Parenteral prostanoids remain the cornerstone of therapy for advanced pulmonary arterial hypertension (PAH). The oral prostanoid receptor selexipag held promise to provide the clinical benefits of this class while avoiding the need for an infusion. Two randomized clinical trials form the basis of guideline recommendations on the use of selexipag in PAH. The GRIPHON trial randomized 1156 patients to selexipag or placebo irrespective of nonprostanoid background therapy and demonstrated a 40% reduction in morbidity and mortality events over a mean of 14 months and a slight increase in walk distance. In contrast, selexipag did not provide therapeutic value when used as part of an up-front triple therapy in conjunction with an endothelin receptor antagonist (ERA) and a phosphodiesterase type 5 inhibitor (PDE5) inhibitor in the TRITON trial. In 2023, a post hoc pooled analysis of patients with incident PAH (within 6 months of diagnosis) who participated in the initial GRIPHON and TRITON trials was completed. In 329 patients who received selexipag, there was a reduction in the risk of disease progression (hazard ratio, 0.48; 95% CI, 0.3-0.56) compared with 320 patients in the control group.
Despite available data and guideline recommendations, the literature suggests that this agent is not routinely prescribed and, relative to other PAH therapies, is more frequently discontinued. Selexipag has a more challenging adverse effect profile, with higher rates of intolerance, and requires dose escalation, making it more cumbersome for patients and clinicians. Dose escalation is balanced against adverse effects, further limiting uptake by patients and clinicians who perceive that patients may have made some initial therapeutic gains with an endothelin receptor antagonist and PDE5 inhibitor. It is also more nuanced to convince patients that while they may not feel better (owing to adverse effects and relatively smaller gains in functional capacity), the data suggest that they will have less risk of disease progression. In the absence of a new prospective trial, a rigorous attempt to address the merits of adding selexipag using clinical data is of potential interest to clinicians who treat these patients.
In that vein, Burger et al used data from the Komodo Health payer dataset representing more than 150 private payers in the US to evaluate the association of selexipag with morbid events in patients with PAH recently prescribed a PDE5 inhibitor and an endothelin antagonist. The authors identified 2966 patients who met eligibility criteria. Using this dataset, the authors attempted to control for available covariates for outcome through inverse probability censoring weights. Additionally, patients who had an event were replicated in both groups prior to initiating selexipag using baseline data from all patients not treated with up-front selexipag as though both populations had a chance of being treated with selexipag at some time during the observation period. Using this approach, the authors found that 351 patients (11.8%) had added oral selexipag and 2615 patients had no exposure to selexipag.
The authors reported that early exposure to selexipag was associated decreased risk for all-cause hospitalization, PAH-related hospitalization, and PAH-related disease progression (initiation of parenteral therapies, PAH-related hospitalization, death, transplant, or septostomy). Similar to results in the post hoc analysis of the GRIPHON trial, there was an advantage to earlier exposure to selexipag (earlier than 6 months), with less benefit seen when selexipag was added after 12 months. Findings seemed stable using different sensitivity analyses with respect to duration of exposure to baseline treatments with endothelin receptor antagonists, PDE5 inhibitors, or both.
Clinical data are messy, sometimes misleading, and often incomplete. For example, clinical features that are incorporated into a risk assessment in patients with PAH were not available in this study. Components of these risk tools include assessments of functional capacity (walk distance and New York Heart Association classification), symptoms, and laboratory data (eg, N-terminal pro-B-type natriuretic peptide and echocardiogram data). These risk tools are established markers of prognosis and are used to evaluate patients who are at risk of worsening and to inform decision-making, such as addition of new therapies or referral for lung transplant. Only 37% of patients reported to have PAH in the study cohort were registered as having dyspnea, the most common symptom in PAH, highlighting shortcomings of the dataset. In the absence of access to this information, important prognostic variables and covariates may have been missed. Patients who were prescribed selexipag early may also have been treated in more expert centers. It is also unclear if patients who did not receive selexipag had any other sociodemographic factors (eg, income or race and ethnicity categories) that may have been associated with time to clinical worsening and hence influenced results. Although the authors cite these drawbacks in their discussion, they remain important limitations of this study.
It is also concerning that only 36% of patients were recorded as having a cardiac catheterization in the 6 months prior to diagnosis. Right heart catheterization remains the cornerstone of correctly diagnosing PAH. This raises the risk that patients in the study cohort may have had other forms of pulmonary hypertension or were patients with prevalent PAH who simply changed insurance companies.
Patients receiving double combination therapy had a mean of 3.13 comorbid conditions (eg, diabetes, hypertension, chronic obstructive pulmonary disease, coronary disease, heart failure, atrial fibrillation, or obesity); an identical frequency was observed in patients given triple therapy. This suggests that selexipag was not preferentially administered to patients without comorbidities, which would have been expected given the evidence base. We are therefore left with uncertainty about variables that may have been associated with therapeutic decisions or access to therapies and the nature of the population being treated. The main reassurance is that patients who were sicker would likely have been exposed to selexipag more frequently and would also have been at greater risk for morbid events. This is directionally opposite to the study findings.
The study cohort represented an older group of patients with a higher number of comorbidities than those enlisted in pivotal selexipag trials. Clinical data provided by Berger et al suggest that the use of PAH therapies may be more liberal and may not be aligned with the population in which the benefit of PAH-directed therapy has been demonstrated in clinical trials. Data from Burger et al are consistent with earlier studies that suggest the benefit of selexipag may be time dependent, being most effective if introduced early, ideally within 6 months of initiating therapy with an ERA and a PDE5 inhibitor. However, the reliability of the observations by Burger et al is influenced by missing data that would normally be used to inform prognosis and clinical decision-making. As clinicians treating patients with PAH, we are also left with addressing the therapeutic elephant in the room: specifically, how sotatercept will be incorporated into practice relative to other add-on therapies.
Corresponding Author: John Granton, MD, Division of Respirology, Faculty of Medicine, Toronto General Hospital, University of Toronto, 585 University Ave, Toronto, ON M5G 2N2 (john.granton@uhn.ca).
Conflict of Interest Disclosures: Dr Granton reported receiving grants from Janssen outside the submitted work. Dr Coghlan reported receiving grants and personal fees from Janssen outside the submitted work.