Topiramate-Induced Acute Liver Failure in a Pediatric Patient: a Case Report and Review of Literature

Drug-induced liver injury (DILI) represents a special type of adverse drug reaction. More than 600 drugs and chemicals have been associated with significant liver injury. Hepatotoxicity is a well-known adverse effect of first-generation antiepileptics. Second-generation antiepileptics generally have a milder adverse effect profile; however, hepatotoxicity and hyperammonemic encephalopathy have been reported in patients on topiramate, in conjunction with other antiepileptics. We report a case of a pediatric patient presenting in acute liver failure (ALF) with remarkable coagulopathy and hyperammonemia that likely resulted from topiramate toxicity, compounded by hypovolemic shock. CASE The patient is an 11-year-old boy with history of cerebral palsy, intellectual disability, and epilepsy who presented with somnolence and diarrhea. Initial evaluation revealed aspartate aminotransferase 5666 U/L, alanine aminotransferase 7890 U/L, gamma-glutamyl transferase 243 U/L, bilirubin 5.9 mg/dL, international normalized ratio 10.8, and ammonia 1350 μg/dL; consistent with ALF. His home medications, since infancy, included topiramate, phenobarbital, diazepam, and baclofen. Both serum antiepileptic blood levels were within their therapeutic ranges, although the trough topiramate level of 21.2 μg/mL (normal range 2–25 μg/mL) was significantly elevated from his previous levels of 5 to 7 μg/mL. In addition, he had acidosis (pH 7.1, lactate 18 mmol/L); acute kidney injury (creatinine 2.2 mg/dL), and hypovolemic shock. At baseline, he was nonverbal but able to communicate with crying/moaning. On admission, he was unresponsive with grade IV hepatic encephalopathy. A computed tomography head showed no intracranial hemorrhage or edema. An electroencephalogram was normal. Topiramate was held given his ALF and hyperammonemia, but phenobarbital, baclofen, and diazepam were continued. Workup for ALF included negative serum viral studies (including Epstein–Barr virus/cytomegalovirus serology and polymerase chain reaction (PCR), hepatitis A, B, C, Coxsackie A and B ab panels, Echovirus Ab, herpes simplex virus 1/herpes simplex virus 2 serology and PCR, Parvovirus B19 ab, human herpesvirus6, Enterovirus PCR from nares); negative urine toxicology screen; undetectable acetaminophen; negative autoimmune workup (antinuclear antibody, smooth muscle antibody, liver kidney microsomal), normal ceruloplasmin, and elevated ferritin but negative iron stains on biopsy. A nasal respiratory viral panel was positive for enterovirus/rhinovirus, and adenovirus. Serum adenovirus PCR and liver biopsy adenoviral stain were negative. A liver ultrasound/Doppler showed good hepatic flow and minimal fatty infiltration. Aminotransferases, ammonia, and coagulopathy rapidly improved with resuscitation, fresh frozen plasma, and discontinuation of topiramate. On day 4 of admission, a transjugular liver biopsy revealed severe acute hepatitis and microvesicular steatosis involving 50% of the liver parenchyma consistent with DILI (Fig. 1). Given the pathology and the clinical improvement after discontinuation of topiramate, the patient's ALF and hyperammonemia were deemed a consequence of topiramate toxicity.FIGURE 1: The liver biopsy demonstrates severe lobular disarray and florid portal and lobular inflammatory cell infiltrates. The inflammatory infiltrates consist mostly of lymphocytes and neutrophils. Microvesicular steatosis is present and involves 50% of liver parenchyma. Intrahepatocellular cholestasis and bile plugs are also present. H&E = hematoxylin and eosin.The patient continued to improve during the following weeks with normalization of laboratory values. He, however, remains ventilator-dependent, and his mental status is further impaired. DISCUSSION ALF in pediatric patients is rare. Approximately 50% of pediatric ALF cases remain without etiology. According to the Pediatric ALF registry between 1999 and 2004, DILI accounted for 19% of cases of ALF of which 5% were idiosyncratic (non–acetaminophen-induced) (1). In adults in the United States, idiosyncratic reactions make up 13% of all ALF cases. Regardless of age, idiosyncratic DILI from any single medication is uncommon (2). Topiramate is an antiepileptic medication also used in migraine prevention. It is metabolized by the cytochrome P450 system and induces CYP3A4 activity, which can alter the levels of other anticonvulsants. Topiramate as a cause of ALF has only rarely been suggested, and the mechanism of DILI is not well understood. Using Roussel Uclaf Causality Assessment Method (RUCAM) causality assessment of a drug in acute liver injury, topiramate as a causal agent for ALF is probable (3). The timing of the elevated topiramate level, the rapid decrease in alanine aminotransferase after discontinuation, and the lack of a concomitant drug trigger add to the argument. The case scores a total of 8 RUCAM points, which puts him on the border of "probably" and "high probably" of topiramate as a cause of hepatotoxicity. The elevated ammonia level also supports topiramate-induced hepatic toxicity. The proposed mechanism for topiramate-induced hyperammonemia is via an inhibition of type V mitochondrial carbonic anhydrase. This inhibition decreases bicarbonate synthesis, which is needed for the synthesis of carbamyl phosphate from free ammonia. This is an essential step in the urea cycle for detoxification of ammonia (4). An ammonia level of >1300, as in our patient, is rare. Both ALF and topiramate can cause hyperammonemia independently. No well-established characteristic pathology for topiramate hepatotoxicity exists. Our patient's pathology showed a hepatocellular and cholestatic pattern of damage with microvesicular steatosis, consistent with DILI. Histology of idiosyncratic DILI in children tends to be diverse, often with inflammatory activity and cholestasis (5). There have been 3 previous case reports of DILI attributed primarily to topiramate. Unlike our case, these incidents occurred after initiating topiramate whereas our patient was on topiramate for the past 10 years. Our patient's topiramate level, however, was >4-fold higher than it had ever been before, with no changes in his medications and no supplements, which, we hypothesize, mimics a "new drug" effect. Bjoro et al (6) described an adult with ALF requiring a liver allograft 4 months after topiramate was added to a stable regimen of carbamazepine and days after a dose increase. Similarly, Doan and Clendenning (7) reported a case of ALF in a young adult for whom topiramate was added to a long-standing valproate regimen. Her liver function tests corrected after discontinuing topiramate. Finally, del Val Antonana et al (8) reported a case of hepatotoxicity in a 16-year-old boy who received topiramate as monotherapy, initiated 10 weeks before presentation. He had elevated transaminases and coagulopathy but no hyperammonemia. His findings resolved 2 weeks after discontinuing topiramate. CONCLUSIONS We report a rare case of ALF likely caused by topiramate toxicity in a pediatric patient, whose hypovlemia could have potentiated the liver toxicity. This case is unique with regards to age, degree of hepatic failure, level of hyperammonemic encephalopathy, and onset of toxicity with a chronic medication. Patients with baseline impairment on antiepileptic drugs need close monitoring for subtle changes in mental status. If a patient presents with liver dysfunction on topiramate, caution should be taken with its continuation and further investigation should be undertaken to rule out drug-associated hepatotoxicity.