Eisenmenger syndrome (ES) is a known complication of congenital heart disease associated with unrepaired systemic to pulmonary shunts. Evidence for use of targeted pulmonary arterial hypertension therapy in ES is limited. The early experience using ambrisentan was evaluated in a cohort of consecutive patients with ES who initiated ambrisentan at Columbia University’s Pulmonary Hypertension Center from January 1, 2007, to August 1, 2008. Effects of ambrisentan on rest and exercise systemic arterial oxygen saturation (S a O 2 ), exercise capacity, functional status, hemoglobin levels, and hemodynamics were evaluated and compared using paired Student’s t tests. Seventeen patients were evaluated at short-term (mean 163 ± 57 days) and longer term (mean 2.5 ± 0.5 years) follow-up. At short-term follow-up, there was an improvement in exercise capacity (6-minute walking distance 389 ± 74 vs 417 ± 77 m, p = 0.03, n = 11) and maintenance of rest S a O 2 (89 ± 7% vs 89% ± 6%, p = 0.75, n = 15), exercise S a O 2 (75 ± 15% vs 77% ± 15%, p = 0.33, n = 11), functional class (improvement in 2 patients, no change in 13), and hemoglobin (16.5 ± 2.8 vs 15.8 ± 1.8 g/dl, p = 0.11, n = 14). At longer term follow-up compared to baseline and short-term follow-up, there was stability of exercise capacity, S a O 2 , functional class, and hemoglobin. In conclusion, in this single-center cohort of patients with ES, ambrisentan was safe and was associated with increasing exercise capacity at short-term follow-up, with patients maintaining S a O 2 , functional class, and hemoglobin, and with no significant evidence of clinical deterioration at longer term follow-up. Additional studies are required to further assess the efficacy of ambrisentan in patients with ES.
Pulmonary arterial hypertension (PAH) is a recognized complication of congenital heart disease (CHD), related to pulmonary vascular remodeling due to nonrestrictive, shunt-related increases in pulmonary blood flow and/or exposure to increased pulmonary artery pressure. Over time, these patients develop predominant right-to-left shunts because of pulmonary arterial resistance exceeding systemic arterial resistance, accompanied by oxygen-unresponsive hypoxemia, identified as Eisenmenger syndrome (ES). Research on the use of targeted PAH agents to treat ES has been limited, in part because of concern for increased right-to-left shunting due to systemic vasodilation. Data exist that agents used to treat idiopathic PAH lead to improvements in hemodynamics, exercise capacity, and systemic arterial oxygen saturation (S a O 2 ) in patients with PAH and CHD. The use of endothelin receptor antagonists in these patients is intriguing, as plasma levels of this potent vasoconstrictor and promoter of fibrosis and inflammation have been noted to be increased in this patient population. Bosentan, a nonselective endothelin receptor antagonist, has been shown to improve exercise capacity and functional class without decreasing S a O 2 in patients with ES. There has been concern for hepatic transaminase elevation with bosentan, with up to 11% development in some studies. Ambrisentan, a selective endothelin-A receptor antagonist, has demonstrated a lower risk for transaminase elevation, and its selectivity for the endothelin-A receptor may provide an advantage due to the vasoconstrictor effects of endothelin-A receptors and the vasodilatory and clearance properties inherent to endothelin-B receptors.
Methods
A retrospective chart review was performed on consecutive patients with ES followed at the Columbia University Pulmonary Hypertension Center who were initiated on ambrisentan between January 2007 and August 2008. Data collected included demographics, type of CHD, and past or present treatment with oral and intravenous advanced therapies for PAH. Comparisons were performed of rest and exercise S a O 2 , 6-minute walking distance, blood hemoglobin level, World Health Organization (WHO) functional class, and hemodynamic catheterization data (if available) at baseline (before treatment with ambrisentan), at short-term follow-up (3 to 9 months after treatment initiation), and at longer term follow-up (>1 year after treatment initiation). Data cutoff for longer term follow-up was August 1, 2010. Data were compared using 2-tailed, paired Student’s t tests, with p values <0.05 considered statistically significant. Adverse events were documented. This study was approved by the institutional review board of the Columbia University College of Physicians and Surgeons.
Results
Seventeen patients with ES were initiated on ambrisentan from January 2007 to August 2008. Demographic data, type of CHD, and associated conditions are listed in Table 1 . The average age at the initiation of treatment was 32.2 ± 11.9 years. Fifteen patients were previously treated with bosentan or the selective endothelin-A receptor antagonist sitaxsentan before ambrisentan ( Table 1 ). In all, 14 patients were treated with sitaxsentan for an average of 2.5 ± 1.4 years, and 10 patients were treated with bosentan for an average of 2.0 ± 1.3 years. Patients treated with intravenous and inhaled prostanoids, as well as with the phosphodiesterase-5 inhibitor sildenafil, before the initiation of ambrisentan are listed in Table 1 as well. All patients were treated concomitantly with conventional therapies, including digoxin, diuretics, and warfarin, if warranted. All patients were initiated on a 5 mg/day dose of ambrisentan and up-titrated to 10 mg/day as tolerated. Fifteen patients had short-term follow-up data available at 3 to 9 months after the initiation of treatment (average time to follow-up 163 ± 57 days; Table 2 ). Longer term follow-up data were available ≥1 year after the initiation of treatment in 13 patients (average time to follow-up 2.5 ± 0.5 years; Table 3 ). All follow-up data were collected before the data cutoff date of August 1, 2010.
| Patient | Gender | Diagnosis | Age (years) | Endothelin Receptor Antagonist ⁎ | Time on Treatment | Prostanoid | Time on Treatment | Sildenafil † |
|---|---|---|---|---|---|---|---|---|
| 1 | M | ASD ‡ , Noonan syndrome | 13 | Bosentan | 17 mos | 0 | 0 | + |
| 2 | F | ASD ‡ | 23 | Bosentan | 19 mos | Inhaled iloprost | 20 mos † | + |
| Sitaxsentan | 3 yrs | |||||||
| 3 | F | ASD | 26 | Sitaxsentan | 2 yrs | 0 | 0 | + |
| Bosentan | 2 yrs | |||||||
| 4 | F | ASD | 34 | Sitaxsentan | 5 yrs | 0 | 0 | + |
| 5 | M | ASD | 36 | 0 | 0 | 0 | 0 | + |
| 6 | F | ASD | 40 | Bosentan | 6 mos ∥ | 0 | 0 | 0 |
| Sitaxsentan | 15 mos | |||||||
| 7 | M | ASD | 42 | Bosentan | 1 mos | IV epoprostenol | 16 mos ⁎ | + |
| Sitaxsentan | 5 yrs | |||||||
| 8 | F | ASD | 48 | 0 | 0 | IV epoprostenol | 18 mos ⁎ | + |
| IV treprostinil | 3 yrs † | |||||||
| 9 | F | ASD | 56 | Bosentan | 4 yrs | IV treprostinil | 9 yrs † | 0 |
| Sitaxsentan | 18 mos | |||||||
| 10 | F | VSD | 15 | Bosentan | 4 yrs | 0 | 0 | + |
| Sitaxsentan | 6 mos | |||||||
| 11 | F | VSD, trisomy 21 | 19 | Bosentan | 15 mos | 0 | 0 | 0 |
| Sitaxsentan | 3 yrs | |||||||
| 12 | M | VSD, trisomy 21 | 22 | Bosentan | 2 yrs | 0 | 0 | + |
| Sitaxsentan | 3 yrs | |||||||
| 13 | M | VSD | 24 | Sitaxsentan | 3 yrs | 0 | 0 | 0 |
| 14 | M | VSD | 29 | Sitaxsentan | 12 mos | 0 | 0 | 0 |
| 15 | F | VSD § | 30 | Sitaxsentan | 3 yrs | Inhaled iloprost | 2 yrs † | 0 |
| 16 | M | VSD | 44 | Bosentan | 3 yrs | IV epoprostenol | 4 yrs ⁎ | + |
| Sitaxsentan | 7 mos | IV treprostinil | 7 mos † | |||||
| 17 | F | Complete atrioventricular canal, trisomy 21 | 37 | Sitaxsentan | 3 yrs | 0 | 0 | 0 |
⁎ Patients were pretreated before initiation of ambrisentan.
† Patients were being treated at the time of initiation of ambrisentan.
‡ Residual ASD after operative repair.
§ Residual VSD after operative repair of complete atrioventricular canal.
∥ Bosentan discontinued because of elevation of hepatic transaminases.
| Variable | n | Baseline | Short-Term Follow-Up | p Value |
|---|---|---|---|---|
| Rest oxygen saturation (%) | 15 | 89 ± 7 | 89 ± 6 | 0.75 |
| Exercise oxygen saturation (%) | 11 | 75 ± 15 | 77 ± 15 | 0.33 |
| 6-minute walking distance (m) | 11 | 389 ± 74 | 417 ± 77 | 0.03 |
| Hemoglobin (g/dl) | 14 | 16.5 ± 2.8 | 15.8 ± 1.8 | 0.11 |
| Variable | n | Baseline | Long-Term Follow-Up | p Value | n | Short-Term Follow-Up | Long-Term Follow-Up | p Value |
|---|---|---|---|---|---|---|---|---|
| Rest oxygen saturation (%) | 13 | 89 ± 5 | 90 ± 6 | 0.47 | 11 | 89 ± 5 | 89 ± 6 | 0.75 |
| Exercise oxygen saturation (%) | 13 | 75 ± 11 | 76 ± 14 | 0.47 | 9 | 79 ± 12 | 80 ± 12 | 0.84 |
| 6-minute walking distance (m) | 12 | 395 ± 91 | 402 ± 70 | 0.65 | 8 | 420 ± 88 | 401 ± 68 | 0.15 |
| Hemoglobin (g/dl) | 12 | 16.5 ± 2.8 | 16.0 ± 2.5 | 0.14 | 9 | 15.8 ± 1.9 | 16.0 ± 2.7 | 0.62 |
Short-term follow-up results ( Table 2 ) demonstrated an improvement in 6-minute walking distance from baseline (389 ± 74 m at baseline vs 417 ± 77 m at short-term follow-up, p = 0.03). Rest and exercise S a O 2 and hemoglobin remained stable from baseline, while WHO functional class improved in 2 patients and remained the same in the other 13. Longer term follow-up results were compared to baseline and short-term follow-up findings. From baseline to longer term follow-up (n = 13), there was stability of 6-minute walking distance, rest and exercise S a O 2 , and hemoglobin ( Table 3 ). WHO functional class improved in 3 patients, worsened in 1, and remained the same in 9 patients. Eleven patients had data available at short-term and longer term follow-up (average time between assessments 2.1 ± 0.4 years). There was stability of 6-minute walking distance, rest and exercise S a O 2 , and hemoglobin between short-term and longer term follow-up in these patients ( Table 3 ). WHO functional class improved in 1 patient and remained the same in the other 10 patients.
Hemodynamic measurements were available within 1 year before and ≥6 months (mean 2.2 ± 1.0 years) after the initiation of treatment in 6 patients ( Table 4 ). There was significant improvement in mean pulmonary arterial pressure (61.8 ± 8.5 mm Hg at baseline vs 55.7 ± 4.6 mm Hg at follow-up, p = 0.05). Improvements in the ratio of pulmonary arterial systolic pressure to systolic arterial blood pressure and indexed pulmonary vascular resistance were not statistically significant. There were no significant effects on mean aortic pressure, systemic vascular resistance, and cardiac output.
