Article Text
Abstract
Objective: To determine the effects of recently available oral therapies, ie, endothelin receptor antagonists (ERAs) and phosphodiesterase-5 inhibitors (PDEIs), in patients with pulmonary arterial hypertension related to connective tissue disease (CTD), mostly systemic sclerosis (SSc).
Method: A systematic literature search was conducted up to April 2007. All randomised controlled trials evaluating the efficacy of bosentan, sitaxsentan and sildenafil vs placebo on exercise capacity were selected. Effect size was calculated in each study to assess the magnitude of treatment effect.
Results: In all, 10 studies were analysed, giving a total of 613 participants (186 with CTD) who received the active treatment and 272 (72 with CTD) who received placebo. The effect sizes of bosentan, sitaxsentan and sildenafil for exercise capacity in the CTD subset of patients were non-significant; 0.31 (95% confidence interval (CI) −0.22 to 0.83), 0.26 (95% CI −0.06 to 0.57) and 0.53 (95% CI −0.02 to 0.89), respectively. In the whole PAH population, these values were significant; 0.61 (95% CI 0.38 to 0.84), 0.33 (95% CI 0.15 to 0.51) and 0.58 (95% CI 0.38 to 0.79), respectively.
Conclusion: This meta-analysis suggests an absence of clinically relevant improvement on exercise capacity in patients with CTD/SSc after 12 to 18 weeks of treatment. A poor therapeutic response, insufficient power of studies or poor sensitivity to change of the 6-min walk test may explain these results. The promising preliminary data on survival of ERAs and the confounding effects of other comorbidities associated with CTD and SSc may support the latter hypothesis.
Statistics from Altmetric.com
Pulmonary arterial hypertension (PAH) is a devastating condition characterised by an increase in pulmonary vascular resistance, which leads to right ventricular failure and, ultimately, death. This disease can be classified into five categories according to the Venice classification,1 including idiopathic PAH (IPAH) and PAH associated with other diseases and conditions. IPAH is the prototypic form and most commonly studied subtype of PAH in phenotypic analyses and clinical trials.2 3 Recently, there has been increasing interest in PAH related to systemic sclerosis (SSc), which is the connective tissue disease most often associated with PAH.4
SSc is a heterogeneous systemic disorder characterised by alterations of the microvasculature, by disturbances of the immune system and by the massive deposition of collagen and other matrix substances in connective tissue. Since the improvement of the prognosis of renal crisis by angiotensin-converting enzyme (ACE) inhibitors, PAH has emerged as a leading cause of death in SSc.5 Estimates of the prevalence of PAH in patients with SSc depend on the criteria used for diagnosis and the method of measurements (cardiac echography or right heart catheterisation). Using cardiac catheterisation, the prevalence of pulmonary hypertension appears to be 8–12%.6 7 Subsequent data suggest that PAH related to SSc is associated with poorer long-term survival compared with the idiopathic form of PAH (IPAH).8 9
Only continuous epoprostenol treatment has demonstrated benefits in improving exercise capacity in PAH specifically related to SSc.10 However, prostacyclin is delivered via continuous intravenous infusion, which may lead to severe side effects such as catheter infection, catheter thrombosis, and paradoxical embolism. Furthermore, pump or catheter malfunction may result in life-threatening loss of haemodynamic effects and prolonged therapies to tachyphylaxis.10 Recent clinical data showed that analogues of prostacyclin were effective in patients with PAH at least for the short term (12 weeks), and have fewer severe side effects.11 However, long-term effects remain to be confirmed.
New oral treatments, as endothelin receptor antagonists (ERAs) and phosphodiesterase-5 inhibitors (PDEIs) have been developed based on improved understanding of SSc pathology as an alternative treatment targeting patients with early disease and functional class II or III. These drugs have proven efficacy over 12 to 18 weeks in improving exercise capacity in patients mainly with IPAH.12 13 However, many data suggest a poorer response to treatment 14 in CTD/SSc subgroups. Unfortunately, no direct comparison between PAH related to SSc and IPAH is available to date.
The objective of this study was to assess the efficacy of oral treatments on exercise capacity, used as the primary outcome in most of the studies in PAH related to CTD, and more particularly SSc. To this end, a meta-analysis of randomised controlled trials (RCTs) was performed.
PATIENTS AND METHODS
Literature research
A literature search was performed on all articles published between 1966 and April 2007 and expanded on Medline, EMBASE, the Cochrane Controlled Trials Register of RCTs and the European League Against Rheumatism (EULAR)/American College of Rheumatology (ACR) congress abstract archives. There was no language of publication restriction. The search was conducted using the following combination for endothelin receptor antagonists: “endothelin receptor antagonists OR bosentan OR sitaxsentan” AND “pulmonary arterial hypertension {MAJR}” and for phosphodiesterase-5 inhibitors: “phosphodiesterase inhibitors OR sildenafil OR tadalafil OR vardenafil” AND “pulmonary arterial hypertension {MAJR}”. In addition, reference lists of the papers initially detected were searched by hand to identify additional relevant reports.
Inclusion criteria
Study design
Randomised controlled trials vs placebo of PAH oral treatments.
Study population
Adults with PAH (idiopathic, secondary to CTD or congenital heart disease) who required medical treatment for their condition and who were anticoagulated. As patients with SSc were the most represented in the CTD subgroup, we extrapolated the results obtained in the CTD subgroup to patients with SSc.
The diagnosis of pulmonary arterial hypertension, defined as mean pulmonary artery pressure greater than 25 mmHg at rest or 30 mmHg with exercise, was based on clinical symptoms with a combination of pulmonary and cardiac imaging and the results of right-heart catheterisation.
Intervention
Any PAH oral treatment of the two pre-defined therapeutic class (ERAs and PDEI) taken alone and compared to placebo. The drugs taken into account were bosentan (endothelin A and B receptor inhibitors), sitaxsentan (selective endothelin A receptor inhibitor) and sildenafil (PDEI).
Outcome measure
Exercise capacity assessed by the 6-min walk test. If an article reported no interpretable results, it was not analysed.
Methodological quality
The articles that fulfilled the inclusion criteria underwent quality appraisal. We used the impact factor of the journal in which the trial was published to assess the quality of RCTs, and it was checked whether the statistics used intention-to-treat analysis and whether the Jadad scale was applied,15 which contains two questions for randomisation and masking and one question evaluating the reporting of withdrawals and drop-outs. Each question entails a yes or no response option. In total, 5 points can be awarded, with higher scores indicating superior quality. Data were also extracted regarding funding sources for the studies.
Data extraction in included studies
Data extraction was performed by two reviewers (JA, JW) on the full texts, not blinded to author and journal, using a predefined extraction sheet (available from the authors on request). Information extracted included first author, publication year, mean age of participants, sex proportion, trial duration, type of PAH treatment, type of comparator, drug dose, number of patients in active and control group, and the outcome measure used to assess efficacy, defined a priori.
Statistical analysis
To measure the magnitude of the treatment effect for the continuous variable, the effect size was calculated. The effect size is a standard way to determine the degree of improvement of a particular treatment after any placebo effect has been accounted for. The effect size is calculated as the ratio of the treatment effect (mean differences in treatment group minus differences in placebo group) to the pooled standard deviation of these differences.16 This calculation entails the use of means for baseline and final data with a measure of variability such as SD. Every effort was made to calculate the effect size in all studies. If the SD was given in only one group it was used as baseline SD for both groups. However if no measure of variability was given the effect size could not be extrapolated. By convention, an effect size <0.2 is usually considered as trivial; >0.2–0.5 as small; >0.5–0.8 as moderate; >0.8–1.2 as important and >1.2 as very important.17 Minus or plus signs indicate direction of difference, not magnitude of difference. An effect size may be considered as significant as its confidence interval does not include zero. Thus, it is reasonably certain that a statistically significant change occurred.
Primary analyses examined the effect size vs placebo of ERAs and PDEIs for exercise capacity in PAH related to the subset of patients with CTD (mainly patients with SSc), as compared to the whole population of patients. Sensitivity analyses were calculated within subgroups decided a priori (nature of the drug, treatment dose, functional class, methodological quality) to assess the robustness of the main conclusions. Quality was analysed as a binary variable: studies scoring 3 or more on the Jadad scale were considered to be of high quality; 2 or less, of low quality. A sensitivity analysis was also performed with the upper limit for low quality changed to 3, then to 4.
Statistical heterogeneity was tested by Q test (χ2).18 All meta-analyses were carried out with use of fixed-effects model, or random-effects model in case of significant heterogeneity.
RESULTS
Included studies
The results of the article selection process are reported in fig 1. From the 544 articles identified, 119 12 19–27 were included since they reported RCTs vs placebo. One study was eventually excluded as no suitable data were provided to calculate the effect size. Four studies21 24 26 27 were subgroup analysis from the pivotal studies of bosentan, sitaxsentan and sildenafil in the subset of patients with CTD, mainly patients with SSc (59%). One study was a subgroup analysis from the Sitaxsentan to Relieve Impaired Exercise (STRIDE)-1 study in a patient population with pulmonary arterial hypertension that met traditional inclusion criteria of previous pulmonary arterial hypertension trials.25 This study was considered only for sensitivity analyses within functional class at baseline. All studies were parallel in design and funded by the pharmaceutical industry.
Screening process
Methodological quality
The methodological quality was satisfactory: the mean (SD) impact factor of the journals in which were published the different trials was 18 (12) and the mean (SD) Jadad score was 4 (0.7). All of these trials used intention-to-treat analyses with the last value carried forward method.
Study population
This systematic review included 885 patients, 77% female, with a mean (SD) age of 49 (12) years old. A total of 248 patients (28%) had PAH related to connective tissue diseases, including 153 patients with SSc (17%; 62% of the CTD subgroup).
In the whole population, the baseline walking distance was not different between groups (344 (76) m in placebo, 330 (75) m in bosentan, 337 (80) m in sitaxsentan and 345 (77) m in sildenafil). In the subset of patients with CTD, the baseline walking distance was significantly decreased in patients treated with bosentan vs the placebo group (312 vs 361 m, p = 0.01), and there was a trend toward significance in patients treated with sitaxsentan enrolled in the STRIDE-1 study (340 vs 414, p = 0.06) compared to placebo. The baseline walking distance was not reported for patients treated with sitaxsentan included in the STRIDE-2 study and patients treated with sildenafil. The characteristics of patients are detailed in table 1.
Effects of oral treatment for exercise capacity
All the 10 placebo-controlled studies analysed provided suitable data on exercise capacity for 613 participants (176 with CTD) who received the active treatment and for 272 (70 with CTD) who received placebo. The mean trial duration was 14 (2.5) weeks (median: 12 weeks, range: 12–18 weeks). The effect sizes for exercise capacity for each of the 10 included studies are provided in table 2.
The effect sizes of ERAs and PDEIs for exercise capacity in the subset of patients with CTD were respectively 0.27 (95% confidence interval, CI −0.01 to 0.54) and 0.53 (95% CI −0.02 to 0.89). These results were homogenous (Q = 0.56 for 3 degrees of freedom (DF); p = 0.9 and Q = 0.44 for 2 DF; p = 0.8 respectively). In comparison, in the whole PAH population, effect sizes of ERAs and PDEIs were 0.44 (95% CI 0.29 to 0.58) and 0.58 (95% CI 0.38 to 0.79) respectively. These results were also homogenous (Q = 5.4 for 6 DF; p = 0.5 and Q = 0.2 for 2 DF; p = 0.9 respectively).
Sensitivity analyses within the nature of the drug in patients with CTD revealed the following effect sizes: 0.31 (95% CI −0.22 to 0.83) for Bosentan, 0.26 (95% CI −0.06 to 0.57) for sitaxsentan and 0.53 (95% CI −0.02 to 0.89) for sildenafil (table 3). In comparison, effect sizes for exercise capacity within the nature of the drug for the whole PAH population were 0.61 (95% CI 0.38 to 0.84), for bosentan, 0.33 (95% CI 0.15 to 0.51) for sitaxsentan and 0.58 (95% CI 0.38 to 0.79) for sildenafil (table 3). No heterogeneity was observed.
Sensitivity analyses showed no changes in the conclusions with the treatment dose for bosentan (Bosentan 62.5 twice a day for 4 weeks then 125 or 250 mg twice a day) and sildenafil (20, 40 or 80 mg three times per day) in patients with CTD and the whole population of PAH (table 3, fig 2). However, only sitaxsentan 100 mg was associated with a significant effect size for exercise capacity in the whole population of PAH (effect size: 0.36, 95% CI 0.10 to 0.62), but not in the subset of patients with CTD (table 3 and fig 2).
Sensitivity analysis within the baseline functional class showed in the sitaxsentan-treated subpopulation that met enrolment criteria of previous pulmonary arterial hypertension trials (functional class III/IV, baseline 6-min walk distance of <450 m) a greater effect size for exercise capacity than seen in the entire STRIDE-1 population (0.70, 95% CI 0.18 to 1.22 vs 0.33, 95% CI 0.15 to 0.51). Unfortunately, it was not possible to confirm these results in the subset of patients with CTD because of the absence of subgroup analysis. All patients treated with bosentan were class III/IV. Studies on sildenafil did not provide suitable data to perform an analysis within the functional class.
DISCUSSION
PAH is a serious condition widely recognised to be a major complication of CTD and SSc ifor limited and diffuse cutaneous subtypes. Once established, severe PAH is difficult to treat and has a poor prognosis.28 There is therefore a need for early detection and treatment.29 Unlike idiopathic PAH, SSc-related PAH occurs in a known population and thus it is a crucial issue to identify patients with early PAH and treat them with effective drugs before the disease becomes irreversible.
This meta-analysis shows that new oral PAH treatments (ERAs and PDEI) are less effective in improving exercise capacity in PAH related to the subset of patients with CTD (and more particularly SSc) than in other patients with PAH. This is highlighted by non-significant effect size values of ERAs and PDEIs for exercise capacity in the subset of patients with CTD, as compared to the whole PAH population. The baseline difference of 6-min walk test between treated and placebo groups could by itself have influenced the results of the studies; the individual values for the correlation between baseline walk test and treatment effect should be analysed to confirm this hypothesis. These results may also be explained by a poor therapeutic response in the subset of patients with CTD/SSc as previously suggested8 14 or by a weaker relevance of the 6-min walk test used to assess efficacy in the subset of patients with CTD. Exercise capacity is commonly chosen as a primary outcome to assess therapeutic efficacy in PAH clinical trials. The 6-min walk test was shown to be an independent predictor of mortality and a reliable tool for the assessment of exercise capacity in patients with IPAH.9 However, primary myocardial involvement,30 pulmonary fibrosis,31 osteoarticular/soft tissue involvements32 and general bad health are other sources of reduction in exercise capacity in patients with SSc, making it difficult to know what the normal value would be in any given patient.33 For example, taking into account the musculoskeletal involvement of the disease, one could suggest standardisation on body mass index, muscular mass or scleroderma Health Assessment Questionnaire is required to improve the accuracy of the test.
Therefore, the validity, reliability and discriminatory capacity of the 6-min walk test may be biased by the co-morbidities of CTD and SSc patients, and this may explain the differences observed between patients with CTD and the whole PAH population. This observation is sustained by the improvement of observed survival (compared to survival of a historical cohort) at 1 and 2 years with ERAs21 34 in CTD related PAH, which contrasts strikingly with the absence of clinical relevant improvement on exercise capacity.
Our analysis had some limitations that merit consideration. The main limitation could be the insufficient power of studies performed on the subset of patients with CTD, which were all subgroup analyses of pivotal randomised controlled trials; this may have influenced the results of the trials and also our meta-analysis. Furthermore, because of the absence of subgroup analyses in the articles reviewed, it was not possible to make direct comparisons between CTD and IPAH subgroups on exercise capacity. It was also not possible to use meta-regressions to assess potential effect modification by type of patient (age, sex, race, disease duration, and so on). Also, the heterogeneity of trials inclusion criteria did not allow us to perform head to head comparison between the different active treatments.
Secondary outcomes as functional class, haemodynamic parameters, time to clinical worsening and survival were not provided in all studies performed on patients with CTD, which did not allow us to confirm the results obtained with the primary outcome (exercise capacity). Another limitation of these results is the handling of missing data: studies performed intention-to-treat analysis with the last value carried forward method. This latter parameter limits the validity of results and growth curve analysis should be preferred.35 Potential publication bias introduced into the process of locating and selecting studies for inclusion cannot be excluded because studies with significant results are more likely to be published than studies without significant results.36 Finally, all the studies were industry funded and there is evidence suggesting that industry-funded studies could overestimate treatment effects.37
Overall, many factors inherent to the methodology of meta-analyses, and also the complexity of SSc, raise the issue of reliability of the results. This systematic review emphasised some implications for practice and future research. There is a need for RCTs assessing the efficacy of ERAs and PDEIs specifically in SSC-related PAH. These studies need to better characterise how these therapies affect the pulmonary vasculature pathologically, biologically and haemodynamically, and whether survival is actually improved.33 Thus, they will require the use of specific outcomes as haemodynamic parameters, functional and dyspnoea class, time to clinical worsening, biological markers38 and survival analysis.
Studies assessing the long-term safety of ERAs and PDEI in the subset of patients with CTD/SSc are also warranted to confirm the promising results on safety obtained in short-term RCTs. Finally, this meta-analysis of RCTs suggests a lack of clinically relevant improvement on exercise capacity in the subset of PAH patients with CTD/SSc after 12 to 18 weeks of treatment with ERAs (dual endothelin receptor antagonist and selective endothelin A receptor antagonist) and PDEIs. A poor therapeutic response, an insufficient power of studies or a bad sensitivity to change of the 6-min walk test may explain these results. The promising preliminary data on survival of ERAs and the confounding effects of other comorbidities associated with CTD, particularly in SSc, may support the latter hypothesis. Thus these data support the need for other tools than the 6-min walk test to assess damages and treatment effects in CTD related PAH.
Further studies with well defined methods and follow-up periods adequate in length are thus warranted in SSC-related PAH with the use of relevant parameters considering the weight of this devastating condition in disease mortality.
REFERENCES
Footnotes
Competing interests: AK has received consulting fees/honoraria from Actelion Pharmaceuticals.