Jurnal Meso
Jurnal Meso
Jurnal Meso
*Division of Gastroenterology and Hepatology, Indiana University School of Medicine, Indianapolis, Indiana; ‡Department of
Medicine, University of Michigan, Ann Arbor, Michigan; §Department of Medicine, University of Southern California, Los Angeles,
California; kLaboratory of Pathology, National Cancer Institute, **Liver Disease Research Branch, National Institute of Diabetes
and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, Maryland; ¶Department of Medicine, University of
North Carolina at Chapel Hill, Chapel Hill, North Carolina; #Duke Clinical Research Institute, Durham, North Carolina
BACKGROUND & AIMS: Rare cases of azithromycin-induced hepatotoxicity have been reported, with variable clinical
and histologic features. We characterized clinical features and outcomes of azithromycin-
induced liver injury.
METHODS: We identified patients with azithromycin-induced liver injury from the Drug-Induced Liver
Injury Network Prospective Study who had causality scores of definite, highly likely, or prob-
able. Demographic, clinical, and laboratory data and 6-month outcomes were examined.
RESULTS: Eighteen patients (72% female; mean age, 37 y) had causality scores of definite (n [ 1), highly
likely (n [ 9), or probable (n [ 8). Common presenting symptoms were jaundice, abdominal
pain, nausea, and/or pruritus. For 16 patients, abnormal results from liver tests were first
detected 14 days after azithromycin cessation (range, 9–20 d). The median duration of azi-
thromycin treatment was 4 days (range, 2–7 d). The pattern of injury was hepatocellular in 10
patients, cholestatic in 6 patients, and mixed in 2 patients. The mean peak level of alanine
aminotransferase was 2127 IU/L, of alkaline phosphatase was 481 IU/L, and of total bilirubin
was 9.2 mg/dL. Liver histology showed ductopenia and veno-occlusive changes in a few pa-
tients. Two individuals had severe hypersensitivity cutaneous reactions. After 6 months, 8
patients had recovered, 4 patients had chronic injury, 1 patient died, and 1 patient underwent
liver transplantation (outcomes were unavailable for 4 patients). Two of the patients who died
or underwent liver transplantation had underlying chronic liver disease.
CONCLUSIONS: Azithromycin-induced liver injury occurs within 1 to 3 weeks after azithromycin initiation and
predominantly is hepatocellular in nature. Although most patients recover fully, severe cuta-
neous reactions, chronic injury, and serious complications leading to death or liver trans-
plantation can occur (ClinicalTrials.gov identifier, NCT00345930).
rug-induced liver injury (DILI) is an important a frequency of 1 to 10 per 100,000 drug prescriptions.6–8
D cause of liver disease in the United States, ac-
counting for 1% to 2% of hospital admissions and 13% of
Among antibiotics, the macrolides (erythromycin, azithro-
mycin, and clarithromycin) are some of the most
cases of acute liver failure.1,2 Hepatotoxicity is one of the commonly used. Their potential for hepatotoxicity initially
most common boxed warning placed on medications and was recognized in animals9 and later was well docu-
is a major cause for failure to receive initial regulatory mented in human beings based on reactions to various
approval or withdrawal after initial approval.3,4 Although erythromycin derivatives.10,11
numerous therapeutic classes of agents can cause idio-
syncratic DILI, antimicrobials are the most common class Abbreviations used in this paper: Alk P, alkaline phosphatase; ALT, alanine
aminotransferase; DILI, drug-induced liver injury; DILIN, Drug-Induced
of drugs implicated and account for 45% of cases in the Liver Injury Network; INR, international normalized ratio; IQR, interquartile
United States.5 The frequency of serious antibiotic- range; SJS, Stevens–Johnson syndrome; TEN, toxic epidermal necrolysis.
induced hepatotoxicity is low when compared with the
© 2015 by the AGA Institute
very large number of prescriptions written each year, 1542-3565/$36.00
but population-based estimates suggest that it occurs at http://dx.doi.org/10.1016/j.cgh.2014.07.054
370 Martinez et al Clinical Gastroenterology and Hepatology Vol. 13, No. 2
ALT] / [alkaline phosphatase (Alk P) value/upper limit of Table 1. Demographic and Clinical Characteristics of
normal Alk P]) at presentation, with R greater than 5 Individuals With Azithromycin-Induced Liver Injury
defined as hepatocellular, R less than 2 defined as
Azithromycin-induced
cholestatic, and R between 2 and 5 defined as mixed.5 All Characteristic liver injury (N ¼ 18)
co-authors had access to the study data and reviewed
and approved the final manuscript. Age, y 37 (1.7–76.1)
Female, n (%) 13 (72.2)
Self-reported race, n (%)
Statistical Analysis Caucasian 17 (94.4)
Black 1 (5.6)
The data analyses were performed by the Duke Hispanic, n (%) 2 (11.1)
BMI, kg/m2 25.2 (13.2–42.4)
Clinical Research Institute, the data coordinating center Latency, median (25th–75th)
of the DILIN. SAS version 9.2 (SAS Institute Inc, Cary, NC) Drug start to onset of symptoms 17 (1–58)
was used to aid all analyses. Continuous variables were Drug start to DILI onset 21 (2–65)
summarized with means and standard deviations or Drug stop to DILI onset 17 (1–57)
medians and interquartile ranges (IQR, 25th–75th). Types of liver injury at onset
Cholestatic 6 (33)
Categoric data were expressed as counts and pro- Mixed 2 (11)
portions. A comparison between the hepatocellular and Hepatocellular 10 (56)
cholestatic groups was performed using a nonparametric Signs and symptoms, n (%)
Mann–Whitney test. Jaundice 16 (89)
Abdominal pain 12 (67)
Nausea 11 (61)
Results Pruritus 10 (56)
Rasha 6 (33)
Fever 5 (28)
Patient Characteristics Recovery of total bilirubin
in those peak 2.5 mg/dL,
Eighteen individuals with definite (n ¼ 1), highly likely median (25th–75th)
(n ¼ 9), and probable (n ¼ 8) DILI constituted the study Days from peak to 50% 12 (8–27)
reduction in total bilirubin level
cohort (Figure 1). Although azithromycin was the
Days from peak to total bilirubin 35 (25–46)
only implicated agent in 13 cases, there was another level <2.5 mg/dL
implicated agent with a “possible” causality score in the Days from peak to 50% 12 (7–26)
remaining 5 cases (trimethoprim-sulfamethoxazole in 2 reduction in total bilirubin
cases, MRC-6 [a body-building supplement] and oral level (nonchronic subjects)
Progression to chronicity, n (%) 4 (28.6)
contraceptives in 1 case, and trimethoprim and ibuprofen
Death from liver disease, n (%) 1 (5.6)
in 1 case each). As a result, the causality score specific for Liver transplant, n (%) 1 (5.6)
azithromycin as the etiology of liver injury was definite in
1, highly likely in 7, and probable in 10 individuals.
NOTE. All values are expressed as mean with range unless otherwise specified.
The mean patient age for the group was 37 years,
BMI, body mass index.
with a range of 1.7 to 76 years (Table 1); 14 patients a
One case of toxic epidermal necrolysis and another of SJS.
were adults and 4 patients were children (ages, 1.7, 8,
11, and 14 years at onset of DILI). The majority were
female (n ¼ 13; 72%) and non-Hispanic white (n ¼ 15; Clinical Presentation and Type of Liver Injury
83%), and the average body mass index was 25.2 kg/m2
(range, 13–42 kg/m2). Alcohol consumption around the All patients were symptomatic at presentation and the
time of liver injury onset was present in 6 patients, with most common presenting symptoms were jaundice
1 patient drinking daily. Five patients had pre-existing (89%), abdominal pain (67%), nausea (61%), and pruri-
liver conditions including nonalcoholic fatty liver dis- tus (56%). The mean duration of azithromycin use was 4
ease in 2 patients, and 1 patient each with alcohol- days (range, 2–7). The median time to onset of symptoms
related liver disease, chronic graft-vs-host disease with after initiating azithromycin was 17 days (range, 1–58 d),
secondary iron overload, or biliary atresia (Table 2). Of and the median duration from azithromycin initiation to
note, 6 subjects reported previous exposure to azi- identification of laboratory abnormalities indicative of
thromycin (Table 2). The latency between starting azi- DILI was 21 days (range, 2–65 d). In 16 patients, liver test
thromycin and onset of DILI was not different among result abnormalities were first identified after stopping
these individuals as compared with 12 subjects who did azithromycin, with a median delay of 14 days (range, 9–20
not report prior exposure to azithromycin. All patients d). Overall, at enrollment, a hepatocellular pattern of liver
tested negative for acute hepatitis A, B, and C. IgG anti- injury (R > 5) was more frequent (56%) than a cholestatic
body to hepatitis E virus was tested in 14 patients, and (33%) or mixed pattern (11%).
IgM antibody to hepatitis E virus was tested in 5 patients, A summary of the age, sex, pattern of liver injury,
the results were negative in all patients. peak laboratory values, DILIN causality and severity
372 Martinez et al Clinical Gastroenterology and Hepatology Vol. 13, No. 2
Table 2. Pattern of Liver Injury With Clinical Outcomes of Each Individual in the Study Cohort (N ¼ 18)
NOTE. The pattern of liver injury was defined based on the R ratio calculated at the onset of liver injury. The DILIN causality score is the overall score, in brackets is
the final azithromycin causality score for cases with multiple drugs implicated. The strength of causal association between the implicated agent(s) and the liver
injury event is graded as definite (1), highly likely (2), and probable (3) using the DILIN causality scoring system. A DILIN severity score is assigned for each case
that ranges from 1 (mild as defined by serum enzyme increases with bilirubin level <2.5 mg/dL and INR <1.5), 2 (moderate, serum enzyme increases and bilirubin
level 2.5 mg/dL or INR 1.5), 3 (moderate-hospitalized, serum enzyme increases, bilirubin level 2.5 mg/dL or INR 1.5 and hospitalization for liver injury), 4
(severe, bilirubin level 2.5 mg/dL and signs of hepatic failure such as INR 1.5, ascites, or hepatic encephalopathy), and 5 (liver transplantation within 6 months
of onset or death as a result of liver injury).22
ANA, antinuclear antibody; Chol, cholestatic; GVHD, graft-vs-host disease; HC, hepatocellular; NA, 6-month follow-up data were not available; NAFLD, nonal-
coholic fatty liver disease; Other meds, other medication(s); TMP/SMZ, trimethoprim-sulfamethoxazole.
a
This child was enrolled as an exception case because she was not yet 2 years old at the time of enrollment.
b
This child subsequently died as a result of respiratory failure.
scores, and 6-month outcome of all 18 cases is shown in and were more likely to have a fatal outcome or require
Table 2. There was no obvious clustering of severity or transplantation (20% vs 0%) and have evidence of
other clinical characteristics according to age or sex. chronic injury 6 months after onset (30% vs 17%).
The comparison among hepatocellular, cholestatic, Features such as rash, fever, and eosinophilia occurred in
and mixed cases showed no differences with respect to both groups at similar rates.
latency, total dose, age, sex, body mass index, alcohol use, Two patients developed a severe cutaneous reaction,
or symptoms. However, there were some differences 1 with Stevens–Johnson syndrome (SJS) (patient 3) and 1
with respect to the presence of autoantibodies, severity, with toxic epidermal necrolysis (TEN) (patient 6). Both
and outcome. Patients with hepatocellular injury were were young (ages, 11 and 20 y), female, and had an
more likely to have a positive antinuclear antibody initial hepatocellular pattern of liver injury (R values, 8.0
(ANA) or anti-smooth muscle antibody (ASMA) than and 7.2) that subsequently evolved into a cholestatic
patients with cholestatic or mixed hepatitis (50% vs 0%), pattern (R value, 1.0 and 1.3). Both patients had evidence
February 2015 Azithromycin–Induced Hepatotoxicity 373
of chronic injury at the 6-month follow-up evaluation, the initial biopsy (Figure 2A and B). Veno-occlusive
and the child with SJS, whose liver injury was improving, changes and/or central venulitis were seen in 3 cases,
died of respiratory failure 9 months after initial pre- which is unusually frequent for DILI (Figure 2C and D).
sentation, which developed as part of her multisystem For example, in a previous report, we observed veno-
adverse drug reaction. occlusive changes in only 4.2% of 249 cases of sus-
pected DILI.25 Three cases showed increased numbers of
Histology intrahepatic eosinophils and none showed increased
numbers of plasma cells. The explant showed massive
Liver biopsies were performed in 8 individuals during hepatocellular necrosis with extensive ductular reaction
the liver injury event at a median duration of 15 days and only mild residual inflammation. Veno-occlusive
(range, 2–497 d) after DILI onset (explant liver sample changes were present but may have been secondary to
was also available in 1 patient). Three cases showed the hepatocellular injury. A follow-up liver biopsy was
hepatitis (1 with mild acute hepatitis, 1 with severe obtained in 3 patients: one, as noted earlier, showed
acute-on-chronic hepatitis, and 1 with moderate chronic ductopenia on a biopsy specimen taken approximately 9
hepatitis) without evidence of visible cholestasis. Three months after the first biopsy, the second showed
cases showed cholestasis with variable degrees of persistence of chronic hepatitis more than a year later,
inflammation: 2 were classified as acute cholestasis and and the third follow-up biopsy was obtained 10 days
1 was classified as cholestatic hepatitis. Of the 2 after the initial biopsy and both showed zonal necrosis
remaining cases, 1 showed zone 3 necrosis with veno- and veno-occlusive changes.
occlusive changes and 1 biopsy showed complex find-
ings including marked hemosiderosis and nodular Outcomes
regenerative hyperplasia that probably was related to a
prior bone marrow transplant. Two cases showed strik- Six-month follow-up evaluation was available for 14
ing ductopenia (<25% of portal areas with ducts)—1 patients, of whom 8 recovered from the liver injury, 4 had
case with acute cholestasis and a second case on a evidence of chronic injury, 1 died of multi-organ failure,
follow-up biopsy of the patient with acute hepatitis on and 1 underwent liver transplantation (Table 2). In
Figure 2. Liver histology in patients with azithromycin hepatotoxicity. (A) Patient 3: an 11-year-old Caucasian girl with azi-
thromycin hepatotoxicity underwent a liver biopsy at 20 days after DILI onset. A ductopenic portal area with mild inflammation
is shown. (B) Zone 3 cholestasis with both hepatocellular and canalicular (arrows) bile accumulation. Panels A and B were
taken from the same case, both H&E, 600. (C) Patient 11: a 45-year-old Caucasian man with azithromycin hepatotoxicity
underwent a liver biopsy at day 139 after DILI onset. Chronic hepatitis with perivenular inflammation and hemorrhage. H&E,
200. P, portal area; V, vein. (D) Patient 10: mild veno-occlusive changes. There is a layer of fibroinflammatory tissue inside the
vein (arrowheads). (Masson trichrome, 600).
374 Martinez et al Clinical Gastroenterology and Hepatology Vol. 13, No. 2
patients who recovered, the resolution was usually those with cholestatic pattern, in many regards the clinical
prompt and evident within 2 to 5 weeks (Table 1). The pattern matched what has been reported from isolated
median time for 50% reduction of peak serum total bili- case reports in the literature. Importantly, the latency to
rubin level was 12 days (range, 8–27 d), and the median onset is 1 to 3 weeks, and in 16 patients it was recognized
time for its normalization was 35 days (range, 25–46 d). 2 weeks after stopping azithromycin. Typical symptoms
Four patients (patients 1, 3, 6, and 11) met the defi- were fatigue, nausea, and abdominal pain, followed by
nition of chronic DILI at 6 months after onset. Both in- itching and jaundice. The onset of symptoms typically
dividuals with severe cutaneous reactions developed occurred after the course of azithromycin had been
chronic DILI. Three of the patients had hepatocellular completed, with only 2 of the 18 individuals developing
injury and jaundice initially. At 6 months, all 3 had jaundice while taking azithromycin. In addition, recovery
normal bilirubin levels but still had mild to moderate usually was prompt, averaging 2 to 5 weeks, and usually
increases in serum ALT or Alk P levels. was followed by complete resolution with completely
The 2 individuals with a fatal/liver transplant normal serum enzyme levels and no symptoms or signs of
outcome within 6 months after DILI onset were elderly, persistent liver disease.
had pre-existing chronic liver disease, and presented with However, there were 3 major exceptions to the usual
jaundice and a hepatocellular pattern of liver injury. In benign outcome of azithromycin-induced liver injury.
the first case (patient 15), a 60-year-old woman with These outcomes took 3 forms: (1) severe hypersensitivity-
underlying nonalcoholic steatohepatitis was treated with related cutaneous reactions with accompanying liver
azithromycin for 3 days for a respiratory tract infection injury, (2) evolution into chronic cholestatic liver injury
and developed fatigue, anorexia, pain, and nausea 4 days with vanishing bile duct syndrome, and (3) acute liver
later, followed by jaundice and worsening of mental sta- failure.
tus. At presentation, her serum bilirubin level was 13.8 Two young women developed severe cutaneous hy-
mg/dL, ALT level was 1579 IU/L, and Alk P level was 142 persensitivity reactions after short courses of azi-
IU/L, with an INR of 2.6. As a result of progressive hepatic thromycin, with 1 patient fulfilling the criteria for SJS and
failure, she underwent liver transplantation 1 month 1 patient fulfilling the criteria for the more severe but
after the initial presentation. The second fatal case was a related syndrome of TEN. These dramatic syndromes
76-year-old man (patient 18) with a history of daily have been reported with several medications, but most
alcohol abuse and chronic obstructive pulmonary disease typically allopurinol, carbamazepine, lamotrigine, nevi-
who was treated with a 4-day course of azithromycin for rapine, phenytoin, and phenobarbital.27–29 These severe
acute bronchitis and developed nausea followed by skin reactions also have been linked to macrolides
jaundice 1 week later, with an initial serum bilirubin level (erythromycin, azithromycin, and clarithromycin), but at
of 7.2 mg/dL, an ALT level of 8787 IU/L, and an Alk P level a much lower frequency.27,29–31 The rash generally arises
of 130 IU/L. Imaging of the liver showed a small liver with within 1 to 28 days of starting therapy and is followed by
heterogeneous texture. He died 1 week later of multi- involvement of other organs, such as liver, lung, kidney,
organ failure. or bone marrow.28 The liver injury associated with SJS
and TEN usually is mild and overshadowed by the
cutaneous manifestations. However, the liver involve-
Discussion ment can be severe and in both of the current cases
evidence of chronic injury was present. One patient died,
Azithromycin is a semisynthetic macrolide antibiotic although not from liver injury. Thus, azithromycin should
widely used in the treatment of mild to moderate sino- be considered as a potential cause of SJS or TEN.
pulmonary infections. Its popularity rests on its broad Another uncommon outcome of azithromycin-induced
spectrum of activity, rapid oral absorption, once-daily liver injury is chronic, vanishing bile duct syndrome,
administration, tissue penetration, and excellent tolera- which typically arises after an episode of severe acute
bility. A typical regimen is 250 to 500 mg given once cholestatic injury.18 After the acute symptoms begin to
daily for 4 to 7 days. Liver injury has been reported in resolve, jaundice and pruritus worsen and persist. Liver
patients taking azithromycin, but only rarely, and the biopsy typically shows marked cholestasis and relative
published literature is confined to isolated case reports. paucity of intralobular bile ducts (ductopenia). The
The reported typical pattern of injury was cholestatic prognosis is variable; some patients ultimately recover
and the clinical course usually was benign, resembling with resolution of jaundice and symptoms, although
reports of liver injury from erythromycin. Only rare frequently with persistence of mild serum enzyme in-
published cases have been severe, fatal, or resulted in creases and liver biopsy findings of relative ductopenia.
chronic liver injury.15,18,26 Other patients with vanishing bile duct syndrome develop
The current study describes the clinical presentation, severe ductopenia and biliary cirrhosis, and ultimately
course, and outcome of 18 cases of clinically apparent liver require liver transplantation because of intractable pru-
injury, prospectively identified and adjudicated as prob- ritus, jaundice, and hepatic dysfunction.32
ably, highly likely, or definitely owing to azithromycin. The most ominous outcome of azithromycin-induced
Although cases with hepatocellular injury outnumbered liver injury is acute liver failure, which seems to occur
February 2015 Azithromycin–Induced Hepatotoxicity 375
very infrequently.2,33 For example, Mindikoglu et al,33 in 4. Lasser KE, Allen PD, Woolhandler SJ, et al. Timing of new black
their analysis of liver transplantation procedures per- box warnings and withdrawals for prescription medications.
formed for drug-induced acute liver failure in the United JAMA 2002;287:2215–2220.
States, did not find any instances of azithromycin-induced 5. Chalasani N, Fontana RJ, Bonkovsky HL, et al. Causes, clinical
features, and outcomes from a prospective study of drug-
acute liver failure, which required liver transplantation. In
induced liver injury in the United States. Gastroenterology
a recent report from the United States Acute Liver Failure
2008;135:1924–1934, 1934 e1–4.
Study Group, Reuben et al2 described 1 patient who
6. Robles M, Toscano E, Cotta J, et al. Antibiotic-induced liver
developed acute liver failure as a result of a combination toxicity: mechanisms, clinical features and causality assess-
of trimethoprim/sulfamethoxazole and azithromycin. In ment. Curr Drug Saf 2010;5:212–222.
the current case series, 2 patients developed acute liver 7. George DK, Crawford DH. Antibacterial-induced hepatotoxicity.
failure with 1 patient dying and another patient requiring Incidence, prevention and management. Drug Saf 1996;15:
liver transplantation, yielding an overall severe adverse 79–85.
event rate of 11%. Both of these patients presented with 8. Andrade RJ, Tulkens PM. Hepatic safety of antibiotics used in
hepatocellular liver injury and jaundice, befitting of Hy’s primary care. J Antimicrob Chemother 2011;66:1431–1446.
Law, which states that the mortality rate of drug-induced 9. Viluksela M, Hanhijarvi H, Husband RF, et al. Comparative liver
liver injury with a hepatocellular pattern and jaundice is toxicity of various erythromycin derivatives in animals. J Antimicrob
likely to be greater than 10%. Indeed, in the current series, Chemother 1988;21(Suppl D):9–27.
10 patients fit the criteria for Hy’s Law, yielding a mor- 10. Braun P. Hepatotoxicity of erythromycin. J Infect Dis 1969;119:
tality/transplant rate of 20%. On the other hand, the 2 300–306.
patients with severe adverse events (1 liver trans- 11. Lloyd-Still JD, Sherman JO, Boggs J, et al. Erythromycin esto-
plantation, 1 death), had underlying, pre-existing liver late hepatotoxicity. Am J Dis Child 1978;132:320.
disease (1 patient was an alcoholic and 1 patient had 12. Hicks LA, Taylor TH Jr, Hunkler RJ. U.S. outpatient antibiotic
prescribing, 2010. N Engl J Med 2013;368:1461–1462.
nonalcoholic steatohepatitis). Both patients were older
13. Viluksela M, Vainio PJ, Tuominen RK. Cytotoxicity of macrolide
than age 60 and presented with a clinical syndrome of
antibiotics in a cultured human liver cell line. J Antimicrob
acute-on-chronic liver failure. Clinical narrative with
Chemother 1996;38:465–473.
biochemical profiles and outcomes of select cases are re-
14. Hopkins S. Clinical toleration and safety of azithromycin. Am J
ported in the Supplementary Material. Med 1991;91:40S–45S.
In conclusion, azithromycin-induced liver injury 15. Lockwood AM, Cole S, Rabinovich M. Azithromycin-induced
typically occurs within 2 to 3 weeks of exposure and liver injury. Am J Health Syst Pharm 2010;67:810–814.
presents with a predominantly hepatocellular pattern of 16. Longo G, Valenti C, Gandini G, et al. Azithromycin-induced
injury and, rarely, severe cutaneous skin reactions. Liver intrahepatic cholestasis. Am J Med 1997;102:217–218.
histology may show ductopenia and veno-occlusive 17. Chandrupatla S, Demetris AJ, Rabinovitz M. Azithromycin-
changes. Two individuals had a severe hypersensitivity induced intrahepatic cholestasis. Dig Dis Sci 2002;47:2186–2188.
cutaneous reaction. Although full recovery occurred in 18. Juricic D, Hrstic I, Radic D, et al. Vanishing bile duct syndrome
the majority of the cases, chronic DILI and serious out- associated with azithromycin in a 62-year-old man. Basic Clin
comes (death and liver transplantation) developed in a Pharmacol Toxicol 2010;106:62–65.
minority of patients. Individuals with underlying chronic 19. Cascaval RI, Lancaster DJ. Hypersensitivity syndrome associ-
liver disease may be at a higher risk for fatal outcomes ated with azithromycin. Am J Med 2001;110:330–331.
from azithromycin-induced liver injury, emphasizing that 20. AZITHROMYCIN. Available: http://livertox.nih.gov/Azithromycin.
drugs should be prescribed to such patients with care htm. Accessed September 25, 2014.
and for good cause. 21. Hoofnagle JH. Drug-induced liver injury network (DILIN). Hep-
atology 2004;40:773.
22. Fontana RJ, Watkins PB, Bonkovsky HL, et al. Drug-Induced
Supplementary Material Liver Injury Network (DILIN) prospective study: rationale, design
and conduct. Drug Saf 2009;32:55–68.
Note: To access the supplementary material accom- 23. Vuppalanchi R, Hayashi PH, Chalasani N, et al. Duloxetine
hepatotoxicity: a case-series from the drug-induced liver injury
panying this article, visit the online version of Clinical
network. Aliment Pharmacol Ther 2010;32:1174–1183.
Gastroenterology and Hepatology at www.cghjournal.org,
24. Rockey DC, Seeff LB, Rochon J, et al. Causality assessment in
and at http://dx.doi.org/10.1016/j.cgh.2014.07.054.
drug-induced liver injury using a structured expert opinion pro-
References cess: comparison to the Roussel-Uclaf causality assessment
1. Lee WM. Acute liver failure in the United States. Semin Liver Dis method. Hepatology 2010;51:2117–2126.
2003;23:217–226. 25. Kleiner DE, Chalasani NP, Lee WM, et al. Hepatic histological
2. Reuben A, Koch DG, Lee WM, et al. Drug-induced acute liver findings in suspected drug-induced liver injury: systematic
failure: results of a U.S. multicenter, prospective study. Hep- evaluation and clinical associations. Hepatology 2014;59:
atology 2010;52:2065–2076. 661–670.
3. Temple RJ, Himmel MH. Safety of newly approved drugs: im- 26. Suriawinata A, Min AD. A 33-year-old woman with jaundice after
plications for prescribing. JAMA 2002;287:2273–2275. azithromycin use. Semin Liver Dis 2002;22:207–210.
376 Martinez et al Clinical Gastroenterology and Hepatology Vol. 13, No. 2
27. Mockenhaupt M. Epidemiology of cutaneous adverse drug re- Rotary Building, Suite 225, Indianapolis, Indiana 46202. e-mail: nchalasa@
iupui.edu; fax: (317) 278-1949.
actions. Chem Immunol Allergy 2012;97:1–17.
28. Mockenhaupt M. The current understanding of Stevens-
Conflicts of interest
Johnson syndrome and toxic epidermal necrolysis. Expert Rev These authors disclose the following: Raj Vuppalanchi has received
Clin Immunol 2011;7:803–813; quiz 814–815. compensation as a consultant regarding drug hepatotoxicity from BMS, has
served on the speaker’s bureau for Gilead, and has received grant support
29. Saha A, Das NK, Hazra A, et al. Cutaneous adverse drug re- from Intercept and Lumena Pharmaceuticals; Robert Fontana has received
action profile in a tertiary care out patient setting in eastern India. grant support from Gilead and Vertex, and has served as a paid consultant
to Tibotec, Merck, and GSK; Naga Chalasani has served as a paid
Indian J Pharmacol 2012;44:792–797.
consultant to Abbvie, Salix, BMS, Aegerion, Lilly, Nimbus, and Merck in the
30. Mittmann N, Knowles SR, Koo M, et al. Incidence of toxic past 12 months; and has received grant support from Intercept, Cumber-
epidermal necrolysis and Stevens-Johnson Syndrome in an HIV land, Gilead, Takeda, and Enterome. The remaining authors disclose no
conflicts.
cohort: an observational, retrospective case series study. Am J
Clin Dermatol 2012;13:49–54.
31. Williams DA. Stevens-Johnson syndrome after erythromycin Funding
The Drug-Induced Liver Injury Network (https://dilin.dcri.duke.edu/) is sup-
therapy while deployed at sea. Mil Med 2000;165:636–637. ported by the National Institute of Diabetes and Digestive and Kidney Dis-
32. Reau NS, Jensen DM. Vanishing bile duct syndrome. Clin Liver eases of the National Institutes of Health as a Cooperative Agreement (U01s)
under grants: 2U01-DK065176-06 (Duke), 2U01-DK065201-06 (University of
Dis 2008;12:203–217, x. North Carolina), 2U01-DK065184-06 (Michigan), 2U01-DK065211-06 (Indi-
33. Mindikoglu AL, Magder LS, Regev A. Outcome of liver trans- ana), 5U01DK065193-04 (University of Connecticut), 5U01-DK065238-08
(University of California at San Francisco California Pacific Medical Center),
plantation for drug-induced acute liver failure in the United
1U01-DK083023-01 (University of Texas Southwestern), 1U01-DK083027-01
States: analysis of the United Network for Organ Sharing data- (Thomas Jefferson Hospital/University of Pennsylvania), 1U01-DK082992-01
base. Liver Transpl 2009;15:719–729. (Mayo), and 1U01-DK083020-01 (USC). Additional funding was provided by
CTSA grants UL1 RR025761 (Indiana), UL1 RR025747 (University of North
Carolina), UL1 RR024134 (University of Pennsylvania), UL1 RR024986
Reprint requests (Michigan), UL1 RR024982 (UTSW), and UL1 RR024150 (Mayo), and by the
Address requests for reprints to: Naga Chalasani, MD, FACG, Division of Intramural Research Program of the National Cancer Institute, National In-
Gastroenterology and Hepatology, Indiana University School of Medicine, 702 stitutes of Health.
February 2015 Azithromycin–Induced Hepatotoxicity 376.e1
ASMA, all of which were negative. Liver ultrasound was The patient was not rechallenged, did not receive
remarkable for hepatomegaly and steatosis. A comput- steroids, and did not require a liver transplant.
erized tomography scan of the abdomen showed ascites,
mild splenomegaly, and mild hepatomegaly. Comment
A liver biopsy was performed 11 months later given
persistent abnormal liver test results; the findings of the This patient developed an acute hepatitis-like syn-
liver biopsy were cirrhosis with features suggesting drome as a result of azithromycin, with a short latency
underlying steatohepatitis (nonalcoholic steatohepati- period and prolonged course. Imaging suggests that he had
tis) and scattered eosinophils. It was possible that the pre-existing cirrhosis, probably as a result of nonalcoholic
patient may have had underlying nonalcoholic fatty steatohepatitis and presented with an acute-on-chronic
liver disease and this may have had accentuated the clinical onset and had residual evidence of liver injury 1
presentation of drug-induced liver injury from and 2 years later. This case was judged as being very likely
azithromycin. (Causality assessment score: 2) owing to azithromycin.
February 2015 Azithromycin–Induced Hepatotoxicity 376.e3
Timea ALT level, U/L AST level, U/L Alk P level, U/L Total bilirubin level, mg/dL INR Comments
Total bilirubin
Timea ALT level, IU/L AST level, IU/L Alk P level, IU/L level, mg/dL INR Comments
Total bilirubin
Timea ALT level, U/L AST level, U/L Alk P level, U/L level, mg/dL INR Comments