The combined OPUS and OrPHeUS dataset is the largest new users database for macitentan, including contemporary real-world data and clinical practices for patients with PAH and the relatively understudied population of patients with CHD-PAH. Here we describe the baseline characteristics, treatment patterns, outcomes and safety of patients with CHD-PAH newly initiating on macitentan and report on three subgroups of CHD-PAH: Eisenmenger syndrome, left-to-right shunts, and small/coincidental CHD, that include patients with a spectrum of disease severities and comorbidities.

Patients with CHD-PAH in OPUS/OrPHeUS were younger in comparison to patients with I/HPAH both at diagnosis and at macitentan initiation, consistent with other registry data both at diagnosis/enrollment [11, 33]. The percentage of patients in WHO FC II and III were broadly similar between the CHD-PAH and I/HPAH groups, while the median 6MWD was higher in the CHD-PAH group compared to I/HPAH, similar to previous studies [11]. This may reflect the younger age of the CHD-PAH patient population and the lower proportion of these patients with comorbidities such as hypertension and obesity compared to the I/HPAH population. Baseline hemodynamics varied slightly between the CHD-PAH subgroups, with the Eisenmenger syndrome subgroup having the highest median PVR at baseline and the small/coincidental group the lowest. These findings contrast with previous reports that suggest patients with small/coincidental CHD have a higher PVR than other CHD subgroups [9]; however, these results need to be interpreted with caution due to the high proportion of missing hemodynamic data at baseline in OPUS and OrPHeUS. Based on previous studies, outcomes for patients with CHD-PAH tend to be better than for patients with I/HPAH [9, 11, 34], although this might not be the case for all CHD-PAH subgroups [10, 14, 34]. Our findings from OPUS/OrPHeUS align with this: outcomes also appeared to be better for patients with CHD-PAH compared to those with I/HPAH; however, the population was younger and more prevalent than the I/HPAH population and these factors were not adjusted for in the analyses. The L to R shunts group appeared to have the worst outcomes of the CHD-PAH subgroups, experiencing hospitalizations at a similar rate to the I/HPAH population. This could be reflective of the more advanced disease of these patients compared to the other CHD-PAH subgroups; they had more patients in WHO FC III/IV and the BNP/NT-proBNP high risk category and were more likely to be on intravenous/ subcutaneous prostanoid therapy. These patients also had higher use of anti-coagulants and switched to macitentan from another ERA more often than in the other CHD-PAH subgroups. Survival in the CHD-PAH group (and its subgroups) was slightly better compared to the I/HPAH group (90.2% versus 82.1% at 2 years after macitentan initiation); 1-year survival rates aligned with those observed for patients with CHD-PAH elsewhere (93.5% versus 91.5–99.0%) [9,10,11, 13]. Within the CHD-PAH group, the L to R shunts subgroup had the lowest survival estimate, while the Eisenmenger syndrome subgroup had the highest, despite worse hemodynamics at macitentan initiation. This observation in the Eisenmenger syndrome subgroup is consistent with earlier observations [9, 15], and may be due to the presence of a cardiac defect resulting in a pulmonary-to-systemic shunt, early development of PAH and subsequent right ventricular adaptation, which helps to maintain sufficient systemic flow and pressure [9, 15].

The CHD-PAH patient population was predominantly prevalent (73.0%) and initiated macitentan considerably later in the course of their disease (38 months from PAH diagnosis to macitentan initiation), especially among patients in the Eisenmenger syndrome (71.0 months) and L to R shunts (36.6 months) subgroups, as compared to the I/HPAH group (7.0 months to macitentan initiation; 52.7% prevalent). This extended duration from PAH diagnosis to macitentan initiation may reflect the initial use of other PAH therapies by these patients. In OPUS/OrPHeUS, nearly 20% of patients with CHD-PAH had previously received bosentan before switching to macitentan, compared to 10% of patients with I/HPAH. Data from the COMPERA registry showed that patients with CHD-PAH use bosentan more often compared to other ERAs [11], so in particular for the Eisenmenger syndrome subgroup, the long time from diagnosis to macitentan initiation is likely due to the 2015 European Society of Cardiology (ESC)/European Respiratory Society (ERS) guideline recommendation for patients to receive bosentan [31, 32]. These real-world data suggest that macitentan is being prescribed to patients with CHD-PAH, even after treatment with another ERA.

Slightly more patients with CHD-PAH (65.0%) initiated macitentan as part of combination therapy, compared to patients with I/HPAH (58.1%,), most commonly in combination with a PDE5i (57.4% and 47.9%, respectively). This may be reflective of the 2015 ESC/ERS treatment guidelines at the time, stating that ERA, PDE5i and prostanoid treatment “should be” considered in patients with CHD-PAH (specifically in patients with Eisenmenger syndrome) and that combination therapy “may be considered” in these patients [31, 32]. Stronger recommendations for combination therapy as initial treatment for repaired CHD-PAH and as sequential treatment for other CHD-PAH groups were made in the updated 2022 treatment guidelines [1, 2].

The proportion receiving combination therapy increased over time in both populations, with 78.9% and 68.5% of patients with CHD-PAH and I/HPAH, receiving combination therapy at 12 months, respectively. More patients with CHD-PAH received combination therapy in OPUS/OrPHeUS than observed elsewhere. In the COMPERA registry, 30% of patients with CHD-PAH initiated combination therapy at diagnosis, which increased to 50% at last observation (median observation time of 45.3 months) [11]. Comparisons between the two CHD-PAH datasets (OPUS/OrPHeUS and COMPERA), however, should be made with caution, due to differences in enrollment criteria (drug versus patient registry), geographical location (US versus Europe), and definition of baseline (macitentan initiation versus diagnosis).

Macitentan discontinuations due to an AE/HAE were similar (in 15.1% of patients with CHD-PAH and 17.6% of patients with I/HPAH, respectively); of the CHD subgroups, discontinuations due to an AE/HAE were lowest for those with Eisenmenger syndrome (in 11.3% of patients). Macitentan discontinuations for reasons other than an AE/HAE were slightly higher for patients with CHD-PAH (22.8%) than I/HPAH (18.6%); as previously described, this is reflective of the non-safety related barriers that impact on PAH therapy adherence, particularly in the US [26]. Overall, the safety profile, including hepatic safety of macitentan in patients with CHD-PAH was similar to patients with I/HPAH and to that reported for patients with PAH from trial data [17]. No new safety signals were detected in the CHD-PAH population.

While the OPUS and OrPHeUS studies capture useful real-world data on the management of patients with PAH newly initiating macitentan, including those with CHD-PAH, there are some limitations associated with their observational design. Firstly, given that OPUS and OrPHeUS were drug registries with the aim of evaluating the safety of macitentan-treated patients, results may not be directly comparable with disease registries that included patients not treated with macitentan and may not fully represent the CHD-PAH population in the US. For this reason, limited data were collected for non PAH-specific therapies used by patients. In both studies, patient follow-up was by routine clinical practice and not according to scheduled visits with protocol-mandated assessments, resulting in incomplete data collection, and higher inter-site and inter-operator differences in assessment. Furthermore, due to the retrospective nature of OrPHeUS, patient information may not have been consistently recorded in patients' medical charts, resulting in a greater amount of missing data in OrPHeUS than in OPUS. Due to the amount of missing data at baseline for important prognostic factors, cohort balancing could not be performed for key outcomes; comparisons between patient populations are therefore descriptive. As CHD-PAH is a subset of a rare disease, the sample sizes of the CHD-PAH subgroups are small, with a small number of events (in time-to-event analyses), hence results should be interpreted with caution. In addition, many parameters were investigator-assessed, and not adjudicated, including etiology and subgroups. The studies did not collect specific data regarding the presence of Down’s syndrome in patients and no echocardiographic data on the left and right ventricle were collected. Finally, no additional information was collected on how cardiac index was measured or how PVR was calculated beyond the values reported by the investigators.