| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Tufts-New England Medical Center (Drs. Steiner, Preston, and Hill), Boston, MA; Rhode Island Hospital (Dr. Klinger), Providence, RI; Temple University Hospital (Dr. Criner), Philadelphia, PA; Massachusetts General Hospital (Dr. Waxman), Boston, MA; and Boston Medical Center (Dr. Farber), Boston, MA.
Correspondence to: M. Kathryn Steiner, MD, Massachusetts General Hospital, 55 Fruit St, Bulfinch Building 148, Boston MA, 02114; e-mail: ksteiner{at}partners.org
Abstract
Study objectives: We assessed the efficacy of bosentan in transitioning from prostacyclin infusions in patients with pulmonary arterial hypertension (PAH).
Methods: Twenty-two PAH patients were recruited from five PAH centers if they had been clinically stable while receiving therapy with IV epoprostenol or subcutaneous treprostinil for at least 3 months. Patients were observed in an open-label prospective trial while bosentan was added to therapy, and then epoprostenol or treprostinil were tapered after 2 months.
Results: Ten of the 22 patients were transitioned off prostacyclin infusion therapy after a mean (± SEM) duration of 6.1 ± 1.2 months. Of those patients, seven patients have continued not receiving prostacyclin infusion therapy for a mean duration of 17.7 ± 5.3 months, with no significant changes in pulmonary artery (PA) pressure estimated by echocardiography, World Health Organization (WHO)/New York Heart Association (NYHA) functional class, 6-min walk distance (6MWD), or Borg dyspnea score. The conditions of three patients deteriorated, necessitating the resumption of prostacyclin therapy, and two patients subsequently died. Twelve patients failed to transition or even lower the prostacylin infusion rate and had worsening of their WHO/NYHA functional class and estimated systolic PA pressures, and had a trend toward deterioration in their mean 6MWD (294 ± 41 to 198 ± 34 m, respectively; p = 0.2). Of these, two patients subsequently died. The baseline characteristics of those who transitioned successfully vs those who transitioned unsuccessfully were a lower prostacyclin infusion rate, and less severe elevations in the mean and estimated systolic PA pressures.
Conclusion: Transitioning from therapy with prostacyclin to bosentan is possible in some PAH patients, mainly in those receiving lower prostacyclin doses and having less pulmonary hypertension at baseline. Careful patient selection and close interim monitoring is needed because the conditions of patients can deteriorate, and they may not respond to the resumption of therapy with prostacyclin.
Key Words: bosentan • epoprostenol • pulmonary arterial hypertension • transitions • treprostinil
Pulmonary arterial hypertension (PAH) is an uncommon disease that is characterized by a progressive increase in pulmonary vascular resistance (PVR) leading to right ventricular failure and death.1 PAH occurs as an idiopathic process or in association with a variety of disease processes. The pathogenesis is unknown but is thought to involve an imbalance between vasoconstrictor and vasodilator mediators, favoring vasoconstriction and predisposing the patient to vessel wall thickening.234
Therapies that are aimed at restoring the vasodilator/vasoconstrictor balance by the infusion of prostacyclin analogues such as epoprostenol improve pulmonary hemodynamics, functional status, and survival.56 However, this therapy is complicated, requiring permanent placement of a tunneled central venous catheter and posing an ever-present risk of line infection.7 Furthermore, abrupt discontinuation of the infusion can be life-threatening because of the short half-life of the drug.7 Treprostinil, which is a prostacyclin analog with a longer half-life than epoprostenol, can be given subcutaneously,8 providing therapy with greater convenience and safety than that with epoprostenol, but most patients are troubled by pain at the infusion site. Treprostinil can also be administered IV, offering convenience and safety advantages over IV epoprostenol, but line sepsis remains a threat.9 Thus, the transition from prostacyclin infusion to oral therapy is an appealing option for many patients with PAH.
Bosentan, which is an orally active nonselective endothelin receptor antagonist, improves the 6-min walk distance (6MWD), pulmonary hemodynamics, and functional status, and delays the time to clinical worsening in patients with class III or IV PAH patients compared to placebo.10 Few studies1112 have examined the possibility that patients receiving infusion therapies can be transitioned to more convenient and less complicated oral therapies such as with bosentan. Even with the lowering of the prostacyclin dose, such an intervention would cut down on expense, reduce the side effects, and minimize the risks of overdosage and high cardiac output failure that are inherent with prostacyclin infusion. Moreover, no characteristics have been identified that can predict the success or failure of therapy transitioning. In this pilot study, we asked whether adding bosentan to prostacyclin infusion therapy would (1) permit a lowering of the prostacyclin dose with the maintenance of functional capacity and (2) allow long-term complete transitioning in some patients. We also sought to identify patient characteristics that would predict successful transitioning.
Materials and Methods
Patients
The study was approved by the institutional review boards at all five participating PAH centers (ie, Tufts-New England Medical Center, Boston Medical Center, Rhode Island Hospital/Brown University, Massachusetts General Hospital, and Temple University) and was overseen by a data safety monitoring board. Patients with PAH receiving therapy either epoprostenol or treprostinil by continuous infusion were offered entry into the trial for potential transitioning to therapy with oral bosentan, and enrolled patients gave written informed consent. Patients who were deemed to be eligible for the study were over age 18 years, had experienced clinically stable PAH (defined as stable symptoms and no evidence of heart failure) for at least 3 months, had a stable World Health Organization (WHO)/New York Heart Association (NYHA) functional score, had experienced a <10% change in 6MWD, and had undergone no changes in their prostacyclin dose for at least the previous month. Patients needed to have had PAH confirmed on a prior right heart catheterization, as determined by a mean pulmonary artery (PA) pressure of 
25 mm Hg at rest and a PA wedge pressure of 
15 mm Hg. In addition to idiopathic PAH, patients with PAH associated with connective tissue disease, congenital heart disease, sarcoidosis, or HIV were also eligible to be enrolled in the study.
Patients were excluded from the study if they were receiving therapy with cyclosporine A or glyburide, had a history of cirrhosis or elevated baseline levels of transaminases more than three times the upper limit of normal, worsening signs or symptoms of pulmonary hypertension (defined as escalating dyspnea, light-headedness, chest pain, WHO/NYHA functional score, or prostacyclin doses, or signs of worsening right heart failure), were receiving another investigational drug, had a current or planned pregnancy, or were nursing. Women were eligible to participate in the trial if they were surgically sterile, at least 1 year postmenopausal, or were using nonhormonal birth control and had a negative urine or serum pregnancy test result at baseline. Patients were enrolled in the study between June 2002 and January 2004.
Transition Process
Enrolled patients were started on therapy with bosentan, 62.5 mg bid for a month, after which the dose was increased to 125 mg bid, unless the liver function test results increased by more than threefold or the patient had another serious adverse reaction. The decrease in the dose of epoprostenol or trepostinil was started on an outpatient basis 8 weeks after the initiation of bosentan therapy, based on the reported time to maximal effect.10 The prostacyclin infusion rate was decreased by 2 ng/kg/min or 5%, whichever was greater, every other day, as tolerated. For symptoms of epoprostenol or trepostinil excess after beginning bosentan therapy, such as increased headache, jaw ache, nausea, diarrhea, and leg pain, the dose was decreased by 2 ng/kg/min hourly until the symptoms abated. For symptoms of epoprostenol or trepostinil withdrawal such as increased dyspnea, chest pain, or fatigue, the infusion rate was increased by steps of 2 ng/kg/min hourly until the symptoms abated. After 2 days, the prostacylin dose could be decreased again, provided that the patients’ symptoms had resolved. Dose reductions continued every other day as tolerated until therapy with prostacyclin was discontinued or symptoms worsened. The rate of dose reduction was slowed to as-tolerated if patients repeatedly complained of symptoms of epoprostenol or trepostinil withdrawal, and attempts at epoprostenol or trepostinil weaning were abandoned if patients had persisting symptoms of withdrawal after multiple attempts (ie, more than three attempts). Once the patient was weaned from epoprostenol completely, the central catheter was removed after at least an additional 2 weeks if the patient remained clinically stable.
Outcome Measures
Patients were monitored monthly via office visits and telephone follow-up as needed for the assessment of clinical status and the titration of prostacyclins. Study end points were assessed before bosentan was added to therapy, at 6 and 12 months after bosentan was added, and periodically thereafter. The primary end point was change in the 6MWD, examined over the first 12 months after the start of bosentan therapy. The 6-min walk tests were performed according to established guidelines.1314 Patients who were unable to walk at baseline for reasons other than dyspnea were not assigned a value in meters. Patients who died or withdrew from the study before week 24 had their last assessment carried forward.
Secondary end points included changes in epoprostenol or trepostinil doses, Borg dyspnea score, WHO/NYHA classification, therapy for PAH (eg, a change in the dose of antithrombotic, diuretic, calcium antagonist, or cardiac glycoside agents), and estimated PA systolic pressure determined by echocardiography. The echocardiography interpretation was performed by the cardiologists at the institution, who were unaware of the study. Serum levels of transaminases and bilirubin were monitored monthly, and US Food and Drug Administration guidelines regarding the management of bosentan were otherwise followed. A chart review was performed in all enrolled patients to confirm their diagnosis by right heart catheterization, and to determine their initial pulmonary hemodynamics, 6MWD, Borg dyspnea score, WHO/NYHA functional class, and estimated systolic PA pressures and right ventricular function by echocardiography at the time of the diagnosis of PAH. Previous right heart catheterizations were required to establish the diagnosis of PAH but not to monitor the transition.
Statistical Analysis
Using the Mann-Whitney test, we compared the medians of baseline characteristics between patients who transitioned and those who could not transition from prostacyclin therapy. Using the Wilcoxon signed rank test, we compared the medians of clinical end points over time (ie, baseline vs 6 months, baseline vs 12 months, and 6 months vs 12 months) of those patients who transitioned successfully and those who could not. All means were compared using nonparametric methods. Using the Wilcoxon signed rank test, we compared medians of the prostaglandin I2 (PGI2) dose over time (ie, baseline vs 12 months) of the nontransitioned patients. For within-group comparisons of the proportion of patients having a WHO/NYHA functional class I and II vs those with functional class III and IV at baseline, at 6 and 12 months, we used the McNemar test, and for between-group comparisons we used the Fisher exact test. To determine whether the likelihood of transitioning therapy in these patients was influenced by one of the clinical end points by taking into account survival, we used the Cox proportional hazards survival regression model for exploratory analyses. The data are shown as the mean ± SEM. Statistical significance was defined as p < 0.05.
Results
Demographics
Twenty-two PAH patients were recruited from a total of 72 patients with PAH who were receiving prostacyclin therapy at five PAH centers. The remaining 50 patients either declined study entry (16 patients) or were not deemed sufficiently stable (34 patients) for study entry. As shown in Table 1
, all enrolled patients were women (age range, 23 to 75 years; mean age, 49 ± 3 years) and predominately white. Comorbid conditions, time from diagnosis, and duration of PGI2 therapy prior to study enrollment did not differ between transitioned and non-transitioned patients. The proportion of patients who successfully transitioned therapy was similar between patients treated with epoprostenol (8 of 17 patients) and those treated with trepostinil (2 of 5 patients) [Table 1].
|
|
|
6MWD and WHO/NYHA Functional Class
Figure 2
, top, A, and Tables 2and 3
show that the 6MWD tended to worsen over the 12 months of the study for both groups, but more for those patients who did not transition (mean difference from baseline to 12 months for transitioned patients, –41 vs –84 m; p = 0.3). However, transitioned patients had longer mean 6MWDs 6 months after the addition of bosentan to therapy than did nontransitioned patients (324 ± 46 vs 223 ± 41 m, respectively; p = 0.04). Borg dyspnea scores that were obtained at the end of the 6MWD were similar at baseline, at 6 months, and at 12 months in patients who successfully transitioned, but tended to worsen at 6 and 12 months in nontransitioned patients. The mean Borg dyspnea score was higher at 12 months in the nontransitioned patients compared to those who transitioned (4.4 ± 0.7 vs 2.9 ± 0.5, respectively; p = 0.05) [Fig 2, bottom, B, Table 2]. Although there were no differences between groups at baseline, the WHO/NYHA functional class was significantly better at 12 months in those who transitioned than in those who could not transition. The condition of one patient in the transitioned group deteriorated from class II to class IV compared to four patients in the nontransitioned group, who changed from class I/II to class III/IV by 12 months (p = 0.02) [Fig 3
, Tables 2, 3].
|
|
|
|
Four patients, two with idiopathic PAH, one with familial PAH, and one with sarcoid-related PAH, died at 12, 14, 14, and 18 months, respectively, from the time of bosentan therapy initiation. Two patients died 5 and 14 months after tapering off epoprostenol, while the other two were unable to transition. At baseline, all four patients were in WHO/NYHA functional class II and had diminished 6MWDs (419, 358, 363, and 232 m, respectively). From prior right heart catheterizations, the mean PA pressures were 39, 43, 53, and 48 mm Hg, respectively, the mean PVRs were 821, 467, 1,500, and 496 dyne · s · cm–5, respectively, and the mean cardiac indexes were 1.63, 3.67, 1.2, and 2.65 L/min/m2, respectively. No indexes were identified that predicted subsequent mortality.
Factors Associated With Successful Transitioning
At baseline, transitioned patients had significantly lower prostacyclin doses, echocardiographic estimates of PA systolic pressure, and mean readings of PA pressure obtained via prior catheterization (Table 1). In addition, baseline WHO/NYHA class, 6MWD, and associated Borg dyspnea score tended to be better in those who transitioned therapy. By exploratory Cox proportional survival hazard regression analysis, baseline estimated PA systolic pressure (hazard ratio [HR], 0.94; 95% confidence interval [CI], 0.89 to 0.99; p = 0.02), mean PA pressure (HR, 0.9; 95% CI, 0.84 to 0.96; p = 0.001), and PGI2 dose (HR, 1; 95% CI, 0.914 to 1; p = 0.05) were predictors associated with successful transitioning. 6MWDs, Borg dyspnea scores, and WHO/NYHA functional class did not reach statistical significance to predict the likelihood of transitioning, but their HRs were elevated (1, 0.9, and 2, respectively). There were no significant differences in the likelihood of being able to transition off PGI2 among the five study sites.
Discussion
Our open-label, multicenter pilot study demonstrates that transitioning from therapy with prostacyclin infusion to bosentan is possible in a minority of PAH patients (10 of 22 patients; 45%). Late failures to transition necessitating the resumption of prostacyclin therapy reduced the number of patients who remained successfully transitioned after 12 months to 7 of 22 (32%). Although transition was considered successful if the patient continued not to receive IV therapy, the lower mean 6MWD at 1 year suggests that the conditions of at least some patients may have clinically worsened. Those who could not transition had no reduction in average prostacyclin dose. Transitioning is possible mainly in patients receiving a relatively low dose of prostacyclin, with no more than mild elevations in PA systolic pressure determined by echocardiogram and WHO/NYHA class I or II. Patients with low mean PA pressures determined by right heart catheterization were also more likely to transition successfully, but these PA pressures were measured at the time of the patient’s diagnosis not when bosentan therapy was initiated. The occurrence of deaths in two transitioned patients who were in WHO/NYHA functional class II at baseline despite the resumption of PGI2 infusion indicates that transitions should be attempted with great caution.
Previous series1112 have reported the successful transitioning of PAH patients from therapy with prostacyclin infusion to oral bosentan. Kim et al11 described three patients whose PA pressures normalized while receiving epoprostenol, all of whom remained stable while receiving bosentan therapy for a mean period of 10 months following the discontinuation of epoprostenol infusion. Suleman and Frost12 reported on 23 PAH patients who had transitioned from prostacyclin infusion (epoprostenol, 17 patients; trepostinil, 6 patients) to bosentan therapy in a prospective 8-week transitioning protocol. At 12 weeks, 15 patients (65%) had successfully transitioned, but over the subsequent 12 months 6 additional patients resumed prostacyclin therapy, 4 patients because of progressive PAH symptoms and 2 patients because of liver function abnormalities, leaving 9 patients (40%) of the original 23. This is slightly higher than our 1-year transition rate of 32%, possibly reflecting the greater impairment of our patient population (mean initial 6MWD: our study, 304 ± 31 m; Suleman and Frost12 study, 392 ± 21 m), but both studies are in agreement that a minority of patients can be transitioned successfully in the long term. Considering that both studies transitioned patients who had begun prostacyclin therapy before bosentan became available, it is likely that at least some of the transitioned patients would have done well receiving bosentan as the initial therapy. In addition, a number of patients had experienced complications of IV infusion or were insistent on attempting transitioning to the new oral medication. Our protocol was devised as a pilot to accommodate these patients and to learn from their experience.
In our study and that of Suleman and Frost,12 patients were selected only if they manifested clinical stability and had no evidence of right heart failure. However, Suleman and Frost12 were unable to identify significant predictors of successful transitioning, presumably because of the small number of patients and missing data. In our study, exploratory survival regression analysis revealed that lower prostacyclin doses and lower PA systolic pressures estimated by echocardiography were significant predictors of a successful transition. These findings suggest that noninvasive testing can be used to aid in selecting patients who are more likely to have successful transitions. Echocardiographic estimates of systolic PA pressures may be inaccurate.1516171819 Echocardiographic indexes of right ventricular function such as mean acceleration time, maximal deceleration, and rate correction before the ejection period may be more reliable,18 but these data were not obtained routinely at the participating centers.
An interesting observation was that 6 of our 22 patients experienced symptoms of prostacyclin excess after first starting bosentan therapy and needed initial rapid reductions in the prostacyclin dose. Whether this effect is related to an alteration of prostacyclin metabolism or some other potentiation of prostacyclin action is unclear from our study, but considering that five of these six patients were subsequently weaned from prostacyclin, it may be an indicator of successful transitioning.
Our study was conceived partly to investigate the technique of weaning. We considered it important to wait at least 2 months after starting bosentan therapy to allow time for significant therapeutic effects10 before weaning from prostacyclin. We initially planned to wean the patients from prostacyclin every other day, so that the average patient should have been weaned from prostacyclin completely after a total time of 4 months after starting bosentan therapy. However, the duration of weaning was actually much longer, averaging 6 months, because patients could not tolerate the planned weaning rate due to the worsening of PAH symptoms, mainly dyspnea. Suleman and Frost12 had weaned their patients by the end of the second month of bosentan therapy, achieving a 65% initial success rate, but it is possible that fewer of their patients would have had late transition failures had they been weaned more slowly. It is also conceivable that the right heart catheterizations performed during transitions might have detected the deterioration earlier. Thus, we suggest that the weaning from prostacyclin be done slowly and as-tolerated, with decreases in dose occurring no more often than once or twice a week.
We anticipated that even if patients were unsuccessful in transitioning completely, the prostacyclin dose would be lower and clinical status might be better because of the possible additive effect of combination therapy. However, we observed no such dose reduction, and, furthermore, we found a significant worsening in their WHO/NYHA functional class and estimated systolic PA pressure with a trend for the worsening of 6MWD despite combination therapy in patients who could not transition. This could have been related either to progression of the underlying PAH or to the attempt at therapy transition accelerating the decline.
Bosentan therapy was well tolerated in our study, with only one patient having to stop receiving therapy because of abnormal liver function test results. However, the late deterioration of several patients who initially transitioned successfully are concerning, particularly considering that some of these patients failed to respond to the resumption of prostacyclin therapy and died, even when treated with doses higher than the pretransition dose. This underlines the need for close monitoring, even after patients have completely transitioned. Early signs of increasing dyspnea, a decrease in 6MWD, or a worsening of WHO/NYHA functional class or estimated systolic PA pressure should prompt consideration of the resumption of prostanoid therapy.
Disturbingly, there were a total of four deaths, two who transitioned and two who could not. With small numbers and lacking a control group, it is difficult to know whether our approach contributed in any way to these deaths. We could identify no specific baseline characteristics that differentiated these patients from those who did well.
Although the lack of control subjects, the small number of patients, and the open-label design are significant limitations of our study, the similarity of our results with those of Suleman and Frost12 strengthen the idea that about one third of PAH patients whose conditions have stabilized with prostacyclin infusion therapy can be transitioned to oral therapy with bosentan. Considering that only stable patients were included in our study, we must acknowledge the possibility that our successfully transitioned patients would have tolerated weaning from epoprostenol therapy, even without the addition of bosentan therapy. However, we consider this unlikely given that all patients had significant functional limitations (WHO/NYHA functional class III-IV) when they started prostacyclin therapy, and the conditions of such patients usually deteriorate without therapy.5
Conclusions
Although transitioning from infusion to oral therapy greatly enhances convenience and is desirable from the patient’s perspective, our findings indicate that such a transition from therapy with infused prostacyclins to bosentan therapy is possible only in a minority of patients. In general, patients who are likely to transition successfully are receiving lower prostacyclin doses and have relatively mild PA pressure elevation, both at the time of diagnosis and at the time of transition. Patients should be transitioned very cautiously if at all because their clinical status can deteriorate and may not recover after the resumption of prostacyclin therapy. Larger controlled studies of longer duration are needed to answer many of the questions raised in this study and to determine whether other factors, like genetic predisposition, are useful in predicting the likelihood of a successful transition of therapy.
Acknowledgements
The authors are grateful to Arlene Schiro, NP, Anne Marie Kuzma, and Barbara S. Smithson, RN BSN, for their assistance in preparing the article.
Footnotes
Abbreviations: CI = confidence interval; HR = hazard ratio; NYHA = New York Heart Association; PA = pulmonary artery; PAH = pulmonary arterial hypertension; PGI2 = prostaglandin I2; PVR = pulmonary vascular resistance; WHO = World Health Organization; 6MWD = 6-min walk distance
This research was supported by Actelion Pharmaceuticals US, Inc, Ipswich, MA.
Drs. Steiner and Criner have reported to the ACCP that they have no significant conflicts of interest with any companies/organizations whose products or services may be discussed in this article. Dr. Preston has reported receiving grants from Actelion, Cotherix, Encysive, Myogen, Pfizer, United Therapeutics, and ICOS-Lilly. Dr. Klinger has reported receiving grants from Actelion, Cotherix, Encysive, Myogen, Pfizer, ICOS-Lilly, and Sanofi-Aventis. Dr. Waxman has reported receiving grants from Actelion, Cotherix, Encysive, Myogen, Pfizer, United Therapeutics, ICOS-Lilly, and Predix. Dr. Farber has reported receiving grants from Actelion, Cotherix, Encysive, Myogen, Pfizer, ICOS-Lilly, and Predix. Dr. Hill has reported the following conflicts of interest: he has worked for the speakers bureau of Actelion; and he has received research grants from Actelion, Cotherix, Encysive, Myogen, Pfizer, United Therapeutics, ICOS-Lilly, and Predix.
Received for publication November 23, 2005. Accepted for publication May 20, 2006.
References
This article has been cited by other articles:
|
|
 |
|
  R. L. Benza, B. K. Rayburn, J. A. Tallaj, S. V. Pamboukian, and R. C. Bourge Treprostinil-Based Therapy in the Treatment of Moderate-to-Severe Pulmonary Arterial Hypertension: Long-term Efficacy and Combination With Bosentan Chest, July 1, 2008; 134(1): 139 - 145. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Shapiro and N. S. Hill Transition From IV to Subcutaneous Prostacyclin: Premature Withdrawal? Chest, September 1, 2007; 132(3): 741 - 743. [Full Text] [PDF]
|
|
|
|
|
|
M. Rubenfire, V. V. McLaughlin, R. P. Allen, G. Elliott, M. H. Park, M. Wade, and R. Schilz Transition From IV Epoprostenol to Subcutaneous Treprostinil in Pulmonary Arterial Hypertension: A Controlled Trial Chest, September 1, 2007; 132(3): 757 - 763. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
= mean ± SEM. The two dark circles with a value of > 200 ng/kg/min are patients who received trepostinil, and the remainder of the patients received epoprostenol.
