Senior Research Specialist Brigham and Women's Hospital Brighton, United States
Background: To mitigate the risk of potential liver toxicity, both ambrisentan and bosentan had liver injury label warning and risk evaluation and mitigation strategies which mandated a liver function test (LFT) before therapy and monthly thereafter in place prior to 2011. In March 2011, the Food and Drug Administration (FDA) removed liver injury warning and LFT requirement for ambrisentan, but not bosentan.
Objectives: To evaluate how March 2011 change affected the use and LFT testing for ambrisentan users compared to bosentan users.
Methods: Using 3 US health insurance claims databases (Optum Clinformatics, Merative Marketscan, Medicaid; 2007-2018), we assessed monthly use as the number of individuals with at least one outpatient dispensing per million eligible enrollees for each month, pooled across the databases. To assess the impact of March 2011 change, we conducted an interrupted time-series analysis comparing monthly use during the 2 years preceding to 2 years following the change. We additionally assessed the percentage of initiators (new users following at least 6 months of no use) who had an LFT ordered (i) within 90 days before initiation and (ii) within 31 days before the first refill. We assessed percentages of initiators with LFTs yearly during the study period (2007-2018). Mean percentages prior to the change (2007-2010) were compared to mean percentages after (2012-2018). Analyses were conducted separately for ambrisentan and bosentan.
Results: Use of ambrisentan increased gradually following its approval in June 2007 to 10 individuals per 1 million eligible enrollees in February 2011. The removal of LFT requirement was associated with an immediate increase in monthly use to 11 individuals per million in April 2011 (p < 0.0001 for level change), as well as an increase in the upward slope compared to before March 2011 (p < 0.01 for trend change). Bosentan use had no level change in March 2011 and decreased during the 2 years following March 2011 (p-value < 0.0001). The percentage of ambrisentan initiators who had LFTs ordered prior to initiation decreased from a mean of 76.6% (standard deviation [SD] 2.7%) before 2011 to 71.4% (SD 3.1%) thereafter, while the percentage of patients who had LFTs ordered prior to their first refill decreased from a mean of 61.0% (SD 2.8%) to a mean of 36.5% (SD 3.1%). There were no changes to LFT testing among bosentan initiators: a mean of 71.4 % and 58.0% had LFTs prior to initiation and first refill, respectively, during 2007-2010 vs 71.7% and 56.6% during 2012-2018.
Conclusions: Removing liver injury warning and LFT requirement for ambrisentan was associated with a rapid increase in use for ambrisentan and a decrease in LFT testing over time. Similar changes were not observed for bosentan.
