Drug-induced liver injury has increasingly become an. Case Report: Fulminant Hepatic Failure Involving Duloxetine Hydrochloride.

Transcription

1 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2006;4: Case Report: Fulminant Hepatic Failure Involving Duloxetine Hydrochloride A. JAMES HANJE,* LINDSAY J. PELL, NICHOLAS A. VOTOLATO,, WENDY L. FRANKEL, and ROBERT B. KIRKPATRICK* *Department of Internal Medicine, Division of Digestive Health; Department of Pharmacy; Department of Psychiatry; and the Department of Pathology, The Ohio State University, Columbus, Ohio Background & Aims: Duloxetine hydrochloride was approved by the Food and Drug Administration in August 2004 for the treatment of major depressive disorder and diabetic peripheral neuropathic pain. Initial product labeling contained a precaution regarding the risk for increases in liver function test results. Recently, postmarketing research has revealed episodes of cholestatic jaundice and increases in transaminase levels to greater than 20 times normal in patients with chronic liver disease. Methods: In this case report, we describe a patient with non-hodgkin s lymphoma in remission and depression treated with duloxetine and mirtazapine. Results: Approximately 6 weeks after increasing her duloxetine dose from 30 to 60 mg daily, she became jaundiced and presented with fulminant hepatic failure. Liver function tests immediately before initiating duloxetine were not available, although the patient carried no prior history of chronic liver disease. A complete work-up for alternate causes failed to reveal another explanation for the patient s clinical presentation. A liver biopsy examination showed histologic changes of subacute injury and the patient s clinical course was consistent with drug-induced liver injury. Despite aggressive measures, the patient s condition deteriorated and the decision was made to withdraw care. Conclusions: This report shows a case of fulminant hepatic failure and death involving duloxetine use. Given recent reports of severe hepatotoxicity associated with the use of duloxetine in patients with pre-existing liver disease, further investigation into the safety of this compound is warranted. Drug-induced liver injury has increasingly become an area of concern for clinicians, regulatory agencies, and pharmaceutical companies pursuing new drug development. Drug-related hepatotoxicity is one of the most common reasons for withdrawal of an approved drug from the marketplace. 1 Occasionally, medicationrelated hepatic injury is not evident during preclinical drug development or during clinical trials owing to the small population sizes studied, the patient populations excluded from these studies, and the short duration of these evaluations. Despite the low rate of adverse drug event reporting by clinicians to the manufacturer and the Food and Drug Administration (FDA), postmarketing surveillance often allows for the identification of serious adverse events, including hepatotoxicity. One such example is nefazodone hydrochloride (Serzone; Bristol- Meyers Squibb, New York, NY), an antidepressant introduced in It was not until December 2001 that a boxed warning was added to the drug s labeling regarding cases of life-threatening hepatic failure. Duloxetine hydrochloride [( )-(S)-N-methyl- -(1- naphthyloxy)-2-thiophenepropylamine hydrochloride)] (Cymbalta; Eli Lilly and Company, Indianapolis, IN) is a selective serotonin and norepinephrine reuptake inhibitor with FDA approval for major depressive disorder and diabetic peripheral neuropathic pain. Common side effects include constipation, decreased appetite, dizziness, dry mouth, fatigue, and nausea. 2 Recently, the concern for hepatotoxicity has also been raised with the use of duloxetine in patients with pre-existing liver dysfunction. Here we describe a case of fulminant hepatic failure from drug-induced liver injury involving duloxetine. Case Report A 56-year-old woman with a history of non- Hodgkin s lymphoma and depression was admitted to the hospital with jaundice and fatigue. The patient originally had been diagnosed with lymphoma in 1997, subsequently treated with chemotherapy for 1 year, and was considered to be in complete remission. She had been treated for her depression with duloxetine and maintained on 30 mg/day for approximately 1 year. Approximately 6 weeks before presentation her duloxetine dose was increased to 60 mg/day and she was prescribed 15 mg/night of mirtazapine for insomnia. Her Abbreviations used in this paper: ALT, alanine transaminase; CT, computed tomography; FDA, Food and Drug Administration by the American Gastroenterological Association Institute /06/$32.00 doi: /j.cgh

2 July 2006 HEPATIC FAILURE AND DULOXETINE 913 past medical history otherwise was significant for chronic low back pain, hypothyroidism, and osteoarthritis. She carried no pre-existing diagnosis of chronic liver disease. She had undergone a prior total abdominal hysterectomy and bilateral salpingo-oopherectomy but had not received any blood transfusions with either of these procedures. She denied any history of alcohol abuse or intravenous drug use. She smoked 1 pack of cigarettes per day. In addition to the duloxetine and mirtazapine, her home medications included levothyroxine 150 mcg daily, controlled-release morphine sulfate 15 mg twice daily, lorazepam.5 mg 4 times daily, promethazine as needed, and hydrocodone/acetaminophen (5/500) as needed for back pain. Her husband reported rare use of the hydrocodone/acetaminophen for breakthrough pain and denied any use of herbal agents or complementary and alternative medicines. Approximately 4 weeks after increasing the duloxetine dose and adding mirtazapine to her daily regimen, the patient began complaining of generalized fatigue. A week later she began complaining of right upper-quadrant pain. Six weeks after her medications were adjusted, she noticed herself to be jaundiced for the first time, prompting her to seek medical attention. Initial examination revealed an alert, oriented woman, visibly jaundiced with scleral icterus. Her abdomen was soft and nontender with no ascites and no palpable hepatosplenomegaly. She had no peripheral edema and the remainder of her examination otherwise was normal. She was admitted to an outside hospital for further evaluation. On admission, her total bilirubin level was 9.9 mg/dl, conjugated bilirubin level was 6.8 mg/dl, alkaline phosphatase level was 307 u/l, aspartate transaminase level was 2477 u/l, alanine transaminase (ALT) level was 277 u/l, albumin level was 3.3 g/dl, and serum protein level was 7.4 g/dl. Her duloxetine was discontinued on admission. Initial evaluation included a right upper-quadrant ultrasound, which was normal. Computed tomography (CT) of her abdomen and pelvis without intravenous contrast revealed a nodular liver contour with an enlarged portal vein and borderline splenomegaly (12 cm). Mild bile duct dilatation and scattered subcentimeter intraperitoneal and retroperitoneal lymphadenopathy also were noted. She was discharged after 3 days in stable condition with the presumptive diagnosis of obstructive jaundice. An outpatient endoscopic retrograde cholangiopancreatography was performed soon after discharge and found to be normal with no evidence of obstruction and no ductal dilatation. She presented 48 hours later to the same outside hospital with new onset of altered mental status according to her husband. Her total bilirubin level at that time was 18.8 mg/dl, aspartate transaminase level was 1644 u/l, ALT level was 783 u/l, and her international normalized ratio was 2.8. Over the next 3 days, her liver function test results continued to worsen and her mental status deteriorated. She was believed to be in fulminant hepatic failure and was transferred subsequently to our institution to be evaluated for emergent liver transplantation. She had been intubated for airway protection because of her deteriorating mental status before transfer. Initial examination after arrival revealed pupils that were dilated but equally reactive and accommodating. Her gag reflex was intact. She was hyperreflexive with clonus and withdrew to painful stimuli in all 4 extremities. She did not respond to verbal stimuli or sternal rub. She had developed new abdominal ascites and peripheral edema, but her examination otherwise was unchanged. Her abdominal and pelvic CT scan was repeated with intravenous contrast, which again showed a nodular liver contour with an enlarged portal vein and borderline splenomegaly (13 cm). Previously reported lymphadenopathy was not seen, and there had been development of abdominal and pelvic ascites. No dominant masses were seen. An abdominal ultrasound also was repeated with Doppler flow that confirmed normal hepatopedal flow in the portal vein and patent intra-abdominal vasculature. Serum and urine toxicology screen was negative. Serologic evaluation for hepatitis A, B, and C, cytomegalovirus, Epstein Barr virus, and human immunodeficiency virus all were negative. Quantitative hepatitis C polymerase chain reaction (Mayo Clinic, Rochester, MN) was undetectable. Antinuclear antibody, antimitochondrial antibody, anti smooth muscle antibody, and antineutrophil cytoplasmic antibody test results were negative. Ceruloplasmin, ferritin, and -1-antitrypsin levels were within their normal limits. Blood and urine culture results were negative. The patient was evaluated immediately by the transplant team for consideration for emergent liver transplantation. With her worsening mental status, an initial head CT without intravenous contrast was performed that was unremarkable for evidence of increased intracranial pressure. An electroencephalogram was performed and showed no evidence of seizure activity with diffuse slowing suggestive of severe nonspecific encephalopathy. She was treated with lactulose and subsequently given both fresh-frozen plasma and recombinant factor VIIa with placement of an intracranial pressure monitor. Repeat head CT confirmed placement of the

3 914 HANJE ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 4, No. 7 Figure 1. Liver biopsy specimen showing centrilobular hepatocellular dropout with hemorrhage and ballooning degeneration (H&E, 10 ). Figure 2. Liver biopsy specimen showing panlobular disarray with portal and lobular mixed inflammatory infiltrate (H&E, 20 ). catheter within the brain parenchyma. Her initial intracranial pressure was 11 mm Hg with a cerebral perfusion pressure of 78 mm Hg. The patient was listed for liver transplantation. Despite continued supportive care, the patient s clinical condition deteriorated over the next 48 hours. An organ did not become available, and the patient was deemed too critical to undergo transplantation and was removed from the transplant list. After extensive consultation with the family, the decision was made to withdraw care. A formal autopsy was refused; however, a percutaneous liver biopsy specimen was obtained. This showed a mixed inflammatory infiltrate within the portal and lobular areas consisting mainly of neutrophils with occasional eosinophils and lymphocytes (Figure 1). Moderate amounts of microsteatosis and macrosteatosis and hepatocellular cholestasis were seen. There was ballooning degeneration of hepatocytes with bridging necrosis and centrilobular hepatocyte dropout (Figure 2). The portal areas showed bile ductular proliferation. Trichrome, reticulin, and elastic stains suggested bridging necrosis and edema rather than cirrhosis (Figure 3). Focal portal and perisinusoidal fibrosis were seen. Although nonspecific, these histologic changes were consistent with subacute liver injury. With the presence of focal fibrosis, chronic underlying liver disease could not be excluded. Discussion Drug-induced liver injury is the leading cause of acute liver failure in the United States. 3 Acetaminophen overdose, both intentional and unintentional, is the most common offending agent, with a recent prospective study showing that acetaminophen accounted for 42% of all drug-induced acute liver failure cases seen in the United States and one third of related deaths. 4 Excluding acetaminophen, idiosyncratic reactions comprise the majority of severe cases of acute drug-induced liver injury. Idiosyncratic reactions to specific drugs occur at therapeutic doses, are infrequent (occurring in 1 in every ,000 patients), and usually show a temporal relationship to initial drug exposure. 5 Diagnosing drug-induced liver injury with certainty can be difficult. Few clinical or laboratory findings exist to implicate a drug specifically as the cause of injury. A number of studies have evaluated the use of uniform factors in objectively assessing causality in suspected drug-related events. Temporal relationship, patient s course after discontinuation of the drug, response to rechallenge, prior response to the specific agent (or similar medications), and alternate explanations for the reaction all have been proposed as important determi- Figure 3. Liver biopsy specimen showing focal portal fibrosis (dark blue) and edema (pale blue) (trichrome stain, 10 ).

4 July 2006 HEPATIC FAILURE AND DULOXETINE 915 nants The majority of cases present as acute hepatocellular injury with increased ALT levels greater than 3 times the upper limit of normal, although a wide array of clinical scenarios have been reported. Individual drugs tend to produce a signature pattern of hepatic injury that is characteristic for the drug regarding time of onset, frequency, and type of hepatotoxicity. 11,12 More severe cases associated with an increase in mortality often are predicted by the emergence of jaundice in the presence of increased serum transaminase levels. 13 Duloxetine was approved by the FDA in August The initial product labeling contained a precaution regarding the risk for increase of liver transaminase levels, accompanied by a recommendation that duloxetine not be prescribed to patients with substantial alcohol use. In placebo-controlled trials in patients with major depressive disorder, increases of ALT levels to greater than 3 times the upper limit of normal occurred in.9% of duloxetine-treated patients. In patients with diabetic peripheral neuropathy, increases of ALT levels to greater than 3 times the upper limit of normal occurred in 1.68% of duloxetine-treated patients. In the full cohort of placebo-controlled trials, there was evidence of a dose-response relationship for ALT level increases greater than 3 times the upper limit of normal. The median time to detection of the laboratory abnormalities was 2 months. 2 In October 2005, a Dear Health Care Professional letter was issued by Eli Lilly and Company. In this letter, the manufacturer described postmarketing reports of cholestatic jaundice and hepatitis with abdominal pain, hepatomegaly, and increase of transaminase levels to greater than 20 times the upper limit of normal with or without jaundice in patients with chronic liver disease or cirrhosis who were treated with duloxetine. As a result of these reports, the product labeling was revised to include the statement that duloxetine should ordinarily not be prescribed to patients with substantial alcohol use or evidence of chronic liver disease. 2 Our patient s clinical course and the temporal relationship between the increase in her duloxetine dose and the onset of symptoms argue in favor of severe druginduced liver injury. The development of jaundice within 6 weeks of dose escalation is consistent with previous controlled trials showing that liver function test result abnormalities were detected at a median time of 2 months. The patient was maintained on a lower dose of duloxetine for almost a year without apparent difficulties, developing clinically significant hepatotoxicity only after the dose was doubled. This also is consistent with earlier trials showing a dose-dependent nature of hepatic injury with duloxetine use. The evaluation of this patient using a variety of objective scales to estimate causality including the Naranjo et al 6 nomogram, the Council for International Organizations of Medical Sciences scale, 7 and the Maria and Victorino 9 clinical scale indicate a probable association between the increase in duloxetine dose and the patient s hepatic injury. The increase in her duloxetine dose and the addition of mirtazapine were the only changes made to her daily medication regimen within the 12 months before presentation. The fact that her clinical course continued to worsen after removal of the drug is not surprising given the severity of her condition on presentation and the likelihood that she had continued taking the medications for weeks after the initial onset of symptoms. The effect of concurrent medications and the possibility of drug interactions also should be considered in this case. Most idiosyncratic drug reactions occur within 90 days of medication initiation, 5 and the addition of mirtazapine during the same time as the increase in her duloxetine dose implicates mirtazapine as a contributing or competing factor in the patient s clinical presentation. By using the same objective scales described earlier, causality scores for mirtazapine also indicate a probable association, although in both the Council for International Organizations of Medical Sciences and Maria and Victorino 9 scales, mirtazapine scored lower than duloxetine because of a lack of bibliographic data. This is because mirtazapine rarely has been associated with clinically significant increases in serum transaminase levels in almost 10 years of postmarketing surveillance and is not thought to have a narrow therapeutic index. 14 Two case reports of mirtazapine-induced hepatotoxicity have been published. 15 Both patients were women and initially presented with jaundice and abnormal liver function test results. The first patient had taken 30 mg/day of mirtazapine for 3 years before presentation and the second patient received 45 mg/day for 1 year before presentation. In both cases, symptoms and laboratory abnormalities resolved after withdrawal of the drug. The longer duration of medication administration before presentation and higher dose of mirtazapine in these case reports contrasts to the relatively short time period between drug initiation and clinical presentation and the lower dose in our patient. The mechanism of injury in this case is unclear. The elimination of duloxetine occurs primarily through hepatic metabolism involving 2 cytochrome P450 isoenzymes, CYP 2D6 and CYP 1A2. Duloxetine is a substrate and an inhibitor for both CYP450 2D6 and CYP450 1A2; thus it could impair its own clearance and

5 916 HANJE ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 4, No. 7 the clearance of other compounds metabolized through these pathways. In patients with pre-existing cirrhosis, the clearance of duloxetine is impaired significantly. In a study of 6 patients with cirrhosis (Child Pugh class B), a single 20-mg dose had a mean plasma clearance of about 15% that of age- and sex-matched healthy subjects. There was an approximate 3-fold increase in both the area under the curve and half-life when compared with controls with a similar peak plasma level. 16 Mirtazapine also is eliminated hepatically through a number of cytochrome P450 isoenzymes, including CYP450 2D6, CYP450 1A2, and CYP450 3A. Mirtazapine does not inhibit any cytochrome enzyme activity significantly. 17 Thus, it is unlikely mirtazapine would have a clinically significant inhibitory effect on the metabolism of duloxetine. Given the low dose of mirtazapine and the relatively rare incidence of hepatotoxicity associated with its use, it seems unlikely that impaired metabolism of mirtazapine was involved in the patient s severe liver injury, although co-administration of these compounds has not been studied specifically. Another concurrent chronic medication, lorazepam, has not been shown to have an effect on the pharmacokinetics of duloxetine. 2 For the year before presentation, no laboratory data were available to determine whether the patient had pre-existing increases in her liver function test results while on the lower dose of duloxetine. Similarly, no liver function test results were available to determine if the patient had pre-existing abnormalities before the initiation of duloxetine, although the patient carried no prior diagnosis of chronic liver disease. Her CT findings were suggestive of underlying portal hypertension, which is associated commonly with chronic liver disease but also can be seen in patients with acute liver failure. 18,19 The presence of a nodular liver on CT often suggests underlying cirrhosis, but is a somewhat nonspecific finding. In this case, the patient s biopsy examination showed only minimal fibrosis without cirrhosis and the degree of necrosis and parenchymal collapse alternatively could explain the appearance of a nodular liver on CT. Thus, it seems unlikely the patient had significant underlying chronic liver disease, but given the lack of data this possibility cannot be excluded. This case report details a published account of fulminant hepatic failure and death associated with duloxetine. Although the patient s clinical presentation strongly suggests drug-induced liver injury, the mechanism remains unclear, as does the possibility of concurrent medications as a contributing or competing factor. The temporal relationship between the dose escalation of duloxetine and the onset of clinically significant increases in transaminase levels is consistent with data obtained from clinical trials before FDA approval. In these studies, however, there were no instances of fulminant hepatic failure or death. It is important to note that the concern for severe liver injury associated with duloxetine use was not truly appreciated until postmarketing reports revealed cases of significant hepatotoxicity in patients with chronic liver disease or ongoing alcohol abuse. This case report highlights the importance of postmarketing surveillance in detecting adverse effects that may not be described before FDA approval. Initial efficacy trials of new antidepressants typically involve more extensive exclusion criteria than what is seen in clinical practice. For example, psychiatric and nonpsychiatric comorbidities, including alcohol abuse or dependence, are common in depressive disorders but are more limited or excluded in patients enrolled in clinical trials. Thus, the effects of many concurrent medications and risk factors cannot be identified by the current approval process for new antidepressants. Depression is a disorder that fails to remit in many patients, and antidepressant medications often are administered for an unlimited period of time. This is in sharp contrast to the shorter duration of antidepressant drug exposure (range, 2 12 mo) seen in clinical trials before FDA approval. The significance of many side effects, including hepatotoxicity, may take a much longer period of time to be understood fully. The experience with nefazodone hydrochloride is one example of a situation in which cases of liver injury and life-threatening hepatic failure preceded the addition of a black box warning by many years. With the established concern for hepatotoxicity associated with the use of duloxetine in patients with alcohol use or chronic liver disease, other factors such as increased dose and duration of therapy and the use of concurrent medications also should be considered when determining potential risk of hepatic injury. Therefore, we recommend increased vigilance in the monitoring of liver function test results before and throughout duloxetine therapy until the extent of potential hepatotoxicity is clarified through postmarketing surveillance. References 1. Wysowski DK, Swartz L. Adverse drug event surveillance and drug withdrawal in the United States : the importance of reporting suspected reactions. Arch Intern Med 2005;165: CYMBALTA (duloxetine hydrochloride) delayed-release capsules. Product labeling. Indianapolis, IN: Eli Lilly and Company, Ostapowicz G, Fontana RJ, Schiodt FV, et al, U.S. Acute Liver Failure Study Group. Results of a prospective study of acute liver

6 July 2006 HEPATIC FAILURE AND DULOXETINE 917 failure at 17 tertiary care centers in the United States. Ann Intern Med 2002;137: Larson AM, Polson J, Fontana RJ, et al, and the Acute Liver Failure Study Group. Acetaminophen-induced acute liver failure: results of a United States multicenter, prospective study. Hepatology 2005;42: Lee WM. Drug-induced hepatotoxicity. N Engl J Med 2003;349: Naranjo CA, Busto U, Sellers EM, et al. A method for estimating the probability of adverse drug reactions. Clin Pharmacol Ther 1981;30: Danan G, Benichou C. Causality assessment of adverse reactions to drugs. I. A novel method based on the conclusions of international consensus meetings: application to drug-induced liver injuries. J Clin Epidemiol 1993;46: Benichou C, Danan G, Flahault A. Causality assessment of adverse reactions to drugs. II. An original model for validation of drug causality assessment methods: case reports with positive rechallenge. J Clin Epidemiol 1993;46: Maria VAJ, Victorino RMM. Development and validation of a clinical scale for the diagnosis of drug-induced hepatitis. Hepatology 1997;26: Kaplowitz N. Causality assessment versus guilt-by-association in drug hepatotoxicity. Hepatology 2001;33: Kaplowitz N. Drug-induced liver injury. Clin Infect Dis 2004;38: Maddrey WC. Clinicopathologic patterns of drug-induced liver disease. In: Kaplowitz N, DeLeve LD, eds. Drug-induced liver disease. New York: Marcel Dekker, 2002: Zimmerman HJ. Hepatotoxicity: the adverse effects of drugs and other chemicals in the liver. 2nd ed. Philadelphia: Lippincott Williams and Wilkins, 1999: Lucena MI, Carvajal A, Andrade RJ, et al. Antidepressant induced hepatotoxicity. Expert Opin Drug Saf 2003;2: Hui CK, Yuen MF, Wong WM, et al. Mirtazapine-induced hepatotoxicity. J Clin Gastroenterol 2002;35: Suri A, Reddy S, Gonzales C, et al. Duloxetine pharmacokinetics in cirrhotics compared with healthy controls. Int J Clin Pharmacol Ther 2005;43: REMERON (mirtazapine) tablets. Product labeling. Roseland, NJ: Oragon USA Incorporated, Navasa M, Garcia-Pagan JC, Bosch J, et al. Portal hypertension in acute liver failure. Gut 1992;33: Valla D, Flejou JF, Lebrec D, et al. Portal hypertension and ascites in acute hepatitis: clinical, hemodynamic, and histological correlations. Hepatology 1989;10: Address requests for reprints to: A. James Hanje, Department of Internal Medicine, Division of Digestive Health, Ohio State University, 515 Doan Hall, 410 W. 10th Avenue, Columbus, Ohio jim.hanje@osumc.edu; fax: (614)