Résumé : BACKGROUND: Because of the challenges associated with conducting large survival studies of patients with pulmonary arterial hypertension (PAH), we analyzed the surrogate markers predictive of long-term survival in a large cohort of patients treated with subcutaneous treprostinil. METHODS: A retrospective review was conducted using data from a total of 811 patients with New York Heart Association Functional Class (NYHA FC) II to IV PAH, who were treated with subcutaneous treprostinil. Patient baseline disease and on-treatment parameters were analyzed by uni- and multivariate analyses for predictive value of 3-year survival with PAH. RESULTS: Among the baseline disease-related factors analyzed, there was a significantly higher risk of death (p < 0.001) associated with connective tissue disease-associated PAH relative to idiopathic PAH (hazard ratio for death [HR] 1.93), NYHA FC IV vs III (HR 2.31), pulmonary vascular resistance index (PVRI) >30 vs ≤16 mm Hg/liter/min/m(2) (HR 2.44) and mixed venous oxygen saturation (SVO(2)) ≤55% vs >55%. The 6-minute walk distance (6MWD) of ≤295 m after 12 weeks of treprostinil treatment was associated with reduced survival at 3 years (58%). A ≥20-m increase from baseline in 6MWD was associated with greater survival (80%) vs smaller walk increments (69%; p = 0.039). Treprostinil dose of ≥40 ng/kg/min (p < 0.001) and every 10-ng/kg/min dose increase (p = 0.009) resulted in improved long-term survival. In a multivariate analysis, only SVO(2), 6MWD and treprostinil dose were significant on-treatment predictors (p < 0.02) of survival. CONCLUSIONS: Disease etiology, baseline factors (NYHA FC, PVRI and SVO(2)) and on-treatment factors (6MWD, SVO(2) and treprostinil dose) were predictors of survival in this study and may be used to aid in treatment optimization.