Dabigatran-Induced Acute Hepatitis

In the March of 2010, a 71-year-old man with a history of chronic atrial fibrillation presented to our institution with a 2-week history of progressive painless icterus, fatigue, and anorexia. He had been initiated on off-label dabigatran for stroke prevention 1 month prior to his hospital admission. Dabigatran was discontinued at the time of admission to hospital. There had been no other recent change to medications or exposure to hepatoxins. He had no premorbid history of hepatic disease.

On examination, he had prominent scleral and dermal icterus in the absence of general stigmata of chronic hepatic disease. There was no asterixis. The abdomen was nontender with no palpable masses or visceromegaly. Biochemical investigations during admission demonstrated an aspartate aminotransferase of 667 IU/L (normal <35), alanine aminotransferase of 407 IU/L (normal 0-28), γ-glutamyltranspeptidase of 479 IU/L (normal 0-32), and total alkaline phosphatase of 680 IU/L (normal 40-120). The conjugated bilirubin was 103 µmol/L (normal 0-5), international normalized ratio was 1.3, and activated partial thromboplastin time was 45 seconds (normal 22-35 s), all indicating mild synthetic dysfunction (Figure 1 ). Screening laboratory investigations for infectious, autoimmune, and metabolic hepatotoxic pathology were unremarkable. Abdominal ultrasonography was technically limited but also unremarkable for any hepatobiliary pathology. Hepatic biopsy was performed with pathology being nondiagnostic due to sample inadequacy. He was treated supportively with a gradual but complete resolution of his hepatitis and return of his clinical condition to his premorbid baseline within 2 weeks of presentation to hospital.

Figure 1.

Liver enzyme trend since starting dabigatran. Previously documented liver enzymes, prior to drug initiation, were all within normal limits.

The introduction of novel anticoagulants, including oral direct thrombin inhibitors, promises highly effective antithrombotic therapy without some of the complexities of warfarin therapy. Ximelgatran, a first-generation direct thrombin inhibitor, demonstrated similar therapeutic efficacy to warfarin with significantly less bleeding when administered to patients with atrial fibrillation. ¹ However, treatment was associated with significantly increased incidence of elevated liver enzymes in comparison to warfarin (6% vs 0.6% with P < .01). Reports of more serious hepatotoxicity also emerged. Based on these data, the drug was not approved by the Food and Drug Administration and further development of ximelgatran was discontinued. ²

Ongoing development has led to the approval of a subsequent oral direct thrombin inhibitor, dabigatran etexilate. A recent randomized controlled trial enrolled 18 000 patients with chronic atrial fibrillation and demonstrated a mean relative risk reduction of 0.66 for stroke or systemic embolization in patients taking dabigatran when compared with warfarin. ³ Dyspepsia was the only adverse event that is more common with dabigatran than with warfarin. In studies published to date, dabigatran has not been identified to cause hepatoxicity. A phase II dose finding trial of dabigatran found only mild transaminase elevation across all dose ranges assessed, although 2 patients demonstrated severe hepatitis (with transaminase levels >5 times the upper limit of normal). ⁴ There were no cases of significant hepatic dysfunction or associated morbidity or mortality. Reviewing the Health Canada database on dabigatran, there have been 5 cases of elevated liver enzymes that may have been associated with dabigatran use. ⁵

Given this information, our case is important in that it identifies a patient who presented with clinically severe hepatic dysfunction that onsets with dabigatran use, in the absence of another identifiable cause and that resolved with only withdrawal of dabigatran. Although not meeting all criteria for Hy’s Law of drug-induced liver injury, ⁶ the relationship between drug and side effect seems quite distinct. As such, it points to the need for ongoing monitoring of patients receiving this drug for rare cases of drug-induced hepatic dysfunction, especially in patients on other potentially hepatotoxic medications.