INTRODUCTION

The disease caused by the SARS-CoV-2 virus, denominated COVID-19, affects primarily the respiratory system. Patients affected by this disease often present fever, cough and dyspnea as their most common symptoms ^1,2. In severe cases, patients may develop pneumonia and associated complications, such as acute respiratory distress syndrome with septic shock and ultimately, death ¹.

The SARS-CoV-2 virus causes a systemic disease, producing injury on the heart, liver, pancreas and kidney, moreover immune system dysfunction and lymphocytes alterations have been reported ^2-5. The liver is one of the most affected organs by this disease and many studies have reported alteration in the levels of liver biochemistry tests such as aspartate transferase (AST), alanine transferase (ALT), total bilirubin and albumin ^6-9. Several degrees of elevation in the level of hepatic enzymes have been reported. For example, alterations of AST and ALT have been reported ranging from 2.5% - 50% and 2.5 - 61.1% of the patients, respectively ¹⁰. Moreover, some studies have found an association between liver biomarker levels and severity of the disease. For example, in a cohort of 1099 patients, Guan et al. ¹¹ observed elevated levels of AST in 18.2% of the patients with non-severe disease and in 39.4% with severe disease. Similarly, Huang et al. ² reported that the proportion of liver damage in intensive care unit (ICU) patients (61.5%) were higher than non-ICU patients (25.0%). In relation to the levels of total bilirubin (TB) these can rise in 0% to 35.3% of the cases ^6,8,11.

Hepatic injury caused by SARS-CoV-2 can arise from various mechanisms, such as direct infection of liver cells, systemic inflammation with production of proinflammatory cytokines, hypoxia, microthrombosis and drug-induced liver injury ^7,9,10. For example, it has been shown that SARS-CoV-2 can directly infect liver cells, given that angiotensin converting enzyme 2 (ACE2), which is an important trans-membrane receptor for viral entrance to the cell, is expressed in the liver and bile duct cells ¹². Data from two independent cohorts revealed ACE2 expression in 2.6% of hepatocytes and 59.7% of cholangiocytes, suggesting that SARS-CoV-2 could directly bind to ACE2-positive cholangiocytes and alter hepatic function ¹³.

Liver biopsy samples from patients who died from COVID-19 showed moderate microvesicular steatosis, as well as, mild lobular and portal inflammatory activity ¹⁰, however, these nonspecific lesions could have been caused by either SARS-CoV-2 or druginduced injury ¹⁴. It has been suggested that drugs toxicity may contribute significantly in the alteration of liver biochemistry tests ⁷. Liver damage can occur after the use of multiple medications, such as antivirals, antibiotics, traditional medicine, antipyretics, and pain relievers ¹⁴.

The aim of this study was to identify the prevalence of alterations in liver function tests in Peruvian patients hospitalized for COVID-19. Also to identify association between severity of the disease and elevation of these biomarkers.

MATERIALS AND METHODS

RESULTS

A total of 1,100 patients were enrolled during the study period. The 81.7% (899/1100) of patients who were admitted to hospital with the diagnosis of pneumonia due to COVID-19 had altered liver function tests. Among these, AST and ALT were altered at admission in 64.7% and 63.7%, of the patients respectively. The majority of our patients with elevated liver enzymes were registered in the age group ≥60 years old (46.8%). Also we could highlight that in the age group 20-39 years old the proportion of patients with elevated liver enzymes was significantly higher than the group with normal transaminases (Table 1). When comparing gender, it was evidenced that 70.6% of male patients had altered liver function tests compared to 29.4% of female patients. Moreover, female patients were less likely to present altered liver tests OR=0.53 (95% CI: 0.39-0.73; p<0.01).

Table 1 Demographic characteristics of 1100 patients admitted for COVID-19 and their association with liver function tests. 

On the other hand, past medical history, co-morbidities and body mass index (BMI) were evaluated in relation to liver function tests. Patients with BMI higher than 25 kg/m² did not show a significant elevation in the liver function tests compared to patients with BMI lower than 25 kg/m². Other factors associated with elevated liver enzymes were: dyslipidemia OR=1.72 (95% CI: 1.102.70; p=0.01) and previous medication OR=1.56 (95% CI: 1.12-2.16, p<0.01). In the case of signs and symptoms on admission, it was evidenced that patients with fever OR=1.43 (95% CI: 1.03-1.199, p=0.03) were more likely to present altered liver function tests. Interestingly, patients with non-alcoholic fatty liver disease (NAFLD) did not show association with alteration in liver function tests.

Table 2 shows the association between each liver function test evaluated and the severity of the disease. It was evidenced that levels of AST and ALT <3ULN did not show significant association with disease severity. On the other hand levels of ≥3ULN were significantly associated with greater severity of the disease. Meanwhile, alkaline phosphatase, total bilirubin and total albumin did not show significant differences.

Table 2 Alterations in liver biochemistry tests according to severity of the COVID-19 disease. 

Table 3 shows the association between selfmedication taken before admission and alteration in liver function tests. We could highlight that 437 patients (39.7%) took self-medication before admission and among these, 63 patients had normal liver function tests and 374 had some degree of alteration. Patients who took self-medication had significantly higher alteration in hepatic function tests, also, self-medication was associated with an OR=1.56 (95% CI: 1.122.16; p<0.01). The most frequently consumed drugs were paracetamol (34.0%), ivermectin (29.1%) and azythromicin/ceftriaxone (27.9%). Patients who took paracetamol had a higher risk to develop alteration in hepatic function with an OR= 1.41 (95% CI: 1.01-1.98; p=0.04). Meanwhile, in patients who took ivermectin and azythromicin/ceftriaxone did not show significant differences among groups. On the other hand, patients taking corticosteroids OR=0.55 (95% CI: 0.38-0.78; p<0.01) and enoxaparin OR=0.53 (95% CI: 0.350.81; p<0.01) showed a significant lower proportion of alteration in hepatic function tests.

Table 3 Associations between prior medication taken before admission and liver function tests. 

Table 4 Frequency of hepatic injury in COVID-19 patients. 

DISCUSSION

COVID-19 is a heterogeneus and complex disease, although it mainly affects the respiratory system, there is growing evidence that it can cause injury on other organs. The gastrointestinal system, particularly the liver, are affected by this disease in variable degrees ¹⁶. Liver dysfunction in the context of COVID-19 may represent an important prognostic variable, therefore the assessment of the liver involvement is essential in patients care. For example, it has been found that increased levels of AST, ALT and AST/ALT ratio are important markers of mortality risk, severity of disease and probability of ICU admission ^17-19.

We performed the largest study evaluating the liver function tests of patients admitted with a diagnosis of COVID-19 in Peru, with 1,100 patients enrolled during the study period. A high prevalence (81.7%) of the patients presented at least one altered liver biochemistry test. Moreover, AST and ALT were altered at admission in 64.7% and 63.7%, of the patients respectively. These findings are significantly higher than those reported in previous literature with almost 3 times higher compared to others studies, as shown in Table 4. One of the largest series of patients was performed by Guan et al. ¹¹, evaluating 1,099 COVID-19 patients in China and among these, 21.3% and 22.2% had elevation of AST and ALT, respectively. We could highlight that this study uses the value of 40 U/L as the upper normal limit for both enzymes, which is similar to our study. On the other hand other studies ^20-22 have used values above 1.5 to 2 times greater the ULN to determine liver tests abnormalities and reported alterations ranging from 13% to 21.6%. It is important to consider that aforementioned studies did not evaluate important variables such as self-medication. One possibility that may explain the high frequency of alteration in liver function tests could be related to the high proportion of patients who took self-medication.

When evaluating the demographic variables in relation to hepatic injury, it was evidenced that some factors such as age, gender, co-morbidities and use of previous medication were associated. Previous evidence has shown that patients older than 50 years are those who are more severely affected by the disease ²³. Similar to our study, in which we report that patients older than 60 years old had a greater prevalence of alteration in liver enzymes. Remarkably, we report that patients in the 20-39 years old group had also a significant higher alteration in liver function tests, which have not been reported previously. Regarding gender, it was evidenced that male patients had elevated liver tests (70.6%) compared to of female patients (29.4%), this difference was statistically significant. Moreover, female were significantly less likely to present abnormal liver function tests.This finding is similar to what has been reported in previous literature, proposing that men are more likely to be affected by the disease and present higher degrees of liver injury ^23,24. Also, the majority of the patients hospitalized in our study were male (68%).

Several co-morbidities have been associated with a poor prognosis, greater severity and mortality in patients affected by COVID-19. Diseases like diabetes mellitus, obesity and cardiovascular diseases are among the most important ²⁴. Moreover, patients with chronic liver diseases may also be at a higher risk of developing serious complications ²⁵. In the current study, no clear association was found between the alteration of liver function tests and co-morbidities like hypertension, diabetes mellitus, chronic liver diseases and obesity. We could highlight a low prevalence of chronic liver diseases in our study, such as cirrhosis (2.8%) and viral hepatitis B/C (2.1%), this finding is in agreement with a previous meta-analysis, showing that COVID-19 patients present a global prevalence of chronic liver diseases of 2.8% ²⁶. In the case of Peru, the leading causes of chronic liver diseases are non-alcoholic fatty liver disease (NAFLD) and nonalcoholic steatohepatitis (NASH), followed by alcholrelated liver disease and viral hepatitis. For example, a study of 200 transplants performed in Peru, reported that the most frequent indication for this procedure was NAFLD with 35% ²⁷. Another study reported the most common causes of chirrosis which included: alcohol (28%), NAFLD (21.3%), Hepatitis B virus (15.2%) and Hepatitis C virus (11.8%) ²⁸. These findings highlight the need to maintain a careful observation on patients with liver diseases and consider them high-risk patients ²⁶.

In the case of past medical history, we found that dyslipidemia, fever on admission and taking medication prior to enrollment were significantly associated with abnormal liver tests. These findings could be explained by a variety of reasons. Firstly, alteration in lipids may predispose patients to hepatic steatosis and NAFLD, making them more susceptible to liver damage ⁸. Nonetheless we observed that patients with NAFLD in the current study did not show signifant differences in the level of transaminases. This could be explained because NAFLD and even NASH do not neccesarily present with elevation of liver enzymes, in fact many patients with hepatic fibrosis may still have normal transminases levels ^29,30. Moreover, taking previous medication can cause variable degrees of elevation in hepatic enzymes, depending on the pharmacologic agent and doses. In the case of fever, this represents a first line response from the body to combat infection, and could be related to the degree of systemic inflammation, which may induce liver injury ^7,16.

An evaluation of disease severity and its association with each individual liver biochemistry test was performed. The levels of AST and ALT were significantly associated with disease severity (p<0.01), particularly levels 3 times greater than the upper limit value. On the other hand, other tests like albumin, alkaline phosphatase and total bilirubin did not show significant differences. A previous review, stratified clinical variables in three levels of risk for developing severe disease, among hepatic function tests it was evidenced that hypoalbuminemia was classified in the highest risk group, meanwhile AST and ALT in the lowest risk group ²³. However, another systematic review revealed that AST, ALT and total bilirubin are strongly correlated with disease severity ³¹. In our study we could also confirm that elevations of AST and ALT three times above the upper limit of normal were associated with disease severity. It is of great importance to mention that only 81/1,100 (7.4%) patients were admitted to ICU in our study, which could explain that minimal rates of hypoalbuminemia found. According to previous literature, this finding is often associated with long hospital stays and tends to appear as a late finding, particularly in patients on ICU ¹⁶. Most of our patients were in regular hospitalization wards and albumin assessment was performed on admission.

COVID-19 may induce liver injury and dysfunction by several mechanisms including: direct cytotoxicity of viral replication, systemic inflammatory response, hypoxia-mediated injury, microthrombi formation, drug-induced injury and reactivation of previous liver diseases ^7,10. Firstly, liver injury caused directly by SARSCoV-2 may be one of the most important mechanisms, given that ACE2 receptors are expressed in many human tissues apart from the lungs. Previous reports have observed the abundant expression of this receptor in enterocytes and cholangiocytes, and in a lesser degree in hepatocytes ^12,13,32. These findings may explain direct viral invasion on these cells, causing elevation in liver biochemistries. Although, cholangiocytes are most likely to be affected by the virus ^12,13, the dysfunction of these cells may cause indirect transaminases elevation. However, alkaline phosphatase, a marker of billiary tract affection, was only elevated in 18% of our patients, with no significant differences in severity. Also, levels of total bilirubin, another marker of cholestasis may be elevated in 0% to 35.3% of the cases. In the current study, we found an elevation of bilirubin in 29.3% of the patients, with no significant differences in severity. Altogether, these findings may indicate a hepatocellular pattern rather than a cholestasic pattern, in patients affected by COVID-19 ^9,10,17.

Finally, given that drug-induced liver injury is an important mechanism for hepatic injury as shown in previous studies ¹⁴, an evaluation on medications taken prior to admission was performed. We could observe that 41.6% of the patients with elevated liver enzymes took self-medication before enrollment. It was evidenced that self-medication was associated with elevated liver enzyme with an OR=1.56 (95% CI: 1.12-2.16; p<0.01). Patients who took paracetamol had a higher risk to develop alteration in hepatic function with an OR= 1.41 (95% CI: 1.01-1.98; p=0.04). Given that a high proportion of patients took self-medication under no medical supervision, it is possible that they have taken inappropriate doses of paracetamol, considering that consumption of paracetamol at therapeutic doses is unlikely to cause an important liver injury ^33-35.

In our study, a variety of other medications were taken by the patients such as Azithromycin/Ceftriaxone, Ciprofloxacin/Levofloxacin, Hydroxychloroquine, Corticosteroids, Enoxapaparin and Ivermectin. We could highlight that consumption of corticosteroids OR=0.55 (95% CI: 0.38-0.78; p<0.01) and enoxaparin OR=0.53 (95% CI: 0.35-0.81; p<0.01) may be protective factors against liver injury. Possible mechanisms that may explain this finding is that corticosteroids could diminish systemic inflammation, avoiding liver involvement by cytokines ^29,36. On the other hand, the administration of anticoagulants such as enoxaparin, may have contributed to prevent thrombi formation, which could in turn generate hypoxia mediated liver injury, as reported previously^37,38 . Widely used medications like Hydroxychloroquine and Ivermectin did not show association with alteration in liver function tests in our study. However, previous studies have shown that they may cause liver injury and should be avoided as they do not provide a clear benefit ^34,36.

These findings highlight the high proportion of patients that take self-medication without medical supervision, which may contribute to later liver dysfunction. Moreover, future studies are required to determine the protective role of corticosteroids and enoxaparin and to examine the proposed mechanisms.

In conclusion we found that 81.7% of patients hospitalized for COVID-19 had alteration in liver function tests. Risk factors for liver injury were being male, history of dyslipidemia, fever on admission and taking prior medication. Severity of the liver injury by COVID-19 was associated with values of AST and ALT three times above the limit level of normality. It was found that patients on self-medication and taking paracetamol were more likely to present liver injury. Finally, corticosteroids and enoxaparin may have a protective role against liver dysfunction in COVID-19 patients. Further studies are required to evaluate alteration in liver function tests as prognostic predictors.