Case Report

Top Case Report Discussion Author & Article Info References

A 75-year-old Asian woman came to the emergency department with jaundice. She had noted the onset of dark urine and pale stools 3 weeks previously, and her family noticed that she was jaundiced 1 week before. She reported nausea and anorexia that had started approximately 8 months before the onset of jaundice, concomitant with the initiation of a course of trazodone therapy, 150 mg daily, prescribed for depression. She had no previous history of jaundice, blood transfusion, intravenous drug abuse, parenteral drug therapy, acupuncture treatment, or alcohol use. She had no significant past medical problem other than a cholecystectomy 15 years previously. The only medication she was taking was trazodone. She reported no history of herbal medicine ingestion.

The physical examination at admission was unremarkable except for jaundice. No hepatosplenomegaly, abdominal tenderness, or ascites were present. Stigmata of chronic liver disease were absent. A rectal examination showed guaiac-negative pale stool.

The hemoglobin concentration was 136 g/L. The prothrombin time and partial thromboplastin times were elevated at 13.9 s and 40.8 s, respectively (normal, < 11.7 s and < 36 s, respectively). The bilirubin level was 306 µmol/L (normal, < 20 µmol/L), with a conjugated bilirubin level of 175 µmol/L (normal, < 5 µmol/L). Initial liver enzyme levels were as follows: alanine aminotransferase, 15.55 µkat/L (normal, < 0.58 µkat/L); aspartate aminotransferase, 14.05 µkat/L (normal, < 0.53 µkat/L); -glutamyltransferase, 6.33 µkat/L (normal, < 0.58 µkat/L); and alkaline phosphatase, 2.4 µkat/L (normal, < 2.4 µkat/L). Results of chest and abdominal radiographs and an abdominal ultrasound scan were normal.

Serologic tests for hepatitis A virus (IgM anti-HAV) and hepatitis B surface and e antigens were negative. An assay for IgM antibody against hepatitis B core antigen was also negative, but IgG antibodies against hepatitis B surface and core antigens were detected. Serologic test results for hepatitis C virus were negative using a first-generation enzyme-linked immunosorbent assay (Ortho Diagnostics, Raritan, New Jersey) and, subsequently, a second-generation recombinant immunoblot assay (Ortho). Serologic tests for Q fever, cytomegalovirus, and herpes simplex, Epstein-Barr, and varicella zoster viruses were all negative. Serologic tests for antimitochondrial, antismooth muscle, and antinuclear antibodies were also negative.

The prothrombin time and partial thromboplastin times remained elevated despite a 3-day course of therapy with subcutaneous vitamin K. Because of the abnormal coagulation factors, a transjugular approach was used for liver biopsy. Measurements of hepatic venous pressures done in conjunction with biopsy showed a modestly elevated hepatic venous pressure gradient (corresponding to portal pressure) of 9 mm Hg (normal, 1 to 4 mm Hg). The liver tissue was fragmented and nodular and showed a pattern consistent with chronic active hepatitis (Figure 1). Between the nodules there was reticulin condensation but no deposition of elastica, indicating that the lesion was relatively recent. The portal tracts were expanded by an infiltrate of lymphocytes and plasma cells, with erosion of limiting plates and portoportal and portocentral bridging necrosis. Eosinophils were rare. The parenchyma showed thickening of liver cell plates, occasional acidophil bodies, mild focal inflammation, low-grade parenchymal and reticuloendothelial hemosiderosis, and no -1-antitrypsin bodies or metallothionein. Centrilobular areas showed liver cell dropout and lipofuscin-laden macrophages that stained strongly positive with diastase-periodic acid-Schiff. Immunostains for hepatitis B surface and core antigens were negative.

View larger version (145K): [in this window] [in a new window]   Figure 1. Liver specimen showing the pattern of chronic active hepatitis. Two parenchymal nodules are separated by an expanded portal tract. Piecemeal necrosis and thickening of liver cell plates are evident.

Trazodone therapy was discontinued and within 1 week the patient's nausea and anorexia had resolved. Ten days after stopping the drug, the aminotransferase enzyme levels had markedly decreased and by 4 weeks had returned to normal. The prothrombin and partial thromboplastin times returned to normal within 2 weeks. The bilirubin and -glutamyltransferase levels were slow to decrease but gradually declined, returning to normal approximately 6 months after the last dose of trazodone. The patient has been followed for 2 years since resolution and has showed no further clinical or laboratory evidence of hepatic dysfunction.

Discussion

Top Case Report Discussion Author & Article Info References

To our knowledge, this is the first reported case of trazodone-induced chronic active hepatitis. Chu and colleagues [1] described a case of acute trazodone-induced hepatotoxicity. Their patient developed an exfoliative, macular, pruritic rash and elevated liver enzyme levels 3 weeks after starting a course of trazodone therapy (500 mg daily). The liver biopsy specimen showed a mixed hepatocellular-cholestatic pattern [1]. Sheikh and Nies [2] described a patient who developed intrahepatic cholestasis 2 weeks after starting trazodone therapy (50 mg daily). In these two patients, symptoms resolved rapidly after discontinuation of therapy, and liver enzyme levels returned to normal within 8 weeks of stopping trazodone. Longstreth and Hershman [3] described a patient who developed jaundice, pruritus, nausea, leukonychia, and elevated liver enzyme levels after a 7-month course of trazodone therapy (200 mg daily). The patient's symptoms and jaundice rapidly resolved after cessation of trazodone. The liver enzyme levels were nearly normal 5 weeks after stopping the drug, but there were an intermittent slight elevations of the aspartate aminotransferase and alkaline phosphatase levels for as long as 6 months after the discontinuation of trazodone therapy [3]. It is possible that the patient had drug-induced chronic hepatitis, especially given the prolonged abnormalities in liver enzyme levels; however, because a liver biopsy was not done, this remains speculative.

In our case, both bilirubin and -glutamyltransferase levels remained elevated until 6 months after cessation of the drug, probably reflecting the chronicity of the hepatotoxic process. Given the extensive damage that the biopsy specimen showed, this slow resolution of all liver chemistry abnormalities was not surprising.

Chronic active hepatitis has many causes. However, we are confident that the chronic hepatitis in our patient was caused by trazodone. Hepatitis B serologic test results were consistent with remote past exposure to hepatitis B but not with ongoing viral infection, which was confirmed by the negative immunostains on the biopsy specimen. The serologic and biopsy findings effectively ruled out other causes of chronic hepatitis. Finally, the temporal relation of the symptoms and liver enzyme abnormalities to the starting and stopping of the drug strongly implicates trazodone as the cause of the chronic hepatitis. We suggest that clinicians be alert to the possibility of hepatic injury in patients taking this medication.