Author + information
- Received September 20, 2005
- Revision received December 23, 2005
- Accepted January 9, 2006
- Published online May 16, 2006.
- Robyn J. Barst, MD⁎,1,⁎ (, )
- David Langleben, MD†,1,
- David Badesch, MD‡,1,
- Adaani Frost, MD§,1,
- E. Clinton Lawrence, MD∥,1,
- Shelley Shapiro, MD¶,1,
- Robert Naeije, MD#,1,
- Nazzareno Galie, MD⁎⁎,1,
- STRIDE-2 Study Group
- ↵⁎Reprint requests and correspondence:
Dr. Robyn J. Barst, Columbia University College of Physicians and Surgeons, 3959 Broadway, BHN 2-255, New York, New York 10032.
Objectives We sought to determine the optimal dose of the selective endothelin A (ETA) receptor antagonist sitaxsentan for the treatment of pulmonary arterial hypertension (PAH); for observation only, an open-label (OL) bosentan arm was included.
Background Endothelin is a mediator of PAH. In a preliminary PAH study, the selective ETAreceptor antagonist sitaxsentan improved six-min walk (6MW) distance, World Health Organization (WHO) functional class (FC), and hemodynamics.
Methods In this double-blind, placebo-controlled 18-week study, 247 PAH patients (idiopathic, or associated with connective tissue disease or congenital heart disease) were randomized; 245 patients were treated: placebo (n = 62), sitaxsentan 50 mg (n = 62) or 100 mg (n = 61), or OL (6MW tests, Borg dyspnea scores, and WHO FC assessments third-party blind) bosentan (n = 60). The primary end point was change in 6MW distance from baseline to week 18. Secondary end points included change in WHO FC, time to clinical worsening, and change in Borg dyspnea score.
Results At week 18, patients treated with sitaxsentan 100 mg had an increased 6MW distance compared with the placebo group (31.4 m, p = 0.03), and an improved WHO FC (p = 0.04). The placebo-subtracted treatment effect for sitaxsentan 50 mg was 24.2 m (p = 0.07) and for OL bosentan, 29.5 m (p = 0.05). The incidence of elevated hepatic transaminases (>3× the upper limit of normal) was 6% for placebo, 5% for sitaxsentan 50 mg, 3% for sitaxsentan 100 mg, and 11% for bosentan.
Conclusions Treatment with the selective ETAreceptor antagonist sitaxsentan, orally once daily at a dose of 100 mg, improves exercise capacity and WHO FC in PAH patients, with a low incidence of hepatic toxicity.
Pulmonary arterial hypertension (PAH) is a progressive disease characterized by vasoconstriction and structural changes in small pulmonary arteries, with increased pulmonary artery pressure and pulmonary vascular resistance ultimately leading to right heart failure and death (1). Endothelin-1 (ET), a 21-amino acid peptide with vasoconstrictor, mitogenic, and profibrotic effects (2), appears to play a key role in the pathobiology of PAH (3,4). Two distinct ET receptor isoforms have been identified, ETAand ETB(5). EndothelinAreceptors are expressed on smooth muscle cells and cardiac myocytes; ETBreceptors are localized on vascular endothelial cells and smooth muscle cells (6). Activation of ETAand ETBreceptors on smooth muscle cells results in vasoconstriction, and cell proliferation and hypertrophy. Activation of ETBreceptors on endothelial cells leads to release of vasodilators, i.e., nitric oxide and prostacyclin, which also appear to have antiproliferative properties (7,8); in addition, endothelial ETBreceptors are involved in the clearance of ET-1, primarily in the vascular beds of the lungs, kidneys, and liver (9).
To date, bosentan, an ETA/ETBreceptor antagonist, is the only approved ET receptor antagonist for the treatment of PAH (10,11). Sitaxsentan sodium is a selective ETAreceptor antagonist, i.e., ETA:ETBratio >6,500:1, with a high oral bioavailability (>90%) and a long duration of action (half-life ∼10 h in PAH patients) (12). Although the Sitaxsentan To Relieve ImpaireD Exercise-1 (STRIDE-1) study demonstrated improved exercise capacity (as assessed by the six-min walk [6MW] test), World Health Organization (WHO) functional class (FC), cardiac index, and pulmonary vascular resistance at both sitaxsentan doses studied, i.e., 100 and 300 mg orally once daily, the safety profile was unacceptable with the 300-mg dose (13). The objective of this study was to confirm the efficacy of sitaxsentan and to determine the optimal dose on the basis of overall risk-to-benefit considerations, i.e., safety and tolerability with therapeutic efficacy. On the basis of the perception that 100 mg was the optimal dose in the STRIDE-1 study, the STRIDE-2 study was powered to show statistical difference of the primary efficacy end point between the sitaxsentan 100 mg and placebo groups. For observation only, an open-label (OL) (6MW tests, Borg dyspnea scores, and WHO FC assessments third-party blind) bosentan arm was included.
Patients in WHO FC II, III, or IV, 12 to 78 years of age, with symptomatic PAH despite treatment with anticoagulants, vasodilators, diuretics, cardiac glycosides, and/or supplemental oxygen (as clinically indicated) were eligible for study participation if they met the following criteria: 1) PAH that was idiopathic, associated with connective tissue disease, or associated with repaired atrial septal defect, ventricular septal defect, or patent ductus arteriosus at least one year before study enrollment; in addition, patients with PAH associated with an unrepaired secundum atrial septal defect (with a resting systemic arterial oxygen saturation of ≥88% in room air) were eligible; 2) mean pulmonary artery pressure ≥25 mm Hg at rest, pulmonary capillary wedge pressure or left ventricular end diastolic pressure ≤15 mm Hg, and pulmonary vascular resistance ≥3 Wood units (obtained by right heart catheterization within six months of study enrollment [to confirm the diagnosis of PAH]); and 3) baseline 6MW distance ≥150 m and ≤450 m. For patients <18 years of age, body weight was ≥50 kg. Patients were excluded if they had significant parenchymal lung disease, portal hypertension, chronic liver disease, human immunodeficiency virus infection, hepatic dysfunction (hepatic transaminase level >1.5× the upper limit of normal [ULN]), renal insufficiency, history of left-sided heart disease, history of obstructive sleep apnea (treated or untreated), previously failed bosentan due to safety or inadequate clinical response (in the judgment of the investigator), were on a prostaglandin, phosphodiesterase inhibitor, or an ET receptor antagonist, or had received any new type of PAH treatment within 30 days before study entry.
The study was conducted according to ethical principles stated in the Declaration of Helsinki (1996) and applicable guidelines on good clinical practice. The protocol was approved by local institutional review committees; written informed consent was obtained from all patients.
Study design and randomization
The study was an international, multicenter, randomized, double-blind, placebo-controlled trial that treated 245 patients between July 2003 and January 2005, at 55 sites. Patients were randomized to receive placebo, sitaxsentan 50 mg, or sitaxsentan 100 mg orally once daily, or OL bosentan for 18 weeks. The OL bosentan arm was included for observation only. The bosentan arm was OL at the standard dose, i.e., 62.5 mg orally twice daily for four weeks, then increasing to the maintenance dose of 125 mg orally twice daily. Because bosentan is only available commercially on a named-patient basis, blinded drug supplies were not available. Patients randomized to bosentan received OL drug (site personnel obtained medication through commercial channels utilizing the bosentan Access Program); however, the efficacy rater was blinded for the 6MW tests, WHO FC assessments, and Borg dyspnea scores, i.e., third-party blind. Randomization was performed centrally according to a computer-generated random number table. Randomization was 1:1:1:1. At the end of the 18-week study, all patients who completed the 18-week trial were eligible to enter an extension study. Patients who had been randomized, double-blind, to sitaxsentan 50 mg or 100 mg in the STRIDE-2 study received sitaxsentan 100 mg in the extension; patients who had been randomized to OL bosentan in the STRIDE-2 study continued OL bosentan in the extension. Patients who had been randomized, double-blind, to placebo in the STRIDE-2 study were randomized to either sitaxsentan 100 mg or OL bosentan in the extension. The STRIDE-2 randomization for patients randomized to sitaxsentan 50 mg or 100 mg in the STRIDE-2 study remained double-blind until the STRIDE-2 study database was locked; the STRIDE-2 randomization for patients randomized to placebo in the STRIDE-2 study and randomized to sitaxsentan 100 mg in the extension study remained double-blind until the STRIDE-2 study database was locked. Patients receiving OL bosentan remained third-party blind throughout the STRIDE-2 study and in the extension until the STRIDE-2 study database was locked.
Patients were evaluated at baseline and at weeks 6, 12, and 18. The primary end point was change from baseline in 6MW distance at week 18. Secondary end points included change in WHO FC, time to clinical worsening, and change in Borg dyspnea score. Safety was assessed by adverse events and laboratory evaluations. Patients receiving sitaxsentan or placebo were discontinued if they had an elevation in alanine aminotransferase (ALT) or aspartate aminotransferase (AST) >5× ULN, or if they had an elevation in total bilirubin >2× ULN with an ALT or AST >3× ULN. For OL bosentan patients, liver function abnormalities were handled according to the bosentan package insert, i.e., a patient was to be discontinued if ALT or AST exceeded 8× ULN, although the patient could undergo bosentan dosage reductions or temporary cessation for lesser elevations in ALT or AST at the discretion of the investigator. Investigators had the option to discontinue patients from the study if patients had a combined deterioration in WHO FC and ≥15% decrease in 6MW distance from baseline. Patients completing the 18-week study (all groups) or prematurely discontinuing (50-mg or placebo groups) were eligible to enter the STRIDE-2X extension study.
Data are reported as mean ± SD unless otherwise stated. The STRIDE-2 study was powered to identify statistical differences in efficacy between the 100-mg sitaxsentan and placebo groups. Comparisons between sitaxsentan and OL bosentan were observational only because of the inability to double-blind the bosentan arm. As defined in the protocol, for reporting statistical significance of efficacy, adjustments were made for multiple comparisons of multiple end points in treatment groups. Statistical significance was reported only if the conditions for the multiple comparison strategy were met; a p value <0.05 was considered statistically significant.
The primary end point was the change in 6MW distance from baseline to week 18. On the basis of the STRIDE-1 study data, an approximate 35-m difference between the 100 mg sitaxsentan and the placebo treatment groups, with a standard deviation of 58 m in each treatment group in the change from baseline in 6MW distance at week 18, was used to calculate the sample size of approximately 60 subjects per treatment group to detect statistically significant differences with 90% power at the significance level of 0.05 using a two-sided Student ttest. The null hypothesis of no difference between treatment groups was tested using the non-parametric analysis of covariance (ANCOVA), adjusted for baseline values. All missing values were imputed using the last observation carried forward method. If the patient was unable to perform the 6MW test because of clinical worsening or death, the 6MW distance was assigned a value of 0 m with a Borg dyspnea score of 10 for the visit.
Secondary end points were change from baseline to week 18 in WHO FC, time to clinical worsening, and Borg dyspnea score. Change from baseline in WHO FC was summarized with frequency counts and percentages, and treatment difference was analyzed using the Cochran-Mantel-Haenszel test, controlling for baseline WHO FC. The Kaplan-Meier method was used to estimate time to clinical worsening. Time to clinical worsening was defined as the number of days between the first dose date and the first date when clinical worsening occurred. Patients who discontinued or completed the study without having a clinical worsening event were censored at their last visit. Clinical worsening was defined as any of the following: hospitalization for PAH, death, transplantation, atrial septostomy, initiation of new chronic PAH treatment, or combined WHO FC deterioration and ≥15% decrease in 6MW distance from baseline. Treatment differences were evaluated using log-rank tests. The change in Borg dyspnea score from baseline to week 18 was analyzed using the non-parametric ANCOVA, adjusted for baseline values.
All efficacy analyses were conducted on the intent-to-treat (ITT) population, which was defined as all randomized patients who took any dose of study drug and performed at least one valid post-baseline 6MW test.
An independent external Data Safety Monitoring Board (DSMB) conducted an interim safety evaluation on the basis of adverse events and laboratory data after 60 patients completed week 12 assessments. In addition, the DSMB reviewed serious adverse events and elevated hepatic transaminases by blinded, masked treatment on an ongoing basis. Only the DSMB was authorized to request unblinding of safety data by treatment groups.
Two hundred forty-seven patients were randomized, 245 patients treated: 62 patients received placebo, 62 patients received sitaxsentan 50 mg, 61 patients received sitaxsentan 100 mg, and 60 patients received OL bosentan. Five patients did not have a valid post-baseline 6MW test; i.e., there were 240 patients in the ITT population for efficacy end points. Thirty-one patients prematurely discontinued the study (Fig. 1).The primary reason given for discontinuation in the placebo group (n = 11) was elevation of AST or ALT >3× ULN and total bilirubin >2× ULN in one patient, WHO FC deterioration and ≥15% decrease in 6MW distance in two patients, initiation of new chronic PAH treatment in five patients, adverse event in two patients, and sponsor decision in one patient. In the sitaxsentan 50 mg group, eight patients discontinued: WHO FC deterioration and ≥15% decrease in 6MW distance in one patient, initiation of new chronic PAH treatment in three patients, adverse event in one patient, patient decision in two patients, and investigator decision in one patient. In the sitaxsentan 100 mg group, four patients discontinued: elevation of AST or ALT >5× ULN in one patient, initiation of new chronic PAH treatment in two patients, and patient decision in one patient. In the OL bosentan group, eight patients discontinued: elevation in ALT or AST >8× ULN in two patients, WHO FC deterioration and ≥15% decrease in 6MW distance in one patient, and initiation of new chronic PAH treatment in five patients. The patients who prematurely discontinued the study were not significantly different than the overall study population, judging from demographic and clinical characteristics at study enrollment.
Patient characteristics for study entry are shown in Table 1.Overall, PAH was idiopathic PAH in 59% of patients, associated with connective tissue disease in 30% of patients, and associated with congenital heart disease (as defined in the inclusion criteria) in 11% of patients. Baseline characteristics were similar for the four treatment groups, with the exception of baseline 6MW distance, age, and height. The distance of 6MW at baseline, therefore, was used as a covariate for efficacy analyses.
At week 18 (Fig. 2),the change from baseline in 6MW distance was 17.8 m in the sitaxsentan 50 mg group, 24.9 m in the sitaxsentan 100 mg group, and 23.0 m in the OL bosentan group. In contrast, a decrease of 6.5 m from baseline occurred in the placebo group at week 18, i.e., the placebo-subtracted treatment effects at week 18 in the sitaxsentan groups were 24.2 m (p = 0.07) for the 50-mg group and 31.4 m (p = 0.03; 95% confidence interval 5.37, 57.44) for the 100-mg group. For the OL third-party blind bosentan-treated patients, the placebo-subtracted treatment effect at week 18 was 29.5 m (p = 0.05).
At week 18 (Fig. 3),the change from baseline in WHO FC was significantly better for the sitaxsentan 100-mg group compared with those receiving placebo (p = 0.04), i.e., 98% of the sitaxsentan 100-mg patients improved WHO FC (13%) or remained unchanged (85%), whereas 87% of the placebo patients improved WHO FC (10%) or remained unchanged (77%). No significant improvements were seen after 18 weeks in WHO FC versus placebo in either the sitaxsentan 50-mg group or the OL bosentan group. Worsening in WHO FC at week 18 (vs. baseline) occurred in 13% of placebo patients, 13% of sitaxsentan 50-mg patients, 2% of sitaxsentan 100-mg patients, and 9% of OL bosentan patients.
Time to clinical worsening
Twenty-nine patients (10 placebo, 6 sitaxsentan 50 mg, 4 sitaxsentan 100 mg, and 9 OL bosentan patients) experienced at least one clinical worsening event; the most frequently reported reasons for clinical worsening were initiation of new chronic PAH treatment (18 patients); combined deterioration in WHO FC and ≥15% decrease in 6MW distance (11 patients); and hospitalization for worsening PAH (10 patients) (Table 2).During the 18-week study, sitaxsentan 100 mg trended toward an increase in the time to clinical worsening versus placebo (p = 0.08); there were no differences for either the sitaxsentan 50-mg group (p = 0.27) or the OL bosentan group (p = 0.80) compared with patients receiving placebo.
Borg dyspnea score
Change from baseline to week 18 in Borg dyspnea score in the sitaxsentan 100-mg group showed a nonsignificant improvement (reduction in score), while the placebo group worsened (−0.01 vs. 0.19; p = 0.6603). Changes from baseline to week 18 in Borg dyspnea score in the sitaxsentan 50-mg and OL bosentan groups also were not significantly different from patients receiving placebo.
The number of patients reporting adverse events in each treatment group was similar; i.e., 90% of placebo patients, 86% of sitaxsentan 50-mg patients, 93% of sitaxsentan 100-mg patients, and 89% of OL bosentan patients experienced at least one adverse event. Treatment-related adverse events were reported in 29% of placebo patients, 36% of sitaxsentan 50-mg patients, 46% of sitaxsentan 100-mg patients, and 56% of OL bosentan patients. Serious adverse events occurred most often in placebo patients. Treatment-related side effects are listed in Table 3;edema, nasal congestion, fatigue, and insomnia were more frequently reported by sitaxsentan patients than by placebo patients, with a frequency that appeared to be dose-related (Table 3). Eighteen patients discontinued the STRIDE-2 study because of an adverse event: six (10%) placebo patients, four (7%) sitaxsentan 50-mg patients, two (3%) sitaxsentan 100-mg patients, and six (10%) OL bosentan patients.
Serious adverse events were more than twice as frequent in the placebo group (31%) compared with any of the other three groups, and were most often related to PAH worsening. Two patients in the placebo group died (due to multiple-organ failure), 5 and 15 days after premature discontinuation from the study. Serious adverse events of elevated hepatic transaminases considered related to a study drug occurred in one sitaxsentan 100-mg patient and in one OL bosentan patient. Bleeding adverse events, most commonly epistaxis, were reported in three placebo patients, eight sitaxsentan 50-mg patients, six sitaxsentan 100-mg patients, and four OL bosentan patients. None of the bleeding events were considered serious or required hospitalization, transfusion, or treatment.
Liver abnormalities have been recognized as a class effect associated with endothelin receptor antagonists (10,11,13,14). The incidence of liver enzyme abnormalities (hepatic transaminase values >3× ULN) was 6% for the placebo group, 5% for those receiving sitaxsentan 50 mg, 3% for the sitaxsentan 100-mg patients, and 11% for the OL bosentan group. Elevations in liver enzymes were associated with drug discontinuation in one placebo patient, one 50-mg patient (>3× ULN, with the reason for discontinuation reported as investigator decision), one 100-mg patient, and two OL bosentan patients. Liver enzyme abnormalities reversed in all sitaxsentan and bosentan cases.
Because of the effect of sitaxsentan on inhibition of the CYP2C9 P450 enzyme, the principal hepatic enzyme involved in warfarin metabolism, an 80% reduction in the warfarin dose for patients not randomized to OL bosentan was used at baseline, with adjustment as needed to maintain a therapeutic international normalized ratio (INR). The proportion of patients with an INR >3.5 was comparable for the sitaxsentan and placebo groups (33% to 42%) and lower for the OL bosentan group (20%) (p = 0.1744). Dosage adjustments occurred with similar frequencies for the four groups (63 to 70 adjustments per group). Daily warfarin dosages at week 18 were: 3.7 ± 2.7 mg/day for placebo patients, 2.8 ±1.2 mg/day for 50-mg sitaxsentan patients, 2.1 ± 1.0 mg/day for 100-mg sitaxsentan patients, and 5.1 ± 2.9 mg/day for OL bosentan patients. Warfarin doses were similar to baseline for placebo patients, lower than baseline in the sitaxsentan patients, and higher than baseline in the bosentan patients.
Decreases in hemoglobin in sitaxsentan- and OL bosentan-treated patients were observed as early as week two and remained stable throughout the 18-week study (mean change from baseline to week 18: placebo, +0.2 g/dl; 50 mg sitaxsentan, −0.2 g/dl; 100 mg sitaxsentan, −0.5 g/dl; and OL bosentan, −0.5 g/dl). A small and similar number of patients, i.e., one to three patients, for each group shifted from the normal laboratory hemoglobin range to below the normal range by week 18.
The DSMB did not request that data be unblinded during the study.
In a preliminary, randomized, controlled PAH trial (STRIDE-1), although both 100- and 300-mg doses of sitaxsentan improved exercise capacity (as assessed by the 6MW test), WHO FC, cardiac index, and pulmonary vascular resistance compared with placebo, the incidence of elevated hepatic transaminases (>3× ULN), serious adverse events, and discontinuation were higher with the 300-mg dose, consistent with a dose response for safety and tolerability. On this basis, 100 mg was chosen as the highest dose to be studied in future trials.
On the basis of overall safety, tolerability, and efficacy, the STRIDE-2 study demonstrated that 100 mg is the optimal dose of sitaxsentan for treating symptomatic PAH. Treatment with sitaxsentan 100 mg significantly improved exercise capacity and WHO FC compared with placebo; 50 mg was subtherapeutic, and based on the results from the STRIDE-1 study, 300 mg was unacceptable for safety. Among the treatment-related side effects, edema, nasal congestion, fatigue, and insomnia were more frequently reported by the sitaxsentan-treated patients than by the placebo patients.
Patients with PAH are frequently anticoagulated; sitaxsentan inhibits warfarin metabolism via inhibition of CYP2C9, and bosentan induces CYP2C9 (15). Dosage adjustments in warfarin were equally frequent in all four groups, with mean daily warfarin doses lower for sitaxsentan-treated patients (in a dose-dependent manner) compared with patients receiving placebo, and higher for OL bosentan-treated patients compared with the placebo group. Although increases in INR >3.5 occurred in a significant number of patients receiving placebo or sitaxsentan, bosentan patients were relatively protected from this effect, consistent with enzyme induction of warfarin metabolism with bosentan. However, in all groups, concomitant therapy with warfarin was well tolerated, and bleeding events were rare, non-serious, and did not require treatment.
Liver function abnormalities have been previously reported with endothelin receptor antagonists and can lead to treatment discontinuation. In the STRIDE-1 study and its extension (median exposure 26 weeks), 5% of patients treated with sitaxsentan 100 mg had >3× ULN hepatic transaminases; all transaminase increases were reversible. Although the STRIDE-2 study was not powered to assess comparison between sitaxsentan 100 mg and OL bosentan, during this 18-week study, the incidence of elevated hepatic transaminases was 3% for sitaxsentan 100 mg and 11% for OL bosentan (i.e., the same as the rate reported in the bosentan package insert).
Pulmonary arterial hypertension is a rare and serious disease. Despite the scientific interest in comparing an ETAselective receptor antagonist with an ETA/ETBreceptor antagonist, it was not possible to conduct a trial of sufficient power (in a reasonable period of time) to directly compare the efficacy of sitaxsentan with bosentan. Only a trial of considerable size, with both treatments administered double-blind, could resolve this question with certainty. Because of a unique distribution program, it was not possible to double-blind the bosentan arm in this study.
Treatment with the selective ETAreceptor antagonist sitaxsentan (at a dose of 100 mg orally once daily) improves exercise capacity and WHO FC in symptomatic PAH patients, with a low incidence of hepatic toxicity.
The authors thank the additional STRIDE-2 Study Group investigators and staff members. For a complete list, please see the online version of this article.
↵1 Drs. Barst, Langleben, Badesch, Lawrence, Frost, Shapiro, Naeije, and Galie have served as consultants to the sponsor, Encysive Pharmaceutics, LP.
Although none of the authors has significant stock ownership or other equity interests or patent licensing arrangements that might pose a conflict of interest in connection with the submitted article
- Abbreviations and Acronyms
- alanine aminotransferase
- analysis of covariance
- aspartate aminotransferase
- Data Safety Monitoring Board
- functional class
- international normalized ratio
- open label
- pulmonary arterial hypertension
- six-min walk
- Sitaxsentan To Relieve ImpaireD Exercise study
- upper limit of normal
- World Health Organization
- Received September 20, 2005.
- Revision received December 23, 2005.
- Accepted January 9, 2006.
- American College of Cardiology Foundation
- Galie N.,
- Manes A.,
- Branzi A.
- Dupuis J.,
- Stewart D.J.,
- Cernacek P.,
- Gosselin G.