Abstract

Introduction

Mifepristone is an antiprogestin medication that is most commonly known for its use in the medical management of first trimester abortion. After a large clinical trial reported clinical and metabolic improvement associated with its use in patients with Cushing’s syndrome, mifepristone was approved by the Food and Drug Administration (FDA) for use in these patients [1]. Drug-induced liver injury (DILI) is associated with multiple mechanisms ranging from hepatocellular injury to intrahepatic cholestasis and is caused by a variety of medications, herbal products and over-the-counter supplements; however, mifepristone-induced liver injury has only been reported in one case report [2].

Case Report

A 35-year-old woman presented to our hospital with generalized abdominal pain for a month. Pain was described as dull, constant and unrelated to eating or activity. She also complained of nausea, vomiting, yellowing of her skin and generalized itching. Her medical history was significant for Cushing’s syndrome, for which she was being treated with mifepristone. She was initially started on 300 mg of mifepristone 3 months prior to presentation, which was increased to 600 mg after 3 weeks, followed by 900 mg 3 more weeks later, and most recently 1,200 mg about a month prior to presentation. Her only other medication was levothyroxine, which she had been taking at a stable dose for more than 10 years. She denied taking any other herbal medications or vitamin supplements. On initial evaluation, she was afebrile with stable hemodynamics. Physical exam revealed jaundice, scleral icterus and diffuse abdominal tenderness. Laboratory findings revealed a total bilirubin (T. Bili) of 20.9 mg/dL, alanine aminotransferase (ALT) level of 27 Units/L, aspartate aminotransferase (AST) level of 49 Units/L and alkaline phosphatase (ALP) level of 258 Units/L. Direct bilirubin was elevated at 18.6 mg/dL. Prothrombin time (PT) and international normalized ratio (INR) were within normal limits. Review of previous laboratory testing from a month prior to presentation showed normal liver enzymes and bilirubin. Abdominal ultrasound, computed tomography (CT) and magnetic resonance imaging (MRI) were normal. Viral hepatitis serologies were negative. Additional tests including anti-smooth muscle antibody (ASMA), antimitochondrial antibody (AMA), antinuclear antibody (ANA), ceruloplasmin level, iron studies and alpha-1 antitrypsin level, were normal. Transjugular liver biopsy was subsequently pursued, which showed diffuse canalicular and cytoplasmic bile stasis, consistent with drug-induced cholestasis (Fig. 1). Diagnosis of mifepristone-associated DILI was made. Mifepristone was discontinued. Ursodeoxycholic acid was started. Her liver enzymes and T. Bili plateaued over the next 2 days and her clinical symptoms completely resolved. She was then discharged in a stable hemodynamic condition. She was seen in an outpatient clinic after 10 days when her symptoms continued to be well controlled. Her laboratory findings had improved (T. Bili 8.1 mg/dL, alkaline phosphatase (ALP) 200 Units/L, AST 56 Units/L, ALT 48 Units/L). At a month’s follow-up, further improvement was noted (T. Bili 1.7 mg/dL, ALP 128 Units/L, AST 44 Units/L, ALT 68 Units/L) (Fig. 2). Mifepristone was then permanently discontinued.

Open in a separate window

Figure 1

Hematoxylin and eosin (H&E) stain of core liver biopsy showing diffuse canalicular (long arrow) and cytoplasmic (short arrow) bile stasis.

Open in a separate window

Figure 2

Graphical representation of patient’s total bilirubin level over time. The bilirubin level initially stabilized and then gradually trended down towards normal over a month’s time after discontinuation of mifepristone.

Discussion

DILI is defined as hepatotoxicity from a drug and can have several manifestations ranging from asymptomatic elevation of liver enzymes to life-threatening acute liver failure (ALF), requiring emergent evaluation for liver transplantation [3]. DILI is, in fact, the most common cause of ALF in the US, with direct acetaminophen toxicity contributing to 39% and various idiosyncratic drug reactions causing 13% of all ALF cases [4, 5]. Of these, majority cases of idiosyncratic drug reactions have been attributed to antimicrobials (45%) and central nervous system agents (15%) [6]. The true incidence of DILI is difficult to establish, owing to the uncertainty of its diagnosis and underreporting of clinically insignificant liver injury [7, 8].

DILI often happens to be the diagnosis of exclusion after evaluating for various toxic, infectious, metabolic and autoimmune causes of liver injury. The lack of a definitive laboratory or radiologic test and presence of nonspecific histopathological findings make DILI an extremely challenging diagnosis. One commonly used attribution system for diagnosis of DILI is the Roussel Uclaf Causality Assessment Method (RUCAM), which includes multiple variables such as the timing of onset of apparent liver injury, presence of risk factors, concomitant use of other medications, established toxicity of the medication in question and pattern of hepatic injury [9]. Injury pattern may be classified as hepatocellular or cholestatic, based on the ratio (R) of ALT to ALP. Hepatocellular injury, indicated by an R > 5, is the most common form (50%). Cholestatic injury (23%) is indicated by an R < 2, while mixed injury has an intermediate R value. Our patient’s laboratory tests were consistent with a cholestatic pattern of injury, eventually confirmed by liver biopsy. We calculated her RUCAM score to be 7, based on the time of onset of liver injury, improvement after discontinuation, absence of other likely causes of liver injury and evidence of a similar presentation of mifepristone-associated DILI in a recent case report, making it a “probable” cause of DILI [2, 9]. However, in the absence of rechallenge, all the categories of RUCAM could not be assessed, possibly leading to under-calculation of the overall RUCAM score, thus, making mifepristone a “highly probable” or “highly likely” cause of DILI [9]. It is important to note that although our patient’s AST and ALT fluctuated through the course of her liver injury as often seen in cases of DILI, her symptoms completely subsided and there was an eventual downtrend in her liver enzymes.

The National Institutes of Health (NIH) has established the LiverTox website (www.livertox.nih.gov) which summarizes the features of DILI and categorizes the likelihood of various medications in causing liver injury [10]. This is reported as a letter-score ranging from “A” to “E”. Mifepristone has been assigned a likelihood score of E, which makes it an unlikely cause of clinically apparent liver injury. This score was established based on its use in a single dose of 200 mg to induce abortion; however, its role in causing hepatotoxicity has not been studied well in larger doses (300 - 1,200 mg daily) that are used in the treatment of Cushing’s syndrome. According to a study involving mice, long-term use of mifepristone in larger amounts was found to cause hepatomegaly and elevation of liver enzymes. These results were more consistent in female mice [11, 12]. Since mifepristone undergoes hepatic metabolism via CYP3A4 isoenzyme, larger doses could possibly be associated with increased production of reactive metabolites causing disturbances in the complex hepatobiliary transporter systems and eventual accumulation of harmful bile constituents resulting in cholestatic liver disease [13]. These effects may be mediated by mifepristone’s complex interactions with hepatic P450 enzymes where it acts as both a substrate and inhibitor of the CYP3A4 isoenzyme [13-15]. The actual mechanism of mifepristone-induced hepatotoxicity, however, continues to remain unclear.

In conclusion, we believe that ours is the second reported case of mifepristone-induced DILI in literature [9]. Although it appears that mifepristone could be a potential hepatotoxin in larger doses, the exact molecular mechanism of mifepristone-associated DILI is poorly understood. Cautious use in higher doses over prolonged periods of time is warranted.

Acknowledgments

References