Abstract
Schistosomiasis-associated pulmonary arterial hypertension (Sch-PAH) is a life-threatening complication of chronic hepatosplenic schistosomiasis. It is suggested to be the leading cause of pulmonary arterial hypertension (PAH) worldwide. However, pathophysiological data on Sch-PAH are scarce. We examined the hypothesis that there are pronounced similarities in pathophysiology, haemodynamics, and survival of Sch-PAH and idiopathic PAH (iPAH).
This systematic review and meta-analysis was registered in the PROSPERO database (identifier CRD42018104066). A systematic search and review of the literature was performed according to PRISMA guidelines for studies published between 01 January 1990 and 29 June 2018.
For Sch-PAH, 18 studies evaluating pathophysiological mechanisms, eight studies on haemodynamics (n=277), and three studies on survival (n=191) were identified. 16 clinical registries reporting data on haemodynamics and survival including a total of 5792 patients with iPAH were included for comparison. Proinflammatory molecular pathways are involved in both Sch-PAH and iPAH. The transforming growth factor (TGF)-β signalling pathway is upregulated in Sch-PAH and iPAH. While there was no difference in mean pulmonary artery pressure (54±17 mmHg versus 55±15 mmHg, p=0.29), cardiac output (4.4±1.3 L·min−1 versus 4.1±1.4 L·min−1, p=0.046), and cardiac index (2.6±0.7 L·min−1·m−2 versus 2.3±0.8 L·min−1·m−2, p<0.001) were significantly higher in Sch-PAH compared to iPAH, resulting in a lower pulmonary vascular resistance in Sch-PAH (10±6 Woods units versus 13±7 Woods units, p<0.001). 1- and 3-year survival were significantly better in the Sch-PAH group (p<0.001).
Sch-PAH and iPAH share common pathophysiological mechanisms related to inflammation and the TGF-β signalling pathway. Patients with Sch-PAH show a significantly better haemodynamic profile and survival than patients with iPAH.
Abstract
Sch-PAH and iPAH share common pathophysiological mechanisms related to inflammation and the TGF-β signalling pathway. Sch-PAH patients show a significantly better haemodynamic profile and survival than iPAH patients. http://bit.ly/2mNwzWZ
Introduction
Schistosomiasis is a parasitic infection caused by trematode worms that are known to cause either acute infection (known as Katayama syndrome) or chronic infection, which primarily results from the host's immune response to schistosome eggs and the granulomatous reaction provoked by their antigens [1, 2]. Schistosomiasis-associated pulmonary arterial hypertension (Sch-PAH) is a life-threatening complication of chronic hepatosplenic schistosomiasis [3–5]. Schistosomiasis pertains to neglected tropical diseases, which affect a large number of people and usually attract insufficient attention, research efforts and financial resources, since neglected tropical diseases primarily concern people of lower socioeconomic status [6]. Hereby, schistosomiasis affects more than 230 million people worldwide, of whom 10% develop hepatosplenic disease [7–9]. Of these, 5% may develop pulmonary arterial hypertension (PAH), suggesting schistosomiasis to be the leading cause of PAH worldwide [3, 5, 10, 11]. Recent data even indicate that these numbers are an underestimation of the actual burden of disease [7, 12–14]. Extrapolating the data from a prevalence study suggests that there are likely to be a total of 425 000 patients with Sch-PAH worldwide [4, 15]. In contrast, idiopathic PAH (iPAH) is rare, with a prevalence of 6–10 per million in the western world and 15–25 per million for idiopathic, familial, drug-induced and associated PAH [16]. To date, Schistosoma mansoni is known to be the primary cause of Sch-PAH, while only a few cases of PAH caused by other schistosomal species have been reported in the literature [17]. In addition, recent preliminary data suggest that Schistosoma japonicum may induce PAH in mice [18].
Despite its high prevalence compared to iPAH, relatively little is known about the underlying pathophysiological mechanisms of Sch-PAH [11]. The following non-mutually exclusive aetiologies have been proposed: 1) mechanical obstruction of the pulmonary circulation by worm eggs; 2) inflammation leading to endothelial cell dysfunction due to the passage of worms or eggs; and 3) portal hypertension leading to pulmonary overflow and thereby endothelial cell dysfunction, comparable with portopulmonary hypertension [11, 19–21]. Compared to Sch-PAH, iPAH has been studied in considerably more detail. For patients with iPAH, targeted therapy is available and effective, whereas there is no targeted treatment approved for Sch-PAH [22]. The anthelmintic drug praziquantel kills adult worms, but is ineffective against immature parasites or eggs. In addition, it has been suggested that treatment of murine Sch-PAH with praziquantel has the potential to reverse vascular remodelling, including obliterative lesions [5, 23]. The vascular remodelling present in Sch-PAH has a point of no return, beyond which, anthelmintic therapies are ineffective to prevent progression to PAH [20].
The histopathologic changes found in the pulmonary vasculature of Sch-PAH are similar to those found in iPAH [3, 24, 25]. Therefore, it has been hypothesised that similar pathophysiological mechanisms are involved in both types of PAH [20, 24].
Understanding the underlying pathophysiologic mechanisms of Sch-PAH in comparison to iPAH may allow future modulation of these pathways with novel targeted therapies [26]. To our knowledge, this is the first systematic review and meta-analysis comparing pathophysiological mechanisms, haemodynamics and survival of patients with Sch-PAH and iPAH.
Methods
Full methodological details are available in the online supplement.
Results
Our database search delivered a total of 533 unique records. Overall, 250 records were excluded after screening of titles and abstracts, resulting in 283 unique papers for full text review. In addition, we identified 17 relevant studies from bibliographies of relevant publications (figure 1).
Pathophysiologic mechanisms in Sch-PAH and iPAH
A total of 18 studies examining the pathophysiological mechanisms of Sch-PAH were eligible for inclusion (table E2) [3, 25, 28–42]. Mechanisms involved in the pathogenesis of iPAH were selected from the proceedings of the 6th World Symposium on Pulmonary Hypertension and compared to pathophysiologic pathways in Sch-PAH (table 1 and figure 2) [74]. No valid studies were found on the following 10 molecules regarding their involvement in the development of Sch-PAH: endothelin-1; prostacyclin; [K+]-channels; macrophage inflammatory protein-1α; von Willebrand factor; matrix metalloproteinase (MMP)-2; membrane-type-MMP1; tissue inhibitor of MMP-1; serotonin transporter/receptor; and chemokine (C-C motif) ligand 5.
Transforming growth factor (TGF)-β isoforms 1, 2, and 3 are expressed in pulmonary hypertension arteries [43, 67]. TGF-β1 enhances cell proliferation in smooth muscle cells (SMCs) and adventitial fibroblasts in the pulmonary arteries of patients with iPAH and has a growth-inhibitory effect on normal cells [31, 43, 75]. In iPAH its expression is dysregulated, showing a loss of function for the TGF-β family signalling in endothelial cells of iPAH, but also a gain of function in SMC growth and AF activation. Patients with Sch-PAH and periportal fibrosis had significantly higher serum levels of TGF-β1 when compared to patients with schistosomal periportal fibrosis only [31]. Furthermore, TGF-β1 and Smad2/3-dependent TGF-β signalling are mandatory for the pathogenesis of experimental pulmonary hypertension caused by S. mansoni [35]. Graham et al. [35] demonstrated a significant increase in the level of TGF-β1 mRNA in S. mansoni-exposed mice compared to unexposed mice. However, there were no changes in TGF-β2 or TGF-β3 expression between the exposed and the unexposed group. The study also showed the increase in TGF-β1 to be dependent on the T-helper (Th)2 cytokines interleukin (IL)-4 and IL-13, which are induced by S. mansoni. Neutralisation of all TGF-β isoforms with anti-pan-TGF-β neutralising antibody resulted in partial protection against increases in right ventricular systolic pressure (RVSP) and media thickening induced by S. mansoni exposure [35]. No significant changes in intima thickness, right ventricular hypertrophy and peri-egg granuloma volume were observed. The same occurred when the effect of blockade of TGF-β receptor-1 kinase activity in S. mansoni-exposed mice was tested: the inhibitors partially prevented the increase in RVSP, but not of medial thickening. In addition, the amounts of IL-4 and IL-13 mRNA were significantly increased after exposure to S. mansoni, but trended towards suppression, when inhibited with anti-pan-TGF-β antibody [35]. These findings, taken together with previously described increased TGF-β signalling in Schistosoma-infected mice lacking the decoy IL-13 receptor IL-13Rα2, point towards a mutually amplifying feed-forward loop of IL-4/IL-13 and TGF-β [33, 35]. Recent data emphasise the importance of the activation of TGF-β by thrombospondin (TSP)-1 in the development of experimental pulmonary hypertension caused by S. mansoni [38]. TSP-1 is upregulated in the lungs of mice after exposure to S. mansoni, via recruitment of circulating monocytes, while TSP-1 inhibition or knockout hinders activation of TGF-β and protects against pulmonary hypertension [38].
Another component of the TGF-β-related family is the BMPR2 gene, coding for bone morphogenetic protein (BMP) receptor type 2 (BMPR-II), which exhibits heterozygous germline mutations in 70–80% of familial cases of PAH [43–45]. Furthermore, BMPR-II expression is reduced in plexiform lesions and remodelled pulmonary arteries, especially in endothelial cells of patients with iPAH [43, 46]. Recent data indicate that cells in the central core of plexiform lesions lack TGF-β and BMP signalling, which would allow plexiform lesions to abnormally proliferate (via loss of cytostatic TGF-β signalling), or facilitate apoptosis of endothelial cells (via loss of BMPR-II) [43]. In addition, it has become clear that BMPR2 mutations in SMCs contribute to abnormal growth responses to BMP and TGF-β [76]. Reduced BMPR-II expression leads to exaggerated TGF-β signalling and altered cellular responses to TGF-β, with the Smad pathway being the mechanism in common of BMP and TGF-β [76]. The combination of SMC proliferation and increased endothelial cell apoptosis has been proposed to be an important mechanism for the development of PAH [76]. In experimental pulmonary hypertension caused by S. mansoni, loss of BMPR-II has shown to be associated with a higher degree of pulmonary vascular remodelling, which could be shown in heterozygous mice with a null BMPR2 mutation (BMPR2+/−) [30, 35]. Moreover, BMPR2+/− mice appeared to have an increased pulmonary egg burden and higher cytokine expression in their lungs compared to wild-type mice [30]. A database study, that analysed RNA-seq data from whole lung tissue of mice with experimental pulmonary hypertension caused by S. mansoni (GSE49114), also concluded BMPR-II, besides Smad9, Eng, and IL-4, to be of importance in the development of pulmonary hypertension [42].
Furthermore, it has been proposed that plexiform lesions represent a process of misguided angiogenesis based on findings in the expression of vascular endothelial growth factor (VEGF), its receptors and hypoxia-inducible factor (HIF)-1α and HIF-1ß [51]. In iPAH, VEGF is among the factors that are suspected to enhance endothelial cell and SMC proliferation, or decrease apoptosis [51]. VEGF is induced by the Th2 inflammatory cytokines IL-4 and IL-13 and can contribute to Th2 responses by forming a proinflammatory feed-forward loop. Thereby VEGF plays an important role in the development of Sch-PAH [37, 77]. In a murine model of pulmonary hypertension caused by S. mansoni, Chabon et al. [29] were able to show that VEGF blockade with a VEGF receptor/tyrosine kinase inhibitor (SU5416) was not only able to partially suppress the expression of IL-4 and IL-13 in lung and liver of infected mice, but also inhibit pulmonary vascular remodelling. Furthermore, an animal model conducted by Araújo et al. [28] showed with the use of the hypoxia marker pimonidazole, that hypoxia occurred in inflammatory cells around eggs and granulomas in liver, spleen and lungs of mice infected with S. mansoni. In addition, positive VEGF and HIF-1α immunostaining were observed in inflammatory regions. However, the study failed to verify a correlation between pimonidazole binding (100% positive) and HIF-1α (21% positive) and VEGF expression (18% positive) in lung tissues and granulomas. The authors hypothesised this to apply because of nonspecific damage and molecular sequelae induced by the passage of schistosomulae through pulmonary vessels in early stages of the disease [28].
Increased serum levels of the proinflammatory cytokine IL-6 have been found in patients with iPAH [52, 78]. An upregulation of the IL-6–STAT3–NFATc2 pathway has also been found in experimental models of pulmonary hypertension [79, 80]. One study utilising whole transcriptome analysis, described an upregulation of the IL-6–STAT3–NFATc2 pathway in lungs of mice exposed to S. mansoni [36]. Immunostaining revealed increased IL-6, STAT3, and NFATc2 levels in peri-egg granulomas and adventitial infiltrates. However, whole lung samples were utilised for transcriptome analysis including multiple tissue compartments like parenchyma and airways and not only vasculature, which definitely represents a major limitation of this study [36]. The same study postulated that in contrast to other models of PH, where IL-6 appears to be pathogenic, IL-6 would have protective effects against the development of experimental PH caused by S. mansoni. However, taking a closer look at the data, the authors found decreased media remodelling in mice with a genetic lack of IL-6 (IL-6−/− mice) stimulated intraperitoneally and intravenously with eggs of S. mansoni [36]. No differences in RVSP, intima thickness, right ventricular hypertrophy, peri-egg granulomas, or clearance of eggs from the lungs between IL-6−/− and IL-6+/+ mice were observed. Additionally, there was a broad spreading of the values of media thickness in IL-6−/− mice.
Haemodynamics of Sch-PAH and iPAH
Eight studies on haemodynamics in Sch-PAH, including a total of 277 patients, were eligible for meta-analysis (table E3) [4, 11, 15, 81–85]. Three case reports assessing invasive haemodynamics were included, but not evaluated in meta-analysis (table E4) [86–88]. Haemodynamic data from 11 clinical registries on iPAH, including a total of 3532 patients, were utilised for comparison (table E5) [16, 60, 81, 89–96].
Haemodynamics from single studies and grand means of all included studies in Sch-PAH are given in table E6; those for iPAH are given in table E7. A comparison of grand means in Sch-PAH and iPAH is provided in table 2.
While there was no difference in mean pulmonary arterial pressure (54±17 mmHg versus 55±15 mmHg, p=0.29), cardiac output (4.4±1.3 L·min−1 versus 4.1±1.4 L·min−1, p=0.046) and cardiac index (2.6±0.7 L·min−1·m−2 versus 2.3±0.8 L·min−1·m−2, p<0.001) were significantly higher in Sch-PAH compared to iPAH and pulmonary vascular resistance was thus significantly lower in Sch-PAH (10±6 Woods units) compared to iPAH (13±7 Woods units, p<0.001).
Survival in Sch-PAH and iPAH
Three studies on survival in Sch-PAH, including 191 patients, were identified (table E8) [15, 81, 82]. For comparison, 11 clinical registries including a total of 3653 iPAH patients were used (table E9) [60, 81, 91, 93, 94, 96–101].
1-year survival was significantly better in the Sch-PAH group (91±9.9%) compared to iPAH (86.1±9.7, p<0.001). The same differential applied for 2-year survival (84.3±20.8% versus 75.7±11.7%, p<0.001) and 3-year survival (73±26.8% versus 67.3±14.1%, p<0.001) (table 3).
5-year survival was significantly better in Sch-PAH (52.5±19.9 versus 79.0±14.1, p<0.001). However, only one study assessed 5-year survival in Sch-PAH (tables E8) [82]. This study evaluated the effect of PAH-targeted therapies on survival in Sch-PAH. Of note, survival at 5 years was significantly better in those receiving targeted therapies (69.2% versus 89.1%, p=0.029).
Discussion
Schistosomiasis-associated PAH is a prevalent and life-threatening disease, with a high socioeconomic burden. Still, Sch-PAH is relatively understudied and data on the effect of schistosomiasis on the pulmonary vasculature are scarce. Therefore, we decided to perform this systematic review and meta-analysis, which sought to compare pathophysiologic patterns, haemodynamics, and survival of patients with Sch-PAH and iPAH from the existing data. Following an extensive search, 29 relevant papers on Sch-PAH were identified. Compared to iPAH, only few data were available on the pathophysiological mechanisms of Sch-PAH. Nevertheless, multiple underlying pathological mechanisms present in iPAH could also be confirmed for Sch-PAH. Furthermore, both entities showed histological similarities and identical plexogenic arteriopathy in lung samples taken at autopsy [3]. Our data show that patients with Sch-PAH have a more favourable haemodynamic profile and longer survival compared to iPAH, even in the absence of targeted therapy.
Fernandes et al. [82] compared the survival of 102 newly diagnosed patients with Sch-PAH treated with PAH-targeted therapies against a group of 50 untreated patients from a historical cohort. The 5-year survival of the untreated group was 69.2%, compared to 89.1% in the treated Sch-PAH group (p=0.029), showing that the use of targeted PAH therapy was associated with improved survival [82].
Based on an absence of a positive acute response to vasoreactivity testing in Sch-PAH [15, 102], as usually observed in ∼15% of patients with iPAH, it has been previously hypothesised that Sch-PAH and portopulmonary hypertension might share pathophysiologic mechanisms [15].
Previous studies have described egg deposition in lung vessels as a key mechanism in the development of pulmonary vascular disease and PAH. Recently, however, Graham et al. [34] have illustrated, via the detection of S. mansoni soluble egg antigen in 18 lung samples collected from necropsy of patients with Sch-PAH, that it was unlikely to detect a significant amount of persistent parasite-derived antigens within the lungs of those individuals. In this study, all lung samples showed histological evidence of pulmonary vascular remodelling with plexiform lesions (100%) and medial thickening (89%) [34]. Pulmonary granulomas were found in only 22% of patients [34]. None of the 18 lungs showed visible eggs, or positive immunofluorescence staining for soluble egg antigen [34]. This is in contrast to historical data on human autopsy studies in Sch-PAH, which reported schistosomal eggs in 100% of lung samples, and an animal study by Crosby et al. [25] which showed that only mice with lung egg burden developed pulmonary hypertension [103]. Another animal study showed that the development of pulmonary hypertension could be prevented by the use of praziquantel, and that established vascular remodelling could be reversed [25]. These findings are controversial with respect to Sch-PAH in humans, where inflammation and vascular lesions are persistent in individuals who died from Sch-PAH despite the lack of soluble egg antigen in lung samples collected at autopsy [34]. Graham et al. [34] hypothesised that the absence of eggs in their study (compared to historical reports) derived from the introduction of anthelmintic treatment with praziquantel. The authors argue that after an initial acute inflammatory response, vascular lesions are established, which persist beyond a “point of no return”, progressing to PAH through altered cellular mechanisms and clonal proliferation [34, 43].
Mauad et al. [3] compared lung tissue samples collected at autopsy of patients who died from Sch-PAH or iPAH and found both entities to be histologically indistinguishable, with both featuring plexogenic arteriopathy as the predominant pattern of pulmonary vasculopathy. Lymphocytic vasculitis with perivascular inflammatory infiltrates of B-cells, T-cells and mast cells was equally present in iPAH and Sch-PAH samples. The presence of S. mansoni eggs could only be observed in two out of 10 cases, which were both associated with active intestinal infection [3].
Overall, it appears that iPAH and Sch-PAH show pronounced similarities in their underlying molecular pathogenic mechanisms: reduced BMPR-II signalling contributes to abnormal growth responses to BMP and TGF-β, leading to a combination of SMC proliferation and increased endothelial cell apoptosis in iPAH [76]. In experimental pulmonary hypertension caused by S. mansoni, loss of BMPR-II is associated with more severe pulmonary vascular remodelling [30, 35]. VEGF enhances SMC and endothelial cell proliferation through increased apoptotic resistance in iPAH [51]. Blockade of VEGF-signalling in murine pulmonary hypertension caused by S. mansoni inhibits pulmonary vascular remodelling [29]. Increased serum levels of IL-6 have been found in patients with iPAH alongside an upregulation of the IL-6–STAT3–NFATc2 pathway in experimental models of pulmonary hypertension [52, 79, 80]. In addition, deficiency of IL-6 in an IL-6 knockout mouse model was shown to be protective against the development of hypoxia-induced pulmonary hypertension compared to wild-type mice [104]. In contrast, one study elucidated that mice lacking IL-6 developed pulmonary hypertension, which was associated with significant intima remodelling without increase in media thickness after S. mansoni exposure [36]. However, the same study reported an upregulation of the IL-6–STAT3–NFATc2 pathway in the lungs of mice exposed to S. mansoni. These contradictory findings could be context-dependent, depending on the inflammatory aetiology of pulmonary hypertension versus hypoxia-induced PH. In conclusion, it appears that data on IL-6 regulation in Sch-PAH is contradictory and that further animal and in vivo investigations are needed in the future.
In iPAH, TGF-β1 is dysregulated and increases cell proliferation of SMC and adventitial fibroblasts, while showing a loss of function in endothelial cells [31, 43, 75]. Patients with Sch-PAH and periportal fibrosis exhibit significantly higher TGF-β1 serum levels than patients with schistosomal periportal fibrosis only (p=0.006) [31]. Furthermore, TGF-β1 expression is enhanced in mice exposed to S. mansoni [33, 35]. Neutralisation of all TGF-β isoforms leads to partial protection against increases in RVSP and media thickening induced by S. mansoni exposure, with the Th2 cytokines IL-4 and IL-13 playing a role [35].
Parasitic TGF-β superfamily members, such as Smad4 are involved in diverse developmental processes in different tissues and different stages throughout the parasite life cycle [40]. Whether these parasitic TGF-β superfamily members, such as S. mansoni inhibin/activin, which is abundantly expressed in ovipositing females and eggs, play a major role in the development of Sch-PAH needs to be determined in future studies [32]. Nevertheless, it has been demonstrated that the schistosomal type II receptor strongly binds human TGF-β, which gives strong evidence for the utilisation of a host ligand in the parasite's growth and development; however, which does not rule out the involvement of a schistosomal self-ligand [41].
It therefore appears that the infringing underlying mechanisms in Sch-PAH and iPAH eventually result in a similarly exaggerated response to injury or chronic inflammatory stimuli (i.e. overstimulated or aberrant wound-healing mechanisms). TGF-β1 is known to function as the profibrotic growth factor in wound healing. Under normal wound-healing conditions TGF-β1 enhances SMC and adventitial fibroblast proliferation and usually induces apoptosis in endothelial cells, while VEGF protects endothelial cells from apoptosis. In Sch-PAH, significantly higher TGF-β1-serum levels have been found compared to patients with schistosomiasis without PAH. We assume that in the pulmonary arteries of patients with Sch-PAH, inadequate SMC proliferation caused by dysregulated TGF-β1 stimulation occurs, which is reflected in the typical media thickening and intima proliferation found in histological samples of patients with Sch-PAH. In addition, SMC proliferation is further enhanced by BMPR2 mutation in SMC as mentioned above [76]. Intact BMPR2 signalling seems to be necessary for the execution of normal pulmonary vascular wound healing by the prevention of apoptosis-induced compensatory cell proliferation [105].
The limitations of this meta-analysis include the imbalance of group sizes between Sch-PAH and iPAH, as only a few data on invasive haemodynamic measurements and survival in Sch-PAH are available. Furthermore, several retrospective clinical studies had to be included for meta-analysis of Sch-PAH, due to the pronounced lack of prospective clinical data on Sch-PAH, and were compared to data on iPAH, which were derived from national registries. In addition, there was a heterogeneity in selected studies on Sch-PAH, as some studies evaluated consecutive cases while others included only newly diagnosed cases, which could have contributed to a seemingly more benign course of disease in Sch-PAH. This issue, however, has also been elicited by the scarcity of haemodynamic data by right heart catheterisation in Sch-PAH. Furthermore, meta-analysis of survival data may be biased by retrospective data and uncontrolled treatments. Additionally, the vast majority of patients with iPAH included in national registries received PAH-specific treatments, while only a small proportion of patients with Sch-PAH were treated. We speculate that survival in Sch-PAH may therefore be underestimated. Unfortunately, the pathophysiological assessment of Sch-PAH is primarily taken from animal or in vitro studies, whereas human patient data are limited. In addition, the majority of included publications were from centres based in Brazil, as very few studies were found from other parts of the world where schistosomiasis is endemic. Publication bias, favouring the publication of positive studies, also cannot be excluded.
Conclusions
Patients with Sch-PAH show a more favourable haemodynamic profile and better survival compared to those with iPAH. Pathophysiologic mechanisms in Sch-PAH and iPAH show pronounced molecular similarities with increased TGF-β signalling playing a key role in both entities. Down regulation of BMPR-II, TGF-βR2 and upregulation of HIF-1α, IL-13, VEGF and TGF-β are involved in both Sch-PAH and iPAH (figure 2). Available data on PAH-targeted therapies in patients with Sch-PAH suggest a beneficial effect on survival but these are scarce and need further evaluation in future randomised controlled trials. Due to the paucity of evidence-based research on Sch-PAH we emphasise the importance of future clinical studies in this field.
Supplementary material
Supplementary Material
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
ERR-0089-2019_Supplementary material ERR-0089-2019_Supplementary_material
Footnotes
This article has supplementary material available from err.ersjournals.com
Submitted article, peer reviewed.
Conflict of interest: D. Knafl has nothing to disclose.
Conflict of interest: C. Gerges reports grants from Bayer HealthCare and Actelion Pharmaceuticals, during the conduct of the study; and personal fees from GlaxoSmithKline, AOPOrphan and Actelion, outside the submitted work.
Conflict of interest: C.H. King has nothing to disclose.
Conflict of interest: M. Humbert reports personal fees from Actelion, Bayer, GSK, Merck and United Therapeutics, during the conduct of the study.
Conflict of interest: A.L. Bustinduy has nothing to disclose.
- Received July 21, 2019.
- Accepted September 17, 2019.
- Copyright ©ERS 2020.
This article is open access and distributed under the terms of the Creative Commons Attribution Non-Commercial Licence 4.0.