Phosphodiesterase type 5 inhibitor plus endothelin receptor antagonist compared to either alone for group 1 pulmonary arterial hypertension

Yuji Oba; Tinashe Maduke; Eddie W Fakhouri; Yohannes Goite

doi:10.1002/14651858.CD015824.pub2

. 2025 Aug 4;2025(8):CD015824. doi: 10.1002/14651858.CD015824.pub2

Phosphodiesterase type 5 inhibitor plus endothelin receptor antagonist compared to either alone for group 1 pulmonary arterial hypertension

Yuji Oba ^1,^✉, Tinashe Maduke ¹, Eddie W Fakhouri ¹, Yohannes Goite ²

Editor: Cochrane Central Editorial Service

PMCID: PMC12320213 PMID: 40757552

Abstract

Rationale

Pulmonary arterial hypertension (PAH), a rare disorder, causes elevated pressure in the pulmonary arteries, leading to heart failure. Untreated PAH has a poor prognosis, emphasising the need for effective intervention. Pharmacological treatment includes pulmonary vasodilators such as endothelin receptor antagonists (ERA), prostacyclin analogues, phosphodiesterase type 5 inhibitors (PDE5i), and soluble guanylate cyclase stimulators, often used together to improve symptoms and quality of life while reducing mortality and risk of hospitalisation.

Objectives

To assess the benefits and harms of combination therapy involving a phosphodiesterase type 5 inhibitor (PDE5i) and an endothelin receptor antagonist (ERA) in adults and adolescents with group 1 pulmonary arterial hypertension (PAH) compared to either agent alone.

Search methods

We searched Cochrane Central Register of Controlled Trials, MEDLINE, Scopus, ClinicalTrials.gov, and the World Health Organization (WHO) International Clinical Trials Registry Platform for randomised controlled trials (RCTs). The most recent searches were conducted on 13 March 2024.

Eligibility criteria

We included published and unpublished RCTs comparing combinations of ERAs and PDE5is versus either agent alone lasting at least 12 weeks. Participants were aged 12 years or older with WHO group 1 PAH meeting specific haemodynamic criteria.

We excluded cluster‐, cross‐over, and quasi‐RCTs, and other PAH‐specific medications.

Outcomes

Critical outcomes were clinical worsening, mortality, and hospitalisation. Important outcomes were changes in six‐minute walk distance (6MWD), WHO functional class, Borg Dyspnea Scale, serious adverse events, and withdrawal from the trial.

Risk of bias

Two review authors independently assessed risk of bias using the Cochrane RoB 2 tool. We resolved disagreements through discussion or consultation. This informed GRADE ratings and summary of findings tables.

Synthesis methods

We used a random‐effects model to address study differences, switching to a fixed‐effect model if there was variation primarily due to random error. We conducted meta‐analyses if deemed meaningful, with data pooled if treatments, participants, and clinical questions were sufficiently similar.

Included studies

We included nine studies with 1807 participants. The median duration of the studies was 16 weeks, ranging from 12 to 129 weeks. Treatment regimens included combinations of medications such as ambrisentan, bosentan, macitentan, tadalafil, and sildenafil.

Synthesis of results

Combination therapy versus endothelin receptor antagonist

Combination therapy reduces clinical worsening compared to ERA alone (risk ratio (RR) 0.53, 95% confidence interval (CI) 0.41 to 0.68; 113 fewer per 1000 participants, 95% CI 141 fewer to 77 fewer; number needed to treat for an additional beneficial effect (NNTB) 9, 95% CI 7 to 13; 5 trials, 1139 participants; high‐certainty evidence). Hospitalisation is likely reduced (RR 0.32, 95% CI 0.19 to 0.55; 70 fewer per 1000 participants, 95% CI 83 fewer to 46 fewer; NNTB 14, 95% CI 12 to 22; moderate‐certainty evidence). Combination therapy may result in little to no difference in mortality (low‐certainty evidence), and a clinically negligible improvement in 6MWD (mean difference (MD) 19.4 m, 95% CI 10.5 to 28.3; moderate‐certainty evidence). There was little to no change in Borg Dyspnea Scale (low‐certainty evidence). The evidence for WHO functional class was very uncertain. There may be little to no difference in serious adverse events and trial withdrawals between the groups (low‐certainty evidence).

Combination therapy versus phosphodiesterase type 5 inhibitor

The evidence is very uncertain about the effect of combination treatment on clinical worsening compared to PDE5i alone (RR 0.68, 95% CI 0.33 to 1.39; 142 fewer per 1000 participants, 95% CI 298 fewer to 173 more; 4 trials, 1372 participants; very low‐certainty evidence), although exclusion of high‐bias studies suggested a potential benefit (RR 0.57, 95% CI 0.44 to 0.73). The evidence on hospitalisations was very uncertainty, while there was little to no difference in mortality (low‐certainty evidence). There was a clinically negligible improvement in 6MWD (MD 20.4 m, 95% CI 10.7 to 30.2; moderate‐certainty evidence) and little to no change in Borg Dyspnea Scale (low‐certainty evidence). The evidence for WHO functional class was very uncertain. Serious adverse events may be comparable (low‐certainty evidence). Combination therapy reduces withdrawal from the trial compared to PDE5i alone (RR 0.84, 95% CI 0.71 to 0.99; 64 fewer per 1000 participants, 95% CI 117 fewer to 4 fewer; NNTB 16, 95% CI 9 to 250; low‐certainty evidence).

Phosphodiesterase type 5 inhibitor versus endothelin receptor antagonist

PDE5i likely results in little to no difference in clinical worsening compared to ERA (RR 0.92, 95% CI 0.71 to 1.20; 3 trials, 644 participants; moderate‐certainty evidence). The evidence is very uncertain for mortality and hospitalisation. PDE5i results in little to no difference in 6MWD compared to ERA (MD 18.4 m, 95% CI −50.2 to 86.9; low‐certainty evidence). The impact on WHO functional class worsening, serious adverse events, and trial withdrawal was also uncertain, with all outcomes supported by very low‐ or low‐certainty evidence. Overall, current data do not provide reliable conclusions on the relative efficacy or safety of PDE5i versus ERA.

Authors' conclusions

Combination therapy for PAH offers benefits over monotherapies, reducing clinical worsening compared to ERA alone (high certainty). Their benefits over PDE5i are less certain, although potentially favourable when studies at high risk of bias are excluded. Hospitalisation rates are likely reduced with combination therapy compared to ERA, but the effect is very uncertain when combination therapy is compared to PDE5i. Uncertainty also persists regarding its impact on mortality and functional outcomes, such as 6MWD and WHO functional class. Serious adverse events and withdrawal rates are similar between combination therapy and monotherapies, with varying levels of certainty, although withdrawals may favour combination therapy over PDE5i. Comparative analyses of PDE5i and ERA provided mixed results with varying levels of certainty. These findings could inform whether initial combination therapy should become the standard of care in people with group 1 PAH with WHO functional class levels of II or III. However, the review's limited representation of Black people raises concerns about generalisability, given the observed differences in response to ERAs between Black and White people with PAH in the literature.

Funding

This review had no dedicated funding.

Registration

Protocol available via DOI10.1002/14651858.CD015824.

Plain language summary

What are the best medications, either alone or combined, for people with group 1 pulmonary arterial hypertension, and do they have any serious unwanted effects?

Key messages

Combination therapies for people with group 1 pulmonary arterial hypertension are more effective at preventing clinical worsening (becoming more ill) and likely reduce hospitalisation rates compared to treatment with an endothelin receptor antagonist alone. It is uncertain whether they offer greater benefit in preventing clinical worsening or hospitalisations compared to a phosphodiesterase type 5 inhibitor alone.
There is also no strong evidence that combination therapy improves people's ability to perform physical tasks or deaths compared to either medication alone.
People taking combination therapy and those taking just one medicine had similar serious unwanted effects and were about as likely to stop treatment. Slightly fewer people stopped treatment when using combination therapy compared to phosphodiesterase type 5 inhibitor.

What is group 1 pulmonary arterial hypertension?

Pulmonary hypertension is high blood pressure in the arteries of the lungs (pulmonary arteries). It is split into five categories (groups 1 to 5), with each requiring a different treatment approach. Group 1 pulmonary arterial hypertension is rare and involves high pressure specifically in the pulmonary arteries, without problems in other parts of the lungs. It can be caused by factors such as genetics, some medications, or other medical conditions. Ignoring or not treating group 1 pulmonary arterial hypertension properly can reduce quality of life, increase the risk of hospitalisation, and increase deaths.

What is the treatment for group 1 pulmonary arterial hypertension?

Medicines for treating group 1 pulmonary arterial hypertension help widen the blood vessels in the lungs, which in turn lowers the pressure in the pulmonary arteries. Medications called endothelin receptor antagonists, phosphodiesterase type 5 inhibitors, soluble guanylate cyclase stimulators, and prostacyclin analogues can be used alone or in combinations. According to guidelines, a common combination is endothelin receptor antagonist with phosphodiesterase type 5 inhibitor.

What did we want to find out?

We wanted to see how well endothelin receptor antagonists and phosphodiesterase type 5 inhibitors work alone or together to treat group 1 pulmonary arterial hypertension. We wanted to know how many people became more ill (called disease worsening), needed hospitalisation, and died. We also wanted to see if there were any serious unwanted effects from treatment.

What did we do?

We looked for studies that compared endothelin receptor antagonist alone, phosphodiesterase type 5 inhibitor alone, or endothelin receptor antagonist plus phosphodiesterase type 5 inhibitor (combination therapy) in people with group 1 pulmonary arterial hypertension. We checked each study to ensure it was fair and reliable, considering factors like study methods and size.

What did we find?

We found nine studies with 1807 people who were monitored for around 16 weeks.

Combination treatment reduces disease worsening compared to endothelin receptor antagonist alone and likely reduces the risk of hospitalisation. It is unclear whether combination therapies are more effective than a phosphodiesterase type 5 inhibitor alone in preventing disease worsening or hospitalisations. There was no strong evidence that combination treatment improves people's ability to perform physical tasks or deaths compared to endothelin receptor antagonist or phosphodiesterase type 5 inhibitor alone. Serious unwanted effects were similar for people using combination or single therapies. Slightly fewer people stopped treatment with combination therapy compared to phosphodiesterase type 5 inhibitor.

What are the limitations of the evidence?

Not all studies reported deaths, which makes it harder for us to be sure about our conclusions for this. Many people dropped out of some studies when comparing combination therapy to single therapies, which could have influenced the results. More research might be needed to understand these aspects better.

How up to date is the evidence?

This evidence is up to date to 13 March 2024.

Summary of findings

Summary of findings 1. Combination therapy versus endothelin receptor antagonist for group 1 pulmonary arterial hypertension.

Combination therapy versus endothelin receptor antagonist for group 1 pulmonary arterial hypertension
Patient or population: group 1 pulmonary arterial hypertension Setting: outpatient Intervention: combination therapy Comparison: ERA
Outcome and follow‐up	Number pf participants (studies)	Relative effect (95% CI)	Absolute effects (95% CI)*			Certainty	What happens?
Outcome and follow‐up	Number pf participants (studies)	Relative effect (95% CI)	ERA	Combination therapy	Difference	Certainty	What happens?
Clinical worsening  Follow‐up: 3–6 months	1139 (5 RCTs)	RR 0.53 (0.41 to 0.68)	239 per 1000	127 per 1000 (98 to 163)	113 fewer per 1000 (from 141 fewer to 77 fewer)	⊕⊕⊕⊕ High	Combination therapy reduces clinical worsening compared to ERA alone.
Mortality  Follow‐up: 3–6 months	748 (4 RCTs)	RR 1.83 (0.61 to 5.48)	11 per 1000	20 per 1000 (7 to 59)	9 more per 1000 (from 4 fewer to 48 more)	⊕⊕⊝⊝ Low^a	Combination therapy may result in little to no difference in mortality compared to ERA alone.
Hospitalisation  Follow‐up: 3–6 months	877 (3 RCTs)	RR 0.32 (0.19 to 0.55)	103 per 1000	33 per 1000 (20 to 57)	70 fewer per 1000 (from 83 fewer to 46 fewer)	⊕⊕⊕⊝ Moderate^b	Combination therapy likely reduces hospitalisations compared to ERA monotherapy.
Change from baseline in 6‐minute walk distance Follow‐up: 3–6 months	787 (5 RCTs)	—	The mean change from baseline was 19.5 m (range 14.1–38.5)^c	—	MD 19.4 (10.5 to 28.3)	⊕⊕⊕⊝ Moderate^b	Combination therapy likely results in a clinically negligible improvement in 6‐minute walk distance compared to ERA alone.
Serious adverse events Follow‐up: 3–6 months	699 (3 RCTs)	RR 1.02 (0.83 to 1.25)	371 per 1000	378 per 1000 (308 to 464)	7 more per 1000 (from 63 fewer to 93 more)	⊕⊕⊝⊝ Low^b,d	Combination therapy results in little to no difference in serious adverse events compared to ERA alone.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; ERA: endothelin receptor antagonist; MD: mean difference; RCT: randomised controlled trial; RR: risk ratio.
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
See the interactive version of this table: https://gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_457580246224668704.

Open in a new tab

^a Downgraded two levels due to very serious imprecision. ^b Total size of fewer than 1000 participants may suggest a small‐study effect [1]. Downgraded one level for publication bias. ^c Data derived from the control group. ^d Downgraded one level due to serious imprecision.

Summary of findings 2. Combination therapy versus phosphodiesterase type 5 inhibitor for group 1 pulmonary arterial hypertension.

Combination therapy versus phosphodiesterase type 5 inhibitor for group 1 pulmonary arterial hypertension
Patient or population: group 1 pulmonary arterial hypertension Setting: outpatient Intervention: combination therapy Comparison: PDE5i
Outcome and follow‐up	Number of participants (studies)	Relative effect (95% CI)	Absolute effects (95% CI)*			Certainty	What happens?
Outcome and follow‐up	Number of participants (studies)	Relative effect (95% CI)	PDE5i	Combination therapy	Difference	Certainty	What happens?
Clinical worsening  Follow‐up: range 3–6 months	1372 (4 RCTs)	RR 0.68 (0.33 to 1.39)	445 per 1000	302 per 1000 (147 to 618)	142 fewer per 1000 (from 298 fewer to 173 more)	⊕⊖⊖⊖ Very low^a,b	The evidence is very uncertain about the effect of combination therapy on clinical worsening compared to PDE5i.
Mortality  Follow‐up: range 3–6 months	1020 (4 RCTs)	RR 0.82 (0.57 to 1.18)	123 per 1000	101 per 1000 (70 to 145)	22 fewer per 1000 (from 53 fewer to 22 more)	⊕⊕⊖⊖ Low^b	Combination therapy may result in little to no difference in mortality compared to PDE5i alone.
Hospitalisation Follow‐up: range 3–6 months	374 (1 RCT)	RR 0.40 (0.18 to 0.90)	9.9%	4.0% (1.8 to 8.9)	6.0% less (8.1 less to 1 less)	⊕⊖⊖⊖ Very low^c,d,e	The evidence is very uncertain about the effect of combination therapy on hospitalisations compared to PDE5i.
Change from baseline in 6‐minute walk distance Follow‐up: range 3–6 months	901 (5 RCTs)	—	The mean change from baseline was 23.5 m (range −6.4 to 43.3)^f	—	MD 20.4 (10.7 to 30.2)	⊕⊕⊕⊖ Moderate^e	Combination therapy likely results in a clinically negligible improvement in 6‐minute walk distance compared to PDE5i alone.
Withdrawal from the trial Follow‐up: range 3–6 months	939 (3 RCTs)	RR 0.84 (0.71 to 0.99)	403 per 1000	338 per 1000 (286 to 399)	64 fewer per 1000 (from 117 fewer to 4 fewer)	⊕⊕⊖⊖ Low^d,e	Combination therapy reduces withdrawal from the trial compared to PDE5i alone.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference; PDE5i: phosphodiesterase type 5 inhibitor; RCT: randomised controlled trial; RR: risk ratio.
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
See the interactive version of this table: https://gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_457600871628380398.

Open in a new tab

^a Downgraded one level due to serious inconsistency. ^b Downgraded two levels due to very serious imprecision. ^c High dropout rates. ^d Downgraded one level due to serious imprecision. ^e Total size of fewer than 1000 participants may suggest a small‐study effect [1]. Downgraded one level for publication bias. ^f Data derived from the control group.

Summary of findings 3. Phosphodiesterase type 5 inhibitor versus endothelin receptor antagonist for group 1 pulmonary arterial hypertension.

Phosphodiesterase type 5 inhibitor versus endothelin receptor antagonist for group 1 pulmonary arterial hypertension
Patient or population: group 1 pulmonary arterial hypertension Setting: outpatient Intervention: PDE5i Comparison: ERA
Outcome and follow‐up	Number of participants (studies)	Relative effect (95% CI)	Absolute effects (95% CI)*			Certainty	What happens?
Outcome and follow‐up	Number of participants (studies)	Relative effect (95% CI)	ERA	PDE5i	Difference	Certainty	What happens?
Clinical worsening Follow‐up: range 3–6 months	644 (3 RCTs)	RR 0.92 (0.71 to 1.20)	286 per 1000	263 per 1000 (203 to 343)	23 fewer per 1000 (from 83 fewer to 57 more)	⊕⊕⊕⊖ Moderate^a	PDE5i likely results in little to no difference in reducing clinical worsening compared to ERA.
Mortality Follow‐up: range 3–6 months	352 (3 RCTs)	RR 3.01 (0.74 to 12.32)	12 per 1000	35 per 1000 (9 to 142)	23 more per 1000 (from 3 fewer to 131 more)	⊕⊖⊖⊖ Very low^b	The evidence is very uncertain about the effect of PDE5i on mortality compared to ERA.
Hospitalisation Follow‐up: range 3–6 months	273 (2 RCTs)	RR 0.63 (0.33 to 1.21)	152 per 1000	96 per 1000 (50 to 184)	56 fewer per 1000 (from 102 fewer to 32 more)	⊕⊖⊖⊖ Very low^c,d	The evidence is very uncertain about the effect of PDE5i on reducing hospitalisation compared to ERA.
Change from baseline in 6‐minute walk distance	351 (3 RCTs)	—	The mean change from baseline was 34.9 m (range 18.8–59.0)^e	—	MD 18.4 (−50.2 to 86.9)	⊕⊕⊖⊖ Low^d,f	PDE5i results in little to no difference in 6‐minute walk distance compared to ERA.
Serious adverse events Follow‐up: range 3–6 months	382 (2 RCTs)	RR 1.09 (0.84 to 1.40)	353 per 1000	385 per 1000 (296 to 494)	32 more per 1000 (from 56 fewer to 141 more)	⊕⊕⊖⊖ Low^a,d	PDE5i results in little to no difference in serious adverse events compared to ERA.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; ERA: endothelin receptor antagonist; MD: mean difference; PDE5i: phosphodiesterase type 5 inhibitor; RCT: randomised controlled trial; RR: risk ratio.
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
See the interactive version of this table: https://gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_457602723057606335.

Open in a new tab

^a Downgraded one level for serious imprecision. ^b Downgraded three levels for extremely serious imprecision. ^c Downgraded two levels due to very serious imprecision. ^d Total size of fewer than 1000 participants may suggest a small‐study effect [1]. Downgraded one level for publication bias. ^e Data derived from the control group.  ^f Variable point estimates, non‐overlapping confidence intervals, and statistical tests of heterogeneity suggest inconsistency. Downgraded one level.

Background

Description of the condition

Pulmonary hypertension is a complex and progressive disorder characterised by a pathological elevation in the pressure of pulmonary arteries, ultimately leading to severe cardiopulmonary impairment. Pulmonary hypertension is defined by a pulmonary artery pressure that exceeds 20 mmHg at rest, which can be clinically determined using right heart catheterisation [2].

Pulmonary arterial hypertension (PAH), a subset of pulmonary hypertension, is categorised as World Health Organization (WHO) group 1. This rare disorder afflicts between one and three individuals per million globally, with a disproportionately higher prevalence in females than in males [3]. While the exact aetiology of PAH remains unclear, it is believed to be multifactorial, resulting from an amalgamation of genetic predisposition, environmental insults, and aberrant vasoconstrictive and proliferative pathways [4].

The pathophysiology of PAH is complex and involves the narrowing and thickening of the pulmonary arteries, leading to an increase in pressure in the pulmonary circulation. This increase in pressure results in an increase in workload on the right ventricle, which over time can lead to right ventricular dysfunction and heart failure. Additionally, PAH leads to an increase in resistance to blood flow through the lungs, causing an increase in right ventricular afterload and reducing right ventricular output [5].

PAH can be classified into several different subtypes based on aetiology, including idiopathic PAH, heritable PAH, PAH associated with connective tissue diseases, PAH associated with congenital heart disease, PAH associated with HIV infection, and drug‐induced PAH. The diagnosis of PAH is based on clinical presentation, imaging studies, and right heart catheterisation, which remains the gold standard for diagnosis. The outcome of untreated PAH can be devastating, including right heart failure and premature mortality, underscoring the seriousness of the disease [6].

Untreated PAH has a poor prognosis. For idiopathic PAH, without intervention, the average survival after diagnosis is only two to three years, dropping to less than six months for people in New York Heart Association (NYHA) functional class IV. The key factor influencing survival is right ventricular function, a primary cause of death in advanced idiopathic PAH cases. In the absence of treatment, median survival stands at 2.8 years, with one‐, three‐, and five‐year survival rates of 68%, 48%, and 34%, respectively, while untreated PAH often leads to heart failure, resulting in a weakening of the heart and reduced circulation throughout the body. Nonetheless, treatment can significantly alleviate symptoms and enhance the quality of life [7].

Description of the intervention and how it might work

Pharmacological treatment for WHO group 1 PAH aims to improve symptoms and exercise capacity and delay disease progression. Treatment typically involves the use of pulmonary vasodilators, which reduce pulmonary vascular resistance and improve blood flow through the pulmonary circulation [8]. These agents include, but are not limited to, the following.

Endothelin receptor antagonists (ERAs): ambrisentan, bosentan, macitentan, and sitaxsentan
Prostacyclin analogues: beraprost, epoprostenol, iloprost, selexipag, and treprostinil
Phosphodiesterase type 5 inhibitors (PDE5i): sildenafil, tadalafil, and vardenafil
Soluble guanylate cyclase stimulator: riociguat

The 2015 European pulmonary hypertension guidelines proposed a risk stratification strategy that divides people into three categories: low risk (less than 5% mortality), intermediate risk (5% to 10% mortality), and high risk (greater than 10% mortality) [9]. This is based on various clinical and haemodynamic parameters. These parameters include WHO functional class, 6‐minute walk distance, levels of B‐type natriuretic peptide or N‐terminal pro B‐type natriuretic peptide, right atrial pressure, cardiac index, and mixed venous oxygen saturation.

Combination therapy may also be used to achieve greater clinical benefits [10]. Other medications, such as diuretics, oxygen, and anticoagulants, may also be used in conjunction with pulmonary vasodilators to manage symptoms and reduce the risk of thromboembolism. Combination therapy with ERA and PDE5i is recommended for people with PAH who are at low or intermediate risk of death within one year of diagnosis, and who have WHO functional class II or III symptoms. Treatment for WHO group 1 PAH requires regular monitoring to assess the efficacy of therapy and adjust the treatment plan as necessary [8].

ERAs work by blocking the action of endothelin‐1, a potent vasoconstrictor that contributes to the pathogenesis of PAH. Prostacyclin analogues are vasodilators that improve blood flow through the lungs and reduce the workload on the right ventricle. PDE5is work by increasing the levels of cyclic guanosine monophosphate, which leads to vasodilation and improved blood flow through the lungs. Soluble guanylate cyclase stimulators work by increasing the production of cyclic guanosine monophosphate and promoting vasodilation [11].

These medications have been shown to improve 6‐minute walk distance, exercise capacity, and quality of life in people with PAH, with a significant reduction in the need for hospitalisation and risk of death [12, 13, 14].

Combination therapy for PAH involves using two or more drugs from different classes with the aim of enhancing the therapeutic effect and minimising the adverse effects of each individual drug. Studies have shown that combination therapy is more effective than monotherapy in improving exercise capacity and delaying disease progression in people with PAH [10, 15].

One common type of combination therapy for PAH involves ERAs plus PDE5is, which improves exercise capacity, delays disease progression, and enhances quality of life in people with PAH. Combining ERAs with PDE5is is generally well tolerated with adverse events consistent with the known side effects of the individual drugs [15]. The AMBITION study demonstrated that initial combination therapy with ambrisentan (ERA) and tadalafil (PDE5i) was superior to either agent alone used as upfront therapy (Galiè 2015 [16]).

Additionally, the combination of ERAs with prostacyclin analogues is also effective in improving exercise capacity and delaying disease progression in people with PAH [13]. Another promising combination therapy is soluble guanylate cyclase stimulators plus ERAs, which improves exercise capacity, haemodynamics, and quality of life in people with PAH [17, 18].

Why it is important to do this review

The use of combination therapy involves carefully selecting drugs from different classes based on the individual needs of each person with PAH and requires close monitoring of disease progression and adverse effects.

Despite promising results from clinical trials, challenges associated with combination therapy for PAH remain, including the cost of treatment, the potential for drug interactions and adverse effects, and the need for close monitoring of disease progression and response to treatment.

The 2022 European Society of Cardiology/European Respiratory Society guidelines recommend initial combination therapy with a PDE5i and an ERA for people with idiopathic, heritable, drug‐associated, or connective tissue disease‐associated PAH who present at low or intermediate risk of death (class I recommendation) [8].

There are multiple medications available in both medication classes, and efficacy may vary across different combinations. It is important to assess the benefits and harms of combining a PDE5i and an ERA compared with either agent alone, and determine if any combination is superior to monotherapy. The guidelines do not provide clear information regarding the evidence supporting their recommendations. Further research is needed to better understand the optimal combination therapy for PAH, including appropriate dosing, treatment duration, and long‐term safety and efficacy.

Objectives

Methods

We followed the Methodological Expectations for Cochrane Intervention Reviews when conducting the review [19] and adhered to the PRISMA 2020 guidelines for reporting [20].

Changes from protocol methods

We made the following change from the protocol [21]. We included the withdrawal from the trial in the summary of findings tables instead of Borg Dyspnea Scale scores, for which we could not conduct a meaningful comparison due to the paucity of data.

Criteria for considering studies for this review

Types of studies

We included randomised controlled trials (RCTs) with a duration of at least 12 weeks of treatment, regardless of whether they were published or unpublished. To minimise the risk of bias or confounding factors in the included studies, we excluded cluster‐, cross‐over, and quasi‐RCTs (i.e. studies where participants were allocated using non‐random methods, such as alternation or accessible information such as date of birth or medical record number, rather than true randomisation) to reduce bias.

Types of participants

Eligible participants aged at least 12 years with a diagnosis of PAH belonging to WHO group 1, including idiopathic PAH, familial PAH, connective tissue disease‐associated PAH, PAH associated with HIV infection, PAH associated with portal hypertension, PAH associated with congenital heart disease, or PAH associated with anorexigen use or toxins. Additionally, participants must have met the following haemodynamic criteria, as determined by right heart catheterisation prior to screening.

Mean pulmonary arterial pressure 20 mmHg or greater
Pulmonary capillary wedge pressure or left ventricular end‐diastolic pressure 15 mmHg or less
Pulmonary vascular resistance 3 Wood units or greater when measured

Furthermore, at the screening visit, participants must have been able to walk 100 m to 500 m, or at least 90% of participants must have belonged to WHO functional class II or III.

Types of interventions

We included studies that compared a combination of an ERA (such as ambrisentan, bosentan, macitentan) and a PDE5i (such as sildenafil, tadalafil, vardenafil) with either agent alone. We selected only studies using dosing approved by regulatory agencies such as the European Medicines Agency (EMA), the National Health Service, and the US Food and Drug Administration (FDA). All formulations of PDE5is and ERAs were combined in the analysis. We allowed the use of anticoagulants, vasodilators, diuretics, cardiac glycosides, and supplemental oxygen, but no other PAH‐specific medications such as prostanoids or soluble guanylate cyclase stimulators. We listed the treatment arms for each study in Table 4 and Supplementary material 2.

1. Demographics of the studies included.

Study ID	Number of participants included in the analysis	Duration of study (weeks)	Therapeutic arms	Mean age (years)	Gender, % female	% White	WHO functional class (%)	PAH type (%)	Mean baseline 6MWD (m)
Study ID	Number of participants included in the analysis	Duration of study (weeks)	Therapeutic arms	Mean age (years)	Gender, % female	% White	I/II/III/IV	Idiopathic/associated	Mean baseline 6MWD (m)
Barst 2011 [62] (PHIRST‐1)^a	206	16	Bosentan vs tadalafil vs bosentan + tadalafil	52.5 (SD 15.5)	76	81	1/33/66/1	62/38	346 (SD 78)
Galiè 2015 (AMBITION)	500	24	Ambrisentan vs tadalafil vs ambrisentan + tadalafil	54.3 (SD 14.6)	78	89	0/31/69/0	56/44	353 (SD 90)
Grünig 2024 [63] (A DUE)	187	16	Macitentan vs tadalafil vs macitentan + tadalafil	50.2 (SD 15.2)	78	88	0/51/49/0	54/46	354 (SD 89)
McLaughlin 2015 (COMPASS‐2)	334	16	Bosentan + sildenafil vs sildenafil	53.9 (SD 15.6)	76	89	0/42/57/1	64/36	360 (SD 76)
Mohammed 2021	27	12	Ambrisentan + sildenafil vs sildenafil	28.8 (SD 12.0)	48	0	Mean NYHA class 2.1 (SD 0.26)	19/81	468 (SD 118)
Pulido 2013 (SERAPHIN)^b	300	115–129	Macitentan + PDE5i vs PDE5i	46.1 (SD 16.0)	77	55	0/51/47/2	54/46	356 (SD 102)
Vizza 2017	103	12	Bosentan + sildenafil vs sildenafil	56.1 (SD 14.5)	76	86	0/34/65/1	65/35	352 (SD 81)
Wilkins 2005 (SERAPH)	26	16	Bosentan vs sildenafil	42.9 (range 27–62)	81	85	NR	92/8	297 (SD 81)
Zhuang 2014	124	16	Ambrisentan + tadalafil vs ambrisentan	51.5 (SD 13.0)	79	0	0/57/39/4	63/37	349 (SD 79)

Open in a new tab

^a The placebo‐only group was excluded. ^b The macitentan 3 mg group was excluded. The PDE5i group included either sildenafil or tadalafil.

6MWD: six‐minute walk distance; NR: not reported; NYHA: New York Heart Association; PAH: pulmonary arterial hypertension; PDE5i: phosphodiesterase type 5 inhibitor; SD: standard deviation; WHO: World Health Organization.

Our original intention was to compare combination therapy with ERA or PDE5i alone. However, we decided to include comparisons of ERA versus PDE5i to be comprehensive, as we observed differences in results when comparing combination therapy versus ERA alone and combination therapy versus PDE5i alone.

Outcome measures

The role of outcomes in determining eligibility for studies in this systematic review was to ensure that selected studies provided data on the specific effects of PDE5is combined with ERAs compared to either treatment alone in people with group 1 PAH. Studies had to include at least one of the predefined clinical outcomes that directly assessed treatment efficacy. These outcomes included, but were not limited to, clinical worsening, mortality, hospitalisations, exercise capacity (e.g. 6MWD), functional status (e.g. WHO functional class), and symptom scores. Additionally, studies were required to report safety outcomes such as serious adverse events or discontinuation rates.

Critical outcomes

Clinical worsening
- Defined as any of the following: all‐cause mortality, lung transplantation, atrial septostomy, hospitalisation due to PAH, worsening functional class, 15% or greater reduction in 6MWD, or the initiation of specific therapies such as parenteral prostacyclin analogue therapy. The timeframe for assessing clinical worsening varied across studies, but the primary time point was typically the end of the intervention period or the longest available follow‐up time.
Mortality
- Assessed as all‐cause death, with the primary time point being the endpoint of the study or at specified follow‐up intervals (e.g. three months, six months).
Hospitalisation
- Assessed as admissions related to PAH or its complications, with the primary time point being the end of the intervention period, though data from interim assessments were also considered when available.

Important outcomes

Change from baseline in 6MWD
- Exercise capacity assessed using the 6MWD. The primary time point was at the end of the intervention period. The measurement was in metres, with a higher value indicating better exercise capacity. We used a minimal important difference of 33 m, based on previous literature [22].
Change from baseline in WHO functional class
- Functional status assessed using the WHO functional class scale, which ranges from I (no symptoms) to IV (severe symptoms). The primary time point was the end of the intervention period, and the minimal important difference defined as a change of at least one class.
Change from baseline in Borg Dyspnea Scale
- Dyspnoea measured using the Borg Dyspnea Scale, a subjective scale assessing breathlessness. The primary time point was the end of the intervention period, with the minimal important difference defined as a change of 0.9 units [23]. The scale ranges from 0 (no dyspnoea) to 10 (extreme dyspnoea).
Serious adverse events
- Safety outcomes included serious adverse events, as defined by each study. A serious adverse event was typically defined as an event that resulted in one or more of the following: death, life‐threatening conditions (i.e. an immediate risk of death), hospitalisation (initial or prolonged), disability or permanent damage, or any other significant medical event that may not necessarily have required hospitalisation but could have jeopardised the individual or required intervention to prevent one of the outcomes listed above. The primary time point for serious adverse events was typically during the study period or at the end of the intervention, depending on the study design.
Withdrawal from the trial
- Withdrawal rates focusing on the proportion of participants who discontinued the trial due to treatment‐related adverse effects or other reasons. The primary time point was at the end of the study period.

By assessing these outcomes, we aimed to comprehensively evaluate both the efficacy and safety of the combined treatment (PDE5is plus ERAs) compared to individual therapies in people with group 1 PAH. The chosen time points for each outcome were those most relevant to understanding the clinical impacts of the interventions within the scope of the studies included in the review.

Search methods for identification of studies

Electronic searches

We identified studies from searches of the following databases and trial registries using the guidance provided in Chapter 4 of the Cochrane Handbook for Systematic Reviews of Interventions [24]. We used Scopus due to a lack of access to Embase, which indexes everything in Embase along with additional content.

Cochrane Central Register of Controlled Trials (CENTRAL; 2024, Issue 1)
MEDLINE (Ovid) (1950 to 13 March 2024)
Scopus (1970 to 13 March 2024)
US National Institutes of Health Ongoing Trials Register ClinicalTrials.gov (https://www.clinicaltrials.gov; searched 13 March 2024)
World Health Organization International Clinical Trials Registry Platform (https://trialsearch.who.int/; searched 13 March 2024)

The MEDLINE search strategy is listed in Supplementary material 1 and was adapted for use in other databases when appropriate. The search strategy was structured to identify articles containing terms for pulmonary hypertension, ERAs, and PDE5is, allowing for searching for all possible comparisons. There were no restrictions on language or publication type.

Searching other resources

We manually searched CENTRAL for conference abstracts and unconventional literature. In addition, we checked the reference lists of all primary studies and review articles for additional references. We also searched relevant manufacturers' websites for study information. To ensure accuracy, we searched PubMed for errata or retractions from included studies published in full text and reported the date of this search within the review. Additionally, we searched the US FDA and the EMA websites for relevant information.

Data collection and analysis

Selection of studies

Two review authors (YO, EF) independently screened the titles and abstracts of the search results and coded them as either 'retrieve' (eligible or potentially eligible or unclear) or 'do not retrieve'. We obtained the full‐text study reports of all potentially eligible studies, and two review authors (YO, EF) independently screened these for inclusion and recorded the reasons for exclusion of ineligible studies. We resolved any disagreements through discussion or by consulting a third review author (TM).

We identified and excluded duplicate publications and collated multiple reports of the same study so that each study, rather than each report, was the unit of interest in the review. We grouped multiple reports and papers related to a single study under a single reference ID. We documented the selection process in sufficient detail to complete a PRISMA flow diagram and Supplementary material 3 [25].

Data extraction and management

We used a data collection form for study characteristics and outcome data that had been piloted on at least one study in the review. Three review authors (YO, TM, EF) extracted the following study characteristics from the included studies.

Methods: study design, total duration of the study, number of study centres and locations, study setting, withdrawals, and the date of the study.
Participants: number of participants, mean age, age range, gender, ethnicity, severity of the condition, diagnostic criteria, baseline 6MWD, inclusion criteria, and exclusion criteria.
Interventions: intervention and comparison used in the study, concomitant medications, and excluded medications.
Outcomes: primary and secondary outcomes specified and collected, time points reported, and the use of end‐of‐study data for all outcomes with available time‐point data for continuous outcomes.
Note: funding for the trial, notable conflicts of interest of trial authors, information needed to assess bias such as any deviations from intended interventions, and whether data were imputed for key outcomes.
We included the information necessary for evaluating GRADE, including the baseline risk for key outcomes in the active control group.

Two review authors (YO, EF) independently extracted outcome data from the included studies. We prioritised the estimated effects of the intervention in the following order: full intention‐to‐treat analysis, modified intention‐to‐treat analysis, per‐protocol analysis [26].

We resolved any disagreements through consensus or with the involvement of a third review author (TM). One review author (YO) was responsible for transferring data into Review Manager [27]. To ensure accuracy, we double‐checked the data by comparing the information presented in the systematic review with that in the study reports. Additionally, a second review author (EF) spot‐checked the study characteristics for accuracy against the study report.

To structure our review, we followed the steps outlined in the Cochrane Handbook for Systematic Reviews of Interventions, which included: selection of studies, data extraction, quality assessment, statistical analysis, and interpretation of results.

For data synthesis, we undertook the following steps: identification of relevant studies, quality assessment of studies, data extraction, data pooling, assessment of heterogeneity, sensitivity analysis, assessment of publication bias, and presentation of results.

We compiled the features of every study contributing to each comparison and presented them in Supplementary material 2.

To prepare the data for meta‐analysis, we converted the data found in the studies into a format suitable for analysis, following the methods described in Chapter 6 of the Cochrane Handbook for Systematic Reviews of Interventions [28].

Risk of bias assessment in included studies

To assess the risk of bias for each study, two review authors (YO, TM) independently used the criteria outlined in Cochrane's RoB 2 tool (https://methods.cochrane.org/risk-bias-2) [29, 30]. We implemented RoB 2 using the RoB 2 Excel tool [31], and presented consensus decisions for signalling questions in a general repository as supplemental data to ensure transparency. We considered the effect of assignment to intervention as the effect of interest on efficacy outcomes and the effect of adherence on safety outcomes.

We assessed the risk of bias in the outcome measures included in the summary of findings tables according to the following domains.

Bias arising from the randomisation process
Bias due to deviations from intended interventions
Bias due to missing outcome data
Bias in the measurement of the outcome
Bias in the selection of the reported result

We categorised each domain as 'low risk', 'some concerns', or 'high risk' using the algorithms in the RoB 2 tool. For each outcome, we summarised the risk of bias judgements across different studies for each of the listed domains. The overall risk of bias in the result was the least favourable assessment across the domains of bias.

We resolved any disagreements through discussion or by consulting a third review author (TM) if necessary. We used the overall risk of bias judgements in the GRADE approach and summary of findings tables, and included figures to illustrate the risk of bias, which were added to the meta‐analysis figures.

Measures of treatment effect

We analysed dichotomous data using risk ratio (RR) or risk difference, and continuous data using mean difference (MD), along with 95% confidence interval (CI). We entered data presented as a scale with a consistent direction of effect.

Unit of analysis issues

For dichotomous outcomes, we used the number of participants as the unit of analysis, rather than events (i.e. the number of participants admitted to the hospital, rather than the number of admissions per participant).

Where outcomes were reported at multiple time points, we extracted and included the latest reported time point while also considering outcomes reported at other time points. We combined data reported at different time points only if we considered it to be clinically appropriate.

If a single study reported multiple trial arms, we included only the relevant arms.

Dealing with missing data

We were able to retrieve the data from websites of regulatory agencies such as the US FDA and the EMA to verify study characteristics or numerical outcome data. We calculated any missing standard deviations (SD) or other required data using the methods outlined in Chapter 6 of the Cochrane Handbook for Systematic Reviews of Interventions [28].

Reporting bias assessment

All analyses included fewer than 10 studies, and we employed the following criteria to assess small‐study bias: if the number of participants per study was less than 50, per pooled analysis was less than 1000, or per arm was less than 100 in specific cases [1, 32].

Synthesis methods

Meta‐analysis of numerical data

We used all eligible studies in our analysis. The primary analysis included studies regardless of their risk of bias. However, we conducted sensitivity analyses, as outlined in Sensitivity analysis, to examine how the inclusion of studies with a higher risk of bias might influence the results.

We used a random‐effects model to account for the uncertainty arising from differences between studies [33]. We used a fixed‐effect model when it appeared that the observed variation across the studies was primarily due to random error or sampling variation rather than systematic differences or heterogeneity. We only performed meta‐analyses if they were deemed meaningful, meaning that the treatments, participants, and clinical questions being compared were similar enough to warrant pooling of the data.

Synthesis using other methods

We did not have to use the alternative synthesis methods described in Chapter 12 of the Cochrane Handbook for Systematic Reviews of Interventions [34].

Investigation of heterogeneity and subgroup analysis

We assessed heterogeneity by considering four main criteria: the similarity of point estimates, the overlap of CIs, statistical tests of heterogeneity such as the Chi² test, and the I² statistic. We interpreted I² statistics in context, considering individual study size, the impact of variability on overall conclusions, and the type of outcomes (i.e. dichotomous versus continuous). We did not rely on popular I² statistic thresholds, such as 30%, 50%, and 75%. Instead, we conducted a comprehensive assessment, considering the entire body of evidence and visually inspecting the similarity in point estimates [35]. We evaluated clinical heterogeneity through the inspection of differences in the baseline characteristics of the study populations, as well as the type and dose of interventions and the outcomes being measured. We scrutinised methodological heterogeneity by examining the types of study, the tools used to measure outcomes (e.g. self‐report questionnaires, clinical examinations), and the methods of data analysis employed (e.g. intention‐to‐treat, per‐protocol). If we identified substantial heterogeneity, we reported it and explored possible causes by subgroup analysis. We summarised key clinical and methodological characteristics and effect modifiers in Table 4 and Supplementary material 2.

We performed subgroup analyses for all critical and important outcomes based on different formulations within the same medication class. Additionally, we conducted a subgroup analysis for 6MWD using varying study durations. We performed the subgroup analyses for two key reasons. First, we examined different formulations within the same medication class to assess whether variations in pharmacokinetics and pharmacodynamics, such as differences in absorption or release rates in oral formulations, could influence treatment efficacy, safety, or tolerability. Second, the subgroup analysis for 6MWD using varying study durations was done to determine whether the treatment effect differs over time. Short‐term studies might show more immediate improvements, while long‐term studies could capture sustained benefits or changes in response as the disease progresses. This analysis helps identify whether the treatment's effect on 6MWD is consistent over different time periods, offering insights into the treatment's long‐term effectiveness.

We investigated potential sources of substantial heterogeneity. We used the formal test for subgroup interactions provided in Review Manager [27]. We took into account the limitations when interpreting the results of subgroup analyses and interpreted them with caution, especially if the results were not consistent with the main findings of the study, as guided in Section 10.11.2 of the Cochrane Handbook for Systematic Reviews of Interventions [36].

Equity‐related assessment

We carefully examined the breakdown of ethnicity and racial groups in the included studies by reviewing the demographic characteristics reported in each study. Specifically, we looked at how participants were categorised in terms of ethnicity and race, considering the classifications used by each study (e.g. White, Black, Asian, Hispanic/Latino, etc.). We recorded the proportion of participants from each racial and ethnic group where this information was available. In cases where studies did not provide detailed breakdowns or used different categorisations, we made note of these variations. Additionally, we evaluated whether the studies adequately represented diverse populations and whether there were any potential biases or limitations in the reporting of these demographics. We then used this information to assess the generalisability of the study findings and to guide the interpretation of the results.

Sensitivity analysis

We performed sensitivity analyses for all critical and important outcomes by excluding studies at high risk of bias from the overall analysis. We used a model that was not used in the primary analysis (fixed‐effect or random‐effects) in a sensitivity analysis. Additionally, we conducted a sensitivity analysis that excluded studies with a large proportion of missing values, such as a dropout rate of 20% or greater [37].

Certainty of the evidence assessment

We constructed a summary of findings table for the following outcomes: clinical worsening, death, hospitalisation, 6MWD, WHO functional class, and serious adverse events. The most important comparisons for decision‐makers are presented in the table. For each prespecified outcome, we evaluated the certainty of the body of evidence from the included studies using the five GRADE considerations: risk of bias, consistency of effect, imprecision, indirectness, and publication bias [35, 38, 39].

To answer the research questions outlined in the objectives, we used the following PICOs (participants, interventions, comparators, and outcomes).

Participants: individuals diagnosed with WHO group 1 PAH and at least 12 years old.
Interventions: combination therapy with an ERA and a PDE5i.
Comparators: monotherapy with either an ERA or a PDE5i.
Outcomes: critical and important outcomes (see Critical outcomes; Important outcomes).

We followed the methods and guidelines outlined in Chapter 14 of the Cochrane Handbook for Systematic Reviews of Interventions [40], employing GRADEpro GDT software [41]. We used the overall RoB 2 judgement to inform the GRADE assessment. Specifically, the RoB 2 assessment provided insights into the confidence we could place in the results of each study, which we then incorporated into the GRADE evaluation of the certainty of the evidence.

We calculated the expected absolute effects in the control group based on relevant RCTs. We provided justification for all decisions to downgrade the certainty of the evidence using footnotes and made additional comments to aid the reader's understanding where necessary. Two review authors (YO, EF) independently performed the GRADE assessment, with any disagreements resolved by discussion or by involving a third review author (TM). We justified, documented, and incorporated our judgements into the reporting of results for each outcome.

Consumer involvement

Members of Cochrane's Consumer Network participated as peer reviewers, providing feedback on protocols and reviews before publication to ensure relevance for consumers.

Results

Description of studies

Results of the search

The database search identified 1665 records. After removing duplicates, 1369 records remained. We excluded an additional 1331 records following title and abstract review. We assessed the remaining 38 articles and excluded an additional 29 articles, with reasons (see Supplementary material 3). We included nine studies in the meta‐analysis as a result. Figure 1 displays the flow diagram detailing the study selection process. The search was conducted up to 13 March 2024. We contacted two study authors regarding missing data on serious adverse events, but received no response (Pulido 2013 [42]; Zhuang 2014 [43]).

Included studies

We included nine RCTs in the analysis with 1807 participants, as presented in Supplementary material 2. The study and participant characteristics, including study durations, treatment arms, demographics, and WHO functional class are presented in Table 4. The median duration of the trials was 16 weeks (range 12 to 129). Seven studies were conducted at multiple centres, whilst two were single‐centre studies. Eight studies were industry‐funded. The mean age of participants was 51.6 years (SD 15.7), with 77% being females. Seven of nine studies reported an ethnicity breakdown, with the median proportion of White participants being 85%. Idiopathic PAH accounted for 62% of participants. The mean baseline 6MWD was 354 m (SD 89).

The interventions across these studies included various monotherapies, such as bosentan, tadalafil, ambrisentan, macitentan, sildenafil, and other PDE5is, as well as combination therapies. The included dosage was within the range of regulatory agency‐approved dosages, and uptitration to the maximum dose may or may not have been achieved in individual studies. The measured outcomes included clinical worsening, mortality, hospitalisations, 6MWD, WHO functional class, Borg Dyspnea Scale, serious adverse events, and withdrawal from the trial. Despite variability in study duration, treatment regimens, and participant characteristics, the studies were generally comparable in their focus on assessing the effectiveness and safety of PAH treatments, particularly in evaluating functional outcomes.

Demographic and clinical diversity in included studies: insights and gaps

Most studies were conducted across multiple centres, reflecting a broad geographical distribution of participants, while two were single‐centre, indicating more localised populations. These studies spanned various regions, although specific details about participants' residences were often not provided. Ethnic diversity varied, with a median of 85% White participants across the studies that reported ethnicity. Notably, Galiè 2015 reported 89% White participants, whereas Mohammed 2021 [44] and Zhuang 2014 included only Indian and Chinese participants, with no other races.

Socioeconomic status was not directly addressed, but the industry‐funded nature of many studies suggested some socioeconomic homogeneity, particularly regarding treatment access. While detailed occupational, educational, or religious data were absent, the studies focused on clinical outcomes such as 6MWD and WHO functional class, emphasising health‐related quality of life and physical capabilities. These physical and functional assessments hinted at participants' social participation.

These characteristics reflect a mix of clinical diversity, but with some gaps in reporting on socioeconomic factors and ethnic diversity, which could limit the generalisability of the findings across all populations.

Excluded studies

We excluded 29 studies after a full‐text review, and the details, along with reasons for exclusion, are documented in Supplementary material 3. Twenty‐six studies had an ineligible study design, two had an ineligible population, and one had an ineligible comparator.

Risk of bias in included studies

The results of the RoB 2 assessment are presented in Figure 2 and Supplementary material 4. Consensus decisions for signalling questions are available at https://doi.org/10.6084/m9.figshare.28826522.v1. The results are summarised as follows.

Clinical worsening: both Galiè 2015 and McLaughlin 2015 [45] had a high risk of bias due to significant attrition, which affected the reliability of results for clinical worsening.

Death: Galiè 2015 and McLaughlin 2015 had a high risk of bias for this outcome due to notable attrition. Zhuang 2014 raised concerns about potential bias from randomisation, which could also have affected the assessment of death.

Hospitalisation: Galiè 2015 and Zhuang 2014 showed potential bias for this outcome. In Galiè 2015, the bias stemmed from attrition, while in Zhuang 2014 concerns arose from the randomisation process.

6MWD: Galiè 2015 and Zhuang 2014 had a high risk of bias for this outcome, with Galiè 2015 primarily affected by attrition and Zhuang 2014 by randomisation issues.

WHO functional class: both Galiè 2015 and McLaughlin 2015 had a high risk of bias due to attrition, and Zhuang 2014 raised concerns about bias due to randomisation.

Borg Dyspnea Scale: Galiè 2015 and McLaughlin 2015 had a high risk of bias for this outcome, primarily due to significant attrition.

We incorporated these risks of bias into the evidence profile to assess the certainty of the evidence. We excluded no studies from the review based on baseline characteristics or poor quality.

Synthesis of results

A summary of pooled analyses is presented in Supplementary material 5. A data package including analysis data, study data, and other references is available in Supplementary material 6. The definition of clinical worsening used in each study is presented in Table 5 and was consistent across the included studies. Therefore, we decided not to downgrade for indirectness.

2. Definition of clinical worsening.

Study ID	Definition of clinical worsening
Barst 2011 (PHIRST‐1)	Death; transplantation; atrial septostomy; hospitalisation due to worsening PAH; initiation of new PAH therapy; worsening of functional class
Galiè 2015 (AMBITION)	Death; hospitalisation for worsening PAH; disease progression; unsatisfactory long‐term clinical response
Grünig 2024 (A DUE)	Not included
McLaughlin 2015 (COMPASS‐2)	Death; hospitalisation; intravenous prostanoid initiation; atrial septostomy; lung transplantation; worsening PAH (based on PGSA scale + new therapy)
Mohammed 2021	Not included
Pulido 2013 (SERAPHIN)	Death; atrial septostomy; lung transplantation; IV/subcutaneous prostanoids; worsening PAH (15% decrease in 6MWD + symptoms + new treatment)
Vizza 2017	Death; heart/lung transplantation; hospitalisation; clinical deterioration requiring therapy
Wilkins 2005 (SERAPH)	Not included
Zhuang 2014	Death; transplantation; atrial septostomy; hospitalisation; initiation of new therapy; worsening WHO functional class

Open in a new tab

6MWD: 6‐minute walk distance; PAH: pulmonary arterial hypertension; PGSA: Patient Global Self‐Assessment; WHO: World Health Organization.

Combination therapy versus endothelin receptor antagonist

See Table 1.

Clinical worsening

Combination therapy reduced clinical worsening compared to ERA alone (RR 0.53, 95% CI 0.41 to 0.68; 113 fewer per 1000 participants, 95% CI 141 fewer to 77 fewer; number needed to treat for an additional beneficial outcome (NNTB) 9, 95% CI 7 to 13; 5 trials, 1139 participants; Analysis 1.1; Figure 3; high‐certainty evidence).

The results remained unchanged after excluding studies with a high risk of bias, those with a large proportion of missing data, or those using a fixed‐effect or random‐effects model. The results remained consistent across the comparisons utilising different ERAs (i.e. combination therapy versus ambrisentan, bosentan, or macitentan) (P = 0.95).

Mortality

Combination therapy may result in little to no difference in mortality compared to ERA alone (RR 1.83, 95% CI 0.61 to 5.48; 4 trials, 748 participants; Analysis 1.2; Figure 4; low‐certainty evidence).

The results remained unchanged after excluding studies with a high risk of bias, those missing a large proportion of data, or those utilising the fixed‐effect or random‐effects model. The results remained consistent across the comparisons utilising different ERAs (i.e. combination therapy versus ambrisentan, bosentan, or macitentan) (P = 0.91).

Hospitalisation

Combination therapy is likely to reduce hospitalisations compared to ERA monotherapy (RR 0.32, 95% CI 0.19 to 0.55; 70 fewer per 1000 participants, 95% CI 83 fewer to 46 fewer; NNTB 14, 95% CI 12 to 22; 3 trials, 877 participants; Analysis 1.3; Figure 5; moderate‐certainty evidence).

Change from baseline in six‐minute walk distance

Combination therapy likely results in a clinically negligible improvement in 6MWD compared to ERA alone (MD 19.4 m, 95% CI 10.5 to 28.3; 5 trials, 787 participants; Analysis 1.4; Figure 6; moderate‐certainty evidence). Four trials assessed the change in 6MWD at three months, and one trial assessed 6MWD at six months. At both time points, there was no clinically meaningful improvement in 6MWD with combination therapy compared to ERA (3 months: MD 16.1 m, 95% CI 2.7 to 29.4; 6 months: MD 14.4 m, 95% CI −17.0 to 45.8; Analysis 1.5; Figure 7).

Change from baseline in 6‐minute walking distance: combination therapy versus endothelin receptor antagonist (ERA) monotherapy

Change from baseline in 6‐minute walking distance (subgroups with different study durations): combination therapy versus endothelin receptor antagonist (ERA) monotherapy

The results remained unchanged after excluding studies with a high risk of bias, those missing a large proportion of data, or those utilising the fixed‐effect or random‐effects model. The results remained consistent across comparisons using different ERAs (i.e. combination therapy versus ambrisentan, bosentan, or macitentan; P = 0.39) or different study durations (P = 0.92) (Analysis 1.4; Figure 6; Analysis 1.5, Figure 7).

Worsening in World Health Organization functional class

The evidence was very uncertain for WHO functional class for combination therapy compared to ERA alone (RR 0.79, 95% CI 0.48 to 1.29; 5 trials, 791 participants; Analysis 1.6; Figure 8; very low‐certainty evidence).

Worsening in WHO functional class: combination therapy versus endothelin receptor antagonist (ERA) monotherapy

Change from baseline in Borg Dyspnea Scale

Combination therapy results in a clinically negligible improvement in Borg Dyspnea Scale compared to ERA alone, although the 95% CIs did not rule out a clinically important improvement with combination therapy (MD −0.82, 95% CI −1.42 to −0.22; 2 trials, 473 participants; Analysis 1.7; Figure 9; low‐certainty evidence).

Serious adverse events

Combination therapy results in little to no difference in serious adverse events compared to ERA alone (RR 1.02, 95% CI 0.83 to 1.25; 3 trials, 699 participants; Analysis 1.8; Figure 10; low‐certainty evidence).

Serious adverse event: combination therapy versus endothelin receptor antagonist (ERA) monotherapy

There were no studies with a high risk of bias or missing a large proportion of data in this comparison. The results remained unchanged for those utilising the fixed‐effect or random‐effects model. The results remained consistent across the comparisons utilising different ERAs (i.e. combination therapy versus ambrisentan or bosentan) (P = 0.72).

Withdrawal from the trial

Combination therapy results in little to no difference in withdrawal from the trial compared to ERA alone (RR 1.01, 95% CI 0.41 to 2.50; 4 trials, 824 participants; Analysis 1.9; Figure 11; low‐certainty evidence).

Combination therapy versus phosphodiesterase type 5 inhibitor

See Table 2.

Clinical worsening

The evidence is very uncertain about the effect of combination therapy on clinical worsening compared to PDE5i (RR 0.68, 95% CI 0.33 to 1.39; 142 fewer per 1000 participants, 95% CI 298 fewer to 173 more; 4 trials, 1372 participants; Analysis 2.1; Figure 12; very low‐certainty evidence).

The results were unchanged using either the random‐effects model (RR 0.68, 95% CI 0.33 to 1.39) or the fixed‐effect model (RR 1.00, 95% CI 0.93 to 1.08). Statistical heterogeneity decreased from an I² of 84% to 0% when McLaughlin 2015, which had a high attrition rate, was excluded. The pooled analysis indicates that combination therapy probably reduces clinical worsening compared with PDE5i alone (RR 0.57, 95% CI 0.33 to 0.93) when McLaughlin 2015 was excluded, but not when Galiè 2015, which also had a high attrition rate, was excluded (RR 0.65, 95% CI 0.15 to 2.85). The subgroup difference (P = 0.0001) across the different formulations of PDE5i also disappeared when McLaughlin 2015 was excluded, but not when Galiè 2015 was excluded.

Mortality

Combination therapy may result in little to no difference in mortality compared to PDE5i alone (RR 0.82, 95% CI 0.57 to 1.18; 4 trials, 1020 participants; Analysis 2.2; Figure 13; low‐certainty evidence).

The results remained unchanged after excluding studies with a high risk of bias, those missing a large proportion of data, or those utilising the fixed‐effect or random‐effects model. The results remained consistent across the comparisons utilising different PDE5is (i.e. combination therapy versus sildenafil or tadalafil) (P = 0.95).

Hospitalisation

The evidence is very uncertain about the effect of combination therapy on hospitalisations compared to PDE5i (RR 0.40, 95% CI 0.18 to 0.90; 1 trial, 374 participants; Analysis 2.3; Figure 14; very low‐certainty evidence).

Sensitivity or subgroup analysis was not feasible because only one study contributed to this comparison.

Change from baseline in six‐minute walk distance

Combination therapy likely results in a clinically negligible improvement in 6MWD compared to PDE5i alone (MD 20.4 m, 95% CI 10.7 to 30.2; 5 trials, 901 participants; Analysis 2.4; Figure 15; moderate‐certainty evidence). Three trials assessed the change in 6MWD at three months, and two trials assessed 6MWD at six months. In both cases, there was no clinically meaningful improvement in 6MWD with combination therapy compared to PDE5i (3 months: MD 8.5 m, 95% CI −8.9 to 25.8; 6 months: MD 26.0 m, 95% CI 14.2 to 37.9; Analysis 2.5; Figure 16).

Change from baseline in 6‐minute walking distance: combination therapy versus phosphodiesterase type 5 inhibitor (PDE5i) monotherapy

Change from baseline in 6‐minute walking distance: combination therapy versus phosphodiesterase type 5 inhibitor (PDE5i) monotherapy (subgroups with different study durations)

The results remained unchanged after excluding studies with a high risk of bias, those missing a large proportion of data, or those utilising the fixed‐effect or random‐effects model. The results remained consistent across the comparisons utilising different PDE5is (i.e. combination therapy versus sildenafil, tadalafil, or various PDE5is) (P = 0.63; Analysis 2.4) or different study durations (P = 0.10; Analysis 2.5).

Worsening in World Health Organization functional class

The evidence is very uncertain about the effect of combination therapy on the prevention of WHO functional class worsening compared to PDE5i alone (RR 1.02, 95% CI 0.65 to 1.59; 4 trials, 886 participants; Analysis 2.6; Figure 17; very low‐certainty evidence).

Worsening in WHO functional class: combination therapy versus phosphodiesterase type 5 inhibitor (PDE5i) monotherapy

The results remained unchanged after excluding studies with a high risk of bias, those missing a large proportion of data, or those utilising the fixed‐effect or random‐effects model. The results remained consistent across the comparisons utilising different PDE5is (i.e. combination therapy versus sildenafil or tadalafil) (P = 0.48).

Change from baseline in Borg Dyspnea Scale

Combination therapy results in little to no difference in Borg Dyspnea Scale compared to PDE5i alone (MD −0.03, 95% CI −0.42 to 0.36; 2 trials, 701 participants; Analysis 2.7; Figure 18; low‐certainty evidence).

The results remained unchanged after excluding studies with a high risk of bias, those missing a large proportion of data, or those utilising the fixed‐effect or random‐effects model. The results remained consistent across the comparisons utilising different PDE5is (i.e. combination therapy versus sildenafil or tadalafil) (P = 0.64).

Serious adverse events

Combination therapy results in little to no difference in serious adverse events compared to PDE5i alone (RR 0.92, 95% CI 0.56 to 1.51; 3 trials, 715 participants; Analysis 2.8; Figure 19; low‐certainty evidence).

Serious adverse event: combination therapy versus phosphodiesterase type 5 inhibitor (PDE5i) monotherapy

There were no studies with a high risk of bias or missing a large proportion of data in this comparison. The results remained unchanged for those utilising the fixed‐effect or random‐effects model. Evidence suggests that combination therapy may lead to a reduction in serious adverse events compared to sildenafil monotherapy, but the opposite was true for tadalafil monotherapy (P = 0.05 for subgroup differences; Figure 19).

Withdrawal from the trial

Combination therapy reduces withdrawal from the trial compared to PDE5i alone (RR 0.84, 95% CI 0.71 to 0.99; 64 fewer per 1000 participants, 95% CI 117 fewer to 4 fewer; NNTB 16, 95% CI 9 to 250; 3 trials, 939 participants; Analysis 2.9; Figure 20; low‐certainty evidence).

There were no studies with a high risk of bias or missing a large proportion of data in this comparison. The results remained unchanged for those utilising the fixed‐effect or random‐effects model. The results remained consistent across the comparisons utilising different PDE5is (i.e. combination therapy versus sildenafil or tadalafil) (P = 0.21).

Phosphodiesterase type 5 inhibitor versus endothelin receptor antagonist

See Table 3.

Clinical worsening

PDE5i likely results in little to no difference in reducing clinical worsening compared to ERA (RR 0.92, 95% CI 0.71 to 1.20; 3 trials, 644 participants; Analysis 3.1; Figure 21; moderate‐certainty evidence).

There were no studies with a high risk of bias or a high attrition rate in this comparison. The results remained unchanged when utilising either the fixed‐effect or random‐effects model.

Mortality

The evidence is very uncertain about the effect of PDE5i on mortality compared to ERA (RR 3.01, 95% CI 0.74 to 12.32; 3 trials, 352 participants; Analysis 3.2; Figure 22; very low‐certainty evidence).

There were no studies with a high risk of bias or a high attrition rate in this comparison. The results remained unchanged when utilising either the fixed‐effect or random‐effects model.

Hospitalisation

The evidence is very uncertain about the effect of PDE5i on reducing hospitalisation compared to ERA (RR 0.63 95% CI 0.33 to 1.21; 2 trials, 273 participants; Analysis 3.3; Figure 23; very low‐certainty evidence).

There were no studies with a high risk of bias or a high attrition rate in this comparison. The results remained unchanged when utilising either the fixed‐effect or random‐effects model.

Change from baseline in six‐minute walk distance

PDE5i results in little to no difference in 6MWD compared to ERA (MD 18.4 m, 95% CI −50.2 to 86.9; 3 trials, 351 participants; Analysis 3.4; Figure 24; low‐certainty evidence). Two trials assessed the change in 6MWD at three months, and one trial assessed 6MWD at six months. PDE5i may result in clinically negligible improvement in 6MWD compared to ERA at three months but no improvement at six months (3 months: MD 31.5 m, 95% CI −143.5 to 206.4; 6 months: MD −4.3 m, 95% CI −19.5 to 10.9; Analysis 3.5; Figure 25). However, the 95% CIs did not rule out a clinically important improvement with PDE5i compared to ERA at three months. Nonetheless, small sample sizes make the above results uncertain.

Change from baseline in 6‐minute walking distance: phosphodiesterase type 5 inhibitor (PDE5i) versus endothelin receptor antagonist (ERA)

Change from baseline in 6‐minute walking distance (subgroups with different study durations): phosphodiesterase type 5 inhibitor (PDE5i) versus endothelin receptor antagonist (ERA)

There were no studies with a high risk of bias or a high attrition rate in this comparison. The results remained unchanged when utilising either the fixed‐effect or random‐effects model.

Worsening in World Health Organization functional class

The evidence regarding the effect of combination therapy on the prevention of WHO functional class worsening compared to ERA alone is very uncertain (RR 1.17, 95% CI 0.62 to 2.22; 3 trials, 401 participants; Analysis 3.6; Figure 26; very low‐certainty evidence).

Worsening in WHO functional class: phosphodiesterase type 5 inhibitor (PDE5i) versus endothelin receptor antagonist (ERA)

The results remained unchanged after excluding studies with a high risk of bias, those missing a large proportion of data, or those utilising the fixed‐effect or random‐effects model.

Change from baseline in Borg Dyspnea Scale

The evidence regarding the effect of PDE5i on Borg Dyspnea Scale compared to ERA is very uncertain (MD 0, 95% CI −1.69 to 1.69; 1 trial, 244 participants; Analysis 3.7; Figure 27; very low‐certainty evidence).

Sensitivity or subgroup analysis was not feasible because only one study contributed to this comparison.

Serious adverse events

PDE5i results in little to no difference in serious adverse events compared to ERA (RR 1.09, 95% CI 0.84 to 1.40; 2 trials, 382 participants; Analysis 3.8; Figure 28; low‐certainty evidence).

Serious adverse event: phosphodiesterase type 5 inhibitor (PDE5i) versus endothelin receptor antagonist (ERA)

There were no studies with a high risk of bias or a high attrition rate in this comparison. The results remained unchanged when utilising either the fixed‐effect or random‐effects model.

Withdrawal from the trial

The evidence is very uncertain about the effect of PDE5i on withdrawal from the trial compared to ERA (RR 1.09, 95% CI 0.77 to 1.53; 3 trials, 408 participants; Analysis 3.9; Figure 29; very low‐certainty evidence).

There were no studies with a high risk of bias or a high attrition rate in this comparison. The results remained unchanged when utilising either the fixed‐effect or random‐effects model.

Equity assessment

The median proportion of White participants amongst the included studies that reported an ethnicity breakdown was 85%. This proportion was consistent with what has been observed in various PAH registries [46]. All studies were conducted in resource‐rich countries.

Reporting biases

Judgements on the risk of bias for individual outcomes are displayed in Figure 2 and Supplementary material 4.

Discussion

Summary of main results

We included nine studies with 1807 participants. The median duration of these studies was 16 weeks, with durations ranging from 12 to 129 weeks. Treatment regimens include combinations of medications such as ambrisentan, bosentan, macitentan, tadalafil, and sildenafil. The mean age of participants was 51.6 years (SD 15.7), and 77% of participants were female. The median proportion of White participants was 85%. Ninety‐nine percent of the participants belonged to either WHO functional classes II (41% of the population) or III (58% of the population). The median proportion of participants with idiopathic PAH was 62%, ranging from 0% to 92%. The mean baseline 6MWD was 354 m (SD 89).

High‐certainty evidence indicates that combination therapy reduces the risk of clinical worsening compared to ERA alone and likely reduces hospitalisations. However, low‐certainty evidence may suggest little to no difference in mortality. Combination therapy likely results in only a negligible, clinically unimportant improvement in 6MWD, based on moderate‐certainty evidence, and may have little to no effect on Borg Dyspnea Scale, based on low‐certainty evidence. The impact on WHO functional class is very uncertain. Rates of serious adverse events and treatment withdrawals appear similar between groups, with low‐certainty evidence suggesting little to no difference.

Compared to PDE5i alone, the benefits of combination therapy are less certain. Although four trials reported inconsistent results regarding clinical worsening, excluding studies with a high risk of bias suggests a potential benefit. Evidence on hospitalisations remains inconclusive, while mortality appears similar between groups, based on low‐certainty evidence. Improvements in 6MWD are clinically negligible, averaging approximately 20 m. There is little to no change in Borg Dyspnea Scale, based on low‐certainty evidence, and evidence related to WHO functional class is very uncertain. Serious adverse events are comparable based on low‐certainty evidence, while withdrawal rates reduced with combination therapy; however, this is also supported by low‐certainty evidence.

Comparative analyses of PDE5i and ERA therapies yielded mixed results with varying levels of certainty.

Limitations of the evidence included in the review

The demographics in the included studies were very similar, except for Mohammed 2021, which primarily consisted of people with congenital heart disease, specifically Eisenmenger's syndrome, classified as Group 1 PAH. In Mohammed 2021, the mean age was much younger and baseline 6MWD was greater compared with others, as shown in Table 4. The only outcome provided in Mohammed 2021 was the change from baseline in 6MWD. Excluding the study did not affect the interpretation of the results. Nevertheless, the conclusions drawn from this review might not extend to people with congenital heart disease and Eisenmenger's syndrome owing to the limited availability of data. As described in the results section, the demographic characteristics reflect clinical diversity but lack comprehensive reporting on socioeconomic factors and ethnic diversity, potentially limiting the generalisability of the findings.

There was no specific information regarding the allowance of concomitant PAH‐specific medications, such as prostanoids, in McLaughlin 2015, Mohammed 2021, and Wilkins 2005 [47], while others did not permit their use. We presumed that concurrent PAH‐specific medications were not permitted in those studies lacking specific information, given the clinical trial's objectives and protocol, and thus, we did not lower their rating due to potential bias.

We conducted a subgroup analysis for 6MWD at different time points (i.e. three and six months) and found that it seems reasonable to combine the data for three and six months, considering the similarity of the results for combination therapy versus monotherapy. However, this may not be applicable for comparing PDE5i versus ERA, given the substantial inconsistency observed when comparing results at three months versus six months (Analysis 1.5; Analysis 2.5; Analysis 3.5). Nonetheless, small sample sizes make the results uncertain.

In Galiè 2015 and McLaughlin 2015, there were high attrition rates in comparisons between combination therapy and monotherapy. Galiè 2015 provided data on clinical worsening, death, hospitalisation, 6MWD, Borg Dyspnea Scale, and WHO functional class, while McLaughlin 2015 contributed data on clinical worsening, death, and WHO functional class. The proportion of information from Galiè 2015 was insufficient to impact the interpretation of the results, and as such, we did not downgrade for risk of bias. However, the inclusion of results from McLaughlin 2015 comparing combination therapy and PDE5i for clinical worsening increased the heterogeneity. Excluding McLaughlin 2015 from the analysis would increase the certainty of evidence that combination therapy reduces clinical worsening compared to PDE5i from very low to low.

The certainty of evidence on GRADE analysis ranged from high to very low (Table 1; Table 2; Table 3). The downgrading of certainty in the evidence using GRADE for various comparisons in clinical outcomes reflects several key considerations.

Imprecision: this was the most frequent reason for downgrading. It was cited as 'serious' or 'very serious' in nearly all comparisons across outcomes, particularly for mortality, hospitalisation, WHO functional class, and withdrawal from the trial. Imprecision often stems from small sample sizes, wide CIs, or both.
Inconsistency: there was inconsistency, described as 'serious' or 'very serious', in some comparisons, such as clinical worsening and withdrawal from the trial. It reflects variability in results across studies that may indicate differences in study populations, interventions, or other factors.
Publication bias: strong suspicion of publication bias was a recurring issue, particularly for hospitalisation, WHO functional class, serious adverse events, and withdrawal from the trial. This suggests a potential for selective reporting of positive results.
Risk of bias: for hospitalisation, we downgraded the combination therapy versus PDE5i comparison due to a 'serious risk of bias', indicating concerns about study design, implementation, or reporting quality.
Small‐study effects: for 6MWD, we downgraded all comparisons due to the influence of small studies, which are more prone to bias and variability.

The certainty of evidence regarding death was low to very low, primarily attributable to the small sample size and the fact that fewer than five studies reported outcomes for each comparison, with the longest duration of study spanning six months. This reflects the relative rarity of group 1 PAH and the challenges associated with recruiting participants, especially when compared to more prevalent chronic conditions such as chronic obstructive pulmonary disease or asthma. The total number of participants for most outcomes was fewer than 1000, especially for comparisons between PDE5i and ERA, which negatively impacted the certainty of evidence due to the possibility of a small‐study effect [1].

Clinical worsening was identified as a critical outcome in this meta‐analysis due to its frequent use, consistent definition, and strong clinical relevance in PAH trials. Most studies defined clinical worsening as a composite outcome encompassing all‐cause death, transplantation, atrial septostomy, hospitalisation due to PAH, worsening functional class, a 15% or greater decrease in 6MWD, or initiation of advanced therapies such as parenteral prostacyclin (Table 5). While the individual components of this composite may differ in clinical importance and frequency, the overall outcome reflects meaningful disease progression and has been validated in multiple high‐quality studies as prognostic of future adverse events. To address interpretability concerns often associated with composite endpoints, we extracted and analysed individual components — specifically death and hospitalisation — separately. Nevertheless, clinical worsening remains a clinically coherent and prognostically significant endpoint that aligns with current trial design standards prioritising morbidity and mortality over surrogate measures. Its prevention directly impacts quality of life, healthcare utilisation, and long‐term outcomes, supporting its designation as a critical outcome in this meta‐analysis [48, 49].

We included various formulations and combinations of ERA and PDE5i. The interaction between bosentan and sildenafil has been well characterised in the literature and may have potentially introduced bias. Chronic administration of bosentan can decrease overall sildenafil plasma concentrations by 40% to 60% due to bosentan inducing the expression of cytochrome P450 3A4, which is involved in sildenafil metabolism. In contrast, adding bosentan to treatment of people receiving long‐term sildenafil can lead to a 50% increase in plasma bosentan concentrations [50, 51]. However, despite these interactions, combining bosentan and sildenafil has been shown to be feasible and well‐tolerated in people with PAH, resulting in clinical improvement and increased exercise tolerance (Vizza 2017 [52]). The subgroup analysis suggests that the bosentan–sildenafil combination may result in a reduction in serious adverse events compared to sildenafil alone, a contrast not observed when comparing combination therapy to tadalafil (Analysis 2.8). Additional studies will be necessary to further investigate this difference, especially considering the very small sample size when comparing combination therapy to sildenafil alone. Subgroup analyses in this review did not otherwise indicate a difference in benefits or harms between the bosentan–sildenafil combination and other ERA–PDE5i combinations.

Limitations of the review processes

The review process revealed limitations in the identification and integration of studies, as well as in the comprehensiveness of data collection. While most studies were multicentred and included a geographically diverse population, gaps in reporting on ethnicity and socioeconomic factors were evident. A median of 85% White participants was reported across studies, with limited representation of other ethnicities and no detailed socioeconomic data, potentially restricting the generalisability of findings. The exclusion of 29 studies due to design, population, or comparator issues further underscores the selective nature of the review. Demographically, participants were predominantly middle‐aged women, with clinical diversity reflected in the inclusion of both idiopathic and associated PAH cases.

Risk of bias assessments identified high attrition rates as a significant concern in key studies, such as Galiè 2015 and McLaughlin 2015, which impacted outcomes such as clinical worsening and functional class. Sensitivity analyses demonstrated the robustness of some results, particularly for combination therapies, despite these biases. For example, combination therapies reduced clinical worsening and hospitalisations compared to monotherapies such as ERA or PDE5i, with high‐certainty evidence. However, results related to mortality, 6MWD, and functional class changes were less definitive, with evidence often ranging from moderate to very low certainty.

Syntheses of findings were complicated by heterogeneity across studies, particularly in subgroup analyses and missing data. The evidence demonstrated that combination therapies generally improved clinical outcomes, such as hospitalisation rates and clinical worsening, compared to monotherapies, although some findings lacked clinical significance. Despite consistent results in subgroup analyses for certain outcomes, gaps in equity assessment, such as the under‐representation of non‐White participants and resource‐limited settings, raise concerns about the broader applicability of the results. Reporting biases and exclusion of potentially relevant studies further highlight the need for more inclusive and methodologically robust reviews.

Agreements and disagreements with other studies or reviews

The results from previous reviews were essentially similar to those of this review, concluding that combination therapy was associated with a reduction in clinical worsening and a clinically negligible improvement in 6MWD [10, 53, 54, 55].

However, there are several differences in this review compared to others. We limited the comparison to the combination of ERA plus PDE5i with either agent alone, while other reviews combined various classes of PAH‐specific medications without considering indirectness. There is at least one previous meta‐analysis that compared the combination of ERA and PDE5i to either agent alone [53]. However, the previous review included retrospective analyses, while only RCTs were considered in this review, enhancing the certainty of evidence.

We included both efficacy and safety outcomes, including hospitalisation, to better assess benefits and harms, which is critical for the Evidence to Decision framework for guideline developers [56]. To our knowledge, no previous meta‐analysis included the outcome of hospitalisation.

Authors' conclusions

Implications for practice

Combination therapy with phosphodiesterase type 5 inhibitor (PDE5i) plus endothelin receptor antagonist (ERA) in pulmonary arterial hypertension (PAH) offers clear benefits over ERA monotherapy, reducing clinical worsening and likely reducing hospitalisations, without increasing serious adverse events. However, there is little to no difference in mortality (low‐certainty evidence), and improvements in physical function — such as six‐minute walk distance (6MWD) — are clinically negligible. The impact on the World Health Organization (WHO) functional class is very uncertain.

Compared with phosphodiesterase type 5 inhibitor (PDE5i) monotherapy, the benefits of combination therapy are less consistent, with only limited evidence suggesting a possible advantage. Although combination therapy improves certain clinical outcomes, the trivial or uncertain gains in function and the lack of a confirmed survival benefit underscore the importance of shared decision‐making and individualised care. These findings could help determine whether initial combination therapy should become the standard of care for people with group 1 PAH with WHO functional class levels II or III [57, 58].

Equity‐related implications for practice

The findings of this review may not be generalisable to Black people due to their under‐representation, comprising less than 10% of the participants in the included studies. Additionally, it has been observed that Black people with PAH exhibit differences in response to ERAs compared to White people [59].

PAH poses significant challenges in resource‐limited countries. The management of PAH in low‐ and middle‐income countries requires a multifaceted approach that addresses the unique challenges of limited resources, lack of awareness, and access to diagnostic and treatment modalities. Innovative, context‐specific solutions are needed to improve outcomes for people with PAH in these regions.

Implications for research

In this review, the primary benefits of the ERA–PDE5i combination over ETA monotherapy are a likely reduction in hospitalisations with clinically negligible improvement in 6MWD and Borg Dyspnea Scale, and uncertain evidence on WHO functional class, suggesting that the combination therapy is most effective in people at higher risk of hospitalisation. The results of the population‐based analysis could be further refined by identifying people at high risk of hospitalisation and conducting a cost analysis to determine whether reduced medical costs resulting from the decrease in hospitalisations would favour combination therapy over monotherapy, despite higher pharmacy expenses. Research findings can inform treatment guidelines and healthcare policies. If certain combination therapies prove more cost‐effective or beneficial, they could influence clinical practice guidelines and reimbursement policies, improving PAH management [60]. The available cost analysis suggests that sildenafil may be the most cost‐effective option amongst monotherapies in people with PAH with either WHO functional class II or III disease. Nevertheless, additional direct comparative studies are necessary to definitively determine the relative cost‐effectiveness of various PDE5is as components of combination therapy in managing PAH [61]. Understanding the drivers behind these cost reductions is essential for optimising resource allocation in PAH management.

Equity‐related implications for research

The disparities observed in the representation of racial minorities and their differential treatment responses underscore the need for targeted interventions to address health equity in PAH research and clinical care. This may include initiatives to improve diversity in research participation, culturally sensitive healthcare delivery, and personalised treatment approaches that account for racial differences in treatment response.

Comprehensive research is necessary to gain a deeper understanding of the burden of PAH, develop low‐cost screening and diagnostic tools, tailor treatment guidelines to local contexts, enhance access to essential medications, and implement integrated care models in resource‐limited settings.

Supporting Information

Supplementary materials are available with the online version of this article: 10.1002/14651858.CD015824.

Supplementary materials are published alongside the article and contain additional data and information that support or enhance the article. Supplementary materials may not be subject to the same editorial scrutiny as the content of the article and Cochrane has not copyedited, typeset or proofread these materials. The material in these sections has been supplied by the author(s) for publication under a Licence for Publication and the author(s) are solely responsible for the material. Cochrane accordingly gives no representations or warranties of any kind in relation to, and accepts no liability for any reliance on or use of, such material.

Supplementary material 1 Search strategies

CD015824-SUP-01-searchStrategy.html^{(44.1KB, html)}

Supplementary material 2 Characteristics of included studies

CD015824-SUP-02-characteristicsOfIncludedStudies.html^{(93.3KB, html)}

Supplementary material 3 Characteristics of excluded studies

CD015824-SUP-03-characteristicsOfExcludedStudies.html^{(62.7KB, html)}

Supplementary material 4 Risk of bias

CD015824-SUP-04-riskOfBias2.html^{(51.7KB, html)}

Supplementary material 5 Analyses

CD015824-SUP-05-analyses.html^{(699.4KB, html)}

Supplementary material 6 Data package

CD015824-SUP-06-dataPackage.zip^{(61.8KB, zip)}

New

Additional information

Acknowledgements

We would like to express our deepest appreciation to Rachel Alexander at the University of Missouri Health Sciences Library for her assistance with the search design and strategy.

Editorial and peer‐reviewer contributions

The following people conducted the editorial process for this review.

Sign‐off Editor (final editorial decision): Rui Providencia, Institute of Health Informatics Research, University College London, London, UK
Managing Editor (provided editorial guidance to authors, collated peer‐reviewer comments, and edited the article): Marwah Anas El‐Wegoud, Cochrane Central Editorial Service
Editorial Assistant (selected peer reviewers, conducted editorial policy checks, and supported the editorial team): Jacob Hester, Cochrane Central Editorial Service
Copy Editor (copy editing and production): Anne Lawson, Cochrane Central Production Service
Peer reviewers (provided comments and recommended an editorial decision): Nuala Livingstone, Cochrane Evidence Production and Methods Directorate (methods); Jo Platt, Central Editorial Information Specialist (search); Brian Duncan (consumer). One additional peer reviewer provided clinical peer review but chose not to be publicly acknowledged.

Contributions of authors

YO: designed and developed, screened, extracted, resolved conflicts, assessed certainty, contributed to writing and editing, approved the final version prior to submission, and is a guarantor of the review.

TM: extracted study characteristics, contributed to writing and editing, and approved the final version prior to submission.

EF: screened, extracted, assessed certainty, contributed to writing and editing, and approved the final version prior to submission.

YG: contributed to writing and editing, and approved the final version prior to submission.

Declarations of interest

YO: is a Cochrane editor but did not participate in the editorial process.

TM: none.

EF: none.

YG: none.

Sources of support

Internal sources

None, Other

N/A

External sources

None, Other

N/A

Registration and protocol

Protocol available via DOI https://doi.org/10.1002/14651858.CD015824 [21].

Data, code and other materials

Data availability statement: as part of the published Cochrane review, the following materials are available for download via figshare (see: https://figshare.com/articles/dataset/_b_Summary_Data_for_PAH_meta-analysis_b_/29602445): data collection forms; data extracted from included studies; data used for all analyses. Analyses and data management were conducted within Cochrane's authoring tool, Review Manager, using the inbuilt computation methods.

History

Protocol first published: Issue 1, 2024

References

1.Dechartres A, Trinquart L, Boutron I, Ravaud P. Influence of trial sample size on treatment effect estimates: meta-epidemiological study. BMJ 2013;346:f2304. [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Frost A, Badesch D, Gibbs JS, Gopalan D, Khanna D, Manes A, et al. Diagnosis of pulmonary hypertension. European Respiratory Journal 2019;53(1):1801904. [DOI: 10.1183/13993003.01904-2018] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Cheron C, McBride SA, Antigny F, Girerd B, Chouchana M, Chaumais MC, et al. Sex and gender in pulmonary arterial hypertension. European Respiratory Review 2021;30(162):200330. [DOI: 10.1183/16000617.0330-2020] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Humbert M, Guignabert C, Bonnet S, Dorfmüller P, Klinger JR, Nicolls MR, et al. Pathology and pathobiology of pulmonary hypertension: state of the art and research perspectives. European Respiratory Journal 2019;53(1):1801887. [DOI: 10.1183/13993003.01887-2018] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Thenappan T, Ormiston ML, Ryan JJ, Archer SL. Pulmonary arterial hypertension: pathogenesis and clinical management. BMJ 2018;360:j5492. [DOI: 10.1136/bmj.j5492] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Simonneau G, Montani D, Celermajer DS, Denton CP, Gatzoulis MA, Krowka M, et al. Haemodynamic definitions and updated clinical classification of pulmonary hypertension. European Respiratory Journal 2019;53(1):1801913. [DOI: 10.1183/13993003.01913-2018] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
7.McLaughlin VV, Shillington A, Rich S. Survival in primary pulmonary hypertension: the impact of epoprostenol therapy. Circulation 2002;106(12):1477-82. [DOI: 10.1161/01.cir.0000029100.82385.58] [PMID: ] [DOI] [PubMed] [Google Scholar]
8.Humbert M, Kovacs G, Hoeper MM, Badagliacca R, Berger RM, Brida M, et al. 2022 ESC/ERS guidelines for the diagnosis and treatment of pulmonary hypertension. European Heart Journal 2022;43(38):3618-731. [DOI: 10.1093/eurheartj/ehac237] [PMID: ] [DOI] [PubMed] [Google Scholar]
9.Hoeper MM, Kramer T, Pan Z, Eichstaedt CA, Spiesshoefer J, Benjamin N, et al. Mortality in pulmonary arterial hypertension: prediction by the 2015 European pulmonary hypertension guidelines risk stratification model. European Respiratory Journal 2017;50(2):1700740. [DOI: 10.1183/13993003.00740-2017] [PMID: ] [DOI] [PubMed] [Google Scholar]
10.Lajoie AC, Lauzière G, Lega JC, Lacasse Y, Martin S, Simard S, et al. Combination therapy versus monotherapy for pulmonary arterial hypertension: a meta-analysis. Lancet Respiratory Medicine 2016;4(4):291-305. [DOI: 10.1016/S2213-2600(16)00027-8] [PMID: ] [DOI] [PubMed] [Google Scholar]
11.Duarte JD, Hanson RL, Machado RF. Pharmacologic treatments for pulmonary hypertension: exploring pharmacogenomics. Future Cardiology 2013;9(3):335-49. [DOI: 10.2217/fca.13.6] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Barnes H, Brown Z, Burns A, Williams T. Phosphodiesterase 5 inhibitors for pulmonary hypertension. Cochrane Database of Systematic Reviews 2019, Issue 1. Art. No: CD012621. [DOI: 10.1002/14651858.CD012621.pub2] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Gale S. The evolving treatment landscape of pulmonary arterial hypertension. American Journal of Managed Care 2021;27(3 Suppl):S42-52. [DOI: 10.37765/ajmc.2021.88610] [PMID: ] [DOI] [PubMed] [Google Scholar]
14.Liu C, Chen J, Gao Y, Deng B, Liu K. Endothelin receptor antagonists for pulmonary arterial hypertension. Cochrane Database of Systematic Reviews 2021, Issue 3. Art. No: CD004434. [DOI: 10.1002/14651858.CD004434.pub5] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Sitbon O, Sattler C, Bertoletti L, Savale L, Cottin V, Jaïs X, et al. Initial dual oral combination therapy in pulmonary arterial hypertension. European Respiratory Journal 2016;47(6):1727-36. [DOI: 10.1183/13993003.02043-2015] [PMID: ] [DOI] [PubMed] [Google Scholar]
16.Galiè N, Barberà JA, Frost AE, Ghofrani HA, Hoeper MM, McLaughlin VV, et al. Initial use of ambrisentan plus tadalafil in pulmonary arterial hypertension. New England Journal of Medicine 2015;373(9):834-44. [DOI: 10.1056/NEJMoa1413687] [PMID: ] [DOI] [PubMed] [Google Scholar]
17.Ghofrani HA, Grimminger F, Grünig E, Huang Y, Jansa P, Jing ZC, et al. Predictors of long-term outcomes in patients treated with riociguat for pulmonary arterial hypertension: data from the PATENT-2 open-label, randomised long-term extension trial. Lancet Respiratory Medicine 2016;4(5):361-71. [DOI: 10.1016/S2213-2600(16)30019-4] [PMID: ] [DOI] [PubMed] [Google Scholar]
18.Rubin LJ, Galiè N, Grimminger F, Grünig E, Humbert M, Jing ZC, et al. Riociguat for the treatment of pulmonary arterial hypertension: a long-term extension study (PATENT-2). European Respiratory Journal 2015;45(5):1303-13. [DOI: 10.1183/09031936.00090614] [PMID: ] [DOI] [PubMed] [Google Scholar]
19.Higgins JP, Lasserson T, Chandler J, Tovey D, Churchill R. Methodological Expectations of Cochrane Intervention Reviews. Available from https://community.cochrane.org/mecir-manual.
20.Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71. [DOI: 10.1136/bmj.n71] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Oba Y, Goite Y, Maduke T, Fakhouri EW. Phosphodiesterase type 5 inhibitor plus endothelin receptor antagonist compared to either alone for group 1 pulmonary arterial hypertension. Cochrane Database of Systematic Reviews 2024, Issue 1. Art. No: CD015824. [DOI: 10.1002/14651858.CD015824] [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Moutchia J, McClelland RL, Al-Naamani N, Appleby DH, Blank K, Grinnan D, et al. Minimal clinically important difference in the 6-minute walk distance for patients with pulmonary arterial hypertension. American Journal of Respiratory and Critical Care Medicine 2023;207(8):1070-9. [DOI: 10.1164/rccm.202208-1547OC] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Khair RM, Nwaneri C, Damico RL, Kolb T, Hassoun PM, Mathai SC. The minimal important difference in Borg Dyspnea Score in pulmonary arterial hypertension. Annals of the American Thoracic Society 2016;13(6):842-9. [DOI: 10.1513/AnnalsATS.201512-824OC] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Lefebvre C, Glanville J, Briscoe S, Featherstone R, Littlewood A, Marshall C, et al. Technical Supplement to Chapter 4: Searching for and selecting studies. In: Higgins JP, Thomas J, Chandler J, Cumpston MS, Li T, Page MJ, Welch VA, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 6.3 (updated February 2022). Cochrane, 2022. Available from https://www.cochrane.org/authors/handbooks-and-manuals/handbook/archive/v6.3.
25.Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, Ioannidis JP, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. Public Library of Science Medicine 2009;6(7):e1000100. [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Hernán MA, Robins JM. Per-protocol analyses of pragmatic trials. New England Journal of Medicine 2017;377:1391-8. [DOI: 10.1056/NEJMsm1605385] [DOI] [PubMed] [Google Scholar]
27.Review Manager (RevMan). Version 7.12.0. The Cochrane Collaboration, 2024. Available at https://revman.cochrane.org.
28.Higgins JP, Li T, Deeks JJ. Chapter 6: Choosing effect measures and computing estimates of effect. In: Higgins JP, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 6.3 (updated February 2022). Cochrane, 2022. Available from https://www.cochrane.org/authors/handbooks-and-manuals/handbook/archive/v6.3.
29.Higgins JP, Savović J, Page MJ, Elbers RG, Sterne JA. Chapter 8: Assessing risk of bias in a randomized trial. In: Higgins JP, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 6.2 (updated February 2021). Cochrane, 2021. Available from https://www.cochrane.org/authors/handbooks-and-manuals/handbook/archive/v6.2.
30.Sterne JA, Savović J, Page MJ, Elbers RG, Blencowe NS, Boutron I, et al. RoB 2: a revised tool for assessing risk of bias in randomised trials. BMJ 2019;366:l4898. [DOI] [PubMed] [Google Scholar]
31.RoB 2 tool. Available at https://www.riskofbias.info/welcome/rob-2-0-tool (accessed prior to 17 July 2025).
32.Nüesch E, Trelle S, Reichenbach S, Rutjes AW, Tschannen B, Altman DG, et al. Small study effects in meta-analyses of osteoarthritis trials: meta-epidemiological study. BMJ 2010;342:c3515. [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Dettori JR, Norvell DC, Chapman JR. Fixed-effect vs random-effects models for meta-analysis: 3 points to consider. Global Spine Journal 2022;12(7):1624-6. [DOI: 10.1177/21925682221110527] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
34.McKenzie JE, Brennan SE. Chapter 12: Synthesizing and presenting findings using other methods. In: Higgins JP, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 6.4 (updated August 2023). Cochrane, 2023. Available from https://www.cochrane.org/authors/handbooks-and-manuals/handbook/archive/v6.4.
35.Guyatt G, Zhao Y, Mayer M, Briel M, Mustafa R, Izcovich A, et al. GRADE guidance 36: updates to GRADE's approach to addressing inconsistency. Journal of Clinical Epidemiology 2023;158:70-83. [DOI: 10.1016/j.jclinepi.2023.03.003] [PMID: ] [DOI] [PubMed] [Google Scholar]
36.Deeks JJ, Higgins JP, Altman DG. Chapter 10: Analysing data and undertaking meta-analyses. In: Higgins JP, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 6.4 (updated August 2023). Cochrane, 2023. Available from https://www.cochrane.org/authors/handbooks-and-manuals/handbook/archive/v6.4.
37.Bell ML, Kenward MG, Fairclough DL, Horton NJ. Differential dropout and bias in randomised controlled trials: when it matters and when it may not. BMJ 2013;346:e8668. [DOI: 10.1136/bmj.e8668] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Guyatt G, Oxman AD, Akl EA, Kunz R, Vist G, Brozek J, et al. GRADE guidelines: 1. Introduction – GRADE evidence profiles and summary of findings tables. Journal of Clinical Epidemiology 2011;64(4):383-94. [DOI: 10.1016/j.jclinepi.2010.04.026] [PMID: ] [DOI] [PubMed] [Google Scholar]
39.Zeng L, Brignardello-Petersen R, Hultcrantz M, Mustafa RA, Murad MH, Iorio A, et al. GRADE Guidance 34: update on rating imprecision using a minimally contextualized approach. Journal of Clinical Epidemiology 2022;150:216-24. [DOI: 10.1016/j.jclinepi.2022.07.014] [PMID: ] [DOI] [PubMed] [Google Scholar]
40.Schünemann HJ, Higgins JP, Vist GE, Glasziou P, Akl EA, Skoetz N, et al. Chapter 14: Completing 'Summary of findings' tables and grading the certainty of the evidence. In: Higgins JP, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 6.3 (updated February 2022). Cochrane, 2022. Available from https://www.cochrane.org/authors/handbooks-and-manuals/handbook/archive/v6.3.
41.GRADEpro GDT. Version accessed 15 March 2024. Hamilton (ON): McMaster University (developed by Evidence Prime), 2024. Available at https://www.gradepro.org.
42.Pulido T, Adzerikho I, Channick RN, Delcroix M, Galiè N, Ghofrani HA, et al. Macitentan and morbidity and mortality in pulmonary arterial hypertension. New England Journal of Medicine 2013;369(9):809-18. [DOI: 10.1056/NEJMoa1213917] [PMID: ] [DOI] [PubMed] [Google Scholar]
43.Zhuang Y, Jiang B, Gao H, Zhao W. Randomized study of adding tadalafil to existing ambrisentan in pulmonary arterial hypertension. Hypertension Research 2014;37(6):507-12. [DOI: 10.1038/hr.2014.28] [PMID: ] [DOI] [PubMed] [Google Scholar]
44.Mohammed S, Vijayvergiya R, Malhotra S, Rohit MK. A randomized, double-blind, placebo-controlled study to evaluate sildenafil, ambrisentan combination therapy in pulmonary hypertension, particularly of Eisenmenger syndrome. Indian Heart Journal 2021;73(5):633-6. [DOI: 10.1016/j.ihj.2021.07.007] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
45.McLaughlin V, Channick RN, Ghofrani HA, Lemarié JC, Naeije R, Packer M, et al. Bosentan added to sildenafil therapy in patients with pulmonary arterial hypertension. European Respiratory Journal 2015;46(2):405-13. [DOI: 10.1183/13993003.02044-2014] [PMID: ] [DOI] [PubMed] [Google Scholar]
46.Marshall J, Noel-Storr AH, Kuiper J, Thomas J, Wallace BC. Machine learning for identifying randomized controlled trials: an evaluation and practitioner's guide. Research Synthesis Methods 2018;9(4):602-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Wilkins MR, Paul GA, Strange JW, Tunariu N, Gin-Sing W, Banya WA, et al. Sildenafil versus endothelin receptor antagonist for pulmonary hypertension (SERAPH) study. American Journal of Respiratory and Critical Care Medicine 2005;171(11):1292-7. [DOI: 10.1164/rccm.200410-1411OC] [PMID: ] [DOI] [PubMed] [Google Scholar]
48.Hassoun PM. Pulmonary arterial hypertension. New England Journal of Medicine 2021;385(25):2361-76. [DOI: 10.1056/NEJMra2000348] [PMID: ] [DOI] [PubMed] [Google Scholar]
49.Frost AE, Badesch DB, Miller DP, Benza RL, Meltzer LA, McGoon MD. Evaluation of the predictive value of a clinical worsening definition using 2-year outcomes in patients with pulmonary arterial hypertension: a REVEAL Registry analysis. Chest 2013;144(5):1521-9. [DOI: 10.1378/chest.12-3023] [PMID: ] [DOI] [PubMed] [Google Scholar]
50.Burgess G, Hoogkamer H, Collings L, Dingemanse J. Mutual pharmacokinetic interactions between steady-state bosentan and sildenafil. European Journal of Clinical Pharmacology 2008;64(1):43-50. [DOI: 10.1007/s00228-007-0408-z] [PMID: ] [DOI] [PubMed] [Google Scholar]
51.Paul GA, Gibbs JS, Boobis AR, Abbas A, Wilkins MR. Bosentan decreases the plasma concentration of sildenafil when coprescribed in pulmonary hypertension. British Journal of Clinical Pharmacology 2005;60(1):107-12. [DOI: 10.1111/j.1365-2125.2005.02383.x] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Vizza CD, Jansa P, Teal S, Dombi T, Zhou D. Sildenafil dosed concomitantly with bosentan for adult pulmonary arterial hypertension in a randomized controlled trial. BioMed Central Cardiovascular Disorders 2017;17(1):239. [DOI: 10.1186/s12872-017-0674-3] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Kirtania L, Maiti R, Srinivasan A, Mishra A. Effect of combination therapy of endothelin receptor antagonist and phosphodiesterase-5 inhibitor on clinical outcome and pulmonary haemodynamics in patients with pulmonary arterial hypertension: a meta-analysis. Clinical Drug Investigation 2019;39(11):1031-44. [DOI: 10.1007/s40261-019-00841-1] [PMID: ] [DOI] [PubMed] [Google Scholar]
54.Liu HL, Chen XY, Li JR, Su SW, Ding T, Shi CX, et al. Efficacy and safety of pulmonary arterial hypertension-specific therapy in pulmonary arterial hypertension: a meta-analysis of randomized controlled trials. Chest 2016;150(2):353-66. [DOI: 10.1016/j.chest.2016.03.031] [PMID: ] [DOI] [PubMed] [Google Scholar]
55.Tan Z, Wu PY, Zhu TT, Su W, Fang ZF. Efficacy and safety of sequential combination therapy for pulmonary arterial hypertension: a meta-analysis of randomized-controlled trials. Pulmonary Pharmacology & Therapeutics 2022;76:102144. [DOI: 10.1016/j.pupt.2022.102144] [PMID: ] [DOI] [PubMed] [Google Scholar]
56.Alonso-Coello P, Oxman AD, Moberg J, Brignardello-Petersen R, Akl EA, Davoli M, et al. GRADE Evidence to Decision (EtD) frameworks: a systematic and transparent approach to making well informed healthcare choices. 2: Clinical practice guidelines. BMJ 2016;353:i2089. [DOI: 10.1136/bmj.i2089] [PMID: ] [DOI] [PubMed] [Google Scholar]
57.Humbert M, Lau EM. Point: should initial combination therapy be the standard of care in pulmonary arterial hypertension? Yes. Chest 2019;156(6):1039-42. [DOI: 10.1016/j.chest.2019.07.005] [PMID: ] [DOI] [PubMed] [Google Scholar]
58.Channick RN, Saggar R. Counterpoint: should initial combination therapy be the standard of care in pulmonary arterial hypertension? No. Chest 2019;156(6):1043-5. [DOI: 10.1016/j.chest.2019.07.004] [PMID: ] [DOI] [PubMed] [Google Scholar]
59.Medrek SK, Sahay S. Ethnicity in pulmonary arterial hypertension: possibilities for novel phenotypes in the age of personalized medicine. Chest 2018;153(2):310-20. [DOI: 10.1016/j.chest.2017.08.1159] [PMID: ] [DOI] [PubMed] [Google Scholar]
60.Pizzicato LN, Nadipelli VR, Governor S, Mao J, Lanes S, Butler J, et al. Real-world treatment patterns, healthcare resource utilization, and cost among adults with pulmonary arterial hypertension in the United States. Pulmonary Circulation 2022;12(2):e12090. [DOI: 10.1002/pul2.12090] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Coyle K, Coyle D, Blouin J, Lee K, Jabr MF, Tran K, et al. Cost effectiveness of first-line oral therapies for pulmonary arterial hypertension: a modelling study. PharmacoEconomics 2016;34(5):509-20. [DOI: 10.1007/s40273-015-0366-8] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]
62.Barst RJ, Oudiz RJ, Beardsworth A, Brundage BH, Simonneau G, Ghofrani HA, et al. Pulmonary arterial hypertension and response to tadalafil (PHIRST) study group. Tadalafil monotherapy and as add-on to background bosentan in patients with pulmonary arterial hypertension. Journal of Heart and Lung Transplantation 2011;30(6):632-43. [DOI: 10.1016/j.healun.2010.11.009] [PMID: ] [DOI] [PubMed] [Google Scholar]
63.Grünig E, Jansa P, Fan F, Hauser JA, Pannaux M, Morganti A, et al. Randomized trial of macitentan/tadalafil single-tablet combination therapy for pulmonary arterial hypertension. Journal of the American College of Cardiology 2024;83(4):473-84. [DOI: 10.1016/j.jacc.2023.10.045] [PMID: ] [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary material 1 Search strategies

CD015824-SUP-01-searchStrategy.html^{(44.1KB, html)}

Supplementary material 2 Characteristics of included studies

CD015824-SUP-02-characteristicsOfIncludedStudies.html^{(93.3KB, html)}

Supplementary material 3 Characteristics of excluded studies

CD015824-SUP-03-characteristicsOfExcludedStudies.html^{(62.7KB, html)}

Supplementary material 4 Risk of bias

CD015824-SUP-04-riskOfBias2.html^{(51.7KB, html)}

Supplementary material 5 Analyses

CD015824-SUP-05-analyses.html^{(699.4KB, html)}

Supplementary material 6 Data package

CD015824-SUP-06-dataPackage.zip^{(61.8KB, zip)}

Data Availability Statement

[CD015824-bib-0001] 1.Dechartres A, Trinquart L, Boutron I, Ravaud P. Influence of trial sample size on treatment effect estimates: meta-epidemiological study. BMJ 2013;346:f2304. [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD015824-bib-0002] 2.Frost A, Badesch D, Gibbs JS, Gopalan D, Khanna D, Manes A, et al. Diagnosis of pulmonary hypertension. European Respiratory Journal 2019;53(1):1801904. [DOI: 10.1183/13993003.01904-2018] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD015824-bib-0003] 3.Cheron C, McBride SA, Antigny F, Girerd B, Chouchana M, Chaumais MC, et al. Sex and gender in pulmonary arterial hypertension. European Respiratory Review 2021;30(162):200330. [DOI: 10.1183/16000617.0330-2020] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD015824-bib-0004] 4.Humbert M, Guignabert C, Bonnet S, Dorfmüller P, Klinger JR, Nicolls MR, et al. Pathology and pathobiology of pulmonary hypertension: state of the art and research perspectives. European Respiratory Journal 2019;53(1):1801887. [DOI: 10.1183/13993003.01887-2018] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD015824-bib-0005] 5.Thenappan T, Ormiston ML, Ryan JJ, Archer SL. Pulmonary arterial hypertension: pathogenesis and clinical management. BMJ 2018;360:j5492. [DOI: 10.1136/bmj.j5492] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD015824-bib-0006] 6.Simonneau G, Montani D, Celermajer DS, Denton CP, Gatzoulis MA, Krowka M, et al. Haemodynamic definitions and updated clinical classification of pulmonary hypertension. European Respiratory Journal 2019;53(1):1801913. [DOI: 10.1183/13993003.01913-2018] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD015824-bib-0007] 7.McLaughlin VV, Shillington A, Rich S. Survival in primary pulmonary hypertension: the impact of epoprostenol therapy. Circulation 2002;106(12):1477-82. [DOI: 10.1161/01.cir.0000029100.82385.58] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0008] 8.Humbert M, Kovacs G, Hoeper MM, Badagliacca R, Berger RM, Brida M, et al. 2022 ESC/ERS guidelines for the diagnosis and treatment of pulmonary hypertension. European Heart Journal 2022;43(38):3618-731. [DOI: 10.1093/eurheartj/ehac237] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0009] 9.Hoeper MM, Kramer T, Pan Z, Eichstaedt CA, Spiesshoefer J, Benjamin N, et al. Mortality in pulmonary arterial hypertension: prediction by the 2015 European pulmonary hypertension guidelines risk stratification model. European Respiratory Journal 2017;50(2):1700740. [DOI: 10.1183/13993003.00740-2017] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0010] 10.Lajoie AC, Lauzière G, Lega JC, Lacasse Y, Martin S, Simard S, et al. Combination therapy versus monotherapy for pulmonary arterial hypertension: a meta-analysis. Lancet Respiratory Medicine 2016;4(4):291-305. [DOI: 10.1016/S2213-2600(16)00027-8] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0011] 11.Duarte JD, Hanson RL, Machado RF. Pharmacologic treatments for pulmonary hypertension: exploring pharmacogenomics. Future Cardiology 2013;9(3):335-49. [DOI: 10.2217/fca.13.6] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD015824-bib-0012] 12.Barnes H, Brown Z, Burns A, Williams T. Phosphodiesterase 5 inhibitors for pulmonary hypertension. Cochrane Database of Systematic Reviews 2019, Issue 1. Art. No: CD012621. [DOI: 10.1002/14651858.CD012621.pub2] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD015824-bib-0013] 13.Gale S. The evolving treatment landscape of pulmonary arterial hypertension. American Journal of Managed Care 2021;27(3 Suppl):S42-52. [DOI: 10.37765/ajmc.2021.88610] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0014] 14.Liu C, Chen J, Gao Y, Deng B, Liu K. Endothelin receptor antagonists for pulmonary arterial hypertension. Cochrane Database of Systematic Reviews 2021, Issue 3. Art. No: CD004434. [DOI: 10.1002/14651858.CD004434.pub5] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD015824-bib-0015] 15.Sitbon O, Sattler C, Bertoletti L, Savale L, Cottin V, Jaïs X, et al. Initial dual oral combination therapy in pulmonary arterial hypertension. European Respiratory Journal 2016;47(6):1727-36. [DOI: 10.1183/13993003.02043-2015] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0016] 16.Galiè N, Barberà JA, Frost AE, Ghofrani HA, Hoeper MM, McLaughlin VV, et al. Initial use of ambrisentan plus tadalafil in pulmonary arterial hypertension. New England Journal of Medicine 2015;373(9):834-44. [DOI: 10.1056/NEJMoa1413687] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0017] 17.Ghofrani HA, Grimminger F, Grünig E, Huang Y, Jansa P, Jing ZC, et al. Predictors of long-term outcomes in patients treated with riociguat for pulmonary arterial hypertension: data from the PATENT-2 open-label, randomised long-term extension trial. Lancet Respiratory Medicine 2016;4(5):361-71. [DOI: 10.1016/S2213-2600(16)30019-4] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0018] 18.Rubin LJ, Galiè N, Grimminger F, Grünig E, Humbert M, Jing ZC, et al. Riociguat for the treatment of pulmonary arterial hypertension: a long-term extension study (PATENT-2). European Respiratory Journal 2015;45(5):1303-13. [DOI: 10.1183/09031936.00090614] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0019] 19.Higgins JP, Lasserson T, Chandler J, Tovey D, Churchill R. Methodological Expectations of Cochrane Intervention Reviews. Available from https://community.cochrane.org/mecir-manual.

[CD015824-bib-0020] 20.Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71. [DOI: 10.1136/bmj.n71] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD015824-bib-0021] 21.Oba Y, Goite Y, Maduke T, Fakhouri EW. Phosphodiesterase type 5 inhibitor plus endothelin receptor antagonist compared to either alone for group 1 pulmonary arterial hypertension. Cochrane Database of Systematic Reviews 2024, Issue 1. Art. No: CD015824. [DOI: 10.1002/14651858.CD015824] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD015824-bib-0022] 22.Moutchia J, McClelland RL, Al-Naamani N, Appleby DH, Blank K, Grinnan D, et al. Minimal clinically important difference in the 6-minute walk distance for patients with pulmonary arterial hypertension. American Journal of Respiratory and Critical Care Medicine 2023;207(8):1070-9. [DOI: 10.1164/rccm.202208-1547OC] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD015824-bib-0023] 23.Khair RM, Nwaneri C, Damico RL, Kolb T, Hassoun PM, Mathai SC. The minimal important difference in Borg Dyspnea Score in pulmonary arterial hypertension. Annals of the American Thoracic Society 2016;13(6):842-9. [DOI: 10.1513/AnnalsATS.201512-824OC] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD015824-bib-0024] 24.Lefebvre C, Glanville J, Briscoe S, Featherstone R, Littlewood A, Marshall C, et al. Technical Supplement to Chapter 4: Searching for and selecting studies. In: Higgins JP, Thomas J, Chandler J, Cumpston MS, Li T, Page MJ, Welch VA, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 6.3 (updated February 2022). Cochrane, 2022. Available from https://www.cochrane.org/authors/handbooks-and-manuals/handbook/archive/v6.3.

[CD015824-bib-0025] 25.Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, Ioannidis JP, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. Public Library of Science Medicine 2009;6(7):e1000100. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD015824-bib-0026] 26.Hernán MA, Robins JM. Per-protocol analyses of pragmatic trials. New England Journal of Medicine 2017;377:1391-8. [DOI: 10.1056/NEJMsm1605385] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0027] 27.Review Manager (RevMan). Version 7.12.0. The Cochrane Collaboration, 2024. Available at https://revman.cochrane.org.

[CD015824-bib-0028] 28.Higgins JP, Li T, Deeks JJ. Chapter 6: Choosing effect measures and computing estimates of effect. In: Higgins JP, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 6.3 (updated February 2022). Cochrane, 2022. Available from https://www.cochrane.org/authors/handbooks-and-manuals/handbook/archive/v6.3.

[CD015824-bib-0029] 29.Higgins JP, Savović J, Page MJ, Elbers RG, Sterne JA. Chapter 8: Assessing risk of bias in a randomized trial. In: Higgins JP, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 6.2 (updated February 2021). Cochrane, 2021. Available from https://www.cochrane.org/authors/handbooks-and-manuals/handbook/archive/v6.2.

[CD015824-bib-0030] 30.Sterne JA, Savović J, Page MJ, Elbers RG, Blencowe NS, Boutron I, et al. RoB 2: a revised tool for assessing risk of bias in randomised trials. BMJ 2019;366:l4898. [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0031] 31.RoB 2 tool. Available at https://www.riskofbias.info/welcome/rob-2-0-tool (accessed prior to 17 July 2025).

[CD015824-bib-0032] 32.Nüesch E, Trelle S, Reichenbach S, Rutjes AW, Tschannen B, Altman DG, et al. Small study effects in meta-analyses of osteoarthritis trials: meta-epidemiological study. BMJ 2010;342:c3515. [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD015824-bib-0033] 33.Dettori JR, Norvell DC, Chapman JR. Fixed-effect vs random-effects models for meta-analysis: 3 points to consider. Global Spine Journal 2022;12(7):1624-6. [DOI: 10.1177/21925682221110527] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD015824-bib-0034] 34.McKenzie JE, Brennan SE. Chapter 12: Synthesizing and presenting findings using other methods. In: Higgins JP, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 6.4 (updated August 2023). Cochrane, 2023. Available from https://www.cochrane.org/authors/handbooks-and-manuals/handbook/archive/v6.4.

[CD015824-bib-0035] 35.Guyatt G, Zhao Y, Mayer M, Briel M, Mustafa R, Izcovich A, et al. GRADE guidance 36: updates to GRADE's approach to addressing inconsistency. Journal of Clinical Epidemiology 2023;158:70-83. [DOI: 10.1016/j.jclinepi.2023.03.003] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0036] 36.Deeks JJ, Higgins JP, Altman DG. Chapter 10: Analysing data and undertaking meta-analyses. In: Higgins JP, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 6.4 (updated August 2023). Cochrane, 2023. Available from https://www.cochrane.org/authors/handbooks-and-manuals/handbook/archive/v6.4.

[CD015824-bib-0037] 37.Bell ML, Kenward MG, Fairclough DL, Horton NJ. Differential dropout and bias in randomised controlled trials: when it matters and when it may not. BMJ 2013;346:e8668. [DOI: 10.1136/bmj.e8668] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD015824-bib-0038] 38.Guyatt G, Oxman AD, Akl EA, Kunz R, Vist G, Brozek J, et al. GRADE guidelines: 1. Introduction – GRADE evidence profiles and summary of findings tables. Journal of Clinical Epidemiology 2011;64(4):383-94. [DOI: 10.1016/j.jclinepi.2010.04.026] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0039] 39.Zeng L, Brignardello-Petersen R, Hultcrantz M, Mustafa RA, Murad MH, Iorio A, et al. GRADE Guidance 34: update on rating imprecision using a minimally contextualized approach. Journal of Clinical Epidemiology 2022;150:216-24. [DOI: 10.1016/j.jclinepi.2022.07.014] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0040] 40.Schünemann HJ, Higgins JP, Vist GE, Glasziou P, Akl EA, Skoetz N, et al. Chapter 14: Completing 'Summary of findings' tables and grading the certainty of the evidence. In: Higgins JP, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 6.3 (updated February 2022). Cochrane, 2022. Available from https://www.cochrane.org/authors/handbooks-and-manuals/handbook/archive/v6.3.

[CD015824-bib-0041] 41.GRADEpro GDT. Version accessed 15 March 2024. Hamilton (ON): McMaster University (developed by Evidence Prime), 2024. Available at https://www.gradepro.org.

[CD015824-bib-0042] 42.Pulido T, Adzerikho I, Channick RN, Delcroix M, Galiè N, Ghofrani HA, et al. Macitentan and morbidity and mortality in pulmonary arterial hypertension. New England Journal of Medicine 2013;369(9):809-18. [DOI: 10.1056/NEJMoa1213917] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0043] 43.Zhuang Y, Jiang B, Gao H, Zhao W. Randomized study of adding tadalafil to existing ambrisentan in pulmonary arterial hypertension. Hypertension Research 2014;37(6):507-12. [DOI: 10.1038/hr.2014.28] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0044] 44.Mohammed S, Vijayvergiya R, Malhotra S, Rohit MK. A randomized, double-blind, placebo-controlled study to evaluate sildenafil, ambrisentan combination therapy in pulmonary hypertension, particularly of Eisenmenger syndrome. Indian Heart Journal 2021;73(5):633-6. [DOI: 10.1016/j.ihj.2021.07.007] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD015824-bib-0045] 45.McLaughlin V, Channick RN, Ghofrani HA, Lemarié JC, Naeije R, Packer M, et al. Bosentan added to sildenafil therapy in patients with pulmonary arterial hypertension. European Respiratory Journal 2015;46(2):405-13. [DOI: 10.1183/13993003.02044-2014] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0046] 46.Marshall J, Noel-Storr AH, Kuiper J, Thomas J, Wallace BC. Machine learning for identifying randomized controlled trials: an evaluation and practitioner's guide. Research Synthesis Methods 2018;9(4):602-14. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD015824-bib-0047] 47.Wilkins MR, Paul GA, Strange JW, Tunariu N, Gin-Sing W, Banya WA, et al. Sildenafil versus endothelin receptor antagonist for pulmonary hypertension (SERAPH) study. American Journal of Respiratory and Critical Care Medicine 2005;171(11):1292-7. [DOI: 10.1164/rccm.200410-1411OC] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0048] 48.Hassoun PM. Pulmonary arterial hypertension. New England Journal of Medicine 2021;385(25):2361-76. [DOI: 10.1056/NEJMra2000348] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0049] 49.Frost AE, Badesch DB, Miller DP, Benza RL, Meltzer LA, McGoon MD. Evaluation of the predictive value of a clinical worsening definition using 2-year outcomes in patients with pulmonary arterial hypertension: a REVEAL Registry analysis. Chest 2013;144(5):1521-9. [DOI: 10.1378/chest.12-3023] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0050] 50.Burgess G, Hoogkamer H, Collings L, Dingemanse J. Mutual pharmacokinetic interactions between steady-state bosentan and sildenafil. European Journal of Clinical Pharmacology 2008;64(1):43-50. [DOI: 10.1007/s00228-007-0408-z] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0051] 51.Paul GA, Gibbs JS, Boobis AR, Abbas A, Wilkins MR. Bosentan decreases the plasma concentration of sildenafil when coprescribed in pulmonary hypertension. British Journal of Clinical Pharmacology 2005;60(1):107-12. [DOI: 10.1111/j.1365-2125.2005.02383.x] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD015824-bib-0052] 52.Vizza CD, Jansa P, Teal S, Dombi T, Zhou D. Sildenafil dosed concomitantly with bosentan for adult pulmonary arterial hypertension in a randomized controlled trial. BioMed Central Cardiovascular Disorders 2017;17(1):239. [DOI: 10.1186/s12872-017-0674-3] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD015824-bib-0053] 53.Kirtania L, Maiti R, Srinivasan A, Mishra A. Effect of combination therapy of endothelin receptor antagonist and phosphodiesterase-5 inhibitor on clinical outcome and pulmonary haemodynamics in patients with pulmonary arterial hypertension: a meta-analysis. Clinical Drug Investigation 2019;39(11):1031-44. [DOI: 10.1007/s40261-019-00841-1] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0054] 54.Liu HL, Chen XY, Li JR, Su SW, Ding T, Shi CX, et al. Efficacy and safety of pulmonary arterial hypertension-specific therapy in pulmonary arterial hypertension: a meta-analysis of randomized controlled trials. Chest 2016;150(2):353-66. [DOI: 10.1016/j.chest.2016.03.031] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0055] 55.Tan Z, Wu PY, Zhu TT, Su W, Fang ZF. Efficacy and safety of sequential combination therapy for pulmonary arterial hypertension: a meta-analysis of randomized-controlled trials. Pulmonary Pharmacology & Therapeutics 2022;76:102144. [DOI: 10.1016/j.pupt.2022.102144] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0056] 56.Alonso-Coello P, Oxman AD, Moberg J, Brignardello-Petersen R, Akl EA, Davoli M, et al. GRADE Evidence to Decision (EtD) frameworks: a systematic and transparent approach to making well informed healthcare choices. 2: Clinical practice guidelines. BMJ 2016;353:i2089. [DOI: 10.1136/bmj.i2089] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0057] 57.Humbert M, Lau EM. Point: should initial combination therapy be the standard of care in pulmonary arterial hypertension? Yes. Chest 2019;156(6):1039-42. [DOI: 10.1016/j.chest.2019.07.005] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0058] 58.Channick RN, Saggar R. Counterpoint: should initial combination therapy be the standard of care in pulmonary arterial hypertension? No. Chest 2019;156(6):1043-5. [DOI: 10.1016/j.chest.2019.07.004] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0059] 59.Medrek SK, Sahay S. Ethnicity in pulmonary arterial hypertension: possibilities for novel phenotypes in the age of personalized medicine. Chest 2018;153(2):310-20. [DOI: 10.1016/j.chest.2017.08.1159] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0060] 60.Pizzicato LN, Nadipelli VR, Governor S, Mao J, Lanes S, Butler J, et al. Real-world treatment patterns, healthcare resource utilization, and cost among adults with pulmonary arterial hypertension in the United States. Pulmonary Circulation 2022;12(2):e12090. [DOI: 10.1002/pul2.12090] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD015824-bib-0061] 61.Coyle K, Coyle D, Blouin J, Lee K, Jabr MF, Tran K, et al. Cost effectiveness of first-line oral therapies for pulmonary arterial hypertension: a modelling study. PharmacoEconomics 2016;34(5):509-20. [DOI: 10.1007/s40273-015-0366-8] [PMID: ] [DOI] [PMC free article] [PubMed] [Google Scholar]

[CD015824-bib-0062] 62.Barst RJ, Oudiz RJ, Beardsworth A, Brundage BH, Simonneau G, Ghofrani HA, et al. Pulmonary arterial hypertension and response to tadalafil (PHIRST) study group. Tadalafil monotherapy and as add-on to background bosentan in patients with pulmonary arterial hypertension. Journal of Heart and Lung Transplantation 2011;30(6):632-43. [DOI: 10.1016/j.healun.2010.11.009] [PMID: ] [DOI] [PubMed] [Google Scholar]

[CD015824-bib-0063] 63.Grünig E, Jansa P, Fan F, Hauser JA, Pannaux M, Morganti A, et al. Randomized trial of macitentan/tadalafil single-tablet combination therapy for pulmonary arterial hypertension. Journal of the American College of Cardiology 2024;83(4):473-84. [DOI: 10.1016/j.jacc.2023.10.045] [PMID: ] [DOI] [PubMed] [Google Scholar]

PERMALINK

Phosphodiesterase type 5 inhibitor plus endothelin receptor antagonist compared to either alone for group 1 pulmonary arterial hypertension

Yuji Oba

Tinashe Maduke

Eddie W Fakhouri

Yohannes Goite

Abstract

Rationale

Objectives

Search methods

Eligibility criteria

Outcomes

Risk of bias

Synthesis methods

Included studies

Synthesis of results

Authors' conclusions

Funding

Registration

Plain language summary

Summary of findings

Summary of findings 1. Combination therapy versus endothelin receptor antagonist for group 1 pulmonary arterial hypertension.

Summary of findings 2. Combination therapy versus phosphodiesterase type 5 inhibitor for group 1 pulmonary arterial hypertension.

Summary of findings 3. Phosphodiesterase type 5 inhibitor versus endothelin receptor antagonist for group 1 pulmonary arterial hypertension.

Background

Description of the condition

Description of the intervention and how it might work

Why it is important to do this review

Objectives

Methods

Changes from protocol methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

1. Demographics of the studies included.

Outcome measures

Critical outcomes

Important outcomes

Search methods for identification of studies

Electronic searches

Searching other resources

Data collection and analysis

Selection of studies

Data extraction and management

Risk of bias assessment in included studies

Measures of treatment effect

Unit of analysis issues

Dealing with missing data

Reporting bias assessment

Synthesis methods

Meta‐analysis of numerical data

Synthesis using other methods

Investigation of heterogeneity and subgroup analysis

Equity‐related assessment

Sensitivity analysis

Certainty of the evidence assessment

Consumer involvement

Results

Description of studies

Results of the search

1.

Included studies

Demographic and clinical diversity in included studies: insights and gaps

Excluded studies

Risk of bias in included studies

2.

Synthesis of results

2. Definition of clinical worsening.

Combination therapy versus endothelin receptor antagonist

Clinical worsening

3.

Mortality

4.

Hospitalisation

5.

Change from baseline in six‐minute walk distance

6.

7.

Worsening in World Health Organization functional class