI symptoms; 83.4% of patients with GI symptoms had non-severe COVID-19 presentations with fewer ICU admissions. There was no difference in FIB-4 scores among patients with and without GI symptoms (Table 7).

This is the first report from Egypt specifically exploring hepatic and GI disturbances in patients with COVID-19. Within our cohort, AST was more frequently elevated (32.10% of cases), compared with ALT (26% of cases). Only 21 (4.91%) and 16 (3.70%) patients, respectively, had a significant liver injury with AST or ALT elevation > 3-fold. The cholestatic pattern of elevated bilirubin > 1.1 mg/dL, elevated ALP >147 U/L, and GGT > 48 U/L was observed in only 8 (1.86%) patients. Our study findings align with those of other studies showing a direct relationship between elevated liver enzymes and COVID-19 disease activity.

Similar to our study, most other studies reveal a predominant pattern of disturbances in liver enzymes in COVID-19 patients, specifically higher AST elevation than ALT elevation[17-19]. A meta-analysis by Wu et al[20] found that low serum albumin and high GGT were the most frequent abnormalities on admission and that ALT elevation occurred most frequently during hospitalization, which they speculate may be due to the inclusion of patients with pre-existing liver disease. In our study, we observed no difference between survivors and non-survivors in serum albumin level. Wu et al[20] and Liu et al[21] found that serum albumin was significantly lower in non-survivors.

Chronic liver disease was reported in 3.29% of our patients. Phipps et al[4] reported that the severity and type of underlying chronic liver disease (e.g., non-alcoholic fatty liver disease and chronic hepatitis B or C) were not significantly associated with hepatic disturbances related to COVID-19, which could be due to the small numbers of those cases with chronic liver diseases in their study.

In our study, AST was significantly higher in non-survivors, but no significant difference was observed in ALT levels between survivors and non-survivors, which is similar to previous meta-analyses[17,18]. Patients with severe COVID-19 and patients admitted to the ICU had higher transaminase levels, as confirmed by previous meta-analyses[17-19,21,22]. Ponziani et al[23] reported that alterations in transaminase levels in patients with COVID-19 were mild to moderate, whereas significant liver injury with transaminases > 3 times the ULN and pure cholestatic injuries occurred in a minority of patients[23]. Phipps et al[4] showed that 6.4% of their patients had a severe liver injury with ALT > 5 times above the ULN. They also found that peak ALT, older age, diabetes mellitus, and intubation were independent predictors of mortality. A meta-analysis by Kulkarni et al[22] found that severe liver injury occurred in 10.7% of 3440 patients with COVID-19: 24.9% among 358 patients with the non-severe disease and 41.5% among 317 patients with severe disease.

Other causes of liver injury in patients with COVID-19 include medication, such as acetaminophen at doses exceeding 7.5–10 g and combined use of antivirals and antibiotics. Systemic effects of SARS-CoV-2 infection can be hepatotoxic. For example, hypoxia due to lung injury and, in severe cases, sepsis, acute respiratory distress syndrome, and multi-organ failure precipitate hypoxia and ischemia of the liver, causing elevated serum ALT, AST, and total bilirubin levels[17].

In this study, baseline FIB-4 was an independent factor affecting mortality, and it was significantly higher in patients admitted to the ICU, those with more severe COVID-19 disease, and non-survivors. FIB-4, which predicts significant hepatic fibrosis in patients with liver disease, comprises four variables: Age, AST, ALT, and platelet count[11]. These factors are deranged in COVID-19 disease due to systemic inflammation, bone marrow suppression, injury of the skeletal muscles, elevated pressure in the right heart system with subsequent liver congestion, and elevated liver stiffness that occurs in the disease[24,25]. Sterling et al[23] showed that gender, diabetes mellitus, and FIB-4 ≥ 2.67 were associated with increased mortality. Another recent study by Li et al[25] found that FIB-4 correlated with mortality and severe COVID-19, noting that FIB-4 normalized in the survivors only. They also found that baseline FIB-4 correlated with SARS-CoV-2 viral load and their studied monocyte/ interferon-I-related cytokines (interleukin-6 and interferon-gamma–induced protein 10)[25]. Rentsch et al[26] and Price-Haywood et al[27] found that FIB-4 scores > 3.25 and ≥ 2.67, respectively, were associated with hospitalization and with ICU admission and mechanical support.

Within our cohort, 60 patients (13.98%) presented with GI symptoms in addition to respiratory symptoms. The prevalence of diarrhoea, nausea, vomiting, and abdominal pain was 9.51%, 2.01%, 3.11%, and 3.84%, respectively. Diarrhoea was the most common GI symptom in our cohort, similar to findings from Wang et al[28].

The prevalence of diarrhoea, nausea/vomiting, and abdominal pain in their systematic review and meta-analysis was 9.1%, 5.2%, and 3.5%, respectively. The predominant GI symptoms differ among countries[28-30]. In China, for example, Luo et al[29] reported anorexia, nausea, and vomiting as the frequent symptoms in two-thirds of their patients, while diarrhoea and abdominal pain presented in 37%and 25% of their patients, respectively. A meta-analysis by Wan et al[30] (included 55 studies from China, 1 from Austria, 1 from the United States, 1 from Spain, and 2 from Singapore) reported the prevalence of GI symptoms as diarrhoea (53 studies, 8604 patients: 11.2%), nausea and vomiting (33 studies, 6165 patients: 10.0%), loss of appetite (15 studies, 2540 patients: 21.3%), and abdominal pain 14 studies, 2203 patients: (4.6%). Another meta-analysis by Parasa et al[31] reported that diarrhoea occurred in 4.3%–12.2%, and nausea, or vomiting occurred in 2.6%–8.0% of their included 4805 patients[31]. A study from the United States showed that GI symptoms were present in 61.3% of their patients (the most common was loss of appetite (34.8%) followed by diarrhoea (33.7%) and nausea (26.4%))[32].

In our study, 83.4% of patients with GI symptoms had non-severe COVID-19. The presence of diarrhoea or other GI symptoms did not impact COVID-19 mortality. Wang et al[28] also found no difference in the prevalence of diarrhoea among severe and non-severe cases, and the presence of GI symptoms did not affect mortality. Nobel et al[33] found that COVID-19 patients with GI symptoms had lower mortality and no difference in ICU admission rates, compared with patients without GI symptoms[33]. In contrast, Zhong et al[8] and Wan et al[30] reported a correlation between diarrhoea and severity of COVID-19, showing that patients with diarrhoea needed more ICU care and ventilator support.

In our cohort, there was no significant difference in transaminases level among patients with and without GI symptoms, which contrasts with a meta-analysis by Wijarnpreecha et al[17] showing higher transaminases in patients with GI symptoms. The results of our study are similar to Zhou et al[34], who found that GI symptoms were more common in females, sore throat was also more common in their patients with GI symptoms, and haemoglobin level was significantly lower in their patients with GI symptoms. In contrast to our study, ALT was higher in their non-medical group of patients with GI symptoms but not in the medical group, and CRP was higher in their patients with GI symptoms. Limitations of our study include inconclusive reports about medications received before hospitalization and the possibility of underlying undiagnosed liver disease, such as non-alcoholic fatty liver disease and unavailability of creatine phosphokinase enzyme results, as it was not tested at the time of the study in the quarantine hospitals affiliated with the Egyptian Ministry of Health.

In conclusion, within our cohort of Egyptian patients with COVID-19, elevated AST, ALT, total bilirubin ,and GGT were present in 32.1%, 26%, 5.8%, and 1.86% of our patients, respectively. Significant liver injury (AST and ALT three times higher than the ULN) affected 4.91% and 3.70% of patients, respectively. Male gender, smoking, hypertension, chronic hepatitis C, and lung involvement were significantly associated with elevated AST and ALT. FIB-4 scores were significantly higher in patients admitted to the ICU, those with more severe COVID-19, and non-survivors. The independent variables affecting outcome were supplementary vitamin C intake, lung consolidation, ICU admission, and FIB-4 score. GI symptoms were significantly more frequent in women, patients with higher body mass index, and those with non-severe COVID-19.