Risk assessment in pulmonary arterial hypertension: Insights from the GRIPHON study

Science and Research

BACKGROUND: Approaches to risk assessment in pulmonary arterial hypertension (PAH) include the noninvasive French risk assessment approach (number of low-risk criteria based on the European Society of Cardiology and European Respiratory Society guidelines) and Registry to Evaluate Early and Long-term PAH Disease Management (REVEAL) 2.0 risk calculator. The prognostic and predictive value of these methods for morbidity/mortality was evaluated in the predominantly prevalent population of GRIPHON, the largest randomized controlled trial in PAH. METHODS: GRIPHON randomized 1,156 patients with PAH to selexipag or placebo. Post-hoc analyses were performed on the primary composite end-point of morbidity/mortality by the number of low-risk criteria (World Health Organization functional class I-II; 6-minute walk distance >440 m; N-terminal pro-brain natriuretic peptide <300 ng/liter) and REVEAL 2.0 risk category. Hazard ratios and 95% confidence intervals were calculated using Cox proportional hazard models. RESULTS: Both the number of low-risk criteria and the REVEAL 2.0 risk category were prognostic for morbidity/mortality at baseline and any time-point during the study. Patients with 3 low-risk criteria at baseline had a 94% reduced risk of morbidity/mortality compared to patients with 0 low-risk criteria and were all categorized as low-risk by REVEAL 2.0. The treatment effect of selexipag on morbidity/mortality was consistent irrespective of the number of low-risk criteria or the REVEAL 2.0 risk category at any time-point during the study. Selexipag-treated patients were more likely to increase their number of low-risk criteria from baseline to week 26 than placebo-treated patients (odds ratio 1.69, p=0.0002); similar results were observed for REVEAL 2.0 risk score. CONCLUSIONS: These results support the association between risk profile and long-term outcome and suggest that selexipag treatment may improve risk profile.