Article Text
Abstract
Background Although sildenafil has been shown to be safe and effective in idiopathic pulmonary arterial hypertension (PAH) and PAH related to connective tissue disease, its effects in Eisenmenger syndrome are less clear.
Objective To investigate whether long-term treatment (12 months) with the phosphodiesterase type 5 inhibitor sildenafil improves clinical and haemodynamic parameters in patients with Eisenmenger syndrome.
Design Prospective, open-label, multicentre study.
Setting Four pulmonary hypertension centres in China.
Patients 84 Eisenmenger syndrome functional class II–IV patients.
Interventions Oral sildenafil 20 mg orally three times a day.
Outcome measures 6-min walk distance (6MWD) test, resting systemic arterial blood oxygen saturation (SaO2) in room air, haemodynamic parameters assessed by right heart catheterisation, safety and tolerability.
Results The overall treatment effects at 12 months versus baseline (mean changes with 95% CIs) were 56 m increase (42 to 69, p<0.0001) in 6MWD, and 2.4% increase (1.8% to 2.9%, p<0.0001) in resting room air SaO2. Improvements were also seen in mean pulmonary arterial pressure and pulmonary vascular resistance index (−4.7 mm Hg (−7.5 to −1.9), p=0.001; and −474 dyn×s×cm−5×m2 (−634 to −314), p<0.0001, respectively). Sildenafil was well tolerated. Most adverse events were mild and transient, and occurred in the first 2 weeks of treatment.
Conclusions Twelve months of oral sildenafil treatment was well tolerated and appeared to improve exercise capacity, systemic arterial oxygen saturation and haemodynamic parameters in patients with Eisenmenger syndrome.
- Sildenafil
- phosphodiesterase inhibitors
- Eisenmenger syndrome
- pulmonary arterial hypertension
- pulmonary embolism
- pulmonary vascular disease
- deep vein thrombosis
- congenital heart disease
Statistics from Altmetric.com
- Sildenafil
- phosphodiesterase inhibitors
- Eisenmenger syndrome
- pulmonary arterial hypertension
- pulmonary embolism
- pulmonary vascular disease
- deep vein thrombosis
- congenital heart disease
Introduction
Pulmonary arterial hypertension (PAH) is a recognised complication of congenital heart disease. The most common congenital heart defect that can lead to PAH is a systemic to pulmonary shunt, such as a ventricular septal defect (VSD), persistent ductus arteriosus (PDA) or, less often, an atrial septal defect (ASD). Unrepaired, these patients can develop a predominantly right to left shunt accompanied by systemic arterial oxygen desaturation. Although Eisenmenger syndrome and idiopathic PAH have different aetiologies, there are parallels in histopathology, pathobiology and clinical characteristics.1 2 While the natural history of patients with Eisenmenger syndrome is better than those with hereditary or idiopathic PAH, despite similar haemodynamic characteristics, life expectancy is still decreased3 and Eisenmenger syndrome is associated with substantial long-term morbidity. Oya et al4 reported survival rates from the time of diagnostic cardiac catheterisation of 98%, 77% and 58% at 1, 5 and 10 years, respectively, in adult patients with Eisenmenger syndrome. In the past decade, PAH-specific drugs have been shown to be safe and effective in PAH, although the majority of patients studied had either idiopathic or hereditary PAH, or PAH associated with connective tissue disease. Patients with Eisenmenger syndrome have been excluded from most randomised clinical trials designed for PAH. However, BREATHE-5, the only randomised controlled trial performed to date in Eisenmenger syndrome, demonstrated the safety and efficacy of bosentan in this PAH subgroup.5
In human pulmonary artery smooth muscle cells, phosphodiesterase type 5 (PDE5) is the principal regulator of cyclic guanosine monophosphate hydrolysis and downstream signalling.6 The increase of PDE5 in small pulmonary artery endothelial cells of patients with PAH suggests that it may have an important role in the pathobiology of PAH.7 In an animal model of the Eisenmenger syndrome—that is, creation of an aorto-pulmonary shunt in newborn lambs, PDE5 expression was increased.8 PDE5, which is abundant in lung tissue, is upregulated in pulmonary hypertension, thus offering a molecular target for PAH treatment.6 In 2005, sildenafil was approved for the treatment of symptomatic PAH in Europe and North America. However, the majority of patients enrolled in clinical trials of sildenafil had either idiopathic PAH, hereditary PAH or connective tissue disease-associated PAH, and patients with Eisenmenger syndrome were excluded.9–11 From 2006, PDE5 inhibitors such as sildenafil and vardenafil were introduced into clinical use in mainland China for patients with PAH, and encouraging effectiveness has been identified in other PAH aetiologies in Chinese patients.12 13 To date, however, there has been little evidence that sildenafil is efficacious in Eisenmenger syndrome.11 14–17 The aim of this study was to evaluate the long-term effects of sildenafil in Eisenmenger syndrome.
Patients and methods
A 12-month, open-label uncontrolled study was conducted at four pulmonary vascular centres in China from April 2007 to July 2009. The protocol was approved by ethics committees at all centres. It was conducted in accordance with the ethical principles set out in the Declaration of Helsinki and with all applicable local laws and requirements for good clinical practice. Written informed consent was obtained from all patients before performing any study-related procedures.
Patients and procedures
Treatment-naïve patients with functional class II–IV Eisenmenger syndrome who were >14 years of age, and considered ‘clinically stable’ for at least 3 months by their treating doctors were eligible for enrolment. Eisenmenger syndrome was established echocardiographically as an unrepaired unrestricted PDA, VSD of ≥1 cm effective diameter or ASD of ≥2 cm effective diameter, or a combination of these defects associated with bidirectional shunting. Eisenmenger physiology was defined clinically—for example, systemic arterial oxygen saturation (SaO2) at rest in room air <95% with cyanosis and/or clubbing. The majority of patients (74%) had an SaO2 <90% (mean 86±5%) at enrolment, and the other 26% had an SaO2 between 90% and 94%. PAH was confirmed by right heart catheterisation: mean pulmonary arterial pressure ≥25 mm Hg, mean pulmonary capillary wedge pressure ≤15 mm Hg and pulmonary vascular resistance >240 dyn×s×cm−5. Conventional treatments were continued during the study, including diuretics, warfarin and digoxin.
Patients were excluded if they had complex congenital heart defects, left ventricular dysfunction (left ventricular ejection fraction <40%), restrictive lung disease (total lung capacity <70% of predicted), obstructive lung disease (forced expiratory volume in 1 s (FEV1) <70% of predicted with a FEV1/forced vital capacity <60%), chronic thromboembolic pulmonary hypertension or coronary artery disease. In all eligible patients, physical examination, functional class assessment, routine blood testing, a 6 min walk distance (6MWD) test, electrocardiogram, echocardiogram and right heart catheterisation were performed.
Sildenafil (Pfizer Pharmaceuticals, New York, NY, USA) was given to all patients at a fixed dosage of 20 mg orally three times a day, and was continued long term unless intolerable adverse events occurred. All patients had follow-up assessments every 3–6 months as long as their functional class did not deteriorate. Repeat right heart catheterisation was recommended for patients with progressive right heart failure or as routine follow-up every 12 months.
The primary efficacy end points of the study were the 6MWD test and resting SaO2 in room air, while secondary end points included haemodynamic parameters, functional class and serum uric acid. Episodes of clinical worsening (defined as any of the following: death, heart or lung transplantation, hospitalisation for PAH and initiation of PAH-specific treatments such as prostanoids or endothelin receptor antagonists) and adverse events were recorded throughout the study.
Right heart catheterisation
Right heart catheterisation was performed under fluoroscopy at baseline and after 12 months via the right internal jugular or left antecubital veins using a 7F or 7.5F Swan–Ganz catheter. Haemodynamic parameters measured included mean right atrial pressure (mRAP), pulmonary arterial pressure (PAP), mean pulmonary arterial pressure (mPAP), mean pulmonary capillary wedge pressure (mPCWP), mean systemic arterial pressure (mSAP), superior vena cava oxygen saturation, pulmonary arterial oxygen saturation and SaO2. Pulmonary venous oxygen saturations were assumed as 98%. Pulmonary (Qpi) and systemic blood (Qsi) flow indexes were calculated by the Fick method using standard formulae and an assumed oxygen consumption index of 125 ml/min/m2. Pulmonary vascular resistance index (PVRi) and systemic vascular resistance index (SVRi) were calculated by the following formulae:
Statistical analysis
A statistical software package (SPSS, V.13.0; SPSS) was applied to data management. Baseline demographic variables, exercise capacity, haemodynamic parameters, serum uric acid and plasma haemoglobin were determined as mean±SD values. The treatment effects on exercise capacity, haemodynamic parameters, serum uric acid and plasma haemoglobin were determined as means with 95% CIs. Gender and functional class were determined as a number and ratio. Paired t tests were used to compare differences between values at baseline and 12 months. A χ2 test (Fisher's exact test) was used to compare differences in functional class distribution between assessments at baseline and 12 months. Independent t tests were performed for ASD and VSD/PDA subgroup analyses at baseline. For all comparisons, p<0.05 was considered significant.
Results
A total of 100 patients with Eisenmenger syndrome were screened. Nine patients with complex congenital heart malformation were excluded and another seven were lost to follow-up. Of the remaining 84 patients, 58 were female and 26 were male with an overall mean age of 28±9 years (range 14–56 years). The patients' aetiologies included: 25 ASD, 34 VSD, 23 PDA, and 2 VSD in combination with a PDA. All patients were unrepaired. Compared with post-tricuspid shunts (VSD or PDA), patients with pre-tricuspid shunts (ASD) had a significantly higher resting SaO2 (89±4% vs 85±5%, respectively; p=0.001) and significantly lower mPAP (70±19 mm Hg vs 83±18 mm Hg, respectively; p=0.004) (table 1).
All patients were functional class II–IV at baseline, with more than 50% (n=44) class II. The mean duration of sildenafil exposure was 12.3±4.3 months.
Exercise and functional capacity
After 12 months of oral sildenafil, the 6MWD increased from 430±101 m at baseline to 486±94 m at 12 months (p<0.0001) (figure 1), and the Borg dyspnoea score decreased from 3.2±1.9 at baseline to 2.2±1.4 at 12 months (p<0.0001) (table 2). SaO2 at rest in room air increased from 86.4±5.2% at baseline to 88.8±4.5% at 12 months (p<0.0001) (table 2), with more patients in a better functional class at 12 months than at baseline (figure 2). The proportions of functional class I, II, III and IV patients were 0%, 53%, 39% and 8%, respectively, at baseline, and 8%, 81%, 10% and 1%, respectively, at 12 months (p<0.001).
Biochemical markers and haemodynamic parameters
As shown in table 2, there were no significant changes in serum uric acid concentrations (387±103 μmol/l to 372±107 μmol/l, p=0.139). Plasma haemoglobin levels decreased from 169±32 g/l to 162±31 g/l (p=0.002).
Changes in haemodynamic parameters were remarkable for improvements in mPAP, Qpi and PVRi at 12 months (table 2). Mean PAP decreased from 79±19 mm Hg at baseline to 74±168 mm Hg at 12 months (p=0.011), Qpi increased from 2.5±0.8 l/min/m2 to 3.1±1.9 l/min/m2 (p=0.001) and PVRi decreased from 2580±1177 dyn×s×cm−5×m2 to 2106±9 dyn×s×cm−5×m2 (p<0.0001). Qp/Qs increased from 0.9±0.3 to 1.1±0.5 (p=0.002). Neither mSAP nor SVRi changed significantly with sildenafil (82±10 mm Hg to 81±9 mm Hg, p=0.248; and 2345±828 dyn×s×cm−5×m2 to 2211±845 dyn×s×cm−5×m2, p=0.099, respectively). However, the PVRi/SVRi ratio was significantly decreased from 1.17±0.56 to 1.05±0.63 (p=0.033).
PAH subgroups
Patients with both pre-tricuspid shunts (ASD) and post-tricuspid shunts (VSD and/or PDA) showed improvements in 6MWD after 12 months (increased from 443±92 m at baseline to 491±79 m in patients with ASD, p=0.001; and from 425±104 m to 483±100 m in patients with VSD and/or PDA, p<0.0001) (figure 1) with decreases in the Borg dyspnoea score (3.4±1.8 to 2.2±1.4 in patients with ASD, p=0.002; and 3.1±2.0 to 2.2±1.4 in patients with VSD and/or PDA, p=0.001). SaO2 was increased both in patients with ASD (from 89.0±3.5% to 90.3±3.3%, p<0.0001) and patients with VSD and/or PDA (from 85.3±5.5% to 88.2±4.8%, p<0.0001).
Haemodynamic responses were similar in patients with ASD (mPAP decreased from 70±19 mm Hg to 65±16 mm Hg, p=0.022; Qpi increased from 2.4±0.6 l/min/m2 to 2.9±0.9 l/min/m2, p=0.011 and PVRi decreased from 2271±879 dyn×s×cm−5×m2 to 1804±791 dyn×s×cm−5×m2, p=0.002) and in patients with VSD and/or PDA (mPAP decreased from 83±18 mm Hg to 78±17 mm Hg, p=0.016; Qpi increased from 2.6±0.8 l/min/m2 to 3.2±2.2 l/min/m2, p=0.009 and PVRi decreased from 2711±1267 dyn×s×cm−5×m2 to 2234±1215 dyn×s×cm−5×m2, p<0.0001). There was no significant effect on mSAP at 12 months in the ASD, VSD and/or PDA subgroups.
Clinical worsening episodes and adverse events
No patients died or received heart or lung transplants during the study. In eight patients (9.5%), six of whom were WHO functional class III and two class II, 10 hospitalisations were necessary, six due to pneumonia, two due to haemoptysis and two due to progressive right ventricular dysfunction. The two patients who had progressive right ventricular dysfunction were treated with intravenous dopamine, diuretics and oxygen in hospital. However, they continued with sildenafil monotherapy after discharge from hospital for economic reasons.
The most common adverse events reported during sildenafil treatment were headache (51%) and flushing (43%). Less common adverse events were nasal congestion (6%), diarrhoea (2%), dyspepsia (2%), limb pain (2%), skin rash (1%) and myalgia (1%). Most adverse events were mild and transient, and most occurred in the first 2 weeks of treatment.
Discussion
This is the first prospective evaluation of sildenafil treatment in a large cohort of patients with Eisenmenger syndrome. In this cohort, sildenafil improved exercise capacity (as assessed by the 6MWD test), functional class and haemodynamic parameters without worsening SaO2. Moreover, sildenafil was safe and well tolerated, suggesting that there may be benefit from its long-term use. Nevertheless, it must be taken into account that these data are open label and uncontrolled, which is a study limitation.
The 6MWD test has been shown to be an independent predictor of mortality in patients with idiopathic PAH18 and Eisenmenger syndrome, and has been used as a primary efficacy end point in many randomised clinical trials in PAH.19 The increase in 6MWD at 12 months compared with baseline was 56 m, and was similar to that achieved with oral bosentan in the placebo-controlled BREATHE-5 study in Eisenmenger syndrome (functional class III),5 and also similar to that achieved with sildenafil in the placebo-controlled SUPER-1 study.11 Both these studies were placebo controlled and therefore of greater rigour; however, in contrast to 12 months in our study, the evaluation period was 16 weeks in BREATHE-5 and 12 weeks in SUPER-1. Additionally, In BREATHE-5, a decrease in the 6MWD was seen in the placebo-treated patients with Eisenmenger syndrome after 16 weeks, suggesting that patients with Eisenmenger syndrome who are functional class III tend to deteriorate without PAH-specific treatment, even over a relatively short time—namely, 16 weeks. Furthermore, the increase of 6MWD (56 m) at 12 months in our study was similar to the results of the BREATHE-5 open-label extension study (in which the overall improvement from baseline was 61±8 m for the ex-bosentan group at 40 weeks).20 Therefore, our 12-month, open-label uncontrolled study is consistent with continued efficacy of bosentan treatment.
Haemodynamic parameters, especially PVRi and SaO2, were critical markers that reflected the treatment effects. In our study, the PVRi was significantly decreased at 12 months by 474 dyn×s×cm−5×m2, which was similar to the reduction in the BREATHE-5 study (472 dyn×s×cm−5×m2 decrease from baseline at week 16). The haemodynamic improvement with sildenafil can be explained by the fact that it may cause vasodilatation, reversal of vascular remodelling, enhancement of right ventricular contractility, or a combination of these effects.21–23
In our study, a substantial increase of SaO2 at rest with long-term sildenafil treatment was identified. Improvement of SaO2 in patients with Eisenmenger syndrome has also been seen during tadalafil administration, another PDE5 inhibitor, both acutely (90 min after tadalafil administration) and at 12 weeks' follow-up.24 In contrast, there was no improvement with bosentan treatment in BREATHE-5 and its extension study. This difference might be explained by the preferential distribution of the PDE5 enzyme, which is located primarily in the penis and lungs,25 26 and leads to a reduction of right-to-left shunting and augmented pulmonary blood flow owing to a reduction in the PVRi/SVRi ratio.
Our study is also the first to examine potential differences in pre-tricuspid shunting defects (ASD) versus post-tricuspid shunting defects (VSD/PDA). PAH associated with an ASD, particularly a small ASD, is often viewed as idiopathic PAH, although the classification of these small defects remains controversial. Our data suggest that patients with PAH associated with ASDs or VSDs/PDAs have comparable improvements in 6MWD, SaO2 and haemodynamic parameters with sildenafil treatment.
PAH-specific treatments for patients with Eisenmenger syndrome aim, at a minimum, to stabilise disease without significant adverse events and preferably, improve the clinical status and outcome. With the introduction of new treatments for PAH in recent years, improvements in long-term outcomes have, as a consequence, been seen in patients with adult idiopathic PAH,27 paediatric PAH28 and Eisenmenger syndrome.29 Dimopoulos and colleagues in the UK reviewed all patients with Eisenmenger syndrome seen at their centre and found that patients treated with PAH-specific treatments had a lower risk of death than patients who did not receive advanced treatments.29 However, in their cohort, only a small subgroup of patients (25%) was treated with sildenafil and no detailed evidence of its therapeutic efficacy could be obtained. As current therapeutic options for patients with Eisenmenger syndrome remain limited, the apparent efficacy of sildenafil treatment reported in this study is useful, despite its open-label, uncontrolled design. However, as we know, the treatment effect might vary in different individuals. Therefore, investigations of markers for predicting the long-term response to PAH-specific treatments would be helpful to better determine therapeutic strategies. Recently, D'Alto and colleagues found that pulmonary vasoreactivity produced by an intravenous epoprostenol challenge was a significant predictor of clinical worsening in a small cohort of patients with congenital heart disease associated with PAH or Eisenmenger syndrome.30 This finding might be explained by vasoconstriction rather than the involvement of irreversible pulmonary vascular remodelling in progressive elevations of pulmonary vascular resistance in these patients. The clinical value of such predictors needs to be further investigated in a larger cohort of patients to verify whether patients with different acute responses to vasodilators need to receive more specific treatments.
Limitations
The open-label uncontrolled nature of our study is a significant limitation. Although effective, bosentan was not an option in our patients as its costs could not be reimbursed by either insurance claims or by the government. Hence, we could not perform a randomised controlled study with background bosentan treatment in our patients with Eisenmenger syndrome. Consequently, we performed the current open-label study as a ‘proof-of-concept’ investigation.
Conclusion
Long-term treatment with sildenafil is safe and well tolerated and seems to be efficacious in patients with Eisenmenger syndrome. Improvements in exercise capacity, resting systemic arterial oxygen saturation, haemodynamics and functional class are seen. These findings suggest that sildenafil may be a useful treatment option in Eisenmenger syndrome, particularly as it is a relatively low-cost drug, an important factor in developing countries.
Acknowledgments
We thank Professor Zi-Sheng AI for his statistical help with our study.
References
Footnotes
Z-NZ was formerly a visiting fellow in the Department of Cardio-Pulmonary Circulation, Shanghai Pulmonary Hospital, Tongji University School of Medicine. She is now working in the Pulmonary Vascular Department, Beijing Shijitan Hospital, Beijing, China.
Z-NZ, XJ and RZ contributed equally to the manuscript. All authors participated in the study, and in the review and approval of the manuscript.
Funding The study was sponsored by the Shanghai New Frontier Project (SHDC12010102), the Program for Young Excellent Talents in Tongji University (2009KJ042) and China National 973 Project (2007CB512008).
Competing interests Z-CJ has served as a consultant on steering committees for clinical trials and has been a member of scientific advisory boards for Actelion, Bayer-Schering, Pfizer and United Therapeutics; these companies provided funding to Tongji University to support his conduct of clinical trials. MG-M has served as a consultant/participant on data safety monitoring boards/steering committees for clinical trials for Actelion, Gilead, Medtronic and Pfizer. Actelion, Gilead, Lilly/Icos, Pfizer, Novartis and United Therapeutics provided funding to the University of Chicago to support her conduct of clinical trials. None of the other authors has any conflict of interest to declare regarding the content of this paper.
Patient consent Obtained.
Ethics approval Ethics Committees of all participating centres.
Provenance and peer review Not commissioned; externally peer reviewed.