Azathioprine induced cholestatic hepatitis.

We report a case of cholestatic hepatitis developed one week after exposure to azathioprine. The subsequent prolonged cholestatic phase was followed by full clinical remission. Current knowledge on pathogenesis and epidemiology and the diagnostic challenges presented by this rare complication are discussed, followed by recommendations for monitoring and management.

Introduction

Azathioprine is a prodrug of the purine mimic antimetabolite, 6-mercaptopurine, widely used in various autoimmune diseases and organ transplantation. Little is known about the pathogenesis and epidemiology of azathioprine-induced cholestatic hepatitis, a rare but dramatic iatrogenic complication.

Case Report

A 73-year-old man presented to the emergency department at a tertiary center in South India with fever, vomiting, pruritus, and dark urine for seven days. Before six weeks, he underwent penetrating keratoplasty of the left eye at another center. Graft rejection was subsequently diagnosed and 15 days prior to the presentation, he was started on azathioprine 50 mg and prednisone 40 mg once daily. Past medical history revealed chronic bronchitis, essential hypertension, dyslipidemia, and left ventricular diastolic dysfunction for two years. The patient was on aspirin, atorvastatin, losartan, hydrochlorothiazide, and inhaled salmeterol plus fluticasone, ipratropium and salbutamol medications for the past two years with no history of adverse drug events or allergies. There was no history of alcohol consumption.

Examination revealed temperature 100.4°F, pulse rate 100/min, blood pressure 106/70 mm Hg, and respiratory rate 20/min. There was prominent icterus with no pallor, lymphadenopathy, pharyngitis, or edema. The abdomen was soft with right hypochondrial tenderness, but no organomegaly or ascites. The patient was well oriented with no focal neurologic deficits or flapping tremor.

Cholangitis, sepsis with cholestasis, acute hepatitis (of viral, drug or autoimmune aetiology), and systemic febrile illnesses like malaria, leptospirosis, and scrub typhus were considered as differential diagnosis. Investigations revealed conjugated hyperbilirubinemia, elevated transaminases and alkaline phosphatase, prolonged prothrombin time, and bilirubinuria [Table 1]. Blood smears for plasmodium and complete blood counts were unremarkable except for mild eosinophilia (549 cell/μl). Blood cultures, hepatitis B surface antigen (HBsAg) and serology for hepatitis A, C and E, human immunodeficiency virus, scrub typhus, and leptospirosis were negative. Azathioprine and atorvastatin were discontinued. An abdominal ultrasound scan revealed a normal liver and biliary tree. The patient was discharged after 72 hours of intravenous antibacterials.

Table 1

Selected serial hematologic and biochemical parameters

Thirty-five days later, he again presented with persisting low-grade fever, severe pruritus, and jaundice with pale stools. Clinical examination and abdominal ultrasound screening revealed no new findings. Worsening of hyperbilirubinemia, normalization of transaminases and prothrombin time, and improvement in alkaline phosphatase were observed [Table 1]. Anti-nuclear antibody (ANA) and liver kidney microsomal (LKM-1) antibody for autoimmune hepatitis and anti-mitochondrial antibody (AMA) for primary biliary cirrhosis were negative. Serum angiotensin-converting enzyme level was 85 U/L (normal range, 8-52 U/L). Percutaneous liver biopsy showed perivenular hepatocanalicular cholestasis with mild lobular inflammation [Figure 1].

Figure 1

Perivenular canalicular cholestasis: bile plugs within dilated canaliculi (arrows) 400×- H&E

Symptomatic treatment for pruritus with ursodiol and hydroxyzine were started from day 35. An ophthalmology consultation was reassuring; topical prednisone for the operated eye was started (the patient had stopped prednisone after the first visit). By day 42, the patient was afebrile and discharged on ursodiol. Pruritus and icterus (with corresponding biochemical parameters) subsided gradually. Follow-up at 4.5 and 7.5 months showed that the patient was asymptomatic with normal liver function tests (LFTs), except for alkaline phosphatase which showed a declining trend. Atorvastatin was restarted with monitoring in the fourth month with no untoward events.

Discussion

Hepatotoxicity due to azathioprine encompasses a spectrum of syndromes including predominant cholestasis, hypersensitivity (characterized by hepatocellular injury) and peliosis hepatis (reflecting endothelial cell injury).[1] Hepatic idiosyncratic drug reactions are believed to be due to genetic polymorphisms in drug-metabolizing pathways, or due to immunologic mechanisms. Both these mechanisms may be applicable in azathioprine-induced hepatic injury.[1]

Azathioprine-induced cholestatic hepatitis has been reported in various settings including renal and cardiac transplantation, Wegener granulomatosis, systemic lupus erythematosus, and inflammatory bowel disease. Romagnuolo et al. estimate a 9:1 male preponderance based on Western literature.[1] Reactions may be more severe in persons with pre-existing liver disease.[2] We are unaware of published literature on Indian patients with this complication.

Similar to our patient, most reported cases had acute hepatitis with cholestasis. The latency from the initial exposure to azathioprine to the onset of jaundice ranges from two weeks to three months in previous reports. One author reports shortening of the latent period on rechallenging with azathioprine, characteristic of hypersensitivity reactions.[2] Although this was our patient's first exposure to azathioprine, a short latency of seven days preceded jaundice and fever. Wide variation is also noted in the time for clinical and biochemical recovery following drug withdrawal. Our patient had peak bilirubin levels seven weeks after starting azathioprine (five weeks after stopping it), and clinical resolution at four months, with gradual subsidence of alkaline phosphatase beyond seven months. Previous authors document shorter recovery periods ranging from two weeks to two months with one case that had abnormal alkaline phosphatase and gamma-glutamyl transpeptidase beyond four months after cessation of azathioprine. Histopathological findings in most cases were canalicular cholestasis and varying degrees of inflammation. All available published cases recovered with no documentation of fulminant hepatocellular failure.[1]

Our patient was not rechallenged because there are no guidelines on rechallenge in azathioprine-related cholestatic hepatitis; the toxicity was severe, and alternative drugs were available. The experience of Davis et al. makes azathioprine rechallenge inadvisable even in milder cases.

The exclusion of other causes of hepatitis with cholestasis in our resource-limited setting was governed by clinical relevance. Epstein-Barr virus has been reported to cause cholestatic hepatitis, but not prolonged cholestasis. Splenomegaly, lymphadenopathy, and pharyngitis were absent in our patient. Cytomegalovirus infection was unlikely in the absence of characteristic nuclear and cytoplasmic inclusion bodies.[3] The temporal profile of fever concurrent with jaundice and the negative serology argue against viral hepatitis. An elevated serum angiotensin-converting enzyme raised the possibility of sarcoidosis; this enzyme is also elevated in cholestatic disorders.[4] The histological absence of granulomas and normal serum calcium levels and chest radiographs decreased the likelihood of sarcoidosis. Hepatic toxicity with fibrosis has been described with losartan.[5] However, the patient had been taking this medication for two years and it was not stopped during the episode. A final diagnosis of probable azathioprine-induced hepatitis with prolonged cholestasis was made.

Our report highlights a rare hepatic complication of azathioprine that has a wide range of latency, severity and duration, and is probably reversible. Recent guidelines on monitoring for azathioprine toxicity require weekly blood counts and LFTs until a maintenance dose is achieved, thereafter reducing to a minimum of once every three months.[6] We would also stress the need for patient education and regular clinical assessment. The syndrome can be managed with monitoring and symptomatic treatment. Rechallenge following an idiosyncratic hepatic reaction to azathioprine is usually unwarranted in the current era of multiple therapeutic alternatives for immunomodulation.