What's the (end) point?

Clinicians caring for children or adults with pulmonary arterial hypertension (PAH) face a number of challenges, not the least of which is determining whether a patient has met treatment targets based on changes in a given test (such as the 6-min walk distance (6MWD)) and whether these changes translate to meaningful improvements in that patient's quality of life or survival. While several treatment benchmarks (such as achieving a 6MWD >380–440 m) have been put forth in recent consensus guidelines, none of the proposed intermediates has been established as a true surrogate for disease outcomes in PAH [1, 2]. A correlate does not a surrogate make; a given end point needs to be validated before being used to inform treatment or prove therapeutic efficacy [3]. A number of criteria are necessary to establish surrogacy [3–5]. A valid surrogate should be reliable, preferably integral to the disease causal pathway, targeted by treatment, and should track with " hard " outcomes (such as survival) across multiple studies. Finally, therapy-induced changes in the end point should explain a significant proportion of the treatment effect (50–75%) in terms of the outcome of interest. PLOEGSTRA et al. [6] recently published a study in the European Respiratory Journal that showed that changes in functional class, N-terminal pro-brain natriuretic peptide, and tricuspid annular plane systolic excursion in patients receiving PAH treatment predicted transplant-free survival in a paediatric cohort. The authors propose that these measures be used as treatment goals in children with PAH. While we appreciate this important addition to the understudied area of paediatric PAH, the recommendation that these markers should serve as guideposts for goal-directed PAH therapy may be premature. Justification for this caution is evidenced by recent studies of the 6MWD and haemodynamics in adults, which have been shown in multiple observational studies to be associated with survival both at baseline and after therapy [7–10]. We and others have found that while changes in these measures were associated with outcomes in PAH clinical trials (as in prior observational studies), the drug-induced changes (compared with placebo) explained very little of the variances in outcomes [11–14]. At most, the changes in 6MWD and pulmonary vascular resistance respectively explained only 22% and 14% of the effect of treatment on clinical events at 12 weeks, falling short of the desired 50% mark and suggesting that they are not valid surrogates for short-term outcomes. These recent observations regarding some of …