Brief Article Open Access
Copyright ©2012 Baishideng Publishing Group Co., Limited. All rights reserved.
World J Gastroenterol. Nov 7, 2012; 18(41): 5972-5978
Published online Nov 7, 2012. doi: 10.3748/wjg.v18.i41.5972
Drug-induced liver injury in hospitalized patients with notably elevated alanine aminotransferase
Hui-Min Xu, Yan Chen, Jie Xu, Quan Zhou
Hui-Min Xu, Quan Zhou, Department of Pharmacy, the 2nd Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou 310009, Zhejiang Province, China
Yan Chen, Department of Gastroenterology, the 2nd Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou 310009, Zhejiang Province, China
Jie Xu, Department of Information Technology, the 2nd Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou 310009, Zhejiang Province, China
Author contributions: Xu HM and Zhou Q conceived and designed research; Xu J collected data from the Hospital Information System; Xu HM and Chen Y conducted literature review and data interpretation; Xu HM and Zhou Q performed the research, analyzed the data and wrote the paper.
Supported by Zhejiang Provincial Bureau of Education, No. 200908690; and Zhejiang Provincial Bureau of Health, No. 2012KYA090
Correspondence to: Quan Zhou, PhD, MHA, Professor, Department of Pharmacy, the 2nd Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou 310009, Zhejiang Province, China. zhouquan142602@zju.edu.cn
Telephone: +86-571-87783891 Fax: +86-571-87783969
Received: April 11, 2012
Revised: May 31, 2012
Accepted: June 8, 2012
Published online: November 7, 2012

Abstract

AIM: To identify the proportion, causes and the nature of drug-induced liver injury (DILI) in patients with notably elevated alanine aminotransferase (ALT).

METHODS: All the inpatients with ALT levels above 10 times upper limit of normal range (ULN) were retrospectively identified from a computerized clinical laboratory database at our hospital covering a 12-mo period. Relevant clinical information was obtained from medical records. Alternative causes of ALT elevations were examined for each patient, including biliary abnormality, viral hepatitis, hemodynamic injury, malignancy, DILI or undetermined and other causes. All suspected DILI cases were causality assessed using the Council for International Organizations of Medical Sciences scale, and only the cases classified as highly probable, probable, or possible were diagnosed as DILI. Comments related to the diagnosis of DILI in the medical record and in the discharge letter for each case were also examined to evaluate DILI detection by the treating doctors.

RESULTS: A total of 129 cases with ALT > 10 ULN were identified. Hemodynamic injury (n = 46, 35.7%), DILI (n = 25, 19.4%) and malignancy (n = 21, 16.3%) were the top three causes of liver injury. Peak ALT values were lower in DILI patients than in patients with hemodynamic injury (14.5 ± 5.6 ULN vs 32.5 ± 30.7 ULN, P = 0.001). Among DILI patients, one (4%) case was classified as definite, 19 (76%) cases were classified as probable and 5 (20%) as possible according to the CIOMS scale. A hepatocellular pattern was observed in 23 (92%) cases and mixed in 2 (8%). The extent of severity of liver injury was mild in 21 (84%) patients and moderate in 4 (16%). Before discharge, 10 (40%) patients were recovered and the other 15 (60%) were improved. The improved patients tended to have a higher peak ALT (808 ± 348 U/L vs 623 ± 118 U/L, P = 0.016) and shorter treatment duration before discharge (8 ± 6 d vs 28 ± 12 d, P = 0.008) compared with the recovered patients. Twenty-two drugs and 6 herbs were found associated with DILI. Antibacterials were the most common agents causing DILI in 8 (32%) cases, followed by glucocorticoids in 6 (24%) cases. Twenty-four (96%) cases received treatment of DILI with at least one adjunctive drug. Agents for treatment of DILI included anti-inflammatory drugs (e.g., glycyrrhizinate), antioxidants (e.g., glutathione, ademetionine 1,4-butanedisulfonate and tiopronin), polyene phosphatidyl choline and herbal extracts (e.g., protoporphyrin disodium and silymarin). Diagnosis of DILI was not mentioned in the discharge letter in 60% of the cases. Relative to prevalent cases and cases from wards of internal medicine, incident cases and cases from surgical wards had a higher risk of missed diagnosis in discharge letter [odds ratio (OR) 32.7, 95%CI (2.8-374.1), and OR 58.5, 95%CI (4.6-746.6), respectively].

CONCLUSION: DILI is mostly caused by use of antibacterials and glucocorticoids, and constitutes about one fifth of hospitalized patients with ALT > 10 ULN. DILI is underdiagnosed frequently.

Key Words: Drug-induced liver injury, Abnormal liver enzyme, Alanine aminotransferase, Underdiagnosis, Adjunctive drugs



INTRODUCTION

Hepatic injury is encountered frequently in clinical practice. Acute hepatic injury can be recognized by the increased activities of aspartate aminotransferase (AST) and/or alanine aminotransferase (ALT) and can be diagnosed by the presence of ALT levels above 10 times upper limit of normal range (ULN). Viral hepatitis and toxic and ischemic hepatic injury are the most common causes of acute hepatic injury. Rarely, Wilson disease and autoimmune hepatitis can also present as acute hepatic injury[1]. The frequency of different causes of acute hepatic injury remains unknown and varies worldwide.

Among these causes of acute hepatic injury, drug-induced liver injury (DILI) represents an important challenge for physicians[2]. DILI is the leading cause of death from acute liver failure and accounts for approximately 13% of cases of acute liver failure in the United States[3,4]. Furthermore, drug-induced hepatotoxicity is one of the main reasons for postmarketing regulatory decisions, including drug withdrawal[5]. However, because there are no specific markers coupled with the highly variable clinical presentations of DILI, the recognition and diagnosis of DILI are often difficult and delayed due to the need to exclude more common competing causes of liver injury[6,7]. The real proportion and seriousness of DILI in patients with acute hepatic injury remain unknown. A previous study showed that DILI was observed in 18%-22% of medical inpatients fulfilling Council for International Organizations of Medical Sciences (CIOMS) laboratory criteria[8]. Since hepatocellular DILI is the most common type of DILI, the proportion of DILI in patients with ALT > 10 ULN might be higher than in medical inpatients.

The aim of this study was to determine the proportion of DILI in patients with ALT > 10 ULN and identify the causes and the nature of DILI. And we also wanted to know the applicability of ALT > 10 ULN as a criterion for screening DILI patients.

MATERIALS AND METHODS
Case identification

This study was carried out in a 2300-bed teaching hospital of Zhejiang University School of Medicine, China. All hospitalized patients with ALT > 10 ULN (reference value 0-50 U/L) were retrospectively identified from a computerized clinical laboratory database at our hospital covering a 12-mo period (January 2010-December 2010). The medical records for all these patients were independently reviewed by two clinical pharmacists. Alternative causes of ALT elevations were examined for each patient, including biliary abnormality, viral hepatitis, hemodynamic injury, malignancy, DILI or undetermined and other causes (such as liver surgery or trauma, and autoimmune disease). Causes of liver injury were determined based on clinical data and results of investigations such as virology detection, ultrasonography, computed tomographic scanning or magnetic resonance imaging of the liver and biliary tree, antinuclear antibody, smooth-muscle antibody, and gamma globulins detection. A history of alcohol consumption or hypotension which might cause ischemic hepatitis was also needed for diagnosis. As chronic hepatitis B is prevalent in China, presence of hepatitis B surface antigen could not exclude diagnosis of DILI, unless it was accompanied by elevated HBV DNA titer > 1 × 103[9].

DILI was suspected based on the following criteria: (1) an appropriate temporal relationship between the intake of the drug and the onset of liver injury, and between the withdrawal of the drug and the course of the reaction; and (2) exclusion of other causes of liver disease[10]. The CIOMS scale was used to assess the suspected DILI cases[11], and only the cases classified as highly probable, probable, or possible were diagnosed as DILI. Immune-mediated DILI is an important component of idiosyncratic DILI, which is characterized by presence of fever, rash, eosinophilia and autoantibodies. Extrahepatic manifestations of immune-mediated DILI such as rash, fever, arthralgia, eosinophilia and cytopenia were also reviewed. Prevalent cases referred to the DILI cases with no documented normal baseline values of aminotransferase during hospitalization. In contrast, incident cases were defined as the patients having at least one normal liver parameter prior to DILI development during the hospitalization.

Type of DILI

According to the CIOMS criteria, cases with ALT > 2 ULN alone or an ALT/alkaline phosphatase (AP) ratio ≥ 5 (ALT and AP expressed as multiples of their upper normal limit) were classified as hepatocellular, cases with ALT/AP ratio ≤ 2 as cholestatic and cases with an ALT/AP ratio between 2 and 5 as mixed liver injuries.

Severity of DILI

The severity of DILI was categorized according to the previous report with minor modifications[12]: mild [ALT was elevated, but total serum bilirubin was < 2.5 mg/dL and International Normalized Ratio (INR) was < 1.5]; moderate (ALT was elevated and serum bilirubin was ≥ 2.5 mg/dL or INR was ≥ 1.5); and severe [ALT was elevated, and total serum bilirubin ≥ 2.5 mg/dL and with at least one of the following: (1) hepatic failure (INR ≥ 1.5, ascites or encephalopathy); (2) other organ failure believed to be caused by DILI event; and (3) death or liver transplantation because of DILI event].

Drug causality assessment

Drug causality for each case of DILI was assessed according to the WHO definitions for adverse drug reactions: The drug causality was classified as probable if only one drug could be identified as causative for the DILI; it was labeled as certain if a rechallenge was positive; and labeled as possible if more than one drug could have caused the DILI. Drugs were sorted and grouped according to the Anatomical Therapeutic Chemical classification.

Outcome

Cases with a documented decline of all liver parameters < 3 ULN before discharge were labeled as recovery, whereas cases with a documented decline but with liver parameters remaining > 3 ULN were labeled as improvement. Cases without a documented decline of liver parameters before discharge were labeled aggravation.

Detection and evaluation of DILI by treating doctors

In order to evaluate DILI detection by the treating doctors, any comments related to the diagnosis of DILI in the medical record and in the discharge letter for each case were examined. With respect to cases in which DILI was mentioned, we checked whether the causative drugs were specified in these documents.

Statistical analysis

Data were expressed in mean ± SD or mean (range) unless otherwise stated, and were analyzed using the SPSS version 13.0 (Chicago, IL, United States). Continuous variables were compared using independent-sample t test between two groups and one-way analysis of variance and Post hoc tests among more than two groups. Categorical variables were compared by χ2 (Fisher’s exact) test and the likelihood ratio test. Differences were reported as statistically significant if P < 0.05.

RESULTS
Causes of increased ALT > 10 ULN

A total of 129 patients with ALT > 10 ULN were identified in this study. Their mean age was 51 years (range, 6-83 years). Table 1 shows the number of patients with ALT > 10 ULN, age, female percentage as well as peak ALT values derived from all causes. Hemodynamic injury (n = 46, 35.7%), DILI (n = 25, 19.4%) and malignancy (n = 21, 16.3%) were the top three causes of acute liver injury. The other causes included biliary abnormality (11.6%), viral hepatitis (9.3%) and undetermined and other causes (7.8%). There was no difference in gender distribution among different cause groups, but patients were older (54.2 ± 17.2 years vs 45.4 ± 16.8 years, P = 0.029) and the magnitude of ALT alteration was higher (32.5 ± 30.7 ULN vs 14.5 ± 5.6 ULN, P = 0.001) in the group caused by hemodynamic injury than in the group caused by DILI.

Table 1 Characteristics of patients with alanine aminotransferase > 10 upper limit of normal range (mean ± SD).
Causesn (%)Age (yr)Female n (%)Peak ALT values (× ULN)
Hemodynamic injury46 (25.7)54.2 ± 17.2a14 (30.4)32.5 ± 30.7b
Drug-induced liver injury25 (19.4)45.4 ± 16.89 (36.0)14.5 ± 5.6
Malignancy21 (16.3)55.8 ± 14.6a2 (9.5)17.5 ± 10.3
Biliary abnormality15 (11.6)52.3 ± 14.47 (46.7)14.3 ± 7.7
Viral hepatitis12 (9.3)47.8 ± 16.84 (33.3)18.6 ± 10.0
Undetermined and others10 (7.8)42.9 ± 12.05 (50)30.9 ± 33.5a
Total129 (100)51.1 ± 16.441 (31.8)23.0 ± 22.7
P (between different causes)-0.0850.150.004
Features of DILI

Among the DILI patients, the median age was 47 years (range, 20-83 years) and 9 (36%) were female. A total of 10 (40%) patients were 50 years or older and 3 (12%) had a history of alcohol consumption. Sixty-eight percent of patients developed DILI during hospitalization. The median duration between first exposure to the implicated agent and DILI recognition was 6 d (range, 1-40 d). The peak values for serum biochemistries (mean ± SD) were as follows: ALT, 733.7 ± 290.3 U/L; AP, 167.8 ± 120.0 U/L; and total bilirubin, 1.2 ± 1.3 mg/dL. A hepatocellular pattern was observed in 23 (92%) cases and mixed in 2 (8%) cases, and 4 (16%) cases had jaundice. Two (8%) cases had rash, 12 (48%) had fever and 6 (24%) had cytopenia. Neither arthralgia nor eosinophilia was observed. The degree of severity of the liver injury was judged to be mild in 84% and moderate in 16%. The median time to biochemical resolution ≥ 50% from the peak values following discontinuation of the offending drug was 6 d (range, 2-12 d). A definite causal relationship was found in one (4%) case, probable in 19 (76%) and possible in 5 (20%) cases. Ten (40%) cases were recovered and the other 15 (60%) were improved before discharge. Clinical features of 25 DILI patients are listed in Table 2. The improved patients tended to have a higher peak ALT and shorter treatment duration before discharge compared with the recovered patients (mean peak ALT, 808 ± 348 U/L vs 623 ± 118 U/L, P = 0.016; and mean treatment duration before discharge, 8 ± 6 d vs 28 ± 12 d, P = 0.008).

Table 2 Demographic, clinical and laboratory parameters of 25 cases of drug-induced liver injury.
VariablesOutcome
RecoveryImprovement
Age, yr44.6 ± 18.445.9 ± 16.4
Female3 (30)6 (40)
Body mass index, kg/m224.3 ± 1.621.6 ± 2.8
Alcohol use2 (20)1 (6.7)
Jaundice2 (20)2 (13.3)
Extrahepatic manifestations
Rash1 (10)1 (6.7)
Fever6 (60)6 (40)
Cytopenia2 (20)4 (26.7)
Incident cases6 (60)11 (73.3)
Surgical wards5 (50)9 (60)
Laboratory parameters, peak values
Alanine aminotransferase (U/L)623 ± 118808 ± 348a
Aspartate aminotransferase (U/L)338 ± 187636 ± 651
Alkaline phosphatase (U/L)169 ± 172167 ± 75
Total bilirubin (mg/dL)1.4 ± 1.61.1 ± 1.0
γ-glutamyl transpeptidase (U/L)190 ± 224213 ± 149
Hepatocellular type9 (90)14 (93.3)
Causality assessment
Highly probable01 (6.7)
Probable9 (90)10 (66.7)
Possible1 (10)4 (26.7)
Time to onset, d5.3 ± 3.88.9 ± 10.2
Time to enzymes resolution, d6.7 ± 2.46.0 ± 3.0
Number of adjunctive drugs2.6 ± 1.62.5 ± 0.8
Adjunctive drugs for treatment
Anti-inflammation7 (70)14 (93.3)
Antioxidants7 (70)11 (61.1)
Phospholipids6 (60)7 (46.7)
Herbs03 (20.0)
Time from treatment of liver injury to discharge, d28 ± 128 ± 6a
Causative agents

DILI was caused by a single prescription medication in 16% of the cases, by multiple agents in 40% and by herbs in 16%. Twenty-two chemical agents and 6 herbs were associated with DILI. Antibacterials were the most common class of agents associated with DILI in 8 (32%) cases. The antibacterials most often encountered were furbenicillin (n = 2), cefminox (n = 2), ornidazole (n = 2), panipenem and betamipron (n = 2), meropenem (n = 1), piperacillin and sulbactam (n = 1) and cefoxitin (n = 1). DILI was caused by glucocorticoids in 6 cases (24%), including 5 cases treated with methylprednisolone, which was the leading individual drug and one case treated with dexamethasone. Implicated causative chemical drugs in 25 subjects with DILI are shown in Table 3. Characteristics of the 6 DILI patients caused by glucocorticoids are presented in Table 4.

Table 3 Implicated causative chemical drugs in 25 subjects with drug-induced liver injury.
Classification (n)Specific drugs (n)
Proton pump inhibitors (2)Omeprazole (2)
Cardiovascular system (3)Amiodarone (1), atorvastatin (1), cinepazide (1), telmisartan (1)
Glucocorticoids (6)Methylprednisolone (5), dexamethasone (1)
Antibacterials for systemic use (8)Furbenicillin (2), cefminox (2), ornidazole (2), panipenem and betamipron (2), meropenem (1), piperacillin and sulbactam (1), cefoxitin (1)
Monoclonal antibodies (1)Bevacizumab (1)
Musculo-skeletal system (4)Diclofenac (1), celecoxib (1), parecoxib (1), baclofen (1)
Nervous system (2)Olanzapine (2), fluoxetine (1)
Table 4 Characteristics of 6 cases of drug-induced liver injury caused by glucocorticoids.
Age/sexPrincipal diseaseDrugDose (mg/d)Treatment duration (d)Peak values
Follow-up
ALT (U/L)AST (U/L)AP (U/L)TB (mg/dL)GGT (U/L)
20/FSLEMP20-16084600138832.31183Recovered
36/FChronic glomerulonephritisMP32-401710854391620.3441Improved
25/MBrain traumaMP20105152571480.61446Improved
49/MBrain traumaMP40105163631081.19190Improved
42/FDrug eruptionMP30-16024668341100.61146Improved
49/MBrain tumorDXM5-15126022822310.52424Improved
Underdiagnosis rate of DILI

DILI was mentioned in medical records of 19 (76%) cases and in the discharge letter of 10 (40%) cases. Those who had missed diagnosis of DILI in medical records also had a higher risk of missed diagnosis of DILI in discharge letters [odds ratio (OR) 1.7, 95%CI (1.1-2.5), P = 0.022]. Among the 19 cases of DILI, there were 9 (47.4%) cases without information of specified causative drugs. As shown in the discharge letter, patients from surgical wards had a higher risk of missed diagnosis of DILI compared with those from internal medicine wards [OR 58.5, 95%CI (4.6-746.6), P < 0.001]. Similarly, incident cases had a higher risk of underdiagnosis of DILI in their discharge letter compared with the prevalent cases [OR 32.7, 95%CI (2.8-374.1), P = 0.001].

Adjunctive drugs for treatment of DILI

Only one case was not treated by any hepatic disease adjunctive drugs. DILI was treated by a single medication in 12% of the cases, and by at least two agents in 84% cases (Table 5). Ten agents were used for treatment of DILI. Anti-inflammatory drugs such as glycyrrhizinate were the most common agents used for treatment of DILI, with a total of 21 (84%) cases. The antioxidants were prescribed for 18 (72%) cases, including glutathione (n = 17), ademetionine 1, 4-butanedisulfonate (n = 6) and tiopronin (n = 1). Polyene phosphatidyl choline was used in 13 (52%) cases.

Table 5 Adjunctive drugs used for treatment of drug-induced liver injury.
Classification (n)Specific drugs (n)
Anti-inflammation (21)Diammonium glycyrrhizinate (9), magnesium isoglycyrrhizinate (8), compound monoammonium glycyrrhetate S (4), compound glycyrrhizin (3)
Antioxidants (18)Glutathione (17), ademetionine 1,4-butanedisulfonate (6), tiopronin (1)
Phospholipids (13)Polyene phosphatidyl choline (13)
Herbal extrats (3)Protoporphyrin disodium (2), silymarin (1)
DISCUSSION

Among the hospitalized patients with ALT > 10 ULN, DILI was the second common cause of liver injury, accounting for 19.4% of cases. Antibacterials and glucocorticoids were the most frequently causative agents for DILI. The degree of severity of DILI was mild or moderate. Ninety-six percent of DILI cases received at least one adjunctive drug for treatment of DILI. All DILI patients were either recovered or improved when they discharged from the hospital. The underdiagnosis rate of DILI in discharge letters was 60%.

ALT and AST are enzymes highly concentrated in the liver and are sensitive indicators of hepatocyte damage. Patients with a marked increase in aminotransferase levels (> 10 ULN) typically have acute hepatic injury. The study of the causes of notably elevated AST of liver origin in the United Kingdom[13] showed that hepatic hypoxia was the most common cause (50%), followed by pancreatobiliary disease. Since viral hepatitis is prevalent in many Asian countries and it is the primary cause in 95%-100% of patients with acute hepatic failure in India[14], viral hepatitis was supposed to be the most common cause of elevated ALT > 10 ULN in China. However, our results showed that hemodynamic injury is the most common cause of notably elevated ALT (36%) similar to the results from the AST study. Viral hepatitis accounts for 9.3% patients with ALT > 10 ULN, which is more than two times higher than the proportion (3.6%) in the AST study. From the results of present study, we should take a new look at causes of acute liver injury in China. Unlike the results from the study of AST, DILI accounts for 19.4% of inpatients with ALT > 10 ULN, which is more than two times higher than the proportion of 8.8% in patients with AST > 10 ULN, indicating that ALT is a more suitable indicator for diagnosis of drug-induced hepatocellular injury than AST.

Although there can be grey areas in which a range of causes overlap, the magnitude and rate of ALT change may provide initial insight into a different diagnosis. A very high ALT (> 75 ULN) are more likely caused by ischemic or toxic liver injury[6]. In our study, we found that the average peak ALT value in patients with hemodynamic liver injury was more than two times higher than the value in patients with DILI and ALT > 50 ULN was less likely caused by DILI. However, the average peak ALT values were similar among patients with DILI, liver malignancy, biliary abnormality and viral hepatitis. Therefore, it was hard to identify these etiologies of liver injury merely according to the ALT levels.

DILI is usually recorded by spontaneous reporting in postmarketing surveillance. However, spontaneous reporting frequencies of DILI to the pharmacovigilance authorities are always low in most countries[15]. The CIOMS laboratory criteria are widely used in studies of DILI. The CIOMS laboratory criteria require at least two determinations of ALT > 2 ULN, conjugated bilirubin > 2 ULN, or combined increases of AST, AP and total bilirubin with one value > 2 ULN, which are too complicated as criteria for screening DILI in real practice. Proportion of DILI in different populations varies depending on different screening laboratory criteria. In our study, among 129 inpatients with ALT > 10 ULN, 25 patients had DILI resulting in a positive predictive value of 19.4% for a case detection based on ALT > 10 ULN. This is similar to the values of 18%-22% in medical inpatients fulfilling CIOMS laboratory criteria[8]. Because of the hepatocellular predominance of ALT, 92% of DILI cases in this study were of hepatocellular injury. Thus, positive predictive rate of drug-induced hepatocellular injury was 18% in cases with ALT > 10 ULN, which was more than two times higher than the values of 8% in patients fulfilling CIOMS laboratory criteria[16]. However, it must be noted that the criteria used in our study increased the positive predictive value for diagnosis of drug-induced hepatocellular injury at a cost of missing drug-induced cholestatic injury.

The underdiagnosis of DILI is well known[8]. In the present study, among 25 cases of DILI, comments about the diagnoses of DILI were given in only 10 (40%) cases in their discharge letters. Nineteen (76%) cases had a diagnosis of DILI during hospitalization, but causative drugs were specified in only less than half of the cases. Underdiagnosis of DILI occured more often in incident cases and cases from surgery wards. Although the diagnosis rate in our study was not optimistic according to the patients discharge letters, the rate of 40% was more than five times higher than the rate of 7% in a previous study by Meier et al[8]. This may be due to different detection criteria used, e.g., CIOMS laboratory criteria were used in that study, but ALT > 10 ULN was used in our study. Doctors might pay more attention to ALT >10 ULN than to ALT > 2 ULN, which is one of the CIOMS laboratory criteria. Underdiagnosis of DILI may lead to a severe underestimation of the prevalence and incidence of DILI, especially for the studies based on a computerized diagnosis database. It is still a worldwide problem with respect to the methods for improving the reporting rate of DILI. One of the aims of our study is to identify whether ALT >10 ULN is a good cut-off point for screening DILI patients. From the results of our study, we can conclude that a cut-off in all patients with ALT >10 ULN for screening DILI has yielded at least the same positive predictive value as the CIOMS laboratory criteria which might be more easily accepted by doctors.

Because inpatients rarely received only a single drug for treatment, which makes it difficult to determine which drug is the most responsible agent inducing liver injury even using the CIOMS criteria, we labeled possible for all suspected drugs. The drugs causing DILI vary among countries. In European countries and the United States, antibiotics, cardiovascular and central nervous system drugs are the most frequent causes of DILI[3,5,10,17]. In Asian countries, herbal and dietary supplements are often the most common causes of DILI[9,18]. The current study revealed that the most common type of drugs was antibacterials, accounting for 32% of all DILI cases. The ratio of DILI caused by antibacterials in our study is very similar to the data in outpatients of a Swedish university hospital and inpatients in Switzerland hospitals[8,15]. Discrepancies in the published data on the category of antibacterials related with DILI may attribute to the different local therapeutic strategies and prescribing behaviors in different countries[15,19]. Similar to previous studies, hepatoxicity induced by NSAIDs, omeprazole, amiodarone, statins and olanzapine was recognized. Unexpectedly, we found that 6 cases of DILI might be caused by glucocorticoids. Glucocorticoid is a relatively uncommon cause of liver injury; moreover, it is the choice of treatment for severe hepatitis. However, some cases of corticosteroid-induced hepatotoxicity, mostly induced by the high-dose methylprednisolone have been reported[20]. In our study, liver injuries were all induced by low-doses of glucocorticoids. The mechanisms of corticosteroid-induced liver injury still remains unclear. Reactivation of HBV infections, excipient of the methylprednisolone preparation and/or idiosyncrasy reactions might all be associated with the liver injury. The significance and mechanism of liver injury caused by glucocorticoids deserve more attention and need more researches. Herbal medicines seem to be the major causes of DILI in Asian countries[9]. In our study, we found that herbal medicines were responsible for 16% of DILI cases which is higher than the proportion rate in Japan[18], but much lower than the rate in Singapore[9]. This may be explained by different herbs consumptions among countries.

Up till now, therapeutic interventions for DILI remain principally limited to the cessation of the causative drugs, supportive therapy and monitoring for acute hepatic failure[21]. However, in our study, adjunctive drugs were used in 96% of DILI patients after cessation of the inducing drugs. Glycyrrhizinate, antioxidants and polyene phosphatidyl choline were the most common agents used for treatment of DILI. Data from some clinical studies showed that these drugs are beneficial to patients with DILI[22-25]. In the current study, none of the patients died from DILI and all patients were either recovered or improved when they were discharged, in spite of the fact that 4 (16%) patients had jaundice. If both drug-induced hepatocellular injury and jaundice occur at the same time, a mortality of at least 10% can be expected according to the Hy’s rule[26]. The reason why we did not detect any fatal cases may somewhat attribute to the use of adjunctive drugs. We did not find any significant difference between recovered patients and improved patients in how many adjunctive drugs were used and which drug was chosen for treating DILI. Although our sample size was small and the follow-up of the patients was done only till discharge, but not till the time of complete recovery, our findings suggest that adjunctive drugs may be effective for treatment of DILI, however, further studies are needed to determine their efficacy and safety in subjects with DILI.

Our study is limited by its retrospective nature. A detailed medical history of the DILI patients before hospitalization could not be obtained if the doctors had not described it in the medical records. For prevalent cases, our data were mainly based on the doctors’ judgments. Some patients had not undergone a complete investigation to exclude other possible causes of liver disease. We found that even among cases of DILI diagnosed by treating doctors, many critical elements needed for a diagnosis of DILI were not available. A checklist of minimum elements required for diagnosis of DILI[19] would be helpful for improving the DILI diagnosis rates and future DILI research.

In conclusion, our study demonstrates that DILI was most frequently caused by use of antibacterials and glucocorticoids and constituted about one fifth of hospitalized patients with ALT > 10 ULN. Underdiagnosis of DILI is common, especially in patients from surgical wards and incident cases. An online warning for diagnosis of DILI when a patient’ ALT was > 10 ULN and a checklist of minimum elements required for diagnosis of DILI may both be helpful for improving DILI diagnosis rates and future DILI research.

COMMENTS
Background

Drug-induced liver injury (DILI) is being increasingly recognized as an important cause of acute hepatic injury. It is the leading cause of acute liver failure in several Western countries and the most common reason for postmarketing regulatory decisions. The recognition and diagnosis of DILI are often difficult and delayed because of the lack of specific diagnostic markers coupled with the highly variable clinical presentations of DILI.

Research frontiers

The real incidence of DILI remains unknown because of the difficulty in diagnosis and the low spontaneous reporting frequency to the pharmacovigilance authorities. The focus of DILI research is how to improve the detection of DILI to better understand the causes, risk factors and nature of DILI. Screening laboratory results of patients and using diagnostic scales, such as the Council for International Organizations of Medical Sciences scale for patients with suspected DILI, may help improve the detection of DILI.

Innovations and breakthroughs

In this study, the authors identified that DILI was the second common cause of alanine aminotransferase (ALT) > 10 times upper limit of normal range (ULN) and a positive predictive value of 19.4% for DILI detection based on ALT > 10 ULN. Antibacterials were the most common causative agents for DILI. Unexpectedly, glucocorticoids, which are commonly used for severe hepatitis treatment, were found to be the second common causative agents. A high rate (60%) of underdiagnosis of DILI was found in patients with ALT > 10 ULN, especially in incident cases and cases from surgical wards.

Applications

The results suggest that underdiagnosis of DILI is frequent and ALT > 10 ULN can be used as a laboratory screening criterion to improve the detection of DILI. The significance and mechanism of causative drugs of DILI, such as glucocorticoids, await further researches.

Terminology

CIOMS scale: This scale is determined by a score based on 7 criteria, including temporal relationship, clinical course (response after withdrawal of drug), risk factors, concomitant drugs, exclusion of other non-drug etiologies, likelihood of a reaction based on package labeling, and rechallenge. The categories of suspicion are definite or highly probable (score > 8), probable (score 6-8), possible (score 3-5), unlikely (score 1-2) and excluded (score ≤ 0).

Peer review

The authors retrospectively investigated the proportion and causality assessment of possible DILI cases in hospitalized patients in a university hospital in China. The research has novelty in identifying and classifying DILI cases. The causality assessment of the DILI cases in this research also provides meaningful information about unique drugs implicated in non-fulminant DILI in the studied patients.

Footnotes

Peer reviewers: Zhang-Xu Liu, MD, PhD, Assistant Professor, Research Center for Liver diseases, Keck School of Medicine, University of Southern California, 2011 Zonal Avenue, HMR101, Los Angeles, CA 90033, United States

S- Editor Lv S L- Editor A E- Editor Zhang DN

References
1.  Dufour DR, Lott JA, Nolte FS, Gretch DR, Koff RS, Seeff LB. Diagnosis and monitoring of hepatic injury. II. Recommendations for use of laboratory tests in screening, diagnosis, and monitoring. Clin Chem. 2000;46:2050-2068.  [PubMed]  [DOI]  [Cited in This Article: ]
2.  Tajiri K, Shimizu Y. Practical guidelines for diagnosis and early management of drug-induced liver injury. World J Gastroenterol. 2008;14:6774-6785.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in CrossRef: 109]  [Cited by in F6Publishing: 97]  [Article Influence: 7.8]  [Reference Citation Analysis (0)]
3.  Chalasani N, Fontana RJ, Bonkovsky HL, Watkins PB, Davern T, Serrano J, Yang H, Rochon J. Causes, clinical features, and outcomes from a prospective study of drug-induced liver injury in the United States. Gastroenterology. 2008;135:1924-1934.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 553]  [Cited by in F6Publishing: 515]  [Article Influence: 39.5]  [Reference Citation Analysis (0)]
4.  Ostapowicz G, Fontana RJ, Schiødt FV, Larson A, Davern TJ, Han SH, McCashland TM, Shakil AO, Hay JE, Hynan L. Results of a prospective study of acute liver failure at 17 tertiary care centers in the United States. Ann Intern Med. 2002;137:947-954.  [PubMed]  [DOI]  [Cited in This Article: ]
5.  Sgro C, Clinard F, Ouazir K, Chanay H, Allard C, Guilleminet C, Lenoir C, Lemoine A, Hillon P. Incidence of drug-induced hepatic injuries: a French population-based study. Hepatology. 2002;36:451-455.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 527]  [Cited by in F6Publishing: 430]  [Article Influence: 26.4]  [Reference Citation Analysis (0)]
6.  Giannini EG, Testa R, Savarino V. Liver enzyme alteration: a guide for clinicians. CMAJ. 2005;172:367-379.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 776]  [Cited by in F6Publishing: 806]  [Article Influence: 45.6]  [Reference Citation Analysis (0)]
7.  Galan MV, Potts JA, Silverman AL, Gordon SC. The burden of acute nonfulminant drug-induced hepatitis in a United States tertiary referral center [corrected]. J Clin Gastroenterol. 2005;39:64-67.  [PubMed]  [DOI]  [Cited in This Article: ]
8.  Meier Y, Cavallaro M, Roos M, Pauli-Magnus C, Folkers G, Meier PJ, Fattinger K. Incidence of drug-induced liver injury in medical inpatients. Eur J Clin Pharmacol. 2005;61:135-143.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 124]  [Cited by in F6Publishing: 111]  [Article Influence: 7.3]  [Reference Citation Analysis (0)]
9.  Wai CT. Presentation of drug-induced liver injury in Singapore. Singapore Med J. 2006;47:116-120.  [PubMed]  [DOI]  [Cited in This Article: ]
10.  García-Cortés M, Lucena MI, Pachkoria K, Borraz Y, Hidalgo R, Andrade RJ. Evaluation of naranjo adverse drug reactions probability scale in causality assessment of drug-induced liver injury. Aliment Pharmacol Ther. 2008;27:780-789.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 108]  [Cited by in F6Publishing: 101]  [Article Influence: 7.7]  [Reference Citation Analysis (0)]
11.  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;46:1323-1330.  [PubMed]  [DOI]  [Cited in This Article: ]
12.  Fontana RJ, Watkins PB, Bonkovsky HL, Chalasani N, Davern T, Serrano J, Rochon J. Drug-Induced Liver Injury Network (DILIN) prospective study: rationale, design and conduct. Drug Saf. 2009;32:55-68.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in F6Publishing: 1]  [Reference Citation Analysis (0)]
13.  Whitehead MW, Hawkes ND, Hainsworth I, Kingham JG. A prospective study of the causes of notably raised aspartate aminotransferase of liver origin. Gut. 1999;45:129-133.  [PubMed]  [DOI]  [Cited in This Article: ]
14.  Acharya SK, Panda SK, Saxena A, Gupta SD. Acute hepatic failure in India: a perspective from the East. J Gastroenterol Hepatol. 2000;15:473-479.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 77]  [Cited by in F6Publishing: 73]  [Article Influence: 3.5]  [Reference Citation Analysis (0)]
15.  De Valle MB, Av Klinteberg V, Alem N, Olsson R, Björnsson E. Drug-induced liver injury in a Swedish University hospital out-patient hepatology clinic. Aliment Pharmacol Ther. 2006;24:1187-1195.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 136]  [Cited by in F6Publishing: 128]  [Article Influence: 9.1]  [Reference Citation Analysis (0)]
16.  Bagheri H, Michel F, Lapeyre-Mestre M, Lagier E, Cambus JP, Valdiguié P, Montastruc JL. Detection and incidence of drug-induced liver injuries in hospital: a prospective analysis from laboratory signals. Br J Clin Pharmacol. 2000;50:479-484.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 42]  [Cited by in F6Publishing: 42]  [Article Influence: 1.9]  [Reference Citation Analysis (0)]
17.  Andrade RJ, Lucena MI, Kaplowitz N, García-Muņoz B, Borraz Y, Pachkoria K, García-Cortés M, Fernández MC, Pelaez G, Rodrigo L. Outcome of acute idiosyncratic drug-induced liver injury: Long-term follow-up in a hepatotoxicity registry. Hepatology. 2006;44:1581-1588.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 208]  [Cited by in F6Publishing: 176]  [Article Influence: 13.0]  [Reference Citation Analysis (0)]
18.  Takikawa H, Murata Y, Horiike N, Fukui H, Onji M. Drug-induced liver injury in Japan: An analysis of 1676 cases between 1997 and 2006. Hepatol Res. 2009;39:427-431.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 84]  [Cited by in F6Publishing: 32]  [Article Influence: 6.5]  [Reference Citation Analysis (0)]
19.  Fontana RJ, Seeff LB, Andrade RJ, Björnsson E, Day CP, Serrano J, Hoofnagle JH. Standardization of nomenclature and causality assessment in drug-induced liver injury: summary of a clinical research workshop. Hepatology. 2010;52:730-742.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 208]  [Cited by in F6Publishing: 190]  [Article Influence: 17.3]  [Reference Citation Analysis (0)]
20.  Topal F, Ozaslan E, Akbulut S, Küçükazman M, Yüksel O, Altiparmak E. Methylprednisolone-induced toxic hepatitis. Ann Pharmacother. 2006;40:1868-1871.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 32]  [Cited by in F6Publishing: 10]  [Article Influence: 2.0]  [Reference Citation Analysis (0)]
21.  Abboud G, Kaplowitz N. Drug-induced liver injury. Drug Saf. 2007;30:277-294.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 197]  [Cited by in F6Publishing: 180]  [Article Influence: 13.1]  [Reference Citation Analysis (0)]
22.  Tsuruoka N, Abe K, Wake K, Takata M, Hatta A, Sato T, Inoue H. Hepatic protection by glycyrrhizin and inhibition of iNOS expression in concanavalin A-induced liver injury in mice. Inflamm Res. 2009;58:593-599.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 24]  [Cited by in F6Publishing: 25]  [Article Influence: 1.8]  [Reference Citation Analysis (0)]
23.  Demirbilek S, Aydin G, Yücesan S, Vural H, Bitiren M. Polyunsaturated phosphatidylcholine lowers collagen deposition in a rat model of corrosive esophageal burn. Eur J Pediatr Surg. 2002;12:8-12.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 17]  [Cited by in F6Publishing: 17]  [Article Influence: 0.9]  [Reference Citation Analysis (0)]
24.  Kaneo Y, Ogawa K, Tanaka T, Fujihara Y, Iguchi S. A protective effect of glutathione-dextran macromolecular conjugates on acetaminophen-induced hepatotoxicity dependent on molecular size. Biol Pharm Bull. 1994;17:1379-1384.  [PubMed]  [DOI]  [Cited in This Article: ]
25.  Lee CH, Park SW, Kim YS, Kang SS, Kim JA, Lee SH, Lee SM. Protective mechanism of glycyrrhizin on acute liver injury induced by carbon tetrachloride in mice. Biol Pharm Bull. 2007;30:1898-1904.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 126]  [Cited by in F6Publishing: 110]  [Article Influence: 8.4]  [Reference Citation Analysis (0)]
26.  Björnsson E, Olsson R. Outcome and prognostic markers in severe drug-induced liver disease. Hepatology. 2005;42:481-489.  [PubMed]  [DOI]  [Cited in This Article: ]  [Cited by in Crossref: 410]  [Cited by in F6Publishing: 346]  [Article Influence: 24.1]  [Reference Citation Analysis (0)]