Methods

A retrospective chart review was performed on all pediatric patients with IPAH or FPAH diagnosed before 2004 (aged <18 years at the time of diagnosis) at Columbia University Medical Center who transitioned off intravenous epoprostenol to oral or inhaled PAH agents from April 2003 to July 2008. All data were collected before the cut-off date of May 1, 2009. Clinical and hemodynamic assessments at baseline reflect data collected at the initial assessment, before the start of calcium channel blockers and/or intravenous epoprostenol, those at peak epoprostenol reflect data at the patients’ maximum stable intravenous epoprostenol dosage, and those after epoprostenol reflect data from the first follow-up performed ≥3 months after the discontinuation of intravenous epoprostenol therapy. All available hemodynamic and exercise data were included at baseline, peak stable epoprostenol dose, and after transition off epoprostenol therapy. Later term data were also collected when available. Statistical comparisons were made between baseline, peak epoprostenol, and post epoprostenol measurements.

Transition off intravenous epoprostenol was considered for patients who clinically and hemodynamically improved with intravenous epoprostenol and at the time of transition had WHO functional class I or II symptoms, had mean pulmonary arterial pressure <35 mm Hg, had normal cardiac index, were aged >6 years, and had families supportive of the transition. If parents were reluctant to transition because of a concern about the potential for clinical worsening off epoprostenol, the intravenous epoprostenol was not weaned. Transition was often delayed until the patients were aged >6 years, because the clinical strategy was to wait until maximal lung growth had occurred. In children aged >6 years, exercise capacity was also used as a tool for assessing response to intravenous epoprostenol wean. The actual transition time from intravenous epoprostenol to oral or inhaled PAH agents was variable and required serial assessments. Overall, patients were challenged first by halting increases of epoprostenol for ≥6 months to determine if they tolerated dose stabilization. If dose stabilization was well tolerated, an alternative agent, such as a phosphodiesterase-5 inhibitor or an endothelin receptor antagonist, was then added to the treatment regimen before decreasing in the epoprostenol dose. Once patients were receiving combination therapy, the epoprostenol dose was slowly decreased over several months. Patients were assessed at least every 3 months while weaning from intravenous epoprostenol, and decreases were generally halted at 10 to 20 ng/kg/min, with reassessment by cardiac catheterization at that dose. If there was any evidence of clinical worsening (e.g., a decrease in WHO functional class, a decrease in exercise capacity, or worsening of hemodynamics), the transition was stopped, and either another nonparenteral agent was added before further weaning or epoprostenol was increased to the previous level. If patients remained clinically stable and assessments remained favorable, epoprostenol was weaned to <10 ng/kg/min and discontinued in the cardiac catheterization laboratory if hemodynamics were stable. Reassessment including cardiac catheterization was targeted was for 3 to 6 months after the discontinuation of intravenous epoprostenol.

Paired Student’s t tests were used for comparisons; p values <0.05 were considered statistically significant. Data are expressed as mean ± SD.

Results

At the Pulmonary Hypertension Center of Columbia University Medical Center, from 1987 through 2000, 104 children (66 female) with IPAH (n = 89) or FPAH (n = 15) who were not responsive to acute vasodilator testing, were started on intravenous epoprostenol for PAH. From April 2003 to July 2008, 14 of these 104 children with IPAH or FPAH (6 female) underwent attempts at transition off continuous intravenous epoprostenol to oral or inhaled targeted PAH agents. Thirteen of the 14 patients treated with intravenous epoprostenol for 7.0 ± 3.4 years (range 1.2 to 11.8 years) were successfully transitioned off epoprostenol to oral or inhaled PAH agents. Fifteen patients, including 3 of 13 in the successfully transitioned group, had FPAH. When available, all WHO functional class, hemodynamic, and exercise data were included for baseline, peak stable epoprostenol therapy, and post epoprostenol therapy, with the exception of data from the 1 patient who failed transition ( Table 1 ).

Baseline data reflect clinical and hemodynamic status before the initiation of calcium channel blockers and/or intravenous epoprostenol. Intravenous epoprostenol therapy was initiated as first-line treatment in 11 patients and started in 3 patients who failed to adequately improve on calcium channel blockade (despite a robust response to acute vasodilator testing). In the 13 successfully transitioned patients, mean age at intravenous epoprostenol initiation was 4.4 ± 3.8 years (range 3 months to 12.2 years). Intravenous epoprostenol was initiated 14.5 + 11.9 months (range 1.6 months to 3.3 years) after the onset of symptoms. At baseline, 8% of these patients had WHO functional class II symptoms, 77% class III, and 15% class IV (n = 13; Figure 1 ). Hemodynamics at baseline were consistent with severe PAH ( Table 2 ): mean pulmonary arterial pressure 58 ± 21 mm Hg, pulmonary vascular resistance index (PVRi) 24 ± 20 U × m ² , PVRi/systemic vascular resistance index (SVRi) ratio 1.2 ± 1.3, cardiac index 3.9 ± 2.1 L/min/m ² (Fick technique with measured oxygen consumption during the procedure), and mixed venous saturation 56 ± 15%. Six-minute walk and cardiopulmonary exercise testing data were available at baseline for only 2 patients, because of the young age of these patients at the time of intravenous epoprostenol initiation.

WHO functional class: baseline versus peak epoprostenol (EPO) versus off epoprostenol follow-up (p <0.003, off epoprostenol versus peak epoprostenol).

Patients were treated with intravenous epoprostenol for 7.0 ± 3.4 years (range 1.2 to 11.8 years, n = 13) at a stable peak epoprostenol dose of 88 ± 32 ng/kg/min (range 33 to 137 ng/kg/min). At peak epoprostenol dose, clinical and hemodynamic status had significantly improved from baseline (peak epoprostenol vs baseline, respectively): pulmonary arterial pressure 25 ± 6 versus 58 ± 21 mm Hg, PVRi 3.5 ± 2.1 versus 24.0 ± 20.0 U × m ² , PVRi/SVRi ratio 0.3 ± 0.1 versus 1.2 ± 1.3, cardiac index 5.6 ± 2.4 versus 3.9 ± 2.1 L/min/m ² , and mixed venous saturation 74 ± 5% versus 56% ± 15% (p <0.05; Table 2 ) WHO functional class also improved on intravenous epoprostenol, with 31% of patients in WHO functional class I and 69% in functional class II (p <0.05; Figure 1 , Table 2 ). Although exercise data (6-minute walk and cardiopulmonary exercise testing) were available for only 2 patients at baseline because of their young age at the time epoprostenol was started, 6 patients had exercise data available at peak epoprostenol and after epoprostenol. At peak epoprostenol, the 6-minute walking distance was 526 ± 76 m (range 415 to 625, n = 6).

In 13 of the 14 patients in whom it was attempted at 11.3 ± 3.8 years of age (range 5.2 to 17.4 years), transition off epoprostenol was successful. The length of active transition time from peak epoprostenol dose to complete termination of intravenous epoprostenol was variable depending on the clinical stability and tolerance of medication adjustments (mean 3.4 ± 3.1 years, range 2 days to 9.7 years). Peak epoprostenol and post epoprostenol hemodynamic data were available for comparison in all 13 successful transitions ( Table 2 ). Post epoprostenol follow-up hemodynamic assessments were performed at 7.0 ± 5.8 months (range 1 to 21, median 5) after the discontinuation of intravenous epoprostenol therapy. Hemodynamic parameters after epoprostenol were unchanged compared with hemodynamic parameters on the stable peak epoprostenol dose ( Table 2 ): pulmonary arterial pressure 26 ± 7 versus 25 ± 6 mm Hg, PVRi 4 ± 3 versus 4 ± 2 U × m ² , PVRi/SVRi ratio 0.3 ± 0.1 versus 0.3 ± 0.1, cardiac index 4.4 ± 1.2 versus 5.4 ± 2.5 L/min/m ² , and mixed venous saturation 77 ± 3% versus 74 ± 5% for post epoprostenol versus peak epoprostenol, respectively (p = NS).

Data from children who were old enough at the time of peak epoprostenol assessment to perform the 6-minute walk test and cardiopulmonary exercise test reliably were included for analysis. Paired 6-minute walk data were available for 6 of the 13 successful transitions, and paired cardiopulmonary exercise data for 5 of the 13 successful transitions. Exercise capacity remained stable after transition off epoprostenol compared to peak epoprostenol: 6-minute walking distance: 521 ± 100 m (range 421 to 654) versus 526 ± 75.5 m (range 415 to 625) (p = NS, n = 6); cardiopulmonary exercise data: peak workload 116 ± 94 versus 53 ± 22 Watts, peak oxygen consumption 30 ± 8 versus 33 ± 8 ml/kg/min, ventilatory equivalent/V co ₂ ratio 32 ± 5 versus 33 ± 6, and end-tidal carbon dioxide 37 ± 3 versus 36 ± 5 torr (p = NS, n = 5).

WHO functional class continued to improve after the discontinuation of intravenous epoprostenol compared to peak epoprostenol: 85% in functional class I versus 31% in functional class I and 15% in functional class II versus 69% in functional class II (p <0.005 off epoprostenol vs peak epoprostenol, n = 13; Figure 1 ). For the 1 patient in whom transition failed, although the patient had functional class II symptoms at peak epoprostenol and remained in functional class II at the 2-month follow-up post epoprostenol assessment (immediately before epoprostenol reinitiation), there was a significant increase in estimated right ventricular systolic pressure after epoprostenol (109 mm Hg on echocardiography), with associated right ventricular dysfunction, prompting the reinitiation of intravenous epoprostenol. After 7 months back on intravenous epoprostenol, estimated right ventricular systolic pressure had decreased to 23 mm Hg. Overall, there were no significant changes in hemodynamic (n = 13), 6-minute walk (n = 6), or cardiopulmonary exercise test (n = 5) parameters after transition off intravenous epoprostenol compared to peak epoprostenol data ( Table 2 ). Data from the 1 patient in whom transition failed were not included in any statistical calculations.

During transition, patients began various combinations of oral and/or inhaled targeted PAH agents. For those patients transitioned when only bosentan was available (2001 to 2004), bosentan was the preferred initial transition agent. When sildenafil became available (2004), the decision to initiate an endothelin receptor antagonist versus a phosphodiesterase-5 inhibitor first was at the discretion of the treating physician. Iloprost was added in 1 patient after transition off epoprostenol. At follow-up after epoprostenol, of the 13 patients successfully transitioned off, 77% of patients were being treated with endothelin receptor antagonists, 69% with phosphodiesterase-5 inhibitors, 38% with calcium channel blockers, and 8% with inhaled iloprost ( Table 1 ). Of these 13 patients, 23% were receiving monotherapy, 62% dual therapy, and 15% triple therapy.

Although there is some overlap with post epoprostenol data, the late-term follow-up in those patients with data available after ≥9 months off epoprostenol continue to demonstrate stability, with maintenance of WHO functional class (median WHO functional class I at late follow-up vs II at peak epoprostenol therapy, n = 7) and hemodynamics (mean pulmonary arterial pressure 25 ± 7 vs 27 ± 7 mm Hg, n = 5) at 3.3 ± 2.3 years (range 0.9 to 6.6, median 2.8). As of the cut-off date (May 1, 2009), all 13 patients were alive and remained successfully off intravenous epoprostenol, maintaining their previous WHO functional classes on peak epoprostenol. There was 100% survival using this strategy, with 93% treatment success (i.e., 1/14 patient resumed epoprostenol because of clinical deterioration off epoprostenol).