1. Introduction
Coronavirus disease 2019 (COVID-19) is caused by the identified severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), initially reported in Wuhan, China.[1] As COVID-19 is an emerging virus, an effective treatment has not been developed for disease resulting from this virus. However, in Wuhan, according to “Diagnosis and Treatment Scheme of New Coronavirus Infected Pneumonia,”[2] lopinavir plus ritonavir, arbidol, ribavirin, mesenchymal stem cells, traditional Chinese medicine and glucocorticoid therapy usage have commenced recruitment in clinical treatment.
Up to now, large volumes of research from epicenters of infection have focused on COVID-19 related liver impairment. In China, some studies showed an increase of transaminases and other liver enzymes in 14% to 53% of infected patients. These abnormalities were generally mild with moderate increase in transaminase levels in symptomatic and/or severe forms of COVID-19 infection.[3] A pathological study of liver biopsy specimens from a patient who died from COVID-19 showed moderate micro vesicular steatosis and mild lobular and portal activity, indicating that SARS-CoV-2 might have led to this liver damage.[4] Additionally, the SARS-CoV-2 might bind to angiotensin-converting enzyme 2 (ACE2) on cholangiocytes, leading to cholangiocyte dysfunction and inducing a systemic inflammatory response leading to liver injury.[5] Up to date, the mechanisms are still mostly unknown. It may involve the role of the virus itself, the role of virus-induced inflammation, liver hypoxia (thrombotic context, severe acute respiratory syndrome), and/or drug-induced liver injury (including paracetamol misuse, anti-infectious agents, etc).[6] Thus, some authors suggest that drug-induced liver injury (DILI) is also a possible contributing factor to laboratory test abnormalities.[7,8] Liver injury may occur after the use of multiple drugs, such as antivirals, antibiotics, traditional Chinese medicine, antipyretics and analgesics. Importantly, DILI during the treatment of coronavirus infection should not be ignored and needs to be carefully investigated. These drugs are all potential causes of liver injury during COVID-19, but not yet being evident.[9,10]
In the study, we retrospectively investigated the liver test abnormalities or liver injury of patients with COVID-19 at admission and during hospitalization in the hospital, respectively. Furthermore, the study was to clarify the relative risk factors of DILI during hospitalization, and evaluate the association between DILI and current medications, and therefore provide a reference for clinical treatment of COVID-19 patients.
2. Materials and Methods
2.1. Study design and participants
The medical records of 415 patients receiving anti-COVID-19 therapies from January 2020 to March 2020 were collected. Patients demographic characteristics, history of liver diseases, blood chemistry laboratory tests, comorbidities and concomitant medications were included.
The clinical criteria of diagnosis and discharge refer to the standards for “Diagnosis and Treatment Scheme of New Coronavirus Infected Pneumonia.”[2] Four hundred-fifteen patients had a history of related epidemiology, most of them had clinical manifestations, such as fever or respiratory symptoms. Four hundred-fifteen patients were diagnosed after examination of SARS-CoV-2 RNA by RT-PCR and chest CT scanning. Of the 415 patients, 112 patients had been excluded due to incomplete liver function test records. The remaining 303 patients were enrolled in the study (Fig. 1).
;)
Figure 1.: Patient selection diagram. COVID-19 = coronavirus disease 2019, DILI = drug-induced liver injury.
The inclusion criteria of DILI for patients were as follows: Hospitalization for at least 7 days; At admission, normal range of liver function tests at baseline; Conduction of follow up laboratory data adequate for assessing hepatic injury. Excluded from study were patients who had evidence of liver injury or liver function abnormalities or based on initial liver function test results, or liver disease, such as viral hepatitis, liver cirrhosis, hepatocellular carcinoma, and so on, at admission.
The usage and dosage of drugs were shown in Table 1. Dose and schedule of drug was modified according to patient’s age and medical condition by each attending physician’s decision.
Table 1 -
Usage and dosage of drugs.
| Drugs |
Usage and dosage |
| Antivirals |
|
| Oseltamivir |
75 mg/bid/po |
| Arbidol |
0.2g/tid/po |
| Ribavirin |
0.5g/bid/ ivgtt |
| Ganciclovir |
250mg/bid/ivgtt |
| Hydroxychloroquine |
0.2g/ bid/po |
| Recombinant human interferon ɑ-2b |
3 million IU/bid/inh |
| Antibacterials |
|
| Moxifloxacin |
0.4g/ qd/ ivgtt |
| Cefoperazone sodium and sulbactam sodium |
3g/ q8h/ ivgtt |
| Levofloxacin |
0.5g/qd/ ivgtt |
| Ceftazidime |
2g/ q12h/ ivgtt |
| Imipenem and cilastatin sodium |
1g/q8h/ ivgtt |
| Meropenem |
1g/q8h/ ivgtt |
| Chinese patent drug |
|
| Lianhua Qingwen Chengfang |
1.4g/ tid/po |
| Xuebijing |
100 mL/ bid/ ivgtt |
| Chinese herbal medicine |
[1] |
| Methylprednisolone |
40 mg/qd/po |
| Cardiovascular drugs |
|
| Atorvastatin |
20 mg/qn/po |
| Aspirin |
0.1g/qd/po |
| Enoxaparin |
40 mg/qd/po |
| Clopidogrel |
75 mg/qd/po |
| Metoprolol |
100 mg/qd/po |
| Nifedipine |
20 mg/bid/po |
| Digestive system agents |
|
| Esomeprazole |
20 mg/qd/po |
| Omeprazole |
40 mg/qd/po |
| Pantoprazole |
40 mg/qd/po |
| Mosapride |
5 mg/ tid/po |
| Smectite |
3 g/ tid/po |
| Respiratory drugs |
|
| Ambroxol |
30 mg/qd/ ivgtt |
| Doxofylline |
0.2 g/qd/po |
| Acetylcysteine |
0.6 g/bid/po |
| NSAIDs |
|
| Loxoprofen |
60 mg/bid/po |
| Etoricoxib |
120 mg/qd/po |
NSAIDs = non-steroidal anti-inflammatory drugs.
[1] National Health Commission of the People’s Republic of China. Diagnosis and Treatment Scheme of New Coronavirus Infected Pneumonia (7th). (2020).
2.2. Severity of COVID-19
All patients were classified into severe or mild cases based on results from chest radiography, clinical examination, and symptoms.[2,11]
2.3. Liver test abnormalities and liver injury
Liver test abnormalities were defined as the elevation of the following liver enzymes in serum: alanine aminotransferase (ALT) > 40U/L, aspartate transaminase (AST) > 35U/L, alkaline phosphatase (ALP) > 100U/L, gamma–glutamyltransferase (GGT) > 45U/L, and total bilirubin (TBIL) > 23µmol/L.
Acute liver injury was defined on biological criteria based on the ULN of ALT, ALP and TBIL, as follows: elevation of ALT ≥ 5 × ULN; Elevation of ALP ≥ 2ULN (in the absence of known bone pathology); Combination of an elevation of ALT ≥ 3ULN with simultaneous elevation of TBIL exceeding 2 × ULN.[3]
Acute liver injury was classified into 3 types according to the ratio of serum ALT to serum ALP: Hepatocellular injury, ALT/ALP ≥ 5, or ALT > 2 × ULN and ALP in normal range; Cholestatic injury, ALT/ALP ≤ 2, or ALP > 2 × ULN and ALT in normal range; Hepatocellular–cholestatic mixed injury, ALT > 2 × ULN, ALP > 1 × ULN and 2 < ALT/ALP < 5.[12]
Acute liver injury was classified as mild (2 × ULN ≤ ALT and/or AST < 5 × ULN), moderate (5 × ULN ≤ ALT and/or AST < 10 × ULN) and severe (10 × ULN ≤ ALT and/or AST, ALP ≥ 2 × ULN, TBIL ≥ 2.5 × ULN).[12]
The causality assessment followed the Roussel Uclaf Causality Assessment Method model (RUCAM).[13]
2.4. Statistical analysis
The statistical analyses were completed using the SPSS 24.0 software (IBM Corp., Armonk, NY). Categorical variables were expressed as frequencies and percentages. Continuous variables were presented as mean ± SD or median (range). The categorical variables were subject to chi-square test, the unidirectional ordered data was subject to Mann-Whitney test of independent samples. Univariate and multivariate logistic regression analysis of risk factors for DILI were performed. Odds ratio (OR) and 95% confidence interval (CI) were calculated from the confidence and standard errors of the model, and Homos–Lemeshow test was used to evaluate the goodness of fit. Statistical significance was defined as a bilateral P value <.05.[14]
2.5. Ethical review and informed consent of patients
The study was approved by the Renmin Hospital Ethics Committee of Wuhan University (ID: WDRY2021–K010). In view of the retrospective and observational nature of the study with no interventions performed, the need for informed consent was waived.
3. Results
3.1. Characteristics of patients with COVID-19 and liver test results at admission
Of 303 patients, 235 (77.6%, 235/303) had mild COVID-19 and 68 (22.4%, 68/ 303) developed severe COVID-19 at admission. One Hundred fifty-five patients were male and 148 patients were female. The mean age was 62 years (53–70 years). Table 2 showed that at admission, 148 patients (48.8%,148/303) had abnormal liver test results and 7 patients (2.4%, 7/303) had liver injury. Of 7 patients with liver injury, 2 were severe COVID-19 and 5 were mild COVID-19. Patients with liver test abnormalities tended to be male (P = .006). Age, smoke, drink and severity of COVID-19 had no significant association with liver test abnormalities (P > .05), respectively. Among these comorbidities, only hyperlipidemia showed significant association with the liver test abnormalities (P = .022). The increase in ALT, AST and GGT at admission appeared to be more pronounced, with 86 (28.4%), 96 (31.7%) and 96 (31.7%) having more than 1 × ULN respectively, and with 3 (1.0%), 2 (0.7%) and 3 (1.0%) having more than 5 × ULN respectively. However, the increases in TBIL and ALP were mild, with 21 (7.0%) and 33 (10.9%) having more than 1 × ULN respectively, and with 2 (0.7%) and 4 (1.3%) exhibiting an elevation to more than 2 × ULN respectively.
Table 2 -
Characteristics of 303 patients with COVID-19 and liver test results at admission.
| Characteristics |
Liver tests |
Total |
P value |
| Normal |
Abnormal |
Liver injury |
| Patients, n (%) |
148 (48.8) |
148 (48.8) |
7 (2.4) |
303 (100.0) |
|
| Age, yr (mean ± SD) |
60.0 ± 13.8 |
62.3 ± 13.5 |
66.1 ± 11.1 |
61.2 ± 13.6 |
.222 |
| Age, yr, n (%) |
|
|
|
|
|
| <65 |
94 (63.5) |
84 (56.8) |
3 (42.9) |
181 (59.7) |
.145 |
| ≥65 |
54 (36.5) |
64 (43.2) |
4 (57.1) |
122 (40.3) |
|
| Gender, n (%) |
|
|
|
|
|
| Male |
62 (41.9) |
90 (60.8) |
3 (42.9) |
155 (51.2) |
.006 |
| Female |
86 (58.1) |
58 (39.2) |
4 (57.1) |
148 (48.8) |
|
| Severity of COVID-19, n(%) |
|
|
|
|
|
| Mild |
121 (81.8) |
109 (73.6) |
5 (71.4) |
235 (77.6) |
.093 |
| Severe |
27 (18.2) |
39 (26.4) |
2 (28.6) |
68 (22.4) |
|
| Smoke, n (%) |
|
|
|
|
|
| No |
139 (93.9) |
137 (92.6) |
6 (85.7) |
282 (93.1) |
.461 |
| Yes |
9 (6.1) |
11 (7.4) |
1 (14.3) |
21 (6.9) |
|
| Drink, n (%) |
|
|
|
|
|
| No |
142 (95.9) |
138 (93.2) |
6 (85.7) |
286 (94.4) |
.182 |
| Yes |
6 (4.1) |
10 (6.8) |
1 (14.3) |
17 (5.6) |
|
| Comorbidities, n (%) |
|
|
|
|
|
| Diabetes |
22 (14.9) |
20 (13.5) |
1 (14.3) |
43 (14.2) |
.764 |
| Hypertension |
45 (30.4) |
44 (29.7) |
4 (57.2) |
93 (30.7) |
.602 |
| Liver disease |
13 (8.8) |
10 (6.8) |
2 (28.6) |
25 (8.2) |
.808 |
| Hyperlipidemia |
1 (0.7) |
5 (3.4) |
1 (14.3) |
7 (2.3) |
.022 |
| ALT, U/L, median (IQR) |
21.0 (14.0–34.0) |
37.0 (22.0–60.0) |
116.0 (40.5–390.5) |
26.0 (17.0–46.0) |
<.001 |
| Normal |
148 (100.0) |
67 (45.3) |
2 (28.6) |
217 (71.6) |
<.001 |
| 1–2 ULN, n (%) |
0 (0.0) |
50 (33.8) |
1 (14.3) |
51 (16.8) |
|
| 2–3 ULN, n (%) |
0 (0.0) |
18 (12.1) |
1 (14.3) |
19 (6.3) |
|
| 3–5 ULN, n (%) |
0 (0.0) |
13 (8.8) |
0 (0.0) |
13 (4.3) |
|
| >5 ULN, n (%) |
0 (0.0) |
0 (0.0) |
3 (42.8) |
3 (1.0) |
|
| AST, U/L, median (IQR) |
21.0 (17.0–29.0) |
28.5 (21.0–44.0) |
60.0 (35.5–427.5) |
28.0 (20.0–41.0) |
<.001 |
| Normal |
148 (100.0) |
57 (38.5) |
2 (28.6) |
207 (68.3) |
<.001 |
| 1–2 ULN, n (%) |
0 (0.0) |
67 (45.3) |
2 (28.6) |
69 (22.8) |
|
| 2–3 ULN, n (%) |
0 (0.0) |
13 (8.8) |
1 (14.2) |
14 (4.6) |
|
| 3–5 ULN, n (%) |
0 (0.0) |
11 (7.4) |
0 (0.0) |
11 (3.6) |
|
| >5 ULN, n (%) |
0 (0.0) |
0 (0.0) |
2 (28.6) |
2 (0.7) |
|
| TBIL, µmol/L, median (IQR) |
10.5 (8.2–12.4) |
12.3 (9.1–17.2) |
12.0 (8.8–19.8) |
10.6 (8.3–14.1) |
.174 |
| Normal |
148 (100.0) |
129 (87.2) |
5 (71.4) |
282 (93.0) |
<.001 |
| 1–2 ULN, n (%) |
0 (0.0) |
17 (11.5) |
2 (28.6) |
19 (6.3) |
|
| 2–3 ULN, n (%) |
0 (0.0) |
2 (1.3) |
0 (0.0) |
2 (0.7) |
|
| 3–5 ULN, n (%) |
0 (0.0) |
0 (0.0) |
0 (0.0) |
0 (0.0) |
|
| >5 ULN, n (%) |
0 (0.0) |
0 (0.0) |
0 (0.0) |
0 (0.0) |
|
| ALP, U/L, median (IQR) |
60.0 (50.0–72.7) |
72.0 (59.0–91.0) |
227.0 (110.0–243.5) |
63.0 (52.0–85.0) |
<.001 |
| Normal |
148 (100.0) |
120 (81.1) |
2 (28.6) |
270 (89.1) |
<.001 |
| 1–2 ULN, n (%) |
0 (0.0) |
28 (18.9) |
1 (14.3) |
29 (9.6) |
|
| 2–3 ULN, n (%) |
0 (0.0) |
0 (0.0) |
3 (42.8) |
3 (1.0) |
|
| –35, ULN, n (%) |
0 (0.0) |
0 (0.0) |
1 (14.3) |
1 (0.3) |
|
| >5 ULN, n (%) |
0 (0.0) |
0 (0.0) |
0 (0.0) |
0 (0.0) |
|
| GGT, U/L, median (IQR) |
21.0 (14.0–27.0) |
49.0 (29.0–73.0) |
137.0 (111.0–348.5) |
30.0 (18.0–57.0) |
<.001 |
| Normal |
148 (100.0) |
58 (40.9) |
1 (14.3) |
207 (68.3) |
<.001 |
| 1–2 ULN, n (%) |
0 (0.0) |
57 (38.4) |
0 (0.0) |
57 (18.8) |
|
| 2–3 ULN, n (%) |
0 (0.0) |
17 (10.7) |
2 (28.6) |
19 (6.3) |
|
| 3–5 ULN, n (%) |
0 (0.0) |
16 (10.0) |
1 (14.3) |
17 (5.6) |
|
| >5 ULN, n (%) |
0 (0.0) |
0 (0.0) |
3 (42.8) |
3 (1.0) |
|
ALP = alkaline phosphatase, ALT = alanine aminotransferase, AST = aspartate transaminase, COVID-19 = coronavirus disease 2019, GGT = gamma– glutamyltransferase, IQR = inter–quartile range, NSAIDs = non-steroidal anti-inflammatory drugs, TBIL = total bilirubinl, ULN = upper limit of normal.
3.2. Characteristics of patients with COVID-19 and liver test results during hospitalization.
Of 303 patients, 144 (47.5%) had mild COVID-19 and 159 (52.5%) had severe COVID-19 during hospitalization. The median treatment during was 24 days (14–36 days). Table 3 showed that 195 patients (64.4%, 195/303) had abnormal liver test results and 17 patients (5.6%, 17/303) had liver injury during hospitalization. In 17 patients with liver injury during hospitalization, 14 were severe COVID-19 and 3 were mild COVID-19. The presence of liver test abnormalities became more pronounced during hospitalization than at admission for ALT, AST and GGT, with 140 (46.2%), 121 (39.9%) and 134 (44.7%) patients exhibiting abnormal levels more than 1 × ULN respectively, and with 8 (2.6%), 10 (3.3%) and 7 (2.3%) patients exhibiting abnormal levels more than 5 × ULN respectively. In the meantime, the increases in TBIL and ALP were also pronounced, with 56 (18.7%) and 76 (25.3%) patients exhibiting abnormal levels more than 1 × ULN respectively, and with 12 (4.0%) and 6 (2.0%) patients exhibiting an elevation to more than 2 × ULN respectively.
Table 3 -
Characteristics of 303 patients with COVID-19 and liver test results during hospitalization.
| Characteristics |
Liver tests |
Total |
P value |
| Normal |
Abnormal |
Liver Injury |
| Patients, n (%) |
91 (30.0) |
195 (64.4) |
17 (5.6) |
303 (100.0) |
|
| Age, yr (mean ± SD) |
58.9 ± 13.8 |
61.7 ± 13.1 |
66.2 ± 16.8 |
61.2 ± 13.6 |
.074 |
| Age, yr, n (%) |
|
|
|
|
|
| <65 |
61 (67.0) |
116 (59.5) |
4 (23.5) |
181 (59.7) |
.004 |
| ≥65 |
30 (33.0) |
79 (40.5) |
13 (76.5) |
122 (40.3) |
|
| Gender, n (%) |
|
|
|
|
|
| Male |
42 (46.2) |
102 (52.3) |
11 (64.7) |
155 (51.2) |
.322 |
| Female |
49 (53.8) |
93 (47.7) |
6 (35.3) |
148 (48.8) |
|
| Severity of COVID-19, n (%) |
|
|
|
|
|
| Mild |
51 (56.0) |
90 (46.2) |
3 (17.6) |
144 (47.5) |
.012 |
| Severity |
40 (44.0) |
105 (53.8) |
14 (82.4) |
159 (52.5) |
|
| Smoke, n (%) |
|
|
|
|
|
| No |
84 (92.3) |
181 (92.8) |
17 (100.0) |
282 (93.1) |
.505 |
| Yes |
7 (7.7) |
14 (7.2) |
0 (0.0) |
21 (6.9) |
|
| Drink, n (%) |
|
|
|
|
|
| No |
86 (94.5) |
183 (93.8) |
17 (100.0) |
286 (94.4) |
.571 |
| Yes |
5 (5.5) |
12 (6.2) |
0 (0.0) |
17 (5.6) |
|
| Comorbidities, n (%) |
|
|
|
|
|
| Diabetes |
14 (15.4) |
24 (12.3) |
5 (29.4) |
43 (14.2) |
.142 |
| Hypertension |
24 (26.4) |
60 (30.8) |
9 (52.9) |
93 (30.7) |
.093 |
| Hyperlipidemia |
3 (3.3) |
4 (2.0) |
0 (0.0) |
7 (2.3) |
.653 |
| Liver disease |
11 (12.1) |
14 (7.2) |
0 (0.0) |
25 (8.2) |
.166 |
| ALT, U/L, median (IQR) |
18.0 (12.8–26.0) |
49.0 (30.0–75.7) |
147.5 (47.2–212.7) |
37.0 (20.0–66.2) |
<.001 |
| Normal |
91 (100.0) |
70 (35.9) |
2 (11.8) |
163 (53.8) |
<.001 |
| 1–2 ULN, n (%) |
0 (0.0) |
82 (42.1) |
0 (0.0) |
82 (27.1) |
|
| 2–3 ULN, n (%) |
0 (0.0) |
27 (13.8) |
2 (11.8) |
29 (9.6) |
|
| 3–5 ULN, n (%) |
0 (0.0) |
16 (8.2) |
5 (29.4) |
21 (6.9) |
|
| >5 ULN, n (%) |
0 (0.0) |
0 (0.0) |
8 (47.0) |
8 (2.6) |
|
| AST, U/L, median (IQR) |
18.5 (15.0–22.0) |
38.0 (25.0–53.7) |
104.0 (33.7–269.0) |
29.5 (20.0–48.2) |
<.001 |
| Normal |
91 (100.0) |
89 (45.6) |
2 (11.8) |
182 (60.1) |
<.001 |
| 1–2 ULN, n (%) |
0 (0.0) |
86 (44.1) |
1 (5.9) |
87 (28.7) |
|
| 2–3 ULN, n (%) |
0 (0.0) |
20 (10.3) |
2 (11.8) |
22 (7.2) |
|
| 3–5ULN, n (%) |
0 (0.0) |
0 (0.0) |
2 (11.8) |
2 (0.7) |
|
| >5 ULN, n (%) |
0 (0.0) |
0 (0.0) |
10 (58.7) |
10 (3.3) |
|
| TBIL, µmol/L, median (IQR) |
10.3 (8.4–13.5) |
15.1 (11.0–22.2) |
47.2 (18.7–60.0) |
14.0 (9.8–20.9) |
<.001 |
| Normal |
88 (100.0) |
147 (75.4) |
9 (52.9) |
244 (81.3) |
<.001 |
| 1–2 ULN, n (%) |
0 (0.0) |
42 (21.5) |
2 (11.8) |
44 (14.7) |
|
| 2–3 ULN, n (%) |
0 (0.0) |
5 (2.6) |
4 (23.5) |
9 (3.0) |
|
| 3–5 ULN, n (%) |
0 (0.0) |
1 (0.5) |
1 (5.9) |
2 (0.7) |
|
| > 5 ULN, n (%) |
0 (0.0) |
0 (0.0) |
1 (5.9) |
1 (0.3) |
|
| ALP, U/L, median (IQR) |
69.0 (51.0–70.0) |
84.5 (66.0–106.0) |
119.5 (92.2–209.0) |
76.0 (60.0–100.0) |
<.001 |
| Normal |
88 (100.0) |
132 (67.7) |
4 (23.5) |
224 (74.7) |
<.001 |
| 1–2 ULN, n (%) |
0 (0.0) |
63 (32.3) |
7 (41.2) |
70 (23.3) |
|
| 2–3 ULN, n (%) |
0 (0.0) |
0 (0.0) |
5 (29.4) |
5 (1.7) |
|
| 3–5 ULN, n (%) |
0 (0.0) |
0 (0.0) |
1 (5.9) |
1 (0.3) |
|
| > 5 ULN, n (%) |
0 (0.0) |
0 (0.0) |
0 (0.0) |
0 (0.0) |
|
| GGT, U/L, median (IQR) |
19.0 (14.0–25.0) |
53.0 (31.0–86.0) |
77.5 (36.2–140.2) |
35.0 (21.0–72.2) |
<.001 |
| Normal |
88 (100.0) |
74 (37.9) |
4 (23.5) |
166 (55.3) |
<.001 |
| 1–2 ULN, n (%) |
0 (0.0) |
76 (39.0) |
4 (23.5) |
80 (26.7) |
|
| 2–3 ULN, n (%) |
0 (0.0) |
28 (14.4) |
1 (5.9) |
29 (9.7) |
|
| 3–5 ULN, n (%) |
0 (0.0) |
13 (6.7) |
5 (29.4) |
18 (6.0) |
|
| > 5 ULN, n (%) |
0 (0.0) |
4 (2.0) |
3 (17.7) |
7 (2.3) |
|
| Concomitant medication, n (%) |
|
|
|
|
|
| Antivirals |
81 (89.0) |
188 (96.4) |
15 (88.2) |
285 (94.0) |
.035 |
| Antibacterials |
56 (61.5) |
139 (71.3) |
15 (88.2) |
210 (69.3) |
.055 |
| Chinese patent drugs |
70 (76.9) |
147 (75.4) |
11 (64.7) |
199 (65.7) |
.562 |
| Chinese herbal medicine |
62 (68.1) |
120 (61.5) |
4 (23.5) |
186 (61.4) |
.002 |
| Methylprednisolone |
17 (18.7) |
78 (40.0) |
8 (47.0) |
103 (34.0) |
.001 |
| Cardiovascular drugs |
27 (29.7) |
142 (72.8) |
13 (76.5) |
182 (60.1) |
<.001 |
| Digestive system agents |
48 (52.7) |
121 (62.0) |
9 (52.9) |
178 (58.7) |
.291 |
| Respiratory drugs |
72 (79.1) |
178 (91.3) |
14 (82.4) |
264 (87.1) |
.014 |
| NSAIDs |
12 (13.2) |
36 (18.5) |
4 (23.5) |
52 (17.2) |
.421 |
ALP = alkaline phosphatase, ALT = alanine aminotransferase, AST = aspartate transaminase, COVID-19 = coronavirus disease 2019, GGT = gamma– glutamyltransferase, IQR = inter–quartile range, NSAIDs = non-steroidal anti-inflammatory drugs, TBIL = total bilirubin, ULN = upper limit of normal.
From Table 3, gender, smoke, drink and comorbidities had no significant association with the liver test abnormalities (P > .05), respectively. However, these variables, such as age (≥65 years) (P = .004) and severity of COVID-19 (P = .012), appeared to be related to liver test abnormalities. Notably, antivirals (P = .035), Chinese herbal medicine (P = .002), methylprednisolone (P = .001), cardiovascular drugs (P < .001) and respiratory drugs (P = .014), were associated with liver test abnormalities, respectively.
3.3. Drug-induced liver injure in COVID-19 patients during hospitalization
In order to study the DILI in COVID-19 patients, of 303 patients, 168 patients had been excluded due to liver diseases and abnormal liver test results or liver injury at admission (Fig. 1). For the remaining 135 patients, 15 (11.1%, 15/135) patients developed DILI, with 6 patients of hepatocellular type, 5 patients of cholestatic type, and 4 patients of mixed type. Of 15 patients with DILI, 2 (13.3%, 2/15) had mild COVID-19 and 13 (86.7%, 13/15) had severe COVID-19 during hospitalization. Patients with DILI had their ALT, AST, ALP, GGT and TBIL values (mean ± SD) were 21.9 ± 9.1U/L, 29.4 ± 8.4U/L, 65.4 ± 20.5U/L, 27.4 ± 14.8U/L and 13.4 ± 8.1µmol/L, respectively, at baseline measurement, and their peak values during treatment were 201.7 ± 95.2U/L, 303.6 ± 285.3U/L, 169.7 ± 78.3U/L, 124.9 ± 98.6U/L, 45.3 ± 53.6 µmol/L, respectively.
Of the 15 patients with DILI, 10 were male and 5 were female, with a male to female ratio of 2:1. The mean age was 66.8 ± 17.8 years. Of 15 cases, further causality assessment revealed that 12 cases (80%, 12/15) were identified as probable and 3 cases (20%, 3/15) as possible. In the study, 6 patients (40%, 6/15) displayed mild injury pattern, 6 patients (40%, 6/15) had moderate injury pattern, and 3 patients (20%, 3/15) had severe injury pattern. Notably, 9 patients (60%, 9/15) had DILI within 1 week, 12 patients (80%, 12/15) had DILI within 2 weeks, and 15 patients (100%, 15/15) developed DILI within 4 weeks. The median time interval from the initiation of treatment to the onset of hepatotoxicity was 6 days(5–11.5days). In the meantime, 15 patients did not stop treatment, of whom twelve were also given a liver protective drug. With regards to outcome of hepatotoxicity, 5 patients (33.3%, 5/15) were cured, 5 patients (33.3%, 5/15) were improved, 5 patients died (33.3%, 5/15).
Table 4 Summarized the clinical and demographic variables between patients who did and did not develop DILI during anti-COVID-19 therapies. Patients were divided into 2 groups based on the presence (n = 15) or absence (n = 120) of DILI. As shown in Table 5, the results of univariate analysis revealed that age (≥65 years) (OR = 8.308, 95% CI [2.216–31.148], P = .002), gender (male) (OR = 3.217, 95% CI [1.034–10.008], P = .044) appeared to be related to liver injury respectively, while smoke and drink did not differ between 2 groups. Among these comorbidities, diabetes (OR = 3.500, 95% CI [1.052–11.645], P = .041) could increase the risk of liver injury, but hypertension and hyperlipidemia seemed to have no effect on the occurrence of liver injury, respectively. Notably, the incidence of DILI in the patients with severe COVID-19 (17.8%, 13/73) was higher than that of DILI in the patients with mild COVID-19 (3.2%, 2/62), meaning that severe COVID-19 (OR = 6.500, 95% CI [1.406–30.051], P = .017) could increase the occurrence of DILI during hospitalization. Among concomitant drugs, methylprednisolone (OR = 4.343, 95% CI [1.437–13.124], P = .009) was associated with a higher risk of DILI. In contrast, Chinese herbal medicine (OR = 0.189, 95% CI [0.057–0.630], P = .007) was associated with a reduced risk of DILI, while other concomitant drugs had no effect on the occurrence of liver injury. Therefore, 6 statistically significant covariates from univariate analysis were included in the final multivariate logistic regression. Interestingly, multivariate analysis showed that only methylprednisolone (OR = 4.177, 95% CI [1.106–15.771], P = .035) was the risk factor for liver injury, not severity of COVID-19 or Chinese herbal medicine. However, Chinese herbal medicine (OR = 0.251, 95% CI [0.063–0.992], P = .049) was a protective factor for liver injury.
Table 4 -
Summary of DILI and non–DILI groups during hospitalization.
| Variable |
Non–DILI group n = 120 (88.9%) |
DILI group n = 15 (11.1%) |
| Age, yr (mean ± SD) |
59.2 ± 13.5 |
66.8 ± 17.2 |
| Age, yr, n (%) |
|
|
| <65 |
81 (67.5) |
3 (20.0) |
| ≥65 |
39 (32.5) |
12 (80.0) |
| Gender, n (%) |
|
|
| Male |
46 (38.3) |
10 (66.7) |
| Female |
74 (61.7) |
5 (33.3) |
| Severity of COVID-19, n (%) |
|
|
| Mild |
60 (50.0) |
2 (13.3) |
| Severe |
60 (50.0) |
13 (86.7) |
| Smoke, n (%) |
|
|
| No |
111 (92.5) |
15 (100.0) |
| Yes |
9 (7.5) |
0 (0.0) |
| Drink, n (%) |
|
|
| No |
114 (95.0) |
15 (100.0) |
| Yes |
6 (5.0) |
0 (0.0) |
| Comorbidities, n (%) |
|
|
| Diabetes |
15 (12.5) |
5 (33.3) |
| Hypertension |
36 (30.0) |
8 (53.3) |
| Hyperlipidemia |
0 (0.0) |
0 (0.0) |
| Concomitant medication, n (%) |
|
|
| Antivirals |
111 (92.5) |
14 (93.3) |
| Antibacterials |
81 (67.5) |
13 (86.7) |
| Chinese patent drug |
94 (78.3) |
10 (66.7) |
| Chinese herbal medicine |
79 (65.8) |
4 (26.7) |
| Methylprednisolone |
31 (25.8) |
8 (53.3) |
| Cardiovascular drugs |
72 (60.0) |
13 (86.7) |
| Digestive system agents |
66 (55.0) |
8 (53.3) |
| Respiratory drugs |
98 (81.7) |
12 (80.0) |
| NSAIDs |
22 (18.3) |
4 (26.7) |
COVID-19 = coronavirus disease 2019, DILI = drug-induced liver injury, NSAIDs = non-steroidal anti-inflammatory drugs.
Table 5 -
Logistic regression analysis of independent risk factors for drug-induced liver injury.
| Variable |
Univariate analysis |
Multivariate analysis*
|
| OR (95% CI) |
P value |
OR (95% CI) |
P value |
B |
| Age (≥65 yr) |
8.308 (2.216–31.148) |
.002 |
3.866 (0.857–17.447) |
.079 |
1.352 |
| Gender (Male) |
3.217 (1.034–10.008) |
.044 |
3.092 (0.816–11.710) |
.097 |
1.129 |
| Severity of COVID-19 |
6.500 (1.406–30.051) |
.017 |
4.196 (0.710–24.784) |
.113 |
1.434 |
| Smoke |
- |
.999 |
|
|
|
| Drink |
- |
.999 |
|
|
|
| Comorbidities |
|
|
|
|
|
| Diabetes |
3.500 (1.052–11.645) |
.041 |
2.550 (0.576–11.290) |
.217 |
0.936 |
| Hypertension |
2.667 (0.899–7.907) |
.077 |
|
|
|
| Hyperlipidemia |
- |
- |
|
|
|
| Concomitant medication |
|
|
|
|
|
| Antivirals |
1.135 (0.134–9.643) |
.908 |
|
|
|
| Antibacterials |
3.130 (0.673–14.554) |
.146 |
|
|
|
| Chinese patent drug |
0.553 (0.174–1.761) |
.316 |
|
|
|
| Chinese herbal medicine |
0.189 (0.057–0.630) |
.007 |
0.251 (0.063–0.992) |
.049 |
−1.383 |
| Methylprednisolone |
4.343 (1.437–13.124) |
.009 |
4.177 (1.106–15.771) |
.035 |
1.430 |
| Cardiovascular drugs |
4.333 (0.936–20.069) |
.061 |
|
|
|
| Digestive system agents |
0.935 (0.319–2.743) |
.903 |
|
|
|
| Respiratory drugs |
0.898 (0.233–3.453) |
.876 |
|
|
|
| NSAIDs |
1.620 (0.471–5.565) |
.444 |
|
|
|
– = no calculation, B<0, OR<1 = protective factors, B=0, OR=1, not affect, B>0, OR>1 = Risk factors, CI = confidence interval, COVID-19 = coronavirus disease 2019, NSAIDs = non-steroidal anti-inflammatory drugs, OR = odds ratio.
*Values are adjusted for age, gender, severity of COVID-19, diabetes, Chinese herbal medicine and methylprednisolone.
4. Discussion
These published studies reporting clinical features of patients with COVID-19 have shown that they may develop different degrees of liver dysfunction. In these studies, the incidence of liver injury ranged from 14.8% to 53%, mainly reported by abnormal levels of ALT and AST accompanied by slightly elevated bilirubin levels.[15–18]
As shown in Table 2 and Table 3, the incidences of abnormal liver function at admission and during hospitalization are 48.8% and 64.4%, respectively, which were higher than previously reported,[15–18] mainly presenting with abnormal levels of ALT, AST and GGT accompanied by moderately elevated bilirubin levels and ALP. However, the incidences of liver injury at admission and during hospitalization were only 2.4% and 5.6%, respectively. Understanding liver injury in patients with COVID-19 is restricted by the difference in study design, study population, monitoring practices and particularly definition of liver injury.
At admission, patients with abnormal liver test results tended to be male (P = .006) and hyperlipidemia (P = .022). Age, smoke and drink had no significant association with liver test abnormalities (P > .05), respectively. The results showed that there were no other factors affecting liver test abnormality, including diabetes, hypertension, liver disease and severity of COVID-19. The mechanism of liver injury in COVID-19 infections is not well-understood yet and seems to be multifactorial,[19] Liver damage in patients with coronavirus infection might be directly caused by the viral infection of liver cells.[17] ACE2, which mainly localized in the heart, kidney, testes, and expressed at a low level in many other tissues, especially in the colon and lung, was found to be the key receptor for SARS-CoV-2 cell entry.[20] SARS-CoV-2 might directly bind to ACE2 positive cholangiocytes and cause liver injury,[5] which may partially explain the contribution of SARS-CoV-2 infection to the abnormal liver test in our patients at admission. In addition, immune-mediated inflammation, such as cytokine storm and pneumonia-associated hypoxia, might also contribute to liver injury or even develop into liver failure, especially in patients with severe COVID-19.[17]
After admission, only age (≥65 years) (P = .004) had significant association with liver test abnormalities. Unlike at admission, the abnormal liver test results had significant difference between mild and severe COVID-19 (P = .012), and the proportion of developing liver injury in severe COVID-19 patients was markedly higher than that in mild COVID-19 patients. Some studies also exhibited that DILI was also a possible contributing factor to laboratory test abnormalities. Fan et al[10] published a retrospective study on COVID-19 and observed that the utilization rate of this drug combination was significantly higher in patients with abnormal liver function tests compared with patients without liver function tests elevations (56.1% vs 25%, P = .009). In this study, 47.3% of the discharged patients showed elevated liver function tests at baseline, and 23.7% developed abnormalities during hospitalization, suggesting emerging liver injury from drugs or during the course of the infection. It was also reported that the use of multiple drugs, such as antibiotics, antivirals, antipyretics and analgesics, and traditional Chinese medicine may cause liver injury in COVID-19 patients.[9] In the study, after admission, antivirals (P = .035), Chinese herbal medicine (P = .002), methylprednisolone (P = .001), cardiovascular drugs (P < .001) and respiratory drugs (P = .014), were associated with liver test abnormalities, suggesting to closely monitor patients who received these particular therapies, especially in those old patients who had abnormal liver test results at admission. Owing to COVID-19 patients during hospitalization might receive drug treatment, the prevalence of abnormal liver tests increased from 48.8% to 64.4%, while the incidence of liver injury also increased from 2.4% to 5.6%.
After excluding these patients with liver disease and liver function abnormalities or liver injury at admission, only 15 patients during hospitalization were identified DILI. In China, drug-induced liver impairment is most frequently reported with antibiotics, and traditional Chinese medicine. [21] However, the study showed that the use of methylprednisolone, but not Chinese herbal medicine or other used drug, was associated with liver injury in patients with SARS-CoV-2 infections in the multivariable regression model. Methylprednisolone is an unexpected cause of DILI. In fact, due to its anti-inflammatory and immunosuppressive activities, it is used as treatment of choice for severe hepatitis, as well as liver disease with autoimmune features.[22] A study in 46 severe patients with COVID-19 showed that early, low-dose and short-term application of methylprednisolone was associated with better clinical outcomes, such as a faster improvement of oxygenation, shorter length of stay and hospitalization in ICU, and lower level of inflammatory indicators including IL-6 and C-reactive protein.[23] Methylprednisolone-associated liver injury usually occurs after pulse therapy, with a high daily dose of 800 mg.[24] The adaptive immune response is believed to play an important role in the mechanism of DILI.[25] Interestingly, the mean daily dosage of methylprednisolone was only 40 mg in the study. On the other hand, Cai et al[11] also suggested that Chinese herbal medicine was not associated with liver injury in patients with SARS-CoV-2 infections, which was consistent with our results. In the study, Chinese herbal medicine was a protective factor for liver injury.
Cai et al[11] reported the results of liver function tests on 417 patients with COVID-19 in Shenzhen, China, concluding that the use of hepatotoxic drugs was the most important risk factor for liver damage. They pointed out that the use of lopinavir/ritonavir was particularly dangerous and increased the chance of liver damage by 4-fold. Another case series from Wuhan reported that 55.4% of patients experienced liver injuries after treatment with lopinavir and ritonavir.[9]
Nevertheless, in the hospital, the used antivirals mainly included arbidol/ribavirin, not lopinavir/ritonavir. These antivirals could not increase the risk of liver injury in the multivariable regression model. Therefore, it warrants further investigation whether arbidol/ribavirin are safer than lopinavir/ritonavir during anti-COVID-19 therapies.
Various investigators across different studies reported a correlation between the severity of COVID-19 and the degree of liver dysfunction.[26] Additionally, severe COVID-19, preexisting liver disease and older age presented a risk for liver injury in COVID-19.[26] In the study, severe COVID-19 could increase the occurrence of liver injury (P = .017), but the further analysis exhibited that severe COVID-19 might not be the risk factor of liver injury in COVID-19 patients during hospitalization, including diabetes and age (≥65 years).
It remains controversial as to whether preexisting liver disease increases susceptibility to idiosyncratic DILI.[27] In general, preexisting hepatic disease might be one of the pathways that can be associated with liver damage in COVID-19 patients.[5] However, the study showed that liver disease had no significant association with the abnormal liver test results of patients with COVID-19, implying that liver disease was not at greater risk for liver injury in COVID-19 patients.
The study might be limited by its retrospective design. The available data were not enough or completely dependent on the case record, and researchers could not follow the patients and manage liver injury during anti-COVID-19 therapies. In addition, data on these drugs used as self–medication before developing COVID-19 pneumonias were not available from these patients. Unfortunately, we did not obtain some liver biopsy specimens of the patients with liver test abnormality or liver injury who died from the COVID-19 to manifest the study results.
5. Conclusions
Abnormal liver function results were in more than half of the patients with COVID-19, and the incidence of DILI in COVID-19 patients was 11.1% during hospitalization. Liver test abnormality or injury in COVID-19 patients might be directly caused by the viral infection or COVID-19 before admission. Severe COVID-19 could increase the occurrence of DILI (P = .017) during hospitalization, but not a risk factor of liver injury. The detrimental effects on liver injury mainly related to certain medications used during hospitalization, particularly methylprednisolone. However, Chinese herbal medicine was a protective factor for liver injury.
Acknowledgements
We thank to Department of Pharmacy, Renmin Hospital of Wuhan University for technical assistance.
Author contributions
Conceptualization: Ying Chen, Cai Shi, Haiyan Zhan, Jian Yang.
Data curation: Ying Chen, Cai Shi, Haiyan Zhan, Boning Yang, Jun Liu, Peipei Rong, Yi Luo, Jian Yang.
Formal analysis: Ying Chen, Cai Shi, Haiyan Zhan, Boning Yang, Jun Liu, Peipei Rong, Yi Luo, Jian Yang.
Funding acquisition: Ying Chen, Cai Shi, Jian Yang.
Investigation: Ying Chen, Cai Shi, Haiyan Zhan, Jian Yang.
Methodology: Ying Chen, Cai Shi, Haiyan Zhan, Boning Yang, Jun Liu, Jian Yang.
Project administration: Ying Chen, Cai Shi, Haiyan Zhan, Jian Yang.
Resources: Jian Yang.
Software: Ying Chen, Jian Yang.
Supervision: Haiyan Zhan, Jian Yang.
Validation: Ying Chen, Cai Shi, Haiyan Zhan, Jian Yang.
Writing – original draft: Ying Chen, Cai Shi, Haiyan Zhan, Boning Yang, Jian Yang.
Writing – review & editing: Ying Chen, Cai Shi, Haiyan Zhan, Boning Yang, Jun Liu, Peipei Rong, Yi Luo, Jian Yang.
References
[1]. Del RC, Malani N. 2019 novel coronavirus-important information for clinicians. JAMA. 2020;323:1039–40.
[2]. National Health Commission of the People’s Republic of China. Diagnosis and treatment scheme of new coronavirus infected pneumonia (7th). 2020.
[3]. Olry A, Meunier L, Délire B, et al. Drug-induced liver injury and COVID-19 infection: the rules remain the same. Drug Saf. 2020;43:615–7.
[4]. Xu Z, Shi L, Wang Y, et al. Pathological findings of COVID-19 associated with acute respiratory distress syndrome. Lancet Respir Med. 2020;8:420–2.
[5]. Chai X, Hu L, Zhang Y, et al. Specific ACE2 expression in cholangiocytes may cause liver damage after 2019-nCoV infection. bioRxiv. 2020.02.03.931766.
[6]. Sanders JM, Monogue ML, Jodlowski TZ, et al. Pharmacologic treatments for coronavirus disease 2019 (COVID-19): a review. JAMA. 2020;323:1824–36.
[7]. Bangash N, Patel J, Parekh D. COVID-19 and the liver: little cause for concern. Lancet Gastroenterol Hepatol. 2020;5:529–30.
[8]. Boeckmans J, Rodrigues RM, Demuyser T, et al. COVID-19 and drug–induced liver injury: a problem of plenty or a petty point? Arch Toxicol. 2020;94:1367–9.
[9]. Cai Q, Huang D, Ou P, et al. COVID-19 in a designated infectious diseases hospital outside Hubei Province, China. Allergy. 2020;75:1742–52.
[10]. Fan Z, Chen L, Li J, et al. Clinical features of COVID-19 related liver damage. Clin Gastroenterol Hepatol. 2020;18:1561–6.
[11]. Cai Q, Huang D, Yu H, et al. COVID-19: abnormal liver function tests. J Hepatol. 2020;73:566–74.
[12]. Xu N, Yang JX, Yang J. Incidence and associated risk factors of antituberculosis drug–induced hepatotoxicity among hospitalized patients in Wuhan, China. Eur J Hosp Pharm. 2022;29:217–21.
[13]. Danan G, Benichou C. Causality assessment of adverse reactions to drugs--I. A novel method based on the conclusions of international consensus meetings: application to drug-induced liver injuries. J Clin Epidemiol. 1993;46:1323–30.
[14]. Wang Y, Yang J, Zhan H, et al. The potential risk factors of nephrotoxicity during vancomycin therapy in Chinese adult patients. Eur J Hosp Pharm. 2021;28:e51–5.
[15]. Huang C, Wang Y, Li X, et al. Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China. Lancet. 2020;395:497–506.
[16]. Guan WJ, Ni ZY, Hu Y, et al. Clinical characteristics of coronavirus disease 2019 in China. N Engl J Med. 2020;382:1708–20.
[17]. Zhang C, Shi L, Wang F-S. Liver injury in COVID-19: management and challenges. Lancet Gastroenterol Hepatol. 2020;5:428–30.
[18]. Xu L, Liu J, Lu M, et al. Liver injury during highly pathogenic human coronavirus infections. Liver Int. 2020;40:998–1004.
[19]. Nurshad A. Relationship between COVID-19 infection and liver injury: a review of recent data. Front Med (Lausanne). 2020;7:458.
[20]. Hoffmann M, Kleine-Weber H, Schroeder S, et al. SARS-CoV-2 cell entry depends on ACE2 and TMPRSS2 and is blocked by a clinically proven protease inhibitor. Cell. 2020;181:271–280.e8.
[21]. Zhu Y, Niu M, Chen J, et al. Hepatobiliary and pancreatic: comparison between Chinese herbal medicine and Western medicine-induced liver injury of 1985 patients. J Gastroenterol Hepatol. 2016;31:1476–82.
[22]. Topal F, Ozaslan E, Akbulut S, et al. Methylprednisolone-induced toxic hepatitis. Ann Pharmacother. 2006;40:1868–71.
[23]. Wang Y, Jiang WW, He Q, et al. A retrospective cohort study of methyl- prednisolone therapy in severe patients with COVID-19 pneumonia. Signal Transduct Target Ther. 2020;5:57.
[24]. Miguel EZ, José PB, Aida OA, et al. Liver injury after methylprednisolone pulses: a disputable cause of hepatotoxicity. A case series and literature review. United European Gastroenterol J. 2019;7:825–37.
[25]. Stephens C, Castiella A, Gomez-Moreno EM, et al. Autoantibody presentation in drug-induced liver injury and idiopathic autoimmune hepatitis: the influence of human leucocyte antigen alleles. Pharmacogenet Genomics. 2016;26:414–22.
[26]. Alqahtani SA, Schattenberg JM. Liver injury in COVID-19: the current evidence. United Eur Gastroent. 2020;8:509–19.
[27]. Katarey D, Verma S. Drug–induced liver injury. Clin Med (Lond). 2016;16:s104–119.
