Trovafloxacin-Induced Acute Hepatitis

Trovafloxacin, a new trifluoronaphthyridone derivative related to fluoroquinolone antimicrobial drugs, appears to be more effective than available quinolones. As a class, the most common adverse effects involve the CNS and gastrointestinal tract [1]. Liver enzyme abnormalities have been noted in 2%-3% of patients, and liver toxicity is reported infrequently [2]. Since use of this drug in the United States was authorized in December 1997, 152 cases of serious hepatic events probably related to trovafloxacin have been reported to the US Food and Drug Administration (140 cases) and to the European Agency for the Evaluation of Medicinal Products (12 cases as of July 1998). Because the frequency of serious adverse reactions to trovafloxacin was well out of proportion to that seen in other drugs in this class, the manufacturer withdrew trovafloxacin from the European markets on 15 June 1999. The drug remains available in the United States for very restricted indications [3]. We report 3 patients who had acute hepatitis after taking trovafloxacin. A 68-year-old man (patient 1) was treated with trovafloxacin (200 mg/d) for 7 days because of a respiratory infection. At the time of initiation, his liver test values were normal. He had been taking pentoxifylline for peripheral vascular disorder. One week after the end of antibiotic therapy, he presented because of a 3-kg weight loss, rash, pruritus, and dark urine, and was admitted to the hospital. Physical examination findings were normal except for jaundice. Serum chemistry indicated acute liver injury (table 1), and serology ruled out viral causes. Screening for autoantibodies was negative, and findings of an abdominal ultrasonographic examination were normal. A liver biopsy specimen showed predominantly centrozonal necrosis and inflammatory eosinophilic infiltrates. Laboratory findings at 45 days were normal. A 33-year-old man (patient 2) was treated with trovafloxacin (200 mg/d) for 7 days because of a flulike syndrome. Ten days after treatment ended, he noticed fever and dark urine. He was prescribed roxithromycin, paracetamol, and carbocisteine. On admission 5 days later, he was febrile (40°C) and jaundiced. Laboratory results indicated hepatocellular injury without evidence of viral causes (table 1). Screening for autoantibodies was negative. Findings of endoscopic retrograde cholangiography were normal. A liver biopsy showed marked centrozonal necrosis. Results of liver tests became normal during the subsequent 2 months. A 58-year-old man (patient 3) started therapy with trovafloxacin (200 mg/d for 21 days) for lobar pneumonia. He was also given codeine for cough, prednisone (30 mg/d), and ranitidine (150 mg/d) for 10 days. Fourteen days after starting therapy, he complained of asthenia and anorexia. Four days after treatment ended he presented because of fever and dark urine. He reported intake of alcohol (30 g/d). Physical examination showed no jaundice or signs of chronic liver disease. Serum chemistry indicated acute liver injury (table 1). Serology ruled out viral causes, and a screening for autoantibodies was negative. Laboratory findings were normal at 45 days. We believe that these are the first published reports of acute hepatitis due to trovafloxacin. In each case, other causes of liver injury were ruled out. Withdrawal of the drug was followed by abatement of liver dysfunction. The histologic picture was similar in biopsy specimens for both patients (patients 1 and 2) from whom a specimen was obtained, and it was consistent with drug-induced hepatic injury. It is noteworthy that there was a delay of 4–10 days between the end of treatment and the onset of hepatitis; such a delay may hinder the diagnosis of iatrogenic liver injury. The association of hepatitis with hypersensitivity manifestations and the presence of eosinophils in the liver biopsy specimen suggest an immunoallergic mechanism. Temafloxacin, another fluoroquinolone with the same core structure as trovafloxacin, was withdrawn worldwide in 1992 after the detection of severe hemolysis, associated in more than one-half of the cases with renal failure and hepatic dysfunction [4]. Both drugs share a unique difluorinated side chain that is not found in the other quinolones and that renders these drugs highly lipophilic. It is surprising that, in light of these structural similarities, no special attention has been given to the approval and postmarketing surveillance of trovafloxacin. Serum concentrations at time of presentation with trovafloxacin-induced acute hepatitis.