Résumé : BACKGROUND: The purpose of this study was to compare patients with pulmonary arterial hypertension enrolled in the AMBITION trial with (excluded from the primary analysis set [ex-primary analysis set]) and without (primary analysis set) multiple risk factors for left ventricular diastolic dysfunction. METHODS: Treatment-naive patients with pulmonary arterial hypertension were randomized to once-daily ambrisentan and tadalafil combination therapy, ambrisentan monotherapy, or tadalafil monotherapy. The primary end point was time from randomization to first adjudicated clinical failure event. RESULTS: Primary analysis set patients (n = 500), compared with ex-primary analysis set patients (n = 105), were younger (mean, 54.4 vs 62.1 years) with greater baseline 6-minute walk distance (median, 363.7 vs 330.5 meters) and fewer comorbidities (e.g., hypertension and diabetes). Treatment effects of initial combination therapy vs pooled monotherapy were directionally the same for both populations, albeit of a lower magnitude for ex-primary analysis set patients. Initial combination therapy reduced the risk of clinical failure compared with pooled monotherapy in primary analysis set patients (hazard ratio, 0.50; 95% confidence interval, 0.35–0.72), whereas the effect was less clear in ex-primary analysis set patients (hazard ratio, 0.70; 95% confidence interval, 0.35–1.37). Overall, primary analysis set patients had fewer clinical failure events (25% vs 33%), higher rates of satisfactory clinical response (34% vs 24%), and lower rates of permanent study drug withdrawal due to adverse events (16% vs 31%) than ex-primary analysis set patients. CONCLUSIONS: Efficacy of initial combination therapy vs pooled monotherapy was directionally similar for primary analysis set and ex-primary analysis set patients. However, ex-primary analysis set patients (with multiple risk factors for left ventricular diastolic dysfunction) experienced higher rates of clinical failure events and the response to combination therapy vs monotherapy was attenuated. Tolerability was better in primary analysis set than ex-primary analysis set patients.