Selection of Endothelin Receptor Antagonists in the Treatment of Pulmonary Arterial Hypertension: A Comprehensive Narrative Review

Abstract

One of the major mechanisms in the pathogenesis of pulmonary arterial hypertension (PAH) is mediated by elevated levels of endothelin (ET)-1, which activates both ET_A and ET_B receptors in the pulmonary vasculature. Endothelin receptor antagonists (ERAs) are established treatments for PAH, and three agents—bosentan, ambrisentan, and macitentan—are approved for use in adults in the USA. All are orthosteric antagonists of ET-1 and bind with high affinity to the ET_A receptor, which is found largely on vascular smooth muscle cells (SMCs) and involved in vasoconstriction. Bosentan and macitentan also bind to the ET_B receptor, which is upregulated on SMCs and downregulated on endothelial cells in PAH, resulting in vasoconstriction, cell proliferation, and vascular remodeling. Studies show all three ERAs are efficacious in treating PAH as monotherapy or in combination with other PAH drugs and are generally well tolerated, but all can cause fetal harm and are contraindicated in pregnancy. However, there are no head-to-head clinical trials providing a comprehensive comparison of the overall efficacy and safety of ERAs in PAH. Consequently, we examined the literature on ERAs in PAH through a targeted search of the PubMed/MEDLINE database. This narrative review explores the role of ET-1 in PAH underlying the rationale for ET receptor antagonism. It also discusses the differing physicochemical and pharmacokinetic properties of each ERA and how these unique characteristics influence their receptor binding and kinetics, mechanisms of action, therapeutic effects, dosing frequency, and safety in PAH. In the absence of head-to-head clinical trials assessing their comparative efficacy and safety, it is important to understand both the similarities and the distinguishing characteristics of the three ERAs approved in PAH, to inform individualized treatment selection.

Plain Language Summary

In pulmonary arterial hypertension, or PAH, blood pressure in the arteries of the lungs is higher than normal, caused by narrowing of blood vessels due to unusually high levels of a hormone called endothelin. The heart can become damaged as it works harder to pump blood through the narrower vessels. Patients with PAH have many treatment options, including endothelin receptor antagonists, or ERAs. These work by blocking endothelin from binding to its receptors. Three ERAs are available for the treatment of PAH in the USA—bosentan (Tracleer®), ambrisentan (Letairis®), and macitentan (Opsumit®). Each binds differently to members of the endothelin receptor family, but they all work well in PAH (either alone or combined with other PAH drugs). All three are generally well tolerated. However, there are some safety concerns: bosentan can be harmful to the liver, requiring regular monitoring of liver function, and no ERAs can be taken during pregnancy. To date, no studies have investigated how well each ERA works compared with the others. This review examines studies of ERAs in PAH included in a trusted database of scientific literature. The aims are to explain the physical and chemical properties unique to each agent and to explore how different characteristics might influence how each ERA acts in the body, how effective they are, and how safe they are in individuals with PAH. Understanding the similarities and differences among ERAs is important for individualized treatment selection and effective shared decision-making between healthcare providers and patients.

Key Summary Points

The endothelin receptor antagonists (ERAs) bosentan, ambrisentan, and macitentan are established treatments for use in adults with pulmonary arterial hypertension (PAH), alone or in combination with other PAH drugs; however, there are no head-to-head clinical trials to provide a comprehensive comparison of their efficacy and safety.

This narrative review describes the unique physicochemical and pharmacokinetic characteristics of each ERA and explores how these properties influence their affinity and selectivity for the endothelin (ET) receptors, ETA and ETB, receptor binding kinetics, mechanisms of action, potency, therapeutic effects, dosing frequency, and safety in PAH.

All three ERAs are antagonists of ET-1, binding with high affinity to the ETA receptor, which is largely found on vascular smooth muscle cells (SMCs) and involved in vasoconstriction; bosentan and macitentan also bind to the ETB receptor, which is upregulated on SMCs and downregulated on endothelial cells in PAH, resulting in vasoconstriction, cell proliferation, and vascular remodeling.

All three ERAs are efficacious in treating PAH and are generally well tolerated, but they are contraindicated in pregnancy, and additional unique safety concerns exist, including the ability of bosentan to inhibit the bile salt export pump, which is associated with a risk of hepatotoxicity.

Our findings highlight the importance of understanding the similarities and distinguishing characteristics of each ERA to inform individualized treatment selection in patients with PAH.

Introduction

Pulmonary arterial hypertension (PAH) is a rare, progressive, debilitating disease [1] with an estimated mean global prevalence of three cases in every 100,000 individuals [2]. PAH is characterized by elevated pressure in the pulmonary arteries that can lead to right ventricular (RV) hypertrophy, right heart failure, and death [1, 2]. Current treatment guidelines recommend initial therapy with a combination of an endothelin (ET) receptor antagonist (ERA) and a phosphodiesterase type 5 inhibitor (PDE5i) for all patients with non-vasoreactive idiopathic PAH, heritable PAH, or drug-/toxin-induced PAH, or PAH associated with connective tissue disease (PAH-CTD) who do not have cardiopulmonary comorbidities [3]. The addition of an intravenous/subcutaneous prostacyclin analogue to the ERA + PDE5i combination is recommended for those considered at high risk. These recommendations were reinforced at the 7th World Symposium on Pulmonary Hypertension, during which the combination of ERA + PDE5i was recognized as first-line therapy for all low- or intermediate-risk patients with PAH [4].

It is now over 20 years since the first ERA, bosentan (Tracleer®), was approved by the US Food and Drug Administration (FDA) for use in patients with PAH in 2001 [5]. Today, in 2025, there are three ERAs—bosentan, ambrisentan (Letairis®), and macitentan (Opsumit®)—approved by the FDA to treat adults with PAH [5–7]; the FDA has also approved bosentan to treat children (aged ≥ 3 years) with PAH [5]. In addition, macitentan and the PDE5i tadalafil (Adcirca®) are available as a single-tablet combination therapy, macitentan + tadalafil (Opsynvi®). In 2024, Opsynvi® was approved by the FDA to treat adults with PAH who are in World Health Organization functional classes (WHO FC) II and III [8].

Currently, there are no head-to-head clinical trials to facilitate a comprehensive comparison between bosentan, ambrisentan, and macitentan in terms of overall efficacy and safety, or in key patient subpopulations. Therefore, understanding the similarities and distinguishing characteristics of each ERA is important for clinicians to make informed decisions for individualized treatment selection.

This narrative review provides an overview of the role of ET in PAH and discusses the physicochemical properties, pharmacokinetics (PK), mechanisms of action, efficacy, and safety of the three ERAs currently approved for the treatment of PAH. To identify relevant published literature, the PubMed/MEDLINE database was searched for articles from inception to April 2025 using targeted keywords (alone or in combination). Search terms included PAH, ERAs (bosentan, ambrisentan, macitentan), ET, ET receptor, mechanism of action, PK, efficacy, and safety. Titles and abstracts of retrieved articles were screened for relevance to the review objective, and manual reference searches were conducted to identify any additional studies. Clinical trials, real-world observational studies, meta-analyses, reviews, guidelines, and opinion pieces that were published in English were included, as appropriate. No systematic analysis of the literature or meta-analysis of the study findings was carried out. Although inclusion of authors affiliated with a manufacturer of PAH therapies may have introduced potential bias in the selection or interpretation of studies, diligent efforts were made to ensure objectivity and provide a comprehensive summary of the literature.

Ethics approval was not required—this narrative review was based on previously published studies and no new studies involving human or animal participants were performed.

Rationale for Dual ET Receptor Antagonism in PAH

ET-1 is a potent vasoconstrictor produced mainly by vascular endothelial cells. It plays a major role in cardiovascular homeostasis and is implicated in the pathophysiology of PAH [9, 10]. ET-1 acts as an endogenous agonist for both the ET_A and ET_B receptors [9], which are G protein-coupled cell surface receptors that are found in various organs and tissues but highly expressed in the vasculature of the lungs [9]. Under normal physiological circumstances in the lungs, ET_A receptors are predominantly expressed on vascular smooth muscle cells (SMCs) in the pulmonary arteries; ET_B receptors are also found on vascular SMCs but are mostly expressed by endothelial cells under normal physiological circumstances [9].

The ET_A and ET_B receptors have 63% sequence homology in humans but differ in their function [9]. Under normal physiological conditions, binding of ET-1 to the ET_A and ET_B receptors on SMCs promotes local vasoconstriction of the vascular smooth muscle. Conversely, when ET-1 binds to and activates ET_B receptors on endothelial cells, release of nitric oxide and prostacyclin promotes vasodilation [11]. Activation of the ET_B receptors can also lead to the removal of ET-1 from the circulation [9, 11–13]. In addition, there is evidence suggesting that ET_A and ET_B receptors can form functional heterodimers that block vasoconstriction, which may underlie the synergistic effects observed with dual ET receptor blockade seen in animal models [14].

In PAH, the levels of ET-1 in the plasma and pulmonary vascular endothelium are elevated. In addition, ET_B receptor expression is upregulated on SMCs and downregulated on endothelial cells [12, 13], which results in sustained vasoconstriction, cell proliferation, and vascular remodeling [9].

ERA Physicochemical Properties Influence Clinical Mechanism of Action in PAH

Bosentan, ambrisentan, and macitentan are competitive orthosteric antagonists of ET-1 [15] that differ in their physicochemical properties (Table 1). Bosentan has a high affinity for the ET_A receptor and can also bind to and inhibit the ET_B receptor, whereas ambrisentan is highly selective for the ET_A receptor only [5, 6]. To improve upon bosentan, an extensive medicinal chemistry program was undertaken. This led to the discovery of macitentan, an alkyl sulfamide-substituted pyrimidine, which exhibited potent inhibition of ET_A and significant affinity for ET_B [16].

Table 1.

Mechanisms of action and physicochemical properties of bosentan, ambrisentan, and macitentan

	Bosentan (Tracleer®)	Ambrisentan (Letairis®)	Macitentan (Opsumit®)
Mechanism of action [5–7]	Dual ETA and ETB receptor antagonist with slightly higher affinity for the ETA vs ETB receptor	High-affinity ETA receptor antagonist with high selectivity for the ETA vs ETB receptor	High-affinity dual ETA and ETB receptor antagonist with sustained ET receptor occupancy
Chemical structure [5–7]	C27H29N5O6S·H2O	C22H22N2O4	C19H20Br2N6O4S
Chemical name [5–7]	4-tert-Butyl-N-[6-(2-hydroxy-ethoxy)-5-(2-methoxy-phenoxy)-[2,2′]-bipyrimidin­4-yl]-benzenesulfonamide monohydrate	(+)-(2S)-2-[(4,6-Dimethylpyrimidin-2-yl)oxy]-3-methoxy-3,3-diphenylpropanoic acid	N-[5-(4-Bromophenyl)-6-[2-[(5-bromo-2-pyrimidinyl]oxy]ethoxy]-4-pyrimidinyl]-N′-propylsulfamide
Class [5, 6, 12, 23]	Sulfonamide Highly substituted pyrimidine derivative	Carboxylic acid Propionic acid derivative	Sulfamide Pyrimidine derivative
IC50 ETA (nM) [17]	9.9	0.6	1.3
IC50 ETB (µM) [17]	–	–	14.5
ETA Ki (nM) [6, 14]	4.7	0.011	–
ETB Ki (nM) [14]	95	> 44	–
Receptor occupancy [12, 15]	≈70 s (≈1 min)	≈40 s (< 1 min)	≈1020 s (≈17 min)
Acidity (pKa) [20]	5.1	3.5/4.0 [6]	6.2
Non-ionized form at physiological pH 7.4 (%) [20]	1	0.01	6
Aqueous solubility	Poorly soluble in water and low pH aqueous solutions [5]	Practically insoluble in water and low pH aqueous solutions [6]	Insoluble in water [7]
Lipid/aqueous phase distribution coefficient D [20]	20:1	1:2.5	800:1
Lipophilicity (log D) [20]	1.3	− 0.4	2.9
ETA/ETB [6, 14, 20]	20:1	> 4000:1	50:1 16:1 (aprocitentana)

ET endothelin, IC₅₀ half maximal inhibitory concentration, K_i inhibitory constant, log D distribution constant, pK_a acid dissociation constant

^aMajor active metabolite of macitentan; was referred to as ACT-132577 in earlier studies

Selectivity of the three ERAs for ET_A and ET_B is determined by specific residues in their receptor binding pockets [17]. Bosentan and ambrisentan display receptor binding kinetics typical of high-affinity compounds (i.e., fast association and dissociation rates) [18]. In contrast, macitentan is characterized by slower receptor binding kinetics when compared with bosentan and ambrisentan, leading to prolonged receptor occupancy [15, 18]. Bosentan and ambrisentan exhibit short (~ 1 min) ET_A receptor occupancy [12, 15], whereas macitentan has an ET_A receptor occupancy that is 15- to 20-fold longer than that of bosentan or ambrisentan [15, 18].

Differences in the mode of ligand–receptor binding underlie sustained receptor occupancy by macitentan. Molecular modeling and cryo-electron microscopy indicate that macitentan forms a compact structure in aqueous solution, which fits tightly and deeply into a sub-pocket of the ET_A receptor binding site [17, 18]. Engagement of macitentan with the ET_A receptor principally involves tight hydrophobic interactions with minimal electrostatic binding, while bosentan and ambrisentan display charge-dependent binding because of their stronger acidic and lower lipophilic properties when compared with macitentan [18]. Bosentan still binds well with a slightly higher affinity for the ET_A receptor than ET_B, but it is more selective for the ET_B receptor than macitentan; this is because the corresponding binding site in the ET_B receptor appears to be more expansive than in the ET_A receptor, thus accommodating the larger bulky hydrophobic group of bosentan, which is not present in macitentan [17, 19]. Ambrisentan has very low affinity for the ET_B receptor, but has high affinity and selectivity for the ET_A receptor, which results from robust ionic and hydrogen bonds that form with its carboxylic acid group and that couple with hydrophobic interactions within the ET_A receptor binding pocket [17]. These differences in receptor binding properties are reflected in the relative selectivity of each ERA for the ET_A versus ET_B receptor—bosentan 20:1 [14], ambrisentan > 4000:1 [6, 14], and macitentan 50:1 [20].

Proof of Concept for Dual ET Receptor Antagonism in Preclinical Models

The physicochemical properties of bosentan, ambrisentan, and macitentan also affect how they act in vivo. With the majority of ET-1 produced by the vascular endothelium and released into the underlying interstitial space and ET receptors found largely on SMCs and in the alveolar walls [21], the ability of ERAs to cross lipid membranes is key. Compared with bosentan and ambrisentan, macitentan is less acidic, more lipophilic with a higher distribution constant (i.e., a greater proportion is in the lipid versus aqueous phase), and has a higher acid dissociation constant (i.e., a higher percentage exists in the non-ionized form) [20], all of which favor tissue penetration (Table 1). Both macitentan and its major active metabolite aprocitentan (referred to as ACT-132577 in earlier studies) showed enhanced lung and RV distribution compared with bosentan in preclinical models [22].

Since the major metabolite of macitentan (i.e., aprocitentan) is pharmacologically active, it is also important to understand its mechanism of action when administered in vivo. Aprocitentan is also a dual ET receptor antagonist, with a relative selectivity of 16:1 for the ET_A versus ET_B receptor (Table 1). The systemic exposure to the active metabolite is three times the exposure to macitentan and is expected to contribute approximately 40% of the total pharmacologic activity [7]. A schematic depicting the proposed mechanism of action for each ERA in PAH is shown in Fig. 1.

Fig. 1.

Proposed mechanism of action for each ERA in pulmonary arterial hypertension.

Adapted from Bedan et al., Basic Clin Pharmacol Toxicol, 2018, 123:103–113, with permission. © 2018 Nordic Association for the Publication of BCPT (former Nordic Pharmacological Society). ET-1 produced by the vascular endothelium is an endogenous agonist for two subtypes of the ET-1 receptor: ET_A (predominantly found in PASMCs) and ET_B (found in both PASMCs and ECs). Binding of ET-1 to the ET_A receptor leads to the activation of PLC and PKC signaling, leading to increased intracellular Ca²⁺ and increased Ca²⁺ sensitivity, thus facilitating vascular smooth muscle cell contraction. Binding of ET-1 to ET_B receptors on ECs leads to increased intracellular Ca²⁺, activation of eNOS, NO release, and vasodilatation. In pulmonary arterial hypertension, the levels of ET-1 in the plasma and pulmonary vascular endothelium are elevated, and ET_B receptor expression is upregulated on PASMCs and downregulated on ECs, resulting in sustained vasoconstriction, cell proliferation, and vascular remodeling. Bosentan, macitentan, and the active metabolite of macitentan (aprocitentan) are dual ERAs, binding to both ET_A and ET_B receptors, blocking the effects of ET-1. In contrast, ambrisentan blocks only the ET_A receptor. DAG diacylglycerol, EC endothelial cell, eNOS endothelial nitric oxide synthase, ERA endothelin receptor antagonist, ET-1 endothelin-1, IP₃ inositol 1,4,5-trisphosphate, IP₃R IP₃ receptor, NO nitric oxide, PASMC pulmonary arterial smooth muscle cell, PLC phospholipase, PIP₂ phosphatidylinositol 4,5-bisphosphate, PKC protein kinase C, SR sarcoplasmic reticulum

Differences in ERA PK Influences Therapeutic Effect and Dosing Frequency

Absorption

The PK of each ERA influences its therapeutic effect and dosing frequency. Macitentan is absorbed slowly as a result of its low aqueous solubility, reaching peak plasma concentrations within ~ 8 h (30 h for its active metabolite) [7, 24], compared with ~ 2 h for ambrisentan [6] and 3–5 h for bosentan [5, 25]. After accounting for the time it takes each drug to reach steady state, along with its clearance and elimination properties (Table 2), bosentan should be administered twice daily for the treatment of PAH, while macitentan and ambrisentan can be administered once daily [5–7].

Table 2.

Key pharmacokinetic properties for bosentan, ambrisentan, and macitentan

	Bosentan (Tracleer®)	Ambrisentan (Letairis®)	Macitentan (Opsumit®)
Main metabolic pathway [5–7, 23]	Hepatic metabolism involving CYP2C9 and CYP3A4 (and possibly CYP2C19)	Hepatic metabolism involving CYP3A4 (and to a lesser extent CYP3A5 and CYP2C19) along with glucuronidation by the UGTs 1A9S, 2B7S, and 1A3S	Hepatic metabolism involving CYP3A4 (with minor contributions of CYP2C8, CYP2C9, and CYP2C19)
Metabolites [5, 7, 23, 26]	Three metabolites, including Ro 48-5033, which is pharmacologically active (contributes to 10–20% of potency)	4-Hydroxymethyl ambrisentan, which is not pharmacologically active	Aprocitentana, which is pharmacologically active (contributes to ~ 40% of potency), and ACT-373898, which is not pharmacologically active
Terminal elimination half-life [5–7, 27]	5 h Not reported (Ro 48-5033)	9–15 h	16 h 48 h (aprocitentana)
Bioavailability [5–7, 23, 26]	~ 50% No food effect	90% No food effect	~ 74% No food effect
Plasma protein-bound [5–7]	98% (mainly albumin)	99%	> 99% (primarily albumin, also alpha-1-acid glycoprotein)
Tmax [5–7, 24]	3–5 h	~ 2 h	~ 8 h 30 h (aprocitentana)
Steady state [5, 23, 24, 28]	3–5 days	4 days	3 days 7 days (aprocitentana)
Vss/F [5, 7, 25]	18 L (following single IV dose) [5, 25] 30 L (following multiple oral doses) [25]	–	50 L 40 L (aprocitentana)
Clearance [5, 6]	4 L/h (following single IV dose in PAH) [5] 17 L/h (following multiple oral doses in HV) [25]	19 mL/min (PAH) 38 mL/min (HV)	–
Elimination [5, 7, 23]	Mainly biliary Urinary < 3%	Biliary ~ 80% Urinary ~ 20%	Urinary ~ 50% Fecal ~ 24%

CYP cytochrome P450, HV healthy volunteers, IV intravenous, PAH pulmonary arterial hypertension, T_max time to peak plasma concentration, UGTs uridine 5′-diphosphate glucuronosyltransferases, Vss/F volume of distribution at steady state

^aMajor active metabolite of macitentan; was referred to as ACT-132577 in earlier studies

Metabolism

All three ERAs undergo hepatic metabolism involving cytochrome P450 (CYP) enzymes, including CYP3A4 (Table 2). Unlike bosentan and macitentan, ambrisentan also undergoes hepatic glucuronidation [6]. Ambrisentan does not have any active metabolites, in contrast to both bosentan and macitentan [5, 7, 23, 26]. Evidence suggests that 10–20% of the potency of bosentan may be due to its pharmacologically active metabolite, Ro 48-5033 [25], while up to 40% of the potency of macitentan may be due to its pharmacologically active metabolite, aprocitentan [7]. This affects how long the drugs remain active after administration. The terminal elimination half-life for bosentan is 5 h, while for ambrisentan and macitentan, it is substantially longer (9–15 h and 16 h, respectively) [5–7]. Furthermore, the active metabolite for macitentan has a terminal elimination half-life of 48 h, which is important for its continued activity as an ERA.

Special Populations

Hepatic Impairment

Bosentan undergoes extensive hepatic metabolism, and systemic exposure to bosentan and its active metabolite increase significantly in patients with moderate hepatic impairment [5]. In general, bosentan should be avoided in patients with moderate or severe liver impairment. Similarly, ambrisentan is not recommended in patients with moderate or severe hepatic impairment and exposure to ambrisentan may be increased in patients with mild hepatic impairment [6]. Evidence shows dose adjustments are not needed for macitentan in patients with mild, moderate, or severe hepatic impairment [7, 29, 30].

Renal Impairment

The effect of severe renal impairment on the PK of bosentan is small with no dosage adjustment required [5]. Ambrisentan has shown no significant impact in patients with mild or moderate renal impairment and no dose adjustment is required for these patients. There are no data on ambrisentan in patients with severe renal impairment [6]. Levels of macitentan and its active metabolite are increased in patients with severe renal impairment compared with healthy subjects, but the increase is not considered clinically relevant and dose adjustments are not required [7, 29, 30].

Drug–Drug Interactions

Differences in metabolism between the ERAs manifest in varying potential for drug–drug interactions with other CYP substrates, including sildenafil, tadalafil, ketoconazole, cyclosporine, glyburide, lopinavir/ritonavir, rifampin, and hormonal contraceptives, with some contraindications or recommended dose adjustments (Table 3). Bosentan has the potential for significant clinically relevant drug–drug interactions [25]. Its use is contraindicated with cyclosporine A and glyburide, and additional monitoring/dose adjustments are required when it is co-administered with hormonal contraceptives, simvastatin, lopinavir/ritonavir, and rifampicin [5]. Despite undergoing both glucuronidation and CYP-mediated metabolism, ambrisentan has little potential for drug–drug interactions [6, 31–34]. Only cyclosporine has been shown to increase exposure to ambrisentan; it is therefore recommended that the daily dose of ambrisentan be limited to 5 mg when co-administered with cyclosporine [6, 31]. Studies with macitentan show some drug interaction potential [24]. Concomitant use with strong inducers of CYP3A4 (e.g., rifampin), inhibitors of CYP3A4 (e.g., ketoconazole and ritonavir), and moderate dual or combined inhibitors of CYP3A4 and CYP2C9 should be avoided [7].

Table 3.

Key drug–drug interactions for bosentan, ambrisentan, and macitentan

	Bosentan (Tracleer®)	Ambrisentan (Letairis®)	Macitentan (Opsumit®)
Drug–drug interactions: effect of ERA on other drugs [5–7]	Decreased exposure of sildenafil (− 63%), cyclosporine A (contraindicated), glyburide (contraindicated), simvastatin, warfarin, norethindrone, and ethinyl estradiol	None	Increased sildenafil exposure (+ 15%)—not clinically relevant
Drug–drug interactions: effect of other drugs on ERA [5–7]	Increased bosentan exposure with sildenafil (+ 50%), cyclosporine A (contraindicated), lopinavir/ritonavir, and ketoconazole Decreased bosentan exposure with rifampin	Increased ambrisentan exposure with cyclosporine (dose limited to 5 mg/day)	Increased macitentan exposure with ketoconazole and ritonavir (avoid strong CYP3A4 inducers) Decreased macitentan exposure with rifampin (avoid strong CYP3A4 inhibitors) Avoid moderate dual or combined CYP3A4 and CYP2C9 inhibitors
Enzymatic induction [5–7]	Induces CYP3A and CYP2C9	No inducing effects on CYP enzymes	No inducing effects on CYP enzymes

CYP cytochrome P450, ERA endothelin receptor antagonist

Clinical Efficacy and Safety of ERAs in PAH

Efficacy in Clinical Trials, Real-World Studies, and Meta-Analyses

The efficacy of bosentan, ambrisentan, and macitentan in PAH has been established in clinical trials, with each ERA administered either as monotherapy or in combination with other therapies (e.g., dual therapy with a PDE5i and triple therapy with a PDE5i and a prostacyclin analogue; Table 4). Real-world studies (i.e., observational/retrospective registries, chart reviews, and post-marketing surveillance studies) and meta-analyses have also been conducted.

Table 4.

Published studies with bosentan, ambrisentan, and macitentan in adults with pulmonary arterial hypertension

Study identifier/citation	PAH subtype	Design	Duration	Treatment	Number	Female sex (%)	Age (years)	FC I/II/III/IV (%)	BL 6MWD (m)	Efficacy	Most common AEs (≥ 10% for ERA)	SAEs	Abnormal hepatic function	Hb
Monotherapy
Bosentan: clinical studies
Study 351 (Channick et al., 2001) [35]	PAH	R, DB, PC	12 W minimum	Bosentan 125 mg BIDa	21	81	52 (12)	0/0/100/0b	360 (86)	Primary: Change from BL in 6MWD at 12 W: + 70 vs − 6 m; P = 0.021*	AEs were similar in nature and frequency between groups (not specified)	–	Transient increases in ALT/AST in 2 patients on bosentan	–
				Placebo	11	100	47 (14)	0/0/100/0b	355 (82)
BREATHE-1 (Rubin et al., 2002) [36]	PAH	R, DB, PC	16 (+ 12) W	Bosentan 125 or 250 mg BIDa	144	79	49 (16)	0/0/90/10b	330 (74)	Primary: Change from BL in 6MWD at 16 W: + 36 vs − 8 m; P < 0.001*	Headache (21% vs 19%), dizziness (11% vs 19%)	–	ALT/AST > 8 × ULN: 10% vs 0%	–
				Placebo	69	78	47 (16)	0/0/94/6b	344 (76)
BREATHE-2 (Humbert et al., 2004) [46]	Severe PAH	R, DB, PC	16 W	Bosentan 125 mg BIDa + epoprostenol	22	77	45 (17)	0/0/77/23c	–	Primary: Change from BL in RHC-determined TPR at 16 W: − 36.3% vs − 22.6%; P = 0.08	Most common AEs (≥ 20% on ERA): pain in jaw (59% vs 91%), diarrhea (55% vs 27%), flushing (27% vs 45%), headache (27% vs 36%), limb pain (23% vs 18%)	14% vs 18%	Hepatic dysfunction: 18% vs 9%	–
				Placebo + epoprostenol	11	55	47 (19)	0/0/73/27c	–
EARLY	WHO FC II PAH	R, DB, PC (Galiè et al., 2008) [47]	6 M	Bosentan 125 mg BIDa	93	76	45 (18)	0/100/0/0	438 (86)	Co-primary: PVR (expressed as % of BL) at 6 M: 83% vs 107.5%; P < 0.0001* Co-primary: Change from BL in 6MWD at 6 M: + 11.2 vs − 7.9 m; P = 0.0758	Most common AEs (≥ 5% on ERA): nasopharyngitis (8% vs 9%), abnormal liver test (8% vs 3%), peripheral edema (6% vs 8%), nausea (5% vs 9%), dizziness (5% vs 5%), chest pain (5% vs 4%)	13% vs 9%	ALT/AST > 3 × ULN: 13% vs 2%	–
				Placebo	92	63	44 (17)	0/100/0/0	431 (91)
		R, DB, PC + OLE (Simonneau et al., 2014) [48]	43 (22) M FU	Bosentan 125 mg BIDa	173d	69	45 (18)	4/93/4/0b	436 (94)	Change from BL in 6MWD at FU: − 3.7 m (95% CI − 17.0, 9.5)	20% had an AE leading to treatment discontinuation, most commonly abnormal liver test (7%) and PAH worsening (4%)	–	ALT/AST > 3 × ULN: 17% > 8 × ULN: 8%	≤ 10 g/dL: 15%
Bosentan: meta-analyses
Chen et al., 2018 [49]	PAH	8 RCTs	12–18 W (6 M in 2 studies)	Bosentan 125 mg BID (250 mg BID in one study, + epoprostenol in one study, and + sildenafil in one study)	536	–	–	–	–	Significantly higher 6MWD (WMD + 35.7, 95% CI 23.1, 48.4; P < 0.001*), reduced mPAP (WMD − 5.7, 95% CI − 8.1, − 3.3; P < 0.001*), increased cardiac index (WMD + 0.4, 95% CI 0.13, 0.70; P = 0.004*), reduced PVR (WMD − 305.1, 95% CI − 526.5, − 83.0; P = 0.007*), and lowered risk of clinical worsening (RR 0.48, 95% CI 0.27, 0.85; P = 0.01*) vs placebo No significant difference in improving FC vs placebo	Incidence of AEs not significantly different vs placebo (RR 1.00, 95% CI 0.93, 1.06; P = 0.89)	Incidence of SAEs not significantly different vs placebo (RR 0.83, 95% CI 0.68, 1.02; P = 0.08)	Higher rate of liver function abnormality vs placebo (RR 2.57, 95% CI 1.64, 4.05)	–
				Placebo	467	–	–	–	–
Lee & Song, 2013 [50]	PAH	7 RCTs	12–16 W (18 M in one study)	Bosentan 125 mg BIDa (250 mg BID in one study and + epoprostenol in another)	410	–	–	–	–	Significantly higher 6MWD (WMD + 46.2, 95% CI 21.2, 71.2; P < 0.0001*), reduced mPAP (WMD − 6.03, 95% CI − 8.8, − 3.3; P < 0.0001*), lower rates of clinical worsening (OR 0.25, 95% CI 0.14, 0.45; P < 0.0001*), and higher FC amelioration (OR 1.65, 95% CI 1.05, 2.60; P = 0.031*) vs placebo No significant difference in all-cause mortality vs placebo (OR 0.84, 95% CI 0.22, 3.30; P = 0.805)	–	Incidence of SAEs not significantly different vs placebo (OR 0.95, 95% CI 0.56, 1.61; P = 0.843)	Significantly more abnormal liver test results vs placebo (OR 2.31, 95% CI 1.02, 5.24; P = 0.045*)	–
				Placebo	296	–	–	–	–
Ambrisentan: clinical studies
Galiè et al., 2005 [51]	PAH	DB, DR	12 W (+ 12 W OLE)	Ambrisentan 1–10 mg QD	64	84	51 (16)	0/36/64/0b	343 (79)	Primary: Change from BL in 6MWD at 12 W: + 36 m; P < 0.0001*	Peripheral edema (25%), nasal congestion (19%), URTI (19%), headache (16%), flushing (13%), nausea (13%)	11%	ALT/AST > 3 × ULN: 3% over 24 W	Change from BL: 0.8 g/dL
ARIES-1 (Galiè et al., 2008) [37]	PAH	R, DB, PC	12 W	Ambrisentan 5 mg QD	67	84	53 (14)	2/30/60/9b	340 (77)	Primary: Placebo-corrected change from BL in 6MWD at 12 W: + 31 m for 5 mg (P = 0.008*) and + 51 m for 10 mg; P < 0.001*	Peripheral edema (27% and 28% vs 10%), headache (18% and 19% vs 21%), nasal congestion (6% and 10% vs 3%)	10% vs 17% (combined groups)	ALT/AST > 3 × ULN: 0% vs 2% (combined groups)	Change from BL: − 0.84 vs + 0.2 g/dL (combined groups)
				Ambrisentan 10 mg QD	67	79	49 (16)	3/33/54/10b	341 (78)
				Placebo	67	88	48 (16)	3/34/61/2b	342 (73)
ARIES-2 (Galiè et al., 2008) [37]	PAH	R, DB, PC	12 W	Ambrisentan 2.5 mg QD	64	75	52 (15)	0/53/45/2b	347 (84)	Primary: Placebo-corrected change from BL in 6MWD at 12 W: + 32 m for 2.5 mg (P = 0.022*) and + 59 m for 5 mg (P < 0.001*)	Headache (13% and 8% vs 6%)
				Ambrisentan 5 mg QD	63	81	50 (16)	2/44/52/2b	355 (84)
				Placebo	65	68	51 (14)	3/37/57/3b	343 (86)
ARIES-E	PAH	OL LTE to ARIES-1 and ARIES-2 (Oudiz et al., 2009) [52]	2 Y	Ambrisentan 2.5 mg QD	96	72	52 (15)	1/53/41/5b	350 (87)	Primary: Change from BL in 6MWD at 2 Y: + 7 m for 2.5 mg, + 23 m for 5 mg, and + 28 m for 10 mg	Peripheral edema, headache, URTI, dizziness	–	ALT/AST > 3 × ULN: 3% > 8 × ULN: < 1%	Change from BL: − 1 g/dL in all groups
				Ambrisentan 5 mg QD	190	81	51 (14)	4/40/48/8b	348 (87)
				Ambrisentan 10 mg QD	97	81	49 (17)	4/36/50/10b	342 (81)
		Post hoc hemodynamic analysis according to RHC FU (Klinger et al., 2011) [53]	2 Y	Ambrisentan 2.5–10 mg QD with RHC FU	68	71	55 (15)	25 (I/II) and 75 (III/IV)b	362 (72)	Change from BL at time of FU RHC: PAP − 7.6 mmHg, PVR − 266 dyn·s/cm5, and cardiac index 0.4 L/min/m2 Changes from BL in mPAP and PVR inversely correlated with change from BL in 6MWD (r = − 0.41 and − 0.43; both P < 0.001*)	–	–	–	–
Ambrisentan: observational studies
Korean PMS study (Jeon et al., 2023) [54]	PAH (new or repeat ambrisentan users)	OL, MC, PMS (2015–2021)	609 (685) day FU	Ambrisentan 5 or 10 mg QD	293	77	51 (15)	0/40/60/0b	–	Significant improvement from BL in 6MWD: + 37 (101) m (P = 0.015*); n = 67	Most common AEs (≥ 5%): dyspnea (9.6%), peripheral edema (6.5%), dizziness (6.1%), headache (5.1%)	21%	ALT increased (n = 3) AST increased (n = 3)	Hb decrease (n = 1)
VOLT (Vachiéry et al., 2017) [55]	PAH	OL, MC, PRO, OBS (2008–2011)	2.2 (1.4) patient-year FU	Ambrisentan 5 mg QD (85%) or 10 mg QD (15%)	998	67	60 (15)	2/26/65/7b	–	–	Peripheral edema (23%), dyspnea (15%), anemia (14%), heart failure (13%)	–	ALT/AST > 3 × ULN: 0.02 per patient-year	–
Macitentan: clinical studies
SERAPHIN (Pulido et al., 2013; Channick et al., 2015; Mehta et al. 2017; Opsumit PI) [7, 40, 56, 57]	PAH	R, DB, PC	Event-driven	Macitentan 10 mg QD	242	80	46 (15)	< 1/50/48/2b	363 (93)	Primary: Time to the first occurrence of event related to PAH or death: 31% vs 46% (HR 0.55, 97.5% CI 0.32, 0.76; P < 0.001*) Risk and rates of all-cause hospitalization reduced by 32.3% (P = 0.0051*) and 33.1% (P = 0.0005*) vs placebo (Channick et al., 2015) [56] Significant improvement vs placebo in 7/8 SF-36 domains and PCS/MCS scores at 6 M and reduced risk of clinically meaningful deterioration in PCS (HR 0.60, 95% CI 0.47, 0.76; P < 0.0001*) and MCS (HR 0.76, 95% CI 0.61, 0.95; P = 0.0173*) at EOT (Mehta et al., 2017) [57]	PAH worsening (22% vs 35%), peripheral edema (18% vs 18%), URTI (15% vs 13%), nasopharyngitis (14% vs 10%), headache (14% vs 9%), anemia (13% vs 3%), bronchitis (12% vs 6%), dizziness (11% vs 11%)	45% vs 55%	ALT/AST > 3 × ULN: 3% vs 5% > 8 × ULN: 2% vs < 1%	≤ 8 g/dL: 4% vs < 1%
				Placebo	250	74	47 (17)	0/52/47/2b	352 (111)
		Hemodynamic substudy (Galiè et al., 2017) [58]	6 M	Macitentan 10 mg QD	57	81	47 (15)	2/42/53/4b	359 (94)	Significant change from BL vs placebo in cardiac index (+ 0.63 L/min/m2), mPAP (− 6.4 mmHg), PVR (− 38.5 dyn·s/cm5), and NT-proBNP (− 303 fmol/mL; all P < 0.05* Baseline and 6 M cardiac index, RAP and NT-proBNP associated with morbidity/mortality events	PAH worsening (20% vs 47%), headache (23% vs 6%), RVF (16% vs 21%), nasopharyngitis (16% vs 15%), viral respiratory tract infection (18% vs 9%), bronchitis (19% vs 4%)	49% vs 63%	ALT/AST > 3 × ULN: 4% vs 2%	≤ 10 g/dL: 5% vs 2%
				Placebo	68	74	48 (16)	0/47/52/2b	360 (112)
		OLE (Souza et al., 2022) [59]	3.8 (2.5) Y FU	Macitentan 10 mg QD	550	80	48 (16)	5/45/32/19b	332 (150)	–	Most common AEs (≥ 15% on ERA): PAH worsening (29%), URTI (23%), peripheral edema (20%), nasopharyngitis (19%), anemia (18%), bronchitis (16%), RVF (16%)	64%	ALT or AST > 3 × ULN: 8%	≤ 8 g/dL: 6%
REPAIR	PAH	Single-arm, OL (Vonk Noordegraaf et al., 2022) [60]	52 W	Macitentan 10 mg QD	71	80	45 (19–71)e	1/48/51/0b	411 (121)	Primary: Change from BL in RVSV at 26 W: + 12 mL; P < 0.0001* Primary: Change from BL in PVR at 26 W: − 38% (geometric mean 0.62 dyn·s/cm5); P < 0.0001*	Peripheral edema (22%), headache (21%), dizziness (14%), cough (12%), Hb decrease (12%), URTI (12%), myalgia (10%) AESIs: edema and fluid retention (30%), anemia (20%)	16%	ALT/AST ≥ 3 × ULN: 6% ALT/AST ≥ 8 × ULN: 0%	≤ 8 g/dL: 4% Decrease from BL ≥ 2 and ≤ 5 g/dL: 58% Decrease from BL ≥ 5 g/dL: 4%
		Echo substudy (Torbicki et al., 2024) [61]	52 W	Macitentan 10 mg QD	45	82	45 (35, 57)f	0/31/69/0b	387 (324, 543)f	Improvements from BL in echo-assessed RVSV, LVSV, LVEDV, RVFAC, TAPSE, and 2D GLRVS at 26 W and 52 W	Peripheral edema (27%), headache (24%), dizziness (20%), cough (16%), diarrhea (16%), myalgia (16%), nasopharyngitis (13%), back pain (11%), nasal congestion (11%)	11%	–	–
Macitentan: observational studies
OPUS / OrPHeUS (McLaughlin et al., 2022) [62]	Newly treated PAH	PRO, OBS, registry (OPUS; 2014–2020) MC, RETRO, chart review (OrPHeUS; 2013–2017)	14.5 (5.2, 29.0) Mf	Macitentan monotherapy (38%), double therapy (48%), and triple therapy (14%)	4626 (OPUS 2264; OrPHeUS 2362)	76	62 (51, 71)f	8/29/57/7b (n = 2513)	293 (195, 378);f n = 1670	≥ 1 hospitalization: 46% Free from hospitalization (KM estimate): 60% (95% CI 58, 62) at 1 Y Survival (KM estimate): 90% (95% CI 89, 91) at 1 Y and 75% (95% CI 73, 77) at 3 Y	OPUS only: dyspnea (23%), headache (12%), peripheral edema (11%), nausea (10%)	–	ALT/AST > 3 × ULN: 3% > 5 × ULN: 1% > 3 × ULN and TBIL > 2 × ULN: 1%	OPUS only: ≥ 1 event of anemia/Hb decrease: 10%

Study identifier/citation	PAH subtype	Design	Duration	Treatment	Number	Female sex (%)	Age (years)	FC I/II/III/IV (%)	BL 6MWD (m)	Efficacy	Most common AEs (≥ 10% for ERA)	SAEs	Abnormal hepatic function	Hb
Combination therapy
Bosentan: clinical studies
COMPASS-1 (Grünig et al., 2009) [63] Sequential combination	PAH	OL, MC, PRO	60 min (acute hemodynamic effects)	Bosentan 125 mg BID (for ≥ 12 W) + sildenafil 25 mg (single dose)	45	76	53 (15)	0/42/58/0b	409 (107)	Primary: Percent change from BL in PVR 60 min after sildenafil administration: − 15%; P < 0.0001*	11 AEs reported in 7 patients	0%	–	–
COMPASS-2 (McLaughlin et al., 2015) [42]	PAH	R, DB, PC	Event-driven	Bosentan 125 mg BID + sildenafil ≥ 20 mg TIDg	159	79	53 (15)	0/45/55/0b	363 (79)	Primary: Time to first morbidity/mortality event:h 43% vs 51% (HR 0.83, 97.31% CI 0.58, − 1.19; P = 0.2508)	Most common AEs (≥ 15% on combo): worsening PAH (25% vs 35%), peripheral edema (19% vs 16%), dyspnea (16% vs 15%), headache (15% vs 14%)	46% vs 59%	ALT/AST > 3 × ULN: 22% vs 6% > 8 × ULN: 5% vs 0%	Decrease to ≤ 10 g/dL: 9% vs 9%
				Placebo + sildenafil ≥ 20 mg TIDg	175	73	55 (16)	0/39/59/1b	358 (73)
COMPASS-3 (Benza et al., 2018) [64] Sequential combination	PAH	OL, MC, exploratory	28 W (add-on sildenafil at W 16 if 6MWD < 380 m at W 16)	Bosentan 125 mg BIDa for 28 W	16	82	58 (21–84)e	1/17/79/3b	274 (152–360)e	Primary: Patients (%) with 6MWD ≥ 380 m: 56% (mono) vs 20% (combo) at 28 W; 31% at 16 and/or 28 W (mono + combo)	Peripheral edema (20%), dyspnea (18%), headache (13%) at 1 Y More frequent on combo vs mono during W 16–28 (70% vs 38%)	30% (most commonly dyspnea 4% and pneumonia 4%) at 1 Y	3 patients withdrew due to abnormal liver function tests	–
				Bosentan 125 mg BIDa for 16 W + sildenafil 20 mg TID W 16–28	76
PHIRST (Barst et al., 2011) [65] Sequential combination	PAH on stable bosentan for ≥ 12 W	R, DB, PC	16 W	Bosentan + tadalafil 20 mg QD	45	80	50 (15)	0/29/71/0b	346 (71)	Primary: Placebo-corrected change from BL in 6MWD + 23 m (95% CI − 0.5, 46) for tadalafil 20 mg and + 23 (95% CI − 2.4, 48) for tadalafil 40 mg	Most common AEs (≥ 10% in combined tadalafil groups): headache (33% vs 20%), dizziness (11% vs 7%), and dyspepsia (10% vs 4%)	n = 22 (combined tadalafil groups) vs n = 7 (placebo)	–	–
				Bosentan + tadalafil 40 mg QD	42	79	50 (13)	2/40/58/0b	361 (75)
				Bosentan + placebo	45	78	52 (16)	0/29/69/2b	349 (85)
Bosentan: observational studies
Sitbon et al., 2014 [66] Upfront combination	Newly diagnosed severe PAH	RETRO, registry	32 (19) M FU	Bosentan 125 mg BIDa + sildenafil 20 mg TID + epoprostenoli	19	89	39 (14)	0/0/42/58c	215 (174)	Significant change from BL in PVR, 6MWD, NYHA FC, and cardiac index and other hemodynamic parameters at 4 M and final FU visit; all P < 0.01*	Most AEs typical of epoprostenol therapy (jaw pain, headache, diarrhea, or flushing)	–	ALT/AST > 8 × ULN: 2 patients	–
Combination real-world study (Dardi et al., 2015) [67] Sequential combination	PAH (suboptimal response to prior bosentan or sildenafil)	SC, RETRO, chart review (2003–2013)	–	Bosentan 125 mg BIDa + sildenafil 20 mg TID	181	60	50 (36, 62)f	48 (I/II), 52 (III/IV)c	425 (307, 491)f	Significant improvements in FC, 6MWD, RAP, mPAP, cardiac index, and PVR at 3–4 M; all P < 0.01*	–	–	Liver test elevations in 5 patients	–
Ambrisentan: clinical studies
AMBITION (Galiè et al., 2015; Vachiéry et al., 2019) [41, 68] Upfront combination	PAH (newly diagnosed)	R, DB, PC	Event-driven	Ambrisentan 10 mg QD + tadalafil 40 mg QD	253	74	55 (14)	0/30/70/0b	354 (88)	Primary: Risk of event related to PAH or death: 18% (combo) vs 34% (ambrisentan mono; HR 0.48, 95% CI 0.31, 0.72; P < 0.001*), 28% (tadalafil mono; HR 0.53, 95% CI 0.34, 0.83), and 31% (pooled mono; HR 0.50, 95% CI 0.35, 0.72; P < 0.001*) Risk of hospitalization for worsening PAH: − 63% reduction with combo vs pooled mono (HR 0.37, 95% CI 0.21, 0.64; P = 0.0002*)	Most common AEs (≥ 20% on combo/ambrisentan/ tadalafil): peripheral edema (45/33/28%), headache (42/33/35%), nasal congestion (21/15/12%), diarrhea (20/23/19%), dizziness (20/19/12%) AESI: hypotension (8/7/7%), syncope (5/6/8%)	36% (combo) vs 36% (ambrisentan) and 41% (tadalafil)	–	–
				Ambrisentan 10 mg QD + placebo	126	79	54 (15)	0/30/70/0b	354 (92)
				Tadalafil 40 mg QD + placebo	121	83	55 (15)	0/34/66/0b	349 (92)
				Pooled monotherapy	247	81	54 (15)	0/32/68/0b	352 (92)
ATHENA-1 (Shapiro et al., 2017) [69] Sequential combination	PAH (suboptimal response to prior PDE5i)	Single-arm, OL	48 W	Ambrisentan 10 mg QD + PDE5ij	33	79	48 (14)	0/0/100/0b	362 (99)	Primary: Change from BL in PVR at 24 W: − 249 dyn·s/cm5 (95% CI − 338, − 160; P < 0.001*; − 32% improvement)	Most common AEs (≥ 15%): nasal congestion (30%), URTI (24%), headache (21%), peripheral edema (21%)	–	–	–
D’Alto et al., 2020 [70] Upfront combination	Newly diagnosed, severe, non-reversible PAH	MC, RETRO, chart review (2014–2018)	24 (14) M FU	Ambrisentan, tadalafil, subcutaneously administered treprostinil	21	76	44 (15)	0/0/57/43b	158 (130)	Significant improvements in 6MWD, NT-proBNP levels, hemodynamic parameters (RAP, mPAP, cardiac input, PVR, and SVO2), and RV function/structure (right-sided atrial area, RV end-diastolic and systolic area, fractional area change, and LV eccentricity index) at 24 M FU (all P < 0.001*); 90% of patients reached WHO FC I/II	Mild-to-moderate peripheral edema (19%), nasal congestion (10%), flushing (14%)	–	ALT/AST > 3 × ULN: 0 patients	–
Zhuang et al., 2014 [71] Sequential combination	PAH on ambrisentan	R, DB, PC	16 W	Ambrisentan + tadalafil 40 mg QD	60	77	52 (12)	0/60/35/5b	356 (87)	Primary: Change from BL in 6MWD + 54 vs + 18 m at 16 W; P < 0.05*	Headache (60% vs 20%), myalgia (10% vs 5%), flushing (8% vs 3%), dyspepsia (8% vs 6%)	–	–	–
				Ambrisentan + placebo	64	81	51 (14)	0/58/42/3b	343 (71)
Macitentan: clinical studies
SERAPHIN (Jansa & Pulido, 2018; Pulido et al., 2013) [40, 72]	PAH	R, DB, PC	Event-driven	Macitentan 10 mg + background therapyk	154	76	45 (15)	0/47/50/3b	364 (97)	Primary: Risk of event related to PAH or death: HR 0.62 (95% CI 0.43, 0.89; P = 0.009*)	Most common AEs (≥ 15% on combo): worsening PAH (23% vs 37%), peripheral edema (20% vs 24%), anemia (16% vs 5%), URTI (16% vs 18%)	–	–	–
				Placebo + background therapyk	154	73	46 (17)	0/51/48/1b	360 (111)
A DUE (Grünig et al., 2024) [73] Fixed-dose combination	PAH (newly diagnosed and prior ERA or PDE5i treated)	R, DB, PC	16 W + 2 Y OLE	Macitentan 10 mg QD + tadalafil 40 mg QDk	107	77	49 (16)	0/61/39/0b	352 (96)	Primary: Change from BL in PVR at 16 W: − 29% for combo vs macitentan mono (P < 0.0001*) and − 28% vs tadalafil mono (P < 0.0001*)	Most common AEs (≥ 10%) for combo vs macitentan mono and tadalafil mono: headache (17% vs 17% and 14%), peripheral edema (13% vs 11% and 11%), diarrhea (5% vs 0% and 14%) AESIs: edema/fluid retention (21% vs 14% and 16%), anemia (19% vs 3% and 2%), hypotension (7.5% vs 0% and 0%), hepatic disorders (1% vs 3% and 9%)	14% vs 9% and 9%	–	Hb decrease (n = 1)
				Macitentan 10 mg QD	35	83	51 (16)	0/31/69/0b	347 (89)
				Tadalafil 40 mg QDl	44	77	51 (14)	0/43/57/0b	362 (70)
TRITON (Chin et al., 2021) [74] Upfront combination	PAH (newly diagnosed)	R, DB, PC	26 W	Macitentan 10 mg QD + tadalafil 20–40 mg QD + selexipag 200–1600 µg BID	123	76	52 (14)	20 (I/II), 80 (II/III)b	342 (121)	Primary: Change from BL in PVR at 26 W: − 54% vs − 52%; P = 0.42	Most common AEs (≥ 30% on triple combo): headache (69% vs 61%), diarrhea (54% vs 32%), nausea (48% vs 25%), peripheral edema (37% vs 36%), pain in extremity (30% vs 16%)	43% vs 32%	–	–
				Macitentan 10 mg QD + tadalafil 20–40 mg QD + placebo	124	76	52 (14)	20 (I/II), 80 (II/III)b	347 (117)
TRITON and REPAIR- combined (McLaughlin et al., 2024) [75] Upfront combination	PAH	Combined post hoc analysis	26 W	Macitentan 10 mg QD + tadalafil 40 mg QD	148	75	51 (38, 65)f	1/19/77/3b	356 (121)	Change from BL in PVR at 26 W: − 53% Change from BL in 6MWD at 26 W: 57 m (95% CI 42, 72)	Most common AEs (≥ 20%): headache (58%), peripheral edema (35%), diarrhea (29%), nausea (23%)	28%	–	–
Macitentan: observational studies
OPUS / OrPHeUS (Chin et al., 2024) [76]	Newly treated PAH	PRO, OBS, registry (OPUS; 2014–2020) MC, RETRO, chart review (OrPHeUS; 2013–2017)	NR	Macitentan and tadalafil double therapy	1336 (453 incident; 837 prevalent)	76	60 (49, 70)f	8/29/58/6b (n = 666)	310 (228, 400);f n = 530	Free from hospitalization (KM estimate): 62% (95% CI 59, 65) at 1 Y, 47% (95% CI 44, 50) at 2 Y Survival (KM estimate): 92% (95% CI 90, 94) at 1 Y, 76% (95% CI 72, 79) at 3 Y	OPUS only: dyspnea (21%), headache (12%), nausea (11%)	OPUS only: 48%	Hepatic events: 5%	OPUS only: ≥ 1 event of anemia/Hb decrease: 10%

The data are presented as % or mean (SD), unless otherwise stated

2D GLRVS 2-dimensional global longitudinal right ventricular strain, 6MWD 6-min walking distance, AE adverse event, AESI adverse event of special interest, ALT alanine aminotransferase, AST aspartate aminotransferase, BID twice daily, BL baseline, CI confidence interval, combo combination, DB double-blind, DR dose-ranging, ERA endothelin receptor antagonist, EOT end of treatment, FC functional class, FU follow-up, Hb hemoglobin, HR hazard ratio, IV intravenous, KM Kaplan–Meier, LTE long-term extension, LV left ventricular, LVEDV left ventricular end diastolic volume, LVSV left ventricular stroke volume, M months, MC multicenter, MCS mental component summary, mono monotherapy, mPAP mean pulmonary arterial pressure, NR not reported, NT-proBNP N-terminal B-type natriuretic peptide, NYHA New York Heart Association, OBS observational, OL open label, OLE open-label extension, OR odds ratio, PAH pulmonary arterial hypertension, PAP pulmonary arterial pressure, PC placebo controlled, PCS physical component summary, PDE5i phosphodiesterase type 5 inhibitor, PMS post-marketing surveillance, PRO prospective, PVR pulmonary vascular resistance, Q quartile, QD once daily, R randomized, RAP right atrial pressure, RCT randomized controlled trial, RETRO retrospective, RHC right heart catheterization, RR risk ratio, RV right ventricular, RVF right ventricular failure, RVFAC right ventricular fractional area change, RVSV right ventricular stroke volume, SAE serious adverse event, SC single-center, SD standard deviation, SF-36 36-item short form questionnaire, SvO₂ mixed venous oxygen saturation, TAPSE tricuspid annular plane systolic excursion, TBIL total bilirubin, TID three times daily, TPR total pulmonary resistance, ULN upper limit of normal, URTI upper respiratory tract infection, W weeks, WHO World Health Organization, WMD weighted mean difference, Y years

*Significant P value

^aPatients received bosentan 62.5 mg BID for the initial 4 weeks

^bWHO FC

^cNYHA FC

^dPatients received > 1 dose of bosentan during the DB or OLE phase

^eMedian (range)

^fMedian (Q1, Q3)

^gAfter 16 weeks, the dose of sildenafil could be adjusted or discontinued, or patients could be switched to tadalafil

^hDefined as all-cause death, hospitalization for PAH worsening or IV prostanoid initiation, atrial septostomy, lung transplant, or PAH worsening

ⁱEpoprostenol was administered by IV infusion to a maximum dose of 16 ng/kg/min according to tolerance

^jPatients received ambrisentan at a dose of 5 mg QD for 4 weeks, then 10 mg for 20 weeks, sildenafil at doses ranging from 20 to 100 mg TID, and tadalafil to a maximum dose of 40 mg QD

^kBackground therapy consisted of a PDE5i in 97.4% of patients

^lTadalafil dose was increased from 20 to 40 mg during a 2-week titration phase in patients not already on a stable dose of PDE5i

Initial registration studies conducted with bosentan [35, 36] and ambrisentan [37] monotherapy evaluated efficacy over the short term, using the 6-min walking distance (6MWD) test as the primary endpoint. Although important in the clinical management of patients with PAH, there have been some suggestions that changes in exercise capacity might have limited prognostic value for clinically relevant outcomes (i.e., death, hospitalization, and clinical worsening), especially in patients receiving combination therapy [38]. In 2008, the Task Force on End Points and Clinical Trial Design recommended the use of a composite endpoint of time to clinical worsening and mortality to provide a more meaningful assessment in PAH studies [39]. Subsequently, pivotal event-driven studies were conducted. In the first of these, SERAPHIN, macitentan significantly reduced the time to the first occurrence of events related to PAH or death when administered alone or on top of background therapy (mostly a PDE5i) [40]. Similarly, in the event-driven AMBITION study, initial combination therapy with ambrisentan and tadalafil resulted in a significantly lower risk of clinical-failure events than ambrisentan or tadalafil monotherapy [41]. However, COMPASS-2 failed to demonstrate an additional benefit of reduced time to first morbidity/mortality event when bosentan was administered in combination with sildenafil [42].

There are currently no head-to-head clinical trials with commercially available ERAs. In their absence, meta-analyses have been conducted in an attempt to compare the efficacy of bosentan, ambrisentan, and macitentan in PAH, although the quality and validity of published evidence synthesis in many analyses is low [43]. More recently (2022), Zhao et al. [44] compared bosentan and ambrisentan monotherapy in a network meta-analysis of five clinical trials (Study 351, BREATHE-1, BREATHE-5, ARIES-1, and ARIES-2). They found no significant difference between bosentan and ambrisentan in terms of 6MWD, Borg Dyspnea Index, and clinical worsening [44]. A recent Bayesian network meta-analysis compared the efficacy of various PAH-targeted therapies, including all three ERAs, as both monotherapy and combination therapy [45]. Among the 50 studies included in the analyses were the event-driven AMBITION, COMPASS-2, and SERAPHIN studies. The authors concluded that ERA + PDE5i combination therapy was the best choice for PAH, with surface under the cumulative ranking curve analysis showing that bosentan in combination with a PDE5i or iloprost performed best on the primary endpoint of 6MWD [45].

Safety and Tolerability

Bosentan, ambrisentan, and macitentan are generally well tolerated over the long term [48, 52, 55, 59, 62, 77–79]. However, as shown in Table 4, each ERA is associated with specific adverse events (AEs) that should be acknowledged and monitored [80].

Peripheral Edema

As well as being a known clinical consequence of PAH, peripheral edema is also recognized as a side effect of ERAs [80]. Edema may arise because of renal or vascular effects of ERAs, including fluid retention due to blocking ET receptor-mediated natriuresis and diuresis in the renal collecting ducts, or changes in precapillary arteriolar vasodilation and capillary permeability [80, 81]. The risk of peripheral edema appears to be higher with ambrisentan than with other ERAs. A network meta-analysis of 10 randomized controlled trials found that ambrisentan significantly increased the risk of peripheral edema versus placebo (31.1% vs 19.0%; risk ratio [RR] 1.62, 95% confidence interval [CI] 1.23, 2.13; P = 0.0001); however, bosentan (15.8% vs 11.9%; RR 1.32, 95% CI 0.87, 2.00; P = 0.19) and macitentan (17.1% vs 18.1%; RR 0.95, 95% CI 0.68, 1.31; P = 0.73) did not [81]. When looking at ambrisentan in combination with a PDE5i (tadalafil), the AMBITION study found that the risk of peripheral edema is more common with ambrisentan + tadalafil (45%) than with ambrisentan (33%) and tadalafil (28%) monotherapy [41]—peripheral edema led to treatment discontinuation in four patients, three patients, and one patient in the respective treatment groups. In contrast, COMPASS-2 showed little difference in the rates of peripheral edema with bosentan + sildenafil (19%) compared with sildenafil monotherapy (16%) [42]. The A DUE study also reported that rates of peripheral edema were similar in patients receiving macitentan + tadalafil (13%) compared with macitentan (11%) or tadalafil (11%) alone; only one patient in the combination therapy group discontinued treatment as a result [73].

Anemia

Decreases in hemoglobin and hematocrit were generally mild and occurred with all ERAs; they occurred particularly within the first few weeks of administration, stabilizing thereafter [80]. These events do not appear to be caused by hemorrhage or hemolysis [5–7]. Preclinical studies in rats suggested that decreases in hemoglobin and hematocrit were larger in the venous versus arterial side, suggesting that plasma volume redistribution may underlie this clinical observation [82]. Monitoring of all ERAs is advised, and in rare cases, transfusion may be required [5–7]. When the findings of 10 randomized controlled trials were examined in a network meta-analysis, macitentan appeared to significantly increase the risk of anemia versus placebo (11.0% vs 3.2%; RR 3.42, 95% CI 1.65, 7.07; P < 0.01), but bosentan (8.3% vs 5.9%; RR 1.39, 95% CI 0.67, 2.86; P = 0.37) and ambrisentan (14.6% vs 9.3%; RR 1.58, 95% CI 0.88, 2.82; P = 0.12) did not [81]. When ERAs are used in combination with PDE5is, there seems to be a higher rate of anemia versus with ERAs alone. For example, in COMPASS-2, 11% of patients receiving bosentan + sildenafil experienced anemia, compared with 7% of patients receiving sildenafil alone [42]. Similarly, in AMBITION, 15% of patients on combination therapy experienced anemia, compared with 6% of patients on ambrisentan monotherapy and 12% on tadalafil monotherapy [41]. In the A DUE study, 19% of patients who received macitentan + tadalafil as a single-tablet combination therapy experienced anemia; this led to withdrawal for one patient who needed a blood transfusion to correct low hemoglobin, but all other cases were mild/moderate in severity [73]. In contrast, 3% of patients on macitentan monotherapy and 2% on tadalafil monotherapy experienced anemia.

Teratogenicity and Fertility

All ERAs can cause embryo-fetal toxicity and are contraindicated during pregnancy [5–7]. Because of the risk of serious birth defects, ERAs had been available only through restricted programs under a Risk Evaluation and Mitigation Strategy (REMS); however, the REMS requirement was lifted in early 2025. To date, access to ambrisentan and macitentan does not require enrollment in a REMS program [6, 7], while bosentan access remains restricted [5]. Despite less restrictive prescribing, women with reproductive potential taking any of the three ERAs must use acceptable methods of contraception before and during treatment, and for 1 month after treatment end. For bosentan, two reliable forms of contraception are required because, unlike ambrisentan and macitentan, it may decrease the efficacy of hormonal contraceptives—these should therefore not be the only contraceptive method used [5–7, 83, 84]. ERAs may also impair fertility in men. Reduced sperm count has been noted in patients taking ERAs, and testicular tubular atrophy, decreased sperm counts (possibly due to effects on spermatogenesis), and/or impaired fertility have also been observed in animal studies [5–7].

Hepatotoxicity

All ERAs undergo hepatic metabolism involving CYP P450 enzymes and must be discontinued in the event of aminotransferase elevations accompanied by signs or symptoms of liver dysfunction/injury or increases in bilirubin ≥ 2 the upper limit of normal [5–7, 23, 26]. Recent evidence from the FDA Adverse Event Reporting System (2004–2022) showed that the reporting odds ratio for the correlation between bosentan and liver injury was 3.38 (95% CI 3.23, 3.53), compared with 0.50 (95% CI 0.47, 0.54) for ambrisentan and 1.00 (95% CI 0.92, 1.09) for macitentan [85]. Bosentan inhibits the bile salt export pump, leading to its accumulation in the liver; this can cause hepatocellular damage as bosentan is a substrate of the hepatic organic anion-transporting polypeptide transport protein [86, 87]. Thus, bosentan should generally be avoided in patients with moderate or severe liver impairment and the drug carries a Black Box Warning for risk of hepatotoxicity [5]. Given this, bosentan is only available through a restricted distribution program with monthly monitoring [5]. As ambrisentan and macitentan do not seem to have the same effect on bile salt transport or drug-induced liver injury as bosentan [34, 87–89], neither of these drugs has this restriction [6, 7, 90].

Elderly Patients

Although PAH predominantly affects women of childbearing age, it is increasingly being diagnosed in the elderly (≥ 65 years old), who show greater clinical deterioration, reduced exercise capacity, poorer prognostic outcomes, lower response to PAH-targeted therapy, and higher rates of PAH drug discontinuation compared with younger patients [91–93]. Elderly patients with PAH often have multiple comorbidities [94] and are generally underrepresented in PAH clinical trials [91]. Indeed, clinical studies with bosentan do not include sufficient numbers of elderly patients to determine whether they respond differently than younger patients [5]. Data for macitentan and ambrisentan in the elderly population are also limited to post hoc subgroup analyses. In ARIES-1 and ARIES-2, 21% of patients were aged ≥ 65 years; among those taking ambrisentan (n = 261), more elderly patients had peripheral edema than those who did not (35% vs 17%) [95]. In SERAPHIN, 14% of the study population were aged ≥ 65 years [7]; the overall incidence and pattern of AEs and serious AEs with macitentan 10 mg observed in these patients (n = 27) was generally similar to that seen among younger patients, i.e., those aged < 35 years (n = 62) or 35–64 years (n = 153) [96].

Efficacy and Safety in PAH Subtypes

Key studies on the efficacy and safety of ERAs in PAH subtypes are shown in Table 5. These subtypes include patients with Eisenmenger syndrome or PAH associated with congenital heart disease (PAH-CHD), portopulmonary hypertension (PoPH), PAH-CTD, and drug-/toxin-induced PAH.

Table 5.

Key studies with bosentan, ambrisentan, and macitentan in PAH subtypes

Study identifier/citation	PAH subtype	Design	Duration	Treatment	Number	Female sex (%)	Age (years)	FC I/II/III/IV (%)	BL 6MWD (m)	Efficacy	Most common AEs (≥ 10% for ERA)	SAEs	Abnormal hepatic function	Hb
Eisenmenger syndrome/PAH-CHD
Pilot study (Ibrahim et al., 2006) [127]	PAH-CHD	OL, MC, PRO	16 W	Bosentan 125 mg BIDa	10	70	32 (11)	0/0/100/0b	304 (79)	Change from BL in 6MWD at 16 W: + 28 m; P = 0.005*	–	20%	No elevated liver enzymes	No reduced Hb
BREATHE-5 (Galiè et al., 2006) [97]	ES	R, DB, PC	16 W	Bosentan 125 mg BIDa	37	62	37 (12)	0/0/100/0b	332 (83)	Primary: Change from BL in SpO2 at rest on room air at 16 W: placebo-corrected effect was 1.0 (95% CI − 0.7, 2.8); non-inferior	Peripheral edema (19% vs 6%), headache (14% vs 12%), palpitations (11% vs 0%)	14% vs 18%	ALT/AST > 5 × ULN: n = 1 on bosentan	–
				Placebo	17	59	44 (9)	0/0/100/0b	366 (68)
MAESTRO (Gatzoulis et al., 2019) [106]	ES	R, DB, PC	16 W	Macitentan 10 mg QD	114	82	33 (12–82)c	0/60/40/0b	369 (75)	Primary: Change from BL in 6MWD at 16 W: + 18 vs + 20 m; P = 0.612	Headache (11% vs 5%) AESIs: anemia-related (5% vs 1%), edema/fluid overload-related (7% vs 5%), hypotension-related (3% vs 3%)	6% vs 2%	ALT/AST > 3 × ULN: 1% vs 1%	Decrease from BL ≥ 2 g/dL: 36% vs 9% Change from BL: − 1.04 vs 0.12 g/dL
				Placebo	112	68	31 (13–62)c	0/59/41/0b	380 (76)
PoPH
Savale et al., 2013 [111]	PoPH	SC, RETRO	5 (2) M FU	Bosentan 125 mg BIDa	34	53	50 (12)	0/12/82/6d	352 (104)	Significant improvement from BL in 6MWD: 403 m (86); P < 0.001*	–	–	–	–
			35 (16) M FU		24	–	–	0/12/80/8d	358 (108)	Significant improvement from BL in 6MWD: 431 m (122); P < 0.05*	–	–	ALT/AST > 3 × ULN: 6%	–
PORTOPULM (Preston et al., 2020) [110]	PoPH	OL, MC, PRO	24 W (+ 24–28 W OLE)	Ambrisentan 5 mg QD (up-titrated to 10 mg QD at W4)	31	58	63 (9)	3/42/52/3b	310 (96)	Co-primary: Change from BL in PVR to 24 W: 7.1 (5) vs 3.8 (1.8) Wood units; P < 0.001* Co-primary: Change from BL in 6MWD to 24 W: 314 (94) vs 336 (108) m	Peripheral edema (38.7%), headache (22.5%)	–	–	No change from baseline
Cartin-Ceba et al., 2011 [113]	PoPH	OBS, PRO	613 (385, 1011) day FUd	Ambrisentan ≤ 10 mg QD	13	46	57 (52, 60)e	III (II, III)b,e	–	Change in mPAP from 58 (37, 63) at BL to 41 (27, 48) mmHg; P = 0.004*d Change in PVR from 445 (329, 834) at BL to 174 (121, 361) dyn·s/cm5; P = 0.008*	–	–	No difference in ALT/AST after 12 M	–
PORTICO	PoPH	R, DB, PC (Krowka et al., 2020) [114]	12 W	Macitentan 10 mg QD	43	49	58 (9)	–	–	Post hoc: At 12 W, significant improvement in risk category for liver transplant perioperative mortality (OR 4.9, 95% CI 1.6, 17.7; P = 0.004*) and decreased number of patients in high-risk category for wait-list mortality (OR 10.5, 95% CI 2.4, 66.8; P = 0.001*)	Peripheral edema (26% vs 12%), headache (16% vs 17%)	21% vs 14%	–	–
				Placebo	42	48	59 (10)	–	–
		R, DB, PC + OLE (Sitbon et al., 2019) [112]	12 W (DB) + 12 W OLE	Macitentan 10 mg QD	43	49	58 (9)	2/63/35/0b	386 (100)	Primary: PVR (expressed as ratio of BL) at 12 W: 0.63 vs 0.98; P < 0.0001*	Peripheral edema (27%), headache (19%), anemia (14%), bronchitis (10%)	30%	ALT/AST > 3 × ULN + bilirubin > 2 × ULN and increased from BL: n = 1	≤ 8 g/dL: 5% Decrease from BL ≥ 2 g/dL: 37% Change from BL: − 1.77 vs − 0.04 g/dL
				Placebo	42	48	59 (10)	2/55/43/0b	383 (109)
OPUS/ OrPHeUS (Kim et al., 2024) [115]	PoPH	PRO, OBS (OPUS; 2014–2020) MC, RETRO, chart review (OrPHeUS; 2013–2017)	11.9 (3.1, 26.0) Mf	Macitentan mono (25%), double therapy (66%), and triple therapy (9%)	206	52	58 (51, 64)e	7/37/52/4b (n = 102)	317 (240, 400);f n = 65	Free from hospitalization (KM estimate): 49% (95% CI 41, 56) at 1 Y Survival (KM estimate): 82% (95% CI 75, 87) at 1 Y	OPUS only: peripheral edema (42%)	OPUS only: 71%	ALT/AST > 3 × ULN: 12%	OPUS only: ≥ 1 event of Hb decrease: 15%
PAH-CTD
Study 351/ BREATHE-1 subgroup analysis (Denton et al., 2006) [117]	PAH-CTD: SSc (79%), SLE (12%)	R, DB, PC	12–16 W	Bosentan 125 or 250 mg BIDa	44	86	58 (13)	0/0/95/5b	312 (73)	Primary: Change from BL in 6MWD at 12–16 W: + 19.5 m (95% CI − 3.2, 42.2) vs − 2.6 m (95% CI − 32, 76); P = NS Survival: 86% at 1 Y, 73% at 2 Y	–	–	–	–
				Placebo	22	77	50 (13)	0/0/95/5b	361 (67)
Denton et al., 2008 [118]	PAH-CTD: SSc (75%), mixed (11%), SLE (9%)	OL, MC, PRO	48 W	Bosentan 125 mg BIDa	53	83	63 (13)	0/0/100/0b	–	Change in WHO FC from BL at 48 W: 27% improved, 16% worsened Free from clinical worsening at 48 W: 68% (95% CI 55, 82) Survival at 48 W: 92% (95% CI 85, 100)	Peripheral edema (17%), liver enzyme elevations (17%), diarrhea (13%), exacerbated dyspnea (13%), nausea (13%)	45% Most common: exacerbated dyspnea (8%) and pneumonia (8%); all unrelated to bosentan	ALT/AST increases (11%)	–
Combination real-world study (Dardi et al., 2015) [67] Sequential combination	PAH-CTD (suboptimal response to prior bosentan or sildenafil)	SC, RETRO, chart review (2003–2013)	–	Bosentan 125 mg BIDa + sildenafil 20 mg TID	29	83	66 (58, 72)e	28 (I/II), 72 (III/IV)c	360 (219, 425)f	Significant improvements in 6MWD, mPAP, cardiac index, and PVR at 3–4 M; all P < 0.05*	–	–	–	–
ARIES-1/2/E subgroup analysis (Fischer et al., 2016) [119]	PAH-CTD: SSc (61%), mixed (20%), SLE 15%	R, DB, PC + OLE	3 Y	Ambrisentan 2.5–10 mg QD or placebo	124	89	56 (13)	2/46/45/7b	333 (88)	Change from BL in 6MWD: + 9.9 m at 1 Y; + 3.5 m at 2 Y; − 5.1 m at 3 Y Increases from BL in 6MWD in 62.6%, 57.3%, and 58.2% of patients through 1, 2, and 3 Y, respectively	Most common AEs (≥ 20%): peripheral edema (44%), anemia (23%), pulmonary hypertension (21%), arthralgia (20%)	–	–	–
AMBITION subgroup analysis	PAH-CTD: SSc (63%), mixed (12%), SLE (9%) (Coghlan et al., 2017) [122]	DB, R, PC (subgroup analysis)	Event-driven	Ambrisentan 10 mg QD + tadalafil 40 mg QD (combo)	103	86	58 (12)	0/25/75/0b	325 (86)	Primary: Risk of event related to PAH or death: combo (19%) vs pooled mono (36%); HR 0.43, 95% CI 0.24, 0.77 Risk of hospitalization for worsening PAH: 71% risk reduction for combo vs pooled mono (HR 0.29, 95% CI 0.12, 0.67)	Combo vs ambrisentan mono and tadalafil mono (AEs ≥ 20% of combo): peripheral edema (47% vs 34% and 33%), headache (33% vs 32% and 38%), diarrhea (29% vs 32% and 25%), dyspnea (22% vs 20% and 25%)	Combo vs ambrisentan mono and tadalafil mono: 44% vs 34% and 50%	–	–
				Pooled mono (ambrisentan 10 mg QD plus tadalafil 40 mg QD)	84	90	58 (13)	0/26/74/0b	328 (98)
	PAH-CTD: SSc (64%), mixed (13%), SLE (9%) (Kuwana et al., 2020) [121]	DB, R, PC (post hoc analysis in mITT population)	Event-driven	Ambrisentan 10 mg QD + tadalafil 40 mg QD (combo)	117	86	59 (12)	0/28/72/0b	326 (89)	Risk of clinical failure was 51.7% lower for combo vs mono (HR 0.48, 95% CI 0.29, 0.82)	Combo vs ambrisentan mono and tadalafil mono (AEs ≥ 20% of combo): peripheral edema (48% vs 38% and 30%), headache (34% vs 35% and 38%), diarrhea (29% vs 33% and 26%), dizziness (18% vs 29% and 26%), dyspnea (21% vs 23% and 21%)	Combo vs ambrisentan mono and tadalafil mono: 45% vs 38% and 51%	–	–
				Pooled mono (ambrisentan 10 mg QD plus tadalafil 40 mg QD)	99	89	58 (13)	0/30/70/0b	330 (96)
ATPAHSS-O	SSc-PAH	OL, MC, PRO (Hassoun et al., 2015) [123]	36 W	Ambrisentan 10 mg + tadalafil 40 mg	24	92	60 (11)	0/35/65/0b	343 (131)	Co-primary: Change from BL in PVR at 36 W: − 55%; P < 0.01* Co-primary: Change from BL in RV mass: − 14%; P < 0.05*	Edema (29%), headache (29%), nasal congestion (16%), dyspnea (12%), abdominal pain (12%)	5 SAEs reported	–	–
		Cardiac function substudy (Sato et al., 2018) [124]	36 W	Ambrisentan 10 mg + tadalafil 40 mg (upfront)	21	91	63 (53–68)c	0/38/62/0d	390 (260–438)e	Significant improvement in both RV and LV systolic and diastolic function	–	–	–	–
Li et al., 2020 [120]	PAH-CTD	OL, PRO (Chinese post hoc analysis)	24 W	Ambrisentan 5–10 mg QD	71	97	40 (19, 59)e	0/45/55/0b	366 (60)	Change from BL in 6MWD at 24 W: + 73 (85) m; P < 0.001*	Flushing (13%), peripheral edema (10%)	9%	ALT/AST increased (6%/4%)
OPUS / OrPHeUS (Channick et al., 2024) [105]	PAH-CTD: SSc (59%), SLE (13%), mixed (10%)	PRO, OBS (OPUS; 2014–2020) MC, RETRO, chart review (OrPHeUS; 2013–2017)	Median 15.8 M	Macitentan mono (35%) and combo (65%)	1192	86	62 (52, 70)e	8/23/63/6b (n = 654)	279 (184, 362);e n = 454	Free from hospitalization (KM estimate): 59% (95% CI 56, 62) at 1 Y Survival (KM estimate): 91% (95% CI 89, 92) at 1 Y	OPUS only: dyspnea (25%), headache (13%), peripheral edema (12%), dizziness (11%), pneumonia (11%), fatigue (11%) AESI: edema (30%), anemia/Hb decrease (13%)	–	ALT/AST > 3 × ULN: 3%	–
Song et al., 2024 [125]	PAH-CTD: SLE (50%), mixed (19%), Sjogren’s syndrome (13%), SSc (13%)	SC (Chinese), RETRO, chart review (2020–2021)	24 W	Macitentan 10 mg QD (28%)	9	67	46 (17)	0/22/67/11b	280 (185, 363)e	Change from BL in 6MWD at 24 W: − 120 vs − 70 m; P = 0.09* Both mono and combo groups improved 6MWD significantly vs BL	–	–	No significant differences from BL in ALT/AST at 24 W	No significant differences from BL in Hb at 24 W
				Macitentan 10 mg QD + sildenafil 25 mg TID (72%)	23	87	42 (13)	0/9/70/22b	290 (180, 328)e
Drug-/toxin-induced PAH
OPUS / OrPHeUS (Ravichandran et al., 2020) [126]	Drug-/toxin-induced PAH	PRO, OBS (OPUS; 2014–2020) MC, RETRO, chart review (OrPHeUS; 2013–2017)	NR	Macitentan (dose NR)	207	72	51 (45, 59)e	–	363 (288, 423)e n = 70	Survival (KM estimate): 93% (95% CI 88, 96) at 1 Y	–	–	ALT/AST > 3 × ULN: 2.4%	–

Data are presented as % or mean (SD), unless otherwise stated

6MWD 6-min walking distance, AE adverse event, AESI adverse event of special interest, ALT alanine aminotransferase, AST aspartate aminotransferase, BID twice daily, BL baseline, CI confidence interval, combo combination therapy, DB double-blind, ERA endothelin receptor antagonist, ES Eisenmenger syndrome, FC functional class, FU follow-up, Hb hemoglobin, HR hazard ratio, KM Kaplan–Meier, LTE long-term extension, LV left ventricular, M months, MC multicenter, mITT modified intention-to-treat, mono monotherapy, mPAP mean pulmonary arterial pressure, NR not reported, NS non-significant, NYHA New York Heart Association, OBS observational, OL open label, OLE open-label extension, OR odds ratio, PAH pulmonary arterial hypertension, PAH-CHD pulmonary arterial hypertension associated with congenital heart disease, PAH-CTD pulmonary arterial hypertension associated with connective tissue disease, PC placebo controlled, PoPH portopulmonary hypertension, PRO prospective, PVR pulmonary vascular resistance, Q quartile, QD once daily, R randomized, RETRO retrospective, RV right ventricular, SAE serious adverse event, SC single-center, SD standard deviation, SLE systemic lupus erythematosus, SpO₂ systemic arterial blood oxygen saturation, SSc systemic scleroderma, TID three times daily, ULN upper limit of normal, W weeks, WHO World Health Organization, Y years

*Significant P value

^aPatients received bosentan 62.5 mg BID for the initial 4 weeks

^bWHO FC

^cMedian (range)

^dNYHA FC

^eMedian (Q1, Q3)

Eisenmenger Syndrome/PAH-CHD

Eisenmenger syndrome is the most severe form of PAH-CHD. Current treatment guidelines recommend bosentan in symptomatic patients with Eisenmenger syndrome to improve exercise capacity [3]. This recommendation is based partly on results of the BREATHE-5 study, which showed that bosentan was well tolerated and improved exercise capacity and hemodynamics in patients with Eisenmenger syndrome [97]; the long-term efficacy and safety of bosentan in Eisenmenger syndrome and PAH-CHD have since been reported in multiple studies [98–100] and meta-analyses [77, 101–103]. In contrast, a small (n = 17) retrospective chart review of patients with Eisenmenger syndrome showed that ambrisentan increased exercise capacity over the short term, but the increase in exercise capacity returned to baseline during longer-term follow-up (mean 2.5 years) [104]. While a combined analysis of the OPUS/OrPHeUS dataset observed a safety profile for macitentan in PAH-CHD (n = 272) similar to that observed in patients with PAH (n = 2498) with 1- and 2-year survival rates of 94% and 90%, respectively [105], macitentan failed to show a significant effect in patients with Eisenmenger syndrome compared with placebo in the randomized MAESTRO study [106].

Treatment guidelines state that if patients with Eisenmenger syndrome or PAH-CHD do not meet their treatment goals, combination therapy should be considered [3]; however, this recommendation is based on limited evidence (Table 5). Results with the combination of bosentan and sildenafil—a PDE5i—have been inconsistent [67, 107]. Further, while a small, randomized, 12-week, placebo-controlled study in 27 patients (five with idiopathic PAH and 22 with Eisenmenger syndrome) found significant improvements from baseline in New York Heart Association functional class (NYHA FC) and mean pulmonary arterial pressure with upfront combination of ambrisentan and sildenafil along with trends toward improvement in various echocardiographic parameters, there were no significant differences in 6MWD or NYHA FC (the primary endpoints) or any secondary outcomes of RV function when compared with sildenafil monotherapy [108].

PoPH

Patients with PoPH have hypertension in both the pulmonary arteries and the portal vein, which carries blood to the liver [109]. Current treatment guidelines recommend monotherapy with an approved PAH treatment, followed by sequential combination therapy, if needed [3]. In these patients, it is also important to account for any underlying liver disease and the possibility of liver transplantation. Thus, these factors should be considered by clinicians and patients when making treatment decisions on the use of ERAs.

Bosentan, ambrisentan, and macitentan have all been studied in patients with PoPH, with each showing improved outcomes (Table 5) [110–113]. However, macitentan is the only ERA to have been evaluated in a randomized, placebo-controlled clinical trial. In the PORTICO study, macitentan significantly improved pulmonary vascular resistance in patients with PoPH, compared with placebo [112]; and a post hoc analysis found that macitentan improved risk categorization for liver transplant mortality [114]. However, PORTICO did not show any functional improvements with macitentan treatment. In terms of safety, macitentan was not associated with any hepatic concerns. A combined analysis of the OPUS/OrPHeUS dataset also observed no unexpected safety findings among patients with PoPH in those who received macitentan (largely as combination therapy), which is important given the underlying liver disease in these patients [115].

PAH-CTD

PAH is a severe vascular complication of CTD, and PAH-CTD is associated with a higher risk of death compared with other types of PAH [48, 116]. The 2022 European Society of Cardiology (ESC)/European Respiratory Society (ERS) pulmonary hypertension treatment guidelines recommend that for PAH-CTD, guidelines for the underlying condition should be adhered to—such as those for systemic sclerosis or systemic lupus erythematosus [3].

Results have been mixed for bosentan in patients with PAH-CTD (Table 5). In a subgroup analysis of 66 patients with PAH-CTD from the Study 351 and BREATHE-1 trials, exercise capacity (6MWD) remained stable over 12–16 weeks with bosentan monotherapy versus a decline with placebo; however, in contrast to the overall study population, the difference between treatment groups in patients with PAH-CTD was not significant [117]. In a 48-week, open-label, prospective study, bosentan monotherapy improved or stabilized clinical status in many patients with PAH-CTD and was associated with a 92% survival rate [118]. In a real-word study, favorable short-term effects on exercise capacity and hemodynamics were observed in patients with PAH-CTD receiving both bosentan and sildenafil as sequential combination therapy; however, survival estimates were significantly poorer in these patients than in those with other PAH subtypes [67]. A meta-analysis of 15 studies of long-term (≥ 1 year) bosentan administration detected a significant difference in mean pulmonary arterial pressure from baseline among patients with PAH-CTD (standardized mean difference [SMD] − 0.86, 95% CI − 1.28, − 0.44; P < 0.0001), but failed to detect a significant change in 6MWD (SMD 0.18, 95% CI − 0.60, 0.95; P = 0.656) [77].

The available evidence for ambrisentan in PAH-CTD is more consistent than that for bosentan. A subgroup analysis of 124 patients with PAH-CTD from ARIES-1, ARIES-2, and ARIES-E demonstrated increases in 6MWD among 62.6%, 57.3%, and 58.2% of patients treated with ambrisentan at 1, 2, and 3 years, respectively; the 3-year survival rate was 76%, and 64% of patients were free from clinical worsening at 3 years [119]. In a post hoc analysis of a phase 3 study in China (n = 71), ambrisentan also showed significant improvements in 6MWD over 24 weeks and no clinical worsening in the majority of patients (98.6%) with PAH-CTD [120]. In a subgroup analysis of the AMBITION study and the ATPAHSS-O trial, upfront combination therapy with ambrisentan and tadalafil was associated with clinical benefits in patients with PAH-CTD (Table 5) [121–124].

For macitentan, real-world evidence from the combined OPUS/OrPHeUS dataset showed that the efficacy, safety, and tolerability of macitentan are similar in both patients with PAH-CTD and those with other PAH subtypes [105]; and in a retrospective study of 32 patients in China, macitentan significantly improved 6MWD from baseline in patients with PAH-CTD, whether it was administered as monotherapy or in combination with sildenafil [125].

Drug-/Toxin-Induced PAH

Some medicines and illicit drugs/toxins, such as methamphetamine, fenfluramine, and dasatinib, can cause PAH [3]. The 2022 ESC/ERS treatment guidelines recommend discontinuing the use of the causative agent and to consider treatment in patients with intermediate-/high-risk PAH, or in low-risk patients if the hemodynamics have not normalized after discontinuation of the causative agent [3]. There are limited data on the efficacy and safety of ERAs in patients with drug-/toxin-induced PAH, whose numbers are small in clinical trials. Registry data provide greater numbers of these patients, and real-world evidence from the OPUS/OrPHeUS dataset showed that the efficacy and safety of macitentan were similar in patients with drug-/toxin-induced PAH and in those with idiopathic/heritable PAH [126].

Transitioning Between ERAs in PAH

Occasionally, patients with PAH may need to transition between ERAs as a result of clinical worsening/inadequate response, if insurance allows, because of AEs or drug interactions. Multiple studies have shown that transitioning from one ERA to another can be both effective and well tolerated (Table 6).

Table 6.

Transitioning from one ERA to another in PAH

Study identifier/citation	PAH subtype	Design	Duration	Treatment	Number	Female sex (%)	Age (years)	FC I/II/III/IV (%)	BL 6MWD (m)	Efficacy	Safety
Bosentan to ambrisentan
Chen et al., 2021 [128]	Adults with PAH	SC, PRO, cohort (Chinese)	6 M	Prior treatment: bosentan 125 mg BID (stable dose for ≥ 6 M) Switched to: ambrisentan 5–10 mg QD	46	72	34 (10)	0/57/43/0a	416 (50)	6MWD: 427 (48) m at 6 M; P = NS No significant differences in WHO FC, echocardiography characteristics, or quality of life after transition, although patients found ambrisentan more convenient; P < 0.001*	Most common AEs after transition: dizziness (44%), headache (37%), and palpitation (33%); mostly mild to moderate
McGoon et al., 2009 [130]	Adults with PAH	OL, MC	12 W	Prior treatment: bosentan (prior median duration: 13.9 W) and/or sitaxsentan (prior median duration of exposure: 28.7 W) Switched to: ambrisentan 2.5–10 mg QD	36b	86	57 (13)	0/36/64/0a	397 (105)	Significant improvements in 6MWD and other efficacy assessments were observed	One patient had a transient ALT/ALT > 3 × ULN that resolved following a temporary dose reduction and did not require discontinuation
Dawson et al., 2018 [129]	Adults with PAH	SC, RETRO, OBS (2013–2016)	20 (12, 26)c M FU	Prior treatment: bosentan Switched to: ambrisentan	43	81	64 (15)	0/5/90/5a	276 (148)	6MWD and WHO FC were stable/improved in 81% and 95% of patients, respectively Hemodynamic parameters were comparable pre- and post-transition Six patients were changed to macitentan and one back to bosentan due to inadequate response	Most common AEs: ankle edema (n = 12) and headache (n = 4) One patient experienced diarrhea and was changed to macitentan One patient transitioned to ambrisentan was changed back to bosentan due to intolerance Hb levels remained stable and ALT/AST remained within the normal range
Bosentan to macitentan
Safdar et al., 2017 [134]	Adults with PAH	RETRO, SC (2013–2015)	5.7 (1.5) M FU	Prior treatment: bosentan 125 mg BID (duration NR) Switched to: macitentan 10 mg QD (within 12–24 h of previous dose)	24	88	58 (13)	0/42/58/0a	344 (106)	6MWD post-switch: 319 (85) m; P = 0.18 Brain natriuretic peptide levels were 91 (170) pg/mL pre-switch and 90 (137) pg/mL post-switch; P = 0.93	ALT/AST remained unchanged post-switch. Two patients did not tolerate the switch to macitentan and had to be returned to bosentan: one patient with PoPH developed elevated ALT/AST and the second patient's macitentan was stopped because of malaise and tachyarrhythmia. One patient who underwent a successful liver transplant had macitentan stopped following the transplant
Aypar et al., 2020 [132]	Children and young adults with PAH	SC, OL, PRO	24 M	Prior treatment: bosentan 125 mg BID (duration not reported) Switched to: macitentan 10 mg QD	27	63	21 (6)	0/93/7/0a	458 (79)	6MWD: + 43 m at 6 M (P < 0.05*), + 56 m at 12 M (P < 0.05*), and + 74 m at 24 M (P < 0.05*) Macitentan did not significantly change functional class, oxygen saturation, and natriuretic levels	None of the patients had anemia, hepatotoxicity, or peripheral edema
Dawson et al., 2018 [129]	Adults with PAH	SC, RETRO, OBS (2013–2016)	20 (12, 26)c M FU	Prior treatment: bosentan Switched to: macitentan	49	73	58 (14)	2/21/70/6a	314 (151)	6MWD and WHO FC were stable/improved in 84% and 91% of patients, respectively Hemodynamic parameters were comparable pre- and post-transition	Most common AEs: ankle edema (n = 7) and headache (n = 4) Hb levels remained stable and ALT/AST remained within the normal range
POTENT (Aldalaan et al., 2022) [131]	Adults with PAH	OL, PRO, single-arm, registry (2014–2018)	12 M	Prior treatment: bosentan 125 mg BID (for ≥ 3 M) Switched to: macitentan 10 mg QD	50	78	35 (11)	8/44/48/0a	331 (130)	Change from BL in 6MWD: + 41 (13) m at 12 M; P = 0.002* Patients were more likely to be in WHO FC I/II (78%) and 22% of the overall cohort moved to a lower risk condition Most hemodynamic parameters decreased over time	There were no incidences of anemia or liver injury (GGT decreased significantly from BL to 12 M) Clinical worsening was observed only in a small group of patients
Verlinden et al., 2020 [133]	Adults with PAH	RETRO, SC (2010–2019)	6.5 (3–13)d M FU	Prior treatment: bosentan 125 mg BID + sildenafil 20–80 mg TID (median 7 Y) Switched to: macitentan + sildenafil (8); macitentan + tadalafil (1); bosentan + tadalafil (2); ambrisentan + sildenafil (2); ambrisentan + tadalafil (1); ambrisentan + riociguat (1); bosentan + riociguat (1)	16	88	57 (15)	35/19/56/0a	369 (79)	Change from BL in 6MWD: + 8 (− 50 to 70)b m at FU (n = 11); P = 0.39 6MWD and WHO FC stable or significantly improved in 82% and 100% of patients, respectively, after transition No significant changes after transition in hemodynamics, NT-proBNP values, or REVEAL risk scores	94% of patients tolerated the transition with minor AEs occurring in 25% of patients
Ambrisentan to macitentan
Chen et al., 2022 [135]	Adults with PAH	OL, PRO, cohort (Chinese)	12 M	Prior treatment: ambrisentan 5–10 mg QD (for ≥ 6 M) Switched to: macitentan 10 mg QD	145	75	32 (26, 42)c	0/59/41/0a	425 (82)	6MWD: 441 (78) m at 12 M; P = 0.004* 6MWD was stable/improved in 94% at 12 M Significant improvements in cardiac function, hemodynamics, and quality of life versus BL at 12 M	There was no statistical difference in the overall incidence of AEs before and after the transition After transition, incidence of peripheral edema and headache was not significantly changed, but anemia increased significantly from 7% at BL to 15% at 12 M; P = 0.024* Four female patients experienced severe anemia, (decrease in Hb to 73–83 g/L) with associated menstrual disorder and discontinued macitentan
Bosentan or ambrisentan to macitentan
Cadenas-Menéndez et al., 2020 [136]	Adults with PAH	RETRO, MC (Dates NR)	12 M	Prior treatment: bosentan or ambrisentan (doses and treatment duration NR) Switched to: macitentan 10 mg QD	12	83	66 (13)	8/50/42/0a	345 (116)	After 3 M, FC had improved in 4 patients, 6MWD increased in 8 patients, and NT-proBNP levels and right atrial area were lowered in 7 and 8 patients, respectively. Similar results were observed after 6 and 12 M	Macitentan was well tolerated, with no PAH hospitalizations, septostomies, transplants, or deaths registered
Tynan et al., 2019 [137]	Adults with PAH	RETRO, SC (2014–2016)	18 M	Prior treatment: bosentan or ambrisentan Switched to: macitentan	14	86	63 (13)	0/29/64/7a	371 (28)	Change from BL in FC: significant improvement in switched (P = 0.004*) and added (P = 0.03*) groups, but not control (P = NS) group Change from BL in 6MWD at 6 M: + 37 m (P = NS) for switched group and + 98 m (P = 0.01*) for added group; P = NS for control group	Macitentan was well tolerated. Overall, 92% of patients continued macitentan throughout the study. One patient in the switched group stopped macitentan due to anemia; another patient in the switched group stopped all therapy, received palliative therapy, and died shortly thereafter
				Added: initiated macitentan or added macitentan to prior therapy with sildenafil	11	73	65 (5)	0/0/82/18a	334 (28)
				Control: remained on prior therapy with bosentan or ambrisentan ± sildenafil	12	83	74 (9)	0/0/50/50a	484 (28)
Ambrisentan to bosentan
Gong et al., 2018 [139]	Adults with PAH	SC, OL, PRO (Chinese)	24 M	Prior treatment: ambrisentan 5 mg QD (for 12 M) Switched to: bosentan 125 mg BIDe	8	75	51 (12)	–	433 (79)	Change from BL in 6MWD: + 10.3 (14.5) for switched group vs − 5.6 (38.9) m for control group at 24 M; P = NS No significant differences between groups in change from BL in hemodynamic or hemolytic parameters, or WHO FC	No significant differences between switched and control groups in change from BL in ALT/AST at 12 or 24 M
				Control: remained on ambrisentan	12	83	39 (17)	–	512 (41)

Data are presented as % or mean (SD), unless otherwise stated

6MWD 6-min walking distance, AE adverse event, ALT alanine aminotransferase, AST aspartate aminotransferase, BID twice daily, BL baseline, ERA endothelin receptor antagonist, FC functional class, FU follow-up, GGT gamma-glutamyl transferase, Hb hemoglobin, LFT liver function test, M months, MC multicenter, NR not reported, NS non-significant, NT-proBNP N-terminal pro-brain natriuretic peptide, OBS observational, OL open label, PAH pulmonary arterial hypertension, PoPH portopulmonary hypertension, PRO prospective, Q quartile, QD once daily, RETRO retrospective, REVEAL Registry to Evaluate Early and Long-Term PAH Disease Management, SC single-center, SD standard deviation, TID three times daily, ULN upper limit of normal, W weeks, WHO World Health Organization, Y years

^aWHO FC

^bPatients previously discontinued bosentan (n = 31), sitaxsentan (n = 2), or both (n = 3)

^cMedian (Q1, Q3)

^dMedian (range)

^ePatients received bosentan 62.5 mg BID for initial 4 weeks

Bosentan to Ambrisentan

Several studies have shown that switching from bosentan to ambrisentan is well tolerated, with no influence on hematologic parameters or heart function (Table 6) [128–130]. Few patients fail transition. Among 49 patients with PAH who transitioned from bosentan to ambrisentan at a single center in China, only three failed transition and reverted to bosentan: two because of side effects/symptoms and one as a result of clinical deterioration [128]. Similarly, among 43 patients who transitioned from bosentan to ambrisentan as part of a nurse-led, home-based program in the UK, only two patients reverted to bosentan: one because of intolerance, and one as a result of inadequate response [129]. However, an additional seven patients were transitioned from ambrisentan to macitentan: one because of diarrhea and six as a result of inadequate response.

Bosentan to Macitentan

Switching from bosentan to macitentan has been shown to be effective and well tolerated (Table 6) [129, 131, 134]. The most recent of these studies (POTENT) was a prospective, open-label study of 50 patients in the SAUDIPH registry whose treatment was changed from bosentan to macitentan [131]. One year after transition, mean 6MWD had increased significantly (by approximately 34 m from baseline to month 12; P = 0.0006) and WHO FC had shifted from III/IV to II/III in 26% of patients. In a UK study in 49 patients with PAH who transitioned from bosentan to macitentan, 6MWD and WHO FC remained stable or improved in 84% and 91% of patients, respectively [129]. Liver function test results and hemoglobin levels were comparable before and after transition.

Ambrisentan to Macitentan

A large prospective study conducted in 157 patients with PAH at a single center in China provides evidence of the safety and efficacy of switching from ambrisentan to macitentan [135]. Most patients had stable or improved 6MWD (94%) and WHO FC (92%) 1 year post-switch and AEs were largely similar before and after transition, with the exception of anemia and menstrual disorders. In total, 12 patients discontinued macitentan; of these, five patients transitioned back to ambrisentan (two for financial reasons; three as a result of psychological dependence), four suspended macitentan because of severe anemia with associated menstrual disorder (macitentan was reintroduced successfully for three patients following iron therapy), two required treatment change due to disease progression, and one died (unrelated to ERA treatment). Those patients completing 12 months of macitentan treatment indicated a significant improvement in overall satisfaction and improvement in activities of daily living compared with ambrisentan.

Bosentan or Ambrisentan to Macitentan

Two studies have examined the effects of transitioning from bosentan or ambrisentan to macitentan in PAH (Table 6) [136, 137]. Both studies were included in a meta-analysis of nine studies (encompassing 408 patients with PAH), which examined the effect of switching from bosentan or ambrisentan to macitentan [138]. In the meta-analysis, conversion to macitentan was associated with a significant increase in 6MWD of > 20 m (across six studies that assessed exercise capacity; P = 0.00001) and a significant improvement in WHO FC of 0.4 (across three studies reporting change in WHO FC; P = 0.028) [138].

Ambrisentan to Bosentan

Only one small study has specifically examined the effect of switching to bosentan from another ERA. This was a prospective study at a single center in China in which eight of 20 patients whose PAH was stable on ambrisentan were switched to bosentan after 1 year [139]. No differences were noted between either group in 6MWD, WHO FC, or hemodynamic or hematologic parameters up to 2 years post-switch, and no patients discontinued treatment as a result of liver function abnormalities.

Future Directions

As of 2025, there are additional new studies actively evaluating the role of ERAs for the treatment of PAH. Some studies are examining the efficacy and safety of ERAs as initial therapy in PAH either as monotherapy or in combination with other therapies, such as PDE5is or prostacyclin analogues (NCT06317805 [TripleTRE], NCT01347216 [COMPERA/COMPERA-KIDS], NCT04039464). A randomized, placebo-controlled clinical trial in China is evaluating treatment with ambrisentan among patients with mild PAH (defined as mean pulmonary arterial pressure 21–24 mmHg, and pulmonary artery wedge pressure < 15 mmHg; NCT04972656). In addition, UNISUS, a phase 3, multicenter, randomized clinical trial, has been designed to compare the efficacy, safety, and tolerability of macitentan 10 mg (the dose currently approved for the treatment of adults with PAH [7]) versus macitentan 75 mg (NCT04273945). Positive outcomes from these studies have the potential to provide healthcare providers with new treatment options and may serve the needs of additional populations, including children with PAH and patients with mild PAH.

Conclusions

PAH pathogenesis is mediated by ET-1 activation of both the ET_A and ET_B receptors; therefore, dual receptor antagonism has been recognized as an important treatment modality in PAH [12]. Because there are no head-to-head clinical trials that compare the efficacy and safety of ERAs in PAH, it is important to understand the similarities and the distinguishing characteristics of each, so that clinicians and patients can make informed, shared decisions for individualized treatment selection. Bosentan, ambrisentan, and macitentan have all proven to be efficacious in treating PAH; however, it is important to consider other factors such as unique safety concerns, continued laboratory monitoring, and dosing frequency when discussing which ERA is best suited for a particular patient. In addition, it is also important to consider when to initiate ERA treatment and whether it should be administered as monotherapy or in combination with other drugs approved for the treatment of PAH.

Author Contributions

Ankita Adhia, Paul Guichard, and Marinella Sandros contributed to the conceptualization. Ankita Adhia contributed to the literature search. All authors (including Naomi G. Habib and David Lopez) commented on previous versions and approved the final manuscript.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: sponsorship for this article as well as all publication charges were funded by Johnson & Johnson, Titusville, NJ, USA. Medical writing and editorial support were funded by Johnson & Johnson, Titusville, NJ, USA.

Data Availability

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.

Declarations

Conflict of Interest

Naomi G. Habib is a consultant and has received speaker fees from Johnson & Johnson, Liquidia, Merck, Pulmovant, and United Therapeutics. Ankita Adhia, David Lopez, and Marinella Sandros are employees of Johnson & Johnson; Paul Guichard received speaker fees from Bayer, Merck, Johnson & Johnson and United Therapeutics.

Ethical Approval

Ethics approval was not required—this narrative review was based on previously published studies and no new studies involving human or animal participants were performed.