
==== Front
JAMA Netw Open
JAMA Netw Open
JAMA Network Open
2574-3805
American Medical Association

39312239
10.1001/jamanetworkopen.2024.34691
zoi241028
Research
Original Investigation
Online Only
Pulmonary Medicine
Early Addition of Selexipag to Double Therapy for Pulmonary Arterial Hypertension
Selexipag Plus Double Therapy for Pulmonary Arterial Hypertension
Selexipag Plus Double Therapy for Pulmonary Arterial Hypertension
Burger Charles D. MD 1
Tang Wenze PhD 2
Tsang Yuen PharmD 2
Panjabi Sumeet PhD 2
1 Division of Pulmonary, Allergy and Sleep Medicine, Department of Internal Medicine, Mayo Clinic, Jacksonville, Florida
2 Actelion Pharmaceuticals US, Inc, a Johnson & Johnson Company, Titusville, New Jersey
Article Information

Accepted for Publication: July 15, 2024.

Published: September 23, 2024. doi:10.1001/jamanetworkopen.2024.34691

Open Access: This is an open access article distributed under the terms of the CC-BY-NC-ND License. © 2024 Burger CD et al. JAMA Network Open.

Corresponding Author: Charles D. Burger, MD, Division of Pulmonary, Allergy and Sleep Medicine, Department of Internal Medicine, Mayo Clinic, 4500 San Pablo Rd, Jacksonville, FL 32224 (burger.charles@mayo.edu).
Author Contributions: Drs Tang and Tsang had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Tang, Tsang, Panjabi.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Tang, Tsang, Panjabi.

Critical review of the manuscript for important intellectual content: All authors.

Statistical analysis: Tang, Tsang, Panjabi.

Obtained funding: Panjabi.

Administrative, technical, or material support: Tsang, Panjabi.

Supervision: Burger, Tsang, Panjabi.

Conflict of Interest Disclosures: Dr Burger reported receiving personal fees from Janssen, Insmed, Merck, and Gossamer Bio during the conduct of the study and grants from Merck, Insmed, Gossamer Bio, Liquidia, and Aerovate Therapeutics outside the submitted work. Dr Tang reported owning shares of Johnson & Johnson stock during the conduct of the study. Dr Tsang reported owning Johnson & Johnson stock outside the submitted work. Dr Panjabi reported owning equity in Johnson & Johnson during the conduct of the study.

Funding/Support: Funding was provided by Actelion Pharmaceuticals US, Inc, a Johnson & Johnson Company.

Role of the Funder/Sponsor: Actelion Pharmaceuticals US, Inc, had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. Drs Tang, Tsang, and Panjabi, who are employees of Actelion Pharmaceuticals US, Inc, participated in each of these activities. Actelion Pharmaceuticals US, Inc, reviewed and approved the study protocol prior to funding.

Data Sharing Statement: See Supplement 2.

Additional Contributions: We would like to thank Kalla Sun, PhD (Actelion Pharmaceuticals US, Inc, a Johnson & Johnson Company), for providing programming support, who was not compensated for this work outside of her regular salary. Medical writing support was provided by Kathryn Quinn, PhD, and Ify Sargeant, DPhil, on behalf of Twist Medical, a specialist medical communications company, which was compensated by Janssen Scientific Affairs, LLC.

23 9 2024
9 2024
23 9 2024
7 9 e243469119 12 2023
15 7 2024
Copyright 2024 Burger CD et al. JAMA Network Open.
https://creativecommons.org/licenses/by-nc-nd/4.0/ This is an open access article distributed under the terms of the CC-BY-NC-ND License.
jamanetwopen-e2434691.pdf

This comparative effectiveness study investigates the association of selexipag addition to double therapy with hospitalization and disease progression in patients with pulmonary arterial hypertension.

Key Points

Question

Is early oral selexipag addition to double oral therapy (DOT) with an endothelin receptor antagonist plus a phosphodiesterase type 5 inhibitor in clinical practice associated with lower risk of hospitalization and disease progression in patients with pulmonary arterial hypertension (PAH)?

Findings

This comparative effectiveness study emulating a hypothetical trial among 2966 commercially insured patients with PAH found lower risks of all-cause hospitalization, PAH-related hospitalization, and PAH-related disease progression when adding selexipag within 6 months or less of initiating DOT vs DOT alone. Estimated risks were even lower when selexipag was added at 3 months or less.

Meaning

This study found that adding selexipag early to DOT was associated with reduced risk of hospitalization and disease progression in patients with PAH.

Importance

A subgroup analysis of a randomized clinical trial established the efficacy of selexipag plus background therapy (monotherapy or double oral therapy [DOT]) vs placebo plus background therapy and found that the addition of selexipag within 6 months had an added benefit. However, the timing of selexipag addition to DOT and the incremental benefit in clinical practice is not well studied.

Objective

To compare triple oral therapy (TOT) consisting of selexipag, endothelin receptor antagonist (ERA), and phosphodiesterase type 5 inhibitor (PDE5i) vs DOT consisting of ERA and PDE5i.

Design, Setting, and Participants

This comparative effectiveness study was conducted using data from the US Komodo claims database to emulate a randomized trial. Patients aged 18 years or older with pulmonary arterial hypertension (PAH) treated with ERA plus PDE5i with records from July 2015 through June 2022 were duplicated to TOT and DOT and artificially censored when observed treatment deviated from assigned treatment. Hypothetical randomization was emulated using inverse probability of treatment weighting, and the study accounted for censoring-induced selection bias using inverse probability of censoring weighting. A pooled logistic model estimated the per-protocol difference between treatment groups. Data were analyzed from November 2022 through July 2023.

Interventions

TOT (addition of selexipag within 3, 6, and 12 months of initiating DOT) vs DOT.

Main Outcomes and Measures

Adjusted risk of all-cause hospitalization, PAH-related hospitalization, and PAH-related disease progression over a 2-year follow-up.

Results

A total of 2966 patients with PAH (mean [SD] age, 54.3 [14.0] years; 2125 female [71.6%]) met eligibility criteria. Adding selexipag within 6 months of ongoing DOT was associated with a reduction in risk for all-cause hospitalization (adjusted hazard ratio [aHR], 0.82; 95% CI, 0.72-0.94), PAH-related hospitalization (aHR, 0.81; 95% CI, 0.70-0.95), and PAH-related progression (aHR, 0.82; 95% CI, 0.70-0.95) vs DOT alone. There were no associations if selexipag was initiated within 12 months for all-cause hospitalization, PAH-related hospitalization, or PAH-related disease progression. The association remained with a greater decrease in risk for disease progression vs DOT for selexipag initiation within 3 months (aHR, 0.74; 95% CI, 0.61-0.90).

Conclusions and Relevance

This study found that early selexipag addition to ERA plus PDE5i was associated with a reduction in risk of hospitalization and disease progression. These findings suggest that delays in selexipag initiation likely contribute to suboptimal patient and health system outcomes.
==== Body
pmcIntroduction

Pulmonary arterial hypertension (PAH) is a rare, life-threatening disease characterized by pulmonary vascular remodeling to increase pulmonary-artery pressure and, ultimately, right heart failure and premature death. Up-front double oral therapy (DOT) with an endothelin receptor antagonist (ERA) and a phosphodiesterase type 5 inhibitor (PDE5i) is considered the standard-of-care treatment for patients at low and intermediate risk.1

Oral selexipag, a selective prostacyclin receptor agonist approved by the Food and Drug Administration in December 2016, should be added as a triple oral therapy (TOT) in patients who deteriorate to greater risk of death while receiving DOT with ERA and PDE5i.1 This escalation strategy requires regular risk assessment of 1-year mortality during follow-up visits (eg, 3-6 months after DOT initiation). However, the use of risk-assessment tools in clinical practice is infrequent.2 More practical guidance on the timing of the addition of oral selexipag is warranted.

Evidence of the treatment effect of TOT with oral selexipag vs DOT among patients with PAH has been generated mostly from randomized clinical trials and focuses on clinical outcomes.3 In a post hoc analysis of the pivotal GRIPHON trial, treatment with added oral selexipag vs existing treatment with ERA, PDE5i, or both significantly lowered the risk of PAH-associated disease progression and hospitalization.4 Exploratory analyses of the TRITON study showed a potential signal for risk reduction in disease progression with up-front TOT vs up-front DOT, with a median follow-up of 77.6 and 75.8 weeks in initial TOT and DOT groups, respectively.5 Early results from the ongoing EXPOSURE study showed a 45% reduction in risk of death with selexipag compared with other PAH therapies.6 In a 2023 pooled analysis of 649 newly diagnosed (≤6 months) GRIPHON and TRITON patients, early addition of selexipag (<6 months) was associated with a 52% reduction in the risk of disease progression (hazard ratio [HR], 0.48; 95% CI, 0.35-0.66).7 However, patients enrolled in randomized clinical trials are often younger and healthier compared with patients seen in clinical practice and may not represent the general population of patients with PAH, who tend to be older and have higher comorbidity levels.4 To our knowledge, the effectiveness of TOT vs DOT in this more heterogeneous group of patients has been assessed only in a previous claims data study,8 which found that selexipag initiation within 12 months of PAH diagnosis was associated with a lower risk for all-cause hospitalizations compared with not receiving a prostacyclin pathway agent, but there was no difference in PAH-related hospitalizations or risk of disease progression. However, the study used a landmark design that can introduce self-inflicted time-related bias and did not account for time-varying confounding.9 TOT with selexipag is often initiated sequentially after DOT that consists of ERA and PDE5i. Because of this, a conventional new user study design identifying 2 cohorts of patients receiving up-front TOT and DOT cannot be used to study the effectiveness of such a dynamic treatment strategy, in which a treatment decision to add additional therapy is based on deterioration of patient disease severity and other evolving patient characteristics.10 In this study, we used a fit-for-purpose target trial framework with data from a large claims database in the US to study the comparative effectiveness of TOT that allows the delayed addition of oral selexipag vs DOT consisting of ERA and PDE5i.11,12

Methods

Study Design

In this comparative effectiveness study, we emulated a hypothetical trial in which patients with PAH stabilized on treatment with DOT consisting of ERA and PDE5i for 60 days could be randomized to 1 of 2 groups: a TOT group in which oral selexipag could be added at any time within the first 6 months after randomization or a placebo group in which patients were continuously treated with DOT only. Time zero was defined as the 61st day of patient use of ERA and PDE5i. The target trial was emulated using closed medical and outpatient, pharmacy-dispensed claims between January 1, 2015, and December 31, 2022, from the Komodo Health payer-complete dataset derived from more than 150 private insurers in the US (commercial, individual, state exchange purchased, Medicare Advantage, and Medicaid managed care). Patient information used for analysis was deidentified to comply with the Health Insurance Portability and Accountability Act, and the New England Institutional Review Board (IRB) determined that the study and use of the dataset did not constitute research involving human participants and therefore was exempt from IRB oversight and informed consent. We followed the International Society for Pharmacoeconomics and Outcomes Research (ISPOR) reporting guideline and Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

Eligibility Criteria

Our trial emulation included patients who used ERA and PDE5i for 60 or more days between July 1, 2015, and June 30, 2022. To be eligible, patients also had to be continuously enrolled with medical and pharmacy benefits during the 6 months prior to time zero and have a confirmed diagnosis of pulmonary hypertension (PH) prior to time zero.13 Patients were excluded if at time zero they had used soluble guanylate cyclase stimulator or prostacyclin pathway agent classes of PAH medications within 6 months before time zero, had a prior diagnosis of chronic thromboembolic PH, were younger than age 18 years, or had received any PDE5i agent intended for erectile dysfunction where the ratio of pill count over days of supply for the PDE5i prescription was less than 1.14

Treatment Strategies and Artificial Censoring

Our primary treatment strategies of interest were TOT with the addition of oral selexipag within 6 months on and after time zero compared with DOT treatment only. The 2 treatment strategies were compared following the 3-stage approach (cloning, censoring, and weighting).10,11,15,16,17 First, all eligible patients were duplicated in the dataset, with each patient assigned to each treatment strategy (TOT or DOT). This is the key step to address the potential survival bias because for patients who never initiated selexipag or those who initiated selexipag after time zero, their treatment group assignment is determined solely based on information at time zero. To illustrate how patient groups were constructed and followed up, we classified all eligible patients into 4 types depending on when their selexipag was initiated (see the type 1 patient in Figure 1).

Figure 1. Examples of Cohort Assignment and Follow-Up Determination Under Target Trial Design

Patient types 1 to 4 are shown. DOT indicates double oral therapy; ERA, endothelin receptor antagonist; PDE5i, phosphodiesterase type 5 inhibitor; TOT, triple oral therapy.

Second, patient copies were artificially censored when their observed treatment trajectory deviated from the assigned treatment strategy (see the type 2a patient in Figure 1). However, the same patients would be artificially censored at the start of selexipag initiation as the observed treatment trajectory deviated from the DOT strategy (see the type 2b patient in Figure 1). By the same logic, patients initiating selexipag after the sixth month, such as at the eighth month, would be artificially censored at the end of the sixth month under the TOT strategy (see the type 3a patient in Figure 1) but would be followed up until the start of the eighth month under the DOT strategy (see the type 3b patient in Figure 1). Patients who never initiated oral selexipag would not be artificially censored under the DOT strategy (see the type 4b patient in Figure 1) but would be censored at the end of the sixth month under the TOT strategy (see the type 4a patient in Figure 1). Lastly, to control for the selection bias introduced by informative censoring, inverse probability of censoring weighting (IPCW) was used to adjust for prognostic factors of censoring (see further explanation in subsequent section).18

Follow-Up and Outcomes

We followed up each patient until the occurrence of the outcome of interest (separate analysis for each outcome), insurance plan disenrollment, or data cutoff or a maximum of 2 years, whichever happened first. The primary study end point was all-cause hospitalization. Secondary study end points included PAH-related hospitalization, identified using claims with PH diagnosis at any position, and PAH-related disease progression, which is a composite end point defined as the first instance of parenteral therapy of epoprostenol or treprostinil (eTable 1 in Supplement 1), PAH-related hospitalization, lung transplant or atrial septostomy (eTable 2 in Supplement 1), or all-cause death.19 Deaths in the Komodo dataset were derived from 1 of the following sources: the Death Master File maintained by the Social Security Office, Center for Medicare & Medicaid Services, obituary data, or claims with a discharge status of “expired.”

Patients were additionally censored if they discontinued the assigned treatment. Discontinuation was defined as a gap in PAH medication claims of more than 90 days in any component of the treatment strategy or as the addition of new PAH medication that was not part of the treatment strategy.20,21,22,23

Statistical Analysis

The primary analysis compared the per-protocol time to all-cause hospitalization under TOT vs under DOT strategies. We estimated (approximated) adjusted hazard ratios (aHRs), their 95% CIs, and adjusted survival curves through a weighted parametric pooled logistic model with an indicator for treatment group and a polynomial, flexible, time-varying intercept. Follow-up time was discretized to months.12,24

We adjusted for baseline confounding using stabilized inverse probability of treatment weighting created by regressing treatment assignment on baseline confounders using a logistic model. Baseline confounders were assessed using the 6-month period prior to time zero and included demographics and clinical characteristics, such as comorbidities, PAH-related symptoms, use of PAH-related laboratory tests and procedures, and comedication use (eTable 3-7 in Supplement 1).25,26 A standard mean difference less than 0.1 would be considered a sufficient baseline covariate balance.27,28

IPCW was created by regressing the censoring indicator on treatment group assignment, follow-up duration, and time-varying confounders, including comorbidity scores, individual comorbidities, PAH-related symptoms, and PAH-related procedure use via the longitudinal data.12,22 Stabilized inverse probability weight was created as the product of inverse probability of treatment weighting and IPCW at the patient-month level. Extreme weights, defined as the top and bottom 0.5th percentile, were trimmed.11,29 For the primary analysis, we used the Bonferroni method to correct for multiple testing by setting the α level to .016 (ie, 3 outcomes).30 Other statistical tests were conducted at an α level of .05. All tests were 2-sided. Data were analyzed using R statistical software version 4.1.2 (R Project for Statistical Computing) from November 2022 through July 2023.

Sensitivity analyses were prespecified and conducted. To gain insight into the effectiveness of early or delayed selexipag initiation, the TOT group was alternatively defined as treating patients additionally with oral selexipag within 3 and 12 months on and after time zero, and the artificial censoring rule was adjusted according to these alternative TOT treatment strategy definitions. Discontinuation of a medication was redefined as having a gap in medication claims of more than 45 days; this discontinuation definition represents a more stringent treatment adherence standard than the greater than 90–day gap. For patients to be eligible, we alternatively required a minimum of a 7- and 30-day treatment with DOT. A cubic spline term, an alternative flexible modeling choice instead of polynomial terms for time intercept, was fitted in the weighted parametric pooled logistic model.

Results

Patient Characteristics

In total, 2966 patients (mean [SD] age, 54.3 [14.0] years; 2125 female [71.6%]) met the eligibility criteria (eTable 8 in Supplement 1). The most prevalent PDE5i and ERA combination used was tadalafil-ambrisentan (970 patients [32.7%]), followed by sildenafil-ambrisentan (715 patients [24.1%]), sildenafil-macitentan (706 patients [23.8%]), and tadalafil-macitentan (460 patients [15.5%]). Most patients resided in the southern (948 patients [32.0%]) and western (940 patients [31.7%]) regions of the US and were enrolled in a health maintenance organization insurance plan (1829 patients [61.7%]). The most common general and PAH-associated comorbidities were systemic hypertension (2059 patients [69.4%]) and connective tissue disease (718 patients [24.2%]), respectively (Table 1).

Table 1. Baseline Patient Characteristics

Characteristic	Patients, No. (%)	SMD	
All eligible (N = 2966)a	DOT (n = 2921)	TOT (n = 2966)	
Sex					
Female	2125 (71.6)	2090 (71.6)	2125 (71.6)	0.002	
Male	841 (28.4)	831 (28.4)	841 (28.4)	
Age, mean (SD), y	54.3 (14.0)	54.3 (14.0)	54.3 (14.0)	0.001	
Geographic region					
Midwest	546 (18.4)	539 (18.5)	546 (18.4)	0.001	
Northeast	532 (17.9)	524 (17.9)	532 (17.9)	
South	948 (32.0)	933 (31.9)	948 (32.0)	
West	940 (31.7)	925 (31.7)	940 (31.7)	
Insurance type					
PPO	667 (22.5)	659 (22.6)	667 (22.5)	0.002	
HMO	1829 (61.7)	1801 (61.7)	1829 (61.7)	
Other	470 (15.8)	461 (15.8)	470 (15.8)	
Baseline all-cause hospitalization	1094 (36.9)	1075 (36.8)	1094 (36.9)	0.002	
Comorbidity score, mean (SD)	3.13 (2.32)	3.13 (2.32)	3.13 (2.32)	<0.001	
Comorbidities					
Diabetes without complication	756 (25.5)	739 (25.3)	756 (25.5)	0.004	
Diabetes with complication	365 (12.3)	355 (12.2)	365 (12.3)	0.005	
Obesity	946 (31.9)	931 (31.9)	946 (31.9)	<0.001	
Congestive heart failure	1690 (57.0)	1660 (56.8)	1690 (57.0)	0.003	
Atrial fibrillation	279 (9.4)	274 (9.4)	279 (9.4)	0.001	
Coronary artery disease	769 (25.9)	749 (25.6)	769 (25.9)	0.007	
COPD	1687 (56.9)	1663 (56.9)	1687 (56.9)	0.001	
Systemic hypertension	2059 (69.4)	2026 (69.4)	2059 (69.4)	0.001	
PAH-associated comorbidities					
Connective tissue disease	718 (24.2)	712 (24.4)	718 (24.2)	0.004	
HIV	100 (3.4)	98 (3.4)	100 (3.4)	0.001	
Portal hypertension	187 (6.3)	186 (6.4)	187 (6.3)	0.003	
Congenital heart disease	365 (12.3)	356 (12.2)	365 (12.3)	0.004	
PAH symptoms					
Syncope	308 (10.4)	302 (10.3)	308 (10.4)	0.001	
Peripheral edema	513 (17.3)	500 (17.1)	513 (17.3)	0.005	
Malaise fatigue	444 (15.0)	435 (14.9)	444 (15.0)	0.002	
Dyspnea	1086 (36.6)	1064 (36.4)	1086 (36.6)	0.004	
Hemoptysis	39 (1.3)	39 (1.3)	39 (1.3)	0.002	
Chest pain	773 (26.1)	757 (25.9)	773 (26.1)	0.003	
Dizziness	261 (8.8)	257 (8.8)	261 (8.8)	<0.001	
Abnormal gait	68 (2.3)	67 (2.3)	68 (2.3)	<0.001	
Cardiomegaly	933 (31.5)	919 (31.5)	933 (31.5)	<0.001	
Ascites	145 (4.9)	145 (5.0)	145 (4.9)	0.003	
Abbreviations: COPD, chronic obstructive pulmonary disease; DOT, double oral therapy; HMO, health maintenance organization; PAH, pulmonary arterial hypertension; PPO, preferred provider organization; SMD, standard mean difference; TOT, triple oral therapy.

a All eligible patients were entered into both DOT and TOT groups independent of their subsequent addition of oral selexipag, except for patients who had up-front TOT initiated at time zero.

A total of 351 patients added oral selexipag any time on or after time zero, with median (IQR) DOT and TOT durations of 5.9 (2.4-13.1) and 6.1 (1.3-14.5) months, respectively. These patients tended to have a greater comorbidity burden, including obesity and diabetes, at the time of selexipag initiation compared with time zero (eTable 9 in Supplement 1). Among these patients, 45 individuals initiated selexipag at time zero and therefore were not assigned to the DOT group, while 128 individuals initiated selexipag between time zero and the end of the sixth month and 178 individuals initiated selexipag after month 6 (Figure 1). Cumulatively, 103 patients (29.3%), 173 patients (49.3%), and 252 patients (71.8%) added selexipag within 3, 6, and 12 months, respectively. The remaining 2615 patients had no oral selexipag use and a median (IQR) DOT duration of 8.7 (3.2-18.2) months.

All 2966 patients’ initial treatment was compatible with the TOT, and 45 patients initiated up-front TOT and were incompatible with DOT. Under the target trial design, all baseline confounding variables achieved a standard mean difference less than 0.1 (Table 1; eTable 10 in Supplement 1).

Unadjusted Analyses After Artificial Censoring

Based on the unadjusted per-protocol analysis, 435 of 2966 patients (14.7%) in the TOT and 742 of 2921 patients (25.4%) in the DOT group had a hospitalization event before censoring. The most common reasons for censoring were discontinuation of ERA (905 patients [31.0%]) and PDE5i (515 patients [17.6%]) in the DOT group and not adding oral selexipag within 6 months in the TOT group (1498 patients [50.5%] by target trial design) (eTable 11 in Supplement 1). The median (IQR) follow-up for DOT and TOT groups was 8.9 (3.5-19.2) and 6.0 (3.8-6.1) months, respectively, after accounting for artificial censoring. For patients initiating TOT within 6 months vs DOT, the unadjusted HR for the occurrence of a first all-cause hospitalization event was 1.03 (95% CI, 0.90-1.17). Similar HRs were observed for unadjusted analyses for PAH-related hospitalization (HR, 1.02; 95% CI, 0.87-1.20) and disease progression (HR, 1.02; 95% CI, 0.86-1.19).

Selection Bias–Adjusted Analyses

Across all adjusted analyses, the stabilized inverse probability weight had a mean of 1.0 and maximum value of 40 or less. Adding selexipag within 6 months was associated with a reduction in the hazard for all-cause hospitalization vs DOT (adjusted HR [aHR], 0.82; 95% CI, 0.72-0.94). Similarly, the addition of oral selexipag to DOT within 6 months was associated with a lower hazard for PAH-related hospitalization (aHR, 0.81; 95% CI, 0.70-0.95) and disease progression (aHR, 0.82; 95% CI, 0.70-0.95) vs DOT alone (Table 2).

Table 2. Stabilized Inverse Probability Weight–Adjusted Analyses

Outcome	aHR (95% CI)	
3 moa	6 moa	12 moa	
All-cause hospitalizationb				
90-d Treatment gap allowed	0.77 (0.63-0.93)	0.82 (0.72-0.94)	0.89 (0.79-1.01)	
45-d Treatment gap allowed	0.79 (0.65-0.97)	0.85 (0.74-0.97)	0.91 (0.80-1.02)	
PAH-related hospitalization				
90-d Treatment gap allowed	0.85 (0.63-1.16)	0.81 (0.70-0.95)	0.97 (0.78-1.20)	
45-d Treatment gap allowed	0.75 (0.61-0.92)	0.84 (0.72-0.99)	0.89 (0.77-1.02)	
PAH-related progressionc				
90-d Treatment gap allowed	0.88 (0.64-1.21)	0.82 (0.70-0.95)	1.00 (0.82-1.21)	
45-d Treatment gap allowed	0.74 (0.61-0.90)	0.84 (0.73-0.97)	0.90 (0.80-1.03)	
Abbreviations: aHR, adjusted hazard ratio; PAH, pulmonary arterial hypertension.

a Analysis is by length of window to allow for adding oral selexipag.

b Primary outcome.

c Defined as parenteral therapy of epoprostenol or treprostinil, PAH-related hospitalization, lung transplant or atrial septostomy, or all-cause death.

Sensitivity Analyses

The aHR for all-cause hospitalization among patients receiving selexipag within 12 months vs DOT was 0.89 (95% CI, 0.79-1.01). Risk reduction for all-cause hospitalization was more pronounced when 3 months were allowed for the addition of selexipag (aHR, 0.77; 95% CI, 0.63-0.93). The aHRs for TOT when adding selexipag within 3 months vs DOT for PAH-related hospitalization and disease progression were similar to those comparing adding selexipag within 6 months vs DOT (Table 2).

When the maximum allowed gap between medication refills was adjusted from 90 to 45 days, adding selexipag within 3 months was associated with lower risk vs DOT for all-cause hospitalization (aHR, 0.79; 95% CI, 0.65-0.97), PAH-related hospitalization (aHR, 0.75; 95% CI, 0.61-0.92), and disease progression (aHR, 0.74; 95% CI, 0.61-0.90), and the magnitude of risk reduction was greater with the 3-month grace period than with the 6-month period. The aHRs for receipt of selexipag within 12 months vs DOT were 0.89 (95% CI, 0.77-1.02) for PAH-related hospitalization and 0.90 (95% CI, 0.80, 1.03) for disease progression.

Adjusted survival curves for all 3 outcomes under primary and some sensitivity analyses are shown in Figure 2 and Figure 3. All sensitivity analyses results are presented in eTable 12 in Supplement 1.

Figure 2. Adjusted Survival Curves by Strategy With 90-d Gap

DOT indicates double oral therapy; PAH, pulmonary arterial hypertension; TOT, triple oral therapy.

Figure 3. Adjusted Survival Curves by Strategy With 45-d Gap

DOT indicates double oral therapy; PAH, pulmonary arterial hypertension; TOT, triple oral therapy.

Discussion

This comparative effectiveness study found that in commercially insured patients with PAH, adding oral selexipag within 6 months of ongoing DOT consisting of ERA and PDE5i in clinical practice was associated with decreased risk for all-cause and PAH-related hospitalization and for disease progression. The greatest risk reduction across all outcomes was observed for escalation to TOT within 3 months with shorter treatment gaps (45 days). This supports the effectiveness of targeting the 3 foundational PAH disease pathways and is in line with current guidance.1

Our findings on the association of TOT with reductions in hospitalization complement the post hoc findings of the GRIPHON trial,4 in which adding oral selexipag to DOT consisting of ERA and PDE5i reduced the composite risk of morbidity and mortality. Our findings are also in line with pooled analysis findings of the GRIPHON and TRITON trials among the subgroup of 285 patients with PDE5i and ERA background therapy, in which early selexipag addition (≤6 months) was found to reduce the risk of disease progression (aHR, 0.52; 95% CI, 0.30-0.92).7 The smaller risk reduction (aHR, 0.82) found in our study could be due to differences in analytical choices (eg, the pooled analysis did not adjust for noninformative censoring) and, more importantly, the more heterogenous patient population included in our study (eg, number of comorbidities, age, and unexpected intolerance to PAH therapies), which may relate to the lower reduction in hospitalization risk with exposure to selexipag in the TOT group. For example, compared with patients from the pooled analyses, our patients were a mean of 7 years older and had higher prevalence of HIV infection and congenital heart disease at baseline. The large percentage of stress electrocardiograms may also indicate that our study cohort included patients with non–PH-related cardiovascular comorbidity.

Understanding long-term health care use and clinical outcomes, such as hospitalization, under various therapies is central for improving the lives of patients with PAH given that hospitalizations among these patients are common,31 costly,32 and highly predicative of mortality risk.33,34 PAH-related hospitalization also accounted for most PAH-related disease progression events, the composite end point in the GRIPHON study and our study.4,19

Studies have suggested that the treatment effect of selexipag on long-term outcomes was impacted by time to initiation of oral selexipag.4,5,10 Our results showed a potential benefit of adding oral selexipag early to an ERA and PDE5i treatment regimen. TOT was associated with the greatest decrease in all-cause hospitalization vs DOT when patients initiated selexipag within 3 months of DOT initiation, and the decrease was smaller or not significant when selexipag was initiated within 6 and 12 months. These results reinforce the potential benefit of early up-front triple therapy of selexipag, macitentan, and tadalafil vs double therapy of macitentan, tadalafil, and placebo in preventing PAH-related progression found in the exploratory TRITON study.5

When a 45-day gap instead of 90 days in therapy was allowed, TOT was associated with a greater reduction in PAH-related hospitalization and disease progression vs DOT if patients initiated oral selexipag within 3 months of DOT initiation. This finding suggests that prolonged treatment gaps, potentially due to challenges in adverse event management,35 may contribute to suboptimal clinical outcomes and is in line with critical aspects of treatment adherence in improving health outcomes in other chronic diseases.36,37,38

To our knowledge, this was the first study to use the target trial framework to evaluate the effectiveness of a time-varying PAH treatment. As of our data extraction, a conventional new user design would have been severely underpowered due to the small number of patients treated with up-front TOT given that many patients with PAH were treated in a stepwise manner. Furthermore, the target trial allowed us to use prevalent patients with prior PAH treatment and compare different time-varying treatment strategies.11,17,20,29,39 More importantly, this novel design avoided time-related bias that is often introduced by design in observational research.9,12

Limitations

This study has several limitations. First, we could not investigate the comparative effectiveness of an up-front TOT strategy owing to insufficient sample size. Second, claims data were collected for the purpose of reimbursement, not research, and were therefore subject to information error. For example, the prescription claim date indicates the dispensing of a medication, not its actual initiation. PAH-related hospitalization may be underestimated given that it is not clear from claims data to what extent an all-cause hospitalization event was precipitated by PAH-related comorbidities. Symptoms and comorbidities may be underreported. However, the misclassification of covariates or outcomes is unlikely to differ across treatment groups and generally biases the result toward the null.40 Additionally, we did not require a right heart catheter record as part of eligibility criteria. A previous validation study13 investigating the algorithm for identification of patients with PAH in claims data also suggests that additionally requiring a right heart catheter was not associated with increased sensitivity or positive predictive value. Without hemodynamic data, we cannot be certain that patients included in this analysis were incident patients with definitive PAH; however, we followed the recommended approach published by Sprecher et al, 2020,13 for identifying patients with PAH in claims data, and we acknowledge that misclassification and false positives were possible. In addition, socioeconomic and race and ethnicity information were not available in this claims dataset. Third, patients included in this study could also have been diagnosed with PH of groups 2 to 5. However, this reflects only the challenge in clinical practice for the diagnosis and treatment of patients where PH can have several possible causes that are not always mutually exclusive.41 Fourth, the study assumed that the outcome associated with selexipag was consistent across its various dosing regimens, as demonstrated in GRIPHON.3 Fifth, the use of IPCW assumed that censoring was noninformative conditional on time-varying variables in the model, and such an assumption is not verifiable.12,18 Notably, there may have been important clinical and laboratory information (eg, clinician characteristics, hemodynamics, functional class, and biomarker levels) that significantly contributed to changes in treatment decisions (eg, adding selexipag) but was unavailable in the data. However, we note that any causal analysis targeting per-protocol effect, including those for a randomized clinical trial, are subject to the same concern. Sixth, a marginal structural model in the form of a pooled logistic model requires specification of risk over time. We avoided model misspecification by using a flexible polynomial function of time and conducting a sensitivity analysis using cubic spline terms. Seventh, the patient group in this study is a convenience sample in nature and may not represent the general PAH population.

Conclusions

The aggregate of evidence in this comparative effectiveness study suggests that the early addition of oral selexipag (within 6 months) to a DOT of ERA and PDE5i was associated with a reduction in the risk of all-cause and PAH-related hospitalization and disease progression in patients with PAH. This reduction was even more pronounced when selexipag was added within 3 months. Conversely, our findings suggest that a long delay (12 months) in adding oral selexipag and substantial treatment gaps may contribute to suboptimal outcomes among patients with PAH. These findings suggest that the value of TOT may not be fully realized given that most patients in the dataset were never escalated to TOT.

Supplement 1. eTable 1. PAH-Indicated Medication

eTable 2. Components of PAH-Related Diagnosis and Outcomes

eTable 3. Quan-Charlson Comorbidity Index

eTable 4. Other Individual Comorbidities Not in Quan-Charlson Comorbidity Index

eTable 5. PAH-Related Symptoms

eTable 6. PAH-Related Procedures

eTable 7. Concomitant Medications

eTable 8. Patient Identification by Eligibility Criteria

eTable 9. Full List of Patient Characteristics Assessed at Time Zero

eTable 10. Reasons for Reaching the End of Follow-Up Under Per-Protocol Analysis for All-Cause Hospitalization

eTable 11. Stabilized Inverse Probability Weight–Adjusted Sensitivity Analyses Results

eReferences.

Supplement 2. Data Sharing Statement
==== Refs
References

1 Humbert M, Kovacs G, Hoeper MM, ; ESC/ERS Scientific Document Group. 2022 ESC/ERS guidelines for the diagnosis and treatment of pulmonary hypertension. Eur Heart J. 2022;43 (38 ):3618-3731. doi:10.1093/eurheartj/ehac237 36017548
2 Wilson M, Keeley J, Kingman M, Wang J, Rogers F. Current clinical utilization of risk assessment tools in pulmonary arterial hypertension: a descriptive survey of facilitation strategies, patterns, and barriers to use in the United States. Pulm Circ. Published online September 1, 2020. doi:10.1177/2045894020950186 33062258
3 Sitbon O, Channick R, Chin KM, ; GRIPHON Investigators. Selexipag for the treatment of pulmonary arterial hypertension. N Engl J Med. 2015;373 (26 ):2522-2533. doi:10.1056/NEJMoa1503184 26699168
4 Coghlan JG, Channick R, Chin K, . Targeting the prostacyclin pathway with selexipag in patients with pulmonary arterial hypertension receiving double combination therapy: insights from the randomized controlled GRIPHON study. Am J Cardiovasc Drugs. 2018;18 (1 ):37-47. doi:10.1007/s40256-017-0262-z 29307087
5 Chin KM, Sitbon O, Doelberg M, . Three- versus two-drug therapy for patients with newly diagnosed pulmonary arterial hypertension. J Am Coll Cardiol. 2021;78 (14 ):1393-1403. doi:10.1016/j.jacc.2021.07.057 34593120
6 Soderberg S, Escribiano Subias P, O’Donovan C, Survival in pulmonary arterial hypertension (PAH) for 669 patients treated with selexipag in clinical practice: insights from EXPOSURE. Paper presented at: the 44th Annual Meeting and Scientific Sessions of the International Society for Heart and Lung Transplantation; April 12, 2024; Prague, Czech Republic.
7 Coghlan JG, Gaine S, Channick R, . Early selexipag initiation and long-term outcomes: insights from randomised controlled trials in pulmonary arterial hypertension. ERJ Open Res. 2023;9 (1 ):00456-02022. doi:10.1183/23120541.00456-2022 36687361
8 Tsang Y, Stokes M, Kim YJ, Tilney R, Panjabi S. Impact of selexipag use within 12 months of pulmonary arterial hypertension diagnosis on hospitalizations and medical costs: a retrospective cohort study. Clin Respir J. 2023;17 (12 ):1209-1222. doi:10.1111/crj.13704 37804160
9 Hernán MA, Sauer BC, Hernández-Díaz S, Platt R, Shrier I. Specifying a target trial prevents immortal time bias and other self-inflicted injuries in observational analyses. J Clin Epidemiol. 2016;79 :70-75. doi:10.1016/j.jclinepi.2016.04.014 27237061
10 Gaine S, Sitbon O, Channick RN, . Relationship between time from diagnosis and morbidity/mortality in pulmonary arterial hypertension: results from the phase III GRIPHON study. Chest. 2021;160 (1 ):277-286. doi:10.1016/j.chest.2021.01.066 33545163
11 Hernán MA, Robins JM. Using big data to emulate a target trial when a randomized trial is not available. Am J Epidemiol. 2016;183 (8 ):758-764. doi:10.1093/aje/kwv254 26994063
12 Hernan MA, Robins JM. Target trial emulation. In: Hernan MA, Robins JM, eds. Causal Inference: What If. Chapman & Hall/CRC; 2023:299-316. Accessed August 14, 2024. https://www.hsph.harvard.edu/miguel-hernan/wp-content/uploads/sites/1268/2024/04/hernanrobins_WhatIf_26apr24.pdf
13 Sprecher VP, Didden EM, Swerdel JN, Muller A. Evaluation of code-based algorithms to identify pulmonary arterial hypertension and chronic thromboembolic pulmonary hypertension patients in large administrative databases. Pulm Circ. Published online November 1, 2020. doi:10.1177/2045894020961713 33240487
14 Kim D, Lee KM, Freiman MR, . Phosphodiesterase-5 inhibitor therapy for pulmonary hypertension in the United States: actual versus recommended use. Ann Am Thorac Soc. 2018;15 (6 ):693-701. doi:10.1513/AnnalsATS.201710-762OC 29485908
15 Zhang Y, Young JG, Thamer M, Hernán MA. Comparing the effectiveness of dynamic treatment strategies using electronic health records: an application of the parametric g-formula to anemia management strategies. Health Serv Res. 2018;53 (3 ):1900-1918. doi:10.1111/1475-6773.12718 28560811
16 Chakraborty B, Moodie EEM. Statistical Methods for Dynamic Treatment Regimes: Reinforcement Learning, Causal Inference, and Personalized Medicine. Springer; 2013. doi:10.1007/978-1-4614-7428-9
17 Hernán MA. How to estimate the effect of treatment duration on survival outcomes using observational data. BMJ. 2018;360 :k182. doi:10.1136/bmj.k182 29419381
18 Cole SR, Hernán MA. Constructing inverse probability weights for marginal structural models. Am J Epidemiol. 2008;168 (6 ):656-664. doi:10.1093/aje/kwn164 18682488
19 Gauthier-Loiselle M, Tsang Y, Lefebvre P, . Development and evaluation of a predictive algorithm for unsatisfactory response among patients with pulmonary arterial hypertension using health insurance claims data. Curr Med Res Opin. 2022;38 (6 ):1019-1030. doi:10.1080/03007995.2022.2049162 35243952
20 Dickerman BA, Madenci AL, Gerlovin H, . Comparative safety of BNT162b2 and mRNA-1273 vaccines in a nationwide cohort of US veterans. JAMA Intern Med. 2022;182 (7 ):739-746. doi:10.1001/jamainternmed.2022.2109 35696161
21 Ahn N, Nolde M, Günter A, . Emulating a target trial of proton pump inhibitors and dementia risk using claims data. Eur J Neurol. 2022;29 (5 ):1335-1343. doi:10.1111/ene.15284 35170158
22 Yoshida K, Liu J, Desai RJ, Glynn RJ, Solomon DH, Kim SC. Comparative safety of gout treatment strategies on cardiovascular outcomes using observational data: clone-censor-weight target trial emulation approach. Epidemiology. 2023;34 (4 ):544-553. doi:10.1097/EDE.0000000000001608 36943798
23 Takeuchi Y, Kumamaru H, Hagiwara Y, . Sodium-glucose cotransporter-2 inhibitors and the risk of urinary tract infection among diabetic patients in Japan: target trial emulation using a nationwide administrative claims database. Diabetes Obes Metab. 2021;23 (6 ):1379-1388. doi:10.1111/dom.14353 33606891
24 D’Agostino RB, Lee ML, Belanger AJ, Cupples LA, Anderson K, Kannel WB. Relation of pooled logistic regression to time dependent Cox regression analysis: the Framingham Heart Study. Stat Med. 1990;9 (12 ):1501-1515. doi:10.1002/sim.4780091214 2281238
25 Quan H, Li B, Couris CM, . Updating and validating the Charlson comorbidity index and score for risk adjustment in hospital discharge abstracts using data from 6 countries. Am J Epidemiol. 2011;173 (6 ):676-682. doi:10.1093/aje/kwq433 21330339
26 Quan H, Sundararajan V, Halfon P, . Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005;43 (11 ):1130-1139. doi:10.1097/01.mlr.0000182534.19832.83 16224307
27 Stuart EA, Lee BK, Leacy FP. Prognostic score-based balance measures can be a useful diagnostic for propensity score methods in comparative effectiveness research. J Clin Epidemiol. 2013;66 (8 )(suppl ):S84-S90.e1. doi:10.1016/j.jclinepi.2013.01.013 23849158
28 Zhang Z, Kim HJ, Lonjon G, Zhu Y; written on behalf of AME Big-Data Clinical Trial Collaborative Group. Balance diagnostics after propensity score matching. Ann Transl Med. 2019;7 (1 ):16. doi:10.21037/atm.2018.12.10 30788363
29 Madenci AL, Wanis KN, Cooper Z, . Strengthening health services research using target trial emulation: an application to volume-outcomes studies. Am J Epidemiol. 2021;190 (11 ):2453-2460. doi:10.1093/aje/kwab170 34089045
30 Shaffer JP. Multiple hypothesis testing. Annu Rev Psychol. 1995;46 :561-584. doi:10.1146/annurev.ps.46.020195.003021
31 Burger CD, Long PK, Shah MR, . Characterization of first-time hospitalizations in patients with newly diagnosed pulmonary arterial hypertension in the REVEAL registry. Chest. 2014;146 (5 ):1263-1273. doi:10.1378/chest.14-0193 24901386
32 Burke JP, Hunsche E, Régulier E, Nagao M, Buzinec P, Drake Iii W. Characterizing pulmonary hypertension-related hospitalization costs among Medicare Advantage or commercially insured patients with pulmonary arterial hypertension: a retrospective database study. Am J Manag Care. 2015;21 (3 )(suppl ):s47-s58.25734573
33 Benza RL, Gomberg-Maitland M, Elliott CG, . Predicting survival in patients with pulmonary arterial hypertension: the REVEAL risk score calculator 2.0 and comparison with ESC/ERS-based risk assessment strategies. Chest. 2019;156 (2 ):323-337. doi:10.1016/j.chest.2019.02.004 30772387
34 McLaughlin VV, Hoeper MM, Channick RN, . Pulmonary arterial hypertension-related morbidity is prognostic for mortality. J Am Coll Cardiol. 2018;71 (7 ):752-763. doi:10.1016/j.jacc.2017.12.010 29447737
35 Gordon T, Kannel WB. Premature mortality from coronary heart disease: the Framingham study. JAMA. 1971;215 (10 ):1617-1625. doi:10.1001/jama.1971.03180230027005 5107681
36 Jimmy B, Jose J. Patient medication adherence: measures in daily practice. Oman Med J. 2011;26 (3 ):155-159. doi:10.5001/omj.2011.38 22043406
37 DiMatteo MR. Variations in patients’ adherence to medical recommendations: a quantitative review of 50 years of research. Med Care. 2004;42 (3 ):200-209. doi:10.1097/01.mlr.0000114908.90348.f9 15076819
38 Vrijens B, De Geest S, Hughes DA, ; ABC Project Team. A new taxonomy for describing and defining adherence to medications. Br J Clin Pharmacol. 2012;73 (5 ):691-705. doi:10.1111/j.1365-2125.2012.04167.x 22486599
39 Hernán MA, Wang W, Leaf DE. Target trial emulation: a framework for causal inference from observational data. JAMA. 2022;328 (24 ):2446-2447. doi:10.1001/jama.2022.21383 36508210
40 Ogburn EL, VanderWeele TJ. On the nondifferential misclassification of a binary confounder. Epidemiology. 2012;23 (3 ):433-439. doi:10.1097/EDE.0b013e31824d1f63 22450692
41 Lang IM, Palazzini M. The burden of comorbidities in pulmonary arterial hypertension. Eur Heart J Suppl. 2019;21 (suppl K ):K21-K28. doi:10.1093/eurheartj/suz205 31857797
