INTRODUCTION

Type 2 diabetes mellitus (T2DM) is a significant risk factor for nonalcoholic fatty liver disease (NAFLD), non-alcoholic fatty liver (NAFL), and nonalcoholic steatohepatitis (NASH).¹ For example, 33-50% of patients with diabetes mellitus have NAFLD.² Recently, the usefulness of sodium glucose cotransporter 2 (SGLT2) inhibitors for animal NASH models has been reported.³ However, because of its relatively short history as an antidiabetic drug, the long-term safety of SGLT2 2 inhibitors has not been established.⁴ Recently, we encountered a case of ipragliflozin-induced liver injury. As there is a possibility of the future use of SGLT2 inhibitors to treat patients with diabetes mellitus presenting alone or with concomitant NAFLD, NAFL, and NASH, this case report will provide important clinical information about their safety.

CASE REPORT

A 75-year-old Japanese male patient had been regularly visiting our hospital for control of his type 2 diabetes mellitus and hypertension since his diagnosis at 64 years of age. He had diabetic peripheral neuropathy but not diabetic retinopathy. This caused slight bilateral peripheral numbness on his legs. A spot urine test showed a urine protein/creatinine ratio of 6.284, and he had an estimated glomerular filtration rate (eGFR) of 63.1 (mL/min/1.73 m²). As he developed hypoglycemic episodes with sulfonylurea, it was replaced with ipragliflozin (50 mg/day) in February 2016. However, 49 days after starting ipragliflozin, he developed easy fatigability. The following laboratory test results led to a diagnosis of clinical liver injury (Table 1A): aspartate aminotransferase (AST) increased from 13 to 622 U/L (normal range, 10-40 U/L), alanine aminotransferase (ALT) increased from 9 to 266 U/L (normal range, 5-45 U/L), alkaline phosphatase (ALP) increased from 239 to 752 U/L (normal range, 100-325 U/L), and γ-glutamyl transpeptidase (γ -GTP) increased from 19 to 176 U/L (normal value, ≤ 80 U/L). Zinc sulfate turbidity test (ZTT) was 3.5 U (normal range, 2-12 U), thymol tur bidity test (TTT) was 0.4 U (normal value, ≤ 4 U), total protein was 6.4 g/dL (normal range, 6.7-8.3 g/dL), albumin was 3.2 g/dL (normal range, 3.8-5.3 g/dL), and total bilirubin was 0.7 g/dL (normal range, 0.2-1.2 mg/dL). Hepatitis A virus antibody (Anti-HAV-IgM), hepatitis B virus antigen, antibody, and DNA (HBsAg, anti-HBc-IgM, HBV-DNA), hepatitis C antibody (anti-HCV), IgM cytomegalovirus (CMV) antibody and Epstein-Barr virus (EB) viral capsid antigen (VCA)-IgM antibody were nega tive. Abdominal computed tomography (CT) showed mild fatty liver but there were no nodular lesions. The lymphocyte transformation test (LTT), which is used to determine whether a patient has developed a T-cell response against a certain drug,⁵ was positive for ipragliflozin (Table 1B). Glycosylated hemoglobin (HbA1c) decreased from 6.8% to 6.4% after starting ipragliflozin treatment without any reported hypoglycemic episodes.

He was admitted to our hospital and iplagliflozin was discontinued. He was discharged from the hospital 18 days later with his AST and ALT levels reduced to 20 U/L and 13 U/L, respectively. Thereafter, his liver function remained within the normal range until the time of this writing.

DISCUSSION

Ipragliflozin reportedly has therapeutic effects on nonalcoholic steatohepatitis in mice.⁶ However, we have identified an individual who developed liver injury 49 days after starting ipragliflozin therapy, and recovered 11 days after its termination (Table 1A). We confirmed that hepatitis A virus, hepatitis B virus, hepatitis C virus, cytomegalovirus, and Epstein-Barr virus did not cause the elevated levels of AST, ALT, ALP, or γ-GTP. We also ruled out the potential of a tumor mass by abdominal CT examination. Since the patient had mild fatty liver, we used the updated Roussel Uclaf Causality Assessment Method (RUCAM) scale to determine if patient had cholestatic or mixed liver injury of DILI (drug induced liver injury) or HILI (herb induced liver injury). This patient had cholestatic liver injury as his R (= ALT/ALP) value was ≤ 2. His updated RUCAM score was 7, according to recent literature,^{7 , 8} (Table 1B) and indicated that our case was probably of a DILI. We also confirmed that drug-induced lymphocyte transformation test was positive for ipragliflozin (Table 1C). Therefore, we concluded that liver injury was caused by ipragliflozin. At the time of writing this report, we are unable to identify any similar case studies in our PubMed searches. Since 3350% of patients with diabetes mellitus have NAFLD,² and NAFLD can be considered as a risk factor for DILI,^{8 , 9} it is prudent to occasionally monitor liver function and assess patients for indications, such as dehydration, ketosis, and urinary tract infections, before prescribing SGLT2 inhibitors.