Rosiglitazone maleate is a member of the thiazolidinedione class of oral antidiabetic medications that act to enhance sensitivity to insulin in skeletal muscle, adipose tissue, and the liver. Troglitazone, the original preparation in this class, has been associated with idiosyncratic hepatotoxicity, sometimes resulting in liver failure, the need for liver transplantation and death. The mechanism of action of these drugs is not well understood, but it is known that rosiglitazone is metabolized primarily in the liver. Premarketing trials of troglitazone demonstrated clinically significant elevations in liver enzyme levels that have not been noted with rosiglitazone, but the structures of the two agents are similar, and hepatotoxicity may be a class effect of thiazolidinediones. Although no evidence of elevated aminotransferase levels or other signs of hepatotoxicity was seen during premarketing trials, Forman and associates and Al-Salman and associates present two case studies of idiosyncratic reactions to rosiglitazone, the causes of which are not definitively understood.

Forman and associates reported a case of a 69-year-old man with coronary artery disease and New York Heart Association class II congestive heart failure who was treated with rosiglitazone for new-onset diabetes. Before therapy, the patient's hepatic biochemical markers were normal, except for the total bilirubin level (1.69 mg per mL [29 mmol per L]). He also stated that he drank two cans of beer daily. Approximately one month before hospital admission, the patient was prescribed rosiglitazone, in a dosage of 4 mg per day. Within seven days, the patient experienced nausea, anorexia, fatigue and abdominal pain. Rosiglitazone therapy was discontinued five days before hospital admission. On admission, the patient was acidotic, and he rapidly became comatose. He was diagnosed with liver failure and had markedly elevated liver enzymes and prothrombin time and a low serum albumin level. Following intensive medical care over the next two weeks, the patient improved and was eventually discharged.

Al-Salman and associates reported a case of a 61-year-old man with poorly controlled type 2 diabetes mellitus who reported nausea, vomiting, abdominal pain and dark urine approximately eight days after starting rosiglitazone therapy. Eight months earlier the patient had taken troglitazone, but that therapy was discontinued after one week because of nausea and an upset stomach. On admission, the patient had elevated liver enzymes and a low serum albumin level. His symptoms resolved after two days of hospitalization and cessation of rosiglitazone therapy. His liver tests almost normalized over eight days, and he had no signs or symptoms of liver failure. He was well after discharge.

The authors conclude that current recommendations for the use of rosiglitazone, including assessment of liver enzymes before therapy is initiated, every two months for the first 12 months and periodically thereafter are probably inadequate. Frequent monitoring of liver enzymes would be prudent, including at the start of therapy, weekly for the first two to four weeks, monthly for the next 11 months and periodically thereafter or if symptoms develop. Rosiglitazone therapy should be discontinued immediately if liver enzymes become elevated or if adverse symptoms, such as abdominal pain, anorexia, fatigue, nausea or jaundice develop.