Severe Hepatotoxicity in a Rheumatoid Arthritis Patient Switched From Leflunomide to Methotrexate

Introduction

A 44-year-old woman with rheumatoid arthritis (RA) was well maintained on methotrexate, gold, and hydroxychloroquine therapy since 1992. In March 2002, with the advice of a general physician, she started leflunomide (20 mg/day) and stopped other drugs. A marked rise in hepatic enzymes was observed after 9 months of leflunomide therapy. Leflunomide was withdrawn and her liver enzymes returned to normal after 2 months. However, the restart of methotrexate led to a significant elevation of hepatic enzymes, and the patient developed yellow discoloration of eyes, vomiting, and abdominal pain.

Early intervention with disease-modifying antirheumatic drugs (DMARDs) is considered the best treatment for RA patients.[1] Commonly used DMARDs are methotrexate, gold, sulfasalazine, antimalarials, etc, of which methotrexate is the most widely used agent. Hepatotoxicity due to methotrexate is well documented in the doses used in RA.[2,3] Leflunomide is a newer DMARD that is used for the treatment of RA, which was approved by the US Food and Drug Administration (FDA) in 1998, and the European Union followed with its approval of the drug in 1999. Clinical trials have shown risk of hepatotoxicity with leflunomide.[4,5] The hepatotoxic risk may increase further when both leflunomide and methotrexate are used in combination because of additive toxicity. Here, we describe a patient who developed severe hepatotoxicity when leflunomide was stopped and methotrexate restarted.

Case Report

A 44-year-old female patient was receiving treatment for RA from Rheumatology Outpatient Department of All India Institute of Medical Sciences, New Delhi, India, since 1992, which included methotrexate (12.5 mg/week orally), hydroxychloroquine (200 mg twice daily orally), and gold (10 mg intramuscularly monthly). She was doing well with the therapy and had no clinical or laboratory toxicity with this regimen. Her serum transaminase levels had been checked periodically. In March 2002, she was advised to start leflunomide (20 mg/day orally) and to stop other medications by a general physician. The treatment was changed primarily due to dissatisfaction of the patient with the previous regimen.

She continued leflunomide for 9 months and stopped all other drugs. During that period, the physician never followed up nor did she come to the outpatient department of All India Institute of Medical Sciences. After 9 months, she returned to our outpatient department. At this time, her laboratory investigations revealed significant elevation of the hepatic enzymes, ie, aspartate aminotransferase (AST) was 320.1 IU/L and alanine aminotransferase (ALT) was 341.7 IU/L, which were more than 6 times the upper limit of normal (ULN). Leflunomide was subsequently discontinued and her liver enzyme levels had normalized (AST, 67.5 IU/L; ALT, 55.5 IU/L) within 2 months. However, after 2 months of discontinuation of leflunomide, due to her disease symptoms (pain and swelling of the joints), methotrexate 7.5 mg/week was restarted along with folate supplementation. She had not been receiving any other drug aside from methotrexate. After 3 months of methotrexate treatment, she had developed significant elevation of the hepatic enzymes, ie, AST 951 IU/L (> 19 times ULN), ALT 771 IU/L (> 15 times ULN), alkaline phosphatase (ALP) 340 IU/L, serum bilirubin 1.2 mg/dL. Therefore, methotrexate was stopped. Over the next week, her condition deteriorated and she developed jaundice, vomiting, and abdominal pain. Her liver function test showed marked elevation of the hepatic enzyme levels with hyperbilirubinemia, ie, AST 1214 IU/L (> 24 times ULN), ALT 1297 IU/L (> 25 times ULN), ALP 321 IU/L, and serum bilirubin 3.5 mg/dL. She underwent an evaluation of her abnormal liver function test. Ultrasound of the abdomen revealed no active liver disease and no evidence of dilated common bile duct or stones in the duct. Hepatitis A, B, and C serologies were negative. Personal history revealed that she was not alcoholic, not having any chronic illness, and there was no family history of the liver disease. Subsequently, 1 month after stopping methotrexate, her liver enzyme levels returned to the normal.

It is important to mention that in the patient washout, the procedure was not performed (with cholestyramine/activated charcoal) for faster elimination of the active metabolite of leflunomide.

Normal values for liver function tests in our laboratory are as follows: AST, 5–50 IU/L; ALT, 5–50 IU/L; ALP, 80–240 IU/L; and serum bilirubin, .8–1.0 mg/dL.

Discussion

Methotrexate is a well-known hepatotoxic drug.[2,3] Leflunomide, which is a newer DMARD, has also been associated with the risk of severe liver disorders, which is evident from clinical trials and postmarketing surveillance worldwide.[6,7] The European Agency for the Evaluation of Medicinal Products had received 296 reports of hepatic reactions with leflunomide, including cirrhosis (2 patients) and liver failure (15 patients), with death occurring in 9 patients. However, the majority of hepatic reactions were associated with other confounding factors, including concomitant use of hepatotoxic drugs.[8] In 2000–2001, the Adverse Drug Reactions Advisory Committee (ADRAC) of Australia had received 32 reports of hepatic disorders with leflunomide. Twenty-six of 32 patients had elevated hepatic enzymes and 2 patients had a fatal outcome.[9]

Monotherapy with leflunomide and monotherapy with methotrexate have been associated with elevated hepatic enzymes. In a randomized, double-blind study of a comparison between leflunomide (n = 192) and methotrexate (n = 192), Strand and colleagues[10] had reported ALT levels > 3 times ULN in 4.4% of patients who were receiving leflunomide and in 2.7% of patients who were receiving methotrexate. These elevations accounted for treatment discontinuation in 7.1% of patients who were receiving leflunomide and 3.3% of patients who were receiving methotrexate. Thus, the incidence of hepatotoxicity and the number of withdrawals were higher with leflunomide as compared with methotrexate.[10]

In contrast, in another randomized, double-blind study comparing leflunomide (n = 501) and methotrexate (n = 498), Emery and colleagues[11] showed a high incidence of hepatotoxicity (hepatic enzymes levels > 3 ULN) with methotrexate (16.3%), 3-fold higher than in patients receiving leflunomide (5.4%). In addition, the number of withdrawals with methotrexate was twice as high as seen with leflunomide.[11]

Because both drugs are hepatotoxic, combination therapy has also been found to be associated with enhanced risk of hepatotoxicity.[12,13] Weinblatt and colleagues[14] had reported that after receiving combination therapy of methotrexate and leflunomide, 17% of the RA patients had ALT levels > 3 ULN, and 10% of the patients were withdrawn from the therapy on account of persistent elevation of hepatic enzymes.[14]

In a double-blind, placebo-controlled trial by Olsen and associates,[15] RA patients were switched from initial treatment (either placebo or leflunomide or methotrexate) to alternate therapy with leflunomide or methotrexate. The incidence of increased hepatic enzyme levels was 2.5 times greater in patients who were switched from leflunomide to methotrexate (8%) as compared with patients who were switched from methotrexate to leflunomide (3%).

In our patient, initially methotrexate was switched to leflunomide, which led to significant elevation in hepatic enzymes and leflunomide was stopped. The patient needed further treatment and methotrexate was restarted; consequently she developed severe hepatic injury. It is important to mention that previous use of methotrexate had not been associated with elevation of hepatic enzymes in the patient. However, resumption of methotrexate after discontinuation of leflunomide led to the development of serious liver injury in the patient. Because both drugs are hepatotoxic, we cannot be sure which drug was responsible for this adverse reaction. It is worth noting that leflunomide is a prodrug with active metabolite, which can persist in the blood for 2 years after withdrawal of leflunomide, and side effects can develop several weeks after discontinuation of the drug. Thus, it was likely present in the blood of the patient when methotrexate was reinitiated. Hence, hepatotoxicity could have been due to the interaction between the active metabolite of leflunomide and methotrexate. In addition, she had no other related medical conditions or risk factors, which could account for the significant elevation of hepatic enzymes. Our case indicates that switching from methotrexate to leflunomide or leflunomide to methotrexate can lead to severe liver disorders in RA patients. This case supports the fact that liver function monitoring is essential whenever these drugs are prescribed. Furthermore, for the prevention and treatment of severe adverse reactions due to the leflunomide, the washout procedure must take place. The washout procedure hastens the elimination of the active metabolite of leflunomide because it undergoes enterohepatic circulation and has a long half-life. These guidelines should be followed if a severe, undesirable effect occurs during leflunomide treatment as well as when therapy has to be changed from leflunomide to another DMARD.[12]