Introduction: Fasiglifam (TAK-875) is a G protein-coupled receptor 40 agonist that was being investigated for treatment of type 2 diabetes mellitus (T2DM). A development program was terminated late in phase III clinical trials due to liver safety concerns.
Methods: The liver safety of fasiglifam was assessed from data based on six phase II and nine phase III double-blind studies and two open-label studies with emphasis on pooled data from 15 double-blind studies from both global and Japanese development programs. Taking into consideration different daily doses of fasiglifam administered in clinical studies, the primary comparisons were between all patients exposed to fasiglifam (any dose) versus placebo, and, where applicable, versus the two active comparators, sitagliptin or glimepiride. A Liver Safety Evaluation Committee consisting of hepatologists blinded to treatment assignments evaluated hepatic adverse events (AEs) and serious AEs (SAEs) for causal relationship to study drug.
Results: The analysis included data from 9139 patients with T2DM in 15 double-blind controlled studies who received either fasiglifam (n = 5359, fasiglifam group), fasiglifam and sitagliptin (n = 123), or a comparator agent (n = 3657, non-exposed group consisting of placebo and other antidiabetic agents). Exposure to treatment for more than 1 year ranged from 249 patients in the placebo arm, to 370 patients in the glimepiride arm and 617 patients in the fasiglifam 50 mg arm. The primary focus of the analysis was on the hepatic safety of fasiglifam. The overall safety profile based on treatment-emergent AEs (TEAEs), SAEs, deaths, and withdrawal due to AEs was similar between fasiglifam and placebo (excluding liver test abnormalities). However, there was an increased incidence rate of serum alanine aminotransferase (ALT) elevations > 3 × upper limit of normal (ULN), 5 × ULN, and 10 × ULN in fasiglifam-treated patients compared with those treated with placebo or active comparators. ALT elevations > 3 × ULN for fasiglifam were 2.7% compared with 0.8 and 0.5% for the active comparators and placebo. There did not appear to be a clear dose response in incidence of ALT elevations between patients receiving 25 or 50 mg daily. The cumulative incidence of elevations in serum ALT > 3 × ULN was higher in the first 6 months of treatment with fasiglifam compared with both placebo and the active comparators, but the rate of new ALT elevations appeared to be similar across all treatment groups thereafter. No demographic or baseline patient characteristics were identified to predict elevations exceeding ALT > 3 × ULN in fasiglifam-treated patients. The pattern of liver injury with fasiglifam was hepatocellular, and there were no reports of liver-related deaths, liver failure or life-threatening liver injury. Most fasiglifam-associated ALT elevations were asymptomatic and resolved promptly upon discontinuing treatment, but in two patients the recovery was prolonged. Importantly, three important serious liver injury cases were identified among fasiglifam-treated patients; one case was adjudicated to be a clear Hy's Law case and the two remaining cases were considered to closely approximate Hy's Law cases.
Conclusions: Although the incidence of overall AEs, SAEs, and deaths was similar between fasiglifam and placebo, a liver signal was identified based primarily on the difference in liver chemistry values in the fasiglifam group compared with the placebo and active comparator groups. Three serious liver injuries were attributed to fasiglifam treatment. Clinical development of fasiglifam was halted due to these liver safety concerns.