Bosentan

Placebo-controlled studies

The BREATHE-2 trial examined the effects of a combination of intravenous epoprostenol with either bosentan or placebo in a randomized trial in 33 patients with pulmonary arterial hypertension [ ]. There were three deaths in the combination treated group during or soon after the end of the study, although the authors concluded that the severity of the patients’ illnesses it precluded definite attribution of the deaths to the treatment. Other adverse effects were mainly those attributed to epoprostenol, although an excess number of patients developed leg edema in the combination group (27% versus 9%).

Bosentan is the most widely studied of the endothelin receptor antagonists. The BUILD-1 study (Bosentan Use in Interstitial Lung Disease) was a randomized, placebo-controlled trial of bosentan in 154 patients with idiopathic pulmonary fibrosis [ ]. Exercise capacity was not increased by bosentan, although there was a trend towards delayed time to death or disease progression. Respiratory complaints made up the majority of adverse events, but were less frequent among those who took bosentan. There were rises in alanine transaminase only in those who took bosentan; they affected 21% and led to withdrawal in 12%. The profile and frequency of the adverse effects of bosentan are similar to those seen in trials for its other indications, such as pulmonary artery hypertension.

Cardiovascular

In a dose-finding study with bosentan (100, 500, 1000, and 2000 mg/day) in 293 hypertensive patients [ ] there were statistically significant falls in diastolic blood pressure with the 500 and 2000 mg/day doses. The effects were similar to that of enalapril 20 mg/day. The lowering of blood pressure was not associated with any changes in heart rate, plasma noradrenaline concentrations, plasma renin activity, or angiotensin II concentrations.

In the ENABLE (Endothelin Antagonist Bosentan for Lowering Cardiac Events in Heart Failure) placebo-controlled study the effects of low-dose bosentan (125 mg bd) were evaluated in 1613 patients with severe heart failure (left ventricular ejection fraction < 35%, New York Heart Association classes IIIb-IV) [ ]. The primary endpoint of all-cause mortality or hospitalization for heart failure was reached in 321 of 808 patients who took placebo and 312 of 805 who took bosentan. Treatment with bosentan conferred an early risk of worsening heart failure necessitating hospitalization, because of fluid retention. These results throw doubt on the potential benefits of non-specific endothelin receptor blockade in heart failure. Preliminary data suggest that in heart failure selective endothelin-A receptor antagonists (BQ-123, sitaxsentan) may be more beneficial than non-selective antagonists, especially when there is associated pulmonary hypertension [ ].

Liver

The major safety issue that has emerged with bosentan, the endothelin receptor antagonist that has been most extensively studied in man, has been dose-dependent reversible impairment of hepatic function (3% with 125 mg, 7% with 250 mg), manifesting as raised transaminases [ ]. The effect of bosentan on hepatocanalicular bile-salt transport has been studied in rats in conjunction with a re-examination of the safety database from two clinical trials (in hypertension and congestive cardiac failure) and measurement of bile-salt concentrations in stored blood samples from these trials [ ]. Hepatic injury was defined as a three-fold increase in alanine transaminase activity. In the hypertension trial there were no cases of hepatic injury with placebo or enalapril. With bosentan, the frequencies were 2, 4, 11, and 8% at dosages of 100, 500, 1000, and 2000 mg/day respectively. There was a dose-dependent increase in bile-salt concentrations. In the study in patients with heart failure (New York Heart Association classes III/IV), liver injury occurred in 4% of 126 patients taking placebo and 18% of 244 patients taking bosentan 500 mg bd. A subgroup analysis showed a higher incidence of hepatic injury in patients taking concomitant bosentan and glibenclamide. Patients with hepatic injury had raised bile-salt concentrations.

In early placebo-controlled trials of bosentan in pulmonary artery hypertension there were cases of raised hepatic transaminases, particularly in association with the highest doses. A 50-year-old woman with pulmonary artery hypertension developed severe hepatotoxicity and consequent cirrhosis while taking bosentan [ ]. This report was cited by the manufacturer of bosentan (Actelion) in a letter informing prescribers of changes to the product labelling and reminding them of the importance of monthly liver function testing during treatment.

Owing to concerns from early clinical trials about the hepatotoxicity of bosentan, the European Medicines Agency (EMA) required a specific post-marketing surveillance (PMS) study before marketing authorization. The observational results were reported in a comprehensive report of nearly 5000 patients taking bosentan [ ]. Rises in transaminase activities were fairly common (crude incidence of 7.6%), especially during the first 6 months of treatment. After 1 year the probability of liver enzyme abnormalities was greatly reduced, reinforcing the need for monthly monitoring of transaminases for the duration of bosentan treatment. No new adverse reactions were identified.

Fatal hepatotoxicity has been reported in a patient who took bosentan for digital ulceration due to mixed connective tissue disease [ ].

• A 70-year-old Japanese woman with mixed connective tissue disease developed worsening digital gangrene and was given bosentan 125 mg/day (initial 2 weeks 62.5 mg/day). After 8 weeks she developed raised transaminases. Over the next 4 days both transaminases increased to a peak of over 5000 IU/l before gradually decreasing. However, the patient developed a gastrointestinal hemorrhage and died from multiple organ failure 10 days after withdrawal of bosentan.

In rats, intravenous bosentan produced a dose-dependent increase in plasma bile salts. The effect was potentiated when glibenclamide was co-administered. In vitro studies in rat canalicular liver plasma membranes confirmed inhibition of bile-salt transport. Three bosentan metabolites were also investigated. The M2 metabolite was more potent than bosentan, whereas the M1 and M3 metabolites produced less inhibition of bile acid transport than bosentan. The effects of bosentan and its main metabolites, both of which are eliminated in the bile, on biliary secretion have been studied in rats with biliary fistulae and with or without a genetic defect in mrp2 [ ]. Intravenous bosentan 0.1–10 mg/kg caused a dose-dependent increase in biliary bilirubin excretion, and doses of 10 mg/kg or over caused a sustained increase in canalicular salt-independent bile flow, combined with significant increases in the concentrations and output of glutathione and of bicarbonate in the bile. Phospholipid and cholesterol secretion were profoundly inhibited and uncoupled from bile-salt secretion. In mrp2-deficient rats, the choleretic effect of bosentan was markedly reduced. Thus, bosentan alters canalicular bile formation mostly via mrp2-mediated mechanisms. Intermittent uncoupling of lipid from bile-salt secretion may contribute to its hepatic adverse effects.

These data suggest that bosentan causes cholestatic liver injury due to inhibition of bile-salt efflux and damage due to intracellular accumulation of bile salts.