Memantine-Induced Liver Injury With Probable Causality as Assessed Using the Roussel Uclaf Causality Assessment Method (RUCAM)

ABSTRACT

Presentations of drug-induced liver injury (DILI) are highly variable. Although biochemical evidence of cholestasis is common, the extent of aminotransferase elevations and patterns of liver injury vary. Patients may be asymptomatic, and many cases may never be diagnosed. We describe a case of memantine-induced hepatotoxicity in an elderly patient with Alzheimer's dementia, with probable causality for drug-induced liver injury, as assessed using the Roussel Uclaf Causality Assessment Method (RUCAM) score.

INTRODUCTION

Drug-induced liver injury (DILI) is a hepatic injury secondary to prescription or over-the-counter medications occurring at recommended doses and not due to another underlying condition.¹ The most common overall cause of DILI is acetaminophen, which causes a dose-dependent and stereotypical reaction in all patients. Idiosyncratic DILI from any thousands of medications is highly variable in its manifestations. These may present differently in susceptible individuals, making the diagnosis challenging, given the lack of objective diagnostic tests.²

Given the inconsistent presentation of DILI, a major challenge is establishing a causative relationship between the hepatic injury and the drug in question. An important tool in the elucidation of causality is the Roussel Uclaf Causality Assessment Method (RUCAM) score, a well-validated clinical scoring system.³ The RUCAM can be applied to all medications that a patient was taking to determine which medication was the most likely causative factor in hepatic injury.⁴

This case describes a rare presentation of DILI secondary to memantine. Memantine is an oral antagonist of the N-methyl-D-aspartate type of glutamate receptor that is well absorbed and partially metabolized by the liver. There is no current literature suggesting that other N-methyl-D-aspartate antagonists cause similar hepatotoxicity, and only one case identifying memantine itself as a causal agent of DILI.⁵

CASE REPORT

An 86-year-old asymptomatic man with a history of coronary artery disease, atrial fibrillation, prostate cancer, hypothyroidism, gout, Alzheimer's disease, and cholelithiasis was referred to the gastroenterology clinic for evaluation of an incidental finding of elevated aminotransferases up to 10 times the upper limit of normal. At presentation, the serum aspartate aminotransferase level was 438 U/L, alanine aminotransferase level was 439 U/L, total and direct bilirubin levels were 0.9 and 0.5 mg/dL, respectively, and alkaline phosphatase level was 169 U/L. All of these values were significantly elevated above baseline levels, last evaluated 5 months ago. These laboratory values equated to an R value of 8.17, indicating a hepatocellular pattern of liver injury, and did not meet the criteria for Hy’s Law.²

The patient's longstanding medications included donepezil, lisinopril, carvedilol, apixaban, atorvastatin, levothyroxine, allopurinol, and omeprazole. The only recent change in his medications was the initiation of memantine 2 months before presentation. The patient had no history of tobacco or alcohol use, and he had no personal or family history of hepatobiliary conditions. His blood pressure was well controlled on lisinopril and carvedilol without any evidence of hypotension.

His physical examination was unremarkable. Memantine was discontinued at this visit because of suspicion for DILI, with continuation of all his other medications. Evaluations for viral and metabolic etiologies for liver injury were unremarkable. Evaluations for autoimmune etiologies were unremarkable as well, to include a negative antinuclear antibody immunofluorescent assay, negative smooth muscle antibodies (<1:10 dilution), and negative mitochondrial antibodies. He demonstrated an immunoglobulin G (IgG) level of 1,683 mg/dL (1.05 times upper limit of normal). A right upper quadrant ultrasound was positive only for cholelithiasis. Three months after memantine was discontinued, the levels of aspartate aminotransferase (168 U/L), alanine aminotransferase (140 U/L), and alkaline phosphatase (58 U/L) had decreased significantly. Ultimately, 6 months after memantine withdrawal, all liver-associated enzymes returned to normal levels. Given the concern for DILI, a RUCAM score was established for all medications that the patient was taking when the elevated liver enzymes were noted. Memantine was assigned the highest RUCAM score at 8:2 for time to onset, 2 for the course after cessation of the drug, 1 for risk factor (age >55 years), 2 for exclusion of both group I and II causes of liver injury, and 1 for a previously published case regarding hepatotoxicity associated with memantine. This did not include a score for readministration. The remainder of the medications was assigned RUCAM scores as follows: lisinopril 4, donepezil 4, carvedilol 4, apixaban 4, atorvastatin 5, levothyroxine 3, allopurinol 5, and omeprazole 4.

DISCUSSION

The diagnosis of DILI is a difficult clinical challenge, requiring a thorough evaluation for alternative causes of hepatic injury including viral, autoimmune, biliary, or infiltrative conditions.⁶ In addition, a strong index of suspicion and meticulous history taking are required. The fact that DILI is a diagnosis of exclusion cannot be emphasized enough. In an analysis of 2,906 cases of liver injury attributed to a specific drug, the diagnosis of DILI was incorrect in at least 14% because of obvious alternative causes. The delay in reaching the correct diagnosis in these patients was also significant, with a median delay of 88.5 days in the inpatient group and 122 days in the outpatient group.⁷ The implications of these missed alternative diagnoses and delay to treatment can be significant, underscoring the importance of a thorough evaluation and careful causality assessment with a tool such as the RUCAM.

DILI can be classified by the type of hepatic injury (hepatocellular, cholestatic, mixed), mechanism of injury (intrinsic, idiosyncratic), clinical manifestations, and histologic findings (if liver biopsy is performed).⁸ Our patient was asymptomatic and presented with a hepatocellular pattern of liver injury, with a resultant negative workup for alternative causes. Although autoimmune hepatitis remained on the differential diagnosis, the decision was made to forego liver biopsy, given the low likelihood for a change in management based on the results because of the patient's age and comorbidities as well as associated risks of the procedure. Given the diagnostic limitations without liver biopsy, autoimmune hepatitis was ultimately unlikely based on the negative autoantibody evaluation, IgG level, and Revised Original Scoring System of the International Autoimmune Hepatitis Group score of 4.⁹

Given the complete resolution of liver-associated enzyme derangements after withdrawal of the drug, a negative alternative workup, and a probable RUCAM score, the patient was given the diagnosis of DILI secondary to memantine. This is the second case of memantine toxicity reported in the literature, and this should be a reminder to all providers of the importance of a thorough history and alternative diagnosis workup in the evaluation of DILI and the important role the RUCAM score has in assessing the causality of suspected medications.⁵

DISCLOSURES

Author contributions: J. Shumar wrote the manuscript and is the article guarantor. S. Ordway, Z. Junga, B. Sadowski, and D. Torres edited the manuscript.

Financial disclosure: None to report.

Informed consent was obtained for this case report.

Previous publication: This case report was presented in part as a poster at the 2018 ACG Annual Meeting; October 5–10, 2018; Philadelphia, PA.