An increase in aspartate aminotransferase levels can predict worsening disease severity in Japanese patients with COVID-19

Introduction

In December 2019, Wuhan, China, experienced an outbreak of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), which was subsequently named coronavirus disease 2019 (COVID-19) by the World Health Organization (WHO) in February 2020. The disease has spread rapidly, leading to a global pandemic that has persisted for three years to date. The first case of COVID-19 pneumonia in Japan was reported in Kanagawa Prefecture on January 15, 2020. The number of cases peaked in early April, confirming an epidemic (the first wave of the COVID-19 pandemic). Subsequently, the number of cases increased among individuals aged 20 to 30 years, reaching its peak in early August (the second wave). Outbreaks occurred in early January 2021 (the third wave), early May 2021 (the fourth wave, primarily caused by α strains), late August 2021 (the fifth wave, predominantly caused by δ strains), and January 2022 (the 6th wave, when Omicron strains rapidly replaced δ strains). Although the greatest number of patients were recorded in the seventh wave (caused by Omicron strains), the number of patients with severe disease requiring respiratory care was relatively small. As of July 21, 2022, in Japan, the Omicron BA.2 strain, a substrain of the B.1.1.529 strain, was believed to have been replaced by the BA.5 strain, which has become the dominant strain. [1]

Although COVID-19 has been reported to cause liver dysfunction in both Japanese and international studies, the pathogenesis, treatment, and prognostic significance of hepatic dysfunction in COVID-19 patients remain unclear. [2] Recent reports indicate that more than half of COVID-19 patients exhibit varying levels of liver dysfunction. Elevated aminotransferase levels have been reported in 14–58% of hospitalized patients with COVID-19. [3] While the degree of aminotransferase level increase is usually mild (<5 times the upper normal limit), severe acute hepatitis and high aminotransferase levels have been documented. However, comprehensive data on the clinical characteristics of liver enzyme elevation and liver failure in Japanese COVID-19 patients are lacking.

Although numerous risk factors for severe COVID-19 have been identified, the impact of liver function abnormalities has yet to be investigated owing to confounding factors and the limited number of reports in large Japanese patient cohorts. In this study, we aimed to retrospectively examine the associations between liver function test values and clinical characteristics in our cohort of Japanese COVID-19 patients who were diagnosed before the 6th wave and had a lower rate of severe disease.

Patients and Methods

This was a retrospective single-center cohort study of consecutive Japanese patients who were diagnosed with COVID-19 through PCR or antigen methods and admitted to our hospital between February 2020 and July 2021. Osaka Prefecturure, the third most populous prefecture in Japan, established the Osaka Prefectural Inpatient Follow-up Center on March 30, 2020, to coordinate hospitalization based on patient symptoms and triaged patients with COVID-19 according to disease severity. This public hospital, located in Toyonaka in Osaka Prefecture, has a population of approximately 400,000 and is primarily designated a general hospital for treating patients with moderate to severe COVID-19. This has been in place since the first wave of the pandemic in Japan. [4-8]

We assessed the severity and treatment strategy for hospitalized COVID-19 patients based on guidance from the Japanese Ministry of Health, Labor, and Welfare available on the COVID-19 website, which has been available since March 7, 2020. The latest version (8.0) was released in July 2022 (in Japanese). [1] The severity of COVID-19 was categorized into four groups: mild disease without respiratory symptoms (SpO₂ >96%); moderate I disease with breathing difficulties or pneumonia without respiratory failure (93%<SpO₂<96%); moderate II disease requiring oxygen support (SpO₂ ≤93%); and severe disease requiring intensive care unit (ICU) treatment or intubation. This Japanese severity classification is generally equivalent to the WHO severity classification for mild, moderate, severe, and critical disease [9], although there are some differences in the definitions. Treatment strategies have been updated with therapeutic advances. The therapeutic agents used in our hospital were hydroxychloroquine, favipiravir, ciclesonide, and dexamethasone from waves 1 to 3 and favipiravir, dexamethasone, heparin, and remdesivir from waves 4 to 5. We reviewed the clinical characteristics of COVID-19 patients with liver involvement by collecting electronic medical records from our hospital (MegaOak Online Imaging System, NEC, Japan). Clinical and laboratory data were collected from all patients at hospital admission. We prospectively used medical templates to assess all hospitalized COVID-19 patients at our infectious disease unit to avoid missing data. Patient height and body weight were obtained from medical interviews to reduce contact with patients. We conducted hepatitis B virus serology and hepatitis C virus antibody tests if they had not been previously conducted. We performed a portable chest radiograph for the initial evaluation of pulmonary complications at admission.

This study was conducted in accordance with the Declaration of Helsinki, and the institutional review board of Toyonaka Municipal Hospital approved this study (2022-10-3). The requirement for informed consent was waived by the same committee using the opt-out method on our hospital website.

In the context of our retrospective study, data were accessed on December 11, 2022, for research purposes.Throughout the research process, including during and after data collection, we had access to information that could potentially identify individual participants.All necessary ethical considerations and precautions were undertaken to maintain confidentiality and comply with applicable data protection regulations.

We investigated several predictive variables associated with the development of severe conditions, including age ≥ 65 years, BMI >30, smoking status, hypertension status, diabetes status, hyperlipidemia status, chronic kidney disease status, chronic lung disease status, solid cancer status, pregnancy status, lymphocyte count, lactate dehydrogenase (LDH) level, and C-reactive protein (CRP) level. We previously reported that the cutoff values for lymphocyte count, LDH levels, CRP levels, and estimated glomerular filtration rate (eGFR) were 980 count/μl, 309 IU/ml, 2.92 mg/dl, and 68 ml/min, respectively, as predictive risk factors in our hospital cohort. [4] In the present study, we estimated that cutoff values of 1000, 300, 3, and 70 ml/min (CKD) were predictive risk factors. The FIB-4 index, which evaluates liver fibrosis, was computed using the formula FIB-4 index = age × aspartate aminotransferase (AST)/platelet count × √ alanine aminotransferase (ALT). [10] We also assessed two liver function parameters, aspartate aminotransferase (AST) and alanine aminotransferase (ALT), and categorized patients into three groups at admission based on their AST and ALT levels: those with normal aminotransferase levels (grade 1, AST/ALT < 30 U/L), those with moderately elevated aminotransferase levels (grade 2, 30 U/L≤ AST/ALT < 60 U/L), and those with highly elevated aminotransferase levels (grade 3, 60 U/L< AST/ALT).

Results

During the study period, 748 COVID-19 patients were enrolled. Of these, 37 readmitted patients, 17 patients transferred from another hospital, 29 asymptomatic patients, 41 patients younger than 16 years, and 10 non-Japanese patients were excluded, as were 10 patients with insufficient data. Ultimately, 604 patients were included in the final analysis (Figure 1).

• Download figure
• Open in new tab

Figure 1: Flow chart of patient enrollment. COVID-19. We enrolled 748 COVID-19 patients during the study period, and 604 patients were included in the final analysis.

The clinical characteristics of the enrolled COVID-19 patients are summarized in Table 1. The median age was 62 years (IQR, 47-78), and 335 patients (55.5%) were men. The median BMI was 23.9 (IQR 21.0, 26.9), and 34.9% and 34.9% of the patients had a history of alcohol consumption and smoking, respectively. Other comorbidities included hypertension in 270 (44.1%) patients, cardiac disease in 103 (17.2%), diabetes in 168 (27.9%), hyperlipidemia in 133 (23.3%), and pregnancy in 23 (3.8%). Regarding medications, 148 (24.5%) patients were taking ACEis/ARBs, 190 (31.6%) were taking calcium blockers, and 110 (18.3%) were taking statins.

The median time from onset to admission was 6 days (IQR: 4–9 days). On admission, 92.7% of the patients had fever, 48% had fatigue, 77.3% had respiratory symptoms, and 75.6% had pneumonia. Twenty-five patients (4.1%) were diagnosed with severe disease on admission, and 141 (23.3%) developed severe disease within a median of 2 days (IQR 1,5) of hospitalization. The median length of hospital stay was 10 days (IQR 7,15), and 43 (7.1%) patients died during hospitalization (Table 2).

Discussion

Blood tests can provide important information on the condition of patients with COVID-19 and aid in predicting their prognosis. Therefore, blood tests are recommended for hospitalized COVID-19 patients who are at risk for severe disease or those with moderate or severe disease. Several studies on biomarkers for disease progression that assist in determining illness severity and predicting patient prognosis have been conducted in Japan and internationally. The application of these markers is expected to enhance the quality of medical care and optimize the utilization of medical resources. However, the impact of liver function tests has not been adequately explored, and there is a paucity of research studies conducted on large cohorts of Japanese patients. In this study, multivariate analysis incorporating established risk factors revealed that elevated AST levels at admission were an independent and significant risk factor for severe disease in hospitalized Japanese COVID-19 patients. Furthermore, of the COVID-19 patients who presented with AST levels of 60 IU/ml or higher on admission, 44.0% progressed to severe disease, which was 1.8 times greater than the rate in the entire cohort.

Regarding the impact of underlying liver disease, no significant findings were obtained due to the low number of patients with viral hepatitis or autoimmune diseases in this cohort. Owing to the constraints of minimizing contact with patients, abdominal ultrasound (US) was not performed; thus, we were unable to verify previous data or identify patients with alcoholic liver disease or nonalcoholic fatty liver disease [11]. Although a reduction in the PLT is indicative of the severity of liver fibrosis [12], it was challenging to determine whether the patients had underlying liver disease based on the PLT in this study, as it has been reported that PLT decreases in patients with COVID-19 infection due to inflammation and intravascular coagulation disorders. [13]

The liver may be vulnerable to SARS-CoV-2 infection due to the presence of angiotensin-converting enzyme 2 (ACE2) receptors on biliary and hepatic epithelial cells. [14] The virus enters and harms target organs by binding to ACE2 receptors. [15,16] Autopsy results have confirmed the presence of viral RNA in liver tissue [17], indicating the possibility of direct hepatocellular damage caused by SARS-CoV-2. In our study, 97 patients received angiotensin receptor blockers (ARBs), 9 received angiotensin-converting enzyme (ACE) inhibitors, and 45% had hypertension. Among the 105 patients who were treated with ARBs and/or ACE inhibitors, 35.2% (37/105) progressed to severe disease, which was significantly greater than the percentage of patients (20.5%; 82/401) who did not receive these medications (P=0.00027, data not shown). In addition, AST severity was significantly greater in ARB and/or ACE inhibitor users than in nonusers (grade 1 in 36.5% of users and 50.1% of nonusers, grade 2 in 48.2% of users and 38.0% of nonusers, grade 3 in 15.5% of users and 11.9% of nonusers; P=0.0478; data not shown). Our findings suggested that ARB/ACE inhibitor use may upregulate ACE2 receptor expression in biliary and hepatic epithelial cells, which could lead to liver dysfunction in ARB/ACE inhibitor users. However, the impact of hypertension as a confounding factor could not be determined. Furthermore, liver dysfunction in COVID-19 patients may also be attributed to inflammation, cytokine storms, the use of therapeutic drugs and hypoxemia associated with respiratory failure.

This study described the first to fifth waves of the COVID-19 outbreak in Japan, during which different mutant strains were prevalent, leading to varying infectivity and symptoms. Since the number of available drugs increased from the fourth period, we classified patients into two groups: those in the earlier period up to the third period and those in the latter period after the fourth period. We investigated differences in the background and frequency of liver injury between these two groups. The frequency of liver injury at presentation was greater in the latter period (Grade 1 (48.8%, 35.6%), Grade 2 (39.9%, 43.4%), and Grade 3 (11.3%, 21.0%)). However, the frequency of AST level upgrades was not significantly different between the two groups (21.8%, 21.0%, P=0.81). Although there was no significant difference in the severity of admission between the two groups, the frequency of progression to severe disease was significantly greater in the latter group (20.0%, 28.8% P=0.013) (Supplementary Table 1). Although it was challenging to assess this due to differences in hospitalization criteria during each wave, it was suggested that the variation in viral strains might have an impact on liver damage.

With the spread of COVID-19, existing drugs have been repurposed, and several treatments have been approved, but liver damage has been reported with COVID-19 treatments. [18-22] Such treatments have also been reported to induce liver injury. [23]. At our hospital, chloroquine, favipiravir, ciclesonide, and dexamethasone were used from the first to the third wave, and favipiravir, decadron, heparin, and remdesivir were used from the fourth wave to the fifth wave. The impact of each drug on liver dysfunction was studied. Univariate logistic analysis of the six drugs, disease progression to severe COVID-19, and worsening AST levels during hospitalization indicated that the use of favipiravir, ciclesonide, and remdesivir was a significant factor in exacerbating COVID-19. Hepatic injury exacerbation after hospitalization was influenced by drug use as well as by worsening COVID-19 symptoms; however, drawing conclusions is challenging due to several confounding factors (Supplementary Table 2).

The present study has several limitations. First, it was a retrospective, single-center study. Second, the search for past liver function abnormalities was limited, as the patients were examined in isolation with minimal contact, leading to potentially inadequate evaluation of some patients with abnormal liver function. Thus, the study could not determine whether the presence or absence of background liver disease impacted the course of the disease. Third, the differences in virus strains and the involvement of the drugs used have not been adequately investigated. The relationship between drugs and liver injury was also examined; however, the backgrounds of patients treated with each drug and those who did not differ in terms of the decision to administer drugs were based on the symptoms of COVID-19, making it challenging to draw conclusions regarding drug effects.

In conclusion, our study demonstrated that AST levels at admission were a significant independent risk factor for severe disease in hospitalized Japanese patients with COVID-19.

Figure legends

Supplementary Figure 1: AST levels on admission and peak AST levels. During the clinical course, 79 patients with grade 1 ASTs were upgraded to a higher AST grade, and 51 patients with grade 2 ASTs were upgraded to grade 3, although the AST grade did not change in 383 patients with grades 1 and 2.

Supplementary Figure 2: The severe disease-free survival times in the three grade groups. The three groups were significantly separated in parallel according to the severity of AST.