Study of risk assessment scores as surrogate endpoints in pulmonary arterial hypertension trials

RATIONALE: Pulmonary arterial hypertension (PAH) clinical trials largely rely on clinically relevant intermediate outcomes, as mortality and lung transplant are infrequent and remote outcomes. The change in 6-minute walk distance (Δ6MWD) is utilized as an intermediate endpoint and is used as a surrogate for clinical worsening events. OBJECTIVES: We investigate whether the change in 4 composite risk assessment scores-REVEAL Lite 2 (ΔRL2), European Four Strata (ΔE4), European Three Strata (ΔE3), and French Noninvasive score (ΔFPHN)-are potentially stronger surrogate endpoints than the single-component Δ6MWD in PAH studies. METHODS: Using data from the AMBITION and FREEDOM-EV trials, we estimated the percentage of treatment effect in reducing clinical worsening events within 1 year that is attributed to the change on each of the 5 scores between baseline and 12-16 weeks' follow-up. We conducted bootstrap analysis to provide head-to-head comparisons. MEASUREMENTS AND MAIN RESULTS: Our analysis shows that all 4 composite scores mediate significantly more of the treatment effect compared to 6MWD (29% for ΔRL2 and ΔE4 compared to 19% for Δ6MWD; both P <10-15). The drug treatment effect accounted for by all 5 scores increases with intermediate/high-risk patients, and ΔRL2 had the highest % mediation (37% for ΔRL2 vs 28% for Δ6MWD; P < 10-15). CONCLUSIONS: RL2 and E4 were the strongest mediators, with RL2 showing the highest mediation in intermediate/high-risk patients. Until better surrogates are found that mediate over 50% of the treatment effect, RL2 or E4 may serve as alternatives to the 6MWD as intermediate endpoints.