Designing a randomized controlled trial (RCT) to investigate a survival benefit in
a rare disease such as pulmonary arterial hypertension (PAH) has considerable logistical
and ethical constraints. In the SERAPHIN RCT, a 23% non-significant reduction in the
risk of all-cause mortality was observed with macitentan 10 mg vs. placebo. As SERAPHIN
enrolled patients in the same time frame as the US REVEAL registry, a prediction model
based on REVEAL data was used to further explore the effect of macitentan on mortality.
From REVEAL (N = 3515), 734 patients who would have met SERAPHIN eligibility criteria
were selected (REVEAL analysis cohort [RAC]). Using the RAC, a prediction model for
time to death up to 3 years was constructed based on ten baseline prognostic variables.
The model was used to predict survival of each of the 742 SERAPHIN patients had they
received real-world treatment in the RAC. The average temporal profile of these patients
was then compared with the observed survival of the macitentan 10 mg group (n = 242)
using a log-rank test and Cox's proportional hazard model. Over 3 years, the risk
of mortality observed with macitentan 10 mg was 35% lower than that predicted (p =
0.01) (Fig). Real-world observational data can complement RCT data to provide a means
of evaluating survival benefits in rare diseases. Merits and limitations of this approach
will be discussed.
Fig. 1:
Survival comparison between macitentan 10 mg and predicted real-world treatment for
the SERAPHIN population
