Liver Injury Caused by Oral Anticoagulants

A Population-based Retrospective Cohort Study

Helgi K. Björnsson; David O. Gudmundsson; Einar S. Björnsson


Liver International. 2020;40(8):1895-1900. 

In This Article

Abstract and Introduction


Background & Aims: Idiosyncratic drug-induced liver injury (DILI) is a rare adverse event. DILI caused by direct oral anticoagulants (DOACs) has been reported, however, data on the risk of DILI are limited. The aim of the study was to evaluate the frequency of DILI caused by oral anticoagulants (OACs) in a population-based setting.

Methods: A computerized database search in The National Prescription Database was performed identifying all patients in Iceland who were prescribed OACs (rivaroxaban, apixaban, dabigatran, edoxaban or warfarin) in 2008–2017. Personal identification numbers of these patients were linked with a database containing laboratory results for all hospitals and most outpatient clinics in Iceland. A medical chart review was performed in all cases where onset of liver injury followed intake of OACs. Patients with other specific causes of liver injury were excluded. Causality assessment with the RUCAM method was undertaken in cases with suspected DILI.

Results: Three cases of suspected DILI were identified. In all cases, rivaroxaban was the implicated agent among patients prescribed this product (n = 3446). All were women with a hepatocellular type of liver injury. One patient developed a suspected drug-induced autoimmune hepatitis and was treated with corticosteroids. No cases of DILI in patients on warfarin (n = 9101), apixaban (n = 1903), dabigatran (n = 1335) and edoxaban (n = 34) were identified.

Conclusions: Rivaroxaban was the only OAC associated with DILI during the 10-year study period. Approximately 1 in 1100 patients treated with rivaroxaban developed DILI. Other OACs were not associated with liver injury in this population-based study.


Drug-induced liver injury (DILI) is an important cause of liver injury and liver failure.[1] Hepatotoxicity is the most common cause of cessation of drug development in phase II/III trials as well as post-marketing withdrawals. An application to the FDA for the first direct oral anticoagulant (DOAC), ximelagatran, was withdrawn in 2006 after several cases of serious hepatotoxicity were detected in clinical trials.[2] Although the direct oral anticoagulants that followed have been widely prescribed, the data on hepatotoxicity of these agents are scarce. The body of evidence consists of case reports and case series.[3–8] A study using the adverse reporting system of the US Food and Drug Administration found a signal for liver injury caused by rivaroxaban.[9] However, the authors suggested that the hypotheses-generating approach of the study called for population-based studies to confirm and quantify the risk of DILI owing to this agent.[9] Studies on estimated risk of liver injuries in patients taking these drugs have been published in the cardiology literature but no assessment of drug causality has been performed.[10,11]

Warfarin has been the most used oral anticoagulant (OAC) for decades after it was first marketed in 1954.[12] Since its marketing only a handful of cases suspected of DILI have been reported[13] and the hepatotoxicity of warfarin can therefore be called into question. A recent register study compared liver-related hospitalizations between all oral anticoagulants in the USA.[10] This study that only relied on diagnostic codes without confirmation of diagnoses or causality assessment found a higher risk of liver injury hospitalizations associated with warfarin treatment compared to DOACs. Since all liver and hepatobiliary diseases were included and no causality assessment was performed, it is impossible to draw any conclusions from these data regarding the risk of DILI caused by these agents.[10]

Limited data exist on the absolute and relative risks of acute and clinically relevant DILI associated with the use of drugs.[14,15] However, it is of major importance to try to quantify the risk of adverse effects such as DILI in order to assess the risk-benefit ratio in commonly used drugs such as oral anticoagulants.

The aims of the current study were to evaluate the frequency of DILI caused by oral anticoagulants in a population-based setting and to determine the number needed to harm.