Introduction: Selecting the first oral agent for pulmonary arterial hypertension (PAH) often comes down to endothelin-receptor antagonists (ERAs) or phosphodiesterase-5 inhibitors (PDE5Is). Randomised trials have not directly compared the two classes, and real-world data remain limited.
Methods: We queried the TriNetX US Collaborative Network (2004-2023), comprising 70 health-care systems and >90 million patients. Adults carrying an ICD-10 code for PAH (I27.0) who started either ERA monotherapy (bosentan, ambrisentan, macitentan) or PDE5I monotherapy (sildenafil, tadalafil) with no prior prostacyclin or combination therapy were eligible. The primary outcome was all-cause mortality at 12 months. Secondary endpoints included three-point major adverse cardiac events (death, myocardial infarction, stroke), heart-failure hospitalisation, first syncope, escalation to parenteral prostacyclin, hepatotoxicity (ALT ≥ 3× baseline), and clinically significant oedema. Outcomes were analysed by Kaplan–Meier estimates and Cox models stratified by matched pairs; proportional-hazards assumptions were verified with Schoenfeld residuals.
Results: After matching, 975 patients remained in each cohort (median age 56 years; 69 % female). Baseline characteristics were well balanced (all standardised differences ≤0.065). At one year, mortality was 11.3 % with ERAs versus 12.1 % with PDE5Is (hazard ratio [HR] 0.92, 95 % CI 0.71-1.19; p = 0.57). Three-point MACE occurred in 12.2 % versus 13.5 % (HR 0.89, 95 % CI 0.69-1.16; p = 0.42). Heart-failure hospitalisations (13.4 % vs 14.1 %; HR 0.95, 95 % CI 0.69-1.31; p = 0.77) and syncope (2.5 % vs 2.1 %; HR 1.18, 95 % CI 0.64-2.16; p = 0.56) were likewise similar.Treatment escalation to parenteral prostacyclins was comparable. Safety outcomes showed similar hepatotoxicity and oedema rates in both groups
Conclusions: n this large multicentre, propensity-matched analysis of treatment-naive PAH, ERAs and PDE5Is produced indistinguishable 12-month survival, cardiovascular outcomes, and adverse-event rates. These findings support current guidelines that either class is an acceptable first-line option; choice may therefore be guided by individual comorbidity profiles, side-effect tolerance, and cost rather than expectations of differential efficacy.
