The Role of Iron in Pulmonary Hypertension

THE ROLE OF IRON IN PULMONARY HYPERTENSION

A thesis presented for the degree of MD (Res) by Dr Geoffrey Watson Centre of Pharmacology and Therapeutics Imperial College London February 2018

I declare that this thesis was conducted and written by myself and the work included within is my own unless otherwise stated.

The copyright of this thesis rests with the author. Unless otherwise indicated, its contents are licensed under a Creative Commons Attribution-Non Commercial 4.0 International Licence (CC BY-NC). Under this licence, you may copy and redistribute the material in any medium or format. You may also create and distribute modified versions of the work. This is on the condition that: you credit the author and do not use it, or any derivative works, for a commercial purpose. When reusing or sharing this work, ensure you make the licence terms clear to others by naming the licence and linking to the licence text. Where a work has been adapted, you should indicate that the work has been changed and describe those changes. Please seek permission from the copyright holder for uses of this work that are not included in this licence or permitted under UK Copyright Law.

Acknowledgements

I would like to say an enormous thank you to my supervisors Dr. Luke Howard, Prof. Lan Zhao and Prof. Martin Wilkins. They have offered fantastic guidance throughout this project and have been incredibly patient with me. Dr Luke Howard has really gone out of his way to support me through this work for which I am eternally grateful.

Special thanks also goes to Dr John Wharton and Dr Chris Rhodes. Both played vital roles in guiding me around the pitfalls of medical statistics and moulding me from clinician into a research fellow.

I would like to thank the support of the cardio-pulmonary exercise test (CPET) laboratory, in particular Dr Kevin Murphy for both conducting the CPETs where data was accrued, as well as his invaluable real-time physiology teaching whilst I supervised these studies. He was pivotal in helping to analyse the CPET results, along with Dr Howard.

I would like to thank the Hammersmith hospital cardiac catheter laboratory team, which without technical, nursing, and radiological support I would not have been able to conduct the invasive right and left cardiac catheterisations that has generated key data to support this thesis.

I would like to thank Dr Chris Baker for his help in teaching me right heart catheterisation via the brachial approach which has changed the access route for research and clinical cases, as a default, locally and has allowed me to perform exercise catheterisations when needed, and in turn intra- oesphageal pressure measurements. Professor Michael Polkey was a key contributor towards the theoretical and technical aspects of measuring and interpreting intra-oesphagealpressure measurements for which I am extremely grateful. In turn, we have had an abstract accepted for the Eurpean Respiratory Society annual congress in September 2018 with our early findings presented in this thesis.

I would like to thank Dr Ben Ariff for his expertise in helping to analyse cardiac MRI studies.

I would like to thank Dr Souad Ali who was instrumental in helping me collect, process and collate research blood samples.

I would like to thank the support of Imperial College Healthcare NHS Trust pathology laboratories in providing accurate patient data for iron status, haemoglobin, red cell distribution width, c-reactive protein, creatinine, B-type natriuretic peptide.

I would like to thank Dr Mark Busbridge and his laboratory team at Charing Cross Hospital who provided soluble transferrin and hepcidin data.

I would like to thank my wife, my parents and my brother. They have been monumentally supportive throughout the highs and lows. I could not have done it without them.

I would like to thank my close friends for their support and solid base, and putting me up in a spare room at last minute during my research trips to London.

Abstract

Introduction: Iron is a critical ion in the regulation of many cellular processes and iron deficiency (ID) has been shown to be a powerful predictor of survival in many diseases. It has been shown that ID, defined by increased soluble transferrin receptor (sTfR) levels impacts on outcomes in idiopathic pulmonary arterial hypertension (IPAH). sTfR is not widely available as an assay. It is also not known how common ID is in other forms of pulmonary hypertension (PH) and whether it impacts on function and survival.

Aims: 1) To identify the best routinely-available biomarker of ID in patients with IPAH; 2) Assess the prevalence of ID in common forms of PH; 3) Explore the relationship between ID and function and survival in PH; 4) Run a double-blind, randomised, placebo-controlled trial of iv iron in IPAH.

Methods: Data were collected historically from patients in the PH service at Hammersmith Hospital and analysed to provide the best clinical criteria to identify patients with ID. Subsequently, a larger cohort of patients with IPAH as well as other forms of PH was analysed to assess the prevalence of ID at diagnosis and the impact on functional status and survival. A crossover trial was run to assess the impact of iv iron on haemodynamics and exercise in IPAH.

Results: ID can be best identified by using red cell distribution width (RDW) in IPAH. RDW is associated with exercise capacity in IPAH, but more so, is associated with functional changes in chronic thromboembolic PH. RDW was a strong marker of survival across all groups but independence could not be established from other markers. The clinical trial did not complete in time, but unblinded data suggest improvements in aerobic exercise, but no change in haemodynamics.

Conclusion: ID is common in all forms of PH, and has functional and survival implications. Iron replacement in IPAH may improve exercise function but complete results from the trial are awaited.

Table of contents

Thesis title page Declaration Acknowledgements Abstract Contents List of figures List of tables List of abbreviations

List of Figures

Chapter 1

Figure 1.1 Vascular remodelling in pulmonary arterial hypertension

Figure 1.15 Regulation of dietary iron uptake and release from from iron-utilising cells by hepcidin.

Figure 1.2 Regulation of hepcidin expression

Figure 1.3. Regulation of phosphate homeostasis

Figure 1.4 . Characterisation of iron status in patients with IPAH

Figure 1.5. Soluble transferrin receptor (sTfR) levels in patients with IPAH

Chapter 2

Figure 2.1. Normalization of cardiac output data

Figure 2.2. Characterization of Iron Status, Hb and inflammation in 153 patients with Idiopathic PAH Cohort 1

Figure 2.3. Relationship between sTfR and mortality

Table 2.3. Iron deficiency and ID

Figure 2.4. sTfR associations using standard iron markers and inflammation (IL-6)

Figure 2.5. Scatter plot showing the relationship of RDW with sTfR

Figure 2.6. ROC curves showing the ability of RDW and RDW corrected for IL-6 to predict sTfR> 28.1 Figure 2.7. Correlation between iron and Tsat in patients with IPAH

Figure 2.8. Iron status and RDW in patients with IPAH , PAH-CTD , PAH-CHD and CTEPH

Figure 2.9. Scatter plot demonstrating the relationship of age (years) with RDW

Figure 2.10. The relationship of RDW with 6MWD

Figure 2.105 ROC models for 3 year survival with 6MWD, RDW and cardiac output in IPAH.

Figure 2.11. Kaplan-Meier curves stratified by median RDW (14.8%) in patients with Idiopathic PAH

Figures 2.12 Kaplan-Meier estimates stratified by median iron (12.0 µmol/L) (A) and transferrin saturation (20%) (B) in patients with Idiopathic PAH.

Figure 2.13. ROC models showing RDW, CRP and age as predictors of 3 year survival in PAH-CTD.

Figure 2.14. Kaplan-Meier estimates stratified by median RDW (16%) and CRP (7 mg/L).

Figure 2.15. Kaplan-Meier estimates stratified by median iron (10 µg/L) and transferrin saturation (16.0 %).

Figure 2.16. Kaplan-Meier estimates stratified by median RDW (14.8%).

Figure 2.17. ROC models for 3 year survival with 6MWD, RDW and cardiac output in CTEPH.

Figure 2.18 Kaplan-Meier estimates stratified by median cut-off for RDW (14.9%).

Figure 2.19 Kaplan-Meier estimates stratified by median cut-offs for iron

Figure 2.20 Kaplan-Meier estimates stratified by median cut-offs for transferrin saturation (22.0%).

Chapter 3

Figure 3.1. Schematic overview of a 36-week, double-blind, randomized, placebo-controlled, crossover study

Figure 3.2. Ferritin and TIBC in IPAH patients before and after the administration of iron or saline

Figure 3.3. Ferritin levels in IPAH patients before and after the administration of iron

Figure 3.4. Iron and transferrin saturation levels in IPAH patients before and after the administration of iron or saline

Figure 3.5. Ferritin, iron and transferrin saturation levels in IPAH patients before and after the infusion of Ferinject or saline

Figure 3.6. Haemoglobin levels in IPAH patients before and after the administration of saline or iron

Figure 3.7. Haematocrit in IPAH patients before and after the administration of saline or iron

Figure 3.8. MCV levels in IPAH patients before and after the administration of saline or iron

Figure 3.9. RDW in IPAH patients before and after the administration of saline or iron

Figure 3.10. Circulating phosphate levels in IPAH patients before and after the administration of iron or saline

Figure 3.11. Changes in mean right atrial and pulmonary artery pressures, and subsequent pulmonary vascular resistance, 12 weeks after receiving either saline or iron infusion

Figure 3.12 Changes in ventricular output measured by cardiac index and stroke volume, and

SVO2 measured 12 weeks after receiving saline or iron infusions Figure 3.13. Changes in 6MWD at 12 weeks either having received intravenous saline or iron treatment

Figure 3.14. Changes in peak VO2 at 12 weeks either having received intravenous saline or iron treatment

Figure 3.15. Changes in peak work and oxygen pulse at 12 weeks either having received intravenous saline or iron treatment

Figure 3.16. Changes in OUES and VE/VCO2 at 12 weeks either having received intravenous saline or iron treatment

Figure 3.17. Changes in time to achieve anaerobic threshold on an incremental CPET, 12 weeks after having received intravenous saline or iron

Figure 3.18. Changes in VO2 at anaerobic threshold on incremental CPET, 12 weeks after having received intravenous saline or iron

Figure 3.19. Peak VO2 and heart rate HR; VO2 at 90 seconds into recovery; heart rate at 60 seconds into recovery; 12 weeks having received either intravenous saline or iron

Figure 3.20. Changes in endurance time and peak VO2 at submaximal exercise, 12 weeks after receiving either intravenous saline or iron

Figure 3.21 Changes in oxygen pulse during steady state submaximal exercise (O2 pulse at 3 mins) and peak oxygen pulse, 12 weeks after receiving either intravenous saline or iron

Chapter 4

Figure 4.1. The proximity of the pleura and mid- to lower oesophagus

Figure 4.2. The anatomical positioning of the oesophageal balloon catheter tip

Figure 4.3. Haemodynamic traces for a female patient, aged 69, with a body mass index of 55 kg/m^2.

Figure 4.4. Relationship between body mass index and intrathoracic pressure as assessed by intra-oesophageal pressure

Figure 4.5. Relationship between intrathoracic pressure as assessed by intra-oesophageal pressure and pulmonary wedge pressure

Figure 4.6. Relationship between change in intrathoracic pressure as assessed by intra- oesophageal pressure and pulmonary wedge pressure during Valsalva manoeuvre (A) and extrathoracic applied pressure (B).

Figure 4.7. Pulmonary wedge pressure before (1) and after (2) fluid challenge in 8 patients.

Figure 4.8. Pulmonary wedge pressure corrected for intra-oesophageal pressure.

List of Tables

Chapter 1

Table 1.1 Classification of pulmonary hypertension

Table 1.2. Intravenous iron preparations

Chapter 2

Table 2.1. Baseline characteristics of Idiopathic PAH patients in Cohort 1

Table 2.2. Baseline characteristics of incident IPAH, PAH-CTD, PAH-CHD and CTEPH patients Cohort 2 Table 2.3. Iron deficiency Table 2.4. Iron and ferritin ROC derived cut-offs predicting sTfR level >28.1 nmol/L

Table 2.5. Transferrin saturation and RDW derived cut-offs predicting sTfR >28.1nmol/L Table 2.6. Comparison of ROC derived cut-offs predicting true iron deficiency in patients with IPAH Table 2.7. Chi-squared (ᵡ2) analysis showed that iron deficiency was predicted with 54% sensitivity and 71% specificity using Ferritin <37.5µg/L or TSAT <21.5% Table 2.8. Chi-squared (ᵡ2) analysis showed that iron deficiency was predicted with 58% sensitivity and 91% specificity using Ferritin <37.5µg/L or TSAT <21.5% Table 2.9. sTfR, soluble transferrin receptor (nmol/L) relationship in 153 idiopathic PAH patients (Cohort 1) with other variables

Table 2.10. Stepwise Multiple Linear Regression model from Cohort 1 showing independent predictors of sTfR

Table 2.11. Stepwise Multiple Linear Regression model from Cohort 1 showing independent predictors of RDW

Table 2.115 Positive and negative predictive values of RDW predicating sTfR

Table 2.12. Stepwise Multiple Linear Regression model from Cohort 1 showing independent predictors of sTfR (dependant variable)

Table 2.13. Frequency of iron deficiency in four subtypes of pulmonary hypertension

Table 2.14. Frequency of iron deficiency in four subtypes of pulmonary hypertension

Table 2.15. Relationship of RDW with gender, age, and serum BNP, CRP, creatinine.

Table 2.16. Relationship of RDW with cardiac magnetic resonance imaging.

Table 2.17. Relationship of RDW with cardiopulmonary haemodynamic function.

Table 2.18. Relationship of RDW with exercise

Table 2.19. Independent associations between invasive haemodynamics and exercise

Table 2.20. Cox regression analysis, showing all parameters (segregated by modality) as predictors of overall mortality in patients with IPAH

Table 2.21. Cox modelling, assessing if RDW is an independent predictor of survival against other variables as a univariate analysis

Table 2.22. Mean survival time based on median cut-offs for RDW, iron and transferrin saturation, in patients with Idiopathic PAH

Table 2.23. Correlation of RDW with gender, age and blood biomarkers.

Table 2.24. Correlation of iron and prognostic markers of cardiac structure and function.

Table 2.25. Correlation of RDW and invasive haemodynamics

Table 2.26. Correlation of RDW and invasive haemodynamics

Table 2.27 Correlation of RDW and exercise

Table 2.28. Independent associations with RVEF, right ventricular ejection fraction (%).

Table 2.29 Cox regression analysis, showing all parameters (segregated by modality) as predictors of overall mortality in patients with PAH-CTD

Table 2.30. Mean survival time following stratification by median cut-offs for RDW, CRP, iron and transferrin saturation

Table 2.31. Associations between RDW, haemoglobin with gender, age, and blood biomarkers prognostic in PAH.

Table 2.32. Associations between RDW and haemoglobin with cardiac magnetic resonance imaging

Table 2.33. Associations between RDW, haemoglobin with invasive haemodynamics

Table 2.34. Associations between RDW, haemoglobin with exercise.

Table 2.35 Independent associations with right atrial pressure, cardiac index, and peak VO2.

Table 2.36. Cox regression analysis, showing all parameters (segregated by modality) as predictors of overall mortality in patients with PAH-CHD

Table 2.36. Cox modelling, assessing if RDW is an independent predictor of survival in PAH-CHD

Table 2.37. Correlations between RDW with gender, age and blood biomarkers

Table 2.38. Correlations between RDW with cardiac magnetic resonance imaging parameters.

Table 2.39. Correlations between RDW, with invasive haemodynamics.

Table 2.40. Correlations between RDW with exercise.

Table 2.41. Multiple linear regression analysis of individual variables with cardiac magnetic resonance outputs, invasive cardiopulmonary haemodynamics and exercise in CTEPH

Table 2.42. Cox regression analysis, showing all parameters (segregated by modality) as predictors of overall mortality in patients with CTEPH

Table 2.43. Cox modelling, assessing if RDW is an independent predictor of survival in CTEPH

Chapter 3

Table 3.1 Visit and assessment schedule

Table 3.2 Guide to last doses of medication prior to catheterisation

Chapter 4

Table 4.1 Demographic data for patients

Table 4.2 Echocardiographic data for patients

Table 4.3 Haemodynamic data for patients

Chapter 1 - Introduction 1.1 The pulmonary circulation – structure and function 1.2 Pulmonary hypertension 1.3 Pulmonary arterial hypertension (PAH) 1.4 Pathobiology of PAH 1.5 Metabolic changes in PAH 1.6 Genetics of PAH and pulmonary hypertension 1.7 Therapeutic options 1.8 Disease severity and progression 1.9 Exercise capacity and testing in pulmonary hypertension 1.10 Six-minute walk test 1.11 Cardiopulmonary exercise testing (CPET) 1.12 Biomarkers in PAH 1.13 Iron homeostasis 1.14 Iron uptake, transport and storage 1.15 Regulation of hepcidin 1.16 Genetic basis of hepcidin and iron dysregulation

1.17 Extra-hepatic expression of hepcidin and iron regulatory genes

1.18 The integrated stress response and iron

1.19 Determination of iron status

1.20 Iron deficiency in cardiovascular disease

1.21 Intravenous iron preparations and safety

1.22 Iron deficiency and phosphate homeostasis

1.23 Influence of iron on pulmonary vascular function and pulmonary hypertension

1.24 Iron deficiency in PAH

1.25 Obesity and pulmonary hypertension

2. Hypothesis 3. Aims

1. Introduction

1.1 The pulmonary circulation – structure and function

The healthy adult pulmonary circulation is a distensible, high-capacity and low-pressure system, with relatively thin-walled blood vessels compared to the systemic circulation. The pulmonary arteries arise from the pulmonary trunk and divide into lobar pulmonary arteries, which run in parallel with the airways and transport mixed venous blood to the lungs for gas exchange. The size and structure of the vessels changes throughout the course of the pulmonary arterial tree, according to their function, and comprises elastic, muscular and non-muscular vessels. The pulmonary arterial wall has three layers – the tunica intima, the tunica media, and the tunica adventitia. In elastic and muscular arteries the intima forms the innermost layer and comprises a single layer of endothelial cells and associated connective tissue lying on an internal elastic lamina. The media forms the middle layer and comprises circumferentially‐arranged pulmonary artery smooth muscle cells and elastic laminae enclosed by an external elastic lamina. The adventitia is the outermost layer and is primarily composed of connective tissue and matrix secreting fibroblasts that serves to anchor the vessels in place. Small distal arteries may be muscular or non-muscular and where the muscle is absent, a single elastic lamina separates the intima from the adventitia. Arterioles subdivide at the level of the alveolar ducts, giving rise to an extensive plexus of capillaries that supply each alveolus and make up the bulk of the alveolar wall. In the alveolar wall the capillary endothelium is directly applied to the basement membrane supporting the alveolar epithelium, providing an interface of minimal thickness for gas exchange. The alveolar capillary bed gives rise to the pulmonary venules and veins, which transport oxygenated blood back to the heart and systemic circulation. The pulmonary veins run in the interlobular septa and converge to form larger veins that then accompany bronchi and pulmonary arteries to the hilum. The muscle layer in the pulmonary veins is relatively sparse and near their junction with the left atrium extra-pulmonary veins are surrounded by a sheath of myocardium.

1.2 Pulmonary hypertension

The normal mean pulmonary arterial pressure (mPAP) at rest is 14±3 mmHg, with an upper normal limit of approximately 20 mmHg. (Kovacs, Berghold et al. 2009, Hoeper, Bogaard et al. 2013) Pulmonary hypertension is a heterogeneous disease that is defined by a resting mPAP ≥25 mmHg determined at right heart catheterization. While mPAP≥25 mmHg defines a diagnosis of pulmonary hypertension, it is known that “borderline pulmonary hypertension” (mPAP 19-24 mmHg) has also

12 been associated with increased mortality and hospitalization. (Maron, Hess et al. 2016) Patients presenting with a resting mPAP between 21–24 mmHg also have also been shown to exhibit an abnormal pulmonary vascular response during exercise, together with reduced exercise capacity and functional status, compared to those with normal mPAP. (Lau, Godinas et al. 2016)

The global incidence of pulmonary hypertension is uncertain but the overall burden is likely to be significant when important risk factors such as residence at high altitude (Penaloza and Arias-Stella 2007) and infectious disorders such as HIV, rheumatic heart disease and chronic schistosomiasis are considered (Graham, Bandeira et al. 2010). In a recent review, the prevalence of pulmonary hypertension was estimated to be about 1% of the global population, rising to 10% in individuals aged more than 65 years (Hoeper, Humbert et al. 2016). The disease has been classified into five main categories, according to clinical pathological, physiological and therapeutic characteristics. The current classification was adopted at the 5th World symposium on pulmonary hypertension held in 2013 in Nice, France (Simonneau, Gatzoulis et al. 2013) (Table1.1). Worldwide, left ventricular and valvular disease and lung diseases are the most frequent causes of pulmonary hypertension and 80% of those affected live in the developing world (Hoeper, Humbert et al. 2016).

The symptoms of pulmonary hypertension are non-specific (unexplained dyspnoea on exertion, syncope, and/or signs of right ventricular dysfunction) and diagnostic algorithms have been developed to help determine the main cause and severity of the disease (Hoeper, Bogaard et al. 2013)(Galie et al., 2015 ESC/ERS Guidelines). These comprise an established series of non-invasive tests, with subsequent invasive right heart catheterization being the mandatory investigation to establish diagnosis. However, because diagnosis may be difficult and symptoms are non- specific, there can be a delay of several years between the onset of symptoms and diagnosis (Wilkens, Grimminger et al. 2010); (Brown, Raina et al. 2011, Strange, Gabbay et al. 2013).

1. Pulmonary arterial hypertension (PAH) 1.1 Idiopathic PAH 1.2 Heritable PAH 1.2.1 BMPR2 1.2.2 ALK-1, ENG, SMAD9, CAV1, KCNK3 1.2.3 Unknown 1.3 Drug and toxin induced 1.4 Associated with: 1.4.1 Connective tissue disease 1.4.2 HIV infection 1.4.3 Portal hypertension 1.4.4 Congenital heart diseases 1.4.5 Schistosomiasis 1′ Pulmonary veno-occlusive disease and/or pulmonary capillary hemangiomatosis 1′′. Persistent pulmonary hypertension of the newborn (PPHN) 2. Pulmonary hypertension due to left heart disease 2.1 Left ventricular systolic dysfunction 2.2 Left ventricular diastolic dysfunction 2.3 Valvular disease 2.4 Congenital/acquired left heart inflow/outflow tract obstruction and congenital cardiomyopathies 3. Pulmonary hypertension due to lung diseases and/or hypoxia 3.1 Chronic obstructive pulmonary disease 3.2 Interstitial lung disease 3.3 Other pulmonary diseases with mixed restrictive and obstructive pattern 3.4 Sleep-disordered breathing 3.5 Alveolar hypoventilation disorders 3.6 Chronic exposure to high altitude 3.7 Developmental lung diseases 4. Chronic thromboembolic pulmonary hypertension (CTEPH) 5. Pulmonary hypertension with unclear multifactorial mechanisms 5.1 Hematologic disorders: chronic hemolytic anaemia, myeloproliferative disorders, splenectomy 5.2 Systemic disorders: sarcoidosis, pulmonary histiocytosis, lymphangioleiomyomatosis 5.3 Metabolic disorders: glycogen storage disease, Gaucher disease, thyroid disorders 5.4 Others: tumoral obstruction, fibrosing mediastinitis, chronic renal failure, segmental PH

Table 1.1 Classification of pulmonary hypertension. BMPR2, bone morphogenic protein receptor type II; CAV1, caveolin-1; ENG, endoglin; HIV, human immunodeficiency virus. From Simonneau et al., JACC, 2013

1.3 Pulmonary arterial hypertension (PAH)

Pulmonary arterial hypertension (PAH) represents the first main category of pulmonary hypertension and is further divided into 5 sub-categories. These comprise: idiopathic PAH (IPAH) for disease with no identified cause or family history; heritable PAH (HPAH); drug- or toxin-induced PAH; and PAH associated with various conditions, including connective tissue diseases and congenital heart disease (Table 1.1). The diagnosis of PAH depends on excluding other forms of pulmonary hypertension and demonstrating pre-capillary pulmonary hypertension at cardiac catheterization, with raised mPAP (≥25 mmHg), normal pulmonary wedge pressure (≤15 mmHg) and elevated pulmonary vascular resistance (>3 Wood units). It is a rare disease. In Europe the incidence rates have been found to be 5-10 cases per million per year and a prevalence of 15-60 cases per million (Humbert et al. AJRCCM 2006; Peacock et al. Eur Resp J 2007; Escribano-Subias et al. Eur Resp J 2012).

Women are more likely to develop PAH, yet have better survival than men (Humbert et al. Circ 2010; Benza et al. Circ 2010). Historically the estimated median survival time for untreated patients with IPAH was 2.8 years, with 1-, 3- and 5-year survival rates of 68%, 48% and 34% respectively (D’Alonzo et al. 1991). Despite advances in the understanding of PAH and improved treatment options over the last two decades, prognosis remains poor. Recent registry studies have reported 3-year survival rates for PAH patients of 67% in France (Humbert et al., Eur Resp J 2010), 73% in the United Kingdom and Ireland (Ling et al. AJRCCM 2012) and 68% in USA (Benza et al. Chest 2012). Longer-term follow- up of patients enrolled in the US Registry to Evaluate Early and Long-term PAH Disease Management (REVEAL) demonstrated 5- and 7-year survival rates of 57% and 49% respectively (Benza et al. Chest 2012; Farber et al., Chest 2015). In the most recent national audit of pulmonary hypertension in the United Kingdom, 5 year survival for patients with idiopathic, heritable and anorexigen-induced PAH was 42% (Health and Social Care Information Centre. National audit of pulmonary hypertension 2017. http://content.digital.nhs.uk/ph). The survival rates of PAH patients with prevalent and newly diagnosed (incident) disease is reported to vary, with prevalent cohorts tending to select patients who have a better prognosis than incident patients (Humbert et al., Eur Resp J 2010; Farber et al., Chest 2015). The demographics of patients have also changed over the last two decades, with PAH now being more frequently diagnosed in patients who are older. Patients with PAH diagnosed after 50-55 years may be phenotypically distinct from younger patients, having a higher incidence of comorbidities such as system hypertension, atrial fibrillation and diabetes (Ling et al. AJRCCM 2012; McGoon et al. JACC 2013).

1.4 Pathobiology of PAH The normal pulmonary circulation is a low resistance, high compliance system that can accommodate the entire cardiac output during rest and exercise. 20 In PAH increased resistance and low compliance leads to increased right ventricular afterload.20-22 Histological analysis of lungs from affected patients demonstrates marked vascular remodelling that usually involves all three layers of the vessel wall and reduces the area of the lumen (Tuder et al. JACC 2013). The structural remodelling involves the proliferation or infiltration of endothelial cells, smooth muscle cells, fibroblasts and inflammatory cells, as well as extracellular matrix components such as collagen, elastin and fibronectin (Figure 1.1). Intimal changes include endothelial injury, endothelial cell proliferation, invasion of the intima by (myo)fibroblast-like cells, enhanced matrix deposition with intimal fibrosis and at times obstruction of the vascular lumen by unique plexiform lesions(Tuder 2014). The presence of plexiform lesions is considered a feature of end-stage disease, reflecting disorganised angiogenesis and the clonal expansion of apoptotic-resistant endothelial cells in IPAH (Lee et al. 1998). Vascular remodelling occurs throughout the pulmonary artery tree, with medial thickening due to smooth muscle hyperplasia and hypertrophy and distal muscularization of normally non-muscular vessels (Rabinovitch 2008). It is estimated that the pulmonary circulation contains ~8 million small arteries of the type affected in PAH disease and ~50% can be lost before there are clinical signs of the disease (Austin et al. Chapt 15.2 in Pul Circ 3rd Edition 2011).

Pulmonary endothelial damage and dysfunction is considered to occur early in the development of PAH and to be an important event leading to vascular remodelling (Sakao et al. 2005; Michelakis 2006). The endothelium has a critical role in the regulation of pulmonary vascular tone and is the primary source of the potent vasodilators prostacyclin (PGI2) and nitric oxide (NO) and vasoconstrictors thromboxane A2, endothelin-1 and serotonin (5-hydroxytryptamine – 5HT)

(Schermuly et al. 2011). In PAH there is a reduction in the pulmonary artery expression of PGI2 synthase (Tuder et al. ARCCM 1999) and circulating PGI2 levels are also reduced, while thromboxane A2 levels are raised (Christman et al., NEJM 1992)(Christman, McPherson et al. 1992, Kreymborg, Uchida et al. 2010).

Figure 1.1 Vascular remodelling in pulmonary arterial hypertension. Putative therapeutic targets are indicated. Abbreviations: 5-HT, 5-hydroxytryptamin; K-and Ca-channels, potassium and calcium channels; AEC, alveolar epithelial cells; BMP, bone morphogenetic protein; cGMP, cyclic guanosine monophosphate; ECM, extracellular matrix; EGF, epidermal growth factor; EPC, endothelial progenitor cells; HIF, hypoxia inducible factor; MMPs, matrix metalloproteinases; NADPH, nicotinamide adenine dinucleotide phosphate; NO, nitric oxide; PDE, phosphodiesterase; PDGF, platelet-derived growth factor; PGI2, prostaglandin I2; Rho-Ki, Rho kinases; ROS, reactive oxygen species; sGC, soluble guanylate cyclase; TGF, transforming growth factor-β; TK, tyrosine kinase; TKi, tyrosine kinase inhibitor; TRPC, transient receptor potential cation channels; VEGF, vascular endothelial growth factor. From Schermuly et al. Nat. Rev. Cardiol. 2011 . (Schermuly, Ghofrani et al. 2011)

The vasodilator activity of NO is mediated via cyclic guanosine monophosphate, the metabolism of which is dependent on phosphodiesterases (Corbin & Francis JBC 1999). Phosphodiesterase type 5 (PDE5) is the most abundant isoform in the pulmonary circulation and its expression is increased in PAH patients (Murray et al. Br J Pharm 2002; Corbin et al. BBRC 2005; Wharton et al. AJCCM 2005). Reduced bioavailability of NO in PAH may also reflect lower expression of endothelial NO synthase (eNOS) in the lung microvasculature (Giaid & Saleh NEJM 1995) and increased levels of the

17 endogenous eNOS inhibitor asymmetric dimethylarginine (Pullamisetti et al. FASEB J 2005). Raised levels of ET-1 and 5-HT are also thought to have an important pathogenic role in PAH, inducing PASMC proliferation as well as vasoconstriction (Giaid et al. NEJM 1993; Eddahibi et al., Circ 2006). The pathological increase in pulmonary vasomotor tone in PAH has also been attributed to abnormalities in potassium and calcium channels Yuan et al. 1998; Yu et al. 2004; Burg et al. 2008) and RhoA/Rho kinase signalling (Oka et al. Circ Res 2007; Mouchaers et al. Eur Resp J 2008). Endothelial dysfunction is also likely to provide a pro-thrombotic environment (Tournier et al. Thromb Res 2010), promoting the in situ thrombosis that is a common feature of PAH (White et al. Am J Physiol Lung Cell Mol Physiol 2007), and contribute to the accumulation/infiltration of inflammatory and immune cells in the lungs of PAH patients (Stacher et al. AJRCCM 2012; Savai et al. AJRCCM 2012). The relationship between inflammation and pulmonary vascular remodelling is well recognised and increasing evidence indicates that inflammatory cytokines and chemokines are important role in the pathogenesis of PAH (Dorfmüller et al. Eur Resp J 2003; Hassoun et al. JACC 2009; Price et al. Chest 2012).

1.5 Metabolic changes in PAH

Increasing evidence indicates that PAH patients have abnormalities in metabolism that involve extra- pulmonary tissues as well as the pulmonary vasculature, affected organs including the right ventricle, immune system, bone marrow and skeletal muscles (Sutendra & Michelakis Cell Metab 2013; Paulin & Michelakis, Circ Res 2014; Ryan & Archer Circ 2015). A variety of metabolic changes have been implicated in the pathogenesis of PAH, but mitochondrial dysfunction is postulated to be the critical feature that is common to the affected tissues, resulting in suppression of oxidative phosphorylation and glucose oxidation and up-regulation of glycolysis. This may reflect generalized abnormalities in mitochondrial structure/function and pathways that supress mitochondrial function, including circulating inflammatory cytokines and mitokines (Durieux et al., Cell 2011; Woo & Shadel. Cell 2011).

Mitochondria are also the primary site of two major pathways that use iron – the biosynthesis of iron-sulfur (Fe/S) clusters and haem formation (Rouault et al. Nat Rev Mol Cell Biol 2005) – and many of the enzymes important for mitochondrial and metabolic function are iron-dependent. Fe/S clusters are essential to key nuclear and cytosolic apoproteins that are vital for DNA synthesis and repair and numerous extra-mitochondrial pathways including protein synthesis and cellular iron regulation. Within mitochondria, the synthesis of Fe/S clusters and their insertion into apoproteins is supported by 17 assembly proteins that function either as scaffolds for Fe/S cluster formation or the

18 transfer and insertion of Fe/S cluster into specific apoproteins. Disturbance of this process impairs the maturation of extra-mitochondrial Fe/S proteins and affects cellular and systemic iron homeostasis (Stehling et al. Biochemie 2014). In keeping with the essential function of mitochondria and Fe/S clusters, genetic mutations in genes encoding the Fe/S cluster assembly proteins lead to severe neurological, hematological and metabolic diseases, which are often fatal in early childhood (Stehling et al. Biochemie 2014; Rouault Nat Rev Mol Cell Biol. 2015). Importantly, pathogenic mutations in two of these genes, encoding the NFU1 iron-sulfur cluster scaffold homolog and iron- sulfur cluster assembly enzyme ISCU, have been associated with pulmonary hypertension (Navarro- Sastre et al. Am J Human Genetics 2011; Ahting et al. Frontiers in Genetics 2015; White et al., EMBO Mol Medicine 2015). Reduced ISCU levels and Fe/S cluster integrity was also found in lung tissues from pulmonary hypertensive mice and patients (White et al., EMBO Mol Medicine 2015).

1.6 Genetics of PAH and pulmonary hypertension

A major breakthrough in our understanding of the molecular pathogenesis of PAH came with the identification germ line mutations in the gene encoding bone morphogenetic protein (BMP) receptor type II (BMPR2) in families with hereditable PAH (Deng et al. Am J Human Gen 2000; Lane et al. Nat Genetics 2000). This receptor is a member of the transforming growth factor-β (TGF-β) superfamily and several hundred different BMPR2 mutations have been identified with a prevalence >70% in families with HPAH (Soubrier et al. JACC 2013; Machado et al. Human Mutat 2015). The mutations result in a loss-of-function and reduced signalling in the SMAD pathway downstream of the receptor as well as the perturbation of SMAD-independent pathways. They are generally inherited in an autosomal dominant manner, although instances of de novo BMPR2 mutation have been reported (Thomson et al. J Med Genetics 2000; Girerd et al. Eur Resp J 2016) and BMPR2 mutations have also been identified in ~20% of IPAH patients without a prior family history of the disease (Thomson et al. J Med Genetics 2000; Koehler et al. J Med Genetics 2004) (Lane, Machado et al. 2000, Machado, Aldred et al. 2006) (Lane, Machado et al. 2000, Machado, Aldred et al. 2006) (Lane, Machado et al. 2000, Machado, Aldred et al. 2006) (Lane, Machado et al. 2000, Machado, Aldred et al. 2006) (Lane, Machado et al. 2000, Machado, Aldred et al. 2006) Compared to patients without BMPR2 mutations, those with BMPR2 mutations present at an earlier age, with more severe disease and have a greater risk of death (Sztrymf et al. AJRCCM 2008; Evans et al. Lancet Resp Med 2016).

Despite the importance of BMPR2 in the pathobiology of PAH, disease penetrance in carriers of BMPR2 mutations is only ~20 % and this has led to the consideration of other genetic and/or

19 environmental factors that may be required for development of the phenotype. As recently reviewed (Sourbrier et al. JACC 2013; Austin & Loyd Circ Res 2014), mutations have also been found less frequently in other genes involved in BMP signalling, such as SMAD1, SMAD4 and SMAD9, and genes belonging to the related TGF-β superfamily such as activin receptor-like kinase-1 (ACVRL1, also known as ALK1) and endoglin (ENG). The use of advanced gene-sequencing methods has led to the discovery of further genes variants in PAH patients, including those encoding caveolin 1 (CAV1), potassium channel subfamily K, member 3 (KCNK3) and eukaryotic translation initiation factor 2 alpha kinase 4 (EIF2AK4, also known as GCN2). Disease causing EIF2AK4 mutations were first identified in patients classified as group 1’ PAH, with pulmonary veno-occlusive disease (PVOD) and/or pulmonary capillary hemangiomatosis (PCH) (Table 1.1), but were recently also found in an aggressive form of PAH in Iberian Gypsies (Tenorio et al., Clin Genetics 2015). In addition to mutations in the nine genes known to be associated with PAH (i.e. BMPR2, SMAD1, SMAD4 and SMAD9, ACVRL1, ENG, CAV1, KCNK3, EIF2AK4 ), somatic mutations have been reported in the lungs of PAH patients (Aldred et al. AJRCCM 2010) and more common polymorphisms found in other genes such as cerebellin 2 (CBLN2) (Germain et al. Nat Genet 2013), potassium channel, voltage gated shaker-related subfamily A, member 3 (KCNA5) (Remillard et al. Am J Physiol 2007;Wang et al. Int J Cardiol 2014), prostacyclin synthase (PGIS) (Stearman et al. AJRCCM 2014), endothelin-1 receptor G-protein (GNG2) (Benza et al. AJRCCM 2015), endostatin (Col18a1/ES) (Damico et al. AJRCCM 2015), and BMP9 (GDF2) (Wooderchak-Donahue Am J Hum Genet 2013; Wang et al. BMC Pul Med 2016). It is thought that these represent modifier genes in PAH and mutations act as a ‘second-hit’ that influence the manifestation and severity of the disease and response to treatment. For example, variants in the PGIS promoter region correlated with transcriptional activity and were postulated to have a protective role in unaffected BMPR2 carriers (Stearman et al. AJRCCM 2014). Mutations in regulatory (non-coding) regions of the BMPR2 gene have also been considered as a potential ‘second-hit’ influencing predisposition to PAH (Aldred et al. AJRCCM 2007; Wang et al. Eur J Hum Genet 2009; Viales et al. PLOSOne 2015). Conversely, a single-nucleotide polymorphism in the GNG2 gene has been identified that might influence the clinical efficacy, associating with better clinical outcomes in PAH patients following endothelin receptor antagonist-based therapy (Benza et al. AJRCCM 2015).

As well as genetic variants, there is also a growing interest in the role of non-coding RNAs such as microRNAs (miRNAs) and epigenetic mechanisms in the pathogenesis of PAH, including DNA methylation and histone modification in the post-transcriptional regulation of gene expression (see

Bienertova-Vasku et al. J Am Soc Hypertens 2015; Zhou et al. AJRCMB 2015; Kim et al. Exp Mol Med 2015).

1.7 Therapeutic options

Most patients with PAH receive supporting therapies, comprising oral anticoagulants, diuretics, oxygen, digoxin and other cardiovascular drugs. A small number of IPAH/HPAH patients, who exhibit pulmonary vasoreactiivty when tested with short acting vasodilators, such as inhaled nitric oxide, respond to calcium channel blockers (CCBs) and show marked improvements in quality of life and survival (Rich et al. 1992; Sitbon et al. 2005), however, the vasodilatory response can decline over time and it does not predict a beneficial outcome following CCB therapy in other types of PAH. Since the 1990s several specific vasodilator therapies have been developed that target abnormalities in the prostacyclin, endothelin and NO pathways (Humbert et al. Circ 2014). These include prostacyclin analogues, endothelin receptor antagonists and PDE5 inhibitors and meta-analysis suggests that their introduction has led to a significant improvement in the symptoms and clinical end-points of patients with PAH (Galie et al. Eur Heart J 2009). More recently, guanylate cyclase stimulators, prostacyclin receptor agonists have been introduced and the efficacy of combination treatment examined (Humbert et al. Circ 2014). Meta-analysis of 4095 patients enrolled in 17 trials has now demonstrated that a combination of these PAH-specific is more effective than monotherapy in reducing clinical worsening and improving functional class and exercise capacity (Lajoie et al., Lancet Resp Med 2016). However, a substantial number of patients do not respond to these vasodilator treatments and responsiveness declines with disease progression. There is a need for additional therapies targeting alternative pathways that can reverse pulmonary vascular remodelling, inhibit disease progression and improve survival. Disappointing results have been found in trials examining compounds such as tyrosine kinase inhibitors, statins, vasoactive intestinal polypeptide and serotonin antagonists, but greater understanding of the cellular and molecular mechanisms that underlie PAH has led to investigations examining a wide range of other therapeutic targets. These include compounds that target mitochondrial and metabolic function (e.g. dichloroacetate, an inhibitor of pyruvate dehydrogenase kinase, and proliferator-activator receptor agonists), inflammation (e.g. antibodies against CD20 and IL-6), RhoA/Rho-kinase signalling, proteases and elastases, intracellular calcium, and epigenetic factors such as histone deacetylase (HDAC) activity and miRNAs (Humbert et al. Circ 2014). In addition, therapies are being examined that may help restore BMPR2 signalling, such as BMP-9 (Long et al., Nat Med. 2015 Jul;21(7):777-85). Device-based therapies are also in development for patients with medication-refractory PAH, including pulmonary

21 artery denervation using radiofrequency ablation / high-energy ultrasound (Chen et al. Circ Cardiovasc Interv. 2015; Rothman et al. Circ Cardiovasc Interv. 2015).

1.8 Assessing disease severity and progression

Regular clinical assessment is required to evaluate disease progression, response to therapy and prognosis. Haemodynamic measurements, both at baseline (diagnosis) and during follow-up, are established predictors of disease severity and survival in PAH patients, with right atrial pressure, cardiac index and mixed venous oxygen saturation being the most robust haemodynamic indices of right ventricular function and prognosis (Sitbon et al., JACC 2002; McLaughlin et al. 2005; Nickel et al. ERJ 2012; Rich et al., ERJ 2013). The World Health Organization (WHO) functional class of patients is also a powerful indicator of disease progression, clinical deterioration and survival in PAH (Sitbon et al., JACC 2002; Benza et al., Circ 2010; Nickel et al. ERJ 2012; Barst et al. Chest 2013).

1.9 Exercise capacity and testing in pulmonary hypertension

In addition to the pulmonary vasculature, pulmonary hypertension affects cardiac function, cardiopulmonary gas exchange, oxygen transport and peripheral muscle metabolism. At rest, there are normally reserves in cardiac function, pulmonary vascular capacity and energy stores and it is only during exercise that these are challenged, with limitations in oxygen uptake, transport and metabolism becoming apparent (Dumitrescu & Oudiz, Chapt 10 Pul Circ 2011). During modest exercise, the uptake of oxygen for the generation of energy (ATP) and muscle contraction matches demand. As exercise intensity increases and the demand for oxygen exceeds the available tissue oxygen content and substrates are metabolized anaerobically rather than aerobically. The exercise level that can be sustained by aerobic metabolism varies between individuals and in patients with compromised cardiopulmonary function the anaerobic threshold is reduced due to impaired oxygen uptake and/ or transport and physical inactivity. There is a limited ability to increase cardiac output and optimize lung ventilation/perfusion during maximal exercise and the point of maximal exercise is reached sooner when cardiopulmonary mechanisms are impaired (Dumitrescu & Oudiz, 2011).

Insulin resistance and metabolic abnormalities are common in PAH patients and involve the skeletal muscles as well as the pulmonary vasculature and right ventricle (Sutendra & Michelakis Cell Metab 2013; Paulin & Michelakis, Circ Res 2014; Ryan & Archer Circ 2015). A variety of abnormalities in

22 both skeletal and respiratory muscles have been found in PAH patients, including a switch from type I to type II muscle fibres, reduced muscle capillary density, impaired mitochondrial function, lower aerobic enzyme activity and excitation-contraction coupling (Bauer et al. Resp Med 2007; Mainguy et al. Thorax 2010; Batt et al., AJRCMB 2014; Breda et al. PLOSOne 2014). Differences in skeletal muscle protein expression have also recently been described that relate to abnormal mitochondrial structure and function (Malenfant et al. J Mol Med 2015). Muscle dysfunction correlates with the impairment of exercise capacity in PAH patients and recent trials indicate that exercise training can improve functional capacity and quality of life (de Man et al., Eur Resp J 2009; Mainguy et al. J Cardiopulm Rehabil Prev 2010; Grünig et al., Eur Resp J 2012; Kabitz et al. Lung 2014).

1.10 Six-minute walk test

The six minute walk test (6MWT) is the most widely used submaximal test of exercise capacity. It is inexpensive, technically simple, well tolerated and assesses two of the most common symptoms of patients with pulmonary hypertension – exercise limitation and dyspnoea – and is an indicator of cardiac impairment. As a result, a change in 6 minute walk distance (6MWD) has been the most commonly used primary end-point in clinical trials of therapies for PAH (Galie et al., Eur Heart J 2009). Because of its relationship with functional class, baseline cardiac output and total pulmonary resistance, 6MWD is generally considered a surrogate for disease severity in PAH (Miyamoto et al. AJRCCM 2000). The results of the 6MWT also reflect exercise capacity determined by maximal cardiopulmonary exercise testing (CPET) and show an independent association with mortality (Miyamoto et al. AJRCCM 2000).

Despite being a useful metric in the evaluation and management of patients, several factors may confound the use of 6MWD as a measure of exercise capacity related to PAH (Rasekaba et al. Intern Med J. 2009). These include the patient’s age, gender, height, weight, comorbidities, oxygen dependency, deconditioning, experience of the test and motivation on the test day. The utility of the 6MWT as a primary end-point in PAH trials has also been challenged (Gaine & Simonneau Eur Resp Rev 2013). While, absolute 6MWD values obtained at baseline (before therapy) and following therapy have prognostic value, the change from baseline 6MWD does not correlate well with long- term outcomes and is not a suitable predictor of clinical worsening or mortality (Macchia et al., Am Heart J. 2007 & 2010, Savarse et al., JACC 2012; Gabler et al., Circ 2012). Composite end-points, such as the time to clinical worsening, have therefore been employed as an end-point to explore the effects of therapies on disease progression (Preston et al., Eur Resp Rev 2013).

1.11 Cardiopulmonary exercise testing

Cardiopulmonary exercise testing (CPET) complements the 6MWT and is considered the ‘gold standard’ means of assessing exercise performance in pulmonary hypertension (Babu et al. Expert Rev Cardiovasc Ther 2013; Galie et al. ERJ 2015). CPET is usually performed as a maximal exercise test and provides quantitative information on exercise capacity, gas exchange, ventilation efficiency and cardiac function in response to exercise. It is a recognised test for the diagnosis and follow up of PAH patients in both Europe and North American( (Guazzi, Adams et al. 2012) Galie et al. ERJ 2015).

The cycle ergometer and treadmill and are the most commonly used testing CPET modalities for assessing patients with PAH. Several protocols have been developed are available and generally those lasting less than 6 minutes give a lower association between oxygen consumption (VO2) and workload, whilst protocols lasting over 12 minutes are more likely to be terminated by the patient due to peripheral musculoskeletal limitations or lack of motivation rather than cardiopulmonary mechanisms (Babu et al. Expert Rev Cardiovasc Ther 2013). Incremental ramp protocols are common place as many PAH patients present with significant exercise limitation. A gradual increase in workload is most appropriate and often provides a better prediction of peak VO2 as well as more uniform gas exchange and haemodynamic responses (Babu et al. Expert Rev Cardiovasc Ther 2013).

CPET software packages produce a wide range of data load and the clinical significance of many variables has not been validated (Guazzi et al. Circ 2012). The most commonly used variables include measurements of ventilation (VE), carbon dioxide output (VCO2), and oxygen uptake (VO2), obtained during a progressive increase in workload (W) until a maximum or peak VO2 (VO2max or

VO2peak) is reached and used to define aerobic exercise capacity. The effort level of the subject is assessed by the respiratory exchange ratio (RER), which is the ratio of VCO2/VO2, and at peak VO2 the RER ≥1.15. Other relevant CPET measurements include VO2 at the anaerobic threshold (AT); the

VE /VCO2 slope; relationship of VO2 to work rate (∆VO2/∆W); ratio between VO2 and heart rate

(oxygen pulse); the end tidal partial pressure of CO2 (PETCO2); physiological dead space; and heart rate recovery (HRR). No single CPET variable is able to establish the diagnosis of PAH, but a combination of variables may be useful in monitoring disease severity, response to treatment and in the stratification of patients (Galie et al. ERJ 2015). Several CPET variables are consistently altered in pulmonary hypertension and class-based recommendations with associated levels of evidence have recently been published for the use of CPET in patients with pulmonary hypertension (Pinkstaff et al., Expert Rev Resp Medicine 2016).

1.12 Biomarkers in PAH

Biomarkers, including biochemical and physical measurements, may be used to distinguish the phenotypic characteristics of patients, assess pathogenic and biological processes, inform the selection of therapy, evaluate the response to treatment and assess long-term outcomes (Rhodes et al. Blood Biomarkers Chapt 11, Pul Circ 3rd Edition 2011). Although right heart catheterization is still the gold standard to establish a diagnosis of PAH, non-invasive imaging techniques, in particular magnetic resonance imaging (MRI), are progressing rapidly and MRI findings provide useful prognostic information (Baggen et al. Eur Radiol 2016). Brain natriuretic peptide (BNP) or the N- terminal fragment of pro-BNP (NT-proBNP) is the only circulating biomarker that is currently recommended for risk stratification in PAH (Galie et al. ERJ 2015). However, changes in the level of this cardiac-derived protein are affected by renal function and are not specific to PAH.

Many other potential circulating biomarkers have been investigated in patients with PAH, as well as other forms of pulmonary hypertension. Of particular interest to the subject of this thesis is red cell distribution width (RDW), which reflects variability in erythrocyte size (anisocytosis) and is commonly reported in automated full blood counts diseases (Danse et al. J Thorac Dis 2015; Savagno et al. Crit Rev Clin Lab Sci 2015). It is expressed as a percentage and defined as:

RDW (%) = (standard deviation of red cell volume/mean cell volume) x 100

Increased RDW is seen in iron-deficiency anaemia, thalassaemias and myelodysplastic syndromes and predicts poor clinical outcomes in several cardiovascular diseases and human disorders, as recently reviewed (Danse et al. J Thorac Dis 2015; Savagno et al. Crit Rev Clin Lab Sci 2015). RDW has also been found to vary in the general population and increase with age (Patel et al. Arch Intern Med. 2009; Söderholm et al. PLOSOne 2015; Lappegård et al. Thromb Haemost 2015). Patients with heart failure have been most extensively studied and increased RDW attributed to factors such as dysfunctional erythropoiesis, chronic inflammation, renal dysfunction, vitamin D Deficiency and oxidative stress (Felker et al. JACC 2007; Fӧrhécz et al. Am Heart J 2009, Al-Najjar et al. Eur J Heart Failure 2009; Allen et al. J Card Fail 2010). Higher RDW values were also associated with iron deficiency, being accompanied by reduced serum iron, ferritin levels, transferrin saturation and increased soluble transferrin receptor concentration (Fӧrhécz et al. Am Heart J 2009). RDW may therefore reflect a combination of factors, many of which are relevant to the prognosis of PAH. In patients with pulmonary hypertension of mixed aetiologies (Hampole et al. Am J Cardiol 2009) and those with IPAH (Rhodes et al. Heart 2011), increased RDW levels correlate with disease severity and

25 independently predict survival, even when measured in combination with 6MWD, NT-proBNP and other established clinical indices.

1.13 Iron homeostasis

Iron is a transition metal that has a major role in most biological systems and metabolic pathways, being a critical component of oxygen transporting proteins and numerous enzymatic processes due to its ability to participate in oxidation/reduction reactions. While iron is essential, its distribution in biological fluids must be closely controlled as free (labile) iron can generate reactive oxygen species (ROS) that cause oxidative damage to proteins, lipids and nucleic acids and inhibit enzymatic activity (Meneghini-R . Free Radical Biol Med 1997). The adult human contains 3-5g of iron, of which 60-80% is present in the haemoglobin of circulating erythrocytes and at sites of iron recycling and storage in macrophages and hepatocytes, and 10-15% in muscle myoglobin and other iron-containing proteins (ferroproteins). Generally, only about 0.1% of the body’s iron is present in the circulation bound to transferrin and 1-2mg is lost per day due to desquamation and blood loss. While the loss of iron is not regulated, the intestinal uptake, transport, and recycling of iron is tightly controlled by the key iron-regulatory hormone hepcidin and ferroportin, the only known iron exporter from cells. Hepcidin binds to ferroportin on the cell membranes of enterocytes and macrophages, rapidly inducing its internalization and lysosomal degradation (Lawen & Lane. Antioxidants & Redox Signalling 2013; Rochette et al. Pharmacology & Therapeutics 2015).

1.14 Iron uptake, transport and storage

Iron absorption by duodenal enterocytes involves the uptake of non-haem iron by the divalent metal ion transporter (DMT1) (Table 1.15, following oxidation to the ferrous ion (Fe2+) by oxidoreductases such as Dcytb, a cytochrome b reductase 1 (Lawen & Lane. Antioxidants & Redox Signalling 2013; Rochette et al. Pharmacology & Therapeutics 2015). In addition to Dcytb, other cellular ferrireductases include cytochrome b561 and stromal cell-derived receptor 2 (SDR2/FRRS1) (Lane et al. Nutrients 2015). While DMT1 and Dyctb have a critical role in dietary iron absorption, they are also expressed on a variety of tissues and have a more general function. Extracellular iron is generally found in complexes with iron-binding proteins or chelators, the most important being transferrin and citrate respectively. Iron uptake by cells occurs via both transferrin-dependent and independent mechanisms. Iron is taken up by cells as Fe2+ and upon its release into the circulation via ferroportin it is oxidized to the ferric (Fe3+) state by ferroxidases such as caeruloplasmin. In the circulation, Fe3+ is specifically bound to the glycoprotein transferrin, which can bind one or two Fe3+ ions to form mono- or diferic- (holo-) transferrin. Saturation of transferrin is typically 20-30% and increased saturation signals an increase in hepatocyte hepcidin expression (Lawen & Lane. Antioxidants & Redox Signalling 2013; Rochette et al. Pharmacology & Therapeutics 2015).

Holotransferin binds with high affinity to the widely expressed membrane protein transferrin receptor 1 (TfR1) and the complex undergoes endocytosis. Before being released in to the cytoplasm, via DMT1 or natural resistance-associated macrophage protein 2 (Nramp2), the Fe3+ ion is reduced to Fe2+ by ferrireducatses such as the six-transmembrane epithelial antigen of prostate-3 (STEAP3) expressed in erythroblasts. Recent data indicate that STEAP3 is an important protein required for the efficient transferrin-dependent uptake of iron in erythroid cells (Ohgami et al. Nat Genetics 2005; Zhang et al. Haematologica 2012; Jabara et al. Nat Genetics 2016). Whereas TfR1 is primarily concerned with the uptake of iron into cells, the second transferrin receptor (TfR2) is mainly expressed in the liver and thought to function as an iron sensor (Goswani et al. JBC 2006). The transmembrane protein HFE (also known as the hereditary haemochromatosis protein), encoded by the high-Fe (HFE) gene, interacts with both transferrin receptors and appears to be important in transmitting information about raised circulating holotransferrin levels and the subsequent upregulation of hepcidin expression (Lawen & Lane. Antioxidants & Redox Signalling 2013; Rochette et al. Pharmacology & Therapeutics 2015).

Virtually all iron in plasma is bound to transferrin under physiological conditions, with a minor proportion existing in complex with haem and ferritin. Besides iron associated with these proteins, another pool of iron, ‘non-transferrin bound iron (NTBI)’, has been identified that is bound to small molecules such as citrate. NTBI is thought to be particularly important in iron overload diseases, where transferrin is saturated, but low levels (<1 µmol/L) are reported in normal individuals (Brissot et al. Biochem et Biophys Acta 2012; Lawen & Lane. Antioxidants & Redox Signalling 2013; Rochette et al. Pharmacology & Therapeutics 2015). Circulating NTBI levels may rise to 10-30 µmol/L immediately after iron treatment, which is sufficient to induce transient changes in endothelial cell function in vitro (Mollet et al. PLOSOne 2016). However, NTBI is rapidly taken up the liver, as well as other tissues, via transporter proteins. In addition to DMT1, proteins such as the Zrt-Irt-like protein ZIP14 (SLC39A14) may also be involved in the uptake of NTBI by hepatocytes (Liuzzi et al. PNAS 2006). Interestingly, ZIP14 has also been implicated in the cellular uptake of iron from transferrin (Zhao et al. JBC 2010) and in the transport of other metals (Liuzzi et al. PNAS 2006; Pinilla-Tenas et al. Am J Physiol 2011). A number of other iron-regulatory proteins (e.g. DMT1, ferroportin) are also associated with the metabolism of metals such as zinc, cobalt, manganese and lead and, like iron, these may influence hepcidin expression (Loréal et al. Frontiers in Pharmacol 2014). Another member of the ZIP family, ZIP12 (SLC39A12) has also been implicated in the development of pulmonary hypertension, regulating the pulmonary vascular response to hypoxia (Zhao et al. Nature 2015).

Figure 1.15. Regulation of dietary iron uptake and release from from iron-utilising cells by hepcidin. Hepcidin internalises ferroportin leading to its degradation and inhibition of dietary iron uptake and release of iron-utilising (storing) cells. Hepcidin is regulated by several factors, through bone morphogenic protein (BMP) signalling through the Smad transcription factors..

Hepcidin expression can be stimulated by inflammation (through interleukin-6, IL-6) and circulating iron (Fe-transferrin). Erythropoiesis (and erythropoietin, EPO) indirectly inhibits hepcidin and circulating erythroid factors such as grown differentiation factor – 15 (GDF -15) and twisted gastrulation (TWSG1), have been proposed to directly inhibit hepcidin. Divalent metal transporter-1 (DMT-1) is responsible for uptake of dietary iron into enterocytes.

Systemic iron homeostasis involves the trafficking of haem, as well as elemental non-haem iron, and a number of cells have the ability to take up and export haem via alternative transporters. These include the haem carrier protein 1 (HCP1; SLC46A1), which is responsible for intestinal uptake by enterocytes (Shayeghi et al. Cell 2005), the fowler syndrome-associated protein FLVCR2 that imports haem (Duffy et al. Mol Cell Biol 2010), and the FLVCR1 isoforms that export haem (Qiugley et al. Cell 2004; Keel et al. Science 2008; Chiabrando et al. JCI 2012; Doty et al. JCI 2015). While cellular haem levels are closely regulated, balancing haem synthesis and catabolism, free haem may increase for example in conditions associated with excess haemolysis, such as sickle cell disease, leading to oxidative stress and inflammation (Sawicki et al. JAHA 2015). The degradation of haem is mediated by two isoforms of haem-oxygenase, converting haem to billiverdin, free iron and carbon monoxide (Sawicki et al. JAHA 2015). Haem-oxygenase-2 is ubiquitously expressed and haem-oxygenase -1 (HO-1) normally expressed in macrophages responsible for systemic haem degradation, but is induced in response to raised circulating haemoglobin and oxidative stress (Ryter et al. Physiol Rev 2006). Haem can influence the expression of other iron-regulatory genes, such as ferroportin and transferrin receptor, via haem response elements and the transcription factor Bach1 (Hintze et al. JBC 2007; Zenke-Kawasaki et al. Mol Cell Biol 2007). The rate-limiting step in haem biosynthesis is mediated by the enzyme 5-aminolevulinic acid synthase 2 (ALAS2) in erythroblasts and ALAS1 in all other cell types. These two isoenzymes are differentially regulated; ALAS1 expression being regulated by haem whereas erythroblast ALAS2 lacks any direct feedback inhibition and the availability of iron drives haem synthesis (Ajioka et al. Biochim Biophys Acta 2006). Ferrochelatase (FECH) catalyses the final step in haem biosynthesis, inserting Fe2+ into protoporphyrin IX to form haem (Ponka et al. Blood 1997).The acute-phase glycoprotein hemopexin scavenges free haem with high affinity, the resulting complex being taken up by the liver (Paoli et al. Nat Struct Biol 1999; Hvidberg et al. Blood 2005). In addition, cell free haemoglobin is sequestered by haptoglobin, the haemoglobin-haptoglobin complex having high affinity for the macrophage CD163 scavenger receptor (Kristiansen et al. Nature 2001; Schaer et al. Blood 2006). Upon binding to this receptor, the whole complex is internalised, degraded and iron released from haem by HO-1.

Most iron storage occurs in hepatocytes with shorter term storage in reticuloendothelial cells, comprising splenic macrophages and Kupffer cells in the liver (Lawen & Lane. Antioxidants & Redox Signalling 2013; Rochette et al. Pharmacology & Therapeutics 2015). After the uptake of iron in to the cytoplasm, iron is mainly stored incorporated in ferritin. When Fe2+ binds to ferritin, it is rapidly oxidized to Fe3+ by its intrinsic ferroxidase activity, thereby preventing the formation of ROS. Like cytoplasmic ferritin, mitochondrial ferritin (FTMT) sequesters iron in mitochondria and is preferentially expressed in cells associated with high energy consumption and usually absent from erythorid cells and the iron storage organs, the spleen and liver. Over expression of FTMT can disrupt iron homeostasis and have pathological consequences

The mitochondria have a central role in iron metabolism and, apart from storage in ferritin, represent the other major destination of iron in mammalian cells. They are the sole site of haem synthesis and the major site of Fe/S cluster formation (see section 1.5 above) and as many of the key proteins in cellular metabolism are dependent on these iron-containing complexes, severe diseases are associated with the dysregulation of mitochondrial iron metabolism (Huang et al. Antioxidants & Redox Signalling 2011; Lane et al. Biochemica et Biophysics Acta 2015). In order to be processed in to haem and Fe/S clusters, iron has to cross the outer and inner mitochondrial membranes. The pathways involved in transport across the outer membrane have not been fully established, whereas one of two mitoferrins – mitoferrin 1 (Mfrn1, also known as SLC25A37) – is responsible for iron transport across the inner mitochondrial membrane (Paradkar et al. Mol Cell Biol 2009). Within the mitochondria, the protein frataxin is considered to act as an iron chaperone modulating the principal mitochondrial iron processing pathways of haem synthesis, Fe/S cluster formation and iron storage (Huang et al. Antioxidants & Redox Signalling 2011; Lane et al. Biochemica et Biophysics Acta 2015). Frataxin is an evolutionarily conserved mitochondrial protein that is encoded in the nucleus and targeted to the mitochondria. Importantly, a GAA-repeat expansion in intron 1 of the frataxin gene (FRDA) leads to a deficiency in frataxin and causes Friedreich’s ataxia (Campuzano et al. Hum Mol Genet 1997). Furthermore, studies using murine models and cells from patients with Friedreich’s ataxia indicate that frataxin deficiency in cardiac and skeletal muscle affects systemic as well as cellular iron homeostasis (Anzovino et al. Br J Pharmacol 2014).

Proteins involved in the uptake, transport and storage of iron are also themselves regulated by intracellular iron levels, both at the transcriptional and post-transcriptional level. Low oxygen pressure, low intracellular iron and ascorbate levels stimulate the transcription of genes regulated by hypoxia-inducible factor-1α (HIF-1α) and HIF-2α, the expression of HIF-1α being ubiquitous whereas HIF-2α displays a more restricted distribution. These transcription factors are post- transcriptionally regulated at the level of protein degradation, in an oxygen- and iron-dependent manner (Mole. Antioxidants & Redox Signalling 2010; Cassavaugh et al. JBC 2011). As oxygen sensors, HIF propyl-4 hydroxylases (propyl hydroxylase domain-containing enzymes – PHDs) use oxygen, in the presence of iron, as a substrate to hydroxylate proline residues in HIF-α subunits, thereby promoting HIF-α subunits binding to the von Hippel-Lindau (VHL) ubiquitin E3 ligase and its subsequent ubiquitination and degradation. Under hypoxic or iron-deplete conditions, PHD activity is reduced and results in the stabilization of HIF-α subunits and formation of heterodimers with the constitutively-expressed HIF1β subunit. The HIF heterodimers bind to hypoxia-response elements in the 5’ and 3’ untranslated regions (UTR) of genes encoding proteins such as erythropoietin, HIF-2α being the predominant transcription factor controlling the renal production of EPO. Other key iron- regulatory proteins are also HIF targets, including transferrin, TfR1, DMT1, Dcytb, ferroportin, and caeruloplasmin (Lawen & Lane. Antioxidants & Redox Signalling 2013; Rochette et al. Pharmacology & Therapeutics 2015). HIF-mediated gene regulation has an important role in the pathogenesis of pulmonary hypertension, both HIF-1α and HIF-2α deficient mice being protected from the development of hypoxia-induced pulmonary hypertension (Yu et al. JCI 1999; Brusselmans et al. JCI 2002; Ball et al. AJRCCM 2014). An activating mutation of HIF-2α (EPAS1) gene (Gale et al. Blood 2008; Tan et al. JBC 2013) and VhL mutation (Smith et al. PLOS Med 2006; Hickey et al. JCI 2010) induces pulmonary hypertension in mice and humans. Recently, PHD2 deficiency in endothelial and haemopoietic cells has also been found to induce obliterative vascular remodelling and severe pulmonary hypertension through HIF-2α activation (Dai et al. Circulation 2016).

The expression of several proteins is also regulated by iron at the post transcriptional level, via the intracellular binding of iron regulatory proteins (IRP1 and IRP2) to iron-responsive elements (IRE) in the 5’UTR or 3’UTR of mRNAs encoding proteins involved in the uptake (TfR1), storage (ferritin) and export of iron (ferroportin). IRPs are the principal regulators of cellular iron and bind to IREs with high affinity in iron-depleted cells, suppressing mRNA translation in the case of ferritin and ferroportin and enhancing the stability of the mRNA sequences encoding proteins such as TfR1 and DMT1 (Andersen et al., Biochimica et Biophysics Acta 2012). The IRE/IRP system, in particular IRP1, also modulates HIF-2α levels via a IRE in the 5’UTR of HIF-2α mRNA (Sanchez et al. Nat Struct Mol Biol. 2007). Although the two IRPs are structurally similar they are differentially regulated. IRP1 is a bifunctional protein that binds to IREs as an apoprotein in iron-deficient conditions, but under iron- replete conditions it acquires a Fe/S cluster and in iron-replete conditions is converted to an aconitase. IRP2 is an active IRE-binding protein in iron-deficient cells and otherwise rapidly degraded in iron-replete conditions (Andersen et al., Biochimica et Biophysics Acta 2012). While systemic and intracellular iron homeostasis are controlled by two distinct systems (i.e. hepcidin/ferroportin and IRE/IRP), they are coordinated and there is crosstalk between the cellular and systemic regulators (Hentze et al. Cell 2010). Mutations in the ferritin IREs are associated with disorders in systemic iron homeostasis (Beeaumont et al. Nat Genet 1995; Kato at al. Am J Hum Genet 2001) and deficiency in the IRPs has been implicated in the development of abnormal iron metabolism as well as pulmonary hypertension, myocardial remodelling and chronic obstructive pulmonary disease (COPD) (Gosh et al. Cell Metab 2013; Anderson et al. Cell Metab 2013; Wilkinson & Pantopoulos Blood, 2013; Cloonan et al. Nat Medicine 2016). In mice with targeted depletion of IRP1, polycythaemia and pulmonary hypertension developed that was exacerbated by a low-iron diet and dependent on the enhanced expression of HIF-2α (Gosh et al. Cell Metab 2013). Conversely, increased expression of the IRP1 (ACO1) and IRP2 (IREB2) genes has been reported in lung tissues obtained from PAH patients undergoing lung transplantation (Zhao et al. PLOSOne 2014).

1.15 Regulation of hepcidin

Hepcidin is a 25 amino acid peptide that was first identified in 2001 (Pigeon et al. JCB 2001; Nicolas et al. PNAS 2001; Park et al. JCN 2001). It is produced mainly by the liver, released in an iron- dependent manner and is the principal regulator of iron absorption, recycling, and tissue distribution in health and disease, as recently reviewed (Ganz Blood 2011, Ganz Physiol Rev 2013). BMP signalling has a key role in controlling the hepatic expression of hepcidin (Figure 1.2). BMPs typically act by binding to heteromeric combinations of type II receptors (BMPR2, activin A receptor type II –

ActR-IIa and ActR-IIb) and four type I receptors (ALK1, ALK2, ALK3, ALK6), leading to the intracellular phosphorylation and nuclear translocation of receptor of activated Smads (Smads1/5/8) and involvement of common (Smad4) and inhibitory Smads (Smads6/7) (Upton et al. Exp Physiol. 2013). A link between BMP/Smad signalling and hepcidin was discovered when liver-specific knockout of Smad4 resulted in mice with inappropriately low hepcidin and iron overload (Wang et al. Cell Metab 2005). BMP administration stimulates hepcidin transcription and reduces serum iron, whereas selective disruption of BMP expression, BMP antagonism or BMP type I receptor blockade all inhibit hepcidin expression and increase serum iron levels (Babitt et al. JCI 2007). Other membrane proteins modulate BMP signalling and are considered important in iron homeostasis. These include hemojuvelin (HJV), which is encoded by the Hfe2 (HJV) gene and is thought to influence hepcidin expression through the bone morphogenetic protein pathway via a neogenin-dependent mechanism (Zhang et al., JBC 2007). HJV is a member of the repulsive guidance molecule family and acts as a co- receptor and promoter of BMP signalling whereas neogenin modulates HJV secretion and BMP/HJV- induced stimulation of hepcidin expression (Babitt et al. Nat Genet 2006; Xia et al. Blood 2008; Zhang et al. JBC 2007, 2009 & 2010; Lee et al. Blood 2010). Additional factors have been identified that mediate inhibition of hepcidin expression, including the proteases matriptase-2 (encoded by the transmembrane protease, serine 6 - or TMPRSS6 gene) and furin. These enzymes cleave HJV from cell membranes, leading to the secretion of soluble HJV (s-HJV) that competitively inhibits BMP signalling and thereby suppresses hepcidin expression in hepatocytes (Figure 1.2). Matriptase-2, is required for iron sensing and loss of expression is accompanied by increased hepcidin expression and iron-deficiency in mice (Silvestri et al. Cell Metab 2008; Du et al. Science 2008) and may be refractory to oral iron therapy in humans (Finberg et al. Nat Genet 2008).

Several BMPs (including BMP2, BMP4, BMP5, BMP6, and BMP9) are expressed in the liver and induce hepcidin expression (Babitt et al. Nat Genet 2006; Truska et al. PNAS 2006; Babitt et al. JCI 2007; Xia et al. Blood 2008). Nevertheless, studies in mice suggest that BMP6 is the key BMP regulating hepcidin expression andiron homeostasis, with BMP6 expression in the liver (Kautz et al. Blood 2008; Meynard et al. Nat Genet 2009; Andriopoulos et al. Nat Genet 2009) and intestine (Arndt et al. Gastroenterology 2010) being modulated by iron. While BMP6 is considered essential for hepcidin expression, the contribution of other BMPs cannot be excluded as expression of hepcidin in HepG2 cells is reduced following siRNA inhibition of endogenous BMP2 and BMP4, and hepatic BMP2 expression also exhibits some sensitivity to iron loading (Kautz et al. Blood 2008).

Figure 1.2 Regulation of hepcidin expression. Iron status, BMP signalling and inflammation are major stimuli of hepcidin expression, whereas hypoxia and erythropoiesis both inhibit hepcidin production. HAMP, hepcidin gene; TfR1, transferrin receptor 1; TfR2, transferrin receptor 1; HFE, hemochromatosis protein; BMP, bone morphogenetic protein; (P), phosphorylation; HJV, hemojuvelin; s-HJV, soluble-hemojuvelin; IL-6, interleukin-6; ERK1/2, extracellular signal-regulated kinases 1/2; MAPK, mitogen-activated protein kinase; JAK-STAT, janus kinase/signal transducer and activator of transcription pathway; HIF, hypoxia-inducible factor; EPO, erythropoietin; GDF-15, growth differentiation factor 15 ; TWSG1, twisted gastrulation protein homolog 1. Courtesy Dr J Wharton

In addition to the well-established regulation of hepcidin by BMP signalling, TGF-β1 also activates hepcidin mRNA expression via TGF-β1 type I (ALK5) and type II (TβRII) receptor signalling and Smad1/5/8 phosphorylation (Chen et al. JBC 2016).

Infection and inflammation rapidly induce hepcidin synthesis, with cytokines such as IL-1 and IL-6 being potent inducers of hepcidin expression. IL-6 activates the Janus kinase/signal transducer and

35 activator of transcription (JAK/STAT) signalling pathway and this has been implicated in the development of anaemia associated with chronic inflammation (Nemeth et al. JCI 2004) and crosstalk found between BMP/Smad and IL-6/JAK/STAT signalling (Darshan et al. Maematologica 2010; Schmidt JCB 2015). The relationship between IL-6 and hepcidin in humans is however based mostly on acute challenges in healthy volunteers and subjects with evident inflammation. While inflammation is a well-recognized feature of PAH (Dorfmuller et al. Eur Resp J 2003; Sanchez et al. AJRCCM 2007), an association between IL-6 and hepcidin may not be apparent in mild proinflammatory or chronic disease states (Ashby et al. Kidney Int 2009; Ferrucci et al. Blood 2010). Compared with IL-6, BMPs are more potent stimulators of hepcidin expression (Truska et al. PNAS 2006), but IL-6 may also modulate BMPR2 expression at the post-transcriptional level (Brock et al. Circ Res 2009).

Whereas iron status and inflammation are major stimuli of hepcidin expression, erythropoiesis and hypoxia both inhibit hepcidin production (Figure 1.2). In situations such as haemorrhage, haemolysis and hypoxia, there is a demand for iron to produce more erythrocytes and this requires increased iron absorption from the gut and mobilization of iron from stores in macrophages and hepatocytes. Hypoxia, increased erythropoiesis, reduced iron stores and anaemia are all reported to suppress hepcidin expression. While HIF stabilization in liver cells can suppress hepcidin expression, it is doubtful if HIF-1α or HIF-2α directly modulates hepcidin transcription (Volke et al., PLOSOne 2009). Instead, HIFs are considered to act indirectly and reduce signals that normally increase hepcidin production. For example, HIF induces furin-mediated release of soluble HJV, which suppresses hepcidin production by antagonizing BMP6 signalling (Silvestri et al. Blood 2008). Matriptase-2 (TMPRSS6) is also reported to be HIF-regulated and thought to supress BMP6/HJV signalling in hypoxic conditions (Lakhal et al., JBC 2010; Maurer et al. Biol Chem 2012).

Hypoxia also induces synthesis of erythropoietin, a potent inhibitor of hepcidin production in the liver, but, as the blockade of erythropoietic activity prevents the inhibitory effect of erythropoietin, it is thought that factors released from erythroblasts in the bone marrow are responsible (Ashby et al., Haematologica 2010). Growth differentiation factor 15 (GDF-15) and twisted gastrulation protein homolog 1 (TWSG1) were originally proposed as erythroid regulators, but neither is necessary for the physiological suppression of hepcidin expression during increased erythropoiesis (Casanovas et al. Haematologia 2013; Kautz et al. Nat Genet 2014). Rather, a growing body of evidence indicates that erythroferrone (also known as myonectin or CTRP15) acts as an erythroid regulator (Kautz et al. Nat Genet 2014; Kautz et al. Blood 2014). After blood loss or hypoxia, erythropoiesis is stimulated

36 by erythropoietin (EPO), resulting in the release of erythroferrone from erythroblasts and subsequent down-regulation of hepcidin. Iron deficiency also regulates erythropoietic activity, both by moderating the production of EPO and erythrocytes and by inhibiting aconitase activity and erythroblast maturation (Wilkinson et al. Blood 2103; Ghosh et al. Cell Metab 2013; Richardson et al. JCI 2013). Transferrin receptor 2 is postulated to have a role in the sensing of systemic iron levels (Goswani & Andrews.JBC 2006) and has recently been implicated in the down regulation of erythroferrone and erythropoiesis (Nai et al. Haematologica 2014; Nai et al. Blood 2015; Wallace et al. Br J Haemtaol 2015).

1.16 Genetic basis of hepcidin and iron dysregulation

Genetic factors contribute to variation in markers of iron status (Whitfield et al. Am J Human Genet 2000; Benyamin et al. Am J Hum Genet 2009) and recent genome-wide association studies (GWAS) have identified gene variants associated with haemoglobin levels and blood cell traits indicative of altered erythropoiesis, as well as iron homeostasis. These include single nucleotide polymorphisms in the TMPRSS6 (matriptase-2) gene that couple with lower iron and haemoglobin concentrations and variation in RDW (Benyamin et al. Nat Genet 2009; Chambers et al., Nat Genet 2009; Tanaka et al. Blood 2010). Matriptase-2 activity is required for the suppression of BMP/Smad signalling and hepcidin expression (Finberg et al. Blood 2010) and deficiency in this protease could lead to inappropriately high hepcidin production as well as impaired iron homeostasis, lowered haemoglobin levels and raised RDW (Tanaka et al. Blood 2010).

Given the complex regulation of iron and critical role of hepcidin and it is not surprising that mutations in iron-regulatory genes have been associated with disorders in systemic and cellular iron regulation. Iron overload (haemochromatosis) results in iron accumulation and subsequent damage to tissues. Hereditary haemochromatosis is commonly associated with the c.845G > A (C282Y) mutation in the HFE gene (hemochromatosis protein) and may also arise due to mutations in at least four other genes – HFE2 (hemojuvelin - HJV), HAMP (hepcidin), TFR2 (transferrin receptor protein2) and SLC40A1 (ferroportin) (Beutler et al., Annu Rev Med 2006). Anaemia can also arise from dysregulation in hepcidin-ferroportin interaction. For example, mutations in the TMPRSS6 gene, encoding the hepcidin inhibitor matriptase-2, are associated with iron-refractory iron deficiency anaemia (Finberg et al., Nat Genetics 2008). Novel GWAS loci have been identified associated with iron homeoastasis in individuals at risk for hemochromatosis (Benyamin et al. Nat Commun 2014).

The importance of iron for erythropoiesis is also reflected in the overlap between genes or loci affecting iron homeostasis and erythrocytes (van der Harst et al. Nature 2012).

The manipulation of the hepcidin-ferroportin axis has become an attractive therapeutic target in iron-regulatory disorders. This includes the development of hepcidin agonists, which mimic the action of hepcidin or target hepcidin-regulatory molecules, and hepcidin antagonists that inhibit hepcidin production (Rochette et al. Pharmcaol & Therapeutics 2015).

1.17 Extra-hepatic expression of hepcidin and iron regulatory genes

In addition to its systemic role, the hepcidin-ferroportin axis may also have an autocrine/paracrine function. Although the liver (hepatocytes) is considered the main source of hepcidin and regulator of systemic iron, lower level hepcidin expression also occurs in extra-hepatic cells, including macrophages, adipocytes, cardiomyocytes and gastric cells (Rochette et al. Pharmacol & Therap. 2015). Several proteins involved in regulating iron metabolism are reported to be expressed in lung, including the iron transporters DMT1 and ferroportin-1, hepcidin (Nguyen et al. Am J Physiol 2006; Oliveira Filho et al. Vet Immunol Immunopathol. 2010), neogenin, hemojuvelin (Rodriguez et al. J Histochem Cytochem 2007), matriptase-2 (Ramsay et al. Front Biosci 2008), and BMP6 (Knittel etal. Exp Cell Res 1997). Local expression of hepcidin and its regulation of ferroportin-1 and iron transport may have a number of cellular effects, including modulating cell proliferation (Pinnix et al. Sci Transl Med 2010), oxidative stress in endothelial cells (Nanami et al., Arterioscler Thromb Vasc Biol 2005) and myocardial hypertrophy (Isoda et al., J Nutr Biochem 2010). Furthermore, the regulation of hepcidin expression has been found to differ in organs such as the heart and liver (Merle et al. Endocrinology 2007) and the effects of hepcidin on iron availability are also reported to be cell-specific (Chung et al. J Nutr 2009). In addition to modulating ferroportin-1 degradation, hepcidin may have other local effects such as activating the Jak2/STAT pathway and thereby transcriptional activation of a number of genes (De Domenico et al. JCI 2010).

1.18 The integrated stress response and iron

The integrated stress response (ISR) and phosphorylation of the alpha-subunit of eukaryotic translation initiation factor 2 (eIF2α) is a fundamental regulatory mechanism that controls global rates of protein synthesis in all eukaryotic cells and is considered to be important in a variety of

38 disorders including pulmonary diseases (van ‘t Wout et al. AJRCMB 2014). Under non-stressed conditions, eIF2α is largely un-phosphorylated and binds with GTP to recruit Met-tRNA, forming a complex required for ribosomal subunit recognition and the initiation of protein translation. Adaptation to various cellular stresses involves phosphorylation of eIF2α, temporarily inhibiting general protein synthesis and selectively activating the translation of stress-inducible transcripts, particularly the expression of transcription factor 4 (ATF4). Downstream targets of ATF4 include amino acid transporter and synthetases, the transcription factor CHOP and the eIF2α phosphatase GADD34 (growth arrest and DNA-damage 34). Four mammalian kinases are known to phosphorylate eIF2α in response to distinct stresses, these comprising (1) Haem-regulated inhibitor (HRI; EIF2AK1); (2) Protein kinase RNA-activated (PKR; EIF2AK2); (3) PKR-like endoplasmic reticulum kinase (PERK; EIF2AK3); and (4) General control non-depressible-2 kinase (GCN2; EIF2AK4) (Donnelly et al. Cell Mol Life Sci 2013). Mammals also possess two genes encoding GADD34 (PPP1R15A) and CReP (PPP1R15B) that form complexes with protein phosphatase 1 (PP1) and de-phosphorylate eIF2α (Choy et al. Cell Reports 2016). Under basal conditions, constitutively expressed CReP-PP1 controls the de-phosphorylation of eIF2α whereas GADD34-PP1 expression is induced in stressed or diseased cells and acts as a feedback loop to restore general protein synthesis and promote cell recovery (Choy et al. Cell Reports 2016). Interestingly, reduced BMPR2 signalling has been associated with increased GADD34-PP1 phosphates activity and impaired stress response, which may lead to enhanced cytokine expression and lung inflammation in PAH (Swada et al. JEM 2014). Rare forms of PAH – pulmonary veno-occlusive disease and capillary haemangiomatosis – are also caused by GCN2 (EIF2AK4) mutations (Eyries et al. Nat Genet 2014; Best et al. Chest 2014; Tenorio et al. Clin genetics 2015) and associated with BMPR2 mutations (Runo et al. AJRCCM 2003; Montani et al. Medicine 2008).

Cellular haem levels are closely regulated, balancing haem synthesis and catabolism, and haem- regulated eIF2α kinase (HRI; EIF2AK1) is considered to have an important role in the production of haemoglobin and formation of red blood cells (Liu et al. JCI 2007). It is activated by haem deficiency and coordinates globin mRNA translation with available iron, preventing the accumulation of misfolded globin chains in the absence of haem. HRI is necessary to promote erythroid differentiation during iron deficiency and has an essential protective role in anaemia’s of iron deficiency (Han et al., EMBO J 2001; Han et al. JCI 2005; Liu et al. Br J Haematology 2008; Suragani et al. Blood 2012). In addition to an abnormal erythroid response to iron and haem deficiency, HRI knockout mice also exhibit alterations in iron homeostasis, reduction in hepcidin expression, lower circulating BMP2 levels, and an attenuated inflammatory response; leading to the suggestion that

HRI has a potential role in systemic and cellular iron homeostasis and the anaemia of inflammation (Liu et JCI 2007; Liu et al. Haematologica 2008). In addition to iron and haem deficiency, HRI is also activated by arsenate-induced oxidative stress, osmotic shock and heat shock, whereas other eIF2α kinases – PERK and GCN2 – respond to ER stress and nutrient starvation respectively (Lu et al. Mol Cell Biol 2001).

1.19 The determination of iron status Established laboratory assays are used to determine iron status and distinguish iron deficiency and related conditions, including functional iron deficiency, iron-deficiency anaemia, iron-refractory iron- deficiency anaemia, and anaemia of chronic diseases (Camaschella NEJM 2015). The serum ferritin level is generally considered the most sensitive and specific test used for the identification of iron deficiency (defined by ferritin <30 μg/L). However, the true iron status can be difficult to determine in diseases such as IPAH, using standard laboratory assays, due to the presence of inflammation. This induces ferritin while repressing serum iron and transferrin saturation (Weiss & Goodnough NEJM 2005). Higher cut-offs for ferritin levels (e.g. <100 µg/L or 100-300 µg/L with transferrin saturation <20%) have therefore been used to in the diagnosis of iron deficiency associated with conditions such as heart failure (Anker et al. NEJM 2009), but universally agreed criteria are lacking (Zimmermann & Hurrell Lancet 2007). Circulating sTfR levels are however largely unaffected by inflammation (Cook et al. Br J Haematol 1990). Using the upper limit of normal (28.1 nmol/L) as a cut-off, raised sTfR levels have been used to demonstrate a high prevalence of iron deficiency in patients with IPAH (Rhodes et al. JACC 2011).

1.20 Iron deficiency in cardiovascular disease

Iron deficiency is very common, with an incidence of approximately two billion cases worldwide (Andrews Nat Rev Genet 2000), and it has adverse effects in disorders such as coronary artery disease and heart failure (von Haehling et al. Nat Rev Cardiology 2015). Although it was originally hypothesised that higher levels of stored iron was associated with an increased incidence of heart disease (Sullivan Lancet 1981), subsequent meta-analyses found no significant association between circulating iron, ferritin or total iron-binding capacity and coronary artery disease or myocardial infarction (Das et al. Atherosclerosis 2015). Instead, there was a negative correlation between transferrin saturation and coronary artery disease and myocardial infarction, leading to the conclusion that high body iron stores could protect against the development of coronary artery disease (Das et al. Atherosclerosis 2015). Iron deficiency occurs frequently in patients with coronary

40 artery disease and independently increases the likelihood of mortality (Grammer et al. Atherosclerosis 2014; Ponikowska et al. Diabetes Care 2013).

Iron deficiency is also prevalent in patients with heart failure, occurring in 30-50% of the patients seen with the condition (Jankowska et al. Eur Heart J 2010, 2014; Okonko et al., JACC 2011; Klip et al. Am Heart J 2013). A series of uncontrolled open-label (Bolger et al. JACC 2006; Usmanov et al. J Nephrol 2008) and randomized, blinded and placebo-controlled studies (Toblli et al. JACC 2007; Okonko et al. JACC 2008; Anker et al. NEJM 2009; Ponikowski et al. Eur Heart J 2015) have sought to determine the therapeutic potential of intravenous iron replacement therapy in patients with heart failure. A seminal multi-centre study was the Ferric Carboxymaltose in Patients with Heart Failure and Iron Deficiency (FAIR-HF) trial conducted by Anker and colleagues (Anker et al. NEJM 2009). They recruited 459 symptomatic patients in NYHA class II or III who were iron deficient, as defined by – ferritin <100 µg/L or 100-300 µg/L with transferrin saturation <20% and haemoglobin levels between 9.5-13.5 g/dL. Patients were randomised at a 2:1 ratio of i.v. ferric carboxymaltose to placebo and received 200 mg ferric carboxymaltose until the iron deficiency was corrected. After 24 weeks follow up, 50% of the participants receiving iron experienced improvement in the self- reported Patient Global Assessment (primary end point of the study), compared to 28% in the placebo group. Other parameters (NYHA class, 6MWD, physical well-being, quality of life) also significantly improved, with differences being seen as early as 4 weeks after treatment started, and improvements occurred irrespective of whether patients were anaemic (Anker et al. NEJM 2009). These results were reinforced in the double-blind, multicentre, prospective, randomized CONFIRM- HF study, with improvements maintained to 52 weeks after treatment started (Ponikowski et al. Eur Heart J 2015).

1.21 Intravenous iron preparations and safety

Various i.v. iron preparations have been produced, combining iron with a carbohydrate in order to reduce the undesirable biological reactivity of free (non-transferrin-bound) iron. The first used high molecular-weight dextran, but was withdrawn following allergic/anaphylactic reactions and superseded by a low molecular-weight iron dextran (Cosmofer). Subsequent preparations used other carbohydrate complexes – ferric gluconate (Ferrlecit) and iron glucose (Venofer). However, these preparations are limited by the stability of the iron-carbohydrate complex and cannot be administered in large doses. The third generation of iron preparations enables much higher single doses to be administered over short intervals and comprise – ferric carboxymaltose (Ferinject), iron

41 isomaltose-1000 (Monofer) and iron polyglucose sorbitol carboxymethylether (Ferumoxytol/Rienso), although the use of Ferumoxytol is limited to patients with chronic kidney disease (Table 1.2).

Name Cosmofer Venofer Ferumoxytol Monofer Ferinject

Manufacturer Pharmacosmos Vifor AMAG Pharmacosmos Vifor Pharmaceuticals Carbohydrate LMW Dextran Sucrose Polyglucose sorbitol Isomaltoside Carboxymaltose carboxymethylether Molecular 165 34-60 750 150 150 weight (KDa) Maximum 100 200 510 20 mg/kg 1,000 if patient approved weighs >66 kg single dose (mg)

Test dose Yes No No No No required

Table 1.2. Intravenous iron preparations.

There are safety concerns regarding the use of i.v. iron preparations, mainly reflecting potential anaphylactic reactions, endothelial damage by free iron and the promotion of infection by supplying iron to pathogenic bacteria. Recent reviews and meta-analyses indicate however that serious adverse events associated with ferric carboxymaltose and iron isomaltose-1000 infusion are rare and i.v. iron therapy is not accompanied by increased risk of serious adverse events or infections (Chertow & Winkelmayer Am J Hematol 2010; Moore et al. BMC Blood Disord 2011Begman & Goodnough Therap Advances in Hematol 2014; Avni et al. Mayo Clin Proc 2015). In addition to appearing safe and effective, i.v. ferric carboxymaltose has also been found to provide a more efficient and rapid correction of haemoglobin and serum ferritin levels, compared with other iron preparations, in iron-deficient patients (Rognoni et al. Clin Drug Invest 2016). The results of in vitro studies have led to the suggestion that increases in non-transferrin-bound iron following iron treatment may be sufficient to induce transient changes in endothelial cell function (Mollet et al. PLOSOne 2016), but it is uncertain if similar effects occur in vivo and have adverse consequences.

The effects of i.v. iron treatments on phosphate homeostasis may vary however; in contrast to iron dextran, sucrose and isomaltoside, ferric carboxymaltose has more frequently been associated with significant hypophosphatemia; phosphate levels falling from normal (0.8-1.45 mmol/L) to moderate (0.32-0.8 mmol/L) or severe (<0.32 mmol/L) hypophosphatemia (Wolf et al. J Bone & Mineral Res 2013; Hardy & Vandemergel Int J Rheumatol 2015; Kalra & Bhandari Int J Nephrol & Renovasc Dis. 2016).

1.22 Iron deficiency and phosphate homeostasis

Phosphate homeostasis is maintained by the combined effects of three endocrine factors – bone- derived fibroblast growth factor 23 (FGF23), parathyroid hormone (PTH) and 1,25 dihydroxyvitamin D3 (calcitrol) – that regulate trans-epithelial phosphate transport via a family of sodium-dependent phosphate co-transporters – NaPi-2a (SLC34a1), NaPi-2b (SLC34a2) and NaPi-2c (SLC34a3) – expressed in the intestine and kidney (Razzaque Nature Reviews Endocrinology 2009). FGF23 is synthesised by osteocytes and osteoblasts that exerts phosphate lowering effects in the kidney, acting via an FGF receptor and essential cofactor (membrane-bound protein Klotho) to suppress renal expression of NaPi-2a and NaPi-2c and inhibit phosphate reabsorption. In addition, FGF23 inhibits 1α-hydroxylase activity, thereby supressing the production of 1,25 dihydroxyvitamin D3, reducing intestinal NaPi-2b activity and absorption of phoshate (Razzaque Nature Reviews Endocrinology 2009). PTH can also induce the expression of 1α-hydroxylase activity and increase 1,25 dihydroxyvitamin D3 production, which in turn can inhibit PTH and 1α-hydroxylase expression (Figure 1.3).

Figure 1.3. Regulation of phosphate homeostasis. Modified from Razzaque Nature Reviews Endocrinology 2009; Brown and Razzaque, BoneKEy 2015

In addition to PTH and 1,25 dihydroxyvitamin D3, a number of factors have been implicated in the skeletal expression of FGF23, including iron. Iron deficiency stimulates FGF23 production, but this is normally coupled with the increased cleavage of the hormone, thereby maintaining normal circulating levels of active FGF23 and normal phosphate levels (Wolf et al. J Bone & Mineral Res 2013; Wolf & White Curr Opinion Nephrol Hypertens 2014). However, when iron deficient women were randomly treated with either i.v. ferric carboxymaltose or iron dextran the effects on circulating FGF23 levels were found to differ. Both preparations inhibited FGF23 cleavage, reflected by a fall in circulating levels of the cleaved C-terminal fragment of the hormone, but ferric carboxymaltose treatment was selectively accompanied by an increase in the level of the active hormone and hypophosphatemia (Wolf et al. J Bone & Mineral Res 2013). The hypophosphatemic effect of i.v. iron is not considered to be a class effect, as it is more frequently associated with the use of ferric carboxymaltose than carbohydrate groups such as iron dextran, sucrose or isomaltoside

(Wolf et al. J Bone & Mineral Res 2013; Hardy & Vandemergel Int J Rheumatol 2015; Kalra & Bhandari Int J Nephrol Tenovasc Dis 2016). The mechanism by which ferric carboxymaltose selectively stimulates the production and/or inhibits the degradation of FGF23 is uncertain and, apart from transient and largely asymptomatic reductions in circulating phosphate levels, ferric carboxymaltose has exhibited a similar safety profile to other i.v. iron preparations in the treatment of iron deficiency anaemia (Wolf et al. J Bone & Mineral Res 2013). Nevertheless, the potential longer-term consequences have not been investigated and acute changes in phosphate levels need monitoring as a number of cases reports have linked hypophosphatemia, especially when severe, with complications such as fatigue, seizures, cardiac dysfunction, demyelination, myopathy and haemolytic anaemia (Hardy & Vandemergel Int J Rheumatol 2015). Interestingly, red cell distribution width (RDW) has also been reported to correlate with circulating levels of the cleaved C-terminal fragment of FGF23, but not the intact active hormone, in patients with chronic kidney disease and heart failure (van Breda et al. PLOSOne 2015).

1.23 Influence of iron on pulmonary vascular function and pulmonary hypertension Iron status is an important factor determining pulmonary vascular tone the phenomenon of hypoxic pulmonary vasoconstriction (Frise & Robbins J Appl Physiol 2015). A number of human studies have demonstrated the importance of iron availability in influencing pulmonary artery pressure and pulmonary vascular response to hypoxia. Infusion of the iron chelator desferrioxamine (DFO) increases serum erythropoietin levels (Ren et al. J Appl Physiol 2000) and mimics hypoxic pulmonary vasoconstriction (Balanos et al. J Appl Physiol 2002). Hypersensitivity to hypoxia, produced by pre- exposing subjects to 8 hours of hypoxia, is inhibited by pre-infusing iron, whereas iron chelation with DFO increases the pulmonary vascular response to an acute hypoxic challenge (Smith et al. J Physiol 2008). Iron availability also affects the pulmonary hypertensive response to hypoxia at high altitude. In high altitude residents with chronic mountain sickness, iron depletion induced by repeated venesection led to a gradual increase in pulmonary artery systolic pressure, although this was not significantly reversed by i.v. iron infusion (Smith et al. JAMA 2009). Recent studies on iron replete and iron deficient volunteers have also demonstrated that that iron deficiency disturbs normal responses to hypoxia, iron deficient subjects exhibiting exaggerated hypoxic pulmonary hypertension that is reversed by the subsequent administration of i.v. ferric carboxymaltose (Frise et al. JCI 2016).

Others have linked myeloid abnormalities with PAH in humans and proposed that iron deficiency could contribute to pulmonary vascular remodelling via the induction of HIF-responsive growth factors (erythropoietin, stem cell factor and hepatocyte) and increased recruitment of circulating CD34+/CD133+progenitor cells to the pulmonary vasculature (Fahra et al. Blood 2011). Iron chelation with DFO is also reported to stimulate the angiogenic activity of endothelial cells, promoting revascularization via an Akt-eNOS-dependent mechanism (Ikeda et al. Atherosclerosis 2011). Together these data indicate that iron may have longer term effects on the pulmonary vascular bed, effecting structure as well as function.

As discussed above (section 1.5), PAH is increasingly being considered as a metabolic disease that involves mitochondrial and metabolic dysfunction. Many of the enzymes and complexes involved are also iron-dependent, suggesting that iron-deficiency could compromise mitochondrial and metabolic function (Levi & Rovida Biochim Biophys Acta 2009). For example, deficiency in Fe/S clusters due to pathogenic gene has been associated with pulmonary hypertension (Navarro-Sastre et al. Am J Human Genetics 2011; Ahting et al. Frontiers in Genetics 2015; White et al., EMBO Mol Medicine 2015) and reduced Fe/S cluster integrity found in lung tissues from pulmonary hypertensive mice and patients (White et al., EMBO Mol Medicine 2015). Importantly, iron deficiency in rats alters mitochondrial function and results in pulmonary hypertension that is accompanied by vascular remodelling, proliferation and apoptosis resistance (Cotroneo et al. Cir Res 2015). Several of the metabolic features of PAH were found in the iron deficient rat lung, including the metabolic switch from glucose oxidation to glycolysis, increased glucose uptake, HIF activation and reduced mitochondrial complex I activity. Iron supplementation also resulted in the partial reversal of the pulmonary hypertension and vascular remodelling (Cotroneo et al. Cir Res 2015). Nevertheless, others have suggested that iron deficiency may protect against experimental pulmonary hypertension, treatment with DFO or dietary iron restriction having been found to attenuate the development of pulmonary hypertension in rats exposed to chronic hypoxia or monocrotaline (Wong et al. Free Radic Biol Med 2012; Naito et al. BBRC 2013).

1.24 Iron deficiency in PAH

Several independent studies have reported that iron deficiency is common in IPAH, affecting 31-63% of patients, and is associated with reduced exercise capacity and increased mortality (Rhodes et al. JACC 2011; Soon et al. Thorax 2011; Ruiter et al. Eur Resp J 2011; van Empel Heart Lung Circ 2014). Due to the presence of inflammation, iron status can be difficult to ascertain in chronic diseases such as IPAH using standard laboratory techniques. Inflammation increases ferritin levels, while

46 repressing serum iron and transferrin saturations (Weiss & Goodnough NEJM 2005), whereas soluble transferrin receptor (sTfR) levels are largely unaffected (Cook et al. Br J Haematol 1990). Rhodes and colleagues studied 98 patients with IPAH, 63% having iron deficiency (as defined by sTfR levels above the normal range of 8.7-28.1 nmol/L) without overt anaemia (Rhodes et al. JACC 2011). Consistent with iron deficiency (raised sTfR levels), iron, ferritin and transferrin saturation levels were reduced (Figure 1.4) and red cell distribution width was increased. However, it should be kept in mind that a number of other factors may affect RDW as well as iron deficiency (see section 1.13).

Figure 1.4 . Characterisation of iron status in patients with IPAH. Dot-plots of soluble transferrin receptor, iron, transferrin saturation, ferritin and haemoglobin levels were all measured in plasma samples by standardised clinical pathology accredited assays. Grey areas delineated by dashed lines indicate normal ranges in healthy. From Rhodes et al. JACC 2011

Importantly, circulating levels of sTfR increased with WHO function class (p<0.05) and correlated negatively with exercise capacity (6MWD; p=0.027) and values >28.1 nmol/L independently predicted worse survival (p=0.022) after adjusting for age, WHO class and 6MWD (Figure 1.5).

Figure 1.5. Soluble transferrin receptor (sTfR) levels in patients with IPAH. (A) Distribution of sTfR levels according to WHO class in patients with IPAH and healthy controls. P value relates to Kruskal-Wallis ANOVA. (B) Kaplan-Meier survival estimates stratified by sTfR level below (solid line) or above (dashed line) the pre-defined upper limit of the normal range, 28.1 nmol/L. (Rhodes, Howard et al. 2011)

The same study also revealed abnormal hepcidin levels, the plasma concentration of hepcidin being significantly higher in IPAH patients, compared with healthy controls, and inappropriately raised in iron-deficient patients (Rhodes, Howard et al. 2011). No correlation was found with inflammatory markers such as IL-6 and CRP, suggesting that the raised hepcidin levels were not due to inflammation.

Using a definition of iron deficiency based on serum iron <10 µmol/L and transferrin saturation <15% in women and <20% in men, Ruiter et al. found that 30/70 (43%) of IPAH patients were iron deficient and had a lower exercise capacity relative to iron replete patients (Ruiter, Lankhorst et al. 2011). In a subset of 18 patients given oral iron, ferritin levels were significantly increased. However, 8 patients only slightly increased their iron storage, suggesting impaired iron absorption. In a third study, the prevalence of iron deficiency study was examined in 85 patients with idiopathic or heritable PAH and 130 with CTEPH, iron deficiency being defined as decreased ferritin levels (<10 μg/L) with elevated or normal transferrin levels and a normal CRP level (Soon et al. Thorax 2011). Unexplained iron

48 deficiency was observed in 50% of premenopausal women with IPAH compared with 8% among CTEPH; 28% of men with IPAH compared with 2% among CTEPH; and in 60% of patients with heritable PAH and BMPR2 mutations. IPAH patients also exhibited lower serum iron and transferrin saturations than those with CTEPH and IL-6 levels correlated with iron levels and transferrin saturations in IPAH but not in CTEPH (Soon et al. Thorax 2011). In another study of IPAH, iron deficiency was defined by raised sTfR levels (>4.4 mg/L in females and >5.0 mg/L in males) and found in 45% (13/39) of patients (van Empel Heart Lung Circ 2014). Iron deficient patients also tended to have a more severe phenotype, with worse functional class, higher mPAP, reduced cardiac index and raised NT-proBNP levels (van Empel Heart Lung Circ 2014).

Prevalence of iron deficiency in other types of pulmonary hypertension is less certain. In scleroderma-related PAH, the prevalence of iron deficiency (46.1%) has been found to be more frequent compared with 16.4% in systemic sclerosis patients without pulmonary hypertension (Ruiter et al. Rheumatology 2014). Several mechanisms are important in the development of iron deficiency, including low iron intake or uptake from the gut and increased iron loss or high iron use due to increased erythropoiesis. In patients with systemic sclerosis, gastrointestinal involvement is common and it is plausible that there is lower iron uptake and more iron loss in the gastrointestinal tract of patients with scleroderma-related PAH, particularly when combined with the use of anticoagulants (Johnson et al. Int J Rheumatol 2011).

More than 30% of patients with cyanotic congenital heart disease are also reported to be iron deficient (Kaemmerer et al. Am J Cardiol 2004; Diller et al. Eur Heart J 2006) and the safety and efficacy of iron replacement therapy has been investigated (Tay et al. Int J Cardiol. 2011). Patients (n=25) with iron deficiency, defined as serum ferritin <30 µg/L or serum ferritin <50 µg/L with transferrin saturation <15%, were studied at baseline and 3 months after a course of oral ferrous fumarate (n=23) or single i.v. dose of 200mg iron sucrose for patients (n=2) unable to tolerate oral iron therapy. There were no adverse events requiring termination of treatment and significant improvements were found in iron status, 6MWD and quality of life score, although peak VO2 was unchanged (Tay et al. Int J Cardiol. 2011). Nevertheless, the traditional criteria for defining iron deficiency do not apply in these patients and there is a safety issue they may have high baseline haemoglobin levels and a dramatic increase in could cause marked erythrocytosis and hyperviscosity (Tay et al. Int J Cardiol. 2011).

Combined with the influence of iron in regulating pulmonary vascular tone pulmonary and the success of i.v. iron replacement therapy in heart failure, the results of the studies conducted to date suggests that iron is an attractive therapeutic target in IPAH, which might lead to improved pulmonary haemodynamics, symptoms and exercise capacity.

Until recently years many IPAH patients receive warfarin and are female, many of whom are premenopausal, and this might implicate blood loss. However, this is not a complete explanation. sTfR levels were also raised in patients not taking warfarin. Women above 60 and men with IPAH also exhibited raised sTfR levels.

This is consistent with data of Ruiter et al. (Ruiter et al. 2010) who observed an equal prevalence of iron deficiency in pre- and post-menopausal women with IPAH and no reported abnormal gastrointestinal or gynaecological blood loss. Soon et al also showed that CTEPH patients, who all receive warfarin, do not demonstrate a high prevalence of iron deficiency (Soon et al. 2011), indicating that warfarin therapy itself does not lead to significant iron deficiency. Furthermore, the distribution of plasma hepcidin levels in the IPAH patient population is not consistent with blood loss as a single cause; iron deficiency from chronic blood loss would normally be associated with reduced circulating hepcidin (Zimmermann et al. 2009).

To date, two open-label proof of concept studies have examined the effect of i.v. iron (ferric carboxymaltose) in patients with PAH (Viethen et al. Int J Cardiol 2014; Ruiter et al. Pul Circ 2015). Viethem and co-workers enrolled 20 PAH patients (13 IPAH) with iron deficiency, defined by serum iron <10 µmol/L, ferritin <150 µg/L and transferrin saturation <15%, in the absence of significant inflammation (CRP <25 mg/L) (Viethen et al. Int J Cardiol 2014). Exercise capacity (6MWD) and quality of life was assessed in patients at baseline and 8 weeks after receiving up to 1000mg ferric carboxymaltose, the results being compared with 20 control PAH subjects (13 IPAH) who were iron replete and did not receive ferric carboxymaltose. Iron supplementation was well tolerated and resulted in an improved iron status at 8 weeks, this being associated with a significant increase in 6MWD (from 346.5±28.3 to 374.0±25.5 m; P=0.007) and quality-of-life score compared with no significant change in the control group (Viethen et al. Int J Cardiol 2014). Ruiter et al. enrolled 15 IPAH patients with iron deficiency, defined as serum iron <10 µmol/L, ferritin <100 µg/L and transferrin saturation <15% in women and <20% in men (Ruiter et al. Pul Circ 2015). Patients were examined before and 12 weeks after receiving 1000mg ferric carboxymaltose and

50 underwent a 6MWT, CPET, CMR imaging, muscle biopsy and quality-of-life test. The primary endpoint (6MWD) was unchanged, but i.v. iron administration was well tolerated and accompanied by an increase in exercise endurance time and aerobic capacity. This was also associated with improved oxygen handling in quadriceps muscle cells and a better quality of life (Ruiter et al. Pul Circ 2015). This raises the question as to whether iron supplementation in IPAH can improve outcome, either through an improvement in cardio-pulmonary function and/or improved exercise capacity.

1.25 Obesity and pulmonary hypertension Obesity is on the increase and may often be a comorbid feature in pulmonary hypertension. It has major contributory effects on cardio-pulmonary disease. Body mass index (BMI) correlates with left ventricular mass and wall thickness but obese patients have also been shown to develop left ventricular dysfunction which is independent of blood pressure, age, gender and left ventricular mass (Watson, Fox. 2010).

We have noted that many patients with significant obesity were presenting with clinical features of pulmonary arterial hypertension and features on echocardiography but did not suggest significant left ventricular dysfunction, namely normal left atrial size and diastolic filling patterns, yet were being classified as post-capillary PH, due to elevated PWP and/or left ventricular end-diastolic pressure (LVEDP).

2. Hypothesis

 Iron deficiency is prevalent in forms of pulmonary hypertension other than idiopathic and heritable PAH  Iron deficiency is an important factor in determining function and prognosis in pulmonary hypertension  Iron replacement is safe and efficacious in improving function in patients with idiopathic/heritable/anorexigen PAH  Obesity leads to misinterpretation of pulmonary vascular haemodyncamics

3. Aims

1. Establish the extent of iron deficiency in patients with incident pulmonary hypertension,

including those with IPAH/HPAH, PAH associated with connective tissue or congenital heart

disease, and CTEPH.

2. Determine the association between iron deficiency and measures of cardiac function,

cardiopulmonary haemodynamics and exercise capacity in different categories of pulmonary

hypertension.

3. Investigate the safety and impact of iron supplementation on cardiopulmonary haemodynamics

and exercise capacity in patients with IPAH.

4. Investigate novel interpretations of invasive haemodynamics in obese patients with pulmonary hypertension

2. Alterations in iron status in pulmonary arterial hypertension and chronic thromboembolic pulmonary hypertension and impact on function and survival.

2.1 Introduction ...... 55 2.1.1 Red cell distribution width (RDW), iron status and mortality ...... 55 2.1.2 Iron deficiency and RDW in idiopathic pulmonary arterial hypertension (IPAH) ...... 56 2.1.2.1 Defining iron deficiency in IPAH ...... 57 2.1.3 Iron deficiency and RDW in pulmonary arterial hypertension secondary to connective tissue disease (PAH-CTD)...... 58 2.1.4 Iron deficiency and RDW in pulmonary arterial hypertension secondary to congenital heart disease (PAH-CHD) ...... 59 2.1.5 Iron deficiency and RDW in chronic thromboembolic pulmonary hypertension (CTEPH) ...... 59 2.1.6 Hypothesis and aims ...... 60 2.2 Methods ...... 61 2.2.1 Subjects, samples and clinical investigations ...... 61 2.2.2 Data presentation and statistical analysis ...... 62 2.3 Results ...... 64 2.3.1 Patient demographics ...... 64 2.3.2 Prevalence of iron deficiency defined by sTfR in IPAH ...... 67 2.3.3 sTfR as a predictor of mortality...... 69 2.3.4 Prediction of iron deficiency from standard clinical laboratory measurements in IPAH ...... 70 2.3.3.1 What is the best single, independent marker in predicting iron deficiency? ...... 75 2.3.3.2 Red cell distribution width and inflammation ...... 78 2.3.4 Iron status in pulmonary hypertension subtypes at diagnosis ...... 82 2.3.5 Prevalence of iron deficiency in pulmonary hypertension subtypes ...... 86 2.3.6 Impact of red cell distribution width on cardiopulmonary physiology and outcomes ...... 87 2.3.6.1 Idiopathic pulmonary arterial hypertension ...... 87 2.3.6.1.1 Cardiopulmonary haemodynamics and exercise physiology ...... 87 2.3.6.1.2 Survival ...... 93 2.3.6.2 Connective tissue disease PAH ...... 101 2.3.6.2.1 Cardiopulmonary haemodynamics and exercise physiology ...... 101 2.3.6.2.2 Survival ...... 104 2.3.6.3 Congenital heart disease PAH ...... 110 2.3.6.3.1 Cardiopulmonary haemodynamics and exercise physiology ...... 110 2.3.6.3.2 Survival ...... 113 2.3.6.4 Chronic thromboembolic pulmonary hypertension ...... 117 2.3.6.4.1 Cardiopulmonary haemodynamics and exercise physiology ...... 117 2.3.6.4.2 Survival ...... 121 2.4 Discussion ...... 129

2.1 Introduction

2.1.1 Red cell distribution width (RDW), iron status and mortality Increased RDW has been identified as an early indicator of iron deficiency (McClure et al. JAMA 1985) and it associates robustly with iron deficiency in patients with heart failure, correlating with low serum iron and ferritin levels, reduced transferrin saturation and increased soluble transferrin [sTfR ]evels (Forchecz et al. Am Heart J 2009; Allen et al., J Card Fail 2010). Iron deficiency is associated with increased risk of death in the general population (Corti et al. Am J Cardiol 1997) and independently increases the likelihood of mortality in cardiovascular conditions such as coronary artery disease (Grammer et al. Atherosclerosis 2014;, and in diabetic patients sTfR strongly predicts 5-year all-cause mortality rates (Ponikowska et al. Diabetes Care 2013). Increasing RDW has been reported to accompany the development of iron deficiency and amplify the risk of mortality in patients with chronic heart failure (Aung et al. Int J Cardiol 2013). RDW is also a powerful predictor of all-cause mortality in the general population (Perlstein et al. Arch Intern Med 2009; Patel et al. Arch Int Med 2009) as well as in a variety of cardiovascular diseases and other disorders including cardiovascular disease, venous thromboembolism, cancer, diabetes, community-acquired pneumonia, chronic obstructive pulmonary disease, liver and kidney failure (Salvagno et al. Crit Rev Clin Lab Sci 2015).

In addition to iron deficiency, the relationship between RDW and prognosis could reflect the effects of other common pathological processes such as oxidative stress and inflammation (Savagno et al. Crit Rev Clin Lab Sci 2015) as well as abnormalities in erythropoiesis, renal function and vitamin D deficiency (Felker et al. JACC 2007; Fӧrhécz et al. Am Heart J 2009, Al-Najjar et al. Eur J Heart Failure 2009; Allen et al. J Card Fail 2010). Studies have demonstrated a positive association between RDW and inflammatory markers, including high-sensitivity C-reactive protein and erythrocyte sedimentation rate (Lippi et al. Arch Pathol Lab Med 2009; Vaya´ et al. Clin Hemorheology & Microcirculation 2015), soluble tumour necrosis factor receptors I and II, interleukin (IL)-6 and fibrinogen levels (Förhécz et al. Am Heart J 2009; Emans et al. J Cardiac Fail 2011). As well as their effects on hepcidin expression and iron metabolism, inflammatory cytokines may contribute to increased RDW by inhibiting erythropoietin and erythropoietin receptor expression, reducing the proliferation of erythroid progenitor cells and the survival of red blood cells (Douglas & Adamson Blood 1975; Weiss & Goodnough NEJM 2005). Nevertheless, RDW has been shown to be an independent prognostic indicator when compared with inflammation in cohorts free of cardiovascular disease, cancer and respiratory disease (Perlstein et al. Arch Intern Med 2009; Veeranna et al. Int J Cardiol 2013).

Oxidative stress could influence anisocytosis by reducing the survival and increasing production of red blood cells (Friedman et al. Blood 2004). Furthermore, red blood cells have a role as scavengers of reactive oxygen and nitrogen species and oxidative damage may compromise both this scavenging function and also its role in tissue oxygen delivery (Kiefer & Snyder Curr Opin Hematol 2000; Minetti et al. Cardiovasc Res 2007). RDW is also raised in conditions of dysfunctional erythropoiesis, after blood infusions and situations of accelerated erythropoiesis such as haemolysis and early after intravenous iron therapy (Akarsu et al. ACta Haematologica 2006; Aung et al. Int J Cardiol 2013; Van Craenenbroeck et al. Eur Heart J 2013).

Overall, whilst a growing number of studies indicate that increased RDW robustly predicts mortality in a variety of clinical disorders, it is uncertain if it represents a biological mechanism that contributes directly to the diseased state. Instead, it has been postulated that RDW reflects a combination of pathological processes that directly influence clinical outcomes (Forchecz et al. Am Heart J 2009; Felker et al. JACC 2007; Aung et al. Int J Cardiol 2013).

2.1.2 Iron deficiency and RDW in idiopathic pulmonary arterial hypertension (IPAH) Iron deficiency is thought to be common in IPAH, affecting 31-63% of patients with prevalent disease, and has been associated with inappropriately raised hepcidin levels, reduced exercise capacity and increased mortality (Rhodes et al. JACC 2011; Soon et al. Thorax 2011; Ruiter et al. Eur Resp J 2011; van Empel Heart Lung Circ 2014). Iron status can be difficult to determine in diseases such as IPAH using standard laboratory assays due to the presence of inflammation, because inflammation induces ferritin while repressing serum iron and transferrin saturation (Weiss & Goodnough NEJM 2005), whereas soluble transferrin receptor (sTfR) levels are largely unaffected (Cook et al. Br J Haematol 1990). Some studies therefore defined iron deficiency in IPAH by raised sTfR levels; i.e., either above the normal range (8.7-28.1 nmol/L) without overt anaemia (Rhodes et al. JACC 2011(Rhodes, Howard et al. 2011)) or >4.4 mg/L (51.7 nmol/L) in females and >5.0 mg/L (58.8 nmol/L) in males (van Empel Heart Lung Circ 2014). Other studies have used a definition based either on serum iron <10 µmol/L and transferrin saturation <15% in women and <20% in men (Ruiter et al. Eur Resp J 2011) or reduced ferritin (<10 µg/L) with elevated/normal transferrin levels and normal CRP (Soon et al. Thorax 2011). Hepcidin, which is normally decreased in iron deficiency to facilitate increased iron uptake from the gut, has been shown to be inappropriately raised in IPAH in relation to iron status in IPAH (Rhodes et al. JACC 2011(Rhodes, Howard et al. 2011)). Despite

56 increased levels of circulating inflammatory markers IL-6 levels, no relationship has been found between hepcidin and IL-6 or CRP concentrations, suggesting that iron deficiency is not being driven by inflammation in IPAH (Rhodes et al. JACC 2011(Rhodes, Howard et al. 2011)). It has been postulated that the expression of hepcidin may be increased due to abnormal signalling in the hepatocyte due to downregulation of BMPR2 (Rhodes et al. JACC 2011(Rhodes, Howard et al. 2011)), which may explain why iron deficiency appears to be more severe in patients with heritable PAH with mutations in BMPR2 (Soon et al. Thorax 2011).

Increased RDW accompanies iron deficiency in IPAH patients, correlating with reduced serum iron, ferritin levels, transferrin saturation and increased soluble transferrin receptor concentration (Rhodes et al. JACC 2011(Rhodes, Howard et al. 2011)). Furthermore, increased RDW associates with disease severity and independently predicts mortality, even when measured in combination with 6MWD, NT-proBNP and other established clinical indices in IPAH patients (Rhodes, Wharton et al. 2011) A similar relationship between RDW and disease severity have also been reported in other studies involving cohorts including patients with IPAH (Hampole et al. Am J Cardiol. 2009(Hampole, Mehrotra et al. 2009); Decker et al. Clin Trans Sci 2011).

Overall, these data suggest that RDW may serve as a surrogate measure of abnormal iron homeostasis and disease severity in IPAH. Nonetheless, there is a need to validate the use of sTfR and RDW in the assessment of iron deficiency in IPAH and examine the potential impact of inflammation and other confounding factors on RDW in this disease. In addition, there is a lack of information about the occurrence of iron deficiency and raised RDW in other types of pulmonary hypertension. The relationship between RDW and mortality risk also needs to be explored in incident patients as the correlation between RDW and survival in pulmonary hypertension may, in part, reflect its association with prevalent disease.

2.1.2.1 Defining iron deficiency in IPAH There is no universally agreed definition of iron deficiency. However, circulating sTfR levels are largely independent of inflammation and an upper limit of normal (28.1 nmol/L) has been used to demonstrate a high prevalence of iron deficiency in patients with IPAH (Rhodes, Howard et al. 2011). Furthermore, 83% of patients with sTfR >28.1 nmol/L may be correctly identified if any one of the following criteria are met: ferritin <37 µg/L, iron < 10.3 µmol/L or transferrin saturation <16.4% (Howard, Watson et al. 2013). The iron status of IPAH patients may also be stratified by RDW, iron deficiency being accompanied by increased RDW.(Rhodes, Wharton et al. 2011) In an analysis of 98

IPAH patients, the median RDW (14.9%, IQR 13.7% to 16.4%) was found to lie towards the upper end of the normal range (8.3% to 17.5%) and in iron deficient patients (sTfR >28.1 nmol/L) the median RDW was higher at 15.7% (14.4% to 17.6%). A RDW cut-off of 15.7% was also shown to be an independent predictor of survival in IPAH (Rhodes, Wharton et al. 2011).

2.1.3 Iron deficiency and RDW in pulmonary arterial hypertension secondary to connective tissue disease (PAH-CTD) Circulating autoantibodies (Isern, Yaneva et al. 1992) and inflammatory cytokines (IL-1 and IL-6) (Humbert, Monti et al. 1995) have been implicated in the development of PAH associated with autoimmune diseases and connective tissue diseases such as systemic sclerosis, systemic lupus erythematosus, and mixed connective tissue disease. Despite similarities in the pathobiology and haemodynamic perturbations, clinical outcomes differ significantly in patients with PAH-CTD compared to other forms of PAH, especially in PAH associated with systemic sclerosis (SSc-PAH) (Kawut, Taichman et al. 2003, Fisher, Mathai et al. 2006, Condliffe, Kiely et al. 2009). Pulmonary hypertension is the leading cause of death in systemic sclerosis, a chronic disease, characterized by progressive collagen production, endothelium dysfunction, chronic inflammation and autoantibody production.(Le Pavec, Humbert et al. 2010, Klein-Weigel, Opitz et al. 2011),(Launay, Hervé et al. 2002)

A study investigating the frequency of i, iron deficiency was more prevalent in patients with scleroderma-related PAH than in patients with systemic sclerosis without pulmonary hypertension (46.1% vs 16.4%) and was associated with a lower survival rate after 4 years (Ruiter, Lanser et al. 2014). Iron deficiency was also associated with exercise impairment in systemic sclerosis patients, irrespective of whether they had pulmonary hypertension. In both groups, iron deficient patients exhibited lower serum iron, transferrin saturation and ferritin levels and significantly lower hepcidin levels compared to non- iron deficient patients. However, IL-6 levels were comparable in both and did not correlate with hepcidin levels(Ruiter, Lanser et al. 2014).

A review of the literature revealed that there have been no reports to date focussing on RDW in patients with PAH-CTD, although several studies have explored RDW in connective tissue diseases, including rheumatoid arthritis (Rodriguez-Carrio et al. Rheumatology 2015 & Atherosclerosis 2015; Hassan et al. Clin Rheumatology 2015), Sjӧgren’s syndrome (Clin Biochm 2014), systemic sclerosis (Farkas et al. Rheumatology 2014), and systemic lupus erythematosus (Vaya et al. Clin Hemorheol

Microcirc 2013). Considered together, the results of these studies demonstrated that RDW was associated with increased inflammatory markers, disease activity and cardiovascular risk.

2.1.4 Iron deficiency and RDW in pulmonary arterial hypertension secondary to congenital heart disease (PAH-CHD) Approximately a third of patients with cyanotic congenital heart disease and/or Eisenmenger syndrome are reported to be iron deficient (Kaemmerer et al. Am J Cardiol 2004; Diller et al. Eur Heart J 2006;(Kaemmerer, Fratz et al. 2004, Diller, Dimopoulos et al. 2006) Van De Bruaene et al. Eur Heart J 2011). In contrast to iron-replete patients, an inverse relationship was found between resting oxygen saturation and haemoglobin levels in iron deficient patients (Diller et al. Eur Heart J 2006) and iron deficiency was associated with a higher risk of adverse outcome (Van De Bruaene et al. Eur Heart J 2011). The use of oral anticoagulation and frequent phlebotomies was independently related to iron deficiency in Eisenmenger patients (Van De Bruaene et al. Eur Heart J 201), but other possible causes were thought to include increased iron consumption through erythropoiesis secondary to chronic hypoxia, bleeding from arteriovenous malformations or collateral vessels, abnormal haemostasis and limited dietary intake or absorption of iron(Tay, Peset et al. 2011). Iron deficiency has been associated with cerebrovascular events in these patients(Ammash and Warnes 1996) as well as reduced exercise capacity, the latter thought to be secondary to reduced oxygen delivery and its effects on skeletal muscle cell metabolism(Suedekum and Dimeff 2005). Iron deficiency is not thought to increase whole blood viscosity, common in cyanotic congenital heart disease due to the secondary erythropoiesis, as previously hypothesised (Broberg, Bax et al. 2006)

Finally, increased RDW is reported to predict adverse outcomes in children undergoing surgery for congenital heart disease (Massin Pediatric Cardiol. 2012; Polat et al. Biomed Res Int. 2014) and to be associated with ferritin levels and mortality in adults with congenital heart disease (Martínez- Quintana & Rodríguez-González Congenit Heart Dis 2013; Miyamoto et al. Circ J 2015).

2.1.5 Iron deficiency and RDW in chronic thromboembolic pulmonary hypertension (CTEPH) Little is known about the iron status of patients with CTEPH, although two studies to date have described no significant association with iron metabolism. Patients with CTEPH are routinely anticoagulated, yet the prevalence of iron deficiency has been found to be significantly higher in IPAH than CTEPH (Soon, Treacy et al. 2011). Iron deficiency, defined as decreased ferritin level (<10

µg/l) with an elevated or normal transferrin (2.2 – 4.0 g/L) and a normal CRP level, was significantly less frequent in males (2% versus 28%) and premenopausal females (8% versus 50%) with CTEPH compared to those with IPAH. Although this was not attributable to differences in warfarin or endothelin receptor antagonist use, the relationship between iron status and IL-6 levels is thought to vary between the two diseases (Soon, Treacy et al. 2011). In a separate case-control study in China comparing 45 CTEPH patients and 36 age- and sex-matched chronic thromboembolic patients without pulmonary hypertension, no difference was found in the iron status of the CTPEH group (Xi et al. Cardioavsc Pathol. 2015).

RDW is postulated to predict early mortality (Zorlu et al. Am J Cardiol 2012; Ozsu et al.ClinApll Thrombosis/Hemostasis 2014) and the development of CTEPH in patients with pulmonary embolism (Abdul et al. Chronic Resp Disease). Increased RDW has also been described in patients with CTEPH compared with healthy controls (Wang et al. Clin Respir 2016).

2.1.6 Hypothesis and Aims I hypothesised that there is a high prevalence of iron deficiency in patients with other forms of pulmonary hypertension and this is associated with increased RDW and adverse clinical features and prognosis.

The aims of this chapter are to: 1. Derive cut-offs using standard laboratory-based measures of iron status (including RDW) to predict iron deficiency in patients with IPAH as defined by raised sTfR. 2. Using the same cut-offs, determine whether iron deficiency and raised RDW is a frequent occurrence in incident patients with IPAH, PAH-CTD, PAH-CHD or CTEPH. 3. Investigate if RDW and iron deficiency are associated with baseline measurements of cardiopulmonary haemodynamics, cardiac function and exercise capacity and predicts survival in these patients.

2.2 Methods

2.2.1 Subjects, samples and clinical investigations In order to validate the finding that iron deficiency is common in patients with IPAH cohort and is associated with increased RDW (Rhodes et al., JACC 2011, Heart 2011), iron parameters and RDW were assessed in blood samples obtained from a distinct cohort of consenting IPAH patients (n=153), during consecutive clinical appointments at the National Pulmonary Hypertension Service, Hammersmith Hospital, between 29th November 2011 and 2nd September 2013.

The wider prevalence and clinical relevance of iron deficiency was subsequently examined in 1318 consecutive patients with incident pulmonary hypertension who attended the National Pulmonary Hypertension Service at Hammersmith Hospital between August 1996 and January 2016. Patients were followed for ten years and survival status was censored on 12 January 2016. The diagnosis of PAH and CTEPH was based on standard criteria and confirmed by right heart catheterization(Hoeper, Bogaard et al. 2013). Clinical measurements were obtained from patients with incident disease, prior to therapy being started and within 30 days of the date of diagnosis. This included all-comers with no exclusion for age or co-morbidities. Circulating iron parameters (plasma ferritin, iron, and transferrin saturation) were taken from the blood sample closest to the date of diagnosis.

Demographic and clinical data – Patient information, including cardiopulmonary haemodynamic data, therapy, co-morbidity, hospitalisation and mortality, was obtained using the TRIPHIC (Translational Research in Pulmonary Hypertension at Imperial College) database. This is a secure source of pseudonymised information on patients referred to the National Pulmonary Hypertension Service at Hammersmith Hospital, which also links clinical data with biological samples and has received approval from both a Research Ethics Committee (REC 13/LO/0695) and the NHS Health Research Authority, Confidential Advisory Group (CAG 4-09(a) 2013). The WHO functional class and exercise capacity (6 minute walk distance) were assessed within 30 days of blood sampling (normally at the same visit).

Cardiac magnetic resonance (CMR) imaging – CMR data was obtained via the picture archiving and communication system (PACS) at Hammersmith Hospital. All CMRs were reported by a consultant cardiac radiologist. However, all raw data were reviewed as part of a validation and sense check.

Cardiopulmonary exercise testing (CPET) – CPET was performed on an upright cycle ergometer using a breath-by-breath system (Master Screen 144 CPX; Jaeger; Hoechberg, Germany) according to the ATS/ACCP Statement on Cardiopulmonary Exercise Testing (AJRCCM 2003). Parameters used for analysis were peak VO2 (percent predicted and absolute, ml/min/kg) and VE/CO2 slope since these variables have been shown to be of prognostic significance in pulmonary hypertension and are included in the ESC / ERS guidelines for risk stratification of pulmonary arterial hypertension. (ESC/ERS Guideline)

Blood samples and circulating iron parameters – Blood samples were routinely drawn from the antecubital fossa of patients attending clinical appointments. Research blood samples were obtained with informed consent and local Research Ethics Committee approval (REC 11/LO/0395). Blood was collected in EDTA- and lithium-heparin Vacutainer tubes, immediately put on ice, centrifuged (1,300g, 15 minutes) within 30 minutes and aliquots of plasma stored at -80⁰C until assayed. Soluble transferrin receptor (sTfR) and IL-6 levels were determined using a commercial enzyme-linked immunosorbent assay (ELISA) kit (R&D Laboratories, Abingdon, Oxfordshire, UK), as previously described (Rhodes et al. JACC 2011). Hepcidin levels were determined by a competitive radioimmunoassay (Busbridge et al. 2009). RDW, Iron, ferritin and transferrin saturation were measured by standard clinical pathology accredited hospital assays.

The normal range of hepcidin was previously determined at 2-55 ng/mL and previously independently validated in a second group of healthy volunteers (Ashby et al. 2009).

2.2.2 Data presentation and statistical analysis Data are presented as percentages, mean (±standard deviation, SD), 95% confidence interval (95% 99 CI) or median and percentile range. Several variables were not normally distributed, as shown by the Kolmogorov-Smirnov test on cardiac output values, and were therefore transformed to their natural logarithm or square root, as appropriate to best normalise the data (Figure 2.1). In turn, normalised values were transformed into z-scores for ease of comparison between variables.

Kolmogorov-Smirnova Statistic Sig. Cardiac Output 0.111 0 CO sqrt 0.073 0.001 CO log 0.038 .200*

Figure 2.1. Normalization of data. In this example, the distribution of cardiac output (CO)

values is skewed and was therefore transformed to their square root (CO sqrt) or natural logarithm (CO log), as appropriate to best normalize the data. CO, cardiac output (L/m).

Correlations were assessed by Spearman’s Rank test and Chi-squared (χ2) analysis was used to assess categorical differences. Receiver Operator Characteristic (ROC) analysis was used to assess possible cut-off points for circulating markers of iron deficiency and prognostic value of measured variables against all-cause mortality at 3 years. This time point was used throughout as it provided the optimum balance between number of patients with follow-up and number of events (deaths). The area under the curve (c statistic) was used to compare parameters and prognostic value. Kaplan- Meier plots illustrated events during follow-up, depicting the relationship between iron deficiency (as assessed by iron, transferrin saturation and RDW) and cumulative survival in years. Cox regression analyses assessed the predictive power of individual variables. Significant predictors (p<0.05) were tested in a multi covariate model for both testing individual variables as well as overall survival, which was developed by analysis of proportional hazard ratios. All statistical calculations were performed using SPSSv22.0 (SPSS, Inc., Chicago, IL) and GraphPad Prism 5 (GraphPad, San Diego).

2.3 Results

2.3.1 Patient demographics

Plasma samples were analysed from 153 incident and prevalent IPAH patients (Cohort 1) with the full complement of iron status including sTfR. They comprised predominately (92%) prevalent patients, receiving targeted PAH therapies at the time of sampling, and 106 (69%) were also on warfarin (Table 2.1). Median time between diagnosis and sampling was 2.8 years (IQR 0.4-6.0 years).

A total of 1,318 patients with incident IPAH/HPAH, PAH-CTD, PAH-CHD or CTEPH were subsequently analysed (Cohort 2). They included patients with IPAH/HPAH (n=427/25); PAH-CTD (n=129; 69 scleroderma and 60 non-scleroderma); PAH-CHD (n=255; 97 Eisenmenger syndrome, 3 Fontan circulation and 155 other forms of congenital heart disease); and CTEPH (n=482). Some individuals were excluded at certain points in the analyses because specific measurements were not available within 30 days of diagnosis. Patients routinely had a 6MWT during their initial diagnostic assessment, but some then had a cardiopulmonary exercise test instead at a follow up appointment. Subject demographics and clinical data are detailed in Table 2.2.

Units Mean (±SD) Demographics Age Yrs 50.3 (±17.9) Gender Female 105 (68.6%) Male 48 (31.4%) Ratio 2.2:1 WHO FC I 2 (1.3%) II 29 (19.0%) III 102 (66.7%) IV 20 (13.0%) Warfarin Yes 53 (34.6%) No 100 (65.4%) BSA m2 1.9 (±0.2) Physiology RAPm mmHg 9.6 (±5.2) PAPm mmHg 52.9 (±13.4) PWPm mmHg 10.9 (±4.7) PVR WU 10.9 (±4.7) PVR index WU.m2 19.9 (±9.5) Cardiac Output L/min 4.5 (±1.7) Cardiac index L/min/m2 2.4 (±0.9) 6MWD m 280.9 (±159.8) Iron and inflammation Hb g/dL 15 (±12) sTfR nmol/L 31.4 (±14.2) RDW % 14.9 (±2.1) Ferritin µg/L 97 (±103.0) Iron µmol/L 14.1 (±6.8) TSAT % 23 (±11.0) Hepcidin ng/ml 28 (±36.0) IL-6 pg/ml 5 (±6.0)

Table 2.1. Baseline characteristics of Idiopathic PAH patients in Cohort 1. Data shown as mean (± standard deviation) or number (%) of patients at the time of diagnosis (haemodynamic measurements) or blood sampling (WHO functional class, 6MWD, iron, ferritin, transferrin saturation, sTfR, RDW, haemaglobin). WHO-FC, World Health Organisation functional class; BSA, body surface area (m2); RVEF, right ventricular ejection fraction (%); RAPm (mmHg), mean right atrial pressure (mmHg), PWPm (mmHg); mean pulmonary wedge pressure (mmHg); PVR (WU), pulmonary vascular resistance (WU); PVRI, pulmonary vascular resistance index (WU.m2); 6MWD,

six minute walk distance (m); VO2 max, maximum oxygen consumption (mls/Kg/min); sTfR, soluble transferrin receptor (nmol/L); RDW, red cell distribution width (%); TSAT, transferrin saturation (%); IL-6, interleukin 6 (pg/ml).

Units IPAH/HPAH PAH-CTD PAH-CHD CTEPH Demographics Number 427/25 129 255 482 Age Yrs 54.5 (±18.5) 60.2 (±14.8) 42 (±21.4) 63.0 (±15.6) Sex Female 305 104 164 240 Male 147 25 91 242 Ratio 2.1:1 4.2:1 1.8:1 1:1 WHO FC I 16 (3.5%) 5 (3.9%) 25 (9.8%) 13 (2.7%) II 63 (13.9%) 21 (16.3%) 66 (25.9%) 73 (15.1%) III 276 (61.1%) 81 (62.8%) 147 (57.6%) 316 (65.6%) IV 97 (21.5%) 22 (17.1%) 16 (6.3%) 80 (16.6%)

CMR imaging Number 125 32 21 259 RVEDV ml 204.5 (±74.4) 181.5 (±49.2) 239 (±91.0) 189.7 (±62.4) RVSV ml 71.0 (±29.6) 73.5 (±23.1) 106.3 (±45.2) 69.6 (±23.0) RVEF % 36.9 (±12.7) 41.7 (±12.0) 46.4 (±14.4) 39.0 (±13.8) RV output l/min 5.3 (±2.2) 5.3 (±1.7) 8.1 (±3.8) 5.4 (±3.1) LVEDV ml 104.3 (±45.6) 105.4 (±30.2) 102.0 (±41.8) 109.8 (±36.2) LVSV ml 62.2 (±29.4) 62.8 (±18.4) 59.9 (±17.8) 66.0 (±22.6) LVEF % 60.4 (±11.0) 59.7 (±7.4) 61.4 (±11.5) 61.7 (±11.2) LV output l/min 5.0 (±4.4) 4.5 (±1.2) 4.6 (±1.5) 11.3 (±22.9)

Cardiopulmonary haemodynamics Number 291 78 26 379 RAPm mmHg 10.4 (±6.3) 10.2 (±5.6) 10.2 (±6.0) 10.3 (±5.5) PAPm mmHg 51.5 (±13.9) 42.5 (±12.2) 41.7 (±15.8) 44.6 (±12.8) PWPm mmHg 11.0 (±4.9) 10.6 (±4.3) 12.1(±4.3) 11.4 (±4.5) PVR WU 11.4 (±6.3) 9.0 (±5.4) 6.7 (±5.1) 9.2 (±6.0) PVR index WU.m2 20.2 (±10.7) 15.5 (±9.8) 11.0 (±8.3) 15.9 (±9.1) Cardiac output l/min 4.3 (±1.8) 4.3 (±2.5) 5.5 (±2.9) 4.3 (±1.6) Cardiac index l/min/m2 2.3 (±0.9) 2.7 (±2.0) 3.0 (±1.3) 2.3 (±0.8)

Exercise Number 6MWD 391 111 86 417 Number CPET 217 64 25 281 6MWD m 244.4 (±156.6) 217.5 (142.8) 295.3 (±148.4) 254.4 (148.6)

Peak VO2 ml/Kg/min 12.4 (±5.5) 10.7 (±3.5) 13.9 (±3.8) 12.8 (±4.1)

Peak VO2 % predicted 52.9 (±16.7) 47.5 (±12.6) 59.0 (±16.3) 58.9 (±17.9)

VE/VCO2 slope l/min/l/min 48.1 (±17.9) 50.9 (±20.7) 45.1 (±17.3) 50.2 (±16.8)

Iron parameters and clinical biochemistry Number 452 129 255 482 Hb g/dL 14.1 (±2.5) 13.0 (±2.0) 15.6 (±3.8) 14.0 (±2.1) RDW % 15.5 (±2.5) 16.4 (2.2) 15.6 (±2.9) 15.7 (±2.5) Ferritin µg/L 154.1 (±392.6) 130.7 (±162.7) 75.8 (±101.2) 142.1 (±426) Iron µmol/L 14.0 (±7.9) 10.8 (5.1) 13.8 (±8.5) 15.3 (±9.1) TSAT % 22.4 (±12.6) 19.0 (10.3) 22.4 (±13.4) 24.3 (±15.0) BNP ng/L 505.0 (±594.6) 612.0 (±659.6) 274.6 (±438.1) 407.7 (±582.9) CRP mg/L 9.8 (±16) 11.2 (12.6) 8.7 (±12.4) 12.3 (±18.9) Creatinine µmol/L 99.8 (±55.7) 91.2 (±62.3) 85.0 (25.9) 93.0 (±31.6)

Table 2.2. Baseline characteristics of incident IPAH, PAH-CTD, PAH-CHD and CTEPH patients (Cohort 2). Data shown as mean (± standard deviation) or number (%) for incident patients at time of diagnosis. IPAH, idiopathic pulmonary arterial hypertension; PAH-CTD, PAH secondary to congenital heart disease; PAH-CHD, PAH secondary to congenital heart disease; CTEPH, chronic thromboembolic pulmonary hypertension; WHO-FC, World Health Organisation functional class; RVEDV, right ventricular end diastolic volume (mls); RVSV, right ventricular stroke volume (mls); RVEF, right ventricular ejection fraction (%); RV, right ventricle; LV, left ventricle; LVEDV, left end diastolic volume (mls); LVSV, left ventricular stroke volume (mls); LVEF, left ventricular ejection fraction (%); RAPm, mean right atrial pressure (mmHg); PAPm, mean pulmonary arterial pressure (mmHg); PWPm, mean pulmonary wedge pressure (mmHg); PVR, pulmonary vascular resistance (WU); PVRI, pulmonary vascular resistance index (WU.m2); 6MWD, six minute walk distance (m);

RER max, maximum respiratory exchange ratio; VO2 max, maximum oxygen consumption

(mls/kg/min); PETCO2, partial pressure of end-tidal carbon dioxide (kPa); HR, heart rate (bpm);

O2/HR, oxygen pulse (mls/beat); OUE slope, oxygen uptake efficiency slope (OUES); VE/VCO2, ventilatory equivalent for carbon dioxide; Hb, haemoglobin (g/L); RDW, red cell distribution width (%); TSAT, transferrin saturation (%); BNP, brain natriuretic peptide (ng/L); CRP, c-reactive protein (mg/L).

2.3.2 Prevalence of iron deficiency defined by sTfR in IPAH Blood samples from Cohort 1 were examined to determine the frequency of iron deficiency in IPAH and possible association with abnormal RDW. Circulating sTfR levels were raised above the normal range (>28.1 nmol/L) in 70 of 153 patients (46%). RDW was above or towards the upper end of the normal range and only five patients had haemoglobin levels below 10 g/dL (Figure 2.2). IL-6 levels were also raised (Figure 2.2.)

A B C 120 350

100 300

250 80

200 60 150

sTfR nmol/L sTfR 40 Hepcidin ng/ml 100

20 50

0 0

D E F 700 50 80

600 40

500 60

g/L 30

 400 mol/L

 40

300 20

Iron Ferritin 200 20

10 saturation (%) Transferin 100

0 0 0

G H 25 50

40 20

30 15

20 IL-6 pg/ml

Haemoglobin (g/dL) 10 10

5 0

Figure 2.2. Characterization of Iron Status, Hb and inflammation in 153 patients with Idiopathic

PAH (Cohort 1) - soluble transferrin receptor (sTfR) (A), Red Cell Distribution Width (RDW) (B)

Hepcidin (C); Ferritin (D);Iron (E), and Transferrin saturation (TSAT) (F); Haemoglobin (Hb) (G);

Interleukin-6 (IL-6). Shaded areas indicate normal ranges in healthy populations for sTfR (8.7-

28.1 nmol/L), RDW (10-16%), ferritin (15-300 µg/L, iron (7-29 µmol/L), transferrin saturation (20- 45%), haemoglobin (11.5-17 g/dL) and IL-6 (pg/ml).

2.3.3 sTfR as a predictor of mortality Being able to predict true iron status (sTfR>28.1nmol/L) may be useful surrogate marker for mortality. An elevated sTfR shows strong correlations with mortality in Cohort 1, figure (2.3), p<0.001.

Mortality N Positive 43 Negative 110 Missing 0

95% CI Variable Area Sig. Lower Upper sTfR 0.693 <0.001 0.604 0.781

Figure 2.3. sTfR and mortality. Receiver operating characteristic (ROC) analysis showing relationship between sTfR and mortality. Positive=died, negative=alive. sTfR, soluble transferrin receptor (nmol/L).

2.3.4 Prediction of iron deficiency from standard clinical laboratory measurements in IPAH

To quantify the relative numbers of patients who would be designated as being iron-deficient in each of the four PH subtypes, established cut-offs used to predict elevated sTfR levels were validated (sTfR >28.1) in IPAH from cohort 1 (IPAH n=153) and compared with established previous derived values (ferritin <37 µg/L, iron < 10.3 µmol/L or transferrin saturation <16.4%).(Howard, Watson et al. 2013)(Howard et al 2013)

sTfR > 28.1 Yes No Total Yes 43 28 71 Iron deficiency No 27 55 82 Total 70 83 153

Table 2.3. Iron deficiency and ID. Chi-squared (χ2) analysis showed that iron deficiency was

predicted with 61% sensitivity and 67% specificity using these criteria in Cohort 1.

ROC analysis was used to compare iron parameters and RDW as markers of iron deficiency (i.e. predictor of sTfR level >28.1 nmol/L). Iron, transferrin saturation and RDW significantly predicted iron deficiency, but RDW was the strongest performer with an area under the curve of 0.764 (p<0.001) (Figure 2.4.)

sTfR> 28.1 N Positive 67 Negative 76 Missing 10

Area Sig (p) 95% CI Lower Upper Hgb 0.568 0.162 0.336 0.528 RDW 0.764 <0.001 0.685 0.843 Ferritin 0.575 0.124 0.33 0.521 Iron 0.698 <0.001 0.217 0.388 TSAT 0.727 <0.001 0.191 0.355 IL-6 pg/ml 0.677 <0.001 0.588 0.766 Figure 2.4. sTfR associations using standard iron markers and inflammation (IL-6). Receiver operating characteristic (ROC) analysis showing relationship between iron parameters, RDW and iron deficiency, as defined by sTfR >28.1, in IPAH, Cohort 1 (n=153). 153 patients with a complete

sTfR, soluble transferrin saturation (nmol/L); iron (µg/L); ferritin (µg/L); TSAT, transferrin saturation (%); Hb, haemoglobin (g/dL); RDW, red cell distribution width (g/L) and IL-6, interleukin- 6 (pg/ml) profiles were analysed. P, significance.

The best-performing ROC derived cut-offs were subsequently derived (i.e. optimum balance between sensitivity and specificity) for sTfR > 28.1 nmol/l using iron, ferritin, transferrin saturation and RDW in Cohort 1 (Tables 2.4 and 2.5). These were then compared with those previously derived

71 selected for iron status(Howard, Watson et al. 2013) (Table 2.6). ROC derived cut-offs were similar to median values of RDW, iron, ferritin and transferrin saturation and comparable with those previously reported in IPAH patients(Howard, Watson et al. 2013).

Using these new derived cut-offs (Tables 2.4 and 2.5) from Cohort 1 further Chi-squared analysis showed that by using Ferritin < 37.5 µg/L or Iron <14.5 µmol/L or TSAT <21.5% resulted in improved specificity (compared to Howard et al. Table 2.3), (Table 2.7).

When RDW is added to this combination sensitivity and specificity is improved in predicting sTfR >28.1; Sensitivity 54 to 58% and specificity 71 to 91% (Table 2.8.).

Iron Sensitivity 1 - Specificity Sensitivity - Specificity Ferritin Sensitivity 1 - Specificity Sensitivity - Specificity 2 1 1 0 19.5 0.836 0.947 0.111 3.5 0.97 1 0.03 20.5 0.806 0.947 0.141 4.5 0.955 1 0.045 21.5 0.806 0.921 0.115 5.5 0.91 1 0.09 22.5 0.791 0.921 0.13 23.5 0.776 0.921 0.145 6.5 0.881 0.987 0.106 24.5 0.761 0.908 0.147 7.5 0.836 0.947 0.111 25.5 0.746 0.908 0.162 8.5 0.731 0.921 0.19 26.5 0.731 0.882 0.151 9.5 0.627 0.882 0.255 27.5 0.716 0.882 0.166 10.5 0.552 0.803 0.251 29 0.701 0.882 0.181 11.5 0.493 0.763 0.27 30.5 0.701 0.868 0.167 12.5 0.403 0.671 0.268 32.5 0.687 0.868 0.181 13.5 0.328 0.592 0.264 34.5 0.657 0.868 0.211 14.5 0.239 0.513 0.274 35.5 0.642 0.855 0.213 15.5 0.164 0.421 0.257 36.5 0.627 0.842 0.215 16.5 0.164 0.382 0.218 37.5 0.612 0.842 0.23 17.5 0.119 0.289 0.17 38.5 0.612 0.829 0.217 18.5 0.06 0.276 0.216 40 0.612 0.816 0.204 19.5 0.06 0.224 0.164 41.5 0.612 0.789 0.177 20.5 0.06 0.211 0.151 42.5 0.612 0.763 0.151 44 0.567 0.737 0.17 21.5 0.06 0.184 0.124 45.5 0.567 0.711 0.144 22.5 0.03 0.132 0.102 47 0.567 0.697 0.13 24 0.015 0.092 0.077 48.5 0.552 0.697 0.145 25.5 0.015 0.079 0.064 49.5 0.552 0.671 0.119 27.5 0.015 0.066 0.051 50.5 0.552 0.658 0.106 32 0.015 0.026 0.011 52 0.537 0.645 0.108 41 0.015 0.013 0.002 53.5 0.522 0.645 0.123 47.5 0.015 0 0.015 55 0.522 0.618 0.096 49 0 0 0 56.5 0.507 0.618 0.111

Table 2.4. Iron and ferritin ROC derived cut-offs predicting sTfR level >28.1 nmol/L. sTfR,

soluble transferrin saturation (nmol/L). Optimum cut-offs for iron (11.5 and 14.5 µg/L) and ferritin (37.5 µmol/L) are highlighted.

TSAT Sensitivity 1 - Specificity Sensitivity - Specificity RDW Sensitivity 1 - Specificity Sensitivity - Specificity 10.5 0.851 0.987 0.136 14.05 0.776 0.487 0.289 11.5 0.821 0.987 0.166 14.15 0.731 0.408 0.323 12.5 0.776 0.974 0.198 14.25 0.716 0.408 0.308 13.5 0.701 0.961 0.26 14.35 0.701 0.395 0.306 14.5 0.687 0.934 0.247 14.45 0.701 0.329 0.372 15.5 0.657 0.908 0.251 14.55 0.672 0.303 0.369 16.5 0.597 0.842 0.245 14.65 0.612 0.224 0.388 17.5 0.552 0.842 0.29 14.75 0.582 0.171 0.411 18.5 0.493 0.776 0.283 14.85 0.567 0.158 0.409 19.5 0.418 0.697 0.279 14.95 0.567 0.118 0.449 20.5 0.373 0.658 0.285 15.05 0.552 0.105 0.447 21.5 0.284 0.632 0.348 15.15 0.493 0.079 0.414 22.5 0.254 0.579 0.325 15.25 0.493 0.066 0.427 23.5 0.224 0.539 0.315 15.35 0.478 0.066 0.412 24.5 0.194 0.487 0.293 15.45 0.433 0.066 0.367 25.5 0.164 0.447 0.283 15.55 0.418 0.053 0.365 26.5 0.119 0.421 0.302 15.65 0.388 0.053 0.335 27.5 0.119 0.355 0.236 15.75 0.388 0.026 0.362 28.5 0.104 0.316 0.212 15.85 0.328 0.026 0.302 29.5 0.075 0.303 0.228 15.95 0.284 0.026 0.258 30.5 0.075 0.276 0.201 16.25 0.269 0.026 0.243

Table 2.5. Transferrin saturation and RDW derived cut-offs predicting sTfR >28.1nmol/L. TSAT, transferrin saturation (%); RDW, red cell distribution width (%). Optimum cut-offs for

transferrin saturation (21.5%) and RDW (14.95%) are highlighted.

Howard et al.(Howard, Watson et al. Cohort 1 2013) Red cell distribution width (%) 14.95 - Transferrin saturation (%) 21.50 16.40 Iron (µmol/L) 14.50 10.30 Ferritin (µg/L) 37.50 37.00

Table 2.6. Comparison of ROC derived cut-offs predicting true iron deficiency (sTfR level >28.1 nmol/L) in patients with IPAH.

sTfR>28.1 Yes no Total Yes 57 48 105 Iron deficiency No 14 34 48 Total 71 82 153

2 Table 2.7. Chi-squared (ᵡ ) analysis showed that iron deficiency was predicted with 54%

sensitivity and 71% specificity using Ferritin <37.5µg/L or TSAT <21.5%.

sTfR>28.1 Yes No Total Yes 68 50 118 Iron deficiency No 3 32 35 Total 71 82 153

2 Table 2.8. Chi-squared (ᵡ ) analysis showed that iron deficiency was predicted with 58% sensitivity and 91% specificity using Ferritin <37.5µg/L or TSAT <21.5%.

2.3.3.1 What is the best single, independent marker in predicting iron deficiency? sTfR is acknowledged as the best predictor for true iron deficiency in pulmonary arterial hypertension (Rhodes et al. JACC). Having now established unique derived cut offs for predicting sTfR>28.1 from ferritin, iron, TSAT and RDW (Table 2.6), the best marker for predicting sTfR was evaluated. Currently the sTfR lab assay is not readily available in everyday clinical medicine therefore a cheap, fast and readily available predictor of sTfR is essential.

Table 2.9. and figure 2.5. show correlations with all markers of iron deficiency but RDW shows the best correlation with sTfR in Cohort 1 (r=0.54, p<0.001). There is no relationship with Hgb (p=0.065) which supports previous findings that anaemia does not have strong correlations with ID in IPAH (Rhodes et al. JACC). Hepcidin is positively correlated to sTfR which has been shown previously. (Rhodes et al. JACC). IL-6 is positively correlated with sTfR therefore inflammation may also trigger these iron indices but this is addressed in section 2.3.3.2.

These data suggest RDW has the closest correlation with sTfR. Multiple linear regression modelling showed that RDW was the only iron marker that was an independent predictor of sTfR in Cohort 1 (p<0.001) (Table 2.10). Hepcidin was the only other variable that was signficantas an independent marker of sTfR (p=0.006).

sTfR Iron Ferritin TSAT RDW Hepcidin IL-6 Hgb - - r 0.308** -0.163* 0.353** 0.540** -0.262** 0.290** -0.15 p <0.001 0.048 <0.001 <0.001 0.001 <0.001 0.065 n 150 148 150 152 153 147 152

Table 2.9. sTfR, soluble transferrin receptor (nmol/L) relationship in 153 idiopathic PAH patients (Cohort 1) with (i) iron status - iron (µmol/L); ferritin (µg/L); TSAT, transferrin saturations (%);

RDW, red cell distribution width (%) ; (ii) hepcidin (ng/ml) ; (iii) inflammation (IL-6, interleukin- 6, (pg/L); (iv) haemoglobin (Hgb, haemoglobin (g/L)). Bivariate correlations assessed using

Spearman’s Rank correlation coefficent used to show associations (r, rho), significances (p), and numbers (n).

Figure 2.5. Scatter plot showing the relationship of RDW, red cell distribution width (%), with

sTfR, soluble transferrin receptor (nmol/L).

A Model Included Coefficient Sig. 95.0% CI Variable Beta Lower Upper 1 (Constant) 0.953 -0.182 0.194 RDW 0.351 <0.001 0.172 0.571 2 (Constant) 0.921 -0.173 0.191 RDW 0.348 <0.001 0.175 0.561 Hepcidin -0.258 0.006 -0.432 -0.074 Dependent Variable: sTfR

B Model Excluded Beta Sig. Variable In 1 Hepcidin -.258b 0.006 IL-6 .131b 0.185 Ferritin -.116b 0.23 Iron -.160b 0.119 TSAT -.140b 0.186 CRP .044b 0.646 2 IL-6 .188c 0.052 Ferritin -.069c 0.471 Iron -.115c 0.255 TSAT -.110c 0.288 CRP .038c 0.682

Table 2.10. Stepwise Multiple Linear Regression model from Cohort 1 showing : A. independent predictors of sTfR, and B. excluded variables as not independent predictors of sTfR. Dependent

variable sTfR, soluble transferrin receptor (nmol/L); iron (µmol/L); ferritin (µg/L); TSAT, transferrin

saturations (%); RDW, red cell distribution width (%) ; hepcidin (ng/ml) ; IL-6, interleukin-6 (pg/L); CRP, c-reactive protein (mg/L)

2.3.3.2 Red cell distribution width and inflammation RDW has shown to be the best single marker of ID in cohort 1, however in heart failure RDW may represent a combination of chronic inflammation, dysfunctional erythropoiesis, kidney dysfunction, oxidative stress and nutritional status [Forhecz et al. Am Heart J 2009].

Further multiple regression analysis was performed with RDW as the dependant variable. Transferrin saturation (TSAT) and IL-6 proved to be independent predictors of RDW (Table 2.11.).

RDW is the favourable predictor of sTfR in Cohort 1 but iron deficiency is consistent with a sTfR >28.1 (Rhodes et al. JACC). RDW shows significant correlation with sTfR>28.1 and to correct RDW for the effect of inflammation on the residuals from a linear regression model, with RDW as the dependent variable, and IL-6 as a predictor (independent variable) were calculated for each patient (Figure 2.6 A). In turn, RDW (corrected for inflammation) was compared against other iron markers and still remained favourable (Figure 2.6 B).

IL-6 independently predicts RDW therefore to clarify if RDW independently correlates with sTfR without a confounding inflammatory bias multiple linear regression between RDW and IL-6 was performed to assess prediction of sTfR. This showed that both were independent predictors of sTfR but RDW was favourable in the stepwise multiple regression model. (Table 2.12).

Model Coefficient Sig. 95.0% CI Beta Lower Upper 1 (Constant) 0.745 -0.2 0.144 TSAT -0.434 <0.001 -0.651 -0.268 2 (Constant) 0.739 -0.196 0.14 TSAT -0.409 <0.001 -0.621 -0.245 IL-6 0.211 0.02 0.029 0.341 Dependent Variable: RDW

Excluded Beta t Sig. Model Variables In 1 IL6 .211b 2.358 0.02 CRP .000b -0.005 0.996 Iron .237b 0.988 0.326 Ferritin .026b 0.255 0.8 2 CRP -.025c -0.282 0.779 Iron .340c 1.439 0.153 Ferritin .032c 0.316 0.753 a Dependent Variable: RDW b Predictors in the Model: (Constant), TSAT c Predictors in the Model: (Constant), TSAT, IL-6

Table 2.11. Stepwise Multiple Linear Regression model from Cohort 1 showing independent predictors of RDW (dependant variable); and excluded variables as not independent predictors of RDW. Dependent variable RDW, red cell distribution width (%); iron (µmol/L); ferritin (µg/L);

TSAT, transferrin saturations (%); IL-6, interleukin-6 (pg/L); CRP, c-reactive protein (mg/L)

A B

sTfR nmol/L (>28.1 nmol/L) N sTfR nmol/L (>28.1 nmol/L) N Positive 70 Positive 68 Negative 82 Negative 76 Missing 1 Missing 9

95% CI 95% CI sTfR > 28.1nmol/L Area Sig (p) Lower Upper sTfR> 28.1 nmol/L Area Sig (p) Lower Lower RDW (%) 0.769 <0.001 0.693 0.845 RDW (%) 0.766 <0.001 0.688 0.844 RDW (%) (corrected RDW (corrected for IL6) (%) 0.708 <0.001 0.62 0.796 for IL6) 0.711 <0.001 0.624 0.797 Iron (µmol/L) 0.701 <0.001 0.214 0.383 Ferritin (µg/L) 0.58 0.096 0.324 0.515 TSAT (%) 0.269 <0.001 0.188 0.35

Figure 2.6.

A – ROC curves showing the ability of RDW and RDW corrected for IL-6 to predict sTfR> 28.1 nmol/L. RDW, Red cell distribution width (%); sTfR, soluble transferrin receptor (nmol/L); IL-6,

interleukin-6 (pg/L); iron (µmol/L); ferritin (µg/L); TSAT, transferrin saturations (%). B – ROC curves showing prediction of patients with sTfR> 28.1 nmol/L. RDW corrected for IL-6 (inflammation) is plotted against other standard iron measures.

sTfR>28.1 sTfR<28.1 RDW>14.8 48 16 RDW<14.8 30 70

Table 2.115. Chi-squared (ᵡ2) analysis showes that RDW has a positive predictive value of 0.75 (75%) of predicting sTfR>28.1 and a negative predictive value of (70%). RDW, red cell distribution width (%); sTfR, soluble transferring receptor (nmol/L)

Model Coefficient Sig. 95.0% CI Beta Lower Upper 1 (Constant) 0.614 -0.125 0.21 RDW 0.351 <0.001 0.192 0.527 2 (Constant) 0.618 -0.123 0.206 RDW 0.301 <0.001 0.139 0.477 IL6 0.209 0.013 0.043 0.366 Dependent Variable: sTfR Model Excluded Beta Sig. Variable In 1 IL6 .209b 0.013 a Dependent Variable: sTfR b Predictors in the Model: (Constant), RDW

Table 2.12. Stepwise Multiple Linear Regression model from Cohort 1 showing : independent predictors of sTfR (dependant variable), and B. excluded variables as not independent predictors of sTfR. Variables in model include RDW, red cell distribution width (%); IL-6, interleukin-6 (pg/L).

2.3.4 Iron status in pulmonary hypertension subtypes at diagnosis Circulating transferrin saturation levels are closely correlated and were reduced in all three subtypes of PAH and in patients with CTEPH (Figure 2.7.), many patients displaying levels below the normal range (Figures 2.8 A-D). Iron parameters were consistent with iron deficiency without significant anaemia and were accompanied by an increased RDW (Figures 2.8 A-D).

Figure 2.7. Demonstrating the strong associations (rho) and significance (p) between iron and

transferrin saturations (TSAT) across the four PH subtypes: IPAH, idiopathic pulmonary arterial

hypertension; PAH-CTD, PAH secondary to connective tissue disease; PAH-CHD, PAH secondary to congenital heart disease; CTEPH, chronic thromboembolic pulmonary hypertension. Iron (µg/L); TSAT, transferrin saturation (%).

A 60 60 50

30 40 mol/L)

30 TSAT%

m Iron ( Iron 20 20 25 10

0 0 IPAH IPAH 25

20 10000 RDW (%) RDW

20 1000

g/L) m

100 15

Hb (g/L) Hb Ferritin ( Ferritin

10 10 IPAH 10

1 IPAH 5 IPAH

30 B 60

40 mol/L)

30 m

TSAT% Iron ( Iron 10 20

25 0 0 PAH-CTD PAH-CTD

20 10000

20 RDW (%) RDW 18 1000 16

15 g/L)

m 14 100

Hb (g/L) Hb Ferritin ( Ferritin 10 10 10 PAH-CTD 8

6 1 PAH-CTD PAH-CTD

C 80 100 70

30 60 80 50 60

mol/L) 40 m

30 TSAT% 40 25 ( Iron 20 20 10

0 0 PAH-CHD PAH-CHD 20 30

RDW (%) RDW 1000

15 100

g/L) 20 m

Hb (g/L) Hb 15 10 10 ( Ferritin PAH-CHD 10

1 5 PAH-CHD PAH-CHD

60 100

D 50 80

30 60

mol/L) 30 m

28 TSAT% Iron ( Iron 40 20 26

24 10 20

22 0 0 CTEPH CTEPH 20 10000 22

RDW (%) RDW 20 18

1000 18 16 16

14 g/L)

m 14 100 12

Hb (g/L) Hb 12 Ferritin ( Ferritin 10 CTEPH 10 10 8

6 1 CTEPH CTEPH

Figure 2.8. Iron status and RDW in patients with IPAH (A), PAH-CTD (B), PAH-CHD (C) and CTEPH (D). IPAH, idiopathic pulmonary arterial hypertension; PAH-CTD, PAH secondary to connective tissue disease; PAH-CHD, PAH secondary to congenital heart disease; CTEPH, chronic thromboembolic pulmonary hypertension; RDW, red cell distribution width (%); iron (µg/L); ferritin (µg/L); TSAT, transferrin saturation (%); Hb, haemoglobin (g/L). Shaded areas indicate normal ranges in healthy populations for RDW (10-16%), iron (7-29 µmol/L), transferrin saturation (20-45%), ferritin (15-300 µg/L) and haemoglobin (11.5-17 g/dL).

2.3.5 Prevalence of iron deficiency in pulmonary hypertension subtypes Both sets of criteria were then used to assess the relative frequency of iron deficiency in the four PH subtypes (cohort 2) using cut-offs derived and validated from cohort 1 (Table 2.13.) and cut-offs previously published (Table 2.14.). Despite differences between the cut-offs, the derived frequencies were similar within each cohort and indicated that iron deficiency is prevalent in patients with PAH- CTD, PAH-CHD and CTEPH as well as those with IPAH.

Subgroup Ferritin Iron TSAT Any one of RDW ferritin, iron or TSAT IPAH 48% (133/277) 25% (61/244) 46% (145/314) 57% (177/309) 44% (133/301) PAH-CTD 72% (64/89) 26% (28/95) 62 % (67/108) 66% (70/106) 72% (64/89) PAH-CHD 46% (39/85) 44% (31/71) 48% (44/92) 51% (46/90) 46% (39/85) CTEPH 51% (176/347) 25% (76/309) 39% (143/371) 49% (180/369) 50% (173/348)

Table 2.13. Frequency of iron deficiency in four subtypes of pulmonary hypertension. IPAH, idiopathic pulmonary arterial hypertension; PAH-CTD, PAH secondary to congenital heart disease; PAH-CHD, PAH secondary to congenital heart disease; CTEPH, chronic thromboembolic pulmonary hypertension. Assessed using either specific cut-offs from Tables 2.4 & 2.5 (ferritin <37.5 µg/L, iron <11.5 µmol/L, transferrin saturation (TSAT) < 21.5%), a combination of any one of these or by red cell distribution width (RDW) alone >14.95%.

Subgroup Ferritin Iron TSAT Any one of RDW ferritin or iron or TSAT IPAH 24% (59/244) 39% (122/314) 38% (116/309) 63% (199/316) 37% (112/301) PAH-CTD 29% (28/95) 54% (58/108) 51% (54/106) 60% (66/110) 52% (46/89) PAH-CHD 44% (31/71) 40% (37/92) 43% (39/90) 52% (49/95) 33% (28/85) CTEPH 22% (68/309) 35% (131/371) 34% (126/369) 42% (162/384) 37% (129/348)

Table 2.14. Frequency of iron deficiency in four subtypes of pulmonary hypertension. IPAH, idiopathic pulmonary arterial hypertension; PAH-CTD, PAH secondary to congenital heart disease; PAH-CHD, PAH secondary to congenital heart disease; CTEPH, chronic thromboembolic pulmonary hypertension. Assessed using either specific cut-offs from Table 2.1.6 (Howard, Watson et al. 2013) (ferritin <37.0 µg/L, iron <10.3 µmol/L, transferrin saturation (TSAT) < 16.4%), a combination of any one of these or by red cell distribution width (RDW) alone >15.7%.

2.3.6 Impact of red cell distribution width on cardiopulmonary physiology and outcomes Having shown that RDW appears to be the best marker for predicting sTfR in cohort 1 when correcting for other markers of iron deficiency, inflammation and anaemia, RDW was used in larger groups in the different pulmonary hypertension subtypes to examine the frequency of iron deficiency. Raised RDW appears to be a frequent occurrence in PAH-CTD, PAH-CHD and CTEPH as well as IPAH; I sought to examine potential associations between iron status and cardiopulmonary haemodynamics, cardiac structure and function, exercise capacity and survival in these four subgroups. As sTfR measurements were not available for these patients, median values were selected as the cut-offs in each diagnostic subgroup. This also maximised the statistical power of the analyses and avoided unwarranted extrapolation from IPAH to other disease subtypes with distinct aetiologies.

2.3.6.1 Idiopathic pulmonary arterial hypertension

2.3.6.1.1 Cardiopulmonary haemodynamics and exercise physiology RDW is independently associated with 6MWD and oxygen uptake efficiency slope in IPAH

RDW showed highly significant (p< 0.001), albeit relatively weak, positive associations with age, BNP, CRP and creatinine levels (Table 2.15, Figure 2.9). No significant associations were found with cardiac structure (Table 2.16). RDW showed associations with haemodynamic variables including RAPm, PAPm, and PWPm. (Table 2.17). In contrast, a highly significant negative association was found between RDW and 6MWD (Table 2.18, Figure 2.10.). RDW also exhibited a negative correlation with several CPET parameters, peak VO2 (mls/kg/min), VE/VCO2 slope, and oxygen uptake efficiency slope (OUES) (Table 2.18). This suggests that increased RDW and iron deficiency relates to a reduced 6MWD and impaired exercise capacity in IPAH.

Figure 2.9. Scatter plot demonstrating the relationship of age (years) with RDW, red cell distribution width (%).

Gender Age BNP CRP Creatinine RDW r .134* .380** .298** .272** .222** p 0.026 <0.001 <0.001 <0.001 <0.001 n 277 277 135 264 273

Table 2.15. Relationship of RDW with gender, age, and serum BNP, CRP, creatinine. RDW, red cell distribution width (%); ferritin (µg/L); TSAT, transferrin saturation (%); Hb, haemoglobin (g/L); age (years); BNP, brain natriuretic peptide (pmol/L); CRP, c-reactive protein (mg/L); creatinine (µmol/L). r, rho; p, significance; n, number.

RVEDV RVSV RVEF LVEF RDW r 0.109 0.124 0.037 0.052 p 0.375 0.309 0.763 0.67 n 69 69 69 70

Table 2.16. Relationship of RDW with cardiac magnetic resonance imaging. RDW, red cell distribution width (%);RVEDV, right ventricular end diastolic volume (ml); RVSV, right ventricular stroke volume (ml); RVEF, right ventricular ejection fraction (%); LVEF, left ventricular ejection fraction (%); r, rho; p, significance; n, number.

RAPm PAPm PWPm PVR CO CI RDW r .204* -.189* .294** -0.138 -0.002 -0.048 p 0.019 0.026 0.001 0.115 0.979 0.641 n 133 140 134 131 132 97

Table 2.17. Relationship of RDW with cardiopulmonary haemodynamic function. RDW, red cell distribution width (%);); RAPm, mean right atrial pressure (mmHg); PAPm, mean pulmonary arterial pressure (mmHg); PWPm, mean pulmonary wedge pressure (mmHg); PVR, pulmonary vascular resistance (WU); PVRI, pulmonary vascular resistance index (WU.m2); CO, cardiac output (L/min); CI, cardiac index (L/min/m2). r, rho; p, significance; n, number.

Peak VO2 Peak VO2 (% 6MWD (ml/kg/min) predicted) VE/VCO2 OUES RDW r -.625** -.434** -0.06 .393** -.349** p <0.001 <0.001 0.627 0.001 0.005 n 103 67 67 66 63

Table 2.18. Relationship of RDW with exercise.. RDW, red cell distribution width (%);6MWD, six minute walk distance (m); peak VO2 , peak oxygen consumption in one minute (mls/Kg/min);

peak VO2, maximum oxygen consumption as percentage predicted (% predicted); VE/VCO2, ventilatory equivalent for carbon dioxide units; r, rho; p, significance; n, number.

Figure 2.10. The relationship of RDW with 6MWD. RDW, red cell distribution width (%); 6MWD, six minute walk distance (m); r, rho; p, significance.

Multiple linear regression analysis was used to assess whether RDW was independently associated with cardiopulmonary haemodynamics or exercise capacity in patients with IPAH. Age and gender are known to be associated with IPAH severity (Swift, Capener et al. 2015) and were therefore controlled for in the model. CRP served as a marker of inflammation. RDW did not independently predict invasive haemodynamics in the IPAH/HPAH cohort. Age was the independent predictor for mean pulmonary artery pressure (p<0.001) and mean pulmonary wedge pressure (p=0.025). CRP was the independent marker for mean right atrial pressure (p=0.013). In exercise, gender was the independent variable with regard to peak VO2 (p=0.006) and VE/VCO2 slope (p=0.006). RDW however was an independent marker of 6MWD (p<0.001) and OUES (p=0.002). Table 2.19.).

IPAH Coefficient 95.0% CI Dependent Final Model Beta Sig. Lower Upper RAPm CRP 0.228 0.013 0.049 0.404 PAPm Age -0.408 <0.001 -0.641 -0.267 PWPm Age 0.208 0.025 0.03 0.449 6MWD RDW -0.374 <0.001 -0.536 -0.212 Age -0.406 <0.001 -0.616 -0.261 Peak VO2 Gender -0.352 0.006 -0.577 -0.099 VE/VCO2 Age 0.462 0.003 0.158 0.740 OUES RDW -0.225 0.002 -0.396 -0.087 CRP -0.142 0.047 -0.315 -0.002

Variables in each model Age Gender RDW CRP

Table 2.19. Independent associations between invasive haemodynamics and exercise. Multiple linear regression modelling showing statistically significant independent predictors of RAPm, mean right atrial pressure (mmHg); mean pulmonary arterial pressure (mmHg); PWPm, mean pulmonary wedge pressure (mmHg); 6MWD, six minute walk distance (m); Peak VO2, maximum

oxygen consumption in one minute (mls/kg/min and %predicted); VE/VCO2, ventilatory equivalent for carbon dioxide units; OUES, OUE slope, oxygen uptake efficiency slope.

2.3.6.1.2 Survival On univariate Cox regression analysis, apart from mean right atrial pressure, no other CMR or invasive cardiopulmonary haemodynamic parameter was found to be associated with overall mortality in this cohort (Table 2.20). In contrast, 6MWD, VO2 peak (% predicted and per kg) were significantly associated with mortality (Table 2.20). Age, gender, haemoglobin, RDW, BNP, CRP and creatinine levels were also all identified as significant (p<0.05) prognostic markers (Table 2.20).

Given the nature of the data in incident patients, a multivariate Cox regression surivial analysis could not be undertaken including all the relevant variables, since the datasets were incomplete. This resulted in only 30 patients having complete data at baseline which is an insufficient number in which to carry out this analysis. Therefore, all the parameters with p<0.05 underwent ‘head-head’ Cox regression analysis with RDW to look for independent associations with overall survival (Table 2.21). RDW predicted survival independent of age, gender and haemoglobin in IPAH. RDW, iron and transferrin saturation were all independent predictors of survival when compared against each other. RDW was also independent of CRP and BNP. RDW did not provide independent prognostic information when compared with exercise parameters such as 6 minute walk distance and peak VO2. This would suggest that RDW is elevated in patients with impaired exercise capacity and is consistent with the notion that iron deficiency may contribute to the limitation of exercise in IPAH.

Patients stratified by the median RDW (14.8%) had a mean survival difference of 2.9 years (Figure 2.11; Table 2.22). RDW showed very significant correlations with survival on Kaplan-Meier analysis (p=0.0000006). Iron and TSAT also showing positive signal (p=0.03, p=0.01) respectively.

Died at 3Y N Positive 82 Negative 195 Missing 175

RDW (%) 95% CI Area Sig. (p) Lower Upper 0.701 <0.001 0.639 0.764

Figure 2.105 ROC models for 3 year survival with 6MWD, RDW and cardiac output in IPAH.

RDW, red cell distribution width (%).

95.0% CI p Exp (B) Lower Upper Gender 0.005 1.474 1.125 1.932 Age <0.001 1.88 1.618 2.184 BSA 0.391 0.815 0.511 1.3 RVEDV 0.79 1.056 0.709 1.572 RVSV 0.409 1.005 0.993 1.018 RVEF 0.855 0.966 0.666 1.401 LVEF 0.73 0.943 0.673 1.319 RAPm 0.008 1.36 1.085 1.706 PAPm 0.21 0.854 0.666 1.093 PWPm 0.296 1.137 0.894 1.445 PVR 0.964 0.994 0.765 1.291 PVRI 0.599 0.887 0.567 1.387 CO 0.266 0.866 0.673 1.116 CI 0.734 0.927 0.597 1.438 6MWD <0.001 0.528 0.409 0.68 Peak VO2 (ml/Kg/min) 0.006 0.823 0.715 0.946 Peak VO2 (% pred) 0.019 0.961 0.993 VE/VCO2 <0.001 1.028 1.31 2.622 OUES 0.003 0.352 0.178 0.695 RDW <0.001 1.534 1.323 1.777 BNP <0.001 1.951 1.456 2.614 CRP <0.001 1.398 1.197 1.632 Creatinine <0.001 1.517 1.331 1.728

Table 2.20. Cox regression analysis, showing all parameters (segregated by modality) as predictors of overall mortality in patients with IPAH. Age (years); BSA, body surface area (m2); RVEDV, right ventricular end diastolic volume (mls); RVSV, right ventricular stroke volume (mls); RVEF, right ventricular ejection fraction (%); LVEF, left ventricular ejection fraction (%); RAPm, mean right atrial pressure (mmHg); PAPm, mean pulmonary artery pressure (mmHg); PWPm, mean pulmonary wedge pressure (mmHg); PVR, pulmonary vascular resistance (WU); PVRI, pulmonary vascular resistance index (WU.m2); CO, cardiac output (L/min); CI, cardiac index 2 (L/min/m ); 6MWD, six minute walk distance (m); PeakVO2, maximum oxygen consumption in one minute (mls/kg/min ); Peak VO2, % predicted; ; OUES, oxygen uptake efficiency slope;

VE/VCO2, ventilatory equivalent for carbon dioxide; RDW, red cell distribution width (%);BNP, brain natriuretic peptide (ng/L); CRP, c-reactive protein (mg/L); Creatinine (mg/L); p, significance.

95.0% CI p Exp(B) Lower Upper RDW <0.001 1.492 1.287 1.731 Gender 0.006 1.652 1.158 2.358 RDW 0.001 1.308 1.112 1.539 Age <0.001 1.857 1.486 2.32 RDW 0.027 1.34 1.034 1.736 RAPm 0.312 1.16 0.87 1.548 RDW 0.696 1.064 0.78 1.45 6MWD 0.001 0.512 0.35 0.749 RDW 0.983 0.366 0.186 0.721 Peak VO2 (mls/kg/ming) 0.004 1.005 0.651 1.551 RDW 0.528 1.142 0.756 1.726 Peak VO2 (% predicted) 0.017 0.526 0.310 0.892 RDW 0.376 1.226 0.781 1.923 VE/VCO2 0.002 1.86 1.244 2.781 RDW 0.794 1.065 0.663 1.713 OUES 0.02 0.41 0.194 0.868 RDW <0.001 1.445 1.206 1.731 Ferritin 0.154 1.112 0.961 1.286 RDW <0.001 1.491 1.263 1.76 Iron 0.446 0.923 0.751 1.134 RDW <0.001 1.538 1.302 1.817 TSAT 0.909 0.988 0.808 1.209 RDW <0.001 1.504 1.278 1.77 Hb 0.465 0.937 0.786 1.116 RDW <0.001 1.481 1.265 1.735 CRP 0.007 1.26 1.065 1.49 RDW 0.003 1.49 1.15 1.93 BNP <0.001 1.766 1.284 2.429 RDW <0.001 1.553 1.319 1.828 Creatinine <0.001 1.469 1.278 1.689

Table 2.21. Cox modelling, assessing if RDW is an independent predictor of survival against other variables as a univariate analysis. . RDW, red cell distribution width (%); Age (years);

RAPm, mean right atrial pressure (mmHg); 6MWD, six minute walk distance (m); peak VO2, maximum oxygen consumption in one minute (mls/kg/min); peak VO2, (% predicted); OUES,

oxygen uptake efficiency slope; VE/VCO2, ventilatory equivalent for carbon dioxide; ferritin (µg/L); iron (µmol/L); TSAT, transferrin saturation (%); Hb, haemoglobin (g/L); CRP, c-reactive protein (mg/L); BNP, brain natriuretic peptide (ng/L); Creatinine (mg/L); p, significance.

N at risk

Years 0 1 2 3 4 5 6 7 8 9 10 11

RDW <14.8% 139 125 110 90 71 58 50 39 23 17 12 3 RDW >14.8% 136 99 75 57 42 28 15 14 8 3 3 2

Figure 2.11. Kaplan-Meier curves stratified by median RDW (14.8%) in patients with Idiopathic PAH. RDW, red cell distribution width (%).

N at risk Years 0 1 2 3 4 5 6 7 8 9 10 11 Iron <12.0 166 129 97 79 60 45 31 23 13 7 5 1 Iron >12.0 146 126 107 84 68 55 47 36 23 16 11 5

N at risk

Years 0 1 2 3 4 5 6 7 8 9 10 11 TSAT <20% 164 128 96 77 60 48 35 23 12 6 5 1 TSAT >20% 143 123 105 83 65 50 41 35 24 17 11 5

Figures 2.12 Kaplan-Meier estimates stratified by median iron (12.0 µmol/L) (A) and transferrin saturation (20%) (B) in patients with Idiopathic PAH. Iron (µmol/L); TSAT, transferrin saturation (%).

Mean survival time 95% CI Years Lower Upper RDW <14.8% 7.639 6.857 8.420 RDW >14.8% 4.746 3.920 5.573 Overall 6.342 5.734 6.949 Iron <12.0 µmol/L 5.556 4.793 6.319 Iron >12.0 µmol/L 7.210 6.404 8.017 Overall 6.353 5.788 6.918 TSAT <20% 5.617 4.851 6.384 TSAT >20% 7.142 6.321 7.963 Overall 6.365 5.795 6.935

Table 2.22. Mean survival time based on median cut-offs for RDW, iron and transferrin saturation, in patients with Idiopathic PAH. RDW, red cell distribution (%); iron (µmol/L);

TSAT, transferrin saturation (%).

100

2.3.6.2 Connective tissue disease PAH

2.3.6.2.1 Cardiopulmonary haemodynamics and exercise physiology RDW is independently associated with RV ejection function, but not exercise capacity in CTD-PAH

RDW showed a moderate positive association with BNP levels (Table 2.23) and negative association RVEF (Table 2.24). Haemoglobin levels showed a significant positive correlation with mPAP and PVR and negative correlation with cardiac output and RVEF (Table 2.24). Interestingly, no correlation was apparent between RDW and CRP levels (Table 2.23) and no significant correlations were observed between iron deficiency and measures of exercise in this PAH subgroup (Table 2.25, Table 2.26).

Multiple linear regression (Table 2.27) reveals RDW remains an independent marker for RVEF when correcting for age, gender and crp (inflammation).

Gender Age BNP CRP Creatinine RDW r 0.014 0.191 .403** 0.196 0.14 p 0.899 0.073 0.004 0.073 0.192 n 89 89 50 85 88

Table 2.23. Correlation of RDW with gender, age and blood biomarkers. RDW, red cell width distribution (%);age (years); BNP, brain natriuretic peptide (ng/L); CRP, c-reactive protein (mg/L); Creatinine (mg/L); p, significance. r, correlations; n, number

RVEDV RVSV RVEF LVEF RDW r 0.113 -0.403 -.556* -0.007 p 0.688 0.137 0.031 0.98 n 15 15 15 15

101

Table 2.24. Correlation of iron and prognostic markers of cardiac structure and function. Bivariate correlations of haemoglobin and iron parameters against cardiac MRI, invasive heamodynamics, and exercise. RDW, red cell width distribution (%); ferritin (µg/L); iron (µmol/L); TSAT, transferrin saturation (%); Hb, haemoglobin (g/L); RVEDV, right ventricular end diastolic volume (ml); RVSV, right ventricular stroke volume (ml); RVEF, right ventricular ejection fraction (%); LVEF, left ventricular ejection fraction (%). p, significance. r, correlations; n, number

RAPm PAPm PWPm PVR CO CI RDW r 0.106 0.163 -0.026 0.096 -0.055 -0.252 p 0.492 0.292 0.869 0.544 0.727 0.246 n 44 44 43 42 42 23

Table 2.25. Correlation of RDW and invasive haemodynamics. RDW, red cell width distribution (%); ferritin (µg/L); RAPm, mean right atrial pressure (mmHg); PAPm, mean pulmonary artery pressure (mmHg); PWPm, mean pulmonary wedge pressure (mmHg); PVR, pulmonary vascular resistance (WU); CO, cardiac output (L/min); CI, cardiac index (L/min/m2). p, significance. r, correlations; n, number

Peak VO2 Peak VO2 OUES 6MWD (ml/kg/min) (% predicted) VE/VCO2 RDW r -0.294 -0.267 -0.24 0.245 -0.249 p 0.108 0.218 0.27 0.272 0.252 n 31 23 23 22 23

Table 2.26. Correlation of RDW and exercise. RDW, red cell width distribution (%);6MWD, six

minute walk distance (m); RER max, maximum respiratory exchange ratio; VO2 max, maximum

oxygen consumption in one minute (mls/kg/min); Load max (watts); PETCO2 max, maximum partial pressure of end-tidal carbon dioxide (kPa); HR, heart rate (bpm); O2/HR, oxygen pulse (mls/beat); OUES, oxygen uptake efficiency slope; VE/VCO , ventilatory equivalent for carbon 2 dioxide; OUES, oxygen uptake efficiency slope; r, rho, correlation; p, significance; n, number.

102

RVEF Final Coefficient 95% CI Model Beta Sig. Lower Upper (Constant) 0.026 28.897 353.065 Age 0.23 0.347 -19.466 49.894 Gender -0.585 0.06 -21.815 0.55 RDW -0.607 0.036 -244.508 -10.414 CRP -0.52 0.082 -32.713 2.332

Table 2.27. Independent associations with RVEF, right ventricular ejection fraction (%). Multiple linear regression assesseing for independent predictors of RVEF. Age, years; Gender, male or female; RDW, red cell distribution width (%); CRP, c-reactive protein, (mg/L).

103

2.3.6.2.2 Survival On univariate Cox regression survival analysis, CMR and cardiopulmonary haemodynamic parameters were poor predictors of overall survival apart from mean right atrial pressure (Table 2.28 A). Age, RDW, BNP and CRP levels were identified as significant prognostic markers (Table 2.28 A). In common with IPAH, multivariate Cox regression could not be undertaken. On head-to-head, RDW was independent of Hgb, CRP and gender, but not age, as a predictor of survival. ROC models indicate that RDW, CRP and age are all significant predictors of survival at 3 years, with CRP exhibiting the largest area under the curve (Figure 2.13).

Kaplan-Meier analysis demonstrated that patients stratified by either median RDW or CRP level had marked differences in their cumulative survival, patients with elevated CRP (> 7mg/L) exhibiting a marked difference in mean survival of 3.8 years (Figure 2.14, Table 2.31). In contrast, stratification by median iron and transferrin saturation showed no significant difference in survival (Figure 2.15, Table 2.31).

95.0% CI

A Sig. Exp(B) Lower Upper Gender 0.428 0.753 0.373 1.521 Age <0.001 2.422 1.734 3.384 BSA 0.282 1.494 0.719 3.105 RVEF 0.659 0.84 0.387 1.824 LVEF 0.976 1.002 0.489 2.093 RAPm 0.003 1.799 1.219 2.655 PAPm 0.161 1.022 0.991 1.055 PWPm 0.768 1.015 0.92 1.12 PVR 0.119 1.048 0.988 1.112 CO 0.073 0.045 0.002 1.332 CI 0.186 0.494 0.173 1.405 6MWD 0.01 0.507 0.302 0.852 Peak VO2 (ml/Kg/min) 0.038 0.402 0.17 0.952 Peak VO2 (% predicted) 0.317 0.647 0.276 1.517 OUES 0.544 1.166 0.71 1.914

VE/VCO2 0.061 1.92 0.97 3.83 RDW 0.002 1.391 1.128 1.716 BNP <0.001 2.543 1.768 3.66 CRP <0.001 1.787 1.384 2.308 Creatinine 0.751 0.962 0.756 1.223

104

95.0% CI

B p Exp(B) Lower Upper RDW 0.003 1.148 1.047 1.259 Gender 0.297 0.578 0.206 1.622 RDW 0.183 1.171 0.928 1.476 Age 0.005 1.846 1.202 2.835 RDW 0.012 2.147 1.185 3.889 RAPm 0.01 1.699 1.138 2.538 RDW 0.1 1.735 0.901 3.343 6MWD 0.212 0.654 0.335 1.274 RDW 0.619 1.14 0.68 1.911 BNP 0.003 2.107 1.289 3.444 RDW 0.008 1.389 1.091 1.769 CRP <0.001 1.714 1.325 2.218

Table 2.28 (A) Cox regression analysis, showing all parameters (segregated by modality) as predictors of overall mortality in patients with PAH-CTD. (B) Cox modelling, assessing if RDW is an independent predictor of survival. Gender, male or female; Age (years); BSA, body surface area (m2); RVEF, right ventricular ejection fraction (%); LVEF, left ventricular ejection fraction (%); RAPm, mean right atrial pressure (mmHg); PAPm, mean pulmonary artery pressure (mmHg);

PWPm, mean pulmonary wedge pressure (mmHg); PVR, pulmonary vascular resistance (WU); CO,

cardiac output (L/min); CI, cardiac index (L/min/m2); 6MWD, six minute walk distance (m); Peak

VO2,maximum oxygen consumption in one minute (mls/kg/min) and percentage predicted (%

predicted)OUES, oxygen uptake efficiency slope (OUES); VE/VCO2, ventilatory equivalent for carbon dioxide; RDW, red cell distribution width (%);BNP, brain natriuretic peptide (ng/L); CRP, c- reactive protein (mg/L); Creatinine (mg/L); p, significance.

105

Died at 3Y N Died at 3Y N Positive 33 Positive 32 Negative 56 Negative 58 Missing 40 Missing 39

95% CI Area Sig Lower Upper RDW 0.662 0.011 0.543 0.781 CRP 0.729 <0.001 0.619 0.839 Died at 3Y N Positive 48 Age 0.675 0.001 0.58 0.77 Negative 81 Missing 0

Figure 2.13. ROC models showing RDW, CRP and age as predictors of 3 year survival in PAH-CTD. Age (years); RDW, red cell distribution width (%); CRP, c-reactive protein (mg/L).

106

N at risk N at risk Years 0 1 2 3 4 5 6 7 8 9 Years 0 1 2 3 4 5 6 7 8 9 RDW<16% 45 42 37 29 21 16 13 9 6 5 CRP<7mg/L 46 40 34 28 23 20 14 10 6 5 RDW>16% 42 26 15 9 6 5 3 3 2 2 CRP>7mg/L 42 28 20 13 7 3 2 2 2 2

Figure 2.14. Kaplan-Meier estimates stratified by median RDW (16%) and CRP (7 mg/L). RDW, red cell distribution width (%); CRP, c-reactive protein (mg/L).

107

N at risk N at risk Years 0 1 2 3 4 5 6 7 8 9 Years 0 1 2 3 4 5 6 7 8 9 Iron <10 µg/L 57 45 30 21 15 9 8 6 5 4 TSAT <16% 22 19 11 7 11 7 6 4 3 3 Iron >10µg/L 49 40 33 26 20 19 10 7 4 4 TSAT >16% 51 42 36 29 22 19 11 7 5 5

Figure 2.15. Kaplan-Meier estimates stratified by median iron (10 µg/L) and transferrin

saturation (16.0 %). Iron (µmol/L); TSAT, transferrin saturation (%).

108

Mean survival

95% CI

Years Lower Upper

RDW <16% 6.001 4.945 7.057

RDW >16% 3.437 2.216 4.658

Overall 4.847 3.992 5.702

CRP <7 mg/L 6.621 5.582 7.66

CRP >7 mg/L 2.856 1.953 3.758

Overall 4.948 4.102 5.794

Iron <10 µmol/L 4.348 3.295 5.402

Iron >10 µmol/L 5.245 4.218 6.273

Overall 4.841 4.084 5.598

TSAT <16% 4.01 2.979 5.04

TSAT >16% 5.339 4.291 6.386

Overall 4.704 3.946 5.463

Table 2.29. Mean survival time following stratification by median cut-offs for RDW, CRP, iron and transferrin saturation. RDW, red cell distribution width (%); CRP, c-reactive protein (mg/L); iron (µmol/L); TSAT, transferrin saturation.

109

2.3.6.3 Congenital heart disease PAH

2.3.6.3.1 Cardiopulmonary haemodynamics and exercise physiology RDW is independently associated with absolute peak VO2 in PAH-CHD

Haemoglobin (Hgb) levels were included in the analysis for congenital heart disease as these patients often have high Hgb levels due to chronic hypoxia (Diller et al. Eur Heart J 2006). RDW exhibited a significant positive association with CRP levels in this cohort (Table 2.30). RDW showed a negative correlation with cardiac index (Table 2.32). Haemoglobin showed significant associations with age, gender and BNP levels (Table 2.30,) and negative associations with CMR measures of right ventricular function (Table 2.31). Moderate associations were also observed for haemoglobin with cardiopulmonary haemodynamics, 6MWD and measurements of exercise capacity (Tables 2.32 and 2.33).

RDW (p=0.012) and age (p=0.017) were independent predictors for Peak VO2 (ml/kg/min) but no single variable showed strong associations for mean right atrial pressure or cardiac index on multiple linear regression analysis (Table 2.34).

110

Gender Age BNP CRP Creatinine RDW r .240* -0.049 0.062 .423** .221* p 0.016 0.625 0.665 <0.001 0.028 n 101 101 51 97 98 Hb r .327** .548** .428** 0.028 0.024 p 0.001 <0.001 0.002 0.788 0.818 n 101 101 51 97 98

Table 2.30. Associations between RDW, haemoglobin with gender, age, and blood biomarkers prognostic in PAH. RDW, red cell distribution width (%); Hb, haemoglobin (g/L); age (years); BNP, brain natriuretic peptide (ng/L); CRP, c-reactive protein (mg/L); Creatinine (mg/L); r, rho; p, significance; n (number).

RVEDV RVESV RVEF LVEDV LVESV LVEF RDW r 0.024 0.227 -0.354 0.055 0.022 0.138 p 0.923 0.349 0.137 0.829 0.932 0.585 n 19 19 19 18 18 18 Hgb r -.628** -0.347 -0.345 -0.063 -0.055 -0.037 p 0.004 0.146 0.148 0.804 0.829 0.883 n 19 19 19 18 18 18

Table 2.31. Associations between RDW and haemoglobin with cardiac magnetic resonance imaging. RDW, red cell distribution width (%) Hb, haemoglobin (g/L); RVEDV, right ventricular end diastolic volume (ml); RVSV, right ventricular stroke volume (ml); RVEF, right ventricular ejection fraction (%); RV Output, right ventricular output (L/min); LVEF, left ventricular ejection fraction (%); LV Output, left ventricular output (L/min); r, rho; p, significance; n (number).

111

RAPm PAPm PCWm PVR CO CI RDW r .495* 0.275 0.209 0.244 -0.329 -.697** p 0.019 0.215 0.376 0.274 0.135 0.006 n 22 22 20 22 22 14 Hgb r -0.233 .467* -0.266 .418* 0.098 0.229 p 0.285 0.025 0.244 0.047 0.657 0.431 n 23 23 21 23 23 14

Table 2.32. Associations between RDW, haemoglobin with invasive haemodynamics. RDW, red cell distribution width Hb, haemoglobin (g/L); RAPm, mean right atrial pressure (mmHg); PAPm, mean pulmonary artery pressure (mmHg); PWPm, mean pulmonary wedge pressure (mmHg); PVR, pulmonary vascular resistance (WU);; CO, cardiac output (L/min); CI, cardiac index (L/min/m2); r, rho; p, significance; n (number).

6MWD Peak VO2 ml/kg/min Peak VO2 % predicted VE/VCO2 OUES RDW r 0.224 -.684** -0.357 0.256 -0.259 p 0.176 <0.001 0.103 0.25 0.244 n 38 22 22 22 22 Hgb r .431** 0.056 -.648** .545** -0.076 p 0.004 0.804 0.001 0.009 0.737 n 42 22 22 22 22

Table 2.33. Associations between RDW, haemoglobin with exercise. RDW, red cell distribution width (%);Hb, haemoglobin (g/L); 6MWD, six minute walk distance (m);; Peak

VO2 , maximum oxygen consumption in one minute (mls/kg/min), and percentage predicted

(% predicted); OUES, oxygen uptake efficiency slope (OUES); VE/VCO2, ventilatory equivalent for carbon dioxide; r, rho; p, significance; n (number).

112

Dependent Final Model Beta Sig. Lower Upper (no independent RAPm variable) (no independent CI variable) Peak VO2 ml/Kg/min Age -0.511 0.017 -1.255 -0.141 RDW -0.566 0.012 -1.486 -0.213

Variables in each model Age Gender RDW CRP

Table 2.34 Independent associations with right atrial pressure, cardiac index, and peak VO2.

Multiple linear regression analysis correcting for age, years; Gender, male or female; RDW, red cell distribution width (%); CRP, c-reactive protein, (mg/L).

2.3.6.3.2 Survival Univariate Cox regression survival analysis indicated that RVEF was the only CMR or invasive cardiopulmonary haemodynamic parameter that predicted overall survival predictor in this cohort (Table 2.35). (p= 0.028), along with 6MWD in exercise (p= 0.042). RDW exhibited a highly significant association with survival (p< 0.001), whereas 6MWD, BNP, CRP and creatinine were less significant (Table 2.35).

A head-to-head analysis of RDW as an independent predictor of overall survival indicated that RDW is independent of age, gender, CRP, creatinine, haemoglobin and 6MWD, but not BNP and RVEF (Table 2.36). The Kaplan-Meier analysis, using a median RDW value (14.8%) to stratify patients, showed that a raised RDW was associated with reduction in mean survival of 2.4 years (Figure 2.16).

113

95.0% CI p Exp(B) Lower Upper Gender 0.808 1.072 0.611 1.881 Age 0.063 1.308 0.986 1.736 RVEDV 0.521 0.774 0.353 1.695 RVSV 0.065 0.465 0.206 1.049 RVEF 0.028 0.409 0.184 0.909 LVEF 0.349 0.7 0.331 1.477 RAPm 0.5 1.311 0.597 2.882 PAPm 0.38 1.523 0.595 3.901 PWPm 0.961 1.024 0.4 2.622 PVR 0.267 1.81 0.634 5.163 CO 0.435 0.575 0.143 2.31 CI 0.324 0.397 0.063 2.488 6MWD 0.011 0.657 0.476 0.907 RER max 0.567 0.823 0.422 1.604 Peak VO2 (mls/kg/min) 0.085 0.685 0.446 1.054 Peak VO2 (% predicted) 0.078 0.931 0.859 1.008 OUES max 0.699 1.159 0.548 2.452

VE/VCO2 0.784 1.089 0.592 2.002 RDW 0.006 1.611 1.146 2.266 Hb 0.539 1.131 0.764 1.674 BNP 0.006 3.313 1.419 7.736 CRP 0.167 1.34 0.885 2.029 Creatinine 0.011 1.672 1.124 2.488

Table 2.35. Cox regression analysis, showing all parameters (segregated by modality) as predictors of overall mortality in patients with PAH-CHD. Age (years); RVEF, right ventricular ejection fraction (%); LVEF, left ventricular ejection fraction (%); RAPm, mean right atrial pressure (mmHg); PAPm, mean pulmonary artery pressure (mmHg); PWPm, mean pulmonary wedge pressure (mmHg); PVR, pulmonary vascular resistance (WU); CO, cardiac output (L/min); CI,

2 cardiac index (L/min/m ); 6MWD, six minute walk distance (m); PeakVO2 , maximum oxygen consumption in one minute (mls/kg/min) and percentage predicted (% predicted); OUES, oxygen

uptake efficiency slope ; VE/VCO2, ventilatory equivalent for carbon dioxide; RDW, red cell distribution width (%);Hb, haemoglobin (g/L); BNP, brain natriuretic peptide (ng/L); CRP, c- reactive protein (mg/L); Creatinine (mg/L); p, significance.

114

95.0% CI Sig. Exp(B) Lower Upper RDW <0.001 1.944 1.354 2.792 Gender 0.12 2.026 0.833 4.928 RDW <0.001 2.069 1.438 2.977 Age 0.281 1.36 0.777 2.382 RDW 0.341 1.39 0.706 2.735 RVEF 0.071 0.459 0.197 1.069 RDW 0.004 2.825 1.39 5.74 6MWD 0.196 1.927 0.714 5.206 RDW <0.001 1.935 1.365 2.743 Hb 0.957 0.989 0.66 1.483 RDW 0.785 1.195 0.331 4.316 BNP 0.009 4.2 1.43 12.334 RDW 0.002 1.827 1.257 2.654 CRP 0.229 1.402 0.808 2.435 RDW 0.001 1.88 1.313 2.692 Creatinine 0.197 1.352 0.855 2.137

Table 2.36. Cox modelling, assessing if RDW is an independent predictor of survival in PAH- CHD. RDW, red cell distribution width (%); Age (years); RVEF, right ventricular ejection fraction (%); 6MWD, six minute walk distance (m); Hb, haemoglobin (g/L); BNP, brain natriuretic peptide (ng/L); CRP, c-reactive protein (mg/L); Creatinine (mg/L).

115

N at risk Years 0 1 2 3 4 5 6 7 8 9 10 11 RDW<14.8% 42 39 36 33 28 26 21 17 13 12 9 3 RDW>14.8% 41 36 29 25 19 15 11 6 5 2 2 1

Mean 95% CI

Years Lower Upper

RDW <14.8% 10.046 9.045 11.046 RDW>14.8% 7.659 6 9.318

Overall 8.926 7.948 9.903

Figure 2.16. Kaplan-Meier estimates stratified by median RDW (14.8%). RDW, red cell

distribution width (%).

116

2.3.6.4 Chronic thromboembolic pulmonary hypertension

2.3.6.4.1 Cardiopulmonary haemodynamics and exercise physiology RDW is independently associated with multiple markers of cardiac function, haemodynamics and exercise capacity in CTEPH

RDW showed a positive association with gender, but not age (Table 2.37). RDW exhibited a consistent negative association with RVEF (Table 2.38, Figure 2.17), cardiopulmonary haemodynamics (Table 2.39), 6MWD and measurements of exercise capacity (Table 2.40).

Multiple linear regression analysis was conducted to assess which factors independantly associate with cardiac MRI, cardiopulmonary haemodynamics and exercise.. RDW consistently appeared in the final models, demonstrating an independent correlation with prognostic markers of pulmonary hypertension when correcting for age, gender, and inflammation (CRP). (Table 2.39). Gender Age BNP CRP Creatinine RDW r -.115* 0.089 .324** .362** -0.064 p 0.033 0.099 <0.001 <0.001 0.238 n 347 347 184 315 341

Table 2.37. Correlations between RDW with gender, age and blood biomarkers. RDW, red cell distribution width (%);Age (years); BSA, body surface area (m2); BNP, brain natriuretic peptide (ng/L); CRP, c-reactive protein (mg/L); Creatinine (mg/L); p, significance.

RVEDV RVSV RVEF LVEF RDW r .160* -.199* -.332** -0.136 p 0.049 0.014 <0.001 0.093 n 152 153 153 154

Table 2.38. Correlations between RDW with cardiac magnetic resonance imaging parameters. RDW, red cell distribution width (%);RVEDV, right ventricular end diastolic volume (ls); RVSV, right ventricular stroke volume (mls); RVEF, right ventricular ejection fraction (%); LVEF, left ventricular ejection fraction (%); p, significance.

117

RAPm PAPm PWPm PVR CO CI RDW r .253** 0.103 -.172** .238** -.307** -.349** p <0.001 0.109 0.008 <0.001 <0.001 <0.001 n 235 244 239 237 237 169

Table 2.39. Correlations between RDW, with invasive haemodynamics. RDW, red cell distribution width (%);RAPm, mean right atrial pressure (mmHg); PAPm, mean pulmonary artery pressure (mmHg); PWPm, mean pulmonary wedge pressure (mmHg); PVR, pulmonary vascular resistance (WU); CO, cardiac output (L/min); CI, cardiac index (L/min/m2); p, significance.

Peak VO2 Peak VO2 6MWD (ml/Kg/min) (% predicted) OUES VE/VCO2 - RDW r -.382** -.397** -.253** .327** .329** p <0.001 <0.001 0.006 <0.001 <0.001 n 119 115 115 115 115

Table 2.40. Correlations between RDW with exercise. RDW, red cell distribution width

(%);6MWD, six minute walk distance (m); Peak VO2, maximum oxygen consumption in one minute (mls/kg/min) and percentage predicted (% predicted); OUES, oxygen uptake efficiency

slope; VE/VCO2, ventilatory equivalent for carbon dioxide; p, significance.

118

CTEPH Coefficient 95.0% CI Upper Dependent Final Model Beta Sig. Lower Bound Bound RVEDV (Constant) <0.001 -1.701 -0.702 Gender 0.413 <0.001 0.52 1.148 RVEF RDW -0.217 0.015 -0.387 -0.043 (Constant) <0.001 -1.718 -0.719 RVSV Gender 0.418 <0.001 0.529 1.157 LVEF (Constant) <0.001 0.447 1.46 Gender -0.323 <0.001 -0.979 -0.343 RAPm Gender 0.137 0.037 0.018 0.55 RDW 0.168 0.015 0.034 0.312 CRP 0.202 0.003 0.072 0.357 PWPm RDW -0.176 0.012 -0.317 -0.04 PVR RDW 0.171 0.015 0.033 0.305 CO Age -0.217 0.001 -0.343 -0.091 RDW -0.244 <0.001 -0.371 -0.11 CI RDW -0.275 0.001 -0.436 -0.108 6MWD (Constant) 0.001 -1.433 -0.401

Age -0.202 0.015 -0.402 -0.044 Gender 0.277 0.001 0.233 0.892 RDW -0.264 0.002 -0.456 -0.105 CRP -0.207 0.014 -0.429 -0.049 Peak VO2 (ml/Kg/min) (Constant) <0.001 -1.577 -0.493

Age -0.356 <0.001 -0.57 -0.213 Gender 0.301 <0.001 0.285 0.955 RDW -0.241 0.006 -0.517 -0.091 Peak VO2 (% predicted) (Constant) 0.001 0.413 1.591

Age 0.23 0.013 0.055 0.443 Gender -0.35 <0.001 -1.072 -0.343 RDW -0.24 0.012 -0.529 -0.066 VE/VCO2 RDW 0.299 0.003 0.124 0.595 OUES (Constant) 0.011 -1.386 -0.184

Age -0.222 0.017 -0.441 -0.045 Gender 0.222 0.017 0.083 0.826 RDW -0.231 0.017 -0.526 -0.053

119

Variables in each model Age Gender RDW CRP

Table 2.41. Multiple linear regression analysis of individual variables with cardiac magnetic resonance outputs, invasive cardiopulmonary haemodynamics and exercise in CTEPH. Age (years); ; RDW, red cell distribution width (%); CRP, c-reactive protein (mg/L); C RVEDV, right ventricular end diastolic volume (mls);; RVEF, right ventricular ejection fraction (%); RVSV, right ventricular stroke volume (mls/beat); RAPm, mean right atrial pressure (mmHg); PVR, pulmonary vascular resistance (WU); CO, cardiac output (L/min); CI, cardiac index (L/min/m2) 6MWD, six

minute walk distance (m); Peak VO2 , maximum oxygen consumption in one minute as

mls/kg/min and percentage predicted (% predicted); VE/VCO2, ventilatory equivalent for carbon dioxide; OUES, oxygen uptake efficiency slope.

120

2.3.6.4.2 Survival Univariate Cox regression analysis shows that multiple factors are predictors of survival in CTEPH, including gender, age, RVEF, LVEF, PVR and cardiac output (Table 2.42). 6MWD and measurements of exercise also predict survival, as does RDW, BNP, CRP and creatinine levels (Table 2.43). Cox modelling indicates that RDW is an independent predictor of survival when compared ‘head-to- head’ with gender, age, Hgb, BNP, CRP, creatinine , RAPm and PVR (Table 2.44). RDW falls short as an independent predictor of survival when compared with RVEF, cardiac output and 6MWD (Figure 2.44).

ROC analysis indicated that RDW, CRP and age are all independent predictors of overall survival in CTEPH (Table 2.43). RDW, 6MWD and cardiac output associated with mortality three years after diagnosis, with 6MWD exhibiting a larger area under the curve than RDW (Figure 2.17).

Kaplan-meier regression models and curves show a strong difference in cumulative survival between the two groups separated by the median RDW cut off of 14.9% (figure x. p= 0.0000002).

Patients stratified by median RDW (14.9%) showed a significant difference in overall survival with a mean survival difference of 2.9 years (Figure 2.18). Stratification by median iron level (14.0 ) and transferrin saturation (22.0%) also indicated a significant difference in the survival of CTEPH pateints (Figure 2.19)

121

Sig. Exp(B) 95.0% CI Lower Upper Gender 0.029 1.409 1.035 1.918 Age <0.001 1.738 1.439 2.101 BSA 0.33 0.791 0.493 1.268 RVEDV 0.024 1.388 1.044 1.844 RVSV 0.031 1.409 1.032 1.924 RVEF 0.02 0.654 0.458 0.936 LVEF 0.007 0.69 0.525 0.905 RAPm 0.159 1.171 0.94 1.46 PAPm 0.338 0.894 0.712 1.124 PWPm 0.719 1.04 0.84 1.289 PVR 0.003 1.384 1.117 1.715

Cardiac output <0.001 0.533 0.394 0.721 Cardiac index 0.188 0.787 0.55 1.125 6MWD <0.001 0.538 0.388 0.747 Peak VO2 (ml/kg/min) 0.046 0.574 0.334 0.989 Peak VO2 (% predicted) 0.005 0.512 0.320 0.819 OUES 0.016 0.504 0.288 0.88

VE/VCO2 0.01 1.531 1.107 2.117 Hb 0.101 0.868 0.734 1.028 RDW <0.001 1.392 1.198 1.617 Ferritin 0.041 1.259 1.009 1.571 Iron 0.023 0.814 0.681 0.972 TSAT 0.198 0.894 0.754 1.06 BNP <0.001 2.165 1.582 2.963 CRP <0.001 1.441 1.201 1.73 Creatinine 0.006 1.306 1.08 1.579

Table 2.42. Cox regression analysis, showing all parameters (segregated by modality) as 2 predictors of overall mortality in patients with CTEPH. Age (years); BSA, body surface area (m ); RVEF, right ventricular ejection fraction (%); LVEF, left ventricular ejection fraction (%); RAPm, mean right atrial pressure (mmHg); PAPm, mean pulmonary artery pressure (mmHg); PWPm, mean pulmonary wedge pressure (mmHg); PVR, pulmonary vascular resistance (WU); CO, cardiac 2 output (L/min); CI, cardiac index (L/min/m ); 6MWD, six minute walk distance (m); Peak VO2 , maximum oxygen consumption in one minute (mls/kg/min) and % predicted OUES , oxygen uptake efficiency slope; VE/VCO2, ventilatory equivalent for carbon dioxide; RDW, red cell distribution width (%); ferritin (µg/L); iron (µmol/L); TSAT, transferrin saturation (%); Hb, haemoglobin (g/L); BNP, brain natriuretic peptide (ng/L); CRP, c-reactive protein (mg/L); Creatinine (mg/L); p, significance.

122

95.0% CI for Exp(B) Variable Sig. Exp(B) Lower Upper RDW <0.001 1.411 1.217 1.636 Gender 0.02 1.564 1.072 2.282 RDW <0.001 1.465 1.255 1.71 Age <0.001 1.734 1.365 2.201 RDW 0.059 1.294 0.991 1.691 RVEDV 0.032 1.382 1.028 1.857 RDW 0.057 1.296 0.992 1.694 RVSV 0.038 1.412 1.02 1.955 RDW 0.099 1.264 0.957 1.669 RVEF 0.036 0.668 0.458 0.973 RDW 0.115 1.241 0.949 1.622 LVEF 0.016 0.705 0.53 0.937 RDW 0.006 1.337 1.085 1.647 RAPm 0.712 1.046 0.823 1.331 RDW 0.021 1.284 1.038 1.589 PVR 0.006 1.429 1.111 1.839 RDW 0.056 1.236 0.995 1.537 CO 0.001 0.55 0.387 0.783 RDW 0.083 1.31 0.965 1.778 6MWD 0.001 0.5 0.335 0.745 RDW 0.028 1.569 1.051 0.2342 Peak VO2 (mls/kg/min) 0.112 0.626 0.352 1.116 RDW 0.021 1.582 1.07 2.34 Peak VO2 (% 0.04 0.594 0.362 0.976 predicted) RDW 0.054 1.504 0.994 2.276 OUES 0.174 0.657 0.359 1.204 RDW 0.065 1.5 0.976 2.305

VE/VCO2 0.077 1.467 0.959 2.245 RDW <0.001 1.444 1.208 1.725 Hb 0.456 1.081 0.881 1.326 RDW <0.001 1.464 1.235 1.734 Ferritin 0.014 1.328 1.058 1.666 RDW 0.001 1.342 1.135 1.588 Iron 0.481 0.932 0.768 1.133 RDW <0.001 1.512 1.237 1.849 TSAT 0.284 1.131 0.903 1.417 RDW 0.003 1.286 1.088 1.521 CRP 0.004 1.335 1.099 1.621

123

RDW 0.011 1.431 1.084 1.888 BNP <0.001 2.063 1.482 2.87 RDW <0.001 1.444 1.234 1.689 Creatinine 0.002 1.341 1.113 1.617

Table 2.43. Cox modelling, assessing if RDW is an independent predictor of survival in CTEPH. RDW, red cell distribution width (%); Age (years); BSA, body surface area (m2); RVEF, right ventricular ejection fraction (%); LVEF, left ventricular ejection fraction (%); RAPm, mean right atrial pressure (mmHg); PAPm, mean pulmonary artery pressure (mmHg); PWPm, mean pulmonary wedge pressure (mmHg); PVR, pulmonary vascular resistance (WU); CO, cardiac output (L/min); 6MWD, six minute walk distance (m); Peak VO2, maximum oxygen consumption in one minute (mls/kg/min) and percentage predicted (%

predicted); VE/VCO2, ventilatory equivalent for carbon dioxide ; OUES, oxygen uptake efficiency slope (OUES);; ferritin (µg/L); iron (µmol/L); TSAT, transferrin saturation (%);BNP, brain natriuretic peptide (ng/L); CRP, c-reactive protein (mg/L); Creatinine (mg/L).

124

Died at 3Y N Died at 3Y N Positive 25 Positive 66 Negative 126 Negative 281 Missing 331 Missing 135

Area Sig. 95% CI Died at 3Y N Lower Upper Positive 37 6MWD 0.701 0.002 0.592 0.81 Negative 250 RDW 0.703 <0.001 0.639 0.767 Missing 195 Cardiac output 0.677 <0.001 0.576 0.779

Figure 2.17. ROC models for 3 year survival with 6MWD, RDW and cardiac output in CTEPH.

6MWD, six minute walk distance (m); RDW, red cell distribution width (%); cardiac output (L/min)

125

N at risk Years 0 1 2 3 4 5 6 7 8 9 10 11 RDW <14.9% 175 151 139 119 92 74 56 42 28 20 11 5 RDW>14.9% 172 133 107 86 70 54 44 30 22 11 5 1

Mean 95% CI Years Lower Upper RDW <14.9% 9.168 8.478 9.857 RDW >14.9% 6.202 5.452 6.951 Overall 7.729 7.179 8.279

Figure 2.18 Kaplan-Meier estimates stratified by median cut-off for RDW (14.9%). RDW, red cell distribution width (%).

126

N at risk Years 1 2 3 4 5 6 7 8 9 10 11 Iron<14 152 126 100 79 63 48 32 24 13 8 3 Iron>14 149 128 112 93 78 63 46 33 18 9 4

Figure 2.19 Kaplan-Meier estimates stratified by median cut-offs for iron (14.0). Iron (µmol/L)

127

N at risk

Years 1 2 3 4 5 6 7 8 9 10 11

TSAT<22% 145 119 95 74 60 47 33 23 13 7 2

TSAT>22% 155 134 116 97 80 63 44 34 18 10 5

Figure 2.23 Kaplan-Meier estimates stratified by median cut-offs for transferrin saturation (22.0%). TSAT, transferrin saturation (%).

2.4 Summary In this chapter, I initially established that RDW is the single best blood marker for predicting sTfR that we have readily available in day-to-day clinical practice. I have subsequently shown that it predicts iron deficiency, independently to inflammation, and has strong correlations with overall survival in all subgroups of PAH.

128

2.5 Discussion

The importance or iron regulation in cardiovascular disease (Anker et al. 2009) and in the control of pulmonary vascular tone is well established (Balanos et al. 2002; Smith et al. 2008; Smith et al. 2009). RDW has proven a consistent marker of all cause mortality in the general population (Perlstein et al. 2009; Patel et al. 2009). Elevation in RDW has been deemed attributable to chronic conditions such as diabetes mellitus (Danse et al. 2015) as well as correlations with c-reative protein (CRP) and erythrocyte sedimentation rate (Lippi et al. 2009, Vaya et al. 2015).

Iron deficiency has been associated with IPAH and in turn mortality in a smaller population of

IPAH/HPAH patients compared with Cohort 1 (Rhodes et al. JACC 2011). This translated into derived

ID criteria used in a recent clinical trial of either iron <10.3, ferritin <37.5 or TSAT <16.4% (Howard et al. 2013). In the current study, these cut-offs have been validated using a larger IPAH/HPAH cohort

(Cohort 1) but further cut-offs were derived using iron, ferritin, TSAT and RDW. On further interrogation, RDW was shown to be the best single marker for predicting sTfR and in turn iron deficiency [sTfR>28.1], even once correcting for inflammation (IL-6). Increases in sTfR has a strong association with overall mortality within Cohort 1.

Using these cut-offs, it can be demonstrated that iron deficiency is likely to be highly prevalent in other subtypes of pulmonary hypertension, thus potentially presenting a target for treatment, not only in IPAH, but also in these other forms. This was further supported by multiple associations with functional cardiopulmonary derangements, most marked in CTEPH. This is somewhat surprising given that iron deficiency may be more related to anticoagulation in this group, thus providing a disconnect between severity and iron deficiency, however, the results are striking in comparison with PAH and merit further validation. It is plausible that anticoagulation may amplify more subtle changes in iron physiology, by accentuating situations where iron absorption is decreased and thus, fails to correct for iron deficiency. It is not known however, whether there is impaired iron absorption in CTEPH due to raised hepcidin levels as there is in IPAH.

The associations of RDW with exercise limitation across all groups of PH studied, other than CTD-

PAH, lends weight to the importance of iron in disease severity. One possible explanation for the

129 failure to see an association in CTD-PAH is that these patients often have comorbidities as a result of

CTD that limit exercise and thus lead to a disconnect between PH severity and measures of exercise capacity. However, these data do not provide any evidence to state whether iron deficiency is merely a marker of disease severity or a target for treatment.

There appears to be a significant difference between peak VO2 mls/kg/min versus peak VO2

(%predicted) in the PAH-CTD and PAH-CHD cohorts when matching to RDW. There is no clear mechanistic reason why this should be, but reassuringly both these markers of oxygen consumption correlate well with RDW in the IPAH and CTEPH groups. This therefore is probably linked to much fewer numbers in the PAH-CTD and PAH-CHD groups (see table 2.2)

The survival analyses showing RDW to be an important prognostic marker are heavily limited by incomplete data. Many datasets which have derived biomarkers have been put together from prevalent cohorts or gathered data from different timepoints. In this study, an attempt was made to study incident patients only with contemporaneous data collection. With many variables being predictive of mortality on univariate analysis, there were insufficient deaths/events to undertake multivariate Cox regression analysis. This is a major limitation of the study and hence head-to-head analyses were undertaken to unpick the potential confounding relationships between RDW and other prognostic variables. It is unclear whether RDW could represent an important biomarker in non-IPAH forms of PH. Although it is not possible to determine this from the present data, the association of RDW / iron deficiency with mortality suggests that it may be important in terms of disease development and may serve as a therapeutic target in IPAH and other forms of PH.

130

3. The safety and impact of iron supplementation on cardiopulmonary haemodynamics and exercise capacity in patients with IPAH and iron deficiency

131

Contents 3.1 Introduction ...... 134 3.2 Methods ...... 135 3.2.1 Overall study design and plan ...... 135 3.2.2 Study outcome measures ...... 135 3.2.3 Study subjects - inclusion and exclusion criteria ...... 136 3.2.4 Study design and rationale ...... 138 3.2.5 Recruitment, randomisation and blinding ...... 141 3.2.6 Right heart catheter ...... 141 3.2.6.1 Timing of cardiac catheterisation in relation to PH drug administration ...... 142 3.2.6.2 Haemodynamic measurements ...... 142 3.2.6.3 Catheter position ...... 142 3.2.7 6 minute walk test ...... 142 3.2.8 Cardiopulmonary exercise testing (CPET) ...... 143 3.2.8.1 Incremental cardiopulmonary exercise testing (ICPET) ...... 143 3.2.8.2 Endurance cardiopulmonary exercise testing (ECPET) ...... 143 3.2.9 Assessment of quality of life ...... 143 3.2.10 Haematology and clinical biochemistry ...... 143 3.2.11 Statistics and data analysis ...... 144 3.3 Results ...... 145 3.3.1 Patient recruitment and characteristics ...... 145 3.3.2 Iron status ...... 145 3.3.3 Haematology ...... 151 3.3.4 RDW status ...... 155 3.3.5 Phosphate levels ...... 156 3.3.6 Cardiopulmonary haemodynamics ...... 157 3.3.7 Exercise ...... 158 3.3.7.1 Six minute walk distance (6MWD) ...... 158 3.3.7.2 Incremental cardiopulmonary exercise testing ...... 159 3.3.7.3 Endurance cardiopulmonary exercise testing ...... 164 3.3.8 Tolerability and adverse events ...... 165 3.3.9 Evolution of study design and endpoints ...... 165 3.4 Discussion ...... 166 3.4.1 Iron status following infusion of Ferinject ...... 166

132

3.4.2 Phosphate levels and RDW following infusion of Ferinject ...... 167 3.4.3 Haemodynamics following intravenous Ferinject ...... 168 3.4.3.1 Invasive haemodynamics following infusion of Ferinject ...... 168 3.4.3.2 Exercise capacity following infusion of Ferinject ...... 169 3.4.4 Safety and tolerance of Ferinject infusion in IPAH ...... 170 3.4.5 Study limitations ...... 171 3.5 Conclusions ...... 171 Appendix 3.1 Right heart catheterisation manual ...... 238 Appendix 3.2 Standard Operating Procedure for Incremental and Endurance Cardiopulmonary Exercise Testing ...... 243 Example Graphical output from metabolic cart ...... 264

133

3.1 Introduction

Several independent studies have demonstrated that iron deficiency is common in idiopathic and heritable PAH, affecting between 31 to 63% of patients (Rhodes et al. JACC 2011; Soon et al. Thorax 2011; Ruiter et al. Eur Resp J 2011; van Empel Heart Lung Circ 2014). Furthermore, iron deficiency was associated with reduced exercise capacity, worse functional class and increased mortality (Rhodes et al. JACC 2011; Ruiter et al. Eur Resp J 2011; van Empel Heart Lung Circ 2014). Furthermore, data presented in the previous chapter provide a basis for considering iron supplementation in IPAH to improve exercise capacity and outcomes.

Given the success of i.v. iron replacement therapy in heart failure and the importance of iron in regulating human pulmonary vascular tone, iron deficiency is an attractive therapeutic target in IPAH. While oral iron replacement therapy may be considered, it is slow to take effect, compliance is a problem and high hepcidin levels are expected to impede absorption (Rhodes et al. JACC 2011). Abnormal hepcidin levels have been found in IPAH, exceeding the plasma concentration of healthy volunteers, and inappropriately raised in iron-deficient IPAH patients (Rhodes et al. JACC 2011). Transportation of orally ingested iron from the gastrointestinal tract to the circulation requires ferroportin, which is down regulated by hepcidin when iron stores are replete or in response to inflammation. Indeed, elevated hepcidin levels may be used to predict which patients with iron deficiency anaemia will not respond to oral iron therapy but respond to ferric carboxymaltose (Ferinject) infusion (Bregman et al. Am J Hematol 2013). A proof-of-concept study is therefore essential to determine the safety and efficacy of intravenous iron in IPAH, not least because of general concerns about the use of i.v. iron preparations, but also the conflicting findings that iron restriction may attenuate (Wong et al. Free Radic Biol Med 2012; Naito et al. BBRC 2013) as well as promote (Cotroneo et al. Circ Res 2015) the development of pulmonary hypertension in experimental animal models.

We sought to examine the safety and efficacy of iv iron supplementation, testing the hypothesis that iron treatment would lead to improved pulmonary haemodynamics, cardiac function, exercise capacity and symptoms in iron deficient patients with IPAH.

134

3.2 Methods

3.2.1 Overall study design and plan A Phase 2 multicentre, proof-of-concept study was designed to examine the safety and potential therapeutic benefit of intravenous ferric carboxymaltose (Ferinject) in patients with IPAH and iron deficiency. The initial design comprised a 36-week double-blind, randomised, placebo-controlled, crossover study to investigate whether a single dose of Ferinject improves cardiopulmonary haemodynamics, cardiac function, exercise capacity and quality of life and is well-tolerated (Figure 4.1).

All patients were followed up for at least 24 weeks after dosing with Ferinject to measure the kinetics of iron therapy. Patients with IPAH and iron-deficiency, who had been stable on their current therapy for the preceding 1 month, would be treated with a single 1000 mg or 15 mg/kg (whichever is the smaller) dose of Ferinject.

The study was established in three centres – Hammersmith Hospital, London, Royal Hallamshire Hospital, Sheffield and Papworth Hospital, Cambridge. Each centre aimed to recruit 20 patients according to the same protocol and follow participants for up to 36 weeks after the first randomised infusion of Ferinject or saline. The study (http://clinicaltrials.gov/show/NCT01447628) was funded by a British Heart Foundation programme grant (RG/10/16/28575) and has received Research Ethics Committee approval (REC 11/LO/0095). Ferinject was provided by Vifor (International) Inc., Rechenstrasse 37, CH-9014 St Gallen, Switzerland. The company remained independent of the study design and had no role data collection or analysis.

3.2.2 Study outcome measures Primary outcome measure: The primary efficacy measure was a change in resting pulmonary vascular resistance from baseline at 12 weeks, measured by cardiac catheterisation.

Secondary measures: Invasive pulmonary haemodynamics, exercise performance, quality of life, WHO functional class, NT-pro-BNP and cardiac function as assessed by magnetic resonance scanning.

135

Measurements that will be made:  Incremental bicycle cardiopulmonary exercise testing according to ATS guidelines (ATS/ACCP Statement on cardiopulmonary exercise testing. AJRCCM 2003) – measurements will include

peak VO2 (ml/min/kg), VO2 at metabolic threshold, VE/VCO2 slope, VE/VCO2 equivalents at

metabolic threshold, VO2/WR slope and O2 pulse.  Endurance time on bicycle cardiopulmonary exercise test at 80% peak work rate (from the incremental test), with measurement gas exchange.  Resting cardiopulmonary haemodynamics – right atrial pressure, pulmonary arterial pressure, pulmonary wedge pressure, cardiac output, stroke volume and pulmonary vascular resistance, measurements indexed where appropriate.

 Exercise cardiopulmonary haemodynamics at work rate equivalent to 40% peak VO2 (derived from the incremental exercise test) at Hammersmith and Sheffield sites and at a fixed resistance and cadence estimated to correspond to 40% peak work rate at Papworth site – including pulmonary arterial pressure, pulmonary capillary wedge pressure, cardiac output, stroke volume and pulmonary artery saturation, measurements indexed where appropriate.  Iron indices – serum iron, transferrin saturation, ferritin, soluble transferrin receptor, iron binding capacity, red cell distribution width, and erythropoietin.  6 minute walk distance and Borg dyspnoea scale conducted according to American Thoracic Society (ATS) guidelines.  WHO functional class.  N-terminal-proBNP.  Quality of life using the CAMPHOR questionnaire (McKenna et al. Qual Life Res 2006) and the self-reported Patient Global Assessment.  Safety - the occurrence of adverse events.  Cardiac MRI (Sheffield and Hammersmith sites only) - right ventricular volumes, mass, ejection fraction, stroke volume and diastolic function.

3.2.3 Study subjects - inclusion and exclusion criteria The aim was to recruit 60 patients with symptomatic pulmonary arterial hypertension (idiopathic, heritable or associated with anorexigens). This required a documented diagnosis of PAH by right heart catheterisation prior to screening , with evidence of a resting mean pulmonary artery pressure ≥25mmHg, pulmonary wedge pressure equal or less than 15mmHg and normal or reduced cardiac output (Galie et al. Eur Resp J 2015).

136

Inclusion criteria:  Males or females aged between 18–75 years old.  PAH that is idiopathic, heritable or associated with anorexigens.  Iron deficiency as defined by any one of the following criteria: ferritin < 37 ug/l or iron < 10.3 umol/l or transferrin saturations < 16.4%.  Documented diagnosis of PAH by right heart catheterisation performed at any time prior to Screening showing: resting mean pulmonary artery pressure >25mmHg, pulmonary capillary wedge pressure ≤ 15 mm Hg and normal or reduced cardiac output.  6 minute walking distance greater than 50m at entry.  Stable on an unchanged PAH therapeutic regime (any combination of endothelin receptor antagonist, phosphodiesterase inhibitor or prostacyclin analogue) for at least 1 month.  Able to provide written informed consent prior to any study-mandated Procedures.  Female subjects of child-bearing potential are eligible to participate if they agree to use appropriate contraception.

Exclusion criteria: Patients will not be enrolled into the study if they meet any of the following criteria:  Unable to provide informed consent.  Clinically-significant renal disease (Creatinine clearance < 30 ml/min per 1.73 m2 calculated from CKD-Epi http://www.qxmed.com/renal/Calculate-CKD-EPI-GFR.php) or liver disease (including serum transaminases >3 times upper limit of normal).  Haemoglobin concentration <10 g/dl.  Patients with moderate to severe hypophosphatemia, defined as <0.65 mmol/L.  Known to have haemoglobinopathy e.g. sickle cell disease, thalassaemia.  Hospital admission related to PAH or changes in PAH therapy within 1 month prior to Screening.  Evidence of left ventricular disease or significant lung disease on high-resolution CT scanning or lung function as judged by the investigator.  Acute or chronic infection or inflammation.  Significant uncontrolled asthma as judged by the investigator, eczema or atopic allergies.  Females who are lactating or pregnant.  Individuals known to have HIV, Hepatitis B or C or Creutzfeld-Jakob disease.  Known hypersensitivity to Ferinject or any of its excipients.

137

 Evidence of disturbances in utilisation of iron.  Significant blood loss (e.g. GI bleed) within the last 3 months or history of menorrhagia.

3.2.4 Study design and rationale The three main considerations in the study design were (i) the definition of iron deficiency, (ii) dose and frequency of administration of intravenous iron and (iii) the optimal endpoints, including timing to capture onset and duration of effect.

(i) Soluble transferrin receptor (sTfR) is the best measure of iron deficiency in inflammatory diseases such as IPAH (>28.1 nmol/l) (Rhodes et al. JACC 2011). As it was not routinely available at all the participating centres, iron deficiency was also defined by standard laboratory measures of ferritin, iron and transferrin saturation. Using earlier profiling data (Rhodes et al. JACC 2011) it was estimated that 83% of IPAH patients with sTfR>28.1 nmol/L would be correctly identified if any one of the following criteria were met: ferritin < 37 µg/L or iron < 10.3 µmol/L or transferrin saturations < 16.4% (Howard et al. Pul Circ 2014).

(ii) In the FERRIC-HF trial of iron supplementation in congestive heart failure, patients received 200mg ferric carboxymaltose (Ferinject) weekly until iron repletion according to the Ganzoni formula, unless ferritin exceeded 500 µg/L, and then 200mg every four weeks until the end of the study (Okonko et al. JACC 2008). By 4 and 16 weeks, 781 ± 94 and 1269 ± 297 mg iron sucrose had been given to the non-anaemic group (Hb > 12.5 g/dL) and 1051 ± 219 and 1583 ± 366 mg had been given to the anaemic group. Ferinject, a slow release preparation, can be given in doses up to 1000mg. As IPAH patients live throughout the UK, to minimise the number of hospital visits, Ferinject would be given as a single 1000mg (or 15 mg/kg) dose and follow up iron profiling used to determine how long this dose maintains iron repletion.

(iii) It is unclear whether iron replacement would have an early effect, perhaps related to vasorelaxation or improved muscle bioenergetics, or a late effect related to vascular remodelling. Exercise capacity was therefore measured by cardiopulmonary exercise testing at 2 weeks and 12 weeks. To capture a meaningful effect on pulmonary haemodynamics, right heart catheterisation was undertaken at 12 weeks. Pulmonary vascular resistance was selected as the primary endpoint, in common with most other pharmacological studies in PAH, to demonstrate an effect on the primary defect in the disease. A cross-over design (Figure 3.1) was selected to improve the power to detect an effect on exercise. A repeat third cardiac catheter after cross-over was considered unacceptable to patient comfort. The visit and assessment schedule is shown in Table 3.1.

138

Figure 3.1. Schematic overview of a 36-week, double-blind, randomized, placebo- controlled, crossover study.

139

Assessment Visit 1# Visit 2 Visit 3 Visit 4 Visit 5 Visit 6 Visit 7 Screening Week 0 Week 2 Week 12 Week 14 Week 24 Week 36 ≤ 28 days Baseline/ (+/- 3 (+/- 3 days) (+/- 3 days) (+/- 3 days) (+/- 3 days) before week 0 randomisation § days) Written informed consent X Medical history (incl. demographics) X Physical examination X X X X X X X Concomitant medication X X X X X X X Modified NYHA (WHO) Functional Class X X X X X X X Vital signs X X X X X X X ECG X X X X X X 6MWT and Borg Dyspnoea Score X X X X CAMPHOR Questionnaire X X X X X X Patient Global Assessment X X X X X Routine laboratory assessments (e.g. iron X X X X X X X status) Research blood sampling X X X X X X Serum pregnancy testing (if applicable) X X X X Cardiac catheterisation - haemodynamics X X Incremental Cardiopulmonary Exercise Test X X X X X X Endurance Cardiopulmonary Exercise Test X X X§ Administration of study drug or placebo X X Adverse events X X X X X X

Cardiac MRI* X X X Table 3.1 Visit and assessment schedule. #; Iron status and IPAH status determined from routine clinical care records. §; Screening procedures may be combined with Visit 1 where data from routine clinical visits are not available. *; Additional exploratory assessments, where research facility for MRI available for this protocol

140

3.2.5 Recruitment, randomisation and blinding Potential participants were identified and screened using data collected during their routine outpatient appointment at each of the Pulmonary Hypertension centres. Patients were randomised on 1:1 ratio to each treatment arm. Randomisation was performed at the Baseline (Week 0) visit using InFormTM. Randomisation was stratified by gender, with an appropriate fixed block size, in order to ensure equal allocation to Ferinject and placebo. The computer-generated randomisation list was prepared by a statistician independent of the study team and provided to Imperial College staff (who are not otherwise involved in the study) responsible for building the InFormTM system.

Ferinject is currently licensed up to doses of 1000 mg as an intravenous infusion in the UK for the treatment of iron deficiency when oral preparations cannot be used or are ineffective. Patients who were eligible to participate in the Ferinject study received either Ferinject or normal saline (placebo) on day 2 at week 0 visit or day 2 at week 12 visit, depending on the treatment allocation. The Ferinject / placebo was dispensed from Pharmacy in a pre-prepared 50ml syringe, covered with a white label to maintain its blinding, and administered via intravenous infusion for 15 minutes after the patient’s cardiac catheterisation. Two unblinded nurses were designated to prepare the Ferinject / placebo infusion and to support patients’ blinding, the infusion set was hidden by a blue removable cover and the patient’s cannula extension covered with a light-protective cover (Gusgear™). To ensure that the patient remained blinded to the treatment they received, they were also blindfolded for a short time during the connection of the infusion. As the study was conducted in a double-blind manner the investigator, study staff, and patients remained blinded to the patient’s Ferinject / placebo treatment until the end of the study. Site personnel involved in the preparation and administration of the study drug were not permitted to undertake any of the study assessments. In order to prevent accidental unblinding of treatment allocation, an unblinded physician was also identified at each participating site to review all iron profile results after patients had been dosed.

3.2.6 Right heart catheter Right heart catheteriscatheterisation was necessary to obtain haemodynamic measurements for the evaluation of the primary endpoint, change in PVR, and secondary endpoints. To ensure standardised condition across all participating centres, investigators had to ensure a recognised, accepted, unifying protocol.

141

3.2.6.1 Timing of cardiac catheterisation in relation to PH drug administration The timing of the catheterisation has been important in relation to PH drug administration. The timing of administration of pulmonary hypertension-specific drugs in relation to the timing of cardiac catheterisation could have significant impacts on the outcome measures and thus needs to be standardized. The table below is a guide to when the last dose of medication should be taken prior to catheterisation (Table 3.2). This should be documented and repeated for the second cardiac catheter.

Time of Last Dose of Medication

Planned Time of Morning List Afternoon List Procedure Sildenafil 00:00 08:00

Tadalafil Omitted 08:00

Bosentan Omitted 08:00

Ambrisentan / Omitted 08:00 Macitentan Nebulised Iloprost >2.5 hours from catheter >2.5 hours from catheter

Table 3.2 Guide to last doses of medication prior to catheterisation

3.2.6.2 Haemodynamic measurements Right atrial pressure, pulmonary arterial pressure, pulmonary wedge pressure, cardiac output, stroke volume and pulmonary vascular resistance, measurements indexed where appropriate. Cardiac output has been achieved by the technique available to investigators at their site (thermodilution (average of 3 measurements) or direct Fick Method).

3.2.6.3 Catheter position Full right heart catheter standard operating procedure (SOP) is described in Appendix 3.1 including both Direct Fick and Thermodilution methods for the determining cardiac output (centre-dependent).

3.2.7 6 minute walk test Distance walked during an unencouraged 6 minute walk test conducted according to ATS guidelines [Arena et al. 2010; J Heart Lung Transplant]. This is a standard tool for the study of functional capacity in PAH patients and is primarily determined by cardiac output and hence right ventricular function.

142

3.2.8 Cardiopulmonary exercise testing (CPET) Exercise testing is increasingly used to evaluate the level of exercise intolerance in patients with lung and heart diseases [Arena et al. 2010; J Heart Lung Transplant].

3.2.8.1 Incremental cardiopulmonary exercise testing (ICPET) Patients undergo cycle ergometry CPET following a ramp protocol (Jaeger Masterscreen CPX) according to ATS guidelines [ATS/ACCP Statement on cardiopulmonary exercise testing. Am J Respir Crit Care Med 167: 211-277, 2003]. Using CPET, we can extract effort-independent data from exercise. Endpoints from CPET include oxygen consumption (VO2 ml/min/kg and % predicted), VO2 at peak and anaerobic (metabolic) threshold, ventilatory equivalen/rate of carbon dioxide consumption (VE/VCO2) slope and oxygen pulse. CPET equipment at each centre has been calibrated by two means. First, a metabolic simulator (Vacumed) will be used with standard gases and volumes to verify calibration of sensors. This has porvied an integrated calibration of volume and gas sensors in addition to bike workload. The Hammersmith site has porvied training for those supervising CPET at the other centres and we have undertataken site visits. In due course, all data will be analysed centrally to ensure similar measurement techniques for CPET variables. For a full desciption of the protocol see Appendix 3.2.

3.2.8.2 Endurance cardiopulmonary exercise testing (ECPET) ECPET has also been undertaken as part of the study at weeks 0, 12, 24. Endurance time, peak oxygen consumption and oxygen pulse, on bicycle cardiopulmonary exercise test at 80% peak work rate (determined from the ICPET), has been measured as well steady-state (at 3 minutes) oxygen pulse and gas exchange. For a full desciption of the protocol see Appendix 3.2.

3.2.9 Assessment of quality of life CAMPHOR is a disease-specific patient-reported outcome instrument which has been developed for patients with pulmonary hypertension [Mckeen et al. 2006; Qual Life Res 15). It is sensitive to small changes in health and quality of ife in this patient group. The self- reported Patient Global Assessment have been used to assess patient-perceive response to intervention.

3.2.10 Haematology and clinical biochemistry BNP and NT-proBNP, as an indicator of cardiac stress; iron status as assessed by serum iron, total iron binding capacity (TIBC), transferrin saturation plasma soluble transferrin receptor (sTfR), ferritin, aand hepcidin levels; EPO levels and RDW to assess erythrogenesis.

143

Inflammation status is known to affect hepcidin expression and will be determined by measuring CRP and IL-6 levels. Possible haemolysis will be assessed by measuring unconjugated bilirubin, LDH, haptoglovin and reticulocytes. Measurements are being made as described above or using standard clinical pathology accredited hospital assays.

3.2.11 Statistics and data analysis This is a proof of concept study, thus the sample size has been chosen with respect to safety (in terms of exposure to the drug and investigations) and feasibility (patient population) and with the aim of measuring effect size. We would expect that a sample size of 60 will be sufficient to detect clinically meaningful changes in pulmonary haemodynamics. We expect to see trends in exercise end-points,NT-proBNP and quality of life in the direction consistent with benefit.

The null hypothesis is that iron replacement has no effect on PVR in patients with iron deficiency and IPAH, with a two-sided alternative that iron replacement changes PVR in

-5 patients with iron deficiency and IPAH. Assuming a standard deviation of 250 dynes.s.cm and a drop-out rate of 10%, 60 patients randomised 1:1 would give this study an 80% power

-5 to detect a 194 dynes.s.cm reduction in PVR with iron treatment at a significance of p=0.05 using a two-sample t test. In previous trials in IPAH the mean treatment effect vs. placebo

-5 using bosentan [Cannick et al. Lancet. 2001] was 415 dynes.s.cm (223 decrease vs. 191

-5 increase) with iloprost [Olschewski et al. N Engl J Med 2002] 335 dynes.s.cm after

-5 inhalation and sildenafil [Galie et al. N Engl J Med 2002] 310 dynes.s.cm .

Data from the three centres has been pooled and will be analysed at Hammersmith. Due to the small sample sizes in each centre, the assessment of consistency of results amongst the three centres will be made on an informal basis.

We have endeavoured to minimise loss of patient follow up and resulting missing outcome data. All causes of missing data have and will be collected and reviewed to determine the type of missingness. If it is missing at random, this should have little bias on results with the proposed analysis. If informative missingness is suspected, we will perform sensitivity analysis to investigate robustness of the results from the proposed analysis base on plausible assumption about the missingness.

144

3.3 Results

3.3.1 Patient recruitment and characteristics This interim blinded analysis is based on the first 23 patients to be enrolled, 18 of whom had completed the study (i.e. followed up for a minimum of 24 weeks after the infusion of Ferinject). One subject withdrew immediately after enrolment (before commencing the study) and 4 had still to complete their 12-36 week assessments. The demographics of the cohort are summarised in Figure 3.2. All patients were already optimally treated with approved PAH therapies, in addition to background therapies including diuretics, for at least 4 weeks prior to enrolment and their treatment was not altered during the study.

3.3.2 Iron status This analysis was undertaken to provide interim results based on estimated therapy. Using lab data, raised ferritin levels after either of the dosing visits has been extrapolated to assume that the patient had received intravenous Ferinject.

The infusion of Ferinject led to an increase in ferritin levels 2 weeks later, whereas no significant change occurred after the infusion of saline in the same subjects (Figure 3.2 A-B). In parallel with the increase in ferritin, TIBC fell significantly at 2 weeks (Figure 3.2 C-D). Most subjects exhibited a marked rise in ferritin levels following the infusion of Ferinject, although the amount varied considerably between individuals and was not associated with the baseline level (Figure 3.3 A-B). Ferritin levels declined by 12-14 weeks and plateaued, remaining significantly elevated up to 24 weeks, with only 3/17 (18%) falling below 37 µg/L at this time point (Figure 3.3 A-B).

145

A 1000 B 60

50 P=0.8751 800 P=0.0002

600

g/L) g/L)

  30

400

Ferritin ( Ferritin ( Ferritin 20

200 10

0 0 Week 0 Week 2 Week 0 Week 2 Ferinject (1g) iv infusion Saline iv infusion

C D 100 100

P=0.0002 P=0.8311 90 90

80 80

mol/L)

mol/L) 

70  70 TIBC ( TIBC 60 ( TIBC 60

50 50

40 40 Week 0 Week 2 Week 0 Week 2 Ferinject (1g) iv infusion Saline iv infusion

Figure 3.2. Ferritin and TIBC in IPAH patients before and after the administration of iron or saline. Ferritin levels (A-B) and TIBC (C-D) was determined at baseline (week 0) and 2 weeks after the random i.v. infusion of 1g Ferinject (A, C; n=13) or saline (B, D; n=11). Pink shaded areas delineate iron deficient range for ferritin (<37 µg/L). Green shaded areas delineate normal TIBC range (45-66 µmol/L). Wilcoxon matched-pairs signed rank test, comparing concentrations at week 0 and 2 weeks post-infusion.

146

1000

800

600

g/L) 

400 Ferritin ( Ferritin

200

0 12 14 24 36 Weeks after Ferinject (1g) iv infusion

800

600

g/L) *** 

400 Ferritin ( Ferritin 200 ** ** *

0 0 2 12 14 24 36 Weeks after Ferinject (1g) iv infusion

Figure 3.3. Ferritin levels in IPAH patients before and after the administration of iron.

A, Ferritin levels for individual subjects (n=17) at baseline (week 0) and 2, 12, 14, 24 and 36 weeks after i.v. infusion of 1g Ferinject. B, Ferritin levels expressed as the median and interquartile range. Pink shaded area delineates iron deficient range for ferritin (<37 µg/L). ***, P<0.001; **, P<0.01; *, P<0.05; Kruskal-Wallis test, Dunn’s multiple comparison with baseline (week 0).

147

In parallel with the reduction in TIBC, there was a corresponding increase in circulating iron and transferrin saturation levels 2 weeks after the infusion of Ferinject (Figure 3.4 A-B). A close correlation was found between circulating iron and transferrin saturation levels, both at the start of the study (week 0) and in the differences measured 2 weeks later (Figure 3.4 C-D).

Analysis of 13 subjects, with a full dataset on completion of the study, demonstrated that ferritin, iron and transferrin levels all remained elevated for at least 12 weeks after the infusion of Ferinject whereas no change was observed following saline infusion (Figure 3.5 A-F)

148

A B

30 60 P=0.0026 P=0.0025

25 50

20 40 mol/L)

 15 30 Iron Iron (

10 20 Transferin saturation (%) saturation Transferin

5 10

0 0 Week 0 Week 2 Week 0 Week 2 Ferrinject (1g) iv infusion Ferrinject (1g) iv infusion

C D

60 25

r 0.8982 (%) r 0.9667 P<0.0001  P<0.0001 20 40

15 20

0 Transferin saturation (%) Transferin 5 saturation Transferin 5 10 15 20 -5 0 5 10 15 20 Iron (mol/L) Iron  (mol/L)

Figure 3.4. Iron and transferrin saturation levels in IPAH patients before and after the administration of iron or saline. Iron (A) and transferrin saturation (B) levels at baseline (week 0) and 2 weeks after the random i.v. infusion of 1g Ferinject (A; n=13) or saline (B; n=13) in the same subjects. Pink shaded areas delineate iron deficient range for iron (<10.3 µmol/L) and transferrin saturation (<16.4%). Wilcoxon matched-pairs signed rank test (A-B), comparing concentrations at week 0 and 2 weeks post-infusion. Correlation (Spearman’s) between iron and transferrin saturation at baseline (C) and the change (∆) in these parameters 2 weeks after iron treatment (D).

149

A B

1000 *** * ** 100

800 80

g/L)

g/L)   600 60

400 40

Ferritin( Ferritin( 200 20

0 0 0 2 12 14 24 36 0 2 12 14 24 36 Weeks after Ferinject (1g) iv infusion Weeks after saline iv infusion

C D 40 40

30 *** ** 30 mol/L)

mol/L) 20 20



 Iron ( Iron Iron ( Iron 10 10

0 0  0 2 12 14 24 36 0 2 12 14 24 36 Weeks after Ferinject (1g) iv infusion Weeks after saline iv infusion

E F 60 *** *** ** 60

40 40

20 20 Transferin saturation (%) Transferin Transferin saturation (%) Transferin 0 0 0 2 12 14 24 36 0 2 12 14 24 36 Weeks after Ferinject (1g) iv infusion Weeks after saline iv infusion

Figure 3.5. Ferritin, iron and transferrin saturation levels in IPAH patients before and after the infusion of Ferinject or saline. Ferritin (A-B), iron (C-D) and transferrin saturation levels (E-F), expressed as the median and interquartile range, at baseline (week 0) and 2, 12, 14, 24 and 36 weeks after the random i.v. infusion of 1g Ferinject or saline in the same subjects (n=13). Pink shaded areas delineate iron deficient range for ferritin (<37 µg/L), iron (<10.3 µmol/L) and transferrin saturation (<16.4%). ***, P<0.001; **, P<0.01; *, P<0.05; Kruskal-Wallis test, Dunn’s multiple comparison with baseline (week 0).

150

Seventeen subjects were judged to have received an infusion of Ferinject, exhibiting a mean 27-fold (range 6- to 94-fold) increase in circulating ferritin levels two weeks after infusion, and fourteen were followed for 24-36 weeks. Out of these participants, six (43%) became iron deficient again (i.e. iron <10.3 µmol/L, ferritin <37.0 µg/L or transferrin saturation <16.4%) during the course of the study (i.e. within 36 weeks) and seven (50%) subjects were not iron deficient until 1-2 years (n=4) or 2-3 years (n=3) after the infusion of Ferinject. At their most recent follow up (76 week) appointment, the remaining subject was still not iron deficient.

3.3.3 Haematology Haemoglobin and haematocrit levels tended to rise, albeit non-significantly, 2 weeks after the infusion of Ferinject and no significant change was observed following saline infusion (Figures 3.6 & 3.7).

151

A B 18 18

16 16

P=0.0830 14 14

12 12

Haemoglobin (g/dL)Haemoglobin (g/dL)Haemoglobin

10 10 0 2 12 24 0 2 Weeks after Ferinject (1g) iv infusion Weeks after Ferinject (1g) iv infusion C D 18 18

16 16

14 14 P=0.3051

12 12

Haemoglobin (g/dL)Haemoglobin Haemoglobin (g/dL)Haemoglobin

10 10 Pre 0 2 12 0 2 Weeks after saline iv infusion Weeks after saline iv infusion

Figure 3.6. Haemoglobin levels in IPAH patients before and after the administration of saline or iron. Haemoglobin levels (median and interquartile range) measured before and after i.v. infusion of 1g Ferinject (A-B; n=13) or saline (C- D; n=11). Pre, preceding baseline (screening) measurements of haemoglobin. Wilcoxon matched-pairs signed rank test (B, D), comparing concentrations at baseline (week 0) and 2 weeks post-infusion.

152

A B

0.52 0.52

0.48 0.48

0.44 0.44 P=0.1217

0.40 0.40

Haematocrit (ratio) Haematocrit Haematocrit (ratio) Haematocrit 0.36 0.36

0 2 12 24 0 2 Weeks after Ferinject (1g) iv infusion Weeks after Ferinject (1g) iv infusion

C D 0.52 0.52

0.48 0.48

0.44 0.44 P=0.9324

0.40 0.40

Haematocrit (ratio) Haematocrit Haematocrit (ratio) Haematocrit 0.36 0.36

Pre 0 2 12 0 2 Weeks after saline iv infusion Weeks after saline iv infusion

Figure 3.7. Haematocrit in IPAH patients before and after the administration of saline or iron. Haematocrit (median and interquartile range) before and after i.v. infusion of 1g Ferinject (A-B; n=13) or saline (C-D; n=11). Pre, preceding baseline (screening) haematocrit values. Wilcoxon matched-pairs signed rank test (B, D), comparing concentrations at baseline (week 0) and 2 weeks post-infusion.

153

MCV levels increased 2 weeks after the infusion of Ferinject infusion (P=0.0128), and remained elevated up to 24 weeks, whereas no significant change was observed following saline infusion (Figure 3.8 A-D).

A B

95 95

90 90 P=0.0128

85 85

MCV (fL) MCV (fL) 80 80

75 75 0 2 12 24 0 2 Weeks after Ferinject (1g) iv infusion Weeks after Ferinject (1g) iv infusion

C D

95 95

90 90

P=0.1484

85 85

MCV (fL) MCV (fL) 80 80

75 75 Pre 0 2 12 0 2 Weeks after saline iv infusion Weeks after saline iv infusion

Figure 3.8. MCV levels in IPAH patients before and after the administration of saline or iron. MCV levels (median and interquartile range) measured before and after i.v. infusion of 1g Ferinject (A-B; n=13) or saline (C-D; n=11). Pre, preceding baseline(screening) measurements. Wilcoxon matched-pairs signed rank test (B, D), comparing values at baseline (week 0) and 2 weeks post-infusion.

154

3.3.4 RDW status Infusion of Ferinject led to a significant increase in RDW values 2 weeks later, whereas no change was observed following saline infusion (Figure 3.9). RDW values returned to baseline (week 0) levels at 12 and 24 weeks post-infusion of Ferinject.

A B 24 24

20 20

P=0.0032

16 16

RDW (%) RDW (%)

12 12

0 2 12 24 0 2 Weeks after Ferinject (1g) iv infusion Weeks after Ferinject (1g) iv infusion

C D 24 24

20 20

P=0.2629

16 16

RDW (%) RDW (%)

12 12

Pre 0 2 12 0 2 Weeks after saline iv infusion Weeks after saline iv infusion

Figure 3.9. RDW in IPAH patients before and after the administration of saline or iron. RDW (median and interquartile range) before and after i.v. infusion of 1g Ferinject (A-B; n=13) or saline (C-D; n=11). Pre, preceding baseline (screening) RDW values. Wilcoxon matched-pairs signed rank test (B, D), comparing concentrations at baseline (week 0) and 2 weeks post-infusion.

155

3.3.5 Phosphate levels Infusion of Ferinject led to a significant reduction in circulating phosphate levels 2 weeks later, whereas no change was observed following saline infusion (Figure 3.10). The hypophosphatemia was moderate (phosphate values in the range 0.32 to 0.8 mmol/L or above) and transient, phosphate levels largely falling in the normal range (0.8 to 1.45 mmol/L) from 12-36 weeks (Figure 3.10).

A 1.6

1.2

0.8 ** 0.4

Phosphate (mmol/L) Ferinject

0.0 0 2 12 14 24 36 Weeks after Ferinject (1g) iv infusion B 1.6

1.2

0.8

0.4

Phosphate (mmol/L) Saline

0.0 Pre 0 2 12 Weeks after saline iv infusion

Figure 3.10. Circulating phosphate levels in IPAH patients before and after the administration of iron or saline. Phosphate levels (median and interquartile range) before and after i.v. infusion of 1g Ferinject (A-B; n=9) or saline (C-D; n=12). Pink shaded area indicates moderate hypophosphatemia (0.32-0.8 mmol/L) and green area the normal range (0.8-1.45 mmol/L). Pre, preceding baseline measurements. **, P<0.01, Kruskal-Wallis test, Dunn’s multiple comparison (B).

156

3.3.6 Cardiopulmonary haemodynamics Infusion of Ferinject made no significant difference to invasive haemodynamics. There has not been any significant changes from baseline measurements at the point of cardiac catheterisation, after intravenous infusion of either saline or Iron (Ferinject). The primary endpoint, PVR, shows no trends in either treatment arm (Figure 3.11). Invasive markers of cardiac function including cardiac index, stroke volume and mixed venous saturations also show limited trends and differences (Figure 3.12).

A B

p = 0.93 20 10 p = 0.92

10 5

0 0

-10

-5 (mmHg) PAPm

 RAPm (mmHg) RAPm

 -20 -10

Iron Iron Saline Saline Presumed drug Presumed drug C p = 0.44 D p = 1.0

10 ) 1 2

5 0

0 -1

PVR (WU)

 PVR index (WU.m index PVR

-5  -2

Iron iron Saline saline Presumed drug Presumed drug

Figure 3.11. Changes in mean right atrial and pulmonary artery pressures, and subsequent

pulmonary vascular resistance, 12 weeks after receiving either saline or iron infusion. Invasive

markers of pulmonary vascular disease (median, interquartile range) compared via Mann-

Whitney U test - Δ RAPm, change in mean right atrial pressure, (A); Δ PAPm, change in mean right

atrial pressure, (B); Δ PVR, change in pulmonary vascular resistance, (C), and pulmonary vascular

resistance index, (D). Saline, n=7; iron, n=4.

157

A B p = 0.61 p = 0.65 40 2

20 ) 2 1 0

0 -20

SV (ml.beat) SV -40 

CI (l.min.m CI -1  -60 -2

Iron Iron Saline Saline Presumed drug Presumed drug

p = 0.51 C 10

saturation (%) saturation 2

-5 SVO

 -10

Iron Saline Presumed drug

Figure 3.12 Changes in ventricular output measured by cardiac index and stroke volume, and mixed venous oxygen saturation measured 12 weeks after receiving saline or iron infusions. Invasive markers of cardio-pulmonary disease (median, interquartile range) compared via Mann-

Whitney U test: Δ CI, change in cardiac index, (A); Δ SV, change in stroke volume, (B); Δ SVO2,

change in mixed venous saturation, (C). Saline, n=7; iron, n=4.

3.3.7 Exercise

3.3.7.1 Six minute walk distance (6MWD) Six minute walk testing showed no differences in walk distance from baseline measurements in either treatment arm (Figure 3.13).

158

p = 0.88 100

-50 6MWD (m) 6MWD

 -100

-150

Iron Saline Presumed drug

Figure 3.13. Changes in six minute walk distance (6MWD) at 12 weeks either having received intravenous saline or iron treatment. 6MWD (median, interquartile range) compared using Wilcoxon matched-pairs signed rank test. n=10.

3.3.7.2 Incremental cardiopulmonary exercise testing Several consistent trends are seen in the CPET data. Presumend iron (Ferinject) treatment improves oxygen consumption (peak VO2), when standardised both for weight (ml/kg/min) and by age, height and gender (% predicted) (Figure 3.14). There is a small (non-signficant) improvement in work achieved (Watts) (Figure 3.15, A). Although insignificant, the trends suggest a change improvement in gas exchange after iron treatment with increases in the oxygen uptake efficiency slope (OUES) and reduction in the minute ventilation/carbon dioxide production (VE/VCO2) slope (Figure 3.16). Time to anaerobic threshold on and incremental ramp CPET protocol has been assessed (Figure 3.17). There is significant improvement in the time to anaerobic threshold (p=0.047). Two types of analysis were run, since in some individual CPET studies the point of AT was very difficult to delineate. In turn, an unpaired, Mann-Whitney U test was also used to maximise the number of data points available (Figure 3.16, B). VO2 at AT shows increased trends after iron therapy (Figure 3.17). Peak heart rate shows no significant changes in either arm of the study. Exercise recovery data are not significant between the two treatment arms (Figure 3.19).

159

A B

p = 0.41 p = 0.29 2 10

0 0 (ml.kg.min)

2 -2 -10

(% predicted) (% 2

-4 -20

Peak VO Peak -6 -30



Peak VO Peak 

Iron Iron Saline Saline Presumed drug Presumed drug

Figure 3.14. Changes in peak VO2 at 12 weeks either having received intravenous saline or iron

treatment. Peak VO2 (median, interquartile range) was assessed and standardised by both weight (ml/kg/min) (A) and % predicted B). Wilcoxon matched-pairs signed rank test. n= 12.

A B

p = 0.48 p = 0.43 20 10 10 5 0 0 -10

pulse (ml.beat) pulse -5 -20 2

-30 -10 Peak work (Watts) work Peak

 -40 -15

Peak O Peak 

Iron Iron Saline Saline Presumed drug Presumed drug

Figure 3.15. Changes in peak work and oxygen pulse at 12 weeks either having received

intravenous saline or iron treatment. Peak work (A) and O2 pulse (B) (median, interquartile

range) compared using Wilcoxon matched-pairs signed rank test. Peak work, n= 11; Peak O2 pulse, n=12.

160

A B p = 0.11 p = 0.13 200 20

10 0 2

OUES

VE/VCO 

-200  -10

-400 -20

Iron Iron Saline Saline Presumed drug Presumed drug

Figure 3.16. Changes in OUES and VE/VCO2 at 12 weeks either having received intravenous

saline or iron treatment. OUES and VE/VCO2(median, interquartile range) compared using Wilcoxon matched-pairs signed rank test. n=10.

161

A B

p = 0.25 (paired) p = 0.047 (unpaired) 200 150

100 * 100 50

0 0

-50 Time to AT (s) AT to Time

Time to AT (s) AT to Time -100   -100

-200 -150

Iron Iron Saline Saline Presumed drug Presumed drug

Figure 3.17. Changes in time to achieve anaerobic threshold on an incremental CPET, 12 weeks after having received intravenous saline or iron. Time to AT in seconds (median, interquartile range) compared: Wilcoxon matched-pairs signed rank test (A), n=4; Mann-Whitney U test (B),

n=9, * p<0.05.

p = 0.32 1.0

0.5

0.0

at AT (l/min) AT at 2

-0.5

VO  -1.0

Iron Saline Presumed drug

Figure 3.18. Changes in VO2 at anaerobic threshold on incremental CPET, 12 weeks after having received intravenous saline or iron. Changes in VO2 (median, interquartile range) compared using Wilcoxon matched-pairs signed rank test. n=10.

162

p = 0.47 B p = 0.56 A 200 20

10 0

0 (ml.min)

2 -200 -10

-400 -20

Peak HR (bpm) HR Peak

 Peak VO Peak -600 -30

Iron Iron Saline Saline Presumed drug Presumed drug

C 2 D p = 0.47 0.2 p = 0.42 0.2

0.1 0.1

0.0 0.0

- recovery 90s)/peak VO 90s)/peak recovery - -0.1 -0.1 2

-0.2 -0.2

(Peak HR-HRR 60s)/Peak HR 60s)/Peak HR-HRR (Peak Iron

Iron 

(Peak VO (Peak Saline Saline  Presumed drug Presumed drug

Figure 3.19. Peak VO2 and heart rate (HR); VO2 at 90 seconds into recovery; heart rate at 60 seconds into recovery; 12 weeks having received either intravenous saline or iron. Peak VO2 (A), peak heart rate (B), VO2 recovery at 90 seconds (C), Heart rate recovery at 60s (D); (median, interquartile range). Wilcoxon matched-pairs signed rank test. HRR, heart rate covery; (peak

VO2-recovery 90s)/peak VO2, VO2 recovery at 90s as a proportion of peak VO2; (peak HR – HRR

60s)/peak HR, heart rate recovery at 60s as a proportion of peak HR. n=12.

163

3.3.7.3 Endurance cardiopulmonary exercise testing Endurance cardiopulmonary exercise (submaximal exercise) testing has shown trends suggesting that iron therapy improves both endurance time and peak VO2 (Figure 3.20). Peak oxygen pulse shows a significant improvement (p=0.02; Figure 3.21, B) following the infusion of iron, as well as positive trends 3 minutes in to steady state exercise (Figure 3.21, A).

A B

p = 0.38 p = 0.28 600 0.3

400 0.2

(l.min) 0.1

200 2 0.0 0 -0.1

-200 -0.2

Peak VO Peak

 Endurance time (s) time Endurance

 -400 -0.3

Iron Iron Saline Saline Presumed drug Presumed drug

164

A B

p = 0.13 p = 0.02 2 4 * 3 1 2

0 1 pulse (ml.beat) pulse

2 0 -1 -1

pulse at 3 mins (ml.beat) mins 3 at pulse -2 Peak O Peak

2 -2

 O

 Iron Iron Saline Saline Presumed drug Presumed drug

Figure 3.21 Changes in oxygen pulse during steady state submaximal exercise (O2 pulse at 3 mins) and peak oxygen pulse, 12 weeks after receiving either intravenous saline or iron. Both 3 minute and peak values (median, interquartile range) compared using Wilcoxon matched-pairs signed rank test. 3 mins, n=7; peak, n=10.

3.3.8 Tolerability and adverse events Two patients were admitted for low phosphate levels and received phosphate infusions. Both subjects were asymptomatic. These were reported as serious adverse events.

3.3.9 Evolution of study design and endpoints Recruitment to the study, in particular at other participating centres, was slower than anticipated and has led to an increase in the number of collaborating centres. Fuwai Hospital in Beijing, China started recruiting in September 2015 and centres in Germany (Cologne and Giessen) are due to begin recruitment before the end of 2016. The primary reasons given by patients for reluctance to enrol were the number of visits, having to stay overnight and undergo cardiac catheterisation.

165

3.4 Summary and Discussion Summary This was a challenging trial to recruit patients for hence it is not fully completed and in turn I have had to use unblinded data and Ferritin levels to try and look for trends in outcome parameters. The signal is weak in invasive haemodynamics but we are seeing some early trends in CPET. Reassuringly it appears intravenous Ferinject does appear to correct iron deficiency in these patients within the study period.

Discussion The infusion of Ferinject led to an improvement in iron status, patients exhibiting an overall increase in iron storage and availability. Despite a transient increase in RDW and reduction in circulating phosphate levels, the infusion of 1g Ferinject appeared to be well tolerated and was not accompanied by other potentially adverse effects. Despite significant caution needing to be applied to these data which remain blinded, this interim analysis suggests that there is lack of response noted on static invasive haemodynamics (cardiac catheterisation) but there have been trends and on occasion, significant changes, in exercise performance, particularly on submaximal exercise testing. No data are available on cardiac MR, sTfR or NT- proBNP since these will be analysed at the end of the study. Quality of life data are locked in the database until trial closure.

3.4.1 Iron status following infusion of Ferinject Numerous open-label studies have reported the effect of Ferinject infusion on iron status in a variety of disease states. The majority of these studies were of 6-12 weeks duration and included patients with heart failure (Bolger et al. JACC 2006; Usmanov et al. J Nephrol 2008;), iron deficiency anaemia (Kulnigg et al. Am J Gastroenrerol. 2008; Seid et al. Am J Obstet Gynecol 2008; Van Wyck et al. Transfusion 2009; Qunibi et al. 2011; Barish et al. Anemia 2012; Charytan et al. nephrol Dial Transplant 2013; Wolf et al. J Bone & Mineral Res 2013; Bregman et al., Am J Hematol 2013; Onken et al. Transfusion 2014), hyperlipidaemia (Schatz et al. Atherosclerosis 2015), and PAH (Viethen et al. Int J Cardiol 2014; Ruiter et al. Pul Circ 2015). In contrast, there have been few randomized, blinded and placebo-controlled studies, these having been conducted in patients with heart failure over 16-24 weeks (Okonko et al. JACC 2008; Anker et al., NEJM 2009; Ponikowski et al. Eur Heart J 2015).

166

The current investigation represents the first randomized, double-blinded and placebo- controlled study to be undertaken in PAH, with patients followed up for at least 24 weeks after the infusion of Ferinject. Ferritin levels were markedly elevated at 2 weeks and remained significantly elevated 24 weeks afterwards. Circulating iron and transferrin saturation levels were also significantly elevated over this period, the changes in iron indices corresponding to those found in the studies described above. Importantly, ~50% of the IPAH patients followed up after completion of the study were still not classified as iron deficient 1-3 years after Ferinject infusion. This is consistent with the loss of iron from the body, which is thought to occur mainly from the gastrointestinal tract and at a rate of approximately 1-2 mg/day (Mole Antioxidants & redox signalling 2010). Thus, in the absence of additional iron loss or absorption, it may take 1-3 years for iron stores to return to pre- infusion levels following the infusion of 1g Ferinject (Bart et al. J Appl Physiol 2016). The persistent elevation of iron stores (ferritin levels), long after ferric carboxymaltose has been eliminated from the blood (half-life 7-12 hours), may also affect pulmonary vasculature function, Ferinject infusion in healthy volunteers being accompanied by a 50% reduction in the pulmonary vasoconstrictor response to hypoxia 6 weeks later (Bart et al. J Appl Physiol 2016).

In contrast to the marked changes in iron status following the infusion of Ferinject, the effects on haematological parameters were relatively modest. Haemoglobin and haematocrit levels tended to rise, but were not significantly affected following Ferinject infusion, while MCV exhibited a significant increase at 2 weeks post infusion. Previous studies have also reported varying effects of Ferinject infusion, finding either an increase in haemoglobin, haematocrit and MCV (Van Wyck et al. Transfusion 2009; Viethen et al. Int J Cardiol. 2014; Favrat et al. PLOSOne 2014; Froessher et al. BMC Preg & Childbirth 2014; Robels-Mezcua et al. Int J Cardiol 2016) or no significant change (Anker et al. NEJM 2009; Wolf et al. JBMR 2013; Ruiter et al. Pul Circ 2015).

3.4.2 Phosphate levels and RDW following infusion of Ferinject The most common adverse event associated with the infusion of Ferinject is a temporary reduction in circulating phosphate levels, but this was not examined in the two open-label proof-of-concept investigations conducted to date in patients with PAH (Viethen et al. Int J Cardiol 2014; Ruiter et al. Pul Circ 2015). In keeping with earlier studies on the treatment of iron deficiency anaemia (Van Wyck et al. Transfusion 2009; Barish et al. Anemia 2012; Wolf

167 et al. J Bone & Mineral Res 2013), a transient hypophosphatemia was observed 2 weeks after a single dose (1g) Ferinject infusion in iron deficient IPAH patients. This was also accompanied by an increase in RDW, both phosphate and RDW returning to baseline levels by 12 weeks. Previous studies have examined the effect of ferric carboxymaltose infusion on RDW in iron deficient children (Akarsu et al. Acta Haematologica 2006) and iron deficient adults with chronic heart failure (van Craenbroeck et al. Eur Heart J 2013). Although the treatment procedures differed from the current protocol, both reported an increase in RDW 1-4 weeks after infusion and a subsequent reduction by 12-24 weeks. The initial early rise in RDW coincided with an increase in MCV and may reflect active erythropoiesis (van Craenbroeck et al. Eur Heart J 2013). However, a temporal association has also been noted between hypophosphatemia and increased RDW following intravenous ferric carboxymaltose infusion for iron deficiency anaemia (Van Wyck et al. Transfusion 2009; Barish et al. Anemia 2012). In addition, a connection between RDW and FGF23 catabolism has recently been reported, independent of phosphate and iron metabolism and inflammation (van Breda et al. PLOSOne 2015). Further studies are therefore required to understand the apparent transient post-infusion relationship between phosphate homeostasis and RDW in IPAH as well as other disease states.

3.4.3 Haemodynamics following intravenous Ferinject

3.4.3.1 Invasive haemodynamics following infusion of Ferinject There are no changes seen in invasive end-points between treatment arms which in turn. This fits with indirect data from the open-label study by Ruiter et al. (Ruiter et al. Pul Circ 2015) showing no changes in right ventricular function which would be expected with significant changes in right ventricular afterload. There may be several reasons why we fail to demonstrate any changes unlike the animal data (Controneo et al.) or data showing changes in pulmonary vascular tone in short-term hypoxia (Smith et al., JAMA 2009). Animal models may not be representative of chronic PAH in humans who demonstrate more advanced and perhaps fixed remodelling. Iron supplementation may not affect pulmonary vascular tone, either because it is a short-lived effect while serum iron levels are high or because advanced vascular remodelling prevents any acute vasodilatation, similar to the studies in high-altitude natives (Smith et al., JAMA 2009). There are no comparable iron studies in PAH or heart failure utilising invasive end points from cardiac catheterisation.

168

3.4.3.2 Exercise capacity following infusion of Ferinject High RDW values have been linked to exercise intolerance in heart failure patients, RDW being a significant negative determinant of peak VO2, which also correlated with maximal workload and VO2-derived prognostic indices such as VO2 at anaerobic threshold and O2 pulse, and was positively associated with the VE/VCO2 slope (Van Craenenbroeck et al. Eur J Heart Failure 2012), an adverse marker in heart failure. Iron deficiency is thought to contribute to increased RDW and exercise intolerance in heart failure patients, treatment with ferric carboxymaltose being associated with an increase in 6MWD and overall reduction in RDW (Van Craenenbroeck et al. Eur J Heart Failure 2013). The increase in functional capacity was significant at 24 weeks and sustained at 1 year follow-up (Ponikowski et al. Eur Heart J 2015). Cardiopulmonary exercise testing is currently being examined in the EFFECT- HF trial (Effect of Ferric Carboxymaltose on Exercise Capacity in Patients with Iron Deficiency and Chronic Heart Failure; NCT01394562), the primary outcome being a change in peak VO2 from baseline to week 24 after ferric carboxymaltose infusion.

The exercise end-points in this present study are encouraging. Six minute walk distance is not significant, comparable with the study by Ruiter et al. Viethen et al. did publish significant changes in 6MWD but it must be noted this was a heterogeneous population with iron replete vs iron deficient patients with PAH, with variations in baseline data and no cross-over arm to the study. However, Ruiter et al. showed no changes in maximal exercise testing.

The present study has shown suggestions in improvement in both maximal and submaximal exercise performance.

Incremental cardiopulmonary exercise test (maximal) We have shown positive trends in maximal exercise (ICPET) data with increases in peak oxygen consumption (peak VO2), work and oxygen pulse following iron therapy. Gas exchange has been a rewarding finding with both OUES and VE/VCO2 showing positive trends following iron treatment. The oxygen uptake efficiency slope (OUES) is a function of VO2 and ventilation (VE). A steeper slope and a higher OUES represents a more efficient oxygen uptake while a lower OUES shows that higher amount of ventilations is required for oxygen uptake [Hollenberg et al. JACC. 2001]. Exercise capacity, as reflected by peak VO2, has been consistently shown to be one of the most powerful prognostic markers in patients with CHF [O’Neill et al. Circulation. 2005].

169

However, peak VO2 has important limitations and may not be accurately obtained, as the plateau may be inaccurate by the patient unable to achieve peak exercise. By contrast, the minute ventilation/carbon dioxide (VE/VCO2) slope is easier to determine objectively than maximal exercise capacity. Numerous studies have confirmed that VE/VCO2 slope has equivalent or even superior prognostic value to the measurement of peak VO2 with CHF (Sarullo et al. 2010. Open Cardio vasc med J). Both previous papers in IPAH showed positive findings in relation to anaerobic threshold. Our findings in relation to the AT are consistent with findings from the two other open-label trials which showed changes in VO2 at AT (Vithien et al) and time to AT (Ruiter et al). CPET is able to identify aerobic metabolism, anaerobic metabolism and the anaerobic threshold because different metabolic pathways have different signatures on gas exchange [Older P, Hall A. Wasserman, K (ed): Gas Exchange in Heart Disease. Armonk, NY: Futura publishing Company, Inc. 1996]. This study has also supported an increase time to AT on a incremental ramp protocol following iron therapy with very some small signal in improvement in VO2 at AT.

Endurance cardiopulmonary exercise test (submaximal)

Submaximal exercise testing has shown trends in improved endurance time and peak VO2, but significant trends in peak oxygen pulse. Steady state (3 minute) oxygen pulse also showing close to significant positive signal. The amount of oxygen consumed per heart beat is termed the oxygen pulse and depends on the size of the stroke volume (SV) and the arteriovenous oxygen content difference (reflecting oxygen utilisation / extraction), the latter also relating to haemoglobin concentration. Apart from haemoglobin, it is not possible to determine which of these components is responsible for a change in oxygen pulse. The study by Ruiter et al. also analysed muscle biopsies pre and post iron supplementation and demonstrated that there was an increase in oxidative metabolism, which may suggest that our findings and theirs relate to improved muscle oxygen extraction, rather than any changes in central cardiopulmonary haemodynamics.

3.4.4 Safety and tolerance of Ferinject infusion in IPAH Potential concerns have been associated with intravenous iron therapy, including hypersensitivity reactions, oxidative stress (which may increase the risk of cardiovascular events), exacerbation of infection, hypophosphataemia and endothelial damage. Nevertheless, recent reviews and meta-analyses, involving several thousand patients,

170 indicate that serious adverse events are rare and not accompanied by increased risk of infections (Chertow & Winkelmayer Am J Hematol 2010; Moore et al. BMC Blood Disord 2011; Begman & Goodnough Therap Advances in Hematol 2014; Avni et al. Mayo Clin Proc 2015). In addition to appearing safe and effective, intravenous ferric carboxymaltose has also been found to provide a more efficient and rapid correction of circulating ferritin levels, compared with other iron preparations (Rognoni et al. Clin Drug Invest 2016). Although the adverse effects of intravenous ferric carboxymaltose on circulating phosphate levels and RDW appear to be transitory it will be necessary to establish whether there are longer term effects, particularly in patients who receive multiple ferric carboxymaltose injections (Hardy & Vandemergel Int J Rheumatol 2015).

3.4.5 Study limitations The interim analysis of patients enrolled to date has limited power as the number of subjects is small and it has been undertaken without access to un-blinded data and research plasma samples. The random infusion of Ferinject and saline was therefore determined indirectly, based on the characteristic marked rise in ferritin levels 2-12 weeks after infusion of Ferinject. The results of appropriate laboratory assays, including measurement of sTfR, hepcidin, IL-6, FGF23, and cleaved FGF23 levels, as well as missing iron indices, were also unavailable and will determined together once the study has closed. Nevertheless, the analysis has confirmed persistent improvements in iron status following Ferinject infusion, accompanied by transient adverse changes in phosphate levels and RDW.

3.5 Conclusions We have undertaken the first randomised, placebo-controlled, double-blind trial of intravenous iron in patients with IPAH who have evidence of iron deficiency. This represents a first also in that this is a biomarker-driven study with patients being recruited on the basis of their iron profile. Two open-label studies have shown improvements in oxygen delivery, but these cold have been affected by the open-label nature of the study, leading to interpretation bias. Nonetheless, early unblinded data, which should be interpreted with great caution, support their findings and suggest that there may be no benefit to central haemodynamics. This may mean that patients gain symptomatic benefit and further studies will need to be undertaken to determine if there are any other longer-term adverse effects. This approach of undertaking an unblinded interim analysis is of course not a typical

171 approach to study evaluation, but due to slow recruitment, was undertaken for the purposes of presentation in this thesis.

172

4. Impact of obesity on right heart catheterisation measurements and diagnosis of pulmonary hypertension

Contents 4.1 Introduction ...... 174 4.2 Methods ...... 176 4.3 Results ...... 180 4.3.1 Demographics ...... 180 4.3.2 Example data ...... 180

173

4.1 Introduction Obesity is on the increase and may often be a comorbid feature in pulmonary hypertension.

It has major contributory effects on cardio-pulmonary disease. Body mass index (BMI) correlates with left ventricular mass and wall thickness but obese patients have also been shown to develop left ventricular dysfunction which is independent of blood pressure, age, gender and left ventricular mass.(Watson and Fox 2010)

Obesity leads to an increase in the work of breathing, neural respiratory drive and ventilatory load, sleep disordered breathing which in turn may lead to hypercapnic respiratory failure.(Steier, Jolley et al. 2009) Steier et al showed that increased intra- abdominal pressure and the consequent increased intrathoracic pressure in obese subjects reduce functional residual capacity (FRC) and expiratory reserve volume (ERV).(Steier, Lunt et al. 2014) Pressures recorded in the pulmonary circulation at right heart catheterisation are made with reference to atmospheric pressure. Haemodynamic definitions of pulmonary hypertension are based on the assumption that atmospheric pressure and intrathoracic pressure are approximately the same at FRC. Consequently, situations where intrathoracic pressure is increased, this may lead to increased intravascular pressure, but normal transmural pressure. There is the potential that this could lead to misclassification of patients.

Pre-capillary pulmonary hypertension (PH) is defined haemodynamically on right heart catheterisation as a mean pulmonary artery pressure (mPAP) of ≥ 25mmHg and pulmonary wedge pressure (PWP) ≤ 15mmHg.(Galie, Humbert et al. 2015) Pre-capillary PH applies to a heterogeneous group of disorders associated with abnormalities in the pulmonary vasculature that restrict blood flow and can lead to right heart failure, but includes pulmonary arterial hypertension (PAH) which is treatable with pulmonary vasodilator therapies.(Galie, Humbert et al. 2015) Post-capillary PH (where mPAP is ≥ 25mmHg and PWP

174

> 15mmHg) is a common condition since left ventricular disease is common in the population, but is treated very differently from pre-capillary PH, by treating the underlying left ventricular disease and avoiding the use of pulmonary vasodilators.

Distinguishing pre-capillary PH from post-capillary PH is therefore critical and requires accurate measurement of PWP at the time of right heart catheterisation. There are many controversies surrounding how best to measure the PWP, but these have largely been settled over the last few years. It is firstly important to obtain an accurate zero level for the transducer, at the mid thoracic level.(Rosenkranz and Preston 2015) Secondly, it is thought best to measure the pressure at FRC, when intrathoracic pressure is close to zero.(Rosenkranz and Preston 2015) However, LeVarge et al. have shown that using measurements of PWP at FRC can lead to misclassification of patients as having post- capillary PH, if an average of the PWP over the whole respiratory cycle is taken as reflective of true PWP.(LeVarge, Pomerantsev et al. 2014) Body mass index correlated with the difference between PWP during inspiration and expiration, which is most likely as a result of increased intrathoracic pressure at end expiration (FRC) due to obesity: to generate inspiration would require a significant generation of negative pressure to overcome such positive intrathoracic pressure and produce inward airflow. It is immediately apparent that simply averaging the PWP over the whole respiratory cycle is unsatisfactory in determining true transmural PWP at end expiration.

In the service at Imperial, it was noted that many patients with significant obesity were presenting with clinical features of pulmonary arterial hypertension and features on echocardiography but did not suggest significant left ventricular dysfunction, namely normal left atrial size and diastolic filling patterns, yet were being classified as post-capillary PH, due to elevated PWP and/or left ventricular end-diastolic pressure (LVEDP). This had the potential to deprive some patients from pulmonary vasodilator therapy when they may

175 derive some benefit. This led to a change in in-house protocol at the time of catheterisation, to include measurements of intra-oesophageal pressure (IOP) to correct for abnormal increases as a result of obesity. The results are presented here.

4.2 Methods

Individuals were clinically obese and in whom non-invasive screening suggested a pre- capillary phenotype underwent invasive right +/- left cardiac catheterisation as part of a standardised investigation algorithm.(Galie, Humbert et al. 2015)

Cardiac catheterisation was undertaken according to our standard protocol, which is outlined in appendix 3.1. In addition, patients were consented for intra-oesophageal pressure measurements.

Siting the balloon catheter

The balloon catheter is passed in the patient in the upright position. We passed the lubricated catheter via the nose into the naso-pharynx where resistance is met. The patient is then instructed to take sips of water, advancing the catheter for each swallow. The catheter is advanced. The removable stilett allows purchase and stiffness in the catheter to ease advancement. The stilet tip is radiolucent to x-rays therefore the tip of the catheter can be accurately delivered to the distal third of the oesophagus utilising fluoroscopic screening in the cardiac catheter laboratory.

At the end of a relaxed exhalation (FRC) and with the mouth open, the alveolar pressure, the pressure at the airway opening, and the atmospheric pressure are equal. In turn, the distending pressure of the lung is equal to pressure inside the alveolar pressure which is equal to atmospheric pressure (as described above) minus the pleural pressure. With the glottis open atmospheric pressure is assumed to equal zero. Ideally measuring pleural pressure would involve passing a catheter directly into the pleural space. Fortunately, the

176 adjacent oesophagus is a passive entity and is able to transmit pleural pressure to a measurement catheter in the lower third of the oesophagus (Figures 4.1 and 4.2).(Benditt

2005)

We aimed to assess whether intra-thoracic pressure (intra-oesophageal pressure) affected cardio-pulmonary haemodynamics therefore we measured a series of pressures during different physiological manoeuvres to look for associations:

 Relaxed breathing and relaxed breath hold (functional residual capacity)

 Valsalva breath hold (raised intrinsic pressure)

 External compression of the abdomen (raised extrinsic pressure)

A Valsalva manoeuvre was performed by moderately forcefully exhaling against a 2.5 ml syringe for five seconds, aiming for a steady state or plateau in the oesophageal pressure with no respiratory variation. Extrinsic compression was done by the second operator (same individual each time to reduce the chance of operator variability) by external pressure delivered to the upper abdomen with two hands for a period of five seconds.

Pressures were measured across the pulmonary circulation during the three separate manoeuvres described. Fluid challenge with rapid saline infusion of 500 ml to 1 litre was administered by the operator if clinically indicated.

177

Figure 4.1. The proximity of the pleura and mid- to lower oesophagus. Adapted from Benditt et al. 2005.

178

Figure 4.2. The anatomical positioning of the oesophageal balloon catheter tip. The catheter has multiple small holes in which is surrounded by a latex balloon that prevents the catheter from being occluded by the oesophageal wall and maintains a column of air in and around the catheter in order to measure pressure in the surrounding structure. The proximal end is attached to pressure transducers that are zeroed to air. Adapted from Benditt et al. 2005.

179

4.3 Results 4.3.1 Demographics Seventeen patients underwent right heart catheterisation with simultaneous intra- oesophageal balloon manometry between Jan 2013 and Dec 2016. Demographics are presented in Table 4.1, with echocardiography data presented in Table 4.2 and haemodynamic data in Table 4.3. Patients have a restrictive pulmonary function, with reduced gas transfer factor, but unlike pure obesity, have normal transfer coefficient.

Overall, the echocardiographic assessment of the patients suggests normal left-sided filling pressures with normal size left atria and no evidence of increase left atrial pressure on the basis of transmitral inflow pattern and E/E’. Haemodynamics are in keeping with a typical diagnosis of post-capillary pulmonary hypertension, however intra-oesophageal pressures are elevated at 12.3 mmHg with normal values in healthy non-obese subjects in the supine position approximately 4 mmHg.(Steier, Lunt et al. 2014)

4.3.2 Example data Representative simultaneous recordings of left ventricular, pulmonary wedge and intra- oesophageal pressure are presented in Figure 4.3. It can be seen how PWP and LV diastolic pressures are strongly influenced by the intra-thoracic pressure, as measured by intra- oesophageal pressure (IOP).

180

Mean SD / range Age (years) 58.15 10.74 WHO FC I/II/III/IV -/-/14/3 Height (metres) 1.68 0.10 Weight (kg) 119.8 19.4 Body Mass Index 42.4 7.0 FEV1 % 54.8 13.2 VC % 60.1 19.2 FEV1/VC 67.3 26.5 TLC % 67.7 16.2 Residual Volume % 90.2 51.0 TLCO% 49.9 15.3 KCO% 82.7 18.3 pH 7.43 7.36-7.56 PaO2 (kPa) 7.7 6.2-10.3 PaCO2 (kPa) 5.3 3.9-8.1

Table 4.1 Demographic data for patients (n=17). Data are presented as mean and standard deviation (SD). World Health Organisation (WHO), functional class (FC), forced expiratory volume in one second (FEV1), vital capacity (VC), total lung capacity (TLC), transfer factor / coefficient of the lung for carbon monoxide (TLCO / KCO).

181

Mean SD LA diameter (cm) 4.12 0.56 Mitral E/E' Lateral 6.91 3.38 Mitral E/E' Septal 11.07 4.13 Mitral E/E' average 8.57 3.61 Transmitral E:A ratio 1.03 0.37 Tricuspid Velocity (cm/s) 360.87 73.68 RVSP (mmHg) 59.15 30.66 LV diastolic dysfunction (grade) 1.25 0.70 LV hypertrophy 2(mild) / 17

Table 4.2 Echocardiographic data for patients (n=17). Data are presented as mean and standard deviation (SD). Left atrium (LA), right ventricular systolic pressure (RVSP), left ventricle (LV).

Mean SD Right atrial pressure (mmHg) 13.9 6.5 RV systolic pressure (mmHg) 69.8 18.8 RV end diastolic pressure (mmHg) 16.6 6.1 PA systolic pressure (mmHg) 69.6 18.7 PA diastolic pressure (mmHg) 27.9 8.3 PA mean pressure (mmHg) 45.3 10.0 PWP mean (mmHg) 19.7 7.5 IOP (mmHg) 12.3 4.7 LV systolic pressure (mmHg) 138.7 25.1 LV end diastolic pressure (mmHg) 18.8 9.7 Pulmonary flow index (l/min/m2) 2.7 1.4 Pulmonary vascular resistance (WU) 5.7 4.2

Table 4.3 Haemodynamic data for patients (n=17). Data are presented as mean and standard deviation (SD). Left atrium (LA), right atrium (RA), right ventricle (RV), intra-oesophageal pressure (IOP), pulmonary artery (PA), pulmonary capillary wedge pressure (PCWP), left ventricle (LV), wood units (WU).

182

Figure 4.3. Haemodynamic traces for a female patient, aged 69, with a body mass index of 55 kg/m^2 showing relaxed breathing. Left ventricle (LV), pulmonary wedge (PCW), intra- oesophageal pressure (LA on haemodynamic panel).

183

4.3.3 Association with intrathoracic pressure and measured PWP No clear linear relationship between BMI and intrathoracic pressure was found (Figure 4.4), although there appeared to be a trend. However, there was a clear direct relationship (almost 1:1) between intrathoracic pressure, as assessed by IOP, and PWP (p < 0.01) (Figure 4.5). Fewer data were collected for LVEDP, although the results were similar (data not shown). When increased thoracic pressure was induced through Valsalva or extrathoracic application of pressure, there was again a tight relationship between IOP and PWP (Figure 4.6).

The phenotype was further challenged with a fluid challenge of 500 ml to 1 litre in 8 patients with no clear changes in PWP (Figure 4.7). Only one patient had a significant rise in PWP with fluid from 6 to 17 mmHg. When PWP was corrected for intrathoracic pressure, by subtracting intra-oesphageal pressure less 4 mmHg (normal pressure), all but three patients had a “corrected” PWP of ≤ 15 mmHg. One patient had a value less than zero and this was the same patient who increased PWP by 11 mmHg with fluid, suggesting that they may have been dry at the time of catheterisation.

184

Figure 4.4. Relationship between body mass index (BMI, kg/m^2) and intrathoracic pressure as assessed by intra-oesophageal pressure (IOP, mmHg).

PWP vs IOP y = 1.0872x + 6.125 R² = 0.374 45 40

30 25 20

PWP(mmHg) 15 10 5 0 0 5 10 15 20 25 IOP (mmHg)

Figure 4.5. Relationship between intrathoracic pressure as assessed by intra-oesophageal pressure (IOP, mmHg) and pulmonary wedge pressure (PWP, mmHg), p < 0.01.

185

Figure 4.6. Relationship between change in intrathoracic pressure as assessed by intra- oesophageal pressure (IOP, mmHg) and pulmonary wedge pressure (PWP, mmHg), during Valsalva manoeuvre (A) and extrathoracic applied pressure (B).

186

Pre and post fluid challenge 45 40 35

30 25 20

PWP(mmHg) 15 10 5 0 1 2

Figure 4.7. Pulmonary wedge pressure (PWP, mmHg) before (1) and after (2) fluid challenge in 8 patients.

PWP "corrected for IOP" 45 40 35

25 20

15 PWP(mmHg) 10 5 0 -5

Figure 4.8. Pulmonary wedge pressure (PWP, mmHg) corrected for intra-oesophageal pressure (IOP). Solid black line represents 15 mmHg cut-off for pre- versus post-capillary PH.

187

4.4 Discussion It has been acknowledged in the context of high intrathoracic pressures present in COPD that spurious elevations in pulmonary wedge pressure may lead to a misdiagnosis of combined pre-capillary and post-capillary pulmonary hypertension.(Boerrigter, Waxman et al. 2014) Since pulmonary hypertension is not treatable in this context in any case as a result of the COPD, misclassification does not have the serious consequences of denying access to pulmonary vasodilator therapy. It is also been recently documented that elevated intrathoracic pressure due to obesity can result in misclassification of pre- versus post- capillary pulmonary hypertension.(Jawad, Tonelli et al. 2017)

The data that have been accrued as part of this study demonstrate more robustly that intrathoracic pressure elevations in obesity may lead to misclassification of pulmonary hypertension. The patients had echocardiographic data supportive of normal left sided filling pressures, with normal left atrial size and normal transmitral filling patterns. Moreover, when the patients were challenged with intravenous fluid, significant increases were generally not observed in pulmonary wedge pressure, furthermore confirming normal left ventricular compliance. 14 out of the 17 patients tested could be reclassified as pre-capillary pulmonary hypertension once correction of intrathoracic pressure was applied, assuming a normal intrathoracic pressure of 4 mmHg.

Of note, it was possible to demonstrate a direct, 1:1, relationship between increases and intrathoracic pressure and intravascular pressure at FRC. This would suggest, in theory that one could estimate the intrathoracic pressure from the relative difference between inspiratory and expiratory wedge pressure and use this to correct the wedge pressure, taking into account a normal pressure of 4 mmHg, without undertaking intra-oesophageal measurements. Further work is required to validate this.

188

Conclusions and future work

In the introduction to this thesis, I have presented some compelling data on the rationale to study the role of iron in pulmonary hypertension. Work in healthy humans shows that iron plays a key role in vascular tone in response to hypoxia and animal models suggest that iron deficiency in itself can induce marked pulmonary vascular remodelling. Previous work has shown a high prevalence of iron deficiency (ID) in patients with idiopathic pulmonary arterial hypertension, but this requires access to a specialised biomarker, the soluble transferrin receptor (sTfR), which is not available outside standard clinical pathology labs. The first aim of my thesis was to find the best routinely-available surrogate to identify ID as defined by elevated sTfR levels. I was able to validate that RDW was indeed a robust predictor of ID in IPAH. Elevated RDW being has been associated with ID in an array of other pathological conditions where there has been similar debate to what is the best single routinely-available clinical marker for ID.

In a distinct population of IPAH iron derived, cut-offs were established using a balance of sensitivity and specificity for other established everyday markers of ID including ferritin, iron, transferrin saturation and RDW. Whilst RDW showed closer association with the other parameters for predicting sTfR, I ensured that other factors such has inflammation were not a key co-founder given the well documented link with RDW and inflammatory conditions and in particular the inflammatory component of PAH. By correcting for IL-6 and CRP I have shown that RDW still outperforms other variables for predicting sTfR. This is important because I demonstrated that sTfR has a very strong correlation with mortality and in turn, RDW may also have survival implications. Consistently throughout these data showed that anaemia was not a major influencer or confounding factor.

The natural progression was to look at a large incident PAH and CTEPH population to see if these were consistent findings. I was interested to understand if ID had strong associations with mechanistic implications to disease severity, and whether these could be more of a core (i.e., cardiopulmonary) or peripheral (i.e, skeletal muscle) mechanisms. In the PAH subtypes RDW was independently associated with exercise, 6MWD and cardiopulmonary exercise testing, but not with haemodynamics or cardiac function on cardiac magnetic resonance (CMR). This does imply that manipulation of iron in IPAH may be more likely to impact on skeletal muscle function, and supports by other open-label studies of iron

189 supplementation, which have shown changes in skeletal muscle oxidative metabolism. The flip side of this is that iron supplementation may not in fact modulate the pulmonary circulation in human IPAH.

In relation to survival, I was only able to undertake univariate analysis and somewhat surprisingly, right heart function and pulmonary haemodynamics were much less impressive than 6MWD and exercise parameters. The intriguing findings were the associations of RDW in CTEPH with functional parameters across CMR, haemodynamics and exercise. To date, the literature is quite limited on the role of iron in CTEPH. This raises an area for future study. It may be that anticoagulation at diagnosis in patients with CTEPH allows for a greater spread in iron status and thus produces more robust associations. When assessing the relationship with survival, however, there are significant difficulties, since a large proportion of patients undergo surgery for CTEPH with near curative results. Consequently, we would need to look in more detail at inoperable patients, who receive medical therapies, to assess if there was a signal in terms of ID and survival. Interpreting the role of iron within this large range of treatment options poses challenging logistics.

Survival analysis in IPAH showed clearly that RDW has an association on survival both on univariate Cox regression and Kaplan-Meier survival curves. However, despite going to extensive lengths to collect this high fidelity data, analysing all cardiac catheter haemodynamic traces and having performed a significant proportion of them, I was unable to perform multi-variate survival analysis as the timings of all investigations for each modality may have not have overlapped sufficiently, in turn producing too few numbers of variables and number of deaths to run a multi-variate Cox regression analysis. This is a major limitation in this study. Many biomarker studies are performed in prevalent disease groups for this reason where there may be more than one opportunity in a patient’s journey to collate all the data required. Another solution is to use data from different time points, but this may result in significant errors in analysis and lead to false conclusions.

The clinical trial provides a follow on to the work described here and earlier work by the group. Having identified and established iron deficiency in IPAH I was the main recruiter for the first randomised, placebo-controlled, double blinded trial of intravenous iron in patients in IPAH, whilst overseeing and undertaking all the cardiac catheterisations and cardiopulmonary exercise testing. Clearly, conclusions at this stage must be drawn

190 cautiously but an early signal suggests that iron supplementation may replete iron stores and may have influence on exercise, also raising a narrative that the beneficial role of iron may be more peripheral than central. Clearly, no further conclusions can be made at this stage but once the trial is complete and blinding lifted it will be fascinating to assess these novel data.

Finally, during my period in clinical research, while in the cardiac catheter laboratory, the challenge of making an accurate diagnosis in obese patients with pulmonary hypertension arose. While outside the theme of this these, Chapter 4 highlights the crucial role of accurate phenotyping both in clinical practice and research, the latter being relevant to all studies where phenotype is being correlated with ‘omics and outcomes. Having performed well over 100 right +/- left cardiac catheterisations, we had the opportunity to look at other invasive markers of interest. Obesity is an epidemic of our time and clearly a major impact on health and health provision. We have demonstrated that obese patients may be being misclassified at the point of cardiac catheterisation having questioned whether intrathoracic pressure had an influence on pulmonary wedge pressures. Although numbers are on the small size the data is suggestive that there is a relationship between obesity, intrathoracic pressure, and pulmonary wedge pressure. This could have major implications on classification of pulmonary hypertension in these patients and potentially could change current guidelines in due course.

191

References

2015. "2015 ESC/ERS Guidelines for the diagnosis and treatment of pulmonary hypertension. The Joint Task Force for the Diagnosis and Treatment of Pulmonary Hypertension of the European Society of Cardiology (ESC) and the European Respiratory Society (ERS)." Nazzareno Galiè, Marc Humbert, Jean-Luc Vachiery, Simon Gibbs, Irene Lang, Adam Torbicki, Gérald Simonneau, Andrew Peacock, Anton Vonk Noordegraaf, Maurice Beghetti, Ardeschir Ghofrani, Miguel Angel Gomez Sanchez, Georg Hansmann, Walter Klepetko, Patrizio Lancellotti, Marco Matucci, Theresa McDonagh, Luc A. Pierard, Pedro T. Trindade, Maurizio Zompatori and Marius Hoeper. Eur Respir J 2015; 46: 903-975. Eur Respir J, 46, 1855-6. ABUL, Y., OZSU, S., KORKMAZ, A., BULBUL, Y., OREM, A. & OZLU, T. 2014. Red cell distribution width: a new predictor for chronic thromboembolic pulmonary hypertension after pulmonary embolism. Chron Respir Dis, 11, 73-81. AGUIRRE, J. I., MORRELL, N. W., LONG, L., CLIFT, P., UPTON, P. D., POLAK, J. M. & WILKINS, M. R. 2000. Vascular remodeling and ET-1 expression in rat strains with different responses to chronic hypoxia. Am J Physiol Lung Cell Mol Physiol, 278, L981-7. AHTING, U., MAYR, J. A., VANLANDER, A. V., HARDY, S. A., SANTRA, S., MAKOWSKI, C., ALSTON, C. L., ZIMMERMANN, F. A., ABELA, L., PLECKO, B., ROHRBACH, M., SPRANGER, S., SENECA, S., ROLINSKI, B., HAGENDORFF, A., HEMPEL, M., SPERL, W., MEITINGER, T., SMET, J., TAYLOR, R. W., VAN COSTER, R., FREISINGER, P., PROKISCH, H. & HAACK, T. B. 2015b. Clinical, biochemical, and genetic spectrum of seven patients with NFU1 deficiency. Front Genet, 6, 123. AJIOKA, R. S., PHILLIPS, J. D. & KUSHNER, J. P. 2006. Biosynthesis of heme in mammals. Biochim Biophys Acta, 1763, 723-36. AKARSU, S., TASKIN, E., YILMAZ, E., YILMAZ, H., KILIC, M. & AYGUN, A. D. 2006. Treatment of iron deficiency anemia with intravenous iron preparations. Acta Haematol, 116, 51- 7. ALDRED, M. A., COMHAIR, S. A., VARELLA-GARCIA, M., ASOSINGH, K., XU, W., NOON, G. P., THISTLETHWAITE, P. A., TUDER, R. M., ERZURUM, S. C., GERACI, M. W. & COLDREN, C. D. 2010. Somatic chromosome abnormalities in the lungs of patients with pulmonary arterial hypertension. Am J Respir Crit Care Med, 182, 1153-60. ALDRED, M. A., MACHADO, R. D., JAMES, V., MORRELL, N. W. & TREMBATH, R. C. 2007. Characterization of the BMPR2 5'-untranslated region and a novel mutation in

192

pulmonary hypertension. Am J Respir Crit Care Med, 176, 819-24. ALLEN, L. A., FELKER, G. M., MEHRA, M. R., CHIONG, J. R., DUNLAP, S. H., GHALI, J. K., LENIHAN, D. J., OREN, R. M., WAGONER, L. E., SCHWARTZ, T. A. & ADAMS, K. F. 2010a. Validation and potential mechanisms of red cell distribution width as a prognostic marker in heart failure. J Card Fail, 16, 230-8. AL-NAJJAR, Y., GOODE, K. M., ZHANG, J., CLELAND, J. G. & CLARK, A. L. 2009. Red cell distribution width: an inexpensive and powerful prognostic marker in heart failure. Eur J Heart Fail, 11, 1155-62. AMMASH, N. & WARNES, C. A. 1996. Cerebrovascular events in adult patients with cyanotic congenital heart disease. J Am Coll Cardiol, 28, 768-72. ANDERSON, C. P., SHEN, M., EISENSTEIN, R. S. & LEIBOLD, E. A. 2012. Mammalian iron metabolism and its control by iron regulatory proteins. Biochim Biophys Acta, 1823, 1468-83. ANDRIOPOULOS, B., CORRADINI, E., XIA, Y., FAASSE, S. A., CHEN, S., GRGUREVIC, L., KNUTSON, M. D., PIETRANGELO, A., VUKICEVIC, S., LIN, H. Y. & BABITT, J. L. 2009. BMP6 is a key endogenous regulator of hepcidin expression and iron metabolism. Nat Genet, 41, 482-7. ANKER, S. D., COMIN COLET, J., FILIPPATOS, G., WILLENHEIMER, R., DICKSTEIN, K., DREXLER, H., LÜSCHER, T. F., BART, B., BANASIAK, W., NIEGOWSKA, J., KIRWAN, B. A., MORI, C., VON EISENHART ROTHE, B., POCOCK, S. J., POOLE-WILSON, P. A., PONIKOWSKI, P. & INVESTIGATORS, F.-H. T. 2009. Ferric carboxymaltose in patients with heart failure and iron deficiency. N Engl J Med, 361, 2436-48. ANZOVINO, A., LANE, D. J., HUANG, M. L. & RICHARDSON, D. R. 2014. Fixing frataxin: 'ironing out' the metabolic defect in Friedreich's ataxia. Br J Pharmacol, 171, 2174-90. ARNDT, S., MAEGDEFRAU, U., DORN, C., SCHARDT, K., HELLERBRAND, C. & BOSSERHOFF, A. K. 2010. Iron-induced expression of bone morphogenic protein 6 in intestinal cells is the main regulator of hepatic hepcidin expression in vivo. Gastroenterology, 138, 372-82. ASHBY, D. R., GALE, D. P., BUSBRIDGE, M., MURPHY, K. G., DUNCAN, N. D., CAIRNS, T. D., TAUBE, D. H., BLOOM, S. R., TAM, F. W., CHAPMAN, R., MAXWELL, P. H. & CHOI, P. 2010. Erythropoietin administration in humans causes a marked and prolonged reduction in circulating hepcidin. Haematologica, 95, 505-8. ASHBY, D. R., GALE, D. P., BUSBRIDGE, M., MURPHY, K. G., DUNCAN, N. D., CAIRNS, T. D., TAUBE, D. H., BLOOM, S. R., TAM, F. W., CHAPMAN, R. S., MAXWELL, P. H. & CHOI, P.

193

2009. Plasma hepcidin levels are elevated but responsive to erythropoietin therapy in renal disease. Kidney Int, 75, 976-81. AUNG, N., LING, H. Z., CHENG, A. S., AGGARWAL, S., FLINT, J., MENDONCA, M., RASHID, M., KANG, S., WEISSERT, S., COATS, C. J., RICHARDS, T., THOMAS, M., WOLDMAN, S. & OKONKO, D. O. 2013. Expansion of the red cell distribution width and evolving iron deficiency as predictors of poor outcome in chronic heart failure. Int J Cardiol, 168, 1997-2002. AUSTIN, E. D. & LOYD, J. E. 2014. The genetics of pulmonary arterial hypertension. Circ Res, 115, 189-202. AUSTIN, E. D., MA, L., LEDUC, C., BERMAN ROSENZWEIG, E., BORCZUK, A., PHILLIPS, J. A., PALOMERO, T., SUMAZIN, P., KIM, H. R., TALATI, M. H., WEST, J., LOYD, J. E. & CHUNG, W. K. 2012. Whole exome sequencing to identify a novel gene (caveolin-1) associated with human pulmonary arterial hypertension. Circ Cardiovasc Genet, 5, 336-43. AVNI, T., BIEBER, A., GROSSMAN, A., GREEN, H., LEIBOVICI, L. & GAFTER-GVILI, A. 2015. The safety of intravenous iron preparations: systematic review and meta-analysis. Mayo Clin Proc, 90, 12-23. AVOUAC, J., AIRÒ, P., MEUNE, C., BERETTA, L., DIEUDE, P., CARAMASCHI, P., TIEV, K., CAPPELLI, S., DIOT, E., VACCA, A., CRACOWSKI, J. L., SIBILIA, J., KAHAN, A., MATUCCI- CERINIC, M. & ALLANORE, Y. 2010. Prevalence of pulmonary hypertension in systemic sclerosis in European Caucasians and metaanalysis of 5 studies. J Rheumatol, 37, 2290-8. BABITT, J. L., HUANG, F. W., WRIGHTING, D. M., XIA, Y., SIDIS, Y., SAMAD, T. A., CAMPAGNA, J. A., CHUNG, R. T., SCHNEYER, A. L., WOOLF, C. J., ANDREWS, N. C. & LIN, H. Y. 2006. Bone morphogenetic protein signaling by hemojuvelin regulates hepcidin expression. Nat Genet, 38, 531-9. BABITT, J. L., HUANG, F. W., XIA, Y., SIDIS, Y., ANDREWS, N. C. & LIN, H. Y. 2007a. Modulation of bone morphogenetic protein signaling in vivo regulates systemic iron balance. J Clin Invest, 117, 1933-9. BABU, A. S., MYERS, J., ARENA, R., MAIYA, A. G. & PADMAKUMAR, R. 2013. Evaluating exercise capacity in patients with pulmonary arterial hypertension. Expert Rev Cardiovasc Ther, 11, 729-37. BAGGEN, V. J., LEINER, T., POST, M. C., VAN DIJK, A. P., ROOS-HESSELINK, J. W., BOERSMA, E., HABETS, J. & SIESWERDA, G. T. 2016. Cardiac magnetic resonance findings

194

predicting mortality in patients with pulmonary arterial hypertension: a systematic review and meta-analysis. Eur Radiol, 26, 3771-3780. BALANOS, G. M., DORRINGTON, K. L. & ROBBINS, P. A. 2002. Desferrioxamine elevates pulmonary vascular resistance in humans: potential for involvement of HIF-1. J Appl Physiol (1985), 92, 2501-7. BALL, M. K., WAYPA, G. B., MUNGAI, P. T., NIELSEN, J. M., CZECH, L., DUDLEY, V. J., BEUSSINK, L., DETTMAN, R. W., BERKELHAMER, S. K., STEINHORN, R. H., SHAH, S. J. & SCHUMACKER, P. T. 2014. Regulation of hypoxia-induced pulmonary hypertension by vascular smooth muscle hypoxia-inducible factor-1α. Am J Respir Crit Care Med, 189, 314-24. BARST, R. J., CHUNG, L., ZAMANIAN, R. T., TURNER, M. & MCGOON, M. D. 2013. Functional class improvement and 3-year survival outcomes in patients with pulmonary arterial hypertension in the REVEAL Registry. Chest, 144, 160-168. BATT, J., AHMED, S. S., CORREA, J., BAIN, A. & GRANTON, J. 2014. Skeletal muscle dysfunction in idiopathic pulmonary arterial hypertension. Am J Respir Cell Mol Biol, 50, 74-86. BAUER, R., DEHNERT, C., SCHOENE, P., FILUSCH, A., BÄRTSCH, P., BORST, M. M., KATUS, H. A. & MEYER, F. J. 2007. Skeletal muscle dysfunction in patients with idiopathic pulmonary arterial hypertension. Respir Med, 101, 2366-9. BEAUMONT, C., LENEUVE, P., DEVAUX, I., SCOAZEC, J. Y., BERTHIER, M., LOISEAU, M. N., GRANDCHAMP, B. & BONNEAU, D. 1995. Mutation in the iron responsive element of the L ferritin mRNA in a family with dominant hyperferritinaemia and cataract. Nat Genet, 11, 444-6. BENDITT, J. O. 2005. Esophageal and gastric pressure measurements. Respir Care, 50, 68-75; discussion 75-77. BENYAMIN, B., FERREIRA, M. A., WILLEMSEN, G., GORDON, S., MIDDELBERG, R. P., MCEVOY, B. P., HOTTENGA, J. J., HENDERS, A. K., CAMPBELL, M. J., WALLACE, L., FRAZER, I. H., HEATH, A. C., DE GEUS, E. J., NYHOLT, D. R., VISSCHER, P. M., PENNINX, B. W., BOOMSMA, D. I., MARTIN, N. G., MONTGOMERY, G. W. & WHITFIELD, J. B. 2009a. Common variants in TMPRSS6 are associated with iron status and erythrocyte volume. Nat Genet, 41, 1173-5. BENYAMIN, B., MCRAE, A. F., ZHU, G., GORDON, S., HENDERS, A. K., PALOTIE, A., PELTONEN, L., MARTIN, N. G., MONTGOMERY, G. W., WHITFIELD, J. B. & VISSCHER, P. M. 2009b. Variants in TF and HFE explain approximately 40% of genetic variation in serum-

195

transferrin levels. Am J Hum Genet, 84, 60-5. BENZA, R. L., GOMBERG-MAITLAND, M., DEMARCO, T., FROST, A. E., TORBICKI, A., LANGLEBEN, D., PULIDO, T., CORREA-JAQUE, P., PASSINEAU, M. J., WIENER, H. W., TAMARI, M., HIROTA, T., KUBO, M. & TIWARI, H. K. 2015. Endothelin-1 Pathway Polymorphisms and Outcomes in Pulmonary Arterial Hypertension. Am J Respir Crit Care Med, 192, 1345-54. BENZA, R. L., MILLER, D. P., BARST, R. J., BADESCH, D. B., FROST, A. E. & MCGOON, M. D. 2012. An evaluation of long-term survival from time of diagnosis in pulmonary arterial hypertension from the REVEAL Registry. Chest, 142, 448-456. BENZA, R. L., MILLER, D. P., GOMBERG-MAITLAND, M., FRANTZ, R. P., FOREMAN, A. J., COFFEY, C. S., FROST, A., BARST, R. J., BADESCH, D. B., ELLIOTT, C. G., LIOU, T. G. & MCGOON, M. D. 2010a. Predicting survival in pulmonary arterial hypertension: insights from the Registry to Evaluate Early and Long-Term Pulmonary Arterial Hypertension Disease Management (REVEAL). Circulation, 122, 164-72. BIENERTOVA-VASKU, J., NOVAK, J. & VASKU, A. 2015. MicroRNAs in pulmonary arterial hypertension: pathogenesis, diagnosis and treatment. J Am Soc Hypertens, 9, 221- 34. BOERRIGTER, B. G., WAXMAN, A. B., WESTERHOF, N., VONK-NOORDEGRAAF, A. & SYSTROM, D. M. 2014. Measuring central pulmonary pressures during exercise in COPD: how to cope with respiratory effects. Eur Respir J, 43, 1316-25. BREDA, A. P., PEREIRA DE ALBUQUERQUE, A. L., JARDIM, C., MORINAGA, L. K., SUESADA, M. M., FERNANDES, C. J., DIAS, B., LOURENÇO, R. B., SALGE, J. M. & SOUZA, R. 2014. Skeletal muscle abnormalities in pulmonary arterial hypertension. PLoS One, 9, e114101. BREGMAN, D. B., MORRIS, D., KOCH, T. A., HE, A. & GOODNOUGH, L. T. 2013. Hepcidin levels predict nonresponsiveness to oral iron therapy in patients with iron deficiency anemia. Am J Hematol, 88, 97-101. BRIDLE, K. R., FRAZER, D. M., WILKINS, S. J., DIXON, J. L., PURDIE, D. M., CRAWFORD, D. H., SUBRAMANIAM, V. N., POWELL, L. W., ANDERSON, G. J. & RAMM, G. A. 2003. Disrupted hepcidin regulation in HFE-associated haemochromatosis and the liver as a regulator of body iron homoeostasis. Lancet, 361, 669-73. BRISSOT, P., ROPERT, M., LE LAN, C. & LORÉAL, O. 2012. Non-transferrin bound iron: a key role in iron overload and iron toxicity. Biochim Biophys Acta, 1820, 403-10. BROBERG, C. S., BAX, B. E., OKONKO, D. O., RAMPLING, M. W., BAYNE, S., HARRIES, C.,

196

DAVIDSON, S. J., UEBING, A., KHAN, A. A., THEIN, S., GIBBS, J. S., BURMAN, J. & GATZOULIS, M. A. 2006. Blood viscosity and its relationship to iron deficiency, symptoms, and exercise capacity in adults with cyanotic congenital heart disease. J Am Coll Cardiol, 48, 356-65. BROCK, M., TRENKMANN, M., GAY, R. E., MICHEL, B. A., GAY, S., FISCHLER, M., ULRICH, S., SPEICH, R. & HUBER, L. C. 2009. Interleukin-6 modulates the expression of the bone morphogenic protein receptor type II through a novel STAT3-microRNA cluster 17/92 pathway. Circ Res, 104, 1184-91. BROWN, R. B. & RAZZAQUE, M. S. 2015a. Dysregulation of phosphate metabolism and conditions associated with phosphate toxicity. Bonekey Rep, 4, 705. BROWN, R. B. & RAZZAQUE, M. S. 2015c. Dysregulation of phosphate metabolism and conditions associated with phosphate toxicity. Bonekey Rep, 4, 705. BROWN, S. B., RAINA, A., KATZ, D., SZERLIP, M., WIEGERS, S. E. & FORFIA, P. R. 2011. Longitudinal shortening accounts for the majority of right ventricular contraction and improves after pulmonary vasodilator therapy in normal subjects and patients with pulmonary arterial hypertension. Chest, 140, 27-33. BROWNLIE, T., UTERMOHLEN, V., HINTON, P. S. & HAAS, J. D. 2004. Tissue iron deficiency without anemia impairs adaptation in endurance capacity after aerobic training in previously untrained women. Am J Clin Nutr, 79, 437-43. BURG, E. D., REMILLARD, C. V. & YUAN, J. X. 2008a. Potassium channels in the regulation of pulmonary artery smooth muscle cell proliferation and apoptosis: pharmacotherapeutic implications. Br J Pharmacol, 153 Suppl 1, S99-S111. BUSBRIDGE, M., GRIFFITHS, C., ASHBY, D., GALE, D., JAYANTHA, A., SANWAIYA, A. & CHAPMAN, R. S. 2009. Development of a novel immunoassay for the iron regulatory peptide hepcidin. Br J Biomed Sci, 66, 150-7. CAIRO, G., BERNUZZI, F. & RECALCATI, S. 2006. A precious metal: Iron, an essential nutrient for all cells. Genes Nutr, 1, 25-39. CAMASCHELLA, C. 2015. Iron-Deficiency Anemia. N Engl J Med, 373, 485-6. CAMPUZANO, V., MONTERMINI, L., LUTZ, Y., COVA, L., HINDELANG, C., JIRALERSPONG, S., TROTTIER, Y., KISH, S. J., FAUCHEUX, B., TROUILLAS, P., AUTHIER, F. J., DÜRR, A., MANDEL, J. L., VESCOVI, A., PANDOLFO, M. & KOENIG, M. 1997. Frataxin is reduced in Friedreich ataxia patients and is associated with mitochondrial membranes. Hum Mol Genet, 6, 1771-80. CASSAVAUGH, J. M., HALE, S. A., WELLMAN, T. L., HOWE, A. K., WONG, C. & LOUNSBURY, K.

197

M. 2011. Negative regulation of HIF-1α by an FBW7-mediated degradation pathway during hypoxia. J Cell Biochem, 112, 3882-90. CHAMBERS, J. C., ZHANG, W., LI, Y., SEHMI, J., WASS, M. N., ZABANEH, D., HOGGART, C., BAYELE, H., MCCARTHY, M. I., PELTONEN, L., FREIMER, N. B., SRAI, S. K., MAXWELL, P. H., STERNBERG, M. J., RUOKONEN, A., ABECASIS, G., JARVELIN, M. R., SCOTT, J., ELLIOTT, P. & KOONER, J. S. 2009. Genome-wide association study identifies variants in TMPRSS6 associated with hemoglobin levels. Nat Genet, 41, 1170-2. CHAOUAT, A., COULET, F., FAVRE, C., SIMONNEAU, G., WEITZENBLUM, E., SOUBRIER, F. & HUMBERT, M. 2004. Endoglin germline mutation in a patient with hereditary haemorrhagic telangiectasia and dexfenfluramine associated pulmonary arterial hypertension. Thorax, 59, 446-8. CHEN, S. L., ZHANG, H., XIE, D. J., ZHANG, J., ZHOU, L., ROTHMAN, A. M. & STONE, G. W. 2015. Hemodynamic, functional, and clinical responses to pulmonary artery denervation in patients with pulmonary arterial hypertension of different causes: phase II results from the Pulmonary Artery Denervation-1 study. Circ Cardiovasc Interv, 8, e002837. CHERTOW, G. M. & WINKELMAYER, W. C. 2010. On the relative safety of intravenous iron formulations: new answers, new questions. Am J Hematol, 85, 643-4. CHIABRANDO, D., MARRO, S., MERCURIO, S., GIORGI, C., PETRILLO, S., VINCHI, F., FIORITO, V., FAGOONEE, S., CAMPOREALE, A., TURCO, E., MERLO, G. R., SILENGO, L., ALTRUDA, F., PINTON, P. & TOLOSANO, E. 2012. The mitochondrial heme exporter FLVCR1b mediates erythroid differentiation. J Clin Invest, 122, 4569-79. CHIN, K. M. & RUBIN, L. J. 2008. Pulmonary arterial hypertension. J Am Coll Cardiol, 51, 1527-38. CHRISTMAN, B. W., MCPHERSON, C. D., NEWMAN, J. H., KING, G. A., BERNARD, G. R., GROVES, B. M. & LOYD, J. E. 1992a. An imbalance between the excretion of thromboxane and prostacyclin metabolites in pulmonary hypertension. N Engl J Med, 327, 70-5. CHRISTMAN, B. W., MCPHERSON, C. D., NEWMAN, J. H., KING, G. A., BERNARD, G. R., GROVES, B. M. & LOYD, J. E. 1992b. An imbalance between the excretion of thromboxane and prostacyclin metabolites in pulmonary hypertension. N Engl J Med, 327, 70-5. CHRISTMAN, J. W., BLACKWELL, T. R., COWAN, H. B., SHEPHERD, V. L. & RINALDO, J. E. 1992c. Endotoxin induces the expression of macrophage inflammatory protein 1

198

alpha mRNA by rat alveolar and bone marrow-derived macrophages. Am J Respir Cell Mol Biol, 7, 455-61. CHUNG, B., CHASTON, T., MARKS, J., SRAI, S. K. & SHARP, P. A. 2009. Hepcidin decreases iron transporter expression in vivo in mouse duodenum and spleen and in vitro in THP-1 macrophages and intestinal Caco-2 cells. J Nutr, 139, 1457-62. CONDLIFFE, R., KIELY, D. G., PEACOCK, A. J., CORRIS, P. A., GIBBS, J. S., VRAPI, F., DAS, C., ELLIOT, C. A., JOHNSON, M., DESOYZA, J., TORPY, C., GOLDSMITH, K., HODGKINS, D., HUGHES, R. J., PEPKE-ZABA, J. & COGHLAN, J. G. 2009. Connective tissue disease- associated pulmonary arterial hypertension in the modern treatment era. Am J Respir Crit Care Med, 179, 151-7. COOK, J. D., DASSENKO, S. & SKIKNE, B. S. 1990. Serum transferrin receptor as an index of iron absorption. Br J Haematol, 75, 603-9. COOL, C. D., KENNEDY, D., VOELKEL, N. F. & TUDER, R. M. 1997. Pathogenesis and evolution of plexiform lesions in pulmonary hypertension associated with scleroderma and human immunodeficiency virus infection. Hum Pathol, 28, 434-42. CORBIN, J. D. & FRANCIS, S. H. 1999. Cyclic GMP phosphodiesterase-5: target of sildenafil. J Biol Chem, 274, 13729-32. CORBIN, J. D., BEASLEY, A., BLOUNT, M. A. & FRANCIS, S. H. 2005. High lung PDE5: a strong basis for treating pulmonary hypertension with PDE5 inhibitors. Biochem Biophys Res Commun, 334, 930-8. CORTI, M. C., GURALNIK, J. M., SALIVE, M. E., FERRUCCI, L., PAHOR, M., WALLACE, R. B. & HENNEKENS, C. H. 1997. Serum iron level, coronary artery disease, and all-cause mortality in older men and women. Am J Cardiol, 79, 120-7. COTRONEO, E., ASHEK, A., WANG, L., WHARTON, J., DUBOIS, O., BOZORGI, S., BUSBRIDGE, M., ALAVIAN, K. N., WILKINS, M. R. & ZHAO, L. 2015. Iron homeostasis and pulmonary hypertension: iron deficiency leads to pulmonary vascular remodeling in the rat. Circ Res, 116, 1680-90. DAI, Z., LI, M., WHARTON, J., ZHU, M. M. & ZHAO, Y. Y. 2016. Prolyl-4 Hydroxylase 2 (PHD2) Deficiency in Endothelial Cells and Hematopoietic Cells Induces Obliterative Vascular Remodeling and Severe Pulmonary Arterial Hypertension in Mice and Humans Through Hypoxia-Inducible Factor-2α. Circulation, 133, 2447-58. DAMICO, R., KOLB, T. M., VALERA, L., WANG, L., HOUSTEN, T., TEDFORD, R. J., KASS, D. A., RAFAELS, N., GAO, L., BARNES, K. C., BENZA, R. L., RAND, J. L., HAMID, R., LOYD, J. E., ROBBINS, I. M., HEMNES, A. R., CHUNG, W. K., AUSTIN, E. D., DRUMMOND, M. B.,

199

MATHAI, S. C. & HASSOUN, P. M. 2015. Serum endostatin is a genetically determined predictor of survival in pulmonary arterial hypertension. Am J Respir Crit Care Med, 191, 208-18. DARSHAN, D., FRAZER, D. M., WILKINS, S. J. & ANDERSON, G. J. 2010. Severe iron deficiency blunts the response of the iron regulatory gene Hamp and pro-inflammatory cytokines to lipopolysaccharide. Haematologica, 95, 1660-7. DAVIE, N., HALEEN, S. J., UPTON, P. D., POLAK, J. M., YACOUB, M. H., MORRELL, N. W. & WHARTON, J. 2002. ET(A) and ET(B) receptors modulate the proliferation of human pulmonary artery smooth muscle cells. Am J Respir Crit Care Med, 165, 398-405. DE DOMENICO, I., ZHANG, T. Y., KOENING, C. L., BRANCH, R. W., LONDON, N., LO, E., DAYNES, R. A., KUSHNER, J. P., LI, D., WARD, D. M. & KAPLAN, J. 2010. Hepcidin mediates transcriptional changes that modulate acute cytokine-induced inflammatory responses in mice. J Clin Invest, 120, 2395-405. DE MAN, F. S., HANDOKO, M. L., GROEPENHOFF, H., VAN 'T HUL, A. J., ABBINK, J., KOPPERS, R. J., GROTJOHAN, H. P., TWISK, J. W., BOGAARD, H. J., BOONSTRA, A., POSTMUS, P. E., WESTERHOF, N., VAN DER LAARSE, W. J. & VONK-NOORDEGRAAF, A. 2009. Effects of exercise training in patients with idiopathic pulmonary arterial hypertension. Eur Respir J, 34, 669-75. DECKER, I., GHOSH, S., COMHAIR, S. A., FARHA, S., TANG, W. H., PARK, M., WANG, S., LICHTIN, A. E. & ERZURUM, S. C. 2011. High levels of zinc-protoporphyrin identify iron metabolic abnormalities in pulmonary arterial hypertension. Clin Transl Sci, 4, 253-8. DELCROIX, M., VONK NOORDEGRAAF, A., FADEL, E., LANG, I., SIMONNEAU, G. & NAEIJE, R. 2013. Vascular and right ventricular remodelling in chronic thromboembolic pulmonary hypertension. Eur Respir J, 41, 224-32. DENG, Z., MORSE, J. H., SLAGER, S. L., CUERVO, N., MOORE, K. J., VENETOS, G., KALACHIKOV, S., CAYANIS, E., FISCHER, S. G., BARST, R. J., HODGE, S. E. & KNOWLES, J. A. 2000. Familial primary pulmonary hypertension (gene PPH1) is caused by mutations in the bone morphogenetic protein receptor-II gene. Am J Hum Genet, 67, 737-44. DILLER, G. P., DIMOPOULOS, K., BROBERG, C. S., KAYA, M. G., NAGHOTRA, U. S., UEBING, A., HARRIES, C., GOKTEKIN, O., GIBBS, J. S. & GATZOULIS, M. A. 2006. Presentation, survival prospects, and predictors of death in Eisenmenger syndrome: a combined retrospective and case-control study. Eur Heart J, 27, 1737-42. DORFMÜLLER, P., GÜNTHER, S., GHIGNA, M. R., THOMAS DE MONTPRÉVILLE, V., BOULATE,

200

D., PAUL, J. F., JAÏS, X., DECANTE, B., SIMONNEAU, G., DARTEVELLE, P., HUMBERT, M., FADEL, E. & MERCIER, O. 2014. Microvascular disease in chronic thromboembolic pulmonary hypertension: a role for pulmonary veins and systemic vasculature. Eur Respir J, 44, 1275-88. DORFMÜLLER, P., PERROS, F., BALABANIAN, K. & HUMBERT, M. 2003a. Inflammation in pulmonary arterial hypertension. Eur Respir J, 22, 358-63. DOTY, R. T., PHELPS, S. R., SHADLE, C., SANCHEZ-BONILLA, M., KEEL, S. B. & ABKOWITZ, J. L. 2015. Coordinate expression of heme and globin is essential for effective erythropoiesis. J Clin Invest, 125, 4681-91. DOUGLAS, S. W. & ADAMSON, J. W. 1975. The anemia of chronic disorders: studies of marrow regulation and iron metabolism. Blood, 45, 55-65. DU, X., SHE, E., GELBART, T., TRUKSA, J., LEE, P., XIA, Y., KHOVANANTH, K., MUDD, S., MANN, N., MORESCO, E. M., BEUTLER, E. & BEUTLER, B. 2008a. The serine protease TMPRSS6 is required to sense iron deficiency. Science, 320, 1088-92. DU, X., SHE, E., GELBART, T., TRUKSA, J., LEE, P., XIA, Y., KHOVANANTH, K., MUDD, S., MANN, N., MORESCO, E. M., BEUTLER, E. & BEUTLER, B. 2008b. The serine protease TMPRSS6 is required to sense iron deficiency. Science, 320, 1088-92. DUFFY, S. P., SHING, J., SARAON, P., BERGER, L. C., EIDEN, M. V., WILDE, A. & TAILOR, C. S. 2010. The Fowler syndrome-associated protein FLVCR2 is an importer of heme. Mol Cell Biol, 30, 5318-24. DURIEUX, J., WOLFF, S. & DILLIN, A. 2011. The cell-non-autonomous nature of electron transport chain-mediated longevity. Cell, 144, 79-91. EDDAHIBI, S., GUIGNABERT, C., BARLIER-MUR, A. M., DEWACHTER, L., FADEL, E., DARTEVELLE, P., HUMBERT, M., SIMONNEAU, G., HANOUN, N., SAURINI, F., HAMON, M. & ADNOT, S. 2006a. Cross talk between endothelial and smooth muscle cells in pulmonary hypertension: critical role for serotonin-induced smooth muscle hyperplasia. Circulation, 113, 1857-64. ELLIOTT, C. G., GLISSMEYER, E. W., HAVLENA, G. T., CARLQUIST, J., MCKINNEY, J. T., RICH, S., MCGOON, M. D., SCHOLAND, M. B., KIM, M., JENSEN, R. L., SCHMIDT, J. W. & WARD, K. 2006. Relationship of BMPR2 mutations to vasoreactivity in pulmonary arterial hypertension. Circulation, 113, 2509-15. EMANS, M. E., VAN DER PUTTEN, K., VAN ROOIJEN, K. L., KRAAIJENHAGEN, R. J., SWINKELS, D., VAN SOLINGE, W. W., CRAMER, M. J., DOEVENDANS, P. A., BRAAM, B. & GAILLARD, C. A. 2011. Determinants of red cell distribution width (RDW) in

201

cardiorenal patients: RDW is not related to erythropoietin resistance. J Card Fail, 17, 626-33. ESCRIBANO-SUBIAS, P., BLANCO, I., LÓPEZ-MESEGUER, M., LOPEZ-GUARCH, C. J., ROMAN, A., MORALES, P., CASTILLO-PALMA, M. J., SEGOVIA, J., GÓMEZ-SANCHEZ, M. A., BARBERÀ, J. A. & INVESTIGATORS, R. 2012. Survival in pulmonary hypertension in Spain: insights from the Spanish registry. Eur Respir J, 40, 596-603. EVANS, J. D., GIRERD, B., MONTANI, D., WANG, X. J., GALIÈ, N., AUSTIN, E. D., ELLIOTT, G., ASANO, K., GRÜNIG, E., YAN, Y., JING, Z. C., MANES, A., PALAZZINI, M., WHEELER, L. A., NAKAYAMA, I., SATOH, T., EICHSTAEDT, C., HINDERHOFER, K., WOLF, M., ROSENZWEIG, E. B., CHUNG, W. K., SOUBRIER, F., SIMONNEAU, G., SITBON, O., GRÄF, S., KAPTOGE, S., DI ANGELANTONIO, E., HUMBERT, M. & MORRELL, N. W. 2016. BMPR2 mutations and survival in pulmonary arterial hypertension: an individual participant data meta-analysis. Lancet Respir Med, 4, 129-37. FARBER, H. W., MILLER, D. P., POMS, A. D., BADESCH, D. B., FROST, A. E., MUROS-LE ROUZIC, E., ROMERO, A. J., BENTON, W. W., ELLIOTT, C. G., MCGOON, M. D. & BENZA, R. L. 2015. Five-Year outcomes of patients enrolled in the REVEAL Registry. Chest, 148, 1043-54. FARHA, S., ASOSINGH, K., XU, W., SHARP, J., GEORGE, D., COMHAIR, S., PARK, M., TANG, W. H., LOYD, J. E., THEIL, K., TUBBS, R., HSI, E., LICHTIN, A. & ERZURUM, S. C. 2011. Hypoxia-inducible factors in human pulmonary arterial hypertension: a link to the intrinsic myeloid abnormalities. Blood, 117, 3485-93. FARKAS, N., SZABÓ, A., LÓRÁND, V., SARLÓS, D. P., MINIER, T., PROHÁSZKA, Z., CZIRJÁK, L. & VARJÚ, C. 2014. Clinical usefulness of measuring red blood cell distribution width in patients with systemic sclerosis. Rheumatology (Oxford), 53, 1439-45. FELKER, G. M., ALLEN, L. A., POCOCK, S. J., SHAW, L. K., MCMURRAY, J. J., PFEFFER, M. A., SWEDBERG, K., WANG, D., YUSUF, S., MICHELSON, E. L., GRANGER, C. B. & INVESTIGATORS, C. 2007. Red cell distribution width as a novel prognostic marker in heart failure: data from the CHARM Program and the Duke Databank. J Am Coll Cardiol, 50, 40-7. FERRUCCI, L., SEMBA, R. D., GURALNIK, J. M., ERSHLER, W. B., BANDINELLI, S., PATEL, K. V., SUN, K., WOODMAN, R. C., ANDREWS, N. C., COTTER, R. J., GANZ, T., NEMETH, E. & LONGO, D. L. 2010. Proinflammatory state, hepcidin, and anemia in older persons. Blood, 115, 3810-6. FINBERG, K. E., HEENEY, M. M., CAMPAGNA, D. R., AYDINOK, Y., PEARSON, H. A., HARTMAN,

202

K. R., MAYO, M. M., SAMUEL, S. M., STROUSE, J. J., MARKIANOS, K., ANDREWS, N. C. & FLEMING, M. D. 2008a. Mutations in TMPRSS6 cause iron-refractory iron deficiency anemia (IRIDA). Nat Genet, 40, 569-71. FINBERG, K. E., WHITTLESEY, R. L., FLEMING, M. D. & ANDREWS, N. C. 2010. Down-regulation of Bmp/Smad signaling by Tmprss6 is required for maintenance of systemic iron homeostasis. Blood, 115, 3817-26. FINCH, C. A., MILLER, L. R., INAMDAR, A. R., PERSON, R., SEILER, K. & MACKLER, B. 1976. Iron deficiency in the rat. Physiological and biochemical studies of muscle dysfunction. J Clin Invest, 58, 447-53. FISHER, M. R., MATHAI, S. C., CHAMPION, H. C., GIRGIS, R. E., HOUSTEN-HARRIS, T., HUMMERS, L., KRISHNAN, J. A., WIGLEY, F. & HASSOUN, P. M. 2006. Clinical differences between idiopathic and scleroderma-related pulmonary hypertension. Arthritis Rheum, 54, 3043-50. FÖRHÉCZ, Z., GOMBOS, T., BORGULYA, G., POZSONYI, Z., PROHÁSZKA, Z. & JÁNOSKUTI, L. 2009a. Red cell distribution width in heart failure: prediction of clinical events and relationship with markers of ineffective erythropoiesis, inflammation, renal function, and nutritional state. Am Heart J, 158, 659-66. FRIEDMAN, J. S., LOPEZ, M. F., FLEMING, M. D., RIVERA, A., MARTIN, F. M., WELSH, M. L., BOYD, A., DOCTROW, S. R. & BURAKOFF, S. J. 2004. SOD2-deficiency anemia: protein oxidation and altered protein expression reveal targets of damage, stress response, and antioxidant responsiveness. Blood, 104, 2565-73. FRISE, M. C. & ROBBINS, P. A. 2015a. Iron, oxygen, and the pulmonary circulation. J Appl Physiol (1985), 119, 1421-31. FRISE, M. C., CHENG, H. Y., NICKOL, A. H., CURTIS, M. K., POLLARD, K. A., ROBERTS, D. J., RATCLIFFE, P. J., DORRINGTON, K. L. & ROBBINS, P. A. 2016. Clinical iron deficiency disturbs normal human responses to hypoxia. J Clin Invest, 126, 2139-50. GABLER, N. B., FRENCH, B., STROM, B. L., PALEVSKY, H. I., TAICHMAN, D. B., KAWUT, S. M. & HALPERN, S. D. 2012a. Validation of 6-minute walk distance as a surrogate end point in pulmonary arterial hypertension trials. Circulation, 126, 349-56. GAINE, S. & SIMONNEAU, G. 2013. The need to move from 6-minute walk distance to outcome trials in pulmonary arterial hypertension. Eur Respir Rev, 22, 487-94. GALE, D. P., HARTEN, S. K., REID, C. D., TUDDENHAM, E. G. & MAXWELL, P. H. 2008. Autosomal dominant erythrocytosis and pulmonary arterial hypertension associated with an activating HIF2 alpha mutation. Blood, 112, 919-21.

203

GALIÈ, N., HOEPER, M. M., HUMBERT, M., TORBICKI, A., VACHIERY, J. L., BARBERA, J. A., BEGHETTI, M., CORRIS, P., GAINE, S., GIBBS, J. S., GOMEZ-SANCHEZ, M. A., JONDEAU, G., KLEPETKO, W., OPITZ, C., PEACOCK, A., RUBIN, L., ZELLWEGER, M., SIMONNEAU, G. & (CPG), E. C. F. P. G. 2009. Guidelines for the diagnosis and treatment of pulmonary hypertension: the Task Force for the Diagnosis and Treatment of Pulmonary Hypertension of the European Society of Cardiology (ESC) and the European Respiratory Society (ERS), endorsed by the International Society of Heart and Lung Transplantation (ISHLT). Eur Heart J, 30, 2493-537. GALIÈ, N., HUMBERT, M., VACHIERY, J. L., GIBBS, S., LANG, I., TORBICKI, A., SIMONNEAU, G., PEACOCK, A., VONK NOORDEGRAAF, A., BEGHETTI, M., GHOFRANI, A., GOMEZ SANCHEZ, M. A., HANSMANN, G., KLEPETKO, W., LANCELLOTTI, P., MATUCCI, M., MCDONAGH, T., PIERARD, L. A., TRINDADE, P. T., ZOMPATORI, M., HOEPER, M., ABOYANS, V., VAZ CARNEIRO, A., ACHENBACH, S., AGEWALL, S., ALLANORE, Y., ASTEGGIANO, R., PAOLO BADANO, L., ALBERT BARBERÀ, J., BOUVAIST, H., BUENO, H., BYRNE, R. A., CARERJ, S., CASTRO, G., EROL, Ç., FALK, V., FUNCK-BRENTANO, C., GORENFLO, M., GRANTON, J., IUNG, B., KIELY, D. G., KIRCHHOF, P., KJELLSTROM, B., LANDMESSER, U., LEKAKIS, J., LIONIS, C., LIP, G. Y., ORFANOS, S. E., PARK, M. H., PIEPOLI, M. F., PONIKOWSKI, P., REVEL, M. P., RIGAU, D., ROSENKRANZ, S., VÖLLER, H. & LUIS ZAMORANO, J. 2016. 2015 ESC/ERS Guidelines for the diagnosis and treatment of pulmonary hypertension: The Joint Task Force for the Diagnosis and Treatment of Pulmonary Hypertension of the European Society of Cardiology (ESC) and the European Respiratory Society (ERS): Endorsed by: Association for European Paediatric and Congenital Cardiology (AEPC), International Society for Heart and Lung Transplantation (ISHLT). Eur Heart J, 37, 67-119. GANZ, T. 2011. Hepcidin and iron regulation, 10 years later. Blood, 117, 4425-33. GANZ, T. 2013. Systemic iron homeostasis. Physiol Rev, 93, 1721-41. GEIGER, R., BERGER, R. M., HESS, J., BOGERS, A. J., SHARMA, H. S. & MOOI, W. J. 2000a. Enhanced expression of vascular endothelial growth factor in pulmonary plexogenic arteriopathy due to congenital heart disease. J Pathol, 191, 202-7. GEIGER, R., BERGER, R. M., HESS, J., BOGERS, A. J., SHARMA, H. S. & MOOI, W. J. 2000b. Enhanced expression of vascular endothelial growth factor in pulmonary plexogenic arteriopathy due to congenital heart disease. J Pathol, 191, 202-7. GERMAIN, M., EYRIES, M., MONTANI, D., POIRIER, O., GIRERD, B., DORFMÜLLER, P., COULET, F., NADAUD, S., MAUGENRE, S., GUIGNABERT, C., CARPENTIER, W., VONK-

204

NOORDEGRAAF, A., LÉVY, M., CHAOUAT, A., LAMBERT, J. C., BERTRAND, M., DUPUY, A. M., LETENNEUR, L., LATHROP, M., AMOUYEL, P., DE RAVEL, T. J., DELCROIX, M., AUSTIN, E. D., ROBBINS, I. M., HEMNES, A. R., LOYD, J. E., BERMAN-ROSENZWEIG, E., BARST, R. J., CHUNG, W. K., SIMONNEAU, G., TRÉGOUËT, D. A., HUMBERT, M. & SOUBRIER, F. 2013. Genome-wide association analysis identifies a susceptibility locus for pulmonary arterial hypertension. Nat Genet, 45, 518-21. GHOFRANI, H. A., DISTLER, O., GERHARDT, F., GORENFLO, M., GRÜNIG, E., HAEFELI, W. E., HELD, M., HOEPER, M. M., KÄHLER, C. M., KAEMMERER, H., KLOSE, H., KÖLLNER, V., KOPP, B., MEBUS, S., MEYER, A., MIERA, O., PITTROW, D., RIEMEKASTEN, G., ROSENKRANZ, S., SCHRANZ, D., VOSWINCKEL, R. & OLSCHEWSKI, H. 2011. Treatment of pulmonary arterial hypertension (PAH): updated Recommendations of the Cologne Consensus Conference 2011. Int J Cardiol, 154 Suppl 1, S20-33. GHOFRANI, H. A., OSTERLOH, I. H. & GRIMMINGER, F. 2006. Sildenafil: from angina to erectile dysfunction to pulmonary hypertension and beyond. Nat Rev Drug Discov, 5, 689-702. GHOSH, M. C., ZHANG, D. L., JEONG, S. Y., KOVTUNOVYCH, G., OLLIVIERRE-WILSON, H., NOGUCHI, A., TU, T., SENECAL, T., ROBINSON, G., CROOKS, D. R., TONG, W. H., RAMASWAMY, K., SINGH, A., GRAHAM, B. B., TUDER, R. M., YU, Z. X., ECKHAUS, M., LEE, J., SPRINGER, D. A. & ROUAULT, T. A. 2013. Deletion of iron regulatory protein 1 causes polycythemia and pulmonary hypertension in mice through translational derepression of HIF2α. Cell Metab, 17, 271-81. GIAID, A. & SALEH, D. 1995. Reduced expression of endothelial nitric oxide synthase in the lungs of patients with pulmonary hypertension. N Engl J Med, 333, 214-21. GIAID, A., YANAGISAWA, M., LANGLEBEN, D., MICHEL, R. P., LEVY, R., SHENNIB, H., KIMURA, S., MASAKI, T., DUGUID, W. P. & STEWART, D. J. 1993. Expression of endothelin-1 in the lungs of patients with pulmonary hypertension. N Engl J Med, 328, 1732-9. GIRERD, B., MONTANI, D., COULET, F., SZTRYMF, B., YAICI, A., JAÏS, X., TREGOUET, D., REIS, A., DROUIN-GARRAUD, V., FRAISSE, A., SITBON, O., O'CALLAGHAN, D. S., SIMONNEAU, G., SOUBRIER, F. & HUMBERT, M. 2010. Clinical outcomes of pulmonary arterial hypertension in patients carrying an ACVRL1 (ALK1) mutation. Am J Respir Crit Care Med, 181, 851-61. GIRERD, B., MONTANI, D., JAÏS, X., EYRIES, M., YAICI, A., SZTRYMF, B., SAVALE, L., PARENT, F., COULET, F., GODINAS, L., LAU, E. M., TAMURA, Y., SITBON, O., SOUBRIER, F., SIMONNEAU, G. & HUMBERT, M. 2016. Genetic counselling in a national referral

205

centre for pulmonary hypertension. Eur Respir J, 47, 541-52. GOETZE, O., SCHMITT, J., SPLIETHOFF, K., THEURL, I., WEISS, G., SWINKELS, D. W., TJALSMA, H., MAGGIORINI, M., KRAYENBÜHL, P., RAU, M., FRUEHAUF, H., WOJTAL, K. A., MÜLLHAUPT, B., FRIED, M., GASSMANN, M., LUTZ, T. & GEIER, A. 2013. Adaptation of iron transport and metabolism to acute high-altitude hypoxia in mountaineers. Hepatology, 58, 2153-62. GOODNOUGH, L. T., NEMETH, E. & GANZ, T. 2010. Detection, evaluation, and management of iron-restricted erythropoiesis. Blood, 116, 4754-61. GORE, B., IZIKKI, M., MERCIER, O., DEWACHTER, L., FADEL, E., HUMBERT, M., DARTEVELLE, P., SIMONNEAU, G., NAEIJE, R., LEBRIN, F. & EDDAHIBI, S. 2014. Key role of the endothelial TGF-β/ALK1/endoglin signaling pathway in humans and rodents pulmonary hypertension. PLoS One, 9, e100310. GOSWAMI, T. & ANDREWS, N. C. 2006. Hereditary hemochromatosis protein, HFE, interaction with transferrin receptor 2 suggests a molecular mechanism for mammalian iron sensing. J Biol Chem, 281, 28494-8. GOUMANS, M. J., LEBRIN, F. & VALDIMARSDOTTIR, G. 2003. Controlling the angiogenic switch: a balance between two distinct TGF-b receptor signaling pathways. Trends Cardiovasc Med, 13, 301-7. GRAHAM, B. B., BANDEIRA, A. P., MORRELL, N. W., BUTROUS, G. & TUDER, R. M. 2010. Schistosomiasis-associated pulmonary hypertension: pulmonary vascular disease: the global perspective. Chest, 137, 20S-29S. GRAMMER, T. B., KLEBER, M. E., SILBERNAGEL, G., PILZ, S., SCHARNAGL, H., TOMASCHITZ, A., KÖNIG, W. & MÄRZ, W. 2014. Hemoglobin, iron metabolism and angiographic coronary artery disease (The Ludwigshafen Risk and Cardiovascular Health Study). Atherosclerosis, 236, 292-300. GRUBER, M., HU, C. J., JOHNSON, R. S., BROWN, E. J., KEITH, B. & SIMON, M. C. 2007. Acute postnatal ablation of Hif-2alpha results in anemia. Proc Natl Acad Sci U S A, 104, 2301-6. GRÜNIG, E., LICHTBLAU, M., EHLKEN, N., GHOFRANI, H. A., REICHENBERGER, F., STAEHLER, G., HALANK, M., FISCHER, C., SEYFARTH, H. J., KLOSE, H., MEYER, A., SORICHTER, S., WILKENS, H., ROSENKRANZ, S., OPITZ, C., LEUCHTE, H., KARGER, G., SPEICH, R. & NAGEL, C. 2012. Safety and efficacy of exercise training in various forms of pulmonary hypertension. Eur Respir J, 40, 84-92. GUAZZI, M., ADAMS, V., CONRAADS, V., HALLE, M., MEZZANI, A., VANHEES, L., ARENA, R.,

206

FLETCHER, G. F., FORMAN, D. E., KITZMAN, D. W., LAVIE, C. J., MYERS, J., EACPR & AHA 2012a. EACPR/AHA Joint Scientific Statement. Clinical recommendations for cardiopulmonary exercise testing data assessment in specific patient populations. Eur Heart J, 33, 2917-27. GUAZZI, M., ADAMS, V., CONRAADS, V., HALLE, M., MEZZANI, A., VANHEES, L., ARENA, R., FLETCHER, G. F., FORMAN, D. E., KITZMAN, D. W., LAVIE, C. J., MYERS, J., REHABILITATION, E. A. F. C. P. & ASSOCIATION, A. H. 2012b. EACPR/AHA Scientific Statement. Clinical recommendations for cardiopulmonary exercise testing data assessment in specific patient populations. Circulation, 126, 2261-74. HACHULLA, E., GRESSIN, V., GUILLEVIN, L., CARPENTIER, P., DIOT, E., SIBILIA, J., KAHAN, A., CABANE, J., FRANCÈS, C., LAUNAY, D., MOUTHON, L., ALLANORE, Y., TIEV, K. P., CLERSON, P., DE GROOTE, P. & HUMBERT, M. 2005. Early detection of pulmonary arterial hypertension in systemic sclerosis: a French nationwide prospective multicenter study. Arthritis Rheum, 52, 3792-800. HAMPOLE, C. V., MEHROTRA, A. K., THENAPPAN, T., GOMBERG-MAITLAND, M. & SHAH, S. J. 2009a. Usefulness of red cell distribution width as a prognostic marker in pulmonary hypertension. Am J Cardiol, 104, 868-72. HÄNZE, J., WEISSMANN, N., GRIMMINGER, F., SEEGER, W. & ROSE, F. 2007. Cellular and molecular mechanisms of hypoxia-inducible factor driven vascular remodeling. Thromb Haemost, 97, 774-87. HARDY, S. & VANDEMERGEL, X. 2015a. Intravenous iron administration and hypophosphatemia in clinical practice. Int J Rheumatol, 2015, 468675. HASSAN, S., ANTONELLI, M. & BALLOU, S. 2015. Red cell distribution width: a measure of cardiovascular risk in rheumatoid arthritis patients? Clin Rheumatol, 34, 1053-7. HASSOUN, P. M., MOUTHON, L., BARBERÀ, J. A., EDDAHIBI, S., FLORES, S. C., GRIMMINGER, F., JONES, P. L., MAITLAND, M. L., MICHELAKIS, E. D., MORRELL, N. W., NEWMAN, J. H., RABINOVITCH, M., SCHERMULY, R., STENMARK, K. R., VOELKEL, N. F., YUAN, J. X. & HUMBERT, M. 2009a. Inflammation, growth factors, and pulmonary vascular remodeling. J Am Coll Cardiol, 54, S10-9. HENTZE, M. W., MUCKENTHALER, M. U., GALY, B. & CAMASCHELLA, C. 2010b. Two to tango: regulation of Mammalian iron metabolism. Cell, 142, 24-38. HICKEY, M. M., RICHARDSON, T., WANG, T., MOSQUEIRA, M., ARGUIRI, E., YU, H., YU, Q. C., SOLOMIDES, C. C., MORRISEY, E. E., KHURANA, T. S., CHRISTOFIDOU-SOLOMIDOU, M. & SIMON, M. C. 2010. The von Hippel-Lindau Chuvash mutation promotes

207

pulmonary hypertension and fibrosis in mice. J Clin Invest, 120, 827-39. HINTZE, K. J., KATOH, Y., IGARASHI, K. & THEIL, E. C. 2007. Bach1 repression of ferritin and thioredoxin reductase1 is heme-sensitive in cells and in vitro and coordinates expression with heme oxygenase1, beta-globin, and NADP(H) quinone (oxido) reductase1. J Biol Chem, 282, 34365-71. HOENDERMIS, E. S., LIU, L. C., HUMMEL, Y. M., VAN DER MEER, P., DE BOER, R. A., BERGER, R. M., VAN VELDHUISEN, D. J. & VOORS, A. A. 2015. Effects of sildenafil on invasive haemodynamics and exercise capacity in heart failure patients with preserved ejection fraction and pulmonary hypertension: a randomized controlled trial. Eur Heart J, 36, 2565-73. HOEPER, M. M., BOGAARD, H. J., CONDLIFFE, R., FRANTZ, R., KHANNA, D., KURZYNA, M., LANGLEBEN, D., MANES, A., SATOH, T., TORRES, F., WILKINS, M. R. & BADESCH, D. B. 2013a. Definitions and diagnosis of pulmonary hypertension. J Am Coll Cardiol, 62, D42-50. HOEPER, M. M., HUMBERT, M., SOUZA, R., IDREES, M., KAWUT, S. M., SLIWA-HAHNLE, K., JING, Z. C. & GIBBS, J. S. 2016. A global view of pulmonary hypertension. Lancet Respir Med, 4, 306-22. HOWARD, L. S., WATSON, G. M., WHARTON, J., RHODES, C. J., CHAN, K., KHENGAR, R., ROBBINS, P. A., KIELY, D. G., CONDLIFFE, R., ELLIOTT, C. A., PEPKE-ZABA, J., SHEARES, K., MORRELL, N. W., DAVIES, R., ASHBY, D., GIBBS, J. S. & WILKINS, M. R. 2013. Supplementation of iron in pulmonary hypertension: Rationale and design of a phase II clinical trial in idiopathic pulmonary arterial hypertension. Pulm Circ, 3, 100- 7. HUANG, M. L., LANE, D. J. & RICHARDSON, D. R. 2011. Mitochondrial mayhem: the mitochondrion as a modulator of iron metabolism and its role in disease. Antioxid Redox Signal, 15, 3003-19. HUMBERT, M. 2010. Update on the European respiratory review. Eur Respir J, 36, 993-4. HUMBERT, M., LAU, E. M., MONTANI, D., JAÏS, X., SITBON, O. & SIMONNEAU, G. 2014. Advances in therapeutic interventions for patients with pulmonary arterial hypertension. Circulation, 130, 2189-208. HUMBERT, M., MONTI, G., BRENOT, F., SITBON, O., PORTIER, A., GRANGEOT-KEROS, L., DUROUX, P., GALANAUD, P., SIMONNEAU, G. & EMILIE, D. 1995. Increased interleukin-1 and interleukin-6 serum concentrations in severe primary pulmonary hypertension. Am J Respir Crit Care Med, 151, 1628-31.

208

HUMBERT, M., SITBON, O., CHAOUAT, A., BERTOCCHI, M., HABIB, G., GRESSIN, V., YAICI, A., WEITZENBLUM, E., CORDIER, J. F., CHABOT, F., DROMER, C., PISON, C., REYNAUD- GAUBERT, M., HALOUN, A., LAURENT, M., HACHULLA, E. & SIMONNEAU, G. 2006. Pulmonary arterial hypertension in France: results from a national registry. Am J Respir Crit Care Med, 173, 1023-30. HUMBERT, M., SITBON, O., CHAOUAT, A., BERTOCCHI, M., HABIB, G., GRESSIN, V., YAÏCI, A., WEITZENBLUM, E., CORDIER, J. F., CHABOT, F., DROMER, C., PISON, C., REYNAUD- GAUBERT, M., HALOUN, A., LAURENT, M., HACHULLA, E., COTTIN, V., DEGANO, B., JAÏS, X., MONTANI, D., SOUZA, R. & SIMONNEAU, G. 2010. Survival in patients with idiopathic, familial, and anorexigen-associated pulmonary arterial hypertension in the modern management era. Circulation, 122, 156-63. HVIDBERG, V., MANIECKI, M. B., JACOBSEN, C., HØJRUP, P., MØLLER, H. J. & MOESTRUP, S. K. 2005. Identification of the receptor scavenging hemopexin-heme complexes. Blood, 106, 2572-9. IKEDA, Y., TAJIMA, S., YOSHIDA, S., YAMANO, N., KIHIRA, Y., ISHIZAWA, K., AIHARA, K., TOMITA, S., TSUCHIYA, K. & TAMAKI, T. 2011a. Deferoxamine promotes angiogenesis via the activation of vascular endothelial cell function. Atherosclerosis, 215, 339-47. ISERN, R. A., YANEVA, M., WEINER, E., PARKE, A., ROTHFIELD, N., DANTZKER, D., RICH, S. & ARNETT, F. C. 1992. Autoantibodies in patients with primary pulmonary hypertension: association with anti-Ku. Am J Med, 93, 307-12. ISODA, M., HANAWA, H., WATANABE, R., YOSHIDA, T., TOBA, K., YOSHIDA, K., KOJIMA, M., OTAKI, K., HAO, K., DING, L., TANAKA, K., TAKAYAMA, T., KATO, K., OKURA, Y., KODAMA, M., OTA, Y., HAYASHI, J. & AIZAWA, Y. 2010. Expression of the peptide hormone hepcidin increases in cardiomyocytes under myocarditis and myocardial infarction. J Nutr Biochem, 21, 749-56. ITOH, T., NAGAYA, N., ISHIBASHI-UEDA, H., KYOTANI, S., OYA, H., SAKAMAKI, F., KIMURA, H. & NAKANISHI, N. 2006. Increased plasma monocyte chemoattractant protein-1 level in idiopathic pulmonary arterial hypertension. Respirology, 11, 158-63. JABARA, H. H., BOYDEN, S. E., CHOU, J., RAMESH, N., MASSAAD, M. J., BENSON, H., BAINTER, W., FRAULINO, D., RAHIMOV, F., SIEFF, C., LIU, Z. J., ALSHEMMARI, S. H., AL-RAMADI, B. K., AL-DHEKRI, H., ARNAOUT, R., ABU-SHUKAIR, M., VATSAYAN, A., SILVER, E., AHUJA, S., DAVIES, E. G., SOLA-VISNER, M., OHSUMI, T. K., ANDREWS, N. C., NOTARANGELO, L. D., FLEMING, M. D., AL-HERZ, W., KUNKEL, L. M. & GEHA, R. S.

209

2016. A missense mutation in TFRC, encoding transferrin receptor 1, causes combined immunodeficiency. Nat Genet, 48, 74-8. JANKOWSKA, E. A., KASZTURA, M., SOKOLSKI, M., BRONISZ, M., NAWROCKA, S., OLEŚKOWSKA-FLOREK, W., ZYMLIŃSKI, R., BIEGUS, J., SIWOŁOWSKI, P., BANASIAK, W., ANKER, S. D., FILIPPATOS, G., CLELAND, J. G. & PONIKOWSKI, P. 2014. Iron deficiency defined as depleted iron stores accompanied by unmet cellular iron requirements identifies patients at the highest risk of death after an episode of acute heart failure. Eur Heart J, 35, 2468-76. JANKOWSKA, E. A., ROZENTRYT, P., WITKOWSKA, A., NOWAK, J., HARTMANN, O., PONIKOWSKA, B., BORODULIN-NADZIEJA, L., BANASIAK, W., POLONSKI, L., FILIPPATOS, G., MCMURRAY, J. J., ANKER, S. D. & PONIKOWSKI, P. 2010. Iron deficiency: an ominous sign in patients with systolic chronic heart failure. Eur Heart J, 31, 1872-80. JAWAD, A., TONELLI, A. R., CHATBURN, R. L., WANG, X. & HATIPOĞLU, U. 2017. Impact of Intrathoracic Pressure in the Assessment of Pulmonary Hypertension in Overweight Patients. Ann Am Thorac Soc, 14, 1861-1863. JOHNSON, S. R., GRANTON, J. T., TOMLINSON, G. A., GROSBEIN, H. A., HAWKER, G. A. & FELDMAN, B. M. 2011. Effect of warfarin on survival in scleroderma-associated pulmonary arterial hypertension (SSc-PAH) and idiopathic PAH. Belief elicitation for Bayesian priors. J Rheumatol, 38, 462-9. JOMOVA, K. & VALKO, M. 2011. Advances in metal-induced oxidative stress and human disease. Toxicology, 283, 65-87. JOYNER, M. J. & JOHNSON, B. D. 2008. Iron lung? New ideas about hypoxic pulmonary vasoconstriction. J Physiol, 586, 5837-8. KABITZ, H. J., BREMER, H. C., SCHWOERER, A., SONNTAG, F., WALTERSPACHER, S., WALKER, D. J., EHLKEN, N., STAEHLER, G., WINDISCH, W. & GRÜNIG, E. 2014. The combination of exercise and respiratory training improves respiratory muscle function in pulmonary hypertension. Lung, 192, 321-8. KAELIN, W. G. & RATCLIFFE, P. J. 2008. Oxygen sensing by metazoans: the central role of the HIF hydroxylase pathway. Mol Cell, 30, 393-402. KAEMMERER, H., FRATZ, S., BRAUN, S. L., KOELLING, K., EICKEN, A., BRODHERR-HEBERLEIN, S., PIETRZIK, K. & HESS, J. 2004. Erythrocyte indexes, iron metabolism, and hyperhomocysteinemia in adults with cyanotic congenital cardiac disease. Am J Cardiol, 94, 825-8.

210

KALRA, P. A. & BHANDARI, S. 2016. Efficacy and safety of iron isomaltoside (Monofer(®)) in the management of patients with iron deficiency anemia. Int J Nephrol Renovasc Dis, 9, 53-64. KATO, J., FUJIKAWA, K., KANDA, M., FUKUDA, N., SASAKI, K., TAKAYAMA, T., KOBUNE, M., TAKADA, K., TAKIMOTO, R., HAMADA, H., IKEDA, T. & NIITSU, Y. 2001. A mutation, in the iron-responsive element of H ferritin mRNA, causing autosomal dominant iron overload. Am J Hum Genet, 69, 191-7. KAUTZ, L., JUNG, G., VALORE, E. V., RIVELLA, S., NEMETH, E. & GANZ, T. 2014. Identification of erythroferrone as an erythroid regulator of iron metabolism. Nat Genet, 46, 678- 84. KAUTZ, L., MEYNARD, D., MONNIER, A., DARNAUD, V., BOUVET, R., WANG, R. H., DENG, C., VAULONT, S., MOSSER, J., COPPIN, H. & ROTH, M. P. 2008. Iron regulates phosphorylation of Smad1/5/8 and gene expression of Bmp6, Smad7, Id1, and Atoh8 in the mouse liver. Blood, 112, 1503-9. KAWUT, S. M., TAICHMAN, D. B., ARCHER-CHICKO, C. L., PALEVSKY, H. I. & KIMMEL, S. E. 2003. Hemodynamics and survival in patients with pulmonary arterial hypertension related to systemic sclerosis. Chest, 123, 344-50. KEEL, S. B., DOTY, R. T., YANG, Z., QUIGLEY, J. G., CHEN, J., KNOBLAUGH, S., KINGSLEY, P. D., DE DOMENICO, I., VAUGHN, M. B., KAPLAN, J., PALIS, J. & ABKOWITZ, J. L. 2008. A heme export protein is required for red blood cell differentiation and iron homeostasis. Science, 319, 825-8. KIEFER, C. R. & SNYDER, L. M. 2000. Oxidation and erythrocyte senescence. Curr Opin Hematol, 7, 113-6. KIM, J. D., LEE, A., CHOI, J., PARK, Y., KANG, H., CHANG, W., LEE, M. S. & KIM, J. 2015. Epigenetic modulation as a therapeutic approach for pulmonary arterial hypertension. Exp Mol Med, 47, e175. KLEIN-WEIGEL, P., OPITZ, C. & RIEMEKASTEN, G. 2011. Systemic sclerosis - a systematic overview: part 1 - disease characteristics and classification, pathophysiologic concepts, and recommendations for diagnosis and surveillance. Vasa, 40, 6-19. KLIP, I. T., COMIN-COLET, J., VOORS, A. A., PONIKOWSKI, P., ENJUANES, C., BANASIAK, W., LOK, D. J., ROSENTRYT, P., TORRENS, A., POLONSKI, L., VAN VELDHUISEN, D. J., VAN DER MEER, P. & JANKOWSKA, E. A. 2013. Iron deficiency in chronic heart failure: an international pooled analysis. Am Heart J, 165, 575-582.e3. KNITTEL, T., FELLMER, P., MÜLLER, L. & RAMADORI, G. 1997. Bone morphogenetic protein-6

211

is expressed in nonparenchymal liver cells and upregulated by transforming growth factor-beta 1. Exp Cell Res, 232, 263-9. KOEHLER, R., GRÜNIG, E., PAUCIULO, M. W., HOEPER, M. M., OLSCHEWSKI, H., WILKENS, H., HALANK, M., WINKLER, J., EWERT, R., BREMER, H., KREUSCHER, S., JANSSEN, B. & NICHOLS, W. C. 2004. Low frequency of BMPR2 mutations in a German cohort of patients with sporadic idiopathic pulmonary arterial hypertension. J Med Genet, 41, e127. KOSKENKORVA-FRANK, T. S., WEISS, G., KOPPENOL, W. H. & BURCKHARDT, S. 2013. The complex interplay of iron metabolism, reactive oxygen species, and reactive nitrogen species: insights into the potential of various iron therapies to induce oxidative and nitrosative stress. Free Radic Biol Med, 65, 1174-94. KOVACS, G., BERGHOLD, A., SCHEIDL, S. & OLSCHEWSKI, H. 2009. Pulmonary arterial pressure during rest and exercise in healthy subjects: a systematic review. Eur Respir J, 34, 888-94. KRASUSKI, R. A., HART, S. A., SMITH, B., WANG, A., HARRISON, J. K. & BASHORE, T. M. 2011. Association of anemia and long-term survival in patients with pulmonary hypertension. Int J Cardiol, 150, 291-5. KREYMBORG, K., UCHIDA, S., GELLERT, P., SCHNEIDER, A., BOETTGER, T., VOSWINCKEL, R., WIETELMANN, A., SZIBOR, M., WEISSMANN, N., GHOFRANI, A. H., SCHERMULY, R., SCHRANZ, D., SEEGER, W. & BRAUN, T. 2010a. Identification of right heart-enriched genes in a murine model of chronic outflow tract obstruction. J Mol Cell Cardiol, 49, 598-605. KRISTIANSEN, M., GRAVERSEN, J. H., JACOBSEN, C., SONNE, O., HOFFMAN, H. J., LAW, S. K. & MOESTRUP, S. K. 2001. Identification of the haemoglobin scavenger receptor. Nature, 409, 198-201. KUNS-HASHIMOTO, R., KUNINGER, D., NILI, M. & ROTWEIN, P. 2008. Selective binding of RGMc/hemojuvelin, a key protein in systemic iron metabolism, to BMP-2 and neogenin. Am J Physiol Cell Physiol, 294, C994-C1003. LAJOIE, A. C., LAUZIÈRE, G., LEGA, J. C., LACASSE, Y., MARTIN, S., SIMARD, S., BONNET, S. & PROVENCHER, S. 2016. Combination therapy versus monotherapy for pulmonary arterial hypertension: a meta-analysis. Lancet Respir Med, 4, 291-305. LAKHAL, S., TALBOT, N. P., CROSBY, A., STOEPKER, C., TOWNSEND, A. R., ROBBINS, P. A., PUGH, C. W., RATCLIFFE, P. J. & MOLE, D. R. 2009. Regulation of growth differentiation factor 15 expression by intracellular iron. Blood, 113, 1555-63.

212

LANE, D. J., BAE, D. H., MERLOT, A. M., SAHNI, S. & RICHARDSON, D. R. 2015a. Duodenal cytochrome b (DCYTB) in iron metabolism: an update on function and regulation. Nutrients, 7, 2274-96. LANE, D. J., MERLOT, A. M., HUANG, M. L., BAE, D. H., JANSSON, P. J., SAHNI, S., KALINOWSKI, D. S. & RICHARDSON, D. R. 2015b. Cellular iron uptake, trafficking and metabolism: Key molecules and mechanisms and their roles in disease. Biochim Biophys Acta, 1853, 1130-44. LANE, K. B., MACHADO, R. D., PAUCIULO, M. W., THOMSON, J. R., PHILLIPS, J. A., LOYD, J. E., NICHOLS, W. C., TREMBATH, R. C. & CONSORTIUM, I. P. 2000c. Heterozygous germline mutations in BMPR2, encoding a TGF-beta receptor, cause familial primary pulmonary hypertension. Nat Genet, 26, 81-4. LAPPEGÅRD, J., ELLINGSEN, T. S., SKJELBAKKEN, T., MATHIESEN, E. B., NJØLSTAD, I., WILSGAARD, T., BROX, J., BRÆKKAN, S. K. & HANSEN, J. B. 2016. Red cell distribution width is associated with future risk of incident stroke. The Tromsø Study. Thromb Haemost, 115, 126-34. LAU, E. M., GODINAS, L., SITBON, O., MONTANI, D., SAVALE, L., JAÏS, X., LADOR, F., GUNTHER, S., CELERMAJER, D. S., SIMONNEAU, G., HUMBERT, M., CHEMLA, D. & HERVE, P. 2016. Resting pulmonary artery pressure of 21-24 mmHg predicts abnormal exercise haemodynamics. Eur Respir J, 47, 1436-44. LAUNAY, D., SITBON, O., HACHULLA, E., MOUTHON, L., GRESSIN, V., ROTTAT, L., CLERSON, P., CORDIER, J. F., SIMONNEAU, G. & HUMBERT, M. 2013. Survival in systemic sclerosis-associated pulmonary arterial hypertension in the modern management era. Ann Rheum Dis, 72, 1940-6. LAUNAY, J. M., HERVÉ, P., PEOC'H, K., TOURNOIS, C., CALLEBERT, J., NEBIGIL, C. G., ETIENNE, N., DROUET, L., HUMBERT, M., SIMONNEAU, G. & MAROTEAUX, L. 2002a. Function of the serotonin 5-hydroxytryptamine 2B receptor in pulmonary hypertension. Nat Med, 8, 1129-35. LAWEN, A. & LANE, D. J. 2013. Mammalian iron homeostasis in health and disease: uptake, storage, transport, and molecular mechanisms of action. Antioxid Redox Signal, 18, 2473-507. LE PAVEC, J., HUMBERT, M., MOUTHON, L. & HASSOUN, P. M. 2010. Systemic sclerosis- associated pulmonary arterial hypertension. Am J Respir Crit Care Med, 181, 1285- 93. LEE, D. H., ZHOU, L. J., ZHOU, Z., XIE, J. X., JUNG, J. U., LIU, Y., XI, C. X., MEI, L. & XIONG, W. C.

213

2010. Neogenin inhibits HJV secretion and regulates BMP-induced hepcidin expression and iron homeostasis. Blood, 115, 3136-45. LEE, S. D., SHROYER, K. R., MARKHAM, N. E., COOL, C. D., VOELKEL, N. F. & TUDER, R. M. 1998. Monoclonal endothelial cell proliferation is present in primary but not secondary pulmonary hypertension. J Clin Invest, 101, 927-34. LEVARGE, B. L., POMERANTSEV, E. & CHANNICK, R. N. 2014. Reliance on end-expiratory wedge pressure leads to misclassification of pulmonary hypertension. Eur Respir J, 44, 425-34. LEVI, S. & ROVIDA, E. 2009. The role of iron in mitochondrial function. Biochim Biophys Acta, 1790, 629-36. LING, Y., JOHNSON, M. K., KIELY, D. G., CONDLIFFE, R., ELLIOT, C. A., GIBBS, J. S., HOWARD, L. S., PEPKE-ZABA, J., SHEARES, K. K., CORRIS, P. A., FISHER, A. J., LORDAN, J. L., GAINE, S., COGHLAN, J. G., WORT, S. J., GATZOULIS, M. A. & PEACOCK, A. J. 2012. Changing demographics, epidemiology, and survival of incident pulmonary arterial hypertension: results from the pulmonary hypertension registry of the United Kingdom and Ireland. Am J Respir Crit Care Med, 186, 790-6. LIPINSKI, B. & PRETORIUS, E. 2013. Iron-induced fibrin in cardiovascular disease. Curr Neurovasc Res, 10, 269-74. LIPPI, G., TARGHER, G., MONTAGNANA, M., SALVAGNO, G. L., ZOPPINI, G. & GUIDI, G. C. 2009. Relation between red blood cell distribution width and inflammatory biomarkers in a large cohort of unselected outpatients. Arch Pathol Lab Med, 133, 628-32. LIU, Q., DAVIDOFF, O., NISS, K. & HAASE, V. H. 2012. Hypoxia-inducible factor regulates hepcidin via erythropoietin-induced erythropoiesis. J Clin Invest, 122, 4635-44. LIUZZI, J. P., AYDEMIR, F., NAM, H., KNUTSON, M. D. & COUSINS, R. J. 2006. Zip14 (Slc39a14) mediates non-transferrin-bound iron uptake into cells. Proc Natl Acad Sci U S A, 103, 13612-7. LONG, L., ORMISTON, M. L., YANG, X., SOUTHWOOD, M., GRÄF, S., MACHADO, R. D., MUELLER, M., KINZEL, B., YUNG, L. M., WILKINSON, J. M., MOORE, S. D., DRAKE, K. M., ALDRED, M. A., YU, P. B., UPTON, P. D. & MORRELL, N. W. 2015. Selective enhancement of endothelial BMPR-II with BMP9 reverses pulmonary arterial hypertension. Nat Med, 21, 777-85. LORÉAL, O., CAVEY, T., BARDOU-JACQUET, E., GUGGENBUHL, P., ROPERT, M. & BRISSOT, P. 2014. Iron, hepcidin, and the metal connection. Front Pharmacol, 5, 128.

214

MA, L., ROMAN-CAMPOS, D., AUSTIN, E. D., EYRIES, M., SAMPSON, K. S., SOUBRIER, F., GERMAIN, M., TRÉGOUËT, D. A., BORCZUK, A., ROSENZWEIG, E. B., GIRERD, B., MONTANI, D., HUMBERT, M., LOYD, J. E., KASS, R. S. & CHUNG, W. K. 2013. A novel channelopathy in pulmonary arterial hypertension. N Engl J Med, 369, 351-61. MACCHIA, A., MARCHIOLI, R., MARFISI, R., SCARANO, M., LEVANTESI, G., TAVAZZI, L. & TOGNONI, G. 2007. A meta-analysis of trials of pulmonary hypertension: a clinical condition looking for drugs and research methodology. Am Heart J, 153, 1037-47. MACCHIA, A., MARCHIOLI, R., TOGNONI, G., SCARANO, M., MARFISI, R., TAVAZZI, L. & RICH, S. 2010. Systematic review of trials using vasodilators in pulmonary arterial hypertension: why a new approach is needed. Am Heart J, 159, 245-57. MACDOUGALL, I. C., BOCK, A. H., CARRERA, F., ECKARDT, K. U., GAILLARD, C., VAN WYCK, D., ROUBERT, B., NOLEN, J. G., ROGER, S. D. & INVESTIGATORS, F.-C. S. 2014. FIND-CKD: a randomized trial of intravenous ferric carboxymaltose versus oral iron in patients with chronic kidney disease and iron deficiency anaemia. Nephrol Dial Transplant, 29, 2075-84. MACHADO, R. D., ALDRED, M. A., JAMES, V., HARRISON, R. E., PATEL, B., SCHWALBE, E. C., GRUENIG, E., JANSSEN, B., KOEHLER, R., SEEGER, W., EICKELBERG, O., OLSCHEWSKI, H., ELLIOTT, C. G., GLISSMEYER, E., CARLQUIST, J., KIM, M., TORBICKI, A., FIJALKOWSKA, A., SZEWCZYK, G., PARMA, J., ABRAMOWICZ, M. J., GALIE, N., MORISAKI, H., KYOTANI, S., NAKANISHI, N., MORISAKI, T., HUMBERT, M., SIMONNEAU, G., SITBON, O., SOUBRIER, F., COULET, F., MORRELL, N. W. & TREMBATH, R. C. 2006a. Mutations of the TGF-beta type II receptor BMPR2 in pulmonary arterial hypertension. Hum Mutat, 27, 121-32. MACHADO, R. D., ALDRED, M. A., JAMES, V., HARRISON, R. E., PATEL, B., SCHWALBE, E. C., GRUENIG, E., JANSSEN, B., KOEHLER, R., SEEGER, W., EICKELBERG, O., OLSCHEWSKI, H., ELLIOTT, C. G., GLISSMEYER, E., CARLQUIST, J., KIM, M., TORBICKI, A., FIJALKOWSKA, A., SZEWCZYK, G., PARMA, J., ABRAMOWICZ, M. J., GALIE, N., MORISAKI, H., KYOTANI, S., NAKANISHI, N., MORISAKI, T., HUMBERT, M., SIMONNEAU, G., SITBON, O., SOUBRIER, F., COULET, F., MORRELL, N. W. & TREMBATH, R. C. 2006c. Mutations of the TGF-beta type II receptor BMPR2 in pulmonary arterial hypertension. Hum Mutat, 27, 121-32. MACHADO, R. D., EICKELBERG, O., ELLIOTT, C. G., GERACI, M. W., HANAOKA, M., LOYD, J. E., NEWMAN, J. H., PHILLIPS, J. A., SOUBRIER, F., TREMBATH, R. C. & CHUNG, W. K. 2009. Genetics and genomics of pulmonary arterial hypertension. J Am Coll Cardiol,

215

54, S32-42. MACHADO, R. D., SOUTHGATE, L., EICHSTAEDT, C. A., ALDRED, M. A., AUSTIN, E. D., BEST, D. H., CHUNG, W. K., BENJAMIN, N., ELLIOTT, C. G., EYRIES, M., FISCHER, C., GRÄF, S., HINDERHOFER, K., HUMBERT, M., KEILES, S. B., LOYD, J. E., MORRELL, N. W., NEWMAN, J. H., SOUBRIER, F., TREMBATH, R. C., VIALES, R. R. & GRÜNIG, E. 2015. Pulmonary Arterial Hypertension: A Current Perspective on Established and Emerging Molecular Genetic Defects. Hum Mutat, 36, 1113-27. MAINGUY, V., MALTAIS, F., SAEY, D., GAGNON, P., MARTEL, S., SIMON, M. & PROVENCHER, S. 2010a. Effects of a rehabilitation program on skeletal muscle function in idiopathic pulmonary arterial hypertension. J Cardiopulm Rehabil Prev, 30, 319-23. MAINGUY, V., MALTAIS, F., SAEY, D., GAGNON, P., MARTEL, S., SIMON, M. & PROVENCHER, S. 2010b. Peripheral muscle dysfunction in idiopathic pulmonary arterial hypertension. Thorax, 65, 113-7. MAINGUY, V., MALTAIS, F., SAEY, D., GAGNON, P., MARTEL, S., SIMON, M. & PROVENCHER, S. 2010. Peripheral muscle dysfunction in idiopathic pulmonary arterial hypertension. Thorax, 65, 113-7. MALENFANT, S., POTUS, F., FOURNIER, F., BREUILS-BONNET, S., PFLIEGER, A., BOURASSA, S., TREMBLAY, È., NEHMÉ, B., DROIT, A., BONNET, S. & PROVENCHER, S. 2015. Skeletal muscle proteomic signature and metabolic impairment in pulmonary hypertension. J Mol Med (Berl), 93, 573-84. MARON, B. A. & RYAN, J. J. 2016. Treatment differences in pulmonary arterial hypertension management. Pulm Circ, 6, 619. MARON, B. A., HESS, E., MADDOX, T. M., OPOTOWSKY, A. R., TEDFORD, R. J., LAHM, T., JOYNT, K. E., KASS, D. J., STEPHENS, T., STANISLAWSKI, M. A., SWENSON, E. R., GOLDSTEIN, R. H., LEOPOLD, J. A., ZAMANIAN, R. T., ELWING, J. M., PLOMONDON, M. E., GRUNWALD, G. K., BARÓN, A. E., RUMSFELD, J. S. & CHOUDHARY, G. 2016. Association of Borderline Pulmonary Hypertension With Mortality and Hospitalization in a Large Patient Cohort: Insights From the Veterans Affairs Clinical Assessment, Reporting, and Tracking Program. Circulation, 133, 1240-8. MARTINELLI, N., TRAGLIA, M., CAMPOSTRINI, N., BIINO, G., CORBELLA, M., SALA, C., BUSTI, F., MASCIULLO, C., MANNA, D., PREVITALI, S., CASTAGNA, A., PISTIS, G., OLIVIERI, O., TONIOLO, D., CAMASCHELLA, C. & GIRELLI, D. 2012b. Increased serum hepcidin levels in subjects with the metabolic syndrome: a population study. PLoS One, 7, e48250.

216

MARTÍNEZ-QUINTANA, E. & RODRÍGUEZ-GONZÁLEZ, F. 2013. Iron deficiency anemia detection from hematology parameters in adult congenital heart disease patients. Congenit Heart Dis, 8, 117-23. MASSIN, M. M. 2012. Relation between red cell distribution width and clinical outcome after surgery for congenital heart disease in children. Pediatr Cardiol, 33, 1021-5. MASTROGIANNAKI, M., MATAK, P., KEITH, B., SIMON, M. C., VAULONT, S. & PEYSSONNAUX, C. 2009. HIF-2alpha, but not HIF-1alpha, promotes iron absorption in mice. J Clin Invest, 119, 1159-66. MCCLURE, S., CUSTER, E. & BESSMAN, J. D. 1985. Improved detection of early iron deficiency in nonanemic subjects. JAMA, 253, 1021-3. MCGOON, M. D., BENZA, R. L., ESCRIBANO-SUBIAS, P., JIANG, X., MILLER, D. P., PEACOCK, A. J., PEPKE-ZABA, J., PULIDO, T., RICH, S., ROSENKRANZ, S., SUISSA, S. & HUMBERT, M. 2013. Pulmonary arterial hypertension: epidemiology and registries. J Am Coll Cardiol, 62, D51-9. MCLANE, J. A., FELL, R. D., MCKAY, R. H., WINDER, W. W., BROWN, E. B. & HOLLOSZY, J. O. 1981a. Physiological and biochemical effects of iron deficiency on rat skeletal muscle. Am J Physiol, 241, C47-54. MCLAUGHLIN, V. V., SITBON, O., BADESCH, D. B., BARST, R. J., BLACK, C., GALIÈ, N., RAINISIO, M., SIMONNEAU, G. & RUBIN, L. J. 2005. Survival with first-line bosentan in patients with primary pulmonary hypertension. Eur Respir J, 25, 244-9. MENEGHINI, R. 1997. Iron homeostasis, oxidative stress, and DNA damage. Free Radic Biol Med, 23, 783-92. MERLE, U., FEIN, E., GEHRKE, S. G., STREMMEL, W. & KULAKSIZ, H. 2007. The iron regulatory peptide hepcidin is expressed in the heart and regulated by hypoxia and inflammation. Endocrinology, 148, 2663-8. MEYNARD, D., KAUTZ, L., DARNAUD, V., CANONNE-HERGAUX, F., COPPIN, H. & ROTH, M. P. 2009. Lack of the bone morphogenetic protein BMP6 induces massive iron overload. Nat Genet, 41, 478-81. MICHELAKIS, E. D. 2006. Spatio-temporal diversity of apoptosis within the vascular wall in pulmonary arterial hypertension: heterogeneous BMP signaling may have therapeutic implications. Circ Res, 98, 172-5. MIMS, M. P. & PRCHAL, J. T. 2005. Divalent metal transporter 1. Hematology, 10, 339-45. MINETTI, M., AGATI, L. & MALORNI, W. 2007. The microenvironment can shift erythrocytes from a friendly to a harmful behavior: pathogenetic implications for vascular

217

diseases. Cardiovasc Res, 75, 21-8. MIYAMOTO, K., INAI, K., TAKEUCHI, D., SHINOHARA, T. & NAKANISHI, T. 2015. Relationships among red cell distribution width, anemia, and interleukin-6 in adult congenital heart disease. Circ J, 79, 1100-6. MIYAMOTO, S., NAGAYA, N., SATOH, T., KYOTANI, S., SAKAMAKI, F., FUJITA, M., NAKANISHI, N. & MIYATAKE, K. 2000. Clinical correlates and prognostic significance of six-minute walk test in patients with primary pulmonary hypertension. Comparison with cardiopulmonary exercise testing. Am J Respir Crit Care Med, 161, 487-92. MOLE, D. R. 2010. Iron homeostasis and its interaction with prolyl hydroxylases. Antioxid Redox Signal, 12, 445-58. MOLLET, I. G., PATEL, D., GOVANI, F. S., GIESS, A., PASCHALAKI, K., PERIYASAMY, M., LIDINGTON, E. C., MASON, J. C., JONES, M. D., GAME, L., ALI, S. & SHOVLIN, C. L. 2016a. Low Dose Iron Treatments Induce a DNA Damage Response in Human Endothelial Cells within Minutes. PLoS One, 11, e0147990. MOORE, J., HAWKINS, S. A., AUSTIN, A. R., KONOLD, T., GREEN, R. B., BLAMIRE, I. W., DEXTER, I., STACK, M. J., CHAPLIN, M. J., LANGEVELD, J. P., SIMMONS, M. M., SPENCER, Y. I., WEBB, P. R., DAWSON, M. & WELLS, G. A. 2011. Studies of the transmissibility of the agent of bovine spongiform encephalopathy to the domestic chicken. BMC Res Notes, 4, 501. MORRELL, N. W. 2006. Pulmonary hypertension due to BMPR2 mutation: a new paradigm for tissue remodeling? Proc Am Thorac Soc, 3, 680-6. MORRIS, T. A., MARSH, J. J., CHILES, P. G., AUGER, W. R., FEDULLO, P. F. & WOODS, V. L. 2006. Fibrin derived from patients with chronic thromboembolic pulmonary hypertension is resistant to lysis. Am J Respir Crit Care Med, 173, 1270-5. MORRIS, T. A., MARSH, J. J., CHILES, P. G., MAGAÑA, M. M., LIANG, N. C., SOLER, X., DESANTIS, D. J., NGO, D. & WOODS, V. L. 2009. High prevalence of dysfibrinogenemia among patients with chronic thromboembolic pulmonary hypertension. Blood, 114, 1929-36. MUKERJEE, D., ST GEORGE, D., COLEIRO, B., KNIGHT, C., DENTON, C. P., DAVAR, J., BLACK, C. M. & COGHLAN, J. G. 2003. Prevalence and outcome in systemic sclerosis associated pulmonary arterial hypertension: application of a registry approach. Ann Rheum Dis, 62, 1088-93. MURRAY, F., MACLEAN, M. R. & PYNE, N. J. 2002. Increased expression of the cGMP- inhibited cAMP-specific (PDE3) and cGMP binding cGMP-specific (PDE5)

218

phosphodiesterases in models of pulmonary hypertension. Br J Pharmacol, 137, 1187-94. NAGENDRAN, J., ARCHER, S. L., SOLIMAN, D., GURTU, V., MOUDGIL, R., HAROMY, A., ST AUBIN, C., WEBSTER, L., REBEYKA, I. M., ROSS, D. B., LIGHT, P. E., DYCK, J. R. & MICHELAKIS, E. D. 2007. Phosphodiesterase type 5 is highly expressed in the hypertrophied human right ventricle, and acute inhibition of phosphodiesterase type 5 improves contractility. Circulation, 116, 238-48. NAITO, Y., HOSOKAWA, M., HAO, H., SAWADA, H., HIROTANI, S., IWASAKU, T., OKUHARA, Y., EGUCHI, A., HIROTA, S., OHYANAGI, M., TSUJINO, T. & MASUYAMA, T. 2013a. Impact of dietary iron restriction on the development of monocrotaline-induced pulmonary vascular remodeling and right ventricular failure in rats. Biochem Biophys Res Commun, 436, 145-51. NAITO, Y., TSUJINO, T., MATSUMOTO, M., SAKODA, T., OHYANAGI, M. & MASUYAMA, T. 2009. Adaptive response of the heart to long-term anemia induced by iron deficiency. Am J Physiol Heart Circ Physiol, 296, H585-93. NANAMI, M., OOKAWARA, T., OTAKI, Y., ITO, K., MORIGUCHI, R., MIYAGAWA, K., HASUIKE, Y., IZUMI, M., EGUCHI, H., SUZUKI, K. & NAKANISHI, T. 2005. Tumor necrosis factor- alpha-induced iron sequestration and oxidative stress in human endothelial cells. Arterioscler Thromb Vasc Biol, 25, 2495-501. NAVARRO-SASTRE, A., TORT, F., STEHLING, O., UZARSKA, M. A., ARRANZ, J. A., DEL TORO, M., LABAYRU, M. T., LANDA, J., FONT, A., GARCIA-VILLORIA, J., MERINERO, B., UGARTE, M., GUTIERREZ-SOLANA, L. G., CAMPISTOL, J., GARCIA-CAZORLA, A., VAQUERIZO, J., RIUDOR, E., BRIONES, P., ELPELEG, O., RIBES, A. & LILL, R. 2011a. A fatal mitochondrial disease is associated with defective NFU1 function in the maturation of a subset of mitochondrial Fe-S proteins. Am J Hum Genet, 89, 656-67. NEMETH, E., RIVERA, S., GABAYAN, V., KELLER, C., TAUDORF, S., PEDERSEN, B. K. & GANZ, T. 2004a. IL-6 mediates hypoferremia of inflammation by inducing the synthesis of the iron regulatory hormone hepcidin. J Clin Invest, 113, 1271-6. NEMETH, E., ROETTO, A., GAROZZO, G., GANZ, T. & CAMASCHELLA, C. 2005. Hepcidin is decreased in TFR2 hemochromatosis. Blood, 105, 1803-6. NEMETH, E., TUTTLE, M. S., POWELSON, J., VAUGHN, M. B., DONOVAN, A., WARD, D. M., GANZ, T. & KAPLAN, J. 2004b. Hepcidin regulates cellular iron efflux by binding to ferroportin and inducing its internalization. Science, 306, 2090-3. NGUYEN, N. B., CALLAGHAN, K. D., GHIO, A. J., HAILE, D. J. & YANG, F. 2006. Hepcidin

219

expression and iron transport in alveolar macrophages. Am J Physiol Lung Cell Mol Physiol, 291, L417-25. NICHOLSON, J. K., HOLMES, E., KINROSS, J. M., DARZI, A. W., TAKATS, Z. & LINDON, J. C. 2012. Metabolic phenotyping in clinical and surgical environments. Nature, 491, 384-92. NICKEL, N., GOLPON, H., GREER, M., KNUDSEN, L., OLSSON, K., WESTERKAMP, V., WELTE, T. & HOEPER, M. M. 2012. The prognostic impact of follow-up assessments in patients with idiopathic pulmonary arterial hypertension. Eur Respir J, 39, 589-96. NICOLAS, G., BENNOUN, M., DEVAUX, I., BEAUMONT, C., GRANDCHAMP, B., KAHN, A. & VAULONT, S. 2001. Lack of hepcidin gene expression and severe tissue iron overload in upstream stimulatory factor 2 (USF2) knockout mice. Proc Natl Acad Sci U S A, 98, 8780-5. NICOLAS, G., CHAUVET, C., VIATTE, L., DANAN, J. L., BIGARD, X., DEVAUX, I., BEAUMONT, C., KAHN, A. & VAULONT, S. 2002. The gene encoding the iron regulatory peptide hepcidin is regulated by anemia, hypoxia, and inflammation. J Clin Invest, 110, 1037- 44. OHGAMI, R. S., CAMPAGNA, D. R., GREER, E. L., ANTIOCHOS, B., MCDONALD, A., CHEN, J., SHARP, J. J., FUJIWARA, Y., BARKER, J. E. & FLEMING, M. D. 2005. Identification of a ferrireductase required for efficient transferrin-dependent iron uptake in erythroid cells. Nat Genet, 37, 1264-9. OKA, M., HOMMA, N., TARASEVICIENE-STEWART, L., MORRIS, K. G., KRASKAUSKAS, D., BURNS, N., VOELKEL, N. F. & MCMURTRY, I. F. 2007. Rho kinase-mediated vasoconstriction is important in severe occlusive pulmonary arterial hypertension in rats. Circ Res, 100, 923-9. OKONKO, D. O., GRZESLO, A., WITKOWSKI, T., MANDAL, A. K., SLATER, R. M., ROUGHTON, M., FOLDES, G., THUM, T., MAJDA, J., BANASIAK, W., MISSOURIS, C. G., POOLE- WILSON, P. A., ANKER, S. D. & PONIKOWSKI, P. 2008. Effect of intravenous iron sucrose on exercise tolerance in anemic and nonanemic patients with symptomatic chronic heart failure and iron deficiency FERRIC-HF: a randomized, controlled, observer-blinded trial. J Am Coll Cardiol, 51, 103-12. OLIVEIRA FILHO, J. P., BADIAL, P. R., CUNHA, P. H., CRUZ, T. F., ARAÚJO, J. P., DIVERS, T. J., WINAND, N. J. & BORGES, A. S. 2010. Cloning, sequencing and expression analysis of the equine hepcidin gene by real-time PCR. Vet Immunol Immunopathol, 135, 34-42. OZSU, S., ABUL, Y., GUNAYDIN, S., OREM, A. & OZLU, T. 2014. Prognostic value of red cell

220

distribution width in patients with pulmonary embolism. Clin Appl Thromb Hemost, 20, 365-70. PALANGE, P., WARD, S. A., CARLSEN, K. H., CASABURI, R., GALLAGHER, C. G., GOSSELINK, R., O'DONNELL, D. E., PUENTE-MAESTU, L., SCHOLS, A. M., SINGH, S., WHIPP, B. J. & FORCE, E. T. 2007. Recommendations on the use of exercise testing in clinical practice. Eur Respir J, 29, 185-209. PAOLI, M., ANDERSON, B. F., BAKER, H. M., MORGAN, W. T., SMITH, A. & BAKER, E. N. 1999. Crystal structure of hemopexin reveals a novel high-affinity heme site formed between two beta-propeller domains. Nat Struct Biol, 6, 926-31. PAPANIKOLAOU, G., SAMUELS, M. E., LUDWIG, E. H., MACDONALD, M. L., FRANCHINI, P. L., DUBÉ, M. P., ANDRES, L., MACFARLANE, J., SAKELLAROPOULOS, N., POLITOU, M., NEMETH, E., THOMPSON, J., RISLER, J. K., ZABOROWSKA, C., BABAKAIFF, R., RADOMSKI, C. C., PAPE, T. D., DAVIDAS, O., CHRISTAKIS, J., BRISSOT, P., LOCKITCH, G., GANZ, T., HAYDEN, M. R. & GOLDBERG, Y. P. 2004a. Mutations in HFE2 cause iron overload in chromosome 1q-linked juvenile hemochromatosis. Nat Genet, 36, 77-82. PARADKAR, P. N., ZUMBRENNEN, K. B., PAW, B. H., WARD, D. M. & KAPLAN, J. 2009. Regulation of mitochondrial iron import through differential turnover of mitoferrin 1 and mitoferrin 2. Mol Cell Biol, 29, 1007-16. PARK, C. H., VALORE, E. V., WARING, A. J. & GANZ, T. 2001. Hepcidin, a urinary antimicrobial peptide synthesized in the liver. J Biol Chem, 276, 7806-10. PARTOVIAN, C., ADNOT, S., EDDAHIBI, S., TEIGER, E., LEVAME, M., DREYFUS, P., RAFFESTIN, B. & FRELIN, C. 1998a. Heart and lung VEGF mRNA expression in rats with monocrotaline- or hypoxia-induced pulmonary hypertension. Am J Physiol, 275, H1948-56. PARTOVIAN, C., LADOUX, A., EDDAHIBI, S., TEIGER, E., RAFFESTIN, B., FRELIN, C. & ADNOT, S. 1998b. Cardiac and lung VEGF mRNA expression in chronically hypoxic and monocrotaline-treated rats. Chest, 114, 45S-46S. PATEL, K. V., FERRUCCI, L., ERSHLER, W. B., LONGO, D. L. & GURALNIK, J. M. 2009. Red blood cell distribution width and the risk of death in middle-aged and older adults. Arch Intern Med, 169, 515-23. PAULIN, R. & MICHELAKIS, E. D. 2014. The metabolic theory of pulmonary arterial hypertension. Circ Res, 115, 148-64. PEACOCK, A. J., MURPHY, N. F., MCMURRAY, J. J., CABALLERO, L. & STEWART, S. 2007. An epidemiological study of pulmonary arterial hypertension. Eur Respir J, 30, 104-9.

221

PENALOZA, D. & ARIAS-STELLA, J. 2007. The heart and pulmonary circulation at high altitudes: healthy highlanders and chronic mountain sickness. Circulation, 115, 1132- 46. PERLSTEIN, T. S., WEUVE, J., PFEFFER, M. A. & BECKMAN, J. A. 2009. Red blood cell distribution width and mortality risk in a community-based prospective cohort. Arch Intern Med, 169, 588-94. PEYSSONNAUX, C., ZINKERNAGEL, A. S., SCHUEPBACH, R. A., RANKIN, E., VAULONT, S., HAASE, V. H., NIZET, V. & JOHNSON, R. S. 2007. Regulation of iron homeostasis by the hypoxia-inducible transcription factors (HIFs). J Clin Invest, 117, 1926-32. PIGEON, C., ILYIN, G., COURSELAUD, B., LEROYER, P., TURLIN, B., BRISSOT, P. & LORÉAL, O. 2001a. A new mouse liver-specific gene, encoding a protein homologous to human antimicrobial peptide hepcidin, is overexpressed during iron overload. J Biol Chem, 276, 7811-9. PIGEON, C., ILYIN, G., COURSELAUD, B., LEROYER, P., TURLIN, B., BRISSOT, P. & LORÉAL, O. 2001b. A new mouse liver-specific gene, encoding a protein homologous to human antimicrobial peptide hepcidin, is overexpressed during iron overload. J Biol Chem, 276, 7811-9. PINILLA-TENAS, J. J., SPARKMAN, B. K., SHAWKI, A., ILLING, A. C., MITCHELL, C. J., ZHAO, N., LIUZZI, J. P., COUSINS, R. J., KNUTSON, M. D. & MACKENZIE, B. 2011. Zip14 is a complex broad-scope metal-ion transporter whose functional properties support roles in the cellular uptake of zinc and nontransferrin-bound iron. Am J Physiol Cell Physiol, 301, C862-71. PINKSTAFF, S. O., BURGER, C. D., DAUGHERTY, J., BOND, S. & ARENA, R. 2016. Cardiopulmonary exercise testing in patients with pulmonary hypertension: clinical recommendations based on a review of the evidence. Expert Rev Respir Med, 10, 279-95. PINNIX, Z. K., MILLER, L. D., WANG, W., D'AGOSTINO, R., KUTE, T., WILLINGHAM, M. C., HATCHER, H., TESFAY, L., SUI, G., DI, X., TORTI, S. V. & TORTI, F. M. 2010. Ferroportin and iron regulation in breast cancer progression and prognosis. Sci Transl Med, 2, 43ra56. POLAT, V., ISCAN, S., ETLI, M., EL KILIC, H., GÜRSU, O., EKER, E. & OZDEMIR, F. 2014. Red cell distribution width as a prognostic indicator in pediatric heart disease and after surgery. Biomed Res Int, 2014, 681679. PONIKOWSKA, B., SUCHOCKI, T., PALECZNY, B., OLESINSKA, M., POWIERZA, S., BORODULIN-

222

NADZIEJA, L., RECZUCH, K., VON HAEHLING, S., DOEHNER, W., ANKER, S. D., CLELAND, J. G. & JANKOWSKA, E. A. 2013. Iron status and survival in diabetic patients with coronary artery disease. Diabetes Care, 36, 4147-56. PONIKOWSKI, P., VAN VELDHUISEN, D. J., COMIN-COLET, J., ERTL, G., KOMAJDA, M., MAREEV, V., MCDONAGH, T., PARKHOMENKO, A., TAVAZZI, L., LEVESQUE, V., MORI, C., ROUBERT, B., FILIPPATOS, G., RUSCHITZKA, F., ANKER, S. D. & INVESTIGATORS, C.- H. 2015. Beneficial effects of long-term intravenous iron therapy with ferric carboxymaltose in patients with symptomatic heart failure and iron deficiency†. Eur Heart J, 36, 657-68. PONKA, P. & RICHARDSON, D. R. 1997. Can ferritin provide iron for hemoglobin synthesis? Blood, 89, 2611-3. PRESTON, I. R., SUISSA, S. & HUMBERT, M. 2013a. New perspectives in long-term outcomes in clinical trials of pulmonary arterial hypertension. Eur Respir Rev, 22, 495-502. PRETORIUS, E. & LIPINSKI, B. 2013. Differences in morphology of fibrin clots induced with thrombin and ferric ions and its pathophysiological consequences. Heart Lung Circ, 22, 447-9. PRICE, L. C., WORT, S. J., PERROS, F., DORFMÜLLER, P., HUERTAS, A., MONTANI, D., COHEN- KAMINSKY, S. & HUMBERT, M. 2012. Inflammation in pulmonary arterial hypertension. Chest, 141, 210-221. PULLAMSETTI, S. S., SAVAI, R., JANSSEN, W., DAHAL, B. K., SEEGER, W., GRIMMINGER, F., GHOFRANI, H. A., WEISSMANN, N. & SCHERMULY, R. T. 2011. Inflammation, immunological reaction and role of infection in pulmonary hypertension. Clin Microbiol Infect, 17, 7-14. PULLAMSETTI, S., KISS, L., GHOFRANI, H. A., VOSWINCKEL, R., HAREDZA, P., KLEPETKO, W., AIGNER, C., FINK, L., MUYAL, J. P., WEISSMANN, N., GRIMMINGER, F., SEEGER, W. & SCHERMULY, R. T. 2005. Increased levels and reduced catabolism of asymmetric and symmetric dimethylarginines in pulmonary hypertension. FASEB J, 19, 1175-7. PULZ, C., DINIZ, R. V., ALVES, A. N., TEBEXRENI, A. S., CARVALHO, A. C., DE PAOLA, A. A. & ALMEIDA, D. R. 2008. Incremental shuttle and six-minute walking tests in the assessment of functional capacity in chronic heart failure. Can J Cardiol, 24, 131-5. QUIGLEY, J. G., YANG, Z., WORTHINGTON, M. T., PHILLIPS, J. D., SABO, K. M., SABATH, D. E., BERG, C. L., SASSA, S., WOOD, B. L. & ABKOWITZ, J. L. 2004. Identification of a human heme exporter that is essential for erythropoiesis. Cell, 118, 757-66. RABINOVITCH, M. 2008. Molecular pathogenesis of pulmonary arterial hypertension. J Clin

223

Invest, 118, 2372-9. RAMSAY, A. J., REID, J. C., VELASCO, G., QUIGLEY, J. P. & HOOPER, J. D. 2008. The type II transmembrane serine protease matriptase-2--identification, structural features, enzymology, expression pattern and potential roles. Front Biosci, 13, 569-79. RANKIN, E. B., BIJU, M. P., LIU, Q., UNGER, T. L., RHA, J., JOHNSON, R. S., SIMON, M. C., KEITH, B. & HAASE, V. H. 2007. Hypoxia-inducible factor-2 (HIF-2) regulates hepatic erythropoietin in vivo. J Clin Invest, 117, 1068-77. RASEKABA, T., LEE, A. L., NAUGHTON, M. T., WILLIAMS, T. J. & HOLLAND, A. E. 2009. The six- minute walk test: a useful metric for the cardiopulmonary patient. Intern Med J, 39, 495-501. REMILLARD, C. V., TIGNO, D. D., PLATOSHYN, O., BURG, E. D., BREVNOVA, E. E., CONGER, D., NICHOLSON, A., RANA, B. K., CHANNICK, R. N., RUBIN, L. J., O'CONNOR, D. T. & YUAN, J. X. 2007a. Function of Kv1.5 channels and genetic variations of KCNA5 in patients with idiopathic pulmonary arterial hypertension. Am J Physiol Cell Physiol, 292, C1837-53. REN, X., DORRINGTON, K. L., MAXWELL, P. H. & ROBBINS, P. A. 2000. Effects of desferrioxamine on serum erythropoietin and ventilatory sensitivity to hypoxia in humans. J Appl Physiol (1985), 89, 680-6. RHODES, C. J., HOWARD, L. S., BUSBRIDGE, M., ASHBY, D., KONDILI, E., GIBBS, J. S., WHARTON, J. & WILKINS, M. R. 2011a. Iron deficiency and raised hepcidin in idiopathic pulmonary arterial hypertension: clinical prevalence, outcomes, and mechanistic insights. J Am Coll Cardiol, 58, 300-9. RHODES, C. J., WHARTON, J., HOWARD, L. S., GIBBS, J. S. & WILKINS, M. R. 2011c. Red cell distribution width outperforms other potential circulating biomarkers in predicting survival in idiopathic pulmonary arterial hypertension. Heart, 97, 1054-60. RICH, J. D., ARCHER, S. L. & RICH, S. 2013. Noninvasive cardiac output measurements in patients with pulmonary hypertension. Eur Respir J, 42, 125-33. RICH, S., KAUFMANN, E. & LEVY, P. S. 1992. The effect of high doses of calcium-channel blockers on survival in primary pulmonary hypertension. N Engl J Med, 327, 76-81. ROBINSON, J. C., GRAHAM, B. B., ROUAULT, T. C. & TUDER, R. M. 2014. The crossroads of iron with hypoxia and cellular metabolism. Implications in the pathobiology of pulmonary hypertension. Am J Respir Cell Mol Biol, 51, 721-9. ROCHETTE, L., GUDJONCIK, A., GUENANCIA, C., ZELLER, M., COTTIN, Y. & VERGELY, C. 2015a. The iron-regulatory hormone hepcidin: a possible therapeutic target? Pharmacol

224

Ther, 146, 35-52. ROCHETTE, L., ZELLER, M., COTTIN, Y. & VERGELY, C. 2015b. Growth and differentiation factor 11 (GDF11): Functions in the regulation of erythropoiesis and cardiac regeneration. Pharmacol Ther, 156, 26-33. RODRIGUEZ, A., PAN, P. & PARKKILA, S. 2007. Expression studies of neogenin and its ligand hemojuvelin in mouse tissues. J Histochem Cytochem, 55, 85-96. RODRÍGUEZ-CARRIO, J., ALPERI-LÓPEZ, M., LÓPEZ, P., ALONSO-CASTRO, S., CARRO-ESTEBAN, S. R., BALLINA-GARCÍA, F. J. & SUÁREZ, A. 2015. Red cell distribution width is associated with endothelial progenitor cell depletion and vascular-related mediators in rheumatoid arthritis. Atherosclerosis, 240, 131-6. ROETTO, A., PAPANIKOLAOU, G., POLITOU, M., ALBERTI, F., GIRELLI, D., CHRISTAKIS, J., LOUKOPOULOS, D. & CAMASCHELLA, C. 2003. Mutant antimicrobial peptide hepcidin is associated with severe juvenile hemochromatosis. Nat Genet, 33, 21-2. ROGNONI, C., VENTURINI, S., MEREGAGLIA, M., MARMIFERO, M. & TARRICONE, R. 2016. Efficacy and Safety of Ferric Carboxymaltose and Other Formulations in Iron- Deficient Patients: A Systematic Review and Network Meta-analysis of Randomised Controlled Trials. Clin Drug Investig, 36, 177-94. ROSENKRANZ, S. & PRESTON, I. R. 2015. Right heart catheterisation: best practice and pitfalls in pulmonary hypertension. Eur Respir Rev, 24, 642-52. ROTHMAN, A. M., ARNOLD, N. D., CHANG, W., WATSON, O., SWIFT, A. J., CONDLIFFE, R., ELLIOT, C. A., KIELY, D. G., SUVARNA, S. K., GUNN, J. & LAWRIE, A. 2015. Pulmonary artery denervation reduces pulmonary artery pressure and induces histological changes in an acute porcine model of pulmonary hypertension. Circ Cardiovasc Interv, 8, e002569. ROUAULT, T. A. & TONG, W. H. 2005. Iron-sulphur cluster biogenesis and mitochondrial iron homeostasis. Nat Rev Mol Cell Biol, 6, 345-51. RUBIN, L. J. 2006. Pulmonary arterial hypertension. Proc Am Thorac Soc, 3, 111-5. RUITER, G., LANKHORST, S., BOONSTRA, A., POSTMUS, P. E., ZWEEGMAN, S., WESTERHOF, N., VAN DER LAARSE, W. J. & VONK-NOORDEGRAAF, A. 2011a. Iron deficiency is common in idiopathic pulmonary arterial hypertension. Eur Respir J, 37, 1386-91. RUITER, G., LANKHORST, S., BOONSTRA, A., POSTMUS, P. E., ZWEEGMAN, S., WESTERHOF, N., VAN DER LAARSE, W. J. & VONK-NOORDEGRAAF, A. 2011. Iron deficiency is common in idiopathic pulmonary arterial hypertension. Eur Respir J, 37, 1386-91. RUITER, G., LANSER, I. J., DE MAN, F. S., VAN DER LAARSE, W. J., WHARTON, J., WILKINS, M.

225

R., HOWARD, L. S., VONK-NOORDEGRAAF, A. & VOSKUYL, A. E. 2014a. Iron deficiency in systemic sclerosis patients with and without pulmonary hypertension. Rheumatology (Oxford), 53, 285-92. RUITER, G., MANDERS, E., HAPPÉ, C. M., SCHALIJ, I., GROEPENHOFF, H., HOWARD, L. S., WILKINS, M. R., BOGAARD, H. J., WESTERHOF, N., VAN DER LAARSE, W. J., DE MAN, F. S. & VONK-NOORDEGRAAF, A. 2015. Intravenous iron therapy in patients with idiopathic pulmonary arterial hypertension and iron deficiency. Pulm Circ, 5, 466-72. RYAN, J. J. & ARCHER, S. L. 2015. Emerging concepts in the molecular basis of pulmonary arterial hypertension: part I: metabolic plasticity and mitochondrial dynamics in the pulmonary circulation and right ventricle in pulmonary arterial hypertension. Circulation, 131, 1691-702. RYTER, S. W., ALAM, J. & CHOI, A. M. 2006. Heme oxygenase-1/carbon monoxide: from basic science to therapeutic applications. Physiol Rev, 86, 583-650. SAKAO, S., TARASEVICIENE-STEWART, L., LEE, J. D., WOOD, K., COOL, C. D. & VOELKEL, N. F. 2005. Initial apoptosis is followed by increased proliferation of apoptosis-resistant endothelial cells. FASEB J, 19, 1178-80. SALVAGNO, G. L., SANCHIS-GOMAR, F., PICANZA, A. & LIPPI, G. 2015a. Red blood cell distribution width: A simple parameter with multiple clinical applications. Crit Rev Clin Lab Sci, 52, 86-105. SANCHEZ, M., GALY, B., MUCKENTHALER, M. U. & HENTZE, M. W. 2007a. Iron-regulatory proteins limit hypoxia-inducible factor-2alpha expression in iron deficiency. Nat Struct Mol Biol, 14, 420-6. SANCHEZ, O., MARCOS, E., PERROS, F., FADEL, E., TU, L., HUMBERT, M., DARTEVELLE, P., SIMONNEAU, G., ADNOT, S. & EDDAHIBI, S. 2007c. Role of endothelium-derived CC chemokine ligand 2 in idiopathic pulmonary arterial hypertension. Am J Respir Crit Care Med, 176, 1041-7. SAVAI, R., PULLAMSETTI, S. S., KOLBE, J., BIENIEK, E., VOSWINCKEL, R., FINK, L., SCHEED, A., RITTER, C., DAHAL, B. K., VATER, A., KLUSSMANN, S., GHOFRANI, H. A., WEISSMANN, N., KLEPETKO, W., BANAT, G. A., SEEGER, W., GRIMMINGER, F. & SCHERMULY, R. T. 2012. Immune and inflammatory cell involvement in the pathology of idiopathic pulmonary arterial hypertension. Am J Respir Crit Care Med, 186, 897-908. SAVARESE, G., PAOLILLO, S., COSTANZO, P., D'AMORE, C., CECERE, M., LOSCO, T., MUSELLA, F., GARGIULO, P., MARCIANO, C. & PERRONE-FILARDI, P. 2012. Do changes of 6- minute walk distance predict clinical events in patients with pulmonary arterial

226

hypertension? A meta-analysis of 22 randomized trials. J Am Coll Cardiol, 60, 1192- 201. SAWICKI, K. T., SHANG, M., WU, R., CHANG, H. C., KHECHADURI, A., SATO, T., KAMIDE, C., LIU, T., NAGA PRASAD, S. V. & ARDEHALI, H. 2015. Increased Heme Levels in the Heart Lead to Exacerbated Ischemic Injury. J Am Heart Assoc, 4, e002272. SCHAER, D. J., SCHAER, C. A., BUEHLER, P. W., BOYKINS, R. A., SCHOEDON, G., ALAYASH, A. I. & SCHAFFNER, A. 2006. CD163 is the macrophage scavenger receptor for native and chemically modified hemoglobins in the absence of haptoglobin. Blood, 107, 373-80. SCHERMULY, R. T., GHOFRANI, H. A., WILKINS, M. R. & GRIMMINGER, F. 2011. Mechanisms of disease: pulmonary arterial hypertension. Nat Rev Cardiol, 8, 443-55. SEMENZA, G. L. 2011. Oxygen sensing, homeostasis, and disease. N Engl J Med, 365, 537-47. SHAH, Y. M., MATSUBARA, T., ITO, S., YIM, S. H. & GONZALEZ, F. J. 2009. Intestinal hypoxia- inducible transcription factors are essential for iron absorption following iron deficiency. Cell Metab, 9, 152-64. SHAYEGHI, M., LATUNDE-DADA, G. O., OAKHILL, J. S., LAFTAH, A. H., TAKEUCHI, K., HALLIDAY, N., KHAN, Y., WARLEY, A., MCCANN, F. E., HIDER, R. C., FRAZER, D. M., ANDERSON, G. J., VULPE, C. D., SIMPSON, R. J. & MCKIE, A. T. 2005. Identification of an intestinal heme transporter. Cell, 122, 789-801. SHINTANI, M., YAGI, H., NAKAYAMA, T., SAJI, T. & MATSUOKA, R. 2009. A new nonsense mutation of SMAD8 associated with pulmonary arterial hypertension. J Med Genet, 46, 331-7. SILVESTRI, L., PAGANI, A. & CAMASCHELLA, C. 2008a. Furin-mediated release of soluble hemojuvelin: a new link between hypoxia and iron homeostasis. Blood, 111, 924-31. SILVESTRI, L., PAGANI, A., NAI, A., DE DOMENICO, I., KAPLAN, J. & CAMASCHELLA, C. 2008b. The serine protease matriptase-2 (TMPRSS6) inhibits hepcidin activation by cleaving membrane hemojuvelin. Cell Metab, 8, 502-11. SIMONNEAU, G., GATZOULIS, M. A., ADATIA, I., CELERMAJER, D., DENTON, C., GHOFRANI, A., GOMEZ SANCHEZ, M. A., KRISHNA KUMAR, R., LANDZBERG, M., MACHADO, R. F., OLSCHEWSKI, H., ROBBINS, I. M. & SOUZA, R. 2013. Updated clinical classification of pulmonary hypertension. J Am Coll Cardiol, 62, D34-41. SINGH, S. J., MORGAN, M. D., SCOTT, S., WALTERS, D. & HARDMAN, A. E. 1992. Development of a shuttle walking test of disability in patients with chronic airways obstruction. Thorax, 47, 1019-24. SITBON, O., HUMBERT, M., JAÏS, X., IOOS, V., HAMID, A. M., PROVENCHER, S., GARCIA, G.,

227

PARENT, F., HERVÉ, P. & SIMONNEAU, G. 2005a. Long-term response to calcium channel blockers in idiopathic pulmonary arterial hypertension. Circulation, 111, 3105-11. SITBON, O., HUMBERT, M., NUNES, H., PARENT, F., GARCIA, G., HERVÉ, P., RAINISIO, M. & SIMONNEAU, G. 2002. Long-term intravenous epoprostenol infusion in primary pulmonary hypertension: prognostic factors and survival. J Am Coll Cardiol, 40, 780- 8. SITBON, O., MCLAUGHLIN, V. V., BADESCH, D. B., BARST, R. J., BLACK, C., GALIÈ, N., HUMBERT, M., RAINISIO, M., RUBIN, L. J. & SIMONNEAU, G. 2005b. Survival in patients with class III idiopathic pulmonary arterial hypertension treated with first line oral bosentan compared with an historical cohort of patients started on intravenous epoprostenol. Thorax, 60, 1025-30. SKIKNE, B. S., FLOWERS, C. H. & COOK, J. D. 1990. Serum transferrin receptor: a quantitative measure of tissue iron deficiency. Blood, 75, 1870-6. SMITH, T. G., BALANOS, G. M., CROFT, Q. P., TALBOT, N. P., DORRINGTON, K. L., RATCLIFFE, P. J. & ROBBINS, P. A. 2008. The increase in pulmonary arterial pressure caused by hypoxia depends on iron status. J Physiol, 586, 5999-6005. SMITH, T. G., BROOKS, J. T., BALANOS, G. M., LAPPIN, T. R., LAYTON, D. M., LEEDHAM, D. L., LIU, C., MAXWELL, P. H., MCMULLIN, M. F., MCNAMARA, C. J., PERCY, M. J., PUGH, C. W., RATCLIFFE, P. J., TALBOT, N. P., TREACY, M. & ROBBINS, P. A. 2006. Mutation of von Hippel-Lindau tumour suppressor and human cardiopulmonary physiology. PLoS Med, 3, e290. SMITH, T. G., TALBOT, N. P., PRIVAT, C., RIVERA-CH, M., NICKOL, A. H., RATCLIFFE, P. J., DORRINGTON, K. L., LEÓN-VELARDE, F. & ROBBINS, P. A. 2009a. Effects of iron supplementation and depletion on hypoxic pulmonary hypertension: two randomized controlled trials. JAMA, 302, 1444-50. SÖDERHOLM, M., BORNÉ, Y., HEDBLAD, B., PERSSON, M. & ENGSTRÖM, G. 2015. Red cell distribution width in relation to incidence of stroke and carotid atherosclerosis: a population-based cohort study. PLoS One, 10, e0124957. SOON, E., TREACY, C. M., TOSHNER, M. R., MACKENZIE-ROSS, R., MANGLAM, V., BUSBRIDGE, M., SINCLAIR-MCGARVIE, M., ARNOLD, J., SHEARES, K. K., MORRELL, N. W. & PEPKE- ZABA, J. 2011. Unexplained iron deficiency in idiopathic and heritable pulmonary arterial hypertension. Thorax, 66, 326-32. SOUBRIER, F., CHUNG, W. K., MACHADO, R., GRÜNIG, E., ALDRED, M., GERACI, M., LOYD, J.

228

E., ELLIOTT, C. G., TREMBATH, R. C., NEWMAN, J. H. & HUMBERT, M. 2013a. Genetics and genomics of pulmonary arterial hypertension. J Am Coll Cardiol, 62, D13-21. STACHER, E., GRAHAM, B. B., HUNT, J. M., GANDJEVA, A., GROSHONG, S. D., MCLAUGHLIN, V. V., JESSUP, M., GRIZZLE, W. E., ALDRED, M. A., COOL, C. D. & TUDER, R. M. 2012. Modern age pathology of pulmonary arterial hypertension. Am J Respir Crit Care Med, 186, 261-72. STEARMAN, R. S., CORNELIUS, A. R., LU, X., CONKLIN, D. S., DEL ROSARIO, M. J., LOWE, A. M., ELOS, M. T., FETTIG, L. M., WONG, R. E., HARA, N., COGAN, J. D., PHILLIPS, J. A., TAYLOR, M. R., GRAHAM, B. B., TUDER, R. M., LOYD, J. E. & GERACI, M. W. 2014. Functional prostacyclin synthase promoter polymorphisms. Impact in pulmonary arterial hypertension. Am J Respir Crit Care Med, 189, 1110-20. STEHLING, O., WILBRECHT, C. & LILL, R. 2014. Mitochondrial iron-sulfur protein biogenesis and human disease. Biochimie, 100, 61-77. STEIER, J., JOLLEY, C. J., SEYMOUR, J., ROUGHTON, M., POLKEY, M. I. & MOXHAM, J. 2009. Neural respiratory drive in obesity. Thorax, 64, 719-25. STORZ, J. F. 2010. Evolution. Genes for high altitudes. Science, 329, 40-1. STRANGE, G., GABBAY, E., KERMEEN, F., WILLIAMS, T., CARRINGTON, M., STEWART, S. & KEOGH, A. 2013. Time from symptoms to definitive diagnosis of idiopathic pulmonary arterial hypertension: The delay study. Pulm Circ, 3, 89-94. SUEDEKUM, N. A. & DIMEFF, R. J. 2005. Iron and the athlete. Curr Sports Med Rep, 4, 199- 202. SUMMER, R., FIACK, C. A., IKEDA, Y., SATO, K., DWYER, D., OUCHI, N., FINE, A., FARBER, H. W. & WALSH, K. 2009. Adiponectin deficiency: a model of pulmonary hypertension associated with pulmonary vascular disease. Am J Physiol Lung Cell Mol Physiol, 297, L432-8. SUTENDRA, G. & MICHELAKIS, E. D. 2014. The metabolic basis of pulmonary arterial hypertension. Cell Metab, 19, 558-73. SWIFT, A. J., CAPENER, D., HAMMERTON, C., THOMAS, S. M., ELLIOT, C., CONDLIFFE, R., WILD, J. M. & KIELY, D. G. 2015. Right ventricular sex differences in patients with idiopathic pulmonary arterial hypertension characterised by magnetic resonance imaging: pair-matched case controlled study. PLoS One, 10, e0127415. SZTRYMF, B., COULET, F., GIRERD, B., YAICI, A., JAIS, X., SITBON, O., MONTANI, D., SOUZA, R., SIMONNEAU, G., SOUBRIER, F. & HUMBERT, M. 2008. Clinical outcomes of

229

pulmonary arterial hypertension in carriers of BMPR2 mutation. Am J Respir Crit Care Med, 177, 1377-83. TALBOT, N. P., LAKHAL, S., SMITH, T. G., PRIVAT, C., NICKOL, A. H., RIVERA-CH, M., LEÓN- VELARDE, F., DORRINGTON, K. L., MOLE, D. R. & ROBBINS, P. A. 2012. Regulation of hepcidin expression at high altitude. Blood, 119, 857-60. TAN, Q., KERESTES, H., PERCY, M. J., PIETROFESA, R., CHEN, L., KHURANA, T. S., CHRISTOFIDOU-SOLOMIDOU, M., LAPPIN, T. R. & LEE, F. S. 2013. Erythrocytosis and pulmonary hypertension in a mouse model of human HIF2A gain of function mutation. J Biol Chem, 288, 17134-44. TANAKA, T., ROY, C. N., YAO, W., MATTEINI, A., SEMBA, R. D., ARKING, D., WALSTON, J. D., FRIED, L. P., SINGLETON, A., GURALNIK, J., ABECASIS, G. R., BANDINELLI, S., LONGO, D. L. & FERRUCCI, L. 2010. A genome-wide association analysis of serum iron concentrations. Blood, 115, 94-6. TANNO, T., BHANU, N. V., ONEAL, P. A., GOH, S. H., STAKER, P., LEE, Y. T., MORONEY, J. W., REED, C. H., LUBAN, N. L., WANG, R. H., ELING, T. E., CHILDS, R., GANZ, T., LEITMAN, S. F., FUCHAROEN, S. & MILLER, J. L. 2007. High levels of GDF15 in thalassemia suppress expression of the iron regulatory protein hepcidin. Nat Med, 13, 1096-101. TANNO, T., PORAYETTE, P., SRIPICHAI, O., NOH, S. J., BYRNES, C., BHUPATIRAJU, A., LEE, Y. T., GOODNOUGH, J. B., HARANDI, O., GANZ, T., PAULSON, R. F. & MILLER, J. L. 2009. Identification of TWSG1 as a second novel erythroid regulator of hepcidin expression in murine and human cells. Blood, 114, 181-6. TARASEVICIENE-STEWART, L., KASAHARA, Y., ALGER, L., HIRTH, P., MC MAHON, G., WALTENBERGER, J., VOELKEL, N. F. & TUDER, R. M. 2001. Inhibition of the VEGF receptor 2 combined with chronic hypoxia causes cell death-dependent pulmonary endothelial cell proliferation and severe pulmonary hypertension. FASEB J, 15, 427- 38. TARASEVICIENE-STEWART, L., SCERBAVICIUS, R., CHOE, K. H., COOL, C., WOOD, K., TUDER, R. M., BURNS, N., KASPER, M. & VOELKEL, N. F. 2006. Simvastatin causes endothelial cell apoptosis and attenuates severe pulmonary hypertension. Am J Physiol Lung Cell Mol Physiol, 291, L668-76. TAY, E. L., PESET, A., PAPAPHYLACTOU, M., INUZUKA, R., ALONSO-GONZALEZ, R., GIANNAKOULAS, G., TZIFA, A., GOLETTO, S., BROBERG, C., DIMOPOULOS, K. & GATZOULIS, M. A. 2011a. Replacement therapy for iron deficiency improves exercise capacity and quality of life in patients with cyanotic congenital heart disease and/or

230

the Eisenmenger syndrome. Int J Cardiol, 151, 307-12. TAY, E. L., PESET, A., PAPAPHYLACTOU, M., INUZUKA, R., ALONSO-GONZALEZ, R., GIANNAKOULAS, G., TZIFA, A., GOLETTO, S., BROBERG, C., DIMOPOULOS, K. & GATZOULIS, M. A. 2011b. Replacement therapy for iron deficiency improves exercise capacity and quality of life in patients with cyanotic congenital heart disease and/or the Eisenmenger syndrome. Int J Cardiol, 151, 307-12. TENORIO, J., NAVAS, P., BARRIOS, E., FERNÁNDEZ, L., NEVADO, J., QUEZADA, C. A., LÓPEZ- MESEGUER, M., ARIAS, P., MENA, R., LOBO, J. L., ALVAREZ, C., HEATH, K., ESCRIBANO-SUBÍAS, P. & LAPUNZINA, P. 2015. A founder EIF2AK4 mutation causes an aggressive form of pulmonary arterial hypertension in Iberian Gypsies. Clin Genet, 88, 579-83. THOMSON, J. R., MACHADO, R. D., PAUCIULO, M. W., MORGAN, N. V., HUMBERT, M., ELLIOTT, G. C., WARD, K., YACOUB, M., MIKHAIL, G., ROGERS, P., NEWMAN, J., WHEELER, L., HIGENBOTTAM, T., GIBBS, J. S., EGAN, J., CROZIER, A., PEACOCK, A., ALLCOCK, R., CORRIS, P., LOYD, J. E., TREMBATH, R. C. & NICHOLS, W. C. 2000. Sporadic primary pulmonary hypertension is associated with germline mutations of the gene encoding BMPR-II, a receptor member of the TGF-beta family. J Med Genet, 37, 741-5. TOBLLI, J. E., CAO, G., RIVAS, C. & KULAKSIZ, H. 2008. Heart and iron deficiency anaemia in rats with renal insufficiency: the role of hepcidin. Nephrology (Carlton), 13, 636-45. TOURNIER, A., WAHL, D., CHAOUAT, A., MAX, J. P., REGNAULT, V., LECOMPTE, T. & CHABOT, F. 2010. Calibrated automated thrombography demonstrates hypercoagulability in patients with idiopathic pulmonary arterial hypertension. Thromb Res, 126, e418-22. TREMBATH, R. C., THOMSON, J. R., MACHADO, R. D., MORGAN, N. V., ATKINSON, C., WINSHIP, I., SIMONNEAU, G., GALIE, N., LOYD, J. E., HUMBERT, M., NICHOLS, W. C., MORRELL, N. W., BERG, J., MANES, A., MCGAUGHRAN, J., PAUCIULO, M. & WHEELER, L. 2001. Clinical and molecular genetic features of pulmonary hypertension in patients with hereditary hemorrhagic telangiectasia. N Engl J Med, 345, 325-34. TRUKSA, J., PENG, H., LEE, P. & BEUTLER, E. 2006. Bone morphogenetic proteins 2, 4, and 9 stimulate murine hepcidin 1 expression independently of Hfe, transferrin receptor 2 (Tfr2), and IL-6. Proc Natl Acad Sci U S A, 103, 10289-93. TUDER, R. M. & VOELKEL, N. F. 1998. Pulmonary hypertension and inflammation. J Lab Clin Med, 132, 16-24.

231

TUDER, R. M. & YUN, J. H. 2008. Vascular endothelial growth factor of the lung: friend or foe. Curr Opin Pharmacol, 8, 255-60. TUDER, R. M. 2014. How do we measure pathology in PAH (lung and RV) and what does it tell us about the disease. Drug Discov Today, 19, 1257-63. TUDER, R. M., ABMAN, S. H., BRAUN, T., CAPRON, F., STEVENS, T., THISTLETHWAITE, P. A. & HAWORTH, S. G. 2009. Development and pathology of pulmonary hypertension. J Am Coll Cardiol, 54, S3-9. TUDER, R. M., ARCHER, S. L., DORFMÜLLER, P., ERZURUM, S. C., GUIGNABERT, C., MICHELAKIS, E., RABINOVITCH, M., SCHERMULY, R., STENMARK, K. R. & MORRELL, N. W. 2013a. Relevant issues in the pathology and pathobiology of pulmonary hypertension. J Am Coll Cardiol, 62, D4-12. TUDER, R. M., CHACON, M., ALGER, L., WANG, J., TARASEVICIENE-STEWART, L., KASAHARA, Y., COOL, C. D., BISHOP, A. E., GERACI, M., SEMENZA, G. L., YACOUB, M., POLAK, J. M. & VOELKEL, N. F. 2001. Expression of angiogenesis-related molecules in plexiform lesions in severe pulmonary hypertension: evidence for a process of disordered angiogenesis. J Pathol, 195, 367-74. TUDER, R. M., COOL, C. D., GERACI, M. W., WANG, J., ABMAN, S. H., WRIGHT, L., BADESCH, D. & VOELKEL, N. F. 1999. Prostacyclin synthase expression is decreased in lungs from patients with severe pulmonary hypertension. Am J Respir Crit Care Med, 159, 1925-32. TUDER, R. M., GROVES, B., BADESCH, D. B. & VOELKEL, N. F. 1994. Exuberant endothelial cell growth and elements of inflammation are present in plexiform lesions of pulmonary hypertension. Am J Pathol, 144, 275-85. UCHIDA, S., SCHNEIDER, A., WIESNET, M., JUNGBLUT, B., ZARJITSKAYA, P., JENNICHES, K., KREYMBORG, K. G., SEEGER, W. & BRAUN, T. 2009a. An integrated approach for the systematic identification and characterization of heart-enriched genes with unknown functions. BMC Genomics, 10, 100. UPTON, P. D. & MORRELL, N. W. 2013. The transforming growth factor-β-bone morphogenetic protein type signalling pathway in pulmonary vascular homeostasis and disease. Exp Physiol, 98, 1262-6. UPTON, P. D., DAVIES, R. J., TREMBATH, R. C. & MORRELL, N. W. 2009a. Bone morphogenetic protein (BMP) and activin type II receptors balance BMP9 signals mediated by activin receptor-like kinase-1 in human pulmonary artery endothelial cells. J Biol Chem, 284, 15794-804.

232

VAN BREDA, F., EMANS, M. E., VAN DER PUTTEN, K., BRAAM, B., VAN ITTERSUM, F. J., KRAAIJENHAGEN, R. J., DE BORST, M. H., VERVLOET, M. & GAILLARD, C. A. 2015. Relation between Red Cell Distribution Width and Fibroblast Growth Factor 23 Cleaving in Patients with Chronic Kidney Disease and Heart Failure. PLoS One, 10, e0128994. VAN CRAENENBROECK, E. M., CONRAADS, V. M., GREENLAW, N., GAUDESIUS, G., MORI, C., PONIKOWSKI, P. & ANKER, S. D. 2013. The effect of intravenous ferric carboxymaltose on red cell distribution width: a subanalysis of the FAIR-HF study. Eur J Heart Fail, 15, 756-62. VAN DE BRUAENE, A., DELCROIX, M., PASQUET, A., DE BACKER, J., DE PAUW, M., NAEIJE, R., VACHIÉRY, J. L., PAELINCK, B., MORISSENS, M. & BUDTS, W. 2011. Iron deficiency is associated with adverse outcome in Eisenmenger patients. Eur Heart J, 32, 2790-9. VAN EMPEL, V. P., LEE, J., WILLIAMS, T. J. & KAYE, D. M. 2014. Iron deficiency in patients with idiopathic pulmonary arterial hypertension. Heart Lung Circ, 23, 287-92. VAYÁ, A., HERNÁNDEZ, V., RIVERA, L., HERNÁNDEZ, J. L., LAGO, A., ESPAÑA, F. & BAUTISTA, D. 2015. Red blood cell distribution width in patients with cryptogenic stroke. Clin Appl Thromb Hemost, 21, 241-5. VEERANNA, V., ZALAWADIYA, S. K., PANAICH, S., PATEL, K. V. & AFONSO, L. 2013. Comparative analysis of red cell distribution width and high sensitivity C-reactive protein for coronary heart disease mortality prediction in multi-ethnic population: findings from the 1999-2004 NHANES. Int J Cardiol, 168, 5156-61. VIALES, R. R., EICHSTAEDT, C. A., EHLKEN, N., FISCHER, C., LICHTBLAU, M., GRÜNIG, E. & HINDERHOFER, K. 2015. Mutation in BMPR2 Promoter: A 'Second Hit' for Manifestation of Pulmonary Arterial Hypertension? PLoS One, 10, e0133042. VIETHEN, T., GERHARDT, F., DUMITRESCU, D., KNOOP-BUSCH, S., TEN FREYHAUS, H., RUDOLPH, T. K., BALDUS, S. & ROSENKRANZ, S. 2014a. Ferric carboxymaltose improves exercise capacity and quality of life in patients with pulmonary arterial hypertension and iron deficiency: a pilot study. Int J Cardiol, 175, 233-9. VOLKE, M., GALE, D. P., MAEGDEFRAU, U., SCHLEY, G., KLANKE, B., BOSSERHOFF, A. K., MAXWELL, P. H., ECKARDT, K. U. & WARNECKE, C. 2009. Evidence for a lack of a direct transcriptional suppression of the iron regulatory peptide hepcidin by hypoxia-inducible factors. PLoS One, 4, e7875. WANG, G. L. & SEMENZA, G. L. 1993. Desferrioxamine induces erythropoietin gene expression and hypoxia-inducible factor 1 DNA-binding activity: implications for

233

models of hypoxia signal transduction. Blood, 82, 3610-5. WANG, G. L. & SEMENZA, G. L. 1995a. Purification and characterization of hypoxia-inducible factor 1. J Biol Chem, 270, 1230-7. WANG, G., FAN, R., JI, R., ZOU, W., PENNY, D. J., VARGHESE, N. P. & FAN, Y. 2016. Novel homozygous BMP9 nonsense mutation causes pulmonary arterial hypertension: a case report. BMC Pulm Med, 16, 17. WANG, G., KNIGHT, L., JI, R., LAWRENCE, P., KANAAN, U., LI, L., DAS, A., CUI, B., ZOU, W., PENNY, D. J. & FAN, Y. 2014. Early onset severe pulmonary arterial hypertension with 'two-hit' digenic mutations in both BMPR2 and KCNA5 genes. Int J Cardiol, 177, e167-9. WANG, H., LI, W., ZHANG, W., SUN, K., SONG, X., GAO, S., ZHANG, C., HUI, R. & HU, H. 2009. Novel promoter and exon mutations of the BMPR2 gene in Chinese patients with pulmonary arterial hypertension. Eur J Hum Genet, 17, 1063-9. WANG, R. H., LI, C., XU, X., ZHENG, Y., XIAO, C., ZERFAS, P., COOPERMAN, S., ECKHAUS, M., ROUAULT, T., MISHRA, L. & DENG, C. X. 2005. A role of SMAD4 in iron metabolism through the positive regulation of hepcidin expression. Cell Metab, 2, 399-409. WATSON, G. M. & FOX, K. F. 2010. Are obesity and left ventricular dysfunction entwined? Eur J Echocardiogr, 11, 422-3. WEIR, E. K., TUCKER, A., REEVES, J. T., WILL, D. H. & GROVER, R. F. 1974. The genetic factor influencing pulmonary hypertension in cattle at high altitude. Cardiovasc Res, 8, 745- 9. WEISS, G. & GOODNOUGH, L. T. 2005. Anemia of chronic disease. N Engl J Med, 352, 1011- 23. WEISS, G. 2009. Iron metabolism in the anemia of chronic disease. Biochim Biophys Acta, 1790, 682-93. WHARTON, J., STRANGE, J. W., MØLLER, G. M., GROWCOTT, E. J., REN, X., FRANKLYN, A. P., PHILLIPS, S. C. & WILKINS, M. R. 2005. Antiproliferative effects of phosphodiesterase type 5 inhibition in human pulmonary artery cells. Am J Respir Crit Care Med, 172, 105-13. WHITE, K., LU, Y., ANNIS, S., HALE, A. E., CHAU, B. N., DAHLMAN, J. E., HEMANN, C., OPOTOWSKY, A. R., VARGAS, S. O., ROSAS, I., PERRELLA, M. A., OSORIO, J. C., HALEY, K. J., GRAHAM, B. B., KUMAR, R., SAGGAR, R., WALLACE, W. D., ROSS, D. J., KHAN, O. F., BADER, A., GOCHUICO, B. R., MATAR, M., POLACH, K., JOHANNESSEN, N. M., PROSSER, H. M., ANDERSON, D. G., LANGER, R., ZWEIER, J. L., BINDOFF, L. A.,

234

SYSTROM, D., WAXMAN, A. B., JIN, R. C. & CHAN, S. Y. 2015. Genetic and hypoxic alterations of the microRNA-210-ISCU1/2 axis promote iron-sulfur deficiency and pulmonary hypertension. EMBO Mol Med, 7, 695-713. WHITE, R. J., MEOLI, D. F., SWARTHOUT, R. F., KALLOP, D. Y., GALARIA, I. I., HARVEY, J. L., MILLER, C. M., BLAXALL, B. C., HALL, C. M., PIERCE, R. A., COOL, C. D. & TAUBMAN, M. B. 2007. Plexiform-like lesions and increased tissue factor expression in a rat model of severe pulmonary arterial hypertension. Am J Physiol Lung Cell Mol Physiol, 293, L583-90. WHITFIELD, J. B., CULLEN, L. M., JAZWINSKA, E. C., POWELL, L. W., HEATH, A. C., ZHU, G., DUFFY, D. L. & MARTIN, N. G. 2000. Effects of HFE C282Y and H63D polymorphisms and polygenic background on iron stores in a large community sample of twins. Am J Hum Genet, 66, 1246-58. WILKENS, H., GRIMMINGER, F., HOEPER, M., STÄHLER, G., EHLKEN, B., PLESNILA-FRANK, C., BERGER, K., RESCH, A. & GHOFRANI, A. 2010. Burden of pulmonary arterial hypertension in Germany. Respir Med, 104, 902-10. WOLF, M. & WHITE, K. E. 2014. Coupling fibroblast growth factor 23 production and cleavage: iron deficiency, rickets, and kidney disease. Curr Opin Nephrol Hypertens, 23, 411-9. WOLF, M., KOCH, T. A. & BREGMAN, D. B. 2013. Effects of iron deficiency anemia and its treatment on fibroblast growth factor 23 and phosphate homeostasis in women. J Bone Miner Res, 28, 1793-803. WONG, C. M., PRESTON, I. R., HILL, N. S. & SUZUKI, Y. J. 2012. Iron chelation inhibits the development of pulmonary vascular remodeling. Free Radic Biol Med, 53, 1738-47. WOO, D. K. & SHADEL, G. S. 2011. Mitochondrial stress signals revise an old aging theory. Cell, 144, 11-2. WOODERCHAK-DONAHUE, W. L., MCDONALD, J., O'FALLON, B., UPTON, P. D., LI, W., ROMAN, B. L., YOUNG, S., PLANT, P., FÜLÖP, G. T., LANGA, C., MORRELL, N. W., BOTELLA, L. M., BERNABEU, C., STEVENSON, D. A., RUNO, J. R. & BAYRAK-TOYDEMIR, P. 2013. BMP9 mutations cause a vascular-anomaly syndrome with phenotypic overlap with hereditary hemorrhagic telangiectasia. Am J Hum Genet, 93, 530-7. XI, Q., LIU, Z., LIU, W., ZHAO, Z., LUO, Q. & HUANG, Z. 2015. Chronic thromboembolic pulmonary hypertension is not associated with iron overload. Cardiovasc Pathol, 24, 76-9. XIA, Y., BABITT, J. L., SIDIS, Y., CHUNG, R. T. & LIN, H. Y. 2008. Hemojuvelin regulates hepcidin

235

expression via a selective subset of BMP ligands and receptors independently of neogenin. Blood, 111, 5195-204. YU, A. Y., SHIMODA, L. A., IYER, N. V., HUSO, D. L., SUN, X., MCWILLIAMS, R., BEATY, T., SHAM, J. S., WIENER, C. M., SYLVESTER, J. T. & SEMENZA, G. L. 1999. Impaired physiological responses to chronic hypoxia in mice partially deficient for hypoxia- inducible factor 1alpha. J Clin Invest, 103, 691-6. YU, P. B., BEPPU, H., KAWAI, N., LI, E. & BLOCH, K. D. 2005. Bone morphogenetic protein (BMP) type II receptor deletion reveals BMP ligand-specific gain of signaling in pulmonary artery smooth muscle cells. J Biol Chem, 280, 24443-50. YU, P. B., HONG, C. C., SACHIDANANDAN, C., BABITT, J. L., DENG, D. Y., HOYNG, S. A., LIN, H. Y., BLOCH, K. D. & PETERSON, R. T. 2008. Dorsomorphin inhibits BMP signals required for embryogenesis and iron metabolism. Nat Chem Biol, 4, 33-41. YU, Y., FANTOZZI, I., REMILLARD, C. V., LANDSBERG, J. W., KUNICHIKA, N., PLATOSHYN, O., TIGNO, D. D., THISTLETHWAITE, P. A., RUBIN, L. J. & YUAN, J. X. 2004. Enhanced expression of transient receptor potential channels in idiopathic pulmonary arterial hypertension. Proc Natl Acad Sci U S A, 101, 13861-6. YUAN, J. X., ALDINGER, A. M., JUHASZOVA, M., WANG, J., CONTE, J. V., GAINE, S. P., ORENS, J. B. & RUBIN, L. J. 1998. Dysfunctional voltage-gated K+ channels in pulmonary artery smooth muscle cells of patients with primary pulmonary hypertension. Circulation, 98, 1400-6. ZENKE-KAWASAKI, Y., DOHI, Y., KATOH, Y., IKURA, T., IKURA, M., ASAHARA, T., TOKUNAGA, F., IWAI, K. & IGARASHI, K. 2007. Heme induces ubiquitination and degradation of the transcription factor Bach1. Mol Cell Biol, 27, 6962-71. ZHANG, A. S. & ENNS, C. A. 2009. Iron homeostasis: recently identified proteins provide insight into novel control mechanisms. J Biol Chem, 284, 711-5. ZHANG, A. S., ANDERSON, S. A., MEYERS, K. R., HERNANDEZ, C., EISENSTEIN, R. S. & ENNS, C. A. 2007. Evidence that inhibition of hemojuvelin shedding in response to iron is mediated through neogenin. J Biol Chem, 282, 12547-56. ZHANG, A. S., GAO, J., KOEBERL, D. D. & ENNS, C. A. 2010. The role of hepatocyte hemojuvelin in the regulation of bone morphogenic protein-6 and hepcidin expression in vivo. J Biol Chem, 285, 16416-23. ZHANG, A. S., YANG, F., WANG, J., TSUKAMOTO, H. & ENNS, C. A. 2009. Hemojuvelin- neogenin interaction is required for bone morphogenic protein-4-induced hepcidin expression. J Biol Chem, 284, 22580-9.

236

ZHANG, F., TAO, Y., ZHANG, Z., GUO, X., AN, P., SHEN, Y., WU, Q., YU, Y. & WANG, F. 2012. Metalloreductase Steap3 coordinates the regulation of iron homeostasis and inflammatory responses. Haematologica, 97, 1826-35. ZHAO, L., OLIVER, E., MARATOU, K., ATANUR, S. S., DUBOIS, O. D., COTRONEO, E., CHEN, C. N., WANG, L., ARCE, C., CHABOSSEAU, P. L., PONSA-COBAS, J., FRID, M. G., MOYON, B., WEBSTER, Z., ALDASHEV, A., FERRER, J., RUTTER, G. A., STENMARK, K. R., AITMAN, T. J. & WILKINS, M. R. 2015. The zinc transporter ZIP12 regulates the pulmonary vascular response to chronic hypoxia. Nature, 524, 356-60. ZHAO, Y., PENG, J., LU, C., HSIN, M., MURA, M., WU, L., CHU, L., ZAMEL, R., MACHUCA, T., WADDELL, T., LIU, M., KESHAVJEE, S., GRANTON, J. & DE PERROT, M. 2014. Metabolomic heterogeneity of pulmonary arterial hypertension. PLoS One, 9, e88727. ZHU, Y. I. & HAAS, J. D. 1998. Altered metabolic response of iron-depleted nonanemic women during a 15-km time trial. J Appl Physiol (1985), 84, 1768-75. ZIMMERMANN, M. B. & HURRELL, R. F. 2007. Nutritional iron deficiency. Lancet, 370, 511- 20. ZORLU, A., BEKTASOGLU, G., GUVEN, F. M., DOGAN, O. T., GUCUK, E., EGE, M. R., ALTAY, H., CINAR, Z., TANDOGAN, I. & YILMAZ, M. B. 2012. Usefulness of admission red cell distribution width as a predictor of early mortality in patients with acute pulmonary embolism. Am J Cardiol, 109, 128-34.

237

Appendix 3.1

Right heart catheterisation manual 1. INTRODUCTION Right heart catheterisation is necessary to obtain haemodynamic measurements for the evaluation of the primary endpoint, change in PVR, and secondary endpoints.

To guarantee standardized conditions across all participating study centers, investigators must ensure that the principles defined in this manual are applied for all study relevant haemodynamic measurements.

2. TIMING OF CATHETERISATION IN RELATION TO PH DRUG ADMINISTRATION The timing of administration of pulmonary hypertension-specific drugs in relation to the timing of cardiac catheterisation could have significant impacts on the outcome measures and thus needs to be standardized. The table below is a guide to when the last dose of medication should be taken prior to catheterisation. This should be documented and repeated for the second cardiac catheter. The timing of diuretics and Nil By Mouth instructions should be managed according to local policy to avoid discomfort / dehydration during the procedure.

Time of Last Dose of Medication

Planned Time of Morning List Afternoon List Procedure Sildenafil 00:00 08:00

Tadalafil Omitted 08:00

Bosentan Omitted 08:00

Ambrisentan / Omitted 08:00 Macitentan Nebulised Iloprost >2.5 hours from catheter >2.5 hours from catheter

3. HAEMODYNAMIC MEASUREMENTS 3.1 DIRECTLY INVASIVE MEASURED PARAMETERS • Mean right atrial pressure (RAPmean, mmHg)

• Systolic and diastolic pulmonary artery pressure (PAPsyst, PAPdiast, mmHg)

• Pulmonary capillary wedge pressure (PCWP, mmHg)

238

• Cardiac output (CO; L/min) *

• Mixed venous oxygen-saturation rate (SvO2, %)

*using the cardiac output technique available to investigators at their site (thermodiluation (average of 3 measurements) or direct Fick Method).

3.2 DIRECTLY NON-INVASIVE MEASURED PARAMETERS • Systolic and diastolic systemic arterial blood pressure (SBP, DBP, mmHg)

• Heart rate (HR, beats/min)

• Weight (W, kg)

• Height (H, cm)

3.3 CALCULATED PARAMETERS Mean systemic arterial blood pressure (MAP):

Mean pulmonary arterial pressure (PAP mean):

Cardiac output (CO) using direct Fick:

VO (ml / min) CO  2 (L/ min) 13.4 Hb(g / dl)(SaO2  SvO2 )

Pulmonary vascular resistance (PVR):

Pulmonary vascular resistance index (PVRI):

Systemic vascular resistance (SVR):

239

System vascular resistance index (SVRI):

Body surface area (BSA):

Cardiac index (CI):

4. PERSONNEL Only personnel with an expertise in right heart catheterisation should be involved in the study related measurements.

5. DOCUMENTATION

• All study relevant data collected in context with the haemodynamic measurements must be documented in the patient file and the electronic CRF.

• Printouts from electronic devices (e.g. paper tracings of the pressure curves, print outs from the thermodilution device) will be signed by the physician who is responsible for the measurements, and need to be stored in the patient file.

6. PROCEDURES 6.1 CATHETER PLACEmENT • All measuring instruments used for haemodynamic measurements must be calibrated according to the hospital standard.

• For right heart catheterisation, a monitor for continuous control of the pressure values, at least one ECG lead, equipment for the registration of the pressure curves and a Statham pressure transducer are required. For the determination of cardiac output through thermodilution technique a cardiac output computer is necessary. If using direct Fick, bedside oximetry must be available as well as measurement of oxygen consumption.

 A balloon-tipped catheter is recommended. • Zero point determination should be performed in supine position at the beginning of each measurement. The zero point is at the mid-thoracic level.

• The catheter is placed using Seldinger technique. After puncture of a peripheral vein, preferably the v. basilica or intermedian cubital vein or alternatively, the internal jugular or femoral vein, the catheter is passed into the right atrium. To avoid

240 ventricular arrhythmia and facilitate the catheter passage to the pulmonary artery, the balloon is inflated with air.

• To determine the catheter position, the pressure is recorded continuously via the distal catheter lumen during the passage of the catheter through the heart. Alternatively, fluoroscopy may be used, where available.

• To avoid clots in the catheter lumen, the flow-guided catheter has to be flushed, at least prior to each pressure registration or at the occurrence of damped pressure curves, using physiological saline solution. The flushing may be performed with bolus heparin, or a saline solution containing heparin. These anti-thrombotic measures should be undertaken according to the local sites standard procedures.

• After passage of the atrium and ventricle, the catheter is placed in the pulmonary artery.

• To assure stable conditions with respect to the pressure recordings, the patient should rest for 10 minutes after catheter placement before the measurements are started. During this time frame any stress (e.g. manipulations on the patient) that might have an impact on the patient should be avoided.

• Meaurements should be taken at held end-expiration, paying attention to avoid a Valsalva manœuvre. On exercise, it may be necessary to take measurements at end expiration (just before the next breath).

• In case that the physician responsible for the measurements has the impression that the results of the measurements are questionable (e.g. measurements not performed under stable conditions), a second series of measurements should be performed after another 10 min.

6.2 SUGGESTED ORDER OF MEASUREMENTS It is suggested to perform the invasive measurements in the following order:

RA pressure -> PAP pressures -> PCW pressure -> CO -> SvO2

6.3 MEASURING OF THE CARDIAC OUTPUT (CO) 6.3.1 THERMODILUTION METHOD • Before the measurement of the CO is started the correct position of the proximal catheter lumen needs to be confirmed (in the right atrium).

• In order to determine the cardiac output according to the thermodilution principle, inject rapidly 10 ml physiologic sodium chloride with a temperature between 2°C and 5°C via the proximal lumen of the Swan-Ganz thermodilution catheter or use the institutional protocol, e.g. with thermistors on the injection syringe and use of room- temperature saline solution. Carry out three measurements of the cardiac output

241

(CO). The mean value of these three measurements must be entered recorded.

• To obtain comparable measurements, the injection of sodium chloride should be performed at the end of the expiration.

• Measurement of systolic and diastolic systemic arterial blood pressure and heart rate (from ECG), needed for the calculation of systemic vascular resistance and cardiac stroke volume, should be performed in close temporal relationship to the determination of cardiac output (e.g. directly before the three cardiac output measurements). The systemic arterial blood pressure can be measured invasively (via an arterial catheter) or non-invasively (via a mercury sphyngomanometer or a validated electronic device).

6.3.2 DIRECT FICK METHOD Measurements required for determination of CO using direct Fick:

Oxygen consumption (ml/min); haemoglobin (g/dl); arterial / peripheral

saturation (SaO2 / SpO2); mixed venous saturation (SvO2).

• Before the measurement of the CO is started the correct position of the distal port needs to be confirmed (in the pulmonary artery).

• In order determine CO using the direct Fick method, oxygen consumption must be recorded with the patient stable for 10 minutes and breathing through the device for at least 5 minutes to allow the patient to settle. Measurements must be taken at least one minute after any breath-holding manœuvres.

 • Measurement of systolic and diastolic systemic arterial blood pressure, heart rate

(from ECG), SvO2 (see 6.4) and arterial or peripheral SO2, needed for the calculation of cardiac output, systemic vascular resistance and cardiac stroke volume, should be performed in close temporal relationship to each other and the measurement of oxygen consumption. 6.4 MEASURING OF MIXED VENOUS OXYGEN-SATURATION (SVO2, %) • The blood for the determination of the mixed venous oxygen-saturation will be collected via the distal catheter lumen, located in the pulmonary artery.

• To avoid incorrect measurements the first 2 ml blood taken from the catheter should be discarded.

242

Appendix 3.2

Standard Operating Procedure for Incremental and Endurance Cardiopulmonary Exercise Testing

1 INTRODUCTION Cardiopulmonary exercise (CPX) testing is part of routine practice for patients with known or suspected pulmonary hypertension. It is used to assess maximal exercise capacity as well as measure effort-independent sub-maximal parameters. As part of the study, patients will undergo incremental and endurance CPX.

2 PERSONNEL AND LOCATION Exercise tests and constant work-rate exercise tests without haemodynamic monitoring will occur in a dedicated temperature-controlled (20-22°C) laboratory suite for cardiopulmonary exercise testing, be conducted by a clinical physiologist experienced in cardiopulmonary exercise testing. Resuscitation equipment and staff trained in advanced life support must be immediately available at all times in the department during testing.

3 CALIBRATION OF EQUIPMENT PRIOR TO EACH TEST  Volume sensor is should be calibrated with a standard volume syringe, eg 3L.

 The gas analyser is calibrated using precision gas-mixtures prior to each test.

MONTHLY  Biological calibration. The same healthy volunteer performs constant workload CPX at 0% (unloaded cycling), 30% and 60% of their maximum work-rate for 5 minutes at each level. Subjects should adhere to the same

pre-testing routine as patients (see below). Ventilation, VO2 and VCO2 should be within 5% of the baseline value. If there is significant drift, then a metabolic calibrator should be used (see below). If there has been no drift in the gas analysers, then biological study should be repeated, and if there has been a consistent drift, the bicycle should be calibrated and serviced by the supplier.

SIX-MONTHLY  Calibration should be performed against an artificial lung (Metabolic Calibrator / Simulator 17056, VacuMed). This is owned by Hammersmith and will be couriered to each site in turn. Standard gas should be available on each site.

243

4 PATIENT CONSIDERATIONS All patients should be given written information regarding the CPX at least 2 days prior to the test (Appendix). If a patient has not performed CPX testing before, it may be helpful to familiarize the patient with the equipment in advance.

4.1 PRIOR TO CPX Patients should

 Avoid heavy meals for 2 hours prior to the test

 Avoid caffeine containing drinks.

 Refrain from smoking on day of test

 Patient should be attired appropriately to facilitate cycling on the ergometer. (see letter in Appendix)

 Take their normal medications

244

5 CPX PROCEDURE Incremental CPX protocol

Continuous measurement of gas exchange and NIRS variables.

Non-invasive blood pressure taken at rest, every 2-mins during exercise, peak exercise and recovery.

Cycling workload increases by 1 watt increments. Rolling Resting Baseline:

Resting

End Test Test End End Start Test Start Start Test Start Baseline No Load. Recovery 3 min 3 min

Endurance CPX Protocol

245

Workload equivalent to 80% of Peak VO2 achieved during maximal incremental CPET

Non-invasive blood pressure taken at rest, every 2-mins during exercise, peak exercise and recovery.

246

6.1 PRE-TEST PROCEDURES 6.1.1 SETTING OF RAMP RATE FOR INCREMENTAL PROCEDURE  This is based on the outcome of the previous CPX or 6MWD using Table 1. The ramp is set to increment in 1 Watt steps.

Table 1 CPX work rate increment for protocol.

6-Minute Walk Distance (m) CPX work-rate increment (W/min)

50-150 5

150-300 10

300-400 15

>400 20

6.1.2 SETUP OF PATIENT AND EQUIPMENT  Electrocardiogram (ECG) electrodes are placed on the patient’s torso as in Fig.1. for 12-lead ECG recording.

 Near-Infrared Spectroscopy (NIRS) optodes are placed on the patient’s left Vastus lateralis muscle (Hammersmith Site only).

 Cycle seat height is adjusted for each patient. The preferred height is determined from the patient, typically with 10-15 degree bend at the knee at full leg extension.

 Exercise protocol is selected and implemented from the metabolic cart.

 Physiological monitoring is initiated as detailed below

6.2 STANDARD PHYSIOLOGICAL MONITORING:  12-lead electrocardiogram (ECG):

 Optimal skin preparation prior to electrode placement improves the fidelity of the ECG tracing. Hair should be removed with a razor where necessary and the skin surface should be cleaned of dirt and oil.

 Leads are positioned as shown below with limb leads placed on the trunk (front or back) to allow unimpaired cycling (Mason-Likar configuration) [1966]. Sweat-resistant adhesive electrodes are placed prior to mounting the cycle and leads attached after the patient is sat comfortably on the cycle

247

R L

C1 C2 C3 C6 C4 C5

N F

Fig 1 ECG Electrode Placement

 A resting ECG is recorded and the tracing viewed by a clinician prior to commencing cycling.

 Full 12-lead ECG is recorded at regular intervals for later review however continuous recordings can be made if there are medical concerns warranting caution without stopping the test such as frequent ectopics.

 Blood Pressure (BP) Monitoring

 Non-invasive blood pressure readings are taken using an arm cuff and a microphone-assisted exercise sphygmomanometer.

 Care should be taken that clothing is removed from the arm, that the cuff size is correct for the patient and that the microphone is positioned above the brachial artery.

 Pre-exercise, resting BP is measured with the patient comfortably seated on the cycle.

 During CPX, BP is measured approximately every 2 minutes and entered into the metabolic cart software manually or recorded separately.

 Peripheral Oxygen Saturations (SpO2)

 Once the patient is seated on the cycle, a pulse oximeter is attached to the index finger of the arm not being used for blood pressure measurements.

248

 False nails and nail polish may have to be removed if there is poor signal.

 Where signal is persistently weak due impaired peripheral circulation, as can occur in patients with connective tissue disease, earlobe probes may be utilised.

 The SpO2 signal is transferred directly to the metabolic cart and incorporated into the real-time graphical output.

 Gas Exchange measurements.

 A mask or mouthpiece with nose-clips is fitted appropriately for the patient. A reasonably airtight seal is important and should be checked with the patient’s assistance.

 The precise mask/mouthpiece must be selected in the metabolic cart software to correct for dead-space.

 Volumetric and gas-exchange data are transferred directly to the metabolic cart software and can be viewed in real-time.

 Metabolic Cart

 Patient identification details, height and weight are entered manually prior to each test.

 Non-invasive blood pressure measurements are entered manually.

 Ventilatory volume and gas exchange readings, peripheral saturation and ECG data are automatically available for real-time graphical display.

 Data can be shown breath-by-breath but for clarity 5-breath averaging is used in the real-time display and 30second averaging in the printed graphical outputs.

6.3 SPIROMETRIC EVALUATION Patients may have contemporaneous data from formal pulmonary function testing however all patients undergo simple spirometry prior to CPX, wearing the same mask/mouthpiece. A resting maximum flow-volume loop and the following measurements are recorded:

 Vital capacity (VC)

 Forced expiratory volume in 1 second (FEV1)

 Predicted maximum ventilation is estimated as FEV1*35.

249

6.4 RUNNING THE CPX 6.4.1 RESTING PHASE Patients rest without cycling, seated upright on the cycle ergometer, for 3 minutes. The physiological measurements are used as a baseline for the patient. Significantly abnormal baseline values may be attributed to clinical conditions or calibration errors.

Commonly encountered causes of abnormal baseline physiology:

 Leak from a patient’s nose or mouth due to poorly fitted masks/clips causes a uniform fall in VE, VCO2 and VO2.

 Anxiety presents as initially increased HR, BF, RER with low PET CO2, which correct once significant work is required of the patient.

Persistent unexplained discrepancies from reference values warrant further enquiry. Normal ranges are provided in Table 2.

250

Table 2. Ranges for Resting Physiology

Parameter Normal Reference Outlier?/Error? Range Check equipment/data

HR 60-100 bpm <40 bpm

VO2 200-400 ml/min < 100 ml/min

VO2/Body weight 3-4 ml/kg/min <2 ml/kg/min

VCO2 140-300 ml/min <100 ml/min

RER 0.7-1.0 <0.5

BF 12-20 breaths/min <5, >30 breaths/min

VE 5-10 L/min <2. >25 L/min

VT 400-600 ml <200ml

PETCO2 4.5-6 kPa <3, >8kPa

6.4.2 UNLOADED CYCLING PHASE Patients are instructed to pedal at 60 rpm without any resistive load from the cycle ergometer for 3 minutes. An LCD display shows current cycling cadence to the patient. Verbal cues should be provided by the operator if necessary to adjust the cycling cadence to 60 rpm.

6.4.3 WORK PHASE At the end of the defined unloading cycling phase, work load is automatically increased according to the CPX protocol selected.

6.4.3.1 INCREMENTAL CPX  Work-load is increased by 5-25 Watts per minute in 1 Watt increments to provide a continuous ramp protocol.

 Work-load increment rate based on estimates should result in approximately 8-12 minute exercise phase of CPX

 The patient should be encouraged to achieve a maximal effort. This is usually indicated by one or more of:

 Heart Rate at maximum predicted value

251

 RER >1.1

 Breathing reserve < 20% of predicted maximum value

 VO2 plateau

6.4.3.2 ENDURANCE CPX Work-load is set at 80% of the workload achieved during a prior maximal incremental test in the same patient. The patient should be encouraged to cycle for the maximal possible duration at this workload as for an incremental test. See 6.7.1 for guidance on when to time-out the test.

6.5 CPX: TERMINATION OF EXERCISE AND RECOVERY 6.5.1 CRITERIA FOR TERMINATION OF EXERCISE Exercise continues until any of the following occur:

 patient is unable to maintain cycling cadence of greater than 40 rpm

 the patient needs to stop due to symptoms

 any symptoms of chest pain or dizziness

 test terminated at the supervising clinician’s discretion when

 systolic blood pressure is above 240 mmHg

 clinically significant ECG abnormalities are seen (e.g. ischaemia, ventricular dysrhythmias, new second or third degree heart block)

 symptomatic hypotension

 neurological impairment (e.g. confusion or altered conscious state)

252

6.7.2 RECOVERY ECG and gas exchange are monitored for 3 minutes of recovery.

The CPX operator records the primary and any secondary factors limiting further exercise, including:

 shortness of breath

 leg fatigue or pain

 chest pain

 musculoskeletal discomfort

 Operator terminated CPX

253

6 CPX: PHYSIOLOGICAL MEASUREMENTS 6.1 DIRECT NON-INVASIVE Visit demographics

 Weight (W, kg)

 Height (H, m)

 Age (yrs)

 Gender.

Cardiovascular parameters:

 Systolic and diastolic systemic arterial pressure (nSBP, nDBP, mmHg)

 Heart rate (HR, beats/min)

Ventilatory parameters:

 Respiratory rate (BF, Breaths/min)

 Tidal Volume (VT, L/breath)

 Ventilation (VE, L/min) at Body Temperature and Pressure Saturated (BTPS)

Gas Exchange parameters (measured breath-by-breath, averaged over 5 breaths) at Standard Temperature and Pressure Dry (STPD):

 Oxygen consumption rate at mouth (VO2, ml/min)

 Carbon dioxide production rate at mouth (VCO2, ml/min)

 End-tidal Partial pressure of carbon dioxide (PETCO2, kPa)

 End-tidal Partial pressure of oxygen (PETO2, kPa)

Near infrared spectroscopic (NIRS) parameters:

 Muscle tissue oxygenation index (TOI, % change from baseline)

 Change in Oxy-Haemoglobin level (OHb, μmol)

254

 Change in Deoxy-Haemoglobin level (HHb, μmol)

 Change in total Haemoglobin level (CHb, μmol)

Miscellaneous parameters:

 Endurance time (TLim, seconds)

255

6.2 AUTOMATICALLY CALCULATED PARAMETERS  Body surface area (BSA, m2): (DuBois formula)

BSA  W 0.425  H 0.725 0.007184

 Body Mass Index (BMI, kg/m2)

W BMI  H 2

 Mean systemic arterial blood pressure (MAP, mmHg)

nSBP  nDBP 2 MAP  3

 Respiratory exchange Ratio (RER)

This is calculated and graphically represented in real-time.

VCO RER  2 VO2

 Ventilatory equivalent for oxygen (EqO2)

This is calculated and graphically represented in real-time.

VO2 EqO2  VE

 Ventilatory equivalent for carbon dioxide (EqCO2)

This is calculated and graphically represented in real-time.

VCO2 EqCO2  VE

 The predicted values of peak VO2 for each patient vary with sex (0,male; 1, female), age (years), and height (Ht, cm). Peak VO2 = 0.046(Ht) – 0.021 (Age) – 0.62(Sex) – 4.31 L/min

6.3 GRAPHICALLY DISPLAYED RELATIONSHIPS Various combinations of variables can be displayed graphically in real-time using the metabolic cart software. Example graphical outputs are provided in the Appendix.

256

6.4 OPERATOR-DERIVED PARAMETERS

 Peak VO2 (ml/min/kg)

Peak VO2 is recorded as the as the highest peak VO2 averaged over any 30 second period, not the final 30 seconds.

 3-minute measures (in endurance test)

Measurements recorded at 3 minutes in the endurance test are the 30- second averages leading up to 3 minutes.

 VO2/WR slope (ml/min/W)

A first-order polynomial is fitted to the data after completion of the test (using the metabolic cart software) to obtain this value.

Start Test

* **  VE/VCO2 slope (L/min [BTPS ]/L/min[STPD ])

The test data is considered from the point where the VE/VCO2 relationship becomes linear near the start of incremental exercise until the respiratory compensation point, when ventilation increases rapidly and

disproportionately to VCO2, just prior to peak exercise.

The CPX operator sets the range over which a first order polynomial is fitted

to the VE/VCO2 data over this interval using the metabolic cart software (Fig 2, Red Line). The slope of this line is noted from the resultant equation.

257

Respiratory Compensation Point

 VE/VCO2 intercept (L/min)

From the equation of the line fitted to the VE/VCO2 slope, the y-intercept is recorded.

 Estimated anaerobic threshold. (AT, ml/min and ml/min/kg)

The estimated anaerobic threshold is estimated over the same data range by examining a combination of relationships

In order of priority, the CPX operator considers

 The point where the VCO2/VO2 graph steepens (the V-Slope Method)

 The point where the EqO2/Time graph steepens (EqCO2 should remain linear at this time point)

 The RER/Time slope beings to steepen.

V-Slope Method to determine AT

The VCO2/VO2 relationship is initially tightly coupled and linear. At the estimated anaerobic threshold, the slope of this relationship steepens.

This point is identified by the CPX operator using the metabolic software display. Determining AT where Original Plot the V-Slope steepens

258

Using Ventilatory Equivalents and end-tidal gases to determine AT

At AT in an incremental CPX, EqO2 increases noticeably while EqCO2 remains low until the respiratory compensation point, which is always later than AT.

At AT, there is a simultaneous increase in pETO2 and decrease in PET CO2

Using RER to determine AT

At AT in an incremental test, the RER begins to increase continuously up to peak exercise. Note there may be fluctuations in RER throughout exercise related to the patient’s ventilatory pattern and AT may not necessarily coincide with the nadir of RER.

* **  Equivalent for CO2 (EqCO2) at AT and at nadir (L/min [BTPS ]/L/min[STPD ])

The plot of EqCO2 over time is used to determine the lowest value of EqCO2 over the duration of the CPX. Generally, this is temporally near or at AT.

259

 Oxygen Uptake Efficiency Slope (OUES)

Oxygen uptake during incremental exercise was plotted against the logarithm of total ventilation to obtain the OUES. This estimates the peak VO2 predicted to occur from the initial data when a CPX is sub-maximal.

2000

1600 VO2 = a + b . log VE

1200

VO2 ml/min VO2 800

400

0 0.500 0.700 0.900 1.100 1.300 1.500 1.700 1.900 Log10 VE

6.5 DOCUMENTATION Printouts from the metabolic cart (e.g. 9-panel plot) will be signed by the physiologist who is responsible for the measurements, and need to be stored in the patient file.

260

6.5 SOURCE OF MEASURED AND DERIVED VARIABLES USED IN CPX

Primary Measurement Used in Derivation of

Flow transducer VE VE/VO2, EqO2

VT VE/VCO2, EqCO2,

RR VD/VT

O2 analyser VO2 AT,

PET-O2 RER

CO2 Analyser VCO2 VO2/HR (O2 pulse)

PET-CO2

Sphygmomanometer Blood Pressure

Pulse Oximeter SpO2, HR

ECG recording HR VO2/HR

Rhythm

Cycle ergometer Work Rate VO2/WR

261

Exercise Laboratory Letter Date

Hammersmith Hospital, Du Cane Road, London W12 0HS

Web site www.imperial.nhs.uk

020 838 32352

Dear ,

An appointment has been made for you to attend for a Metabolic Exercise Test. This test will be carried out in Ward C2, which is accessed from the main corridor at Hammersmith Hospital. Please ask at the front desk for detailed instructions on how to find the centre.

Appointment Date:

Time:

If this is particularly inconvenient, please telephone 020 8846 7183 to arrange another appointment. If you find yourself unable to attend, for any reason, please let us know as soon as possible, so that your appointment time does not go to waste.

Preparation for your metabolic exercise test

The Metabolic Exercise Test involves you exercising on a bicycle ergometer at a level based on your ability, and should last for about 10 minutes exercise. We will monitor your breathing and heart rate throughout the exercise, and the work rate will gradually increase until you can no longer keep up. The object of the test is to determine how much exercise you can tolerate, and to see how well your heart and lungs respond to the increase in demand. You will always be in control of how much exercise you undertake, and will be free to stop at any point, should you choose.

1. Shoes It is advisable to wear comfortable, flat soled shoes.

Trainers or running shoes are ideal. 262

2. Clothing Women: Loose fitting top.

Comfortable bra, but no slip, body or corset.

Loose fitting trousers or shorts, no long skirts.

Men: Loose fitting top.

Loose fitting trousers or shorts.

3. Meals Please do not eat a heavy meal for 2 hours prior to the test and avoid caffeine containing drinks.

Please refrain from smoking on the day of the test.

4. Medication Continue to take all medication as directed by your doctor.

If you have any queries regarding your medication, please seek advice from your doctor.

Please bring a list of your current medications with you .

263

EXAMPLE GRAPHICAL OUTPUT FROM METABOLIC CART I) NINE-PANEL PLOT

264

Other standard plots:

265